The peptide Pro130–Thr–Ser–Thr–Pro–Thr–Ile–Glu–Ala–Val–Glu140–Ser–Thr–Val–Ala–Thr–Leu–Glu–Ala–Ser–Pro150–Glu–Val–Ile, which corresponds to residues 130–153 of κ-casein B, was synthesized and the conformation of the peptide in solution investigated by circular dichroism (CD) spectroscopy, structure prediction algorithms and 1H-nuclear magnetic resonance spectroscopy. In a solution containing the structure-enhancing solvent trifluoroethanol the CD spectrum was typical of a peptide in the α-helical conformation and nuclear magnetic resonance showed that the amino acids between Ile136and Ser149 (κ-casein numbering) were predominantly in the α-helical conformation, but that Pro130 to Thr135 and Pro150 to Ile153 were not. In addition, Thr133–Pro134 and Ser149–Pro150 were primarily in the trans conformation, the residues from Thr131 to Thr135 were in unordered structures and the residues from Glu151 to Ile153 were in an extended conformation. Residues Glu137 to Glu140 and Thr145 to Ala148 also displayed some 310-helix character. When the peptide was dissolved in 10 mM-cetyltrimethylammonium chloride solution at pH 6, the CD spectra indicated that the proportion of helical structure was comparable to that of the peptide in trifluoroethanol solution (400 ml/l), whereas when the peptide was dissolved in buffer alone or in 10 mM-SDS solution, the CD spectra were consistent with a low helical content. Acidification of these solutions to pH 2·85 resulted in a slight increase in the helical content of the peptide in buffer and more markedly in buffer containing SDS. When the peptide was in 5 mM-CaCl2 solution at neutral pH, the CD spectrum indicated that some ordered structure was present. Taken together these results indicate that the ionizable residues Glu137, Glu140, Glu147 and Glu151 could be important in determining the stability of the putative helix. The structure predictions found that the sequence from Glu137 to Pro150 would be more likely to be in a helical than any other conformation in the intact bovine protein, but that pig, sheep and goat κ-caseins did not give a prediction of a strongly helical region in this part of the molecule.

Capillary electrophoresis using a hydrophilically coated capillary and a low pH buffer containing urea has been used to follow the proteolytic action of plasmin and chymosin on isolated casein fractions, whole casein and individual milk samples selected on the basis of their genetic variants. Several of the main casein breakdown products were identified. These included, among others, γ1-casein (CN) A1, γ1-CN A2, γ1-CN B, γ1-CN C, γ2-CN A, γ1-CN A3, γ2-CN B, γ3-CN A and γ3-CN B, as well as proteose peptones, arising from the action of plasmin on the different genetic variants of β-CN. αs1-I-CN and αs1-CN f(1-23) from αs1-CN, and para-κ-CN and caseinomacropeptide from κ-CN produced by chymosin action were also separated. The knowledge of their migration times provided information on the extent and origin of casein hydrolysis in both milk and cheese, as found in samples of proteolysed milk or fresh cheese.

We propose an alternative method for casein genotyping, generally carried out using polymerase chain reaction followed by restriction fragment length polymorphism analysis. Application of the single-strand conformation polymorphism technique detects nucleotide changes in the fragment amplified by means of polymerase chain reaction and thus avoids the use of restriction enzymes. A 453 bp fragment from exon IV of κ-casein has been amplified. The two variants (A and B), found with the highest frequencies in most bovine breeds and included in some dairy cattle selection schemes, can be discriminated using single-strand conformation polymorphism analysis of heat denatured fragments in acrylamide–bis-acrylamide (100[ratio ]1) gels followed by silver staining. κ-Casein genotyping is therefore simplified, although variants A and E on the one hand, and B and C on the other, are not distinguishable with this technique.

The performances of falling film evaporators used in the New Zealand dairy industry for concentrating whole milk were evaluated by determining kg steam used/kg water evaporated, and the heat transfer coefficient of each pass in the evaporators. A specially written computer program was used to calculate the results. The heat transfer coefficients varied from 0·3 to 3·0 kW/m2K, and the steam consumption from 0·10 to 0·39 kg steam/kg evaporation, depending on the number of effects. The steam consumptions for whole and skim milk were similar. The momentum of the vapours passing down the tubes, the temperature difference across the tubes, the viscosity of the feed and the liquid loading were found to be the main factors controlling the heat transfer coefficient. A correlation between the heat transfer coefficient and these factors is presented, and other factors likely to have an influence on the performance are discussed. The correlation is compared with that obtained for skim milk.

Aminopeptidase P was purified 65·3-fold from the cytoplasm of Lactococcus lactis subsp. cremoris AM2 with a 5·8% yield. The purified enzyme was found to consist of one polypeptide chain with a relative molecular mass of 41600. Metal chelating agents were found to be inhibitory and Mn2+ and Co2+ stimulated activity 7-fold and 6-fold respectively. The purified enzyme removed the N-terminal amino acid from peptides only where proline (and in one case alanine) was present in the penultimate position. No hydrolysis was observed either with dipeptides even when proline was present in the C-terminal position or when either N-terminal proline or pyroglutamate was present preceding a proline residue in the penultimate position of longer peptides. On the basis of this substrate specificity either aminopeptidase P or post-proline dipeptidyl aminopeptidase are necessary along with a broad specificity aminopeptidase to effect complete hydrolysis of casein-derived peptides containing a single internally placed proline residue. However, both aminopeptidase P and post-proline dipeptidyl aminopeptidase would be required together with a broad specificity aminopeptidase in order to completely hydrolyse casein-derived peptides that contain two internally placed consecutive proline residues. As bitter casein-derived peptides are likely to contain either single prolines or pairs of prolines, aminopeptidase P appears to be an important enzyme for debittering.

A rapid procedure for isolating genomic DNA from milk samples has been devised, based on the use of Chelex resin. By using this protocol, genomic DNA was extracted from milk samples from 15 cows and 15 goats. The suitability of these DNA preparations as a template for performing the polymerase chain reaction (PCR) was tested by amplifying three different loci of the bovine genome: exon 4 of the κ-casein gene and the INRA5 and INRA23 microsatellites, together with two others: exon 19 of the αs1-casein gene and exon 2 and part of intron 2 of the DRB gene of the caprine genome. No amplification products could be obtained from any samples at 30 cycles. In contrast, at 45 cycles the number of amplified samples ranged from 86 to 100% and at 65 cycles all the DNA targets were amplified, indicating that the number of cycles was a critical factor to be optimized for obtaining the desired PCR target. These results suggest that this method may be a useful tool for analysing genetic polymorphism at the DNA level by PCR and relating it to milk composition and other traits of economic interest.

Heavy coagulum formation in lactic casein manufacture results in high casein concentration, high starter bacteria concentration and low pH on the base of the coagulation vessel. No heavy coagulum was formed during slow coagulation of skim milk with glucono-δ-lactone, indicating that starter bacteria were necessary. Skim milk preheat treatments severe enough to cause denaturation of the more heat- stable whey proteins, α-lactalbumin and β-lactoglobulin, also caused a reduction in heavy coagulum formation, suggesting the heat inactivation of a co-factor involved in the process. These results suggest a mechanism for heavy coagulum formation similar to that for minor sludge formation in cottage cheese manufacture. Starter bacteria would clump together, probably assisted by a co-factor. These starter clumps would generate a region of low pH causing isoelectric precipitation of casein in and around the clumps. The clumps would then settle forming a layer on the base of the coagulation vessel which might become further compressed by gravity.

Simultaneous measurement of viscosity and degree of proteolysis of κ-casein, α, in skim milk at three different concentrations of rennet and three different concentrations of casein showed that viscosity was a function of α until the former reached its minimum value. Then the relation between viscosity and α depended on the enzyme concentration. This is considered to be caused by casein micelles starting to aggregate at this point. The minimum in viscosity was found when α was between 0·6 and 0·7, indicating that casein micelles started aggregating at a lower degree of proteolysis than was believed earlier. The rate of gel firming was found to depend on the concentration of added rennet even after κ-casein was completely proteolysed. The results demonstrated that both proteolysis of κ-casein and aggregation of casein micelles must be taken into account when modelling the renneting reaction.

Two sandwich enzyme-linked immunosorbent assays (ELISA) have been developed for quantitation of bovine milk plasminogen and plasmin. The assays used two monoclonal antibodies, one specific for plasminogen and the other specific for plasminogen plus plasmin. Plasmin concentration was obtained by subtracting the first concentration from the second. The assays were sensitive (linear range, 5–75 ng/ml), repeatable (CV, 8 and 5% for plasmin and plasminogen titration respectively), specific (no cross reactivity with the major milk proteins) and directly applicable to skimmed milk with no particular pretreatment of the sample. However, ELISA did not permit complete titration of milk plasminogen and plasmin (only 60–90% could be quantified). This lack of accuracy was due to casein interfering with the plasmin–plasminogen titration by ELISA. Results obtained with ELISA or an enzymic technique on 20 milk samples collected from individual cows milked throughout pregnancy showed that the ELISA was particularly suitable for analysis of late lactation or mastitic milk where proteinase inhibitors interfered with enzymic quantitation.

The growth of four Bacillus cereus strains producing diarrhoeal toxin at 32°C (F4433/73 and 29.155, isolated on the occasion of foodborne outbreaks, and F4581/76L and F4581/76R, two variants of a clinical strain), a weakly toxigenic strain isolated in routine analysis of food (3505M) and an emetic isolate (F3502/73) was investigated at low temperature. Biomass was determined by protein assay. Generation times were: for strain F3502/73, which grew at [ges ]12°C, 8·71 h (at 12°C); for other strains, which grew at [ges ]10°C, 10·2 to ∼18·9 h (at 10°C). Toxin production during growth was evaluated by a commercial kit (Oxoid) and by a toxicity test on Chinese hamster ovary cells. Strains F4433/73 and F4581/76, secreting high levels of diarrhoeal toxin during the exponential phase at 32°C, produced high levels of toxicity at 10°C until the stationary phase. Strain 29.155 had decreased toxin production at 10°C. Toxicities for cellular extracts remained low when compared with culture filtrates. A correlation was found between the toxicity values given by the two detection methods tested, and the suitability of both methods for the detection of potential poisoning isolates is discussed.

Recent findings suggest a protective role for folates in the reduction of neural tube defects and possibly also coronary heart disease and cancer. Consequently, an increase in the daily intake of folates is warranted, which emphasizes the need for quantitative as well as qualitative measurements of dietary folates. Milk plays an important part in the food chain in many Western countries today. Several studies suggest that folate-binding proteins might have an impact on folate absorption and therefore their concentrations are also important. The mean concentration of the predominant form of folate, 5-methyltetrahydrofolate (5-CH3THF), was determined using HPLC in thirteen selected dairy products; skim milk powder, two pasteurized milks, UHT milk, two fermented milks, three whey products and four different cheeses. All results were corrected for recovery by spiking the samples with 5-CH3THF. Effects of storage of dairy products on 5-CH3THF concentrations were also investigated; generally small and insignificant fluctuations were found, except for hard cheese, in which 5-CH3THF decreased significantly. There was a significant seasonal variation in the folate concentration of pasteurized milk which peaked in the summer months. The concentrations of folate-binding protein in skim milk powder and pasteurized milk analysed using an enzyme-linked immunosorbent assay were similar. UHT milk and fermented milk, both of which are processed at temperatures >90°C, contained significantly lower concentrations of folate-binding protein.

In all, 4379 isolates from 35 products, including 24 artisanal cheeses, were surveyed with a view to identifying strains that could be used as starters in commercial dairy fermentations. Of the isolates, 38% were classified as Lactococcus, 17% as Enterococcus, 14% as Streptococcus thermophilus, 12% as mesophilic Lactobacillus, 10% as Leuconostoc and 9% as thermophilic Lactobacillus. Acid production by the isolates varied considerably. Of the 1582 isolates of Lactococcus and 482 isolates of mesophilic Lactobacillus tested, only 8 and 2% respectively produced sufficient acid to lower the pH of milk to <5·3 in 6 h at 30°C. In contrast, 53, 32 and 13% of Str. thermophilus, thermophilic Lactobacillus and Enterococcus isolates respectively reduced the pH to 5·3. These isolates were found only in some French, Italian and Greek cheeses. Bacteriocins were produced by 11% of the 2257 isolates tested and 26 of them produced broad-spectrum bacteriocins which inhibited at least eight of the ten target strains used, which included lactic acid bacteria, clostridia and Listeria innocua. The most proteolytic of the 2469 isolates tested were Str. thermophilus from Fontina cheese followed by Enterococcus from Fiore Sardo and Toma cheese and thermophilic Lactobacillus from all sources. Exopolysaccharides were produced by 5·3% of the 2224 isolates tested.

The susceptibility to penicillin-G of Staphylococcus aureus strains that cause mastitis was tested in milk and in Iso-sensitest broth (ISB). The minimal inhibitory concentrations (MIC) of β-lactamase-positive strains in milk were 10–100-fold those in ISB, whereas the MIC of β-lactamase-negative strains in milk were some 10-fold those in ISB; β-lactamase production was induced by milk in β-lactamase-positive strains. Much of the increase in resistance to penicillin-G caused by milk can be attributed to milk fat globules; the increase in resistance was related to the binding capacity of the bacteria to milk fat globules as well as to capsule formation by the bacteria. It appears that the binding of the staphylococci to the fat globules and bacterial capsule formation resulted in a biofilm type of growth. In this case, the staphylococci behaved differently from the planktonic type of growth in artificial broth medium in which antibiotic susceptibility testing is usually carried out.

Escherichia coli O157[ratio ]H7 in spiked samples of raw milk and ice-cream was enriched in tryptic soy broth for 4 h, captured by immunomagnetic separation, subjected to amplification by polymerase chain reaction of parts of the verotoxin genes (SLT-I and SLT-II), and detected by agarose gel electrophoresis. Using this method, as few as 1 cfu Esch. coli O157[ratio ]H7/g food could be detected in <10 h.

Confocal laser scanning microscopy was used to examine components of whole milk powder. Powder was dispersed in a glycerol suspension containing a fat-soluble fluorescent dye which was excited by laser light to produce images of the fat at the surface of and inside the whole milk powder particle. Lactose crystals, produced in the milk prior to spray drying, were seen to be in the diamond-like, α-hydrate form. The surface coverage and the stability of a wetting agent sprayed on to the surface of whole milk powder particles were examined. The wetting agent was shown to be located as a thin surface layer that partly migrated into creases and folds on the particle surface during storage at 37°C.

Whole cells of Lactobacillus delbrueckii subsp. bulgaricus CNRZ 397 (Lb. bulgaricus CNRZ 397) are able to hydrolyse α- and β-caseins. We have isolated a mutant of Lb. bulgaricus altered for growth in milk and unable to hydrolyse α- or β-casein. Normal growth was restored by adding amino acids or tryptone to milk. No significant difference between the peptidase activities of parent and mutant strains was observed. The cell surface caseinolytic activities of three lactobacilli species and Lactococcus lactis subsp. lactis (Lc. lactis) were compared. As expected, the characteristics of the cell surface proteinase activity of Lb. casei were similar to those of Lc. lactis. We showed that the cleavage specificities of the cell surface proteinase activities from lactobacilli were species-dependent and at least three types of activity were distinguished. The regulation of the biosynthesis of cell surface proteinase activities was medium-dependent and different within the Lactobacillus genus and even within the Lb. delbrueckii species. In contrast to Lb. bulgaricus, the cell surface proteinase activity of Lb. lactis was totally inhibited in a medium rich in peptides or amino acids. In contrast, the cell surface of Lb. helveticus probably displayed two proteinases with different cleavage specificities and with a biosynthesis regulation sensitive to different medium components.

Our aim was to develop an Emmental juice-like medium (EJLM) for the growth of propionibacteria. Cheese juices were extracted by pressing three different Emmental cheeses before entering the warm room (i.e. before propionic fermentation) and their composition was determined. They contained (g/kg) lactate 34·6–36·9, protein 23·8–29·6, NaCl 15–23, Ca 5·9–7·3 and no residual sugar, and had a pH of 5·4. In order to simulate this composition, EJLM was constituted from the microfiltrate of a milk, enriched in native casein and fermented by a Lactobacillus helveticus strain until lactose exhaustion. PO43−, K+, Mg2+, Ca2+ and Cl− were added to reach a mineral balance and an ionic strength close to those of juice. The main difference was the nitrogen content: that of juice was ∼3·5 times that of EJLM. Four Propionibacterium freudenreichii strains grew with similar growth rates in EJLM and in juice at 24°C in anaerobiosis and reached >3×109 cfu/ml. Using EJLM, Prop. freudenreichii TL 173 grew at pH 5·4 or 6·0, at 24 or 30°C, in semi-anaerobiosis (static culture) or anaerobiosis (CO2 atmosphere), showing a synergic inhibition of growth at low pH and temperature and in semi-anaerobiosis. EJLM provided a way to study propionibacteria in an environment similar to that in Emmental cheese.

A total of 50 isolates of Bacillus spp. and one reference strain were investigated for their growth at 6·5°C for 10 d, 30°C for 3 d and 40°C for 2 d. The results obtained differentiated three physiological groups: one clearly psychrotrophic (able to grow at 6·5°C in 10 d, but not at 40°C in 2 d), one intermediate in psychrotrophy (it grew at both 40 and 6·5°C) and one mesophilic (capable of growth at 30 and 40°C, but not at 6·5°C). The proportion of strains in the second group was higher among isolates of B. cereus than for other Bacillus spp. However, the proportion of real mesophilic strains was lower for B. cereus isolates. Psychrotrophic B. cereus grew better at both 6·5 and 30°C than other psychrotrophic Bacillus spp. Using eight strains, a correlation between differential growth at mesophilic temperatures (count at 30°C minus count at 40°C) and a standard psychrotrophic count at 6·5°C for 10 d (r=0·95) was obtained in mixed cultures when the psychrotrophic flora count was [les ]1 log (cfu/ml) lower than the mesophilic count.

Growth of Leuconostoc mesenteroides in milk was studied with respect to the proteinase and peptidase activities of the strains and their nutritional requirements. Ln. mesenteroides grew poorly in milk since none of the 14 strains studied exceeded 5×108 cfu/ml at the end of growth. Few strains displayed proteinase activity, and this did not contribute much to growth. The pattern of peptidase activities varied with the strain. Nitrogen starvation and a high requirement for Mn2+ were involved in the cause of growth deficiencies. Addition of amino acids, 50 mg Mg2+/l and 0·08–0·49 mg Mn2+/l stimulated growth of most leuconostoc strains up to 5×108 cfu/ml. Addition of 5 g glucose/l to milk containing amino acids, Mg2+ and Mn2+ or yeast extract stimulated the growth of seven and eight strains respectively up to 109 cfu/ml. No growth advantage was found in a N2 atmosphere. However, the addition of small amounts of Mn2+ to milk suppressed the inhibitory effect of aeration on the growth of Ln. mesenteroides UD23, suggesting a protective role of Mn2+ against O2 toxicity.

Several peptides were isolated from the diafiltration retentate, prepared using 10 kDa membranes, of the water-soluble extract from a commercial mature Cheddar cheese and identified by amino acid sequencing and mass spectrometry. Most of the peptides were from the N-terminal half of β-casein, but peptides from αs1- and αs2-caseins were also identified; the extract also contained α-lactalbumin. Identified peptides showed the important role played by lactococcal cell envelope proteinases in the degradation of primary proteolytic products from αs1- and β- caseins, produced by chymosin and plasmin respectively. Plasmin seemed to be involved in the hydrolysis of αs2-casein. Several phosphopeptides were identified and the action of phosphatase on these peptides was evident.