Part of the raw cheese base used for manufacturing high-fat processed cheese spread Was replaced by cows' (C) or buffalo (B) butter residue at levels of up to 20%. These residues were produced during the conversion of butter to butter oil by traditional moderate (T) or severe (M) heat treatments. The hypocholesterolaemic properties of butter residues incorporated into processed cheese were tested with rats (serum and liver cholesterol) and human beings (serum cholesterol). When rats were given the control cheese with no butter residue there were significant increases in both serum and liver cholesterol compared with rats given stock diet. However, when part of the raw cheese base was replaced by butter residue these rises were reduced by amounts proportional to the level of butter residue. The hypocholesterolaemic properties of the residues were in order CT > BT, CM > BM, and incorporations of 10% CT, 15% BT, 15% CM and 20% BM were found to nullify the hypercholesterolaemic effects of control cheese and restore cholesterol levels to their normal values. A small human trial gave similar results. Incorporation of butter residue slightly affected the chemical composition of the cheeses. Fat content and pH were little different; however, cheeses with butter residue tended to have higher total protein but lower soluble protein, tyrosine and tryptophan, and total volatile fatty acids. Rheological properties were not altered significantly. Incorporation of butter residue improved the organoleptic properties, particularly flavour intensity. Storage at 5 °C for 2 months did not alter the organoleptic properties and changes in chemical composition followed the normal pattern.

The proteolytic specificity of chymosin (EC 3.4.23.4) on bovine αs1-casein at 30°C in phosphate buffer, pH 6·5 and at pH 5·2 in the presence of 5% (w/v) NaCl was investigated. Peptides (pH 4·6-soluble) were isolated by reversed-phase HPLC and identified from their amino acid sequence; the identity of some peptides was confirmed by mass spectrometry and/or amino acid composition. The small peptides produced at pH 6·5 were Arg1–Phe23, Phe24–Phe28, Phe24–Leu40(?), Phe150–Phe153, Phe150–Leu156, Tyr154–Tyr159, Tyr154–Trp164, Asp157–Trp164 and Tyr165–Trp199. The same peptides, except Tyr154–Trp164, were produced at pH 5·2 in the presence of NaCl and, in addition, the peptides Arg1–Leu11, Phe24–Phe32, Lys102–Leu142, Ala143-Leu149 and Tyr165-Phe179. The rates of production of individual peptides differed under the two conditions studied but Arg1-Phe23 and Tyr165–Trp199 were the first and second peptides produced under both conditions. Pathways are proposed to interpret the proteolysis of αs1-casein in solution under the conditions of this study.

Spores of a known toxigenic and psychrotrophic dairy isolate of Bacillus cereus (HRM 44) were unable to grow and produce diarrhoeagenic toxin at 6 °C in creams and dairy-based products. These findings suggest that the production of B. cereus diarrhoeagenic toxin is unlikely to occur in creams and dairy-based products maintained within the cold chain. Growth and toxin production were readily demonstrated in creams and some desserts stored at 21 °C. Growth in creams was associated with obvious spoilage. However, in the flavoured desserts, spoilage was not always obvious before significant growth of B. cereus and toxin production had occurred. Dairy desserts with high sugar content and/or low pH did not support toxin production and these findings are discussed.

Broth cultures supplemented with high levels of sugars, particularly glucose at > 50 g/1, did not support diarrhoeagenic toxin production by psychro-trophic Bacillus cereus despite growth to high counts ( ∼ 107/ml) over a 4 d period of incubation at 21 °C. In contrast, starch levels of 10 and 50 g/l actually enhanced toxin production. Toxin production was also affected by pH levels of broth cultures, and was concomitant with alterations in bacterial growth. These findings help to explain variations in toxin levels previously found in some dairy desserts, which were thought to be associated with pH and sugar content (Sutherland, 1993).

Using polymer producing (ropy) strains of lactic acid bacteria it was possible to reduce considerably the syneresis of yogurt, even with 12% total milk solids. The viscosities obtained with these strains were also similar to those obtained using normal strains and milk with 17% total solids content. The concentration of milk and the polymer produced by ropy starters had a synergic effect in increasing viscosity. Polymer production was not affected in most cases by milk concentration. One type of ropy culture (Wiesby) seemed to produce a different kind of polymer as it could not be determined by alcohol precipitation, in spite of being able to reduce syneresis and increase viscosity in yogurt. A limited number of yogurts were evaluated organoleptically, one prepared with a ropy starter strain (NCFB at 12, 14·5 and 17% total solids) and one prepared with a non-ropy strain (LL-I at 17% total solids). The results suggest that the ropy strain yogurts had different mouthfeel from the non-ropy strain yogurts; the most acceptable product overall was the ropy strain made with 12% total solids.

Comparing the primary sequences of bovine and ovine milk proteins, some short peptide fragments are cow-specific, in particular the 141–148 fragment of bovine αs1-casein, which is deleted in its ovine counterpart. The 140–149 peptide was chemically synthesized on a solid phase matrix and directly used as an immunogen to produce polyclonal monospecific antibodies in rabbits. These antibodies recognized this fragment both on the peptidyl resin and in the native protein. They appeared to be monospecific, since no antigen–antibody complex was formed with homologous ovine or caprine proteins. Subsequently, a competitive enzyme-linked immuno-sorbent assay was successfully developed for the detection of defined amounts of cows' milk in sheep's milk from 0·125 to 64% (v/v) and in cheese from 0·5 to 25% (v/v) that was not influenced by heat treatment of milk or the degree of ripening of cheese.

We report the isolation and immobilization of a nisinogenic strain (NZ1) of Lactococcus lactis subsp. lactis, active on gas-forming lactate-fermenting clostridia responsible for late blowing of Asiago and Montasio cheeses. The bacteriocin (nisin) produced by strain NZ1 is pronase-sensitive and is released in culture media during the growth phase. Using the sensitive indicator strain Lactobacillus delbrueckii subsp. bulgaricus NCDO 1489, a rapid microtitre plate based assay was developed for quantitative determination of the bacteriocin produced by NZ1 cells, either free or immobilized in gel beads. Scanning electron microscopy of cells immobilized in calcium alginate coated beads and viable counts of the surrounding medium showed that no cell leakage occurred during a 24 h assay. The bacteriocin released from immobilized cells reached, after 5 and 24 h, concentrations comparable to that of the free cell system after 3–4 h incubation in culture media.

Listeria monocytogenes survived and, under most conditions, multiplied when inoculated directly into the cheese milk of laboratory made Camembert cheeses. The rate and extent of growth was reduced at lower storage temperatures. Significantly higher rates of growth occurred at the surface compared with the centre of the cheeses, and these were probably associated with increased pH and proteolysis at the cheese surface due to the mould ripening process. Similar results were obtained with Camembert cheeses surface inoculated after manufacture. There was also temperature-dependent growth of List, monocytogenes on a range of inoculated commercially manufactured soft cheeses. Significant growth occurred in Cambazola, French and English Brie, blue and white Lymeswold, French Camembert and Brie with garlic. Little if any growth occurred in blue and white Stilton, Mycella, Chaume and full fat soft cheese with garlic and herbs at the temperatures examined.

Sixteen rennet casein samples and the corresponding sixteen cheese analogues made from these caseins were obtained from a local processor. Following storage at 4°C for 6 months, pH 4·6-soluble N contents of the cheese analogues increased, while αs- and β-casein contents, elastic moduli and apparent relaxation times decreased. The β-casein and pH 4·6-soluble N contents of the 6 month old products were significantly correlated, both were significantly correlated with the plasmin (EC 3.4.21.7) contents of the corresponding rennet caseins from which the analogues were made and all three were significantly correlated with the rheological properties of the stored analogues. Of the compositional characteristics, only percent moisture in non-fat substance and percent fat in dry matter were significantly correlated (P < 0·05) with any of the proteolytic or rheological properties. The sixteen cheese analogues were also assayed at one time, when they ranged in age from 1 to 35 weeks, and highly significant interrelationships between plasmin activity, proteolysis and rheological properties were established. The results indicate that the plasmin content of rennet caseins had a significant influence on the proteolytic and rheological changes that occurred during storage of cheese analogues made from these caseins.

A computerized, coaxial cylinder viscometer, controlled by a function synthesizer, was used to study the stress characteristics of acid fresh cheeses containing 7·5–8·5% protein and 0–20% fat in total solids, as a function of shearing time (0–3600 s) and of shear rate (0–4·5 s–1). For a given shear rate the stress diminished with time, following an exponential equation. The hypothetical final stress level (Sf for the shearing time t = ∞) represented about half its initial value (Si). For four cheese samples, the time required for the stress to become equal to Sf + (Sf – Sf)/e was 340–560 s. The stress-shear rate relation indicated shear thinning behaviour. The effect of shearing time showed that the material was thixotropic. The cheeses exhibited plastic flow. The initial and final stress levels were hyperbolic functions of the shear rate. Bingham final yield stress (for t = ∞) was 14–50 Pa. It was 15–25% higher for increasing than for decreasing shear rate. Bingham viscosity (for t = ∞) was between 2 and 6 Pa s. It was 10–20% higher for decreasing shear rate. The initial level of Bingham viscosity was ˜ 30% higher than its final value.