The conversion of the two-dimensional size distribution of casein micelles, observed by electron microscopy in a plane section, to the three dimensional distribution is discussed and the average size parameters evaluated by several methods are compared. It is shown that parameters containing the −1 moment of the two-dimensional distribution, i.e. Dn, the number of micelles per unit volume and the width of the size distribution, are sometimes uncertain. The occurrence of negative numbers in some of the classes of the distribution is discussed and remedies are suggested. Sections were made by freeze-fracturing skim milk samples; the pH of the milk was between 5·5 and 6·7.

Raw skim milk was diluted 1000-fold using distilled water or various salt solutions as specified. Smooth, hyperbolic profiles of coagulation rate v. pH for casein were calculated from recordings of turbidity (400 nm) with time. The effects of pH, cation type, anion type and cleavage of k−casein by chymosin (EC 3.4.23.4) were determined. The maximum of pH-coagulation rate profiles decreased by 63, 85 and 94% when the skim milk diluent was changed from water to salt solutions of NaCl (100 mM), CaCl2 (50 mM) or MgCl2 (50 mM). The maximum of the pH–coagulation rate profile was 15 times greater when the Ca salt was changed from CaCl2 to Ca(SCN)2 (50 mM). The highest pH at which casein coagulation occurred increased from 4·45 to > 6·0 when Cu2+ (1 mM) was included with casein micelles dispersed in CaCl2 solution (50 mM). The addition of chymosin to casein micelles suspended in CaCl2 solution (70 mM) eliminated the inhibition of casein coagulation by Ca2+ at pH 4·5. It is proposed that ions such as Mg2+, Ca2+, and Na+, which generally associate with casein phosphate and carboxylate groups, increased the H+ concentration required to initiate the coagulation of casein, because H+ must displace bound Ca2+, Mg2+ or Na+ to reduce repulsive hydration forces between casein micelles, allowing attractive hydration forces (e.g. hydrophobic phenomena) to cause casein coagulation. Furthermore, it is proposed that ions such as Cl−, Br−, and SCN− bind to lysine, arginine and histidine groups and thereby decrease repulsive hydration forces between cationic casein micelles.

The functioning of an inorganic microfiltration membrane (Membralox-SCT) with an average pore diameter of 0·8 μm, used for processing milk, acidified milk and lactic coagulum, was studied using dynamic and static conditions. Milk processing (pH 6·62) gave acceptable rates of protein retention but at low flow rates. Study of the effect of operating parameters during microfiltration of coagulum led to determination of optimal operating conditions leading to satisfactory performance and adequate retention rate. The results are explained using a model of the formation and development of internal and external fouling of a mineral membrane. Observations using scanning electron microscopy related to the pH of milk were used to verify the model by showing that internal fouling occurred only when the pH was higher than 5·50. This internal fouling was closely linked with the physicochemical state of milk proteins. Indeed, the fouling particles visualized at the alumina grains of the membrane support changed progressively from individual micelles (pH 6·62) to a chain or clustered structure (pH 5·60). Electrophoretic study confirmed this observation and showed that casein was present in milk permeates (pH 6·62) but not in coagulum permeates (pH 4·40).

Enzymic modification with peptidylarginine deiminase (EC 3.5.3.15) enabled five out of six arginyl residues in αs1-casein to be convcrted to citrullyl residues, only the N-terminal arginyl residue remaining unaffected. An increase in the net negative charge was confirmed by PAGE. The isoelectric point was decreased from 4·46 for the intact αsl-casein to 4·30 for the deiminated type, while simultaneously lowering the acid-precipitation starting point from pH 5·17 to pH 4·62. The deiminated αs1-casein self-associated less in the absence of Ca and was less Ca-sensitive than the native type, although its Ca-binding ability was slightly enhanced. In the presence of 25 mM-CaC12 and K-casein, Ca-induced precipitation of αs1-casein did not occur, the solution of the mixture remaining transparent. Deimination of αs1-casein resulted in altering its characteristics, possibly by interfering with interactions through hydrophobicity and/or hydrogen bonding. The positive charge of the arginyl residues might play an important role in casein micelle formation.

The antioxidant activity of casein and Maillard reaction products obtained by reaction of casein with glucose or lactose was studied. Antioxidant activity was evaluated in a model System containing methyl linoleate with haemoglobin as a pro-oxidant. Casein was antioxidative and heating casein in the presence of glucose or lactose resulted in enhancement of antioxidant activity. The development of antioxidant activity in reacted casein–sugar mixtures was determined as a function of initial casein and sugar concentration. The observed antioxidant activity of reacted casein–sugar mixtures was due to casein itself and Maillard reaction products resulting from reacting casein with sugar.

Recombined concentrated milks (18% SNF, 8% fat) at their natural pH and at pH values in the range 6·28–6·68, made with powders subjected to different preheat conditions (high heat (85°C, 30 min), indirect UHT (120°C, 2 min) and direct UHT (120°C, 2 min)), were sterilized at 120°C for 13 min. The heat stabilities of recombined concentrated milks were dependent on preheat treatment. Ca2+ activity and pH of sterilized recombined concentrated milks, measured l h after sterilization, were lower than those of corresponding unsterilized recombined concentrated milks. The magnitude of the decreases in Ca2+ activity and pH induced on sterilization were dependent on the pH of the unsterilized recombined concentrated milk but were not markedly influenced by the type of preheat treatment applied during powder manufacture. The results suggest that differences in heat stability of high heat, indirect UHT and direct UHT powders are unlikely to be due to Ca2+ activity.

The effect of heat treatments, at various water activities (αw), on digestibility and on the availabilities of amino acids of whey protein samples in the presence of lactose was estimated by an in vitro digestion method with continuons dialysis. Four αw (0·3, 0·5, 0·7 and 0·97), three temperatures (75, 100 and 121 °C) and three heating periods (50, 500 and 5000 s) were selected. The initial lysine: lactose molar ratio was 1:1. Amino acid profiles showed that excessive heating of whey (121 °C, 5000 s) destroyed a significant proportion of cystine at all αw, lysine at αw 0·3, 0·5 and 0·7, and arginine at αw 0·5 and 0·7. At αw 0·3, 0·5 and 0·7, protein digestibility decreased (P < 0·05) as the temperature increased from 75 to 121 °C for a heating period of 5000 s, and as the heating time was prolonged from 500 to 5000 s at 121 °C. Excessive heating also decreased (P < 0·05) the availabilities of ail amino acids at αw 0·3, 0·5 and 0·7. The availabilities of lysine, proline, aspartic acid, glutamic acid, threonine, alanine, glycine and serine were particularly affected. Severe heating at αw 0·97 did not seem to favour the Maillard reaction, but the availabilities of cystine, tyrosine and arginine were decreased, probably as a result of structural modifications of the protein upon heating. Heating whey protein concentrates in the presence of lactose not only affected lysine, but also impaired enzymic liberation of other amino acids, according to the severity of heat treatments and αw.

Highly purified lactoferrin was isolated from ovine colostrum by sequential purification on CM-Sephadex C-50 and Blue-Sepharose, with overall yield of 55%. The ovine lactoferrin was characterized by SDS-PAGE, its amino acid composition and N-terminal sequence to residue 30. Homology with bovine and human lactoferrins was greater than 80 and 50% respectively. Antibodies to ovine lactoferrin were raised in rabbits and used to develop an enzyme-linked immuno-sorbent assay (ELISA). The antiserum was not cross reactive with other colostrum proteins.

A turbidimetric method, based on light reflection, was used to study acid coagulation of reconstituted skim milk at low temperature. Capillary viscosimetry, gelograph and laser granulometer techniques were also employed. Acidification of milk was produced by hydrolysis of glucono-δ-lactone. The general shape of the turbidimetric signal as a function of pH or time can be divided into three stages: a lag phase followed by a significant decrease and then a final rise. Two factors have a great influence on the development of milk turbidity, pH and temperature. Dynamic viscosity measurements can be related to the turbidimetric signal while laser granulometrie measurements cannot be correlated with changes in turbidity: the micelle size distribution remains constant until the first signs of gelation. As previous work showed, dynamic viscosity diminishes with acidification until a particular pH is reached (pH 5·9 at 15°C and pH 5·75 at 20°C). We have related this latter period to the turbidimetric lag phase. As milk turbidity became lower than its initial value (pH 5·75 at 15°C and pH 5·55 at 20°C), dynamic viscosity increased significantly. The release of material from micelles (β-casein and Ca) could explain this phenomenon. In the same way, further increase of turbidity at a particular pH value (pH 5·3 at both 15 and 20°C) coulcl be partly due to the reincorporation of soluble casein monomers in the micelle framework. As the onset of gelation was approached, turbidity still increased as a result of gel network formation.

Milk powders were subjected during manufacture to the following treatments: low-, medium- or high-heat, indirect or direct UHT preheat, or no preheat. The Ca2+ activity–pH profiles of reconstituted milks (9% total solids, TS) in the pH range 6·06–6·98 and reconstituted concentrated milks (19·6% TS) and recombined concentrated milks (26% TS, 18% SNF, 8% fat) in the pH range 6·27–6·69 were determined. Statistical analysis of the results using a (– In Ca2+ activity) transformation was used to quantify the effects of pH, preheat treatment and fat incorporation on Ca2+ activity. pH had a dominant effect on Ca2+ activity and the effects of preheat treatment and fat incorporation were small in comparison. Depending on the batch of milk used for preparation of powders, pH accounted for 98·8–98·9%, 96·4–96·7% and 93·7–97·3% of the total variation in Ca2+ activity in reconstituted milks, reconstituted concentrated milks and recombined concentrated milks. The corresponding contributions of preheat treatment to variation in Ca2+ activity were 0·26–0·47%, 0·34–1·67% and 0·24–0·40%. The addition of fat to reconstituted concentrated milk to yield recombined concentrated milk with the same SNF: water ratio resulted in 0·03–3·71 % variation in Ca2+ activity. The results suggested that differences in heat stability among powders subject to different preheat treatments are not likely to be linked to Ca2+ activity.

The 31P nuclear magnetic resonance spectra of milk from goats, sheep, a rabbit, a baboon and a human were examined and compared with the spectrum of cows' milk. Most of the peaks observed could be assigned to phosphorus compounds already known to be present in such milks. However, rabbit milk was found to contain unexpectedly large amounts of phosphate esters, and to have only low levels of alkaline phosphatase (EC 3.1.3.1) activity. In addition, an unknown compound was detected in baboon milk and in some samples of sheep and goat colostrum. This compound was not identified, but may be a mixed anhydride such as 1,3 diphosphoglyceric acid. The results show that changes during lactation may be followed by 31P nuclear magnetic resonance and that the technique may be of value in authentification of samples.

Cation-exchange chromatography on a Mono S column (Pharmacia) was used to separate macropeptide from whey proteins. Macropeptide was eluted by 0·1 M-NaCl in a 20 mM-KCl–HCl buffer, pH 2. This technique was suitable for quantitative determination of macropeptide in rennet whey and also for following the action of chymosin on κ-casein in skim milk. Precipitation at pH 4·6 was used to remove residual caseins and to keep macropeptide in solution. In comparison with other methods for determining macropeptide, the present one eliminates the need for pretreatment of samples with trichloroacetic acid (TCA) and allows the recovery of all the macropeptide. Quantitative determination of macropeptide in the 8% TCA-soluble fraction by cation-exchange chromatography showed that only 50–75% of the macropeptide was recovered. This chromatographic technique could also be applied for isolating and producing whole macropeptide on a preparative scale.

Bovine milk Hydrophobic fraction of proteose-peptone was prepared by hydrophobic interaction fast protein liquid chromatography. This method has several advantages such as high rapidity, simple good reproducibility and less denaturation. The proteose-peptone was eluted from a TSK-Phenyl-5PW column with a 1 M-0 M ionic strength gradient of NaH2PO4, pH 6·8, using a 6 ml/min flow rate for 56 min. The quantity of protein injected was 62·5 mg; however, it could be increased up to 100 mg. The elution order was β-CN-4P < BSA (1·6% of total N) < β-CN-5P < β-CN-1P. The hydrophobic fraction was obtained in pure water at the end of the gradient (17·3% of total N). A proteose-peptone cartograph was achieved by bidimensional electrophoresis. This hydrophobic fraction represented three principal zones of Mr 30000–28000, 19000 and 11000, which were respectively composed of 13, 4 and 2 principal spots distributed between 4·9 and 6·1 isoelectric points (IP). These spots corresponded to glycoproteins. ·7, 5·0 and 5·1 IP which migrated to Mr 18000 while β-CN-1P was identified as Mr 9000 in two spots of 5·1 and 5·3 IP.

Six batches of Casar de Cáceres raw ewes' milk cheeses were examined throughout their ripening. Three were made in winter, the others in spring. The microbial counts in the interior of the cheeses were ∼1·5 log units higher in the winter cheeses for total viable counts, lactic streptococci and leuconostocs, and >2 log units higher for total enterobacteria, coliforms and faecal coliforms in the 60 d samples. Staphylococci were not detected after day 15 of maturation in spring cheeses, but reached 1 log unit at day 60 in the interior of winter cheeses. In the surface samples, differences in microbial counts were more marked. Values of some physicochemical indices, such as pH, % moisture content and % NaCl were also determined.

Photoacoustic and proton-induced X-ray emission measurements were performed on iron-enriched milk protein concentrates. The possibility of measuring Fe content by photoacoustic measurements in the visible region of the spectrum is discussed.

Moisture sorption isotherms of khoa were determined at 15, 25, 35 and 45 °C over a water activity (aw) range of O·l 1–0·97. The isotherms were sigmoid (Type II) and were fitted to the Guggenheim–Anderson–de Boer (GAB) equation (Bizot, 1983). The aw of freshly made khoa was found to be 0·96. The aw of khoa increased with increasing temperature up to 0·9; above this the effect of temperature on aw diminished. The monolayer moisture content of khoa decreased with increasing temperature. The equations describing the temperature dependence of GAB constants were determined in the form of the Clausius–Clapeyron equation. The net isosteric heat of desorption of khoa decreased rapidly until a moisture level of 0·1 g/g solids, and approached a constant value of 0·43 kJ/mol above a moisture level of 0·25 g/g solids.

Electrophoretic analysis of the action of two commercial enzymes, Neutrase 0·5 and MKC Fungal Protease, on whole casein and αs-, β- and κ-caseins from cows' and ewes' milk showed that Neutrase 0·5 chiefly degraded β-casein, giving rise to peptides soluble at pH 4·6 detectable by PAGE. In contrast, although MKC Fungal Protease caused intense hydrolysis of bovine β-casein, in ovine casein it resulted in more active degradation of αs- than β-casein. The latter enzyme did not produce peptides soluble at pH 4·6 detectable by PAGE. Both enzymes degraded κ-casein, yielding a breakdown product that exhibited an electrophoretic mobility similar to that of the breakdown product produced by the action of commercial rennet.

Skim milks were prepared from skim milk powder at several concentrations between 10 and 25% total solids and portions were pH-adjusted to between pH 6·3 and 7·1 and heated at 120°C. After ultracentrifugation (88000 g for 90 min), the supernatants were analysed using gel electrophoresis to determine the concentrations of β-lactoglobulin, α-lactalbumin and κ-casein. The dissociation of κ-casein from the micelles was dependent on both the pH and the total solids content of milk before heating. Both higher pH (in the range 6·5–7·l) and higher concentration increased the extent of dissociation. A further series of samples were heated for 2–11 min at 120°C at pH 6·55. κ-Casein dissociation increased with concentration and with heating time. It was concluded that as the milk increased in concentration, the pH at which micellar κ-casein dissociated on heating was lowered.

This work aimed to relate changes in rheological behaviour with structural details of different sweetened condensed milk samples as well as to clarify the role of additives in delaying gelation. To study the process of gelation, rheological techniques were applied to fresh and stored condensed milk samples for times and at temperatures known to favour gelation (45 d at 37 °C). Electron microscopy techniques were used to relate the rheological changes with observed structural details. Examination by electron microscopy showed that micellar aggregation and association of casein micelles with fat globules increased during storage. The whey proteins appeared to play an important role in the thickening process.

Volatile components of four batches of Manchego cheese were studied throughout their ripening process. Samples were submitted to a micro simultaneous distillation-extraction procedure, and then analysed by capillary gas chromatography (GC) and combined GC and mass spectrometry. Among other components, four homologous series of compounds were found: free fatty acids, methyl ketones, ethyl and methyl esters of even fatty acids. Ripening stage can be approximately estimated from the concentrations of some components.