Colostrum and milk were collected from ten sows at frequent intervals from before farrowing until 9 d after farrowing. Ionized calcium, pH, and total concentrations of citrate, calcium, phosphate and magnesium were measured in whole milk. The diffusible fraction of the mammary secretion was separated by ultrafiltration and was used for the measurement of diffusible citrate, calcium, phosphate and magnesium. The pH before farrowing was 5·7, and increased to 6·5 on day 4 as total calcium and phosphate also increased. Before farrowing, total and diffusible citrate were 7·8 and 7·3 mM respectively, while diffusible phosphate was 11·9 mM, and these concentrations all decreased during the study period. Total magnesium ranged between 3·3 and 4·1 mM, while diffusible magnesium ranged between 2·0 and 3·1 mM. While these concentrations and patterns of change of diffusible calcium and citrate are quite different from those of women's milk during the first week after birth, theoretical physicochemical relationships between diffusible calcium and citrate, and ionized calcium and HPO42− were corroborated by these results. We conclude that diffusible citrate plays an important role in the determination of the concentration of diffusible calcium. However, while citrate may be the major determinant of the total concentration of calcium in women's milk, this is not the case in sows' milk.

7-Ketocholesterol and 7-ketositosterol were chosen as reliable markers of the oxidation of cholesterol and phytosterols in infant milk formulas and infant milk cereals. A reversed-phase HPLC method was developed to measure them simultaneously in infant formulas. This method was then tested on a wide range of infant milk formulas and milk cereals on sale in Italy whose lipid composition is representative of the most common commercial formulas. The analytical results revealed no significant differences in the extent of oxidation of cholesterol and sitosterol. As the level of 7-ketocholesterol often followed the cholesterol level, a cholesterol content similar to that of human milk produced amounts of cholesterol oxides with possible negative effects on infant health. In contrast, the low cholesterol content of milk cereals never produced amounts of cholesterol oxides high enough to cause concern. The contents of phytosterols and hence their oxides were always low.

From hydrolysis experiments carried out on αs1-caseins A and F at pH 5·2 in the presence of 30 g NaCl/l, i.e. the conditions encountered in many young goats' cheeses, it was found that minima of 19 and 9 bonds were sensitive to chymosin in variants A and F respectively. Variant A was hydrolysed faster than variant F and the proteolytic pattern (reversed-phase HPLC and polyacrylamide agarose gel electrophoresis) differed between the variants. Hydrolysates from both variants had a number of cleavage sites in common (Leu20–Leu21, Phe23–Ala24 and Phe32–Arg33 in both variants, Leu101–Lys102 and Leu64–Lys65, Leu120–His121 and Leu83–His84, Leu142–Ala143 and Leu105–Ala106, Leu149–Phe150 and Leu112–Phe113, Leu156–Asp157 and Leu119–Asp120, Trp164–Tyr165 and Trp127–Tyr128 in variants A and F respectively), while other bonds were split only in variant A (Leu16–Asn17, Glu18–Asn19, Phe28–Pro29, Ile44–Gly45, Tyr80–Ile81, Gln82–Lys83, Tyr91–Leu92, Tyr94–Leu95, Leu109–Glu110 and Phe179–Ser180). Major cleavage sites appeared to be at Phe23–Val24, Leu142–Ala143 and Trp164–Tyr165 for variant A, and Phe23–Val24 and Leu64–Lys65 for variant F. Cleavage site Phe23–Val24 could be the origin of the first breakdown product from goat αs1-caseins A and F visible in polyacrylamide agarose gel electrophoresis.

The effect of graded levels of dietary rumen-protected methionine (rp-Met) or lysine (rp-Lys) on the fatty acid composition of Comisana ewes' milk was investigated. The animals lambed in autumn and were separated from their lambs 7 weeks after parturition; they were divided into five groups of nine which were given different rations. The control group was fed on a mixture of vetch and oat hay with a pelleted concentrate; the other groups were given the control ration supplemented with rp-Met (3·5 or 7·0 g/kg) or rp-Lys (10·5 or 21·0 g/kg). Milk fatty acid composition was determined by gas chromatography, and these values were subjected to principal components analysis. A correlation was found between fatty acid composition and diet. We demonstrated that giving the experimental diets significantly affected milk fat, to an extent that was independent of the levels of the supplementary amino acids. The milk fat of the control group contained largely caprylic, capric and lauric acids, whereas that of the groups given the supplementary amino acids contained largely palmitic, palmitoleic, stearic and oleic acids. Thus supplementation resulted in an increase in the proportion of 16[ratio ]0–18[ratio ]3 (primarily palmitic and stearic) and a reduction in the proportion of 4[ratio ]0–12[ratio ]0 fatty acids. There was also a modest increase in the unsaturated: saturated fatty acid ratio.

The aim of this study was to examine the effect of bovine colostral whey proteins from cows immunized with Streptococcus mutans/Strep. sobrinus on the adherence and aggregation of caries-inducing bacteria, i.e. mutans streptococci. Both adherence and aggregation are important phenomena in the bacterial colonization of the human oral cavity. In all adherence experiments there was a significant difference between treatments by immune product (IP; from immunized cows) and a control product (CP; a similar product from non-immunized cows). The adherence of 35S-labelled Strep. mutans cells (serotype c) to parotid saliva-coated hydroxyapatite (SHA) was dose-dependently inhibited by both IP and CP if SHA was coated with either product before exposure to bacteria, but markedly lower concentrations of IP than CP were effective. When instead of SHA the bacterial cells were pretreated with IP or CP, only IP strongly and dose-dependently inhibited streptococcal adherence. When bacteria, IP or CP, and SHA were incubated simultaneously, a significant difference between IP and CP treatments was again found. Further, IP effectively aggregated both Strep. mutans and Strep. sobrinus cells, whereas hardly any effect was seen with CP. Both IP and CP aggregated the control bacterium Strep. sanguis, which affected the adherence of the pretreated bacteria.

Four minor protein components were detected in whey from Romagnola cows' milk by polyacrylamide gel isoelectric focusing and two dimensional gel electrophoresis. Individual protein spots were transferred by electroblotting on to a polyvinylidene difluoride membrane and isolated by cutting out the relevant area. After in situ trypsinolysis, a portion of the digest was analysed directly by matrix-assisted laser desorption ionization–time of flight mass spectrometry. The mass profile allowed us to establish a correlation between β-lactoglobulins A and B and the four minor whey protein components. They were identified as C-terminally truncated β-lactoglobulin A and B variants with missing N-terminal peptides, beyond residues in the range 100–123 and 136–147 respectively. Two of the minor components were related to β-lactoglobulin A and two to β-lactoglobulin B.

Our aim was to determine the effect of growth hormone on non-insulin-mediated glucose disposal in lactating dairy cows. Following 5 d of subcutaneous injections of either saline or growth hormone, insulin, somatostatin or insulin plus somatostatin were infused for 2 h each, in a series of experiments. Coincident with this, unlabelled glucose was infused at a variable rate to maintain a constant plasma glucose concentration. Glucose, doubly labelled with deuterium, was also infused for the calculations of glucose turnover. Plasma insulin levels were reduced to nearly zero by the infusion of somatostatin; under such conditions whole body glucose disposal should be non-insulin-mediated. Dairy cows treated with growth hormone, which had significantly increased milk yields on the day before the experimental infusions, did not have different levels of whole body non-insulin-mediated glucose disposal when expressed in absolute terms. Growth hormone did not affect non-mammary non-insulin-mediated glucose uptake estimated by calculation. Growth hormone significantly inhibited insulin-mediated glucose uptake when plasma insulin levels were elevated. Glucose uptake during insulin plus somatostatin infusion was not significantly different from that of the insulin only infusion.

The pressure-induced (100–400 MPa) changes in casein micelles from bovine, caprine and ovine milks were investigated by studying the distribution of minerals and proteins after separation of the micellar and serum phases by ultracentrifugation. The results showed that pressurization markedly increased the levels of non-sedimentable casein in the three species, as well as the levels of Ca, P and Mg in the serum, but led to only small increases in the Ca2+ concentration. The dissociation rates of individual caseins corresponded to the ester phosphate contents, in the order κ-casein, β-casein>αs1-casein >αs2-casein. The concentration of non-sedimentable casein was highest in pressurized ovine milk, which also contained higher levels of pressure-liberated serum minerals than the milks from the other two species. In the case of bovine and caprine milks, maximum dissociation from the micelle was found in milks treated at 300 MPa, while in ewes' milk dissociation increased with pressure up to 400 MPa. Denatured β-lactoglobulin (β-lg) was present in the ultracentrifugation pellets of the pressurized milks of the three species, but pressure-denatured β-lg also remained in the soluble fraction. Despite the increases in the content of soluble proteins, all casein and most β-lg were incorporated in the rennet curds of the pressurized milks from the three species.

Whey protein concentrate (WPC) solutions containing 10, 30, 60 and 120 g dry powder/kg were heated at 75°C and whey protein aggregation was studied by following the changes in the distribution of β-lactoglobulin, α-lactalbumin and bovine serum albumin, using one dimensional and two dimensional PAGE. The one dimensional PAGE results showed that a minimal quantity of large aggregates was formed when 10 g WPC/kg solutions were heated at 75°C for up to 16 min whereas appreciable quantities were formed when 30, 60 and 120 g WPC/kg solutions were similarly treated. The two dimensional PAGE analysis showed that some disulphide-linked β-lactoglobulin dimers were present in heated 10 g WPC/kg solution, but very little was present in heated 120 g WPC/kg solution. By contrast, SDS was able to dissociate monomeric protein from high molecular mass aggregates in heated WPC solution of 120 g/kg but not in 10 g WPC/kg solution heated for 30 min. The rates of loss of native-like and SDS-monomeric β-lactoglobulin, α-lactalbumin and bovine serum albumin during heating increased as the WPC concentration was increased from 10 to 120 g/kg. In 120 g WPC/kg solution heated at 75°C, the amounts of SDS-monomeric β-lactoglobulin in each sample were greater than the quantities of native-like protein. However, in WPC solutions of 10, 30 and 60 g/kg, the differences between the amounts of native-like and SDS-monomeric proteins were slight. The loss of the native-like or SDS-monomeric proteins was consistent with a first or second order reaction. In each case, the apparent reaction rate constant appeared to be concentration-dependent, suggesting a change of aggregation mechanism in the more concentrated solutions. Overall, these results indicate that in addition to disulphide-linked aggregates, hydrophobic aggregates involving β-lactoglobulin, α-lactalbumin and bovine serum albumin were formed in heated WPC solution at high protein concentration, as suggested by model studies using binary mixtures of these proteins.

An analytical procedure, based upon HPLC, has been used to determine the degree of isomerization of retinol (vitamin A) in various dairy products. In raw milks, which are not subjected to thermal processing, there was no conversion of the predominant all-trans-isomers to cis-isomers in samples from various species. Pasteurized milks with mild heat treatment (high quality milk) had an average 13-cis[ratio ]all-trans ratio of 2·6%, while pasteurized milk treated for 15 s at temperatures ranging from 72 to 76°C had an average ratio of 6·4%. Milk subjected to more severe heat treatments had a higher degree of isomerization (UHT milk, 15·7%; sterilized milk, 33·5%), consistent with increased thermal conversion of the retinol isomers. In pasteurized and UHT cream, the increase in 13-cis-isomer was also a consequence of heat treatment (pasteurized cream, 3·0%; UHT cream, 14·4%). The presence of cis isomers in fermented milk suggests that fermentation processes, directly or indirectly, can induce cis–trans isomerization. In the cheeses analysed, the extent of retinol isomerization ranged from 7·6 to 35·0%. Our results confirm the importance of individual isomers in evaluating the vitamin A status of dairy products.

Plasma concentrations of glucagon-like peptide-1(7–36)amide (GLP) and gastric inhibitory polypeptide (GIP) were determined at fortnightly intervals for over a year throughout the pregnancy–lactation cycle of goats. Both GIP and GLP concentrations were elevated during lactation and fell rapidly when milk secretion was terminated. At the onset of lactation GLP concentrations rose rapidly whereas GIP concentrations showed a delayed response. GLP concentrations remained high throughout lactation but those of GIP declined linearly as milk yields fell. Serum insulin concentrations correlated positively with plasma glucose concentrations but not with either GIP or GLP concentrations. Negative correlations were found between serum insulin concentrations and milk yield and plasma non-esterified fatty acid concentrations. The results are consistent with plasma GIP and GLP concentrations being determined by other factors in addition to nutrient intake and absorption. Changes in GIP concentrations mirrored reported changes in the hypertrophy and atrophy of the intestine in ruminants while GLP concentrations may be more dependent on the neural and endocrine factors associated with lactation. The elevated concentrations of both peptides indicated a specific role in lactation independent of their normal anabolic and insulinotropic effects.

Lactobacillus helveticus and Propionibacterium freudenreichii are essential starters in Swiss cheesemaking and the release of their intracellular enzymes through autolysis could significantly influence ripening. To provide evidence of this lysis, cheese made from microfiltered thermized milk inoculated with Lb. helveticus ITGLH77, Prop. freudenreichii ITGP23 and a commercial Streptococcus thermophilus was assayed. Starter viability was determined and lysis was monitored during ripening by protein analysis with SDS-PAGE of aqueous cheese extracts and by immunoblot detection of intracellular proteins: dipeptidase (PepD) for Lb. helveticus and methylmalonyl coenzyme A mutase for Prop. freudenreichii. We verified that the species specificity of these lysis markers was towards the cytoplasms of all the species currently used in Swiss cheese. Lb. helveticus exhibited an almost complete loss of viability (99·9%) from the beginning of ripening in the cold room; concomitantly PepD appeared in the cheese extracts and was detected until the end of ripening. Damaged Lb. helveticus cells were also visualized by scanning electron microscopy. In addition, free PepD was also successfully detected in commercial Swiss-related cheeses. All these results clearly demonstrated the autolysis of Lb. helveticus in Swiss cheese. Prop. freudenreichii ITGP23 grew during warm room ripening and no loss of viability was detected after maximal growth (109 cfu/g cheese). Free methylmalonyl-coenzyme A mutase was detected at the end of ripening during cold storage, when the cheese extracts were concentrated 20-fold, demonstrating that the autolysis of Prop. freudenreichii was tardy and limited.

Growth factors are thought to play a decisive role in the course of inflammatory processes. The aim of the present study was to characterize a potential interrelationship between the concentrations of insulin-like growth factor-1 (IGF-1), basic fibroblast growth factor (bFGF) and somatic cell count (SCC) in normal milk, and to investigate the presence of these growth factors in mammary secretions of cows suffering from clinical and subclinical mastitis. Quarter secretions of cows with spontaneous acute clinical mastitis and of cows with subclinical mastitis were analysed radioimmunologically for their concentrations of IGF-1 and bFGF. During two relocation trials with normally lactating Brown Swiss cows, dramatic changes in milk somatic cell count were obtained following a short-term change (5 d) of location and housing system. The animals were relocated from their familiar loose housing system with concrete slatted floor to a separate stanchion barn with long stalls and straw bedding, and vice versa. The concentration profile of IGF-1, but not of bFGF, corresponded well with SCC during the relocation trials, the positive correlation between the characteristics being highly significant, as determined by regression analysis (r=0·60; P<0·001). The results provide evidence that significant changes in SCC and growth factor content may be caused by environmental factors other than infection. The concentrations of both IGF-1 and bFGF were greatly elevated in secretions of quarters affected by acute clinical mastitis compared with the corresponding clinically healthy quarters. Subclinically affected quarters with high SCC, as compared with non-affected quarters with low SCC, also had elevated milk IGF-1, but unchanged bFGF. Measuring of growth factor profiles in milk may have value in the near future in monitoring the state of udder health in addition to SCC.

The cysteine proteinase from Micrococcus sp. INIA 528 was added at three concentrations to pasteurized cows' milk to study its effect on the ripening process of Hispanico cheese. Dry matter was higher in whey from experimental cheeses than in whey from control cheese manufactured from milk with no added proteinase. No differences in cheese proteolysis were found 24 h after manufacture. Residual αs- and β-caseins were significantly lower, and N soluble at pH 4·6 and in trichloroacetic acid significantly higher, in experimental cheeses than in control cheese from day 15. Hydrophobic and hydrophilic peptides in the water-soluble fraction of cheese measured by absorption at 280 nm were significantly higher in experimental cheeses than in control cheese from days 1 and 15 respectively. Texture was less firm in experimental than control cheese from day 30. Bitterness, detected in cheese with the highest levels of added cysteine proteinase, adversely affected cheese flavour. Flavour intensity was significantly higher from day 30 in cheeses with added proteinase than in control cheese.

Microfiltration (MF) of milk with a ceramic membrane (0·1 μm mean pore diam.) for the separation of casein micelles from soluble proteins was studied. Experiments were performed at constant flux density (J=65 or 75 l h−1 m−2) and wall shear stress (τw=100 or 110 Pa) with milks containing particles and solutes with different sizes and charges, produced by physicochemical change (heat and mechanical treatment, pH, ionic strength, addition of ions). Membrane separation performance was limited by the build up of a cake with characteristics that depended on the size of particles and soluble proteins. Best performance (higher permeability and whey protein transmission) was obtained with milk containing fat on the one hand and calcium phosphate on the other. An optimal value of the particle size was found, close to 0·5–3·0 μm; above this separation performance decreased. In addition, the present study confirmed that the transfer of charged solutes is the consequence of both size and ionic exclusion. Better performance was achieved with higher ionic strength (1 M).

HPLC size exclusion chromatography experiments showed that during inactivation at 40–70°C of the extracellular proteinase from Pseudomonas fluorescens 22F small molecular mass fragments were formed, indicating that autoproteolysis was at least one of the major causes of inactivation. The formation of small molecular mass fragments and the reaction order indicated that intermolecular autoproteolysis was more likely than intramolecular autodigestion. This was confirmed by computer simulations. The rate constants and the activation enthalpy (ΔH[Dagger]) and entropy (ΔS[Dagger]) for the reactions of the intermolecular autoproteolysis model were derived from computer simulations. ΔH[Dagger] and ΔS[Dagger] of the unfolding reaction were 504 kJ mol−1 and 1252 J mol K−1 respectively. ΔH[Dagger] and ΔS[Dagger] of the refolding reaction were strongly temperature dependent. The estimates for the enthalpy (ΔH0) and entropy (ΔS0) difference between the folded and unfolded state as derived from the reaction rate constants of unfolding and refolding were subject to large deviations, owing to accumulation of errors in the estimation of the kinetic characteristics.

We have investigated accelerating the ripening of Pecorino Umbro cheese by adding crude cytoplasmic extract from Pseudomonas fluorescens, non-starter lactic acid bacteria (NSLAB) or cheese slurry. Microbiological and biochemical analyses and sensory evaluation were carried out on control and experimental cheeses over 28 d ripening. In the cheeses containing NSLAB or slurry, counts of mesophilic lactobacilli ranged from log 7·6 at day 1 to ∼log 8·6 cfu/g after 28 d ripening, ∼2 log cycles higher than in the control cheese. All the experimental cheeses contained higher levels than the control of total free amino acids and N soluble at pH 4·6 and in 120 g trichloroacetic acid/l. Compared with the control, higher aminopeptidase and dipeptidase activities were found in the cheeses containing NSLAB and slurry, and especially in those containing the Pseudomonas enzyme. The cheeses containing NSLAB or slurry were characterized by an accumulation of short peptides (Mr<2000) detected by FPLC. Although the cheese containing enzyme had an atypical flavour, the addition of mesophilic lactobacilli reduced from 60 to 28 d the ripening period of Pecorino Umbro cheese, without the appearance of off flavour.

To study the possible presence of staphylococcal enterotoxin A in raw goats' milk lactic cheese, milk was inoculated with an enterotoxigenic Staphylococcus aureus strain to a final concentration of 4, 5 and 6 log(cfu/ml). Cheese was prepared following industrial specifications and ripened for 42 d. Detection of the enterotoxins was by the Vidas Staph enterotoxin test (BioMérieux) and by an indirect double-sandwich ELISA technique using anti-enterotoxin monoclonal antibodies. Staphylococcal counts declined markedly after draining, and by the end of ripening they had disappeared from some cheeses. In contrast, aerobic mesophilic organisms grew well. The level of staphylococcal enterotoxin A recovered varied from 1 to 2·5 ng/g cheese made with an initial population of 105 or 106 cfu/ml. Only traces of enterotoxin A (0·5 ng/g) were detected in cheeses made with the lowest Staph. aureus inoculum used in this study. Enterotoxin A was also detected in cheeses from which Staph. aureus had disappeared.

The effects of heat treatment of ewes' milk at 65–85°C for 10–30 min and subsequent storage at 20°C for 90 min on the changes in ionic calcium and the distribution of calcium, magnesium and phosphorus between the diffusible and micellar phases were studied. As milk was heated there was a decrease in ionic calcium of ∼36 mg/l, but there was a rapid 86–90% recovery during the first 30 min after heating. Although the rate of reversal on subsequent storage decreased with the intensity of the heat treatment, there were no significant differences between samples after 90 min storage. Similar results were obtained for cows' milk subjected to the same treatments. Diffusible calcium, magnesium and phosphorus in heated samples were lower than in raw milk. Recoveries of initial values decreased with the severity of the heat treatment and were 80–95% for calcium, 85–97% for magnesium and 87–98% for phosphorus. Heat treatment under milder conditions (85°C for 20 s) was associated with almost complete recoveries. Recoveries of the minerals studied were slightly lower in ewes' than in cows' milk.

Analysis of the concentrations of free amino acids in milk during growth of Lactococcus lactis subsp. lactis revealed a biphasic pattern of change during the logarithmic phase. During the first period there was little overall change in the total concentration of amino acids in the medium. The second phase was characterized by increased net liberation of free amino acids. There were also qualitative differences in the amino acids that were taken up and utilized during each period. The concentrations of Val, Leu and Ile decreased only during the early phase, while those of Ser, Arg, Thr and Met decreased only during the second phase. Gly and Ala were utilized throughout logarithmic growth. Gly uptake appeared to be greater during the second period and accounted for the largest proportion of free amino acid utilization at this time. It is possible that the biphasic nature of amino acid nutrition was due to increased consumption in late log phase of peptides derived from milk proteins by proteolysis. Increased activity of the arginine deiminase pathway during late log phase was inferred from increased utilization of Arg and liberation of citrulline and ornithine.