Partly purified extracellular aminopeptidase from Brevibacterium linens was used to accelerate Cheddar cheese ripening. It was found that although the aminopeptidase was unstable in acidic buffer, it was highly stable in Cheddar cheese; negligible amounts of the enzyme activity were lost during 3 months' maturation. A better score for flavour in sensory analysis of enzyme-treated cheese was obtained by the combination of the aminopeptidase and a commercially available metalloproteinase (Neutrase) than by the metalloproteinase alone.

Although proteolysis is regarded as the most important biochemical event in the ripening of most cheese varieties (Rank et al. 1985; Grappin et al. 1985; Law, 1987), current methods for the fractionation of cheese N, yield very heterogeneous fractions (Rank el al. 1985; Grappin et al. 1985; Fox, 1989). This communication describes the results to date of a study undertaken to develop a comprehensive scheme for the fractionation of cheese N.

There are estimated to be 600000 camels (Camelus dromedarius) in Kenya (Wandera, 1985). Almost 80% of these are kept by pastoral tribes living in arid areas in eastern and north-eastern parts of the country. In these regions, camels are important dairy animals. A camel in north-east Kenya can be expected to yield about 4 kg milk daily as compared with 0·5–1·5 kg for a cow in the same area. Most of the camel milk is consumed in the form of fermented milk. The milk is allowed to ferment naturally at ambient temperature and without prior heat treatment until it turns sour. The resulting fermented camel milk is known as Susa. Due to the spontaneous nature of the fermentation, this traditional method results in a product with varying taste and flavour and is often of poor hygienic quality, in addition, because of the limited scale of production, the product can be sold only in the immediate vicinity of the herd. For production of fermented milk under controlled conditions, thermophilic or mesophilic lactic acid cultures are normally used. In warm countries, mesophilic lactic cultured milk offers some advantages, as it can be incubated at ambient temperature (20–30 °C) and the fermentation stops at 1–1·2% lactic acid, eliminating the need for cooling to stop further souring as occurs in the case of yoghurt (Kurwijila, 1980).

The interactive effect of two of the major Cheddar cheese flavour compounds, H2S and methanethiol, was studied in milk-fat-coated microcapsules containing Brevibacterium linens, cysteine and methionine. Two substrates, cysteine and methionine, were either co-encapsulated with B. linens or encapsulated separately with B. linens and mixed together. The production of H2S from cysteine in microcapsules was slightly reduced by the methanethiol produced from methionine. However, the presence of H2S stabilized methanethiol produced from methionine in milk-fat-coated microcapsules and subsequently increased the level of methanethiol. Without H2S, the methanethiol in milk-fat-coated microcapsules disappeared rapidly after 48 h. By making separate microcapsules at each optimum environment and mixing together, the cells could operate in the protected and ideal environment of the microcapsules to form products that diffused out and acted as reactants with products from other microcapsules.

Lactodynamographic analysis is commonly employed to evaluate the clotting aptitude of milk (Frentz, 1965; Tarodo de la Fuente et al. 1969). The individual lactodynamographic response can therefore be used in selection as a quantitative trait. The aim of this work was to estimate either the biological repeatability of the lactodynamographic response or the repeatability of the analytical method itself.

The water activity of fresh cheese made from cows', ewes' and goats' milk, as well as curd cheese and whey cheese, responds to the equation aw = 0·9719−0·0044[NaCl] + 0·0041 pH, with a square of multiple correlation coefficient R2 = 0·95. This equation was obtained by multiple regression analysis of data point sets of the variables x (g NaCl/100 g moisture), y (pH) and z (water activity, aw) determined as the average of values provided by six samples of each unripe product. Application of this equation to 30 individual samples of unripe dairy products investigated yielded calculated aw values within ±0·005 and ±0·002 aw units of the experimental water activities in 86·7% and 53·3% of the samples, respectively. These results are closer to the experimental values than those provided by earlier predictive equations.

Milk contains indigenous plasmin-like and thrombin-like proteinases and also exogenous proteinases, mainly extracellular enzymes of psychrotrophic bacteria. Their activities lead to protein hydrolysis which may have important effects on milk quality and cheese yield. It is also useful to estimate proteolysis and its development during cheese ripening.

Milk opacity or turbidity due to calcium phosphocaseinate micelles and fat globules necessitates preliminary sample treatment, i.e. precipitation and filtration before spectrophotometric measurement. It is possible to eliminate these firs steps of sample treatment by dispersing the colloidal constituents. Some authors have already used different dissolving mixtures for sample analysis. Nakai & Anh Chi Le (1970) were the first to determine protein contents of milk and dairy products by absorbance measurement at 280 nm after clarification. Bosset & Blanc (1977) used the Biuret method. Owen & Andrews (1984) measured free amino groups with 2,4,6-trinitrobenzene sulphonic acid (TNBS). Humbert et al. (1982) and Stead (1987) used synthetic or specific substrates for proteinase determination.

In this work, we report the application of a new dissolving reagent to measure free amino groups in milk and cheese.

Proteinase, peptidase and esterase activities were detected in cell-free extracts of four Lactobacillus delbrueckii subsp. bulgaricus and four Streptococcus salivarius subsp. thermophilus strains. Post-electrophoretic detection was based on hydrolysis of L-leucine-βnaphthylamide and α-naphthylacetate. The substrates L-leucine-p–nitroanilide, N-acetyl-L-alanine-p–nitroanilide, 2-nitrophenylbutyrate and 4-nitrophenylbutyrate were used for the spectrophotometric detection of the enzymes. Estimation of total proteolytic activity was based on hydrolysis of bovine whole casein. Interesting differences were observed between the two French strains and those isolated from traditional Greek yogurt.

Titration is an old method used for the determination of acidic and basic groups of a molecule. Several metabolic end-products in cheese are compounds with acidic or basic groups, e.g. propionic, acetic and lactic acids, amino acids, amines and ammonia. A number of colorimetric and fluorometric methods have been used to study proteolysis in cheese; some are time-consuming with several steps, and certain of the reagents are hazardous.

We have used acid-base titration for analysis of the products of proteolysis in Swiss-type cheese. The titration system was that originally developed for use in silage analysis (Moisio & Heikonen, 1989) and applied to cheese it has proved to be useful, easy and rapid with an accuracy similar to that achieved by colorimetry.