During a routine inspection of Cheddar cheese manufactured at a commercial factory in New Zealand, some lots of 6-month-old cheese were found to have developed a pinkish colouration on the surface of the 20 kg blocks of cheese. Colouration did not always occur uniformly on all six faces of the rectangular cheese block, or even on a single face of the block. Furthermore, not all blocks from within the same day's manufacture were equally affected. When an affected block was removed from its bag and cut across, colouration was sometimes found to penetrate approximately 1–2 cm down into the cheese. In those blocks where a plug of cheese had been removed previously, a pinkish zone surrounded the plug-hole cavity.

The pinkish colouration was observed to fade slowly (over about 12–24 h) when the cheese surface was exposed to air.

Annatto, known to cause pink discolouration in “coloured” Cheddar cheese (Govindarajan & Morris, 1973) and in processed cheese made using coloured Cheddar, was not used in the manufacture of the present cheeses and could therefore be excluded as a cause of the colouration.

The flavour profiles of all affected cheeses were considered by experienced industry cheese graders to be easily within the normal range of flavour profile expected for a cheese of this type i.e. there was no evidence of any off-flavour development.

The present short communication describes the microbiological and chemical investigations carried out to determine the origin and nature of the pinkish colouration in Cheddar cheese.

β-Lactoglobulin (β-lg), which is found in the milk of several mammal species, is the most abundant protein in bovine whey (McKenzie, 1971). The interaction properties of β-lg with a large variety of small hydrophobic ligands have been extensively studied (Sawyer et al. 1998). β-Lg belongs to the super-family of hydrophobic molecule transporters called the lipocalins, which characteristically bind hydrophobic ligands inside a central calyx (Godovac-Zimmermann, 1988; Perez et al. 1989; Brownlow et al. 1997; Wu et al. 1997; Qin et al. 1998).

At present, whey protein concentrates containing a high percentage of β-lg are commercially available in a very large quantity, but most of the processes used in milk technology involve heat-treatments that are known to affect the initial conformational state of β-lg. Such conformational changes have consequences for both the physicochemical and functional properties in food systems, including a decrease in the availability of lysine, due to Maillard reactions (Léonil et al. 1997) and a decrease in the affinity constant for binding to retinol (Laligant et al. 1991) and to flavour compounds (O'Neil & Kinsella, 1988). In the field of protein-volatile compound interactions, we have recently reported a significant enhancement of foaming properties of β-lg solutions (50 mM-NaCl, pH 6) in the presence of aroma compounds such as isoamyl acetate (Marin & Relkin, 1999), benzaldehyde (BZA; Marin & Relkin, 2000) and vanillin (Relkin & Vermersh, 2000). The observed increase in foaming properties of β-lg was postulated to be due to formation of surface active complexes between β-lg and aroma compounds.

In the present work we have investigated the interaction between a β-lg concentrate (prepared by ultra-diafiltration on an industrial scale) and BZA; particularly, evidence was sought for the presence of covalently bound monomers, dimers and lactolated monomers.

Molecular biology techniques have been used for species identification in food of animal origin in relatively recent years. A polymerase chain reaction (PCR) based method, the multiplex PCR, was recently applied to species identification in meat and meat products. It allows co-amplification of separate regions of a single gene or specific fragments, each typical of a different animal species in a single PCR reaction, using different pairs of primers in the same reaction mix. In the present paper, the duplex-PCR technique is proposed to identify bovine and water buffalo DNA in a single PCR assay in milk and mozzarella cheese (a typical Italian cheese, originally made from pure water buffalo milk). Because of its lower cost, undeclared bovine milk is added to water buffalo milk for making different kinds of mozzarella cheese. The results of this experiment indicate the applicability of this method, which showed an absolute specificity for the two species and a high sensitivity even down to low DNA concentrations (1 pg). In bovine and water buffalo mixtures of both milk and mozzarella cheese, the minimum concentration tested was 1% of bovine in water buffalo milk and water buffalo in bovine milk. The importance of the somatic cell content in raw milk is also discussed with special reference to the evaluation of mixtures (milk or cheese) of the two species.

Milk ejection and milk removal is considerably influenced by the sympathetic nervous system. Stimulation of α-adrenergic receptors by administration of α-adrenergic agonists inhibits alveolar milk ejection and milk removal in dairy cows due to smooth muscle contraction (Blum et al. 1989; Bruckmaier et al. 1991). However, contraction of the teat in response to α-adrenergic receptor stimulation has no influence on milk flow as long as milk is available in the cistern (Bruckmaier et al. 1997). Therefore, α-adrenergic stimulation causes inhibition of transport of alveolar milk into the cistern. On the contrary, the stimulation of β-adrenergic receptors facilitates milk ejection and milk removal in dairy cows (Bernabé & Peeters, 1980; Bruckmaier et al. 1991) because of muscle relaxation. Therefore, the distribution of α- and β-adrenergic receptors plays an important role in the milkability of dairy cows. However, from these in vivo studies it is not possible to distinguish between the different α1- and α2- and β2-receptor subtypes owing to the non-specific nature of the pharmacological agents used.

To date, the precise tissue distribution of these different subtypes, in bovine mammary tissue, is unknown. Using molecular techniques, we were interested in the expression of genes that encode α1c and β2 as a preliminary study towards the understanding of noradrenergic receptor-gene expression and regulation in this important system.

In addition, α1c- and β2-adrenergic receptors were determined in front and rear quarters of the mammary gland to investigate differences in receptor distribution within the udder and possible relations between adrenergic receptor distribution and the higher milk flow rates in rear than in front quarters (Rothenanger et al. 1995).

Numerous methods based on DNA analysis have been employed in the food industry to monitor adulterations of food products of animal origin. Among them the most frequently used are: polymerase chain reaction (PCR) amplification of a marker gene fragment(s) with universal primers, or amplification of DNA with species-specific primers. PCR-products of different origin can be discriminated by size, restriction fragment length polymorphism (RFLP) or single stranded conformational polymorphism (SSCP) analysis. These methods have been used for identification, and differentiation between, the animal origins of raw or heat-treated meat and meat products (Chikuni et al. 1994; Meyer et al. 1994, 1995; Zehner et al. 1998; Behrens et al. 1999; Guoli et al. 1999; Hopwood et al. 1999; Matsunaga et al. 1999; Wolf et al. 1999). These approaches are also applicable to the analysis of dairy products. However, adulterations of goats' milk and its products are traditionally tested by immunological and/or electrophoretic methods (Amigo et al. 1992; Levieux & Venien, 1994; Mimmo & Pagani, 1998). So far, only a few DNA-based techniques designed to detect the presence of bovine DNA in goats' milk have been described (Plath et al. 1997; Branciari et al. 2000). This paper presents a one-step PCR procedure for detection of adulteration of goats' milk with cows' milk. The method, employing bovine-specific primers for amplification of a 274 bp fragment of cytochrome b DNA, seems to be simple, fast, specific and sensitive.