The present study describes the purification of a post-proline dipeptidyl aminopeptidase from the cytoplasm of Streptococcus cremoris AM2. On the basis of its elution from a calibrated Sephadex G200 column, the enzyme had a molecular weight of 117000 and exhibited a broad pH optimum activity between 6·0 and 9·0. The activity was most comprehensively inhibited by phenylmethylsulphonylfluoride and more modestly inhibited by 1,10-phenanthroline and 8-hydroxyquinoline but not by EDTA. A range of peptides containing either proline or alanine as the penultimate amino acid residue could act as substrates. The presence of proline on the carboxy side of the scissile bond prevented hydrolysis. However the enzyme could release Pro-Pro from Pro-Pro-Gly-Phe-Ser-Pro. The significance of this substrate specificity is considered in the context of removal of either single proline residues or prolylproline sequences from oligopeptides during cheese ripening.

A total of 116 molecular species of triglycerides were identified in milk fat, using a combination of HPLC and GLC. Triglyceride composition was predicted from the random composition, which was calculated on the basis of the mole fractions of the main fatty acids making up the total triglyceride fraction. The qualitative composition of the milk fat was similar in cows', ewes' and goats' milk. In all three milks the partition number of the main triglycerides was 46, but the proportions of the triglycerides with partition numbers of 34, 38, 42, and 48 exhibited substantial differences among the milks of the three species.

Both whey protein isolate (WPI) and Na caseinate gave similar emulsion capacity-concentration profiles at pH 7. The emulsion capacity of aqueous solutions of WPI was relatively independent of pH while the values for caseinate fell in the isoelectric region. Saline (7·59 g/1) compared with aqueous solutions improved emulsion capacity, particularly above pH 7. At low concentration, WPI had greater emulsifying capacity and gave finer globule dispersion than sodium caseinate. Increase in emulsifier concentration, within a certain range, increased stability, improved dispersion and reduced capacity. Ethanol up to about 20% w/w improved the emulsion capacity of both proteins while at 50% the phases separated. Heat treatment (> 65 °C) of WPI solution before emulsion formation impaired capacity and stability. Heat treatment of preformed WPI emulsions did not have the same effect. Na caseinate and WPI each formed a thin continuous layer of relatively uniform thickness over the oil globule surfaces while denatured WPI appeared aggregated and relatively unevenly distributed at the oil-aqueous interface.

Samples of UHT milk were contaminated at two levels (102 and 104 cfu/ml) with 52 bacterial strains associated with spoilage. Samples were assayed, using various tests, to determine efficiency in detecting non-sterility after an incubation period shorter than that advised in standard procedures. The tests performed were colony count, titratable acidity and pH modification, organoleptic assessment, stability towards ethanol (68, 80 and 88% v/v), resazurin and nitrate reduction. After 3 d incubation at 30 °C every case was detected by the resazurin test with a good statistical probability (P = 0·999). Reliability was not so high in all other tests, even after 14 d incubation.

Acid casein was prepared under various conditions of pH, concentration and wash regimens (temperature, time, agitation and wash water ratio) according to a response surface experimental design. The resulting product was analysed for water-holding capacity and loss of fines, and the structure of the curd assessed by scanning electron microscopy. Concentration, pH and temperature were the major factors in terms of influencing fines, water-holding capacity and the microstructure of the curd. Fines could be minimized by selecting appropriate pH, concentration and agitation conditions. The final moisture content of the curd was mainly controlled by pH, concentration and temperature and may aid in the selection of parameters to obtain good dewatcring and drying efficiencies. Microstructural analysis visualized the effects of pH, concentration, temperature and their interactions on the casein structural network and the effects of the process parameters are discussed in relation to those changes. Response surface methodology was shown to be an effective tool for examining which variables used in the process were significant and provided guidance as to the type of physicochemical mechanisms affecting the system.

Photoacoustic signals in the near infrared region (~ 2·75 μm) have been investigated using amplitude and phase information to determine their effectiveness in estimating the water content of milk. The results obtained showed a good correlation between the photoacoustic signal and water concentration (R2 = 0·98) even though photoacoustic saturation was present in the range of wavelengths studied. However, this is not recommended as a fast technique for water determination because of inherent phase fluctuations in the samples analysed.

The formation of hydrophobic sites on the surface of casein micelles as a consequence of rennet action has been followed through the binding rate of a fluorescent probe and its distribution between a free and an ‘aggregated’ protein fraction. The variation of this parameter has been related to clotting time and curd hardening kinetics.

Results show that a first aggregation of casein through hydrophobic sites interaction began as soon as rennet was added to milk. At the natural pH of milk, the sol-gel transition occurred when all the casein micelles were already involved in large aggregates. This was not the case with slightly acidified milk (pH 6·5 and 6·3) where clotting occurred well before the first aggregation step had been completed. The surface hydrophobicity of casein continued to increase in the curd due to continuing enzymic action and structural rearrangements. When this process has been completed the hardening of curd proceeds at an accelerated pace until it reached its maximum asymptotic value.

Prolidase activity from cytoplasm of Lactococcus lactis subsp. cremoris (Streptococcus cremoris) AM2 was partially purified. The enzyme had Mr 42000 and optimum activity between pH 7·35 and 8·25 in citrate, phosphate and borate buffers while in a universal buffer system an optimum pH between 8·3 and 9·0 was observed. The activity was strongly inhibited by the chelating agents E.DTA, 1,10-phenanthroline and 8-hydroxyquinoline. Inhibition was also noted with dithio-threitol, N-ethylmaleimide and bacitracin. The enzyme was active on all amino-acylproline substrates tested except Gly-Pro and Gip-Pro and also showed activity against Pro-Pro. While most prolyl amino acids tested were not hydrolysed, hydrolysis was noted with Pro-Ala and Pro-Val. Km values of 20 mM and 10 mM were obtained with Phe-Pro and Met-Pro respectively; however, substrate inhibition was observed with Ile-Pro and Leu-Pro.

The effects of chromatographically pure bovine chymosin and pepsin A on cheese curd firmness have been investigated. The enzymes were studied in parallel experiments, using a Formagraph, under which the following parameters were tested and varied: reconstituted skim milk powder, bulk raw milk, CaCl2 addition, enzyme concentration and temperature. Chymosin had a slightly better rate of curd firming (P ≤ 0·001) and better curd firmness 30 min after gelling point compared to pepsin under all of the conditions investigated. However, as regards curd firmness in cheese manufacture, this difference is only a question of time, since pepsin will give the same degree of curd firmness of the cheesemilk as chymosin after some minutes. The biggest disadvantage with pepsin is its strong pH dependence, which can lead to very long gelling times and weak curd if the pH is too high and the rennet is dominated by pepsin.

Effect of heating and of some chemical treatments on the proteolysis of skim milk caused by its major proteinase system, plasmin, was investigated. Degree of proteolysis was expressed as the increase in γ-casein content of skim milk (mg/ml) after incubation with 0·02% (w/v) NaN3 at 37 °C for 20 h. Apart from reduced proteolysis in skim milk heated at > 70 °C for 10 min or at 75 °C for ∼ 5 min, some enhancement (10–80%) was observed on heating at 55–63 °C for 10 min or at 60–65 °C for 15 s. Heating at pasteurization conditions (63 °C/30 min or 72 °C/15 s) had no apparent effect on the amount of proteolysis. Addition of ascorbic acid (0·2 mg/100 ml) or H2O2 (1/100 ml) to unheated skim milk also increased the amount of proteolysis by 38 and 28%, respectively. However, the extent of this increase diminished with increasing amounts of both chemicals. Furthermore, there was a steady increase (∼ 33%) in proteolysis with time of exposure to light for up to 3 h. These results suggest that oxidative conditions lead skim milk to enhanced proteolysis by inactivating an inhibitor(s) operating in the plasmin system in milk.

The effect of threonine on the biochemical activity of yogurt starter was investigated by the addition of L-threonine to cows' and goats' milk in two concentrations (5 and 10mg/100g). A separate experiment was conducted to investigate the effect of free glycine on the formation of acetaldehyde from yogurt starter. Increasing amounts of glycine (1, 2, 3, 4 and 5 mg/100 g) were added to samples of cows' milk. Addition of threonine to milk samples did not influence acid production, total bacterial growth, or the balance between cocci and rods in the yogurts. Raised levels of threonine resulted in increased production of acetaldehyde. This increase was more pronounced in goats' milk than in cows' milk. The production of acetaldehyde in cows' milk without added threonine was significantly greater than in goats' milk without addition of threonine. Addition of threonine did not influence the formation of diacetyl and α-acetolactic acid, whereas production of acetoin in both milks was reduced. No marked difference in the formation of C02 was observed in samples with or without addition of threonine. Addition of free glycine to cows' milk had a marked negative effect on the production of acetaldehyde by yogurt starter. Addition of 3 mg glycine/100 g cows' milk gave approximately the same amount of acetaldehyde as obtained in goats' milk yogurt. It is suggested that the relatively low amount of acetaldehyde usually observed in goats' milk yogurt is caused by a feed-back inhibition of threonine aldolase produced by the relatively high amount of free glycine present in goats' milk.

The protective effect of feeding milk fermented with a mixture of Lactobacillus casei sp. and Lb. acidophilus sp. against Salmonella typhimurium infection in mice was compared with that obtained feeding milks fermented with these microorganisms individually. The survival rate obtained after oral infection with Sal. typhimurium was 100% in mice pretreated by feeding during 8 d with the mixture of Lb. casei and Lb. acidophilus fermented milks. Similar treatments with the individual milks were ineffective. Moreover, mice became more susceptible to infection with Sal. typhimurium after such treatment. The colonization of liver and spleen with the pathogen was markedly inhibited by the pretreatment with the mixture of fermented milk, while such inhibition was not observed using the Lb. casei and Lb. acidophilus milks. The highest levels of anti-salmonellae antibodies in serum and in intestinal fluid were found in the group of mice fed with the mixture and with Lb. casei fermented milk respectively. However, this latter milk was not effective in protecting against Sal. typhimurium. When the mice were first infected with Sal. typhimurium and then fed with the mixture of fermented milks, pathogen colonization was not prevented. The results suggest that the augmentation of resistance to salmonellae caused by the treatment with Lb. casei-+Lb. acidophilus-fermented milk was due to the anti-salmonellae protective immunity mainly mediated by the mucosal tissue. Milk fermented with this mixture could be used as an immunobiological method to prevent gastrointestinal infection.

Levels of plasmin activity in various commercial cheeses are reported. The effects of some processing conditions (pH, cooking temperature and method of salting) on plasmin activity in cheese were investigated. Neither draining pH nor method of salting affected the activity. However, plasmin activity in cheese, rennet curd and micellar casein dispersions increased with increasing cooking temperature. The results suggest that plasmin activity in cheese depends on cooking temperature used during manufacture.

The influence of chromatographically pure bovine chymosin and pepsin A on cheese curd syneresis has been studied. The enzymes were compared in experiments using a microscale syneresis assay, where the light absorbance of the released whey was used as a measurement of the syneresis. The following parameters were investigated: reconstituted skim milk powder (RSM), bulk raw milk, CaCl2 addition and cutting time of the curd. The gelling (clotting) time was held constant under all experimental conditions to eliminate its influence on syneresis. The differences in influence on syneresis of the two enzymes were neither obvious nor uniform. As regards the syneresis of RSM, the influence of chymosin and pepsin seemed to be very similar, while the syneresis of bulk raw milk seemed to be more enhanced by chymosin. However, results were not sufficiently clear to prove any distinct differences between the enzymes. As long as the cheese curd is cut at a proper firmness, there are probably no practical problems in using bovine rennets with varying chymosin: pepsin ratio.

Physical properties of aerated, high moisture and low calorie table spreads are considerably influenced by the phase volumes of the mix components. Using a statistical approach developed for mixture experiments, the firmness values of milk fat-based table spreads at three different temperatures were quantified in terms of the phase volumes of the fat, water and gas present. A simple equation was found which related changes in relative firmness values with composition at any temperature. Absolute firmness values may be obtained from relative values by multiplying by a factor, F, which is the firmness value of the similarly treated milk fat. This factor is related to the solid fat content by the formula F = 10ks+c, where F is the firmness value in kg cm–2 of milk fat having a percentage solids content, s, k is the slope of a semi-logarithmic plot and c is a constant.

Properties of interest to cheesemaking were investigated with milks concentrated by ultrafiltration and heated at 90 °C or 95–97 °C for 15 s. The effects of light homogenization before concentration and of addition of CaCl2 after heating were assessed. No changes in casein–fat or casein–whey protein interactions were detected by electron microscopy. The heat denaturation of whey protein increased linearly as the milk became more concentrated. Coagulability by rennet after heat treatment increased with concentration to close to the unheated value in 3·5 to 4-fold concentrates. The decrease in curd firming rate by heat treatment was slightly affected by concentration and addition of CaCl2 restored some of the difference. Heat treatment reduced the rate of whey loss and slightly improved the curd structure but did not affect fat losses. Light homogenization slightly reduced heat denaturation of whey protein and whey loss.

Cheddar cheeses were made from milks concentrated approximately 2·3-and 3·4-fold by ultrafiltration, then heated at 90 °C for 15 s. They were compared with cheeses made from pasteurized unconcentrated milks and unheated 5·3-fold concentrates. The recoveries of fat and protein were comparable in the cheeses from heated 2·3-fold concentrates and unheated 5·3-fold concentrates and both were higher than the controls. However, fat losses were high in curds from 3·4-fold heated concentrates. The use of paired fast and slow starter strains and coagulation at 26 °C optimized the curd composition. Proteolysis and flavour development were only slightly reduced in all the cheeses from concentrated milks, perhaps aided by higher starter levels in the curd than in the controls. The structures of the curds from the heated concentrated milks were coarser than those of the controls and the curd particles fused poorly. This may explain the grainy, dry, crumbly texture in the cheeses from the heated concentrates. The texture was not improved by adding a bacterial proteinase to the curd.

Changes in a Greek traditional soft cheese, Kopanisti, were followed during ripening. Mature cheese contained > 52% moisture, 22·7% fat and 19·3% protein. The as-casein content was hydrolysed faster than the β-casein so that in mature cheese only 23% and 35% respectively of these proteins remained intact. Soluble N, TCA-soluble N, amino acid N and NH3 were present to the extent of 28·9, 24·6, 14·3 and 6·6% of total N respectively. Leucine, γ-aminobutyric acid, valine and alanine were the dominant free amino acids in the mature cheese. Lipolysis was intense and the total free fatty acid content of ripe cheese was ∽ 50 g/kg cheese. The rate of liberation of short-chain (C4–C8) fatty acids was higher than that of the long-chain. The characteristic rich flavour and peppery taste appeared after 16 d ripening and the best overall cheese quality was produced after 32 d maturation.

Twenty Lactococcus lactis subsp. lactis strains isolated from cheese made from raw milk have been, studied and compared with well known starter strains. Many of them produced acid bacteriocins resembling those of other lactic streptococci and their patterns of antimicrobial agent susceptibility were also similar to previously studied dairy strains. On the basis of their ability to clot milk they could be divided into two groups. Some strains were rapid fermenters, with high β-phosphogalactosidase and proteolytic activities. The other group had low β-phosphogalactosidase activity and coagulated milk slowly, but seemed to have proteolytic activity. All the strains had plasmid DNA in amounts and sizes similar to starter strains. A β-phosphogalactosidase probe was constructed with plasmid DNA from a Lc. lactis subsp. lactis NCDO 712 derivative. Southern hybridization located homologous sequences in chromosomal and plasmid DNA from the wild isolates that fermented lactose efficiently, but the degree of homology was lower than that observed between starter strains.

Oestrone sulphate is quantitatively a major oestrogen in pregnant goats rising gradually in concentration in milk and plasma as placental function develops (Heap et al. 1984). We report here that its concentration in mammary secretions increases sharply at the time of parturition. The results support the idea that oestrone sulphate in mammary secretions could be used as a predictor of the onset of parturition in dairy animals.