Sarda ewes, ∼4·5 million animals producing 500 000 tonnes milk annually, are one of the most important breeds of dairy sheep in the Mediterranean area. Several studies (Casu & Labussière, 1972; Labussière et al. 1981; Gallego et al. 1983; Rebello de Andrade et al. 1989; Bencini, 1993) have shown that milk production is influenced by mammary gland size and cistern dimension. The size of the mammary cistern affects both milk secretion rate and milk emission kinetics during milking.

Milk secretion rate is controlled at the mammary gland level mainly by a protein feedback inhibitor of lactation (FIL), which is produced by mammary epithelial cells and secreted together with milk into the alveoli (Wilde & Peaker, 1990). As the alveoli are the site of action of the FIL (Henderson & Peaker, 1984), the FIL affects the rate of secretion when the milk is stored in the secretory tissue, whereas it is inactive in the milk stored in the cistern. As a consequence, the action of the FIL should be less in animals with a greater cistern volume, because a large proportion of milk is stored in the mammary cistern and the time during which the milk is in contact with the alveoli is reduced. This hypothesis is supported by the finding that the milk production of cows (Dewhurst & Knight, 1992; Knight & Dewhurst, 1992, 1994) and sheep (Karam et al. 1971; Enne et al. 1972) with large cistern storage capacities was almost unaffected by changes in the frequency of milking.

Cistern volume also affects milk emission kinetics and the proportion of stripped milk obtained at milking (Labussière, 1988). Cisternal milk is immediately available for removal, whereas alveolar milk is available only after operation of the ejection reflex, necessary in dairy ewes for complete udder emptying (Bruckmaier et al. 1997). Therefore, in animals readily able to expel alveolar milk into the cistern before the whole cisternal milk fraction is removed (Pazzona et al. 1978; Bruckmaier et al. 1997), a larger cistern volume enables milking time to be reduced by eliminating or restricting the need for stripping.

On the basis of the above considerations, the volume of the mammary gland cistern could be proposed as a selection objective to improve milk production and milking ability of dairy ewes. For this purpose, a quick, accurate and economic method for measuring it is needed. The ultrasound technique allows the internal structure of the mammary gland cistern to be observed clearly in sheep (Ruberte et al. 1994; Pulina et al. 1996; Bruckmaier et al. 1997), cows (Bruckmaier et al. 1994b) and goats (Bruckmaier et al. 1994a). Cistern size has been measured by ultrasound in dairy cows (Bruckmaier et al. 1994b) and sheep (Pulina & Nudda, 1996), where a positive correlation between milk yield and cistern area calculated from the ultrasound images of mammary glands was found. However, area estimation requires the use of expensive ultrasound equipment or of a digitizing tablet. In both cases, area measurement is difficult owing to the irregular shape of the cistern.

The aim of this study was to test the use of linear measurements taken directly from ultrasound images to estimate cistern size in dairy ewes.

The autolysin, N-acetyl muramidase (AcmA), of six commercial Lactococcus lactis subsp. cremoris starter strains and eight Lc. lactis subsp. cremoris derivatives or plasmid-free strains was shown by renaturing SDS-PAGE (zymogram analysis) to be degraded by the cell envelope proteinase (lactocepin; EC 3.4.21.96) after growth of strains in milk at 30 °C for 72 h. Degradation of AcmA was less in starter strains and derivatives producing lactocepin I/III (intermediate specificity) than in strains producing lactocepin I. This supports previous observations on AcmA degradation in derivatives of the laboratory strain Lc. lactis subsp. cremoris MG1363 (Buist et al. Journal of Bacteriology180 5947–5953 1998). In contrast to the MG1363 derivatives, however, the extent of autolysis in milk of the commercial Lc. lactis subsp. cremoris starter strains in this study did not always correlate with lactocepin specificity and AcmA degradation. The distribution of autolysins within the cell envelope of Lc. lactis subsp. cremoris starter strains and derivatives harvested during growth in milk was compared by zymogram analysis. AcmA was found associated with cell membranes as well as cell walls and some cleavage of AcmA occurred independently of lactocepin activity. An AcmA product intermediate in size between precursor (46 kDa) and mature (41 kDa) forms of AcmA was clearly visible on zymograms, even in the absence of lactocepin I activity. These results show that autolysis of commercial Lc. lactis subsp. cremoris starter strains is not primarily determined by AcmA activity in relation to lactocepin specificity and that proteolytic cleavage of AcmA in vivo is not fully defined.

Mastitis can be caused by a number of bacteria (Philpot & Pankey, 1975), among which Staphylococcus aureus is one of the most important, responsible for considerable economic loss to the dairy industry (Jasper et al. 1982). Carrier rates have been studied extensively in dairy herds suffering from mastitis (Davidson, 1961, 1963).

Subtyping bacteria is an important epidemiological tool; for example, antimicrobial susceptibility patterns (antibiograms) have been used for typing Staph. aureus in human medicine (Gillespie et al. 1990). Staph. aureus of bovine origin can be divided into several categories by biotyping (Devriese, 1984). Phage typing has proved useful in differentiating mastitis strains (Mackie et al. 1987) and plasmid profiling has been valuable in epidemiological studies of bovine Staph. aureus (Baumgartner et al. 1984).

During the last 5 years genomic fingerprinting of Staph. aureus became a powerful tool for epidemiological typing. Numerous techniques for comparison of staphylococcal isolates have been developed and are becoming important in investigations of strain origin, clonal relatedness and epidemiology.

The aim of this study was to determine the epidemiology of Staph. aureus isolates from mammary glands of cows from a single herd using modern molecular typing techniques, and to assess whether the same strain constantly colonizes the udder or cows are repeatedly infected by different strains. We investigated 26 Staph. aureus isolates obtained from quarter milk samples from 16 cows. These isolates were characterized biochemically, by their antibiotic resistance pattern, by two polymerase chain reaction (PCR) methods on the basis of coagulase (Coa) gene and protein A (Spa) gene (X region) polymorphism and by macrorestriction analysis of chromosomal DNA using pulsed-field gel electrophoresis (PFGE).

The respiratory burst activity of polymorphonuclear leucocytes (PMN) was evaluated in eight Holstein cows from 8 weeks before until 6 weeks after calving by chemiluminescence (CL). The CL response started to decrease 1 week before parturition, reaching a minimum during the first 2 weeks after calving. From week 3 of lactation, CL increased again and returned to original levels by week 6 of lactation. Plasma concentrations of 3-hydroxybutyric acid, total bilirubin and bovine pregnancy-associated glycoprotein started to increase before parturition to reach a maximum during the first or second week of lactation. The concentrations of glutamic–oxaloacetic transaminase, lactate dehydrogenase, non-esterified fatty acids and bilirubin increased after calving, reaching a maximum during the second week. A small decrease in plasma cholesterol during the week before and after calving was followed by an increase. The CL response of the PMN showed significant temporal relationships with the plasma concentrations of 3-hydroxybutyric acid, bovine pregnancy-associated glycoprotein, bilirubin, glutamic–oxaloacetic transaminase, non-esterified fatty acids; that with cholesterol was nearly significant. This means that the change in the CL response with time coincided with the changes in plasma concentrations of these substances with time and that these changes were significantly related with each other. The results of this study show that the decreased respiratory burst activity of bovine PMN around parturition may be related to the extent of the metabolic and hormonal changes. Although the causative relationships are not proven, these results support earlier results suggesting that 3-hydroxybutyric acid and bovine pregnancy-associated glycoprotein may directly affect neutrophil function, whereas non-esterified fatty acids, cholesterol, bilirubin, and liver enzymes may have potential as diagnostic markers of impaired neutrophil function and consequently increased disease risk around parturition.

Intracellular peptidases of Lactobacillus helveticus may play a major role in the proteolysis of Swiss cheeses, provided that they are released through bacterial lysis. Experimental Swiss cheeses were manufactured on a small scale from thermized and microfiltered milk using as starters (in addition to Streptococcus thermophilus and Propionibacterium freudenreichii) one of two Lb. helveticus strains, ITGLH1 and ITGLH77, which undergo lysis to different extents in vitro. All the cheeses were biochemically identical after pressing. The viability of Lb. helveticus ITGLH1 and ITGLH77 decreased to a similar extent (96–98%) while in the cold room, but the concomitant release of intracellular lactate dehydrogenase in cheeses made with strain ITGLH1 was 5–7-fold that in cheeses made with ITGLH77. Protein profiles and immunoblot detection of the dipeptidase PepD confirmed a greater degree of lysis of the ITGLH1 strain. Free active peptidases were detected in aqueous extracts of cheese for both strains, and proteolysis occurred principally in the warm room. Reversed-phase HPLC revealed a more extensive peptide hydrolysis for ITGLH1, which was confirmed by the greater release of free NH2 groups (+33%) and free amino acids (+75%) compared with ITGLH77. As the intracellular peptidase activities of ITGLH1 and ITGLH77 have previously been shown to be similar, our results indicated that the extent of lysis of Lb. helveticus could have a direct impact on the degree of proteolysis in Swiss cheeses.

Several studies have reported associations between κ-casein (κ-CN) genetic variants A and B and milk composition and properties (for review, see Ng-Kwai- Hang, 1998). For instance, genetic variant κ-CN B was associated with higher protein, casein and κ-CN contents. The objective of the present study was to demonstrate a differential expression of κ-CN genetic variants A and B at the mRNA level in mammary epithelial cells during lactation.

The sterol fraction of milk is of nutritional interest because high levels of cholesterol in plasma (modulated by the cholesterol ingested) are associated with an increasing risk of cardiovascular disease. In addition, some sterols (ergosterol and 7-dehydrocholesterol) are provitamins (D2 and D3 respectively). At the same time, through the study of the sterol fraction, vegetable fats can be detected in milk and dairy products. Sterols are a minor fraction of total milk fat, the main sterol being cholesterol (3 mg/g fat, equivalent to 100 mg/l cows' milk). Small quantities of other sterols (7-dehydrocholesterol, 22-dehydrocholesterol, ergosterol, fucosterol, lanosterol, lathosterol, 24-methylenecholesterol) and several phytosterols have been reported in cows' milk (Walstra & Jennes, 1984). International Dairy Federation (1992) states that in the sterol profile of genuine milk fat there may appear, in addition to the peak of 7-dehydrocholesterol which ranges from 0·7 to 4% of total sterols, < 1% of minor sterols with retention times corresponding to phytosterols.

Values for the cholesterol content of goats' milk vary considerably, from 211 mg/l (Pantulu et al. 1975) to 125 mg/l (Lu, 1993), partly owing to the use of different analysis techniques. Some of these values were obtained using non-specific colorimetric methods, which are inaccurate in the presence of cholesterol precursors or phytosterols (Clark et al. 1983; Haugh & Harzer, 1984). Some minor peaks have been assumed to be sterols but have not been identified (García-Olmedo & Barrera, 1985).

Conventional methods of sample preparation for sterol analysis prior to gas chromatography (GC), which involve saponification of fat with or without isolation of the sterol fraction by thin layer chromatography, are tedious and time-consuming. Transesterification with KOH–methanol has been successfully used as a rapid alternative for obtaining the unsaponifiable fraction.

This paper describes the identification of sterols (cholesterol and other minor sterols) in goats' milk fat using an alkali-catalysed transesterification procedure prior to GC and GC–mass spectrometry (GC–MS) analysis.

We have investigated the influence of partial hydrolysis with an immobilized proteinase from Bacillus licheniformis on the thermal gelation of isolated β-lactoglobulin B. Gelation behaviour was determined by dynamic rheological measurements (small deformation) and the gels were characterized with respect to microstructure and water-holding properties. A fine-stranded gel with a complex modulus of ∼ 2000 Pa was formed from β-lactoglobulin (50 g/l in 75 mM-Tris-HCl, pH 7·5). Limited hydrolysis prior to thermal gelation resulted in coarser gels with thicker protein strands and larger pores. Gel structure correlated with its permeability, proton mobility and water-holding capacity. Total gel stiffness increased with low degrees of hydrolysis, but decreased after prolonged hydrolysis. Maximal gel stiffness was 1·5-fold that of gels made from unhydrolysed β-lactoglobulin. This was much lower than the stiffening effect obtained after partial hydrolysis of whey protein isolate, showing that the gel strengthening effect of partial hydrolysis was dependent on the protein composition and/or the hydrolysis and gelation conditions. A mechanism to explain the observed effects of hydrolysis on gelation and gel properties is presented.

Among specific genes that may affect economically important traits in sheep, the β-lactoglobulin (LGB) locus has been extensively studied. Polymorphism has been detected in several breeds, but studies of the effect of LGB alleles on milk production traits have given conflicting results. Some found that LGB polymorphism significantly affects milk yield (Bolla et al. 1989; Herget et al. 1995; Fraghì et al. 1996), fat and protein content (Garzon & Martínez 1992; Giaccone et al. 1997; Kukovics et al. 1998), only fat content (Pirisi et al. 1998) and cheese yield and composition (Di Stasio et al. 1997; Rampilli et al. 1997). However, other studies failed to detect any effect of the gene on milk production traits (Barillet et al. 1993; Recio et al. 1997). These inconsistencies, similar to those reported for dairy cattle, can be explained by breed differences, population size, frequency distribution of the genetic variants and a failure to consider relationships among animals (Sabour et al. 1996).

Moreover, both the production data considered and the methods used for statistical analysis could be further causes of conflicting results (Ng-Kwai-Hang, 1997). Investigations of the relationships between milk protein polymorphism and milk production usually consider accumulated yields for standardized lactation lengths, assuming that environmental effects average out over a lactation. Such an assumption is not always valid, because there can be marked effects peculiar to individual test day (TD) measures that may not average out (Jamrozik & Schaeffer, 1997). The direct modelling of TD measures offers the advantage of a more accurate removal of environmental variation from phenotypic observations (Stanton et al. 1992). However, particular attention to the temporal dependence of the covariance structure among TD is required. In TD analysis performed by mixed linear models a simple covariance structure, known as compound symmetry, is usually assumed. This structure assumes an equal variance for all TD and an equal correlation between all pairs of TD within each lactation. An initial drawback of this assumption arises because of the heterogeneity of variance throughout lactation. Moreover, since TD values within a lactation are a sequence of repeated measures taken on the same experimental unit (Van der Werf & Schaeffer, 1997), measures close in time are likely to be more highly correlated than measures far apart in time. All these potential patterns of correlation and variation may combine to produce a complicated structure of covariance among TD that, when ignored, may result in inadequate analysis or incorrect conclusions (Littel et al. 1998). In particular, there can be marked differences in the estimates of the fixed factors considered in the analysis; such a bias is enhanced when the data structure is highly unbalanced, as in the case of studies on relationships between milk protein polymorphisms and milk production traits.

A possible solution can be found in the property of mixed linear models to assume different (co)variance structures in order to find the one that best fits experimental data. The aim of the present study was to test the possible influence of the statistical model used on the results when the relationships between β-lactoglobulin polymorphism and milk production traits in dairy ewes were analysed. With this aim in view, TD measures were directly modelled with mixed linear models and the effects of alternative (co)variance structures on fixed factors estimates were compared.

The last few years have seen a number of advances in the understanding of genetic polymorphisms of caprine caseins, and especially in molecular characterization of the allelic variants and analysis of their frequencies in Spanish and other breeds (Grosclaude et al. 1994; Jordana et al. 1996). Although more is being discovered about the influence of these polymorphisms on the yield and characteristics of cheeses (Remeuf, 1993; Pirisi et al. 1994), little work has been done on the quantitation of different casein fractions of goats' milk throughout lactation. Measurements have been made using SDS-PAGE (Quiles et al. 1990) and cation- exchange FPLC (Brown et al. 1995), and Recio et al. (1997a) have demonstrated the potential of capillary electrophoresis (CE) for the analysis and quantitation of milk proteins. Use of CE has resulted in the development of rapid automated analysis with very high resolution, requiring very small sample and buffer volumes and with a significantly reduced amount of solvent waste.

The aim of the present study was to determine the influence of herd and milking period on the contents of the various nitrogen and casein fractions, the latter being determined by CE, in milk from goats of the Murciana-Granadina breed.

The aim of this work was to study the spin–spin (T2) relaxation components of one hard cheese and three soft cheeses to characterize fat and water states. NMR signals were measured at 6 °C with a 0·47 T NMR device. The transverse relaxation decay was fitted using the Marquardt method. The T2 relaxometric behaviour of the cheeses under consideration was characterized by four relaxation components. To understand the chemical composition of each NMR component, we studied anhydrous milk fat extracted from each cheese analysed. At 6 °C, the fat was 60% crystalline. In cheese, the solid fat was found mainly in the shorter relaxation component with a T2 of 17 μs. The intensity of the NMR relaxation with a T2 > 1 ms was explained by the amount of water, liquid fat and proteins, and the associated relaxation time varied as a function of the process used. The composition of each relaxation component was confirmed by the temperature effect and the influence of the fat content on the NMR cheese signal. NMR relaxometry was able to provide information on water behaviour (i.e. the quantity and level of interactions with proteins) and on the solid[ratio ]liquid ratio of anhydrous milk fat in the cheese.

Volatile concentrates were obtained by vacuum distillation from both natural and creamy Gorgonzola cheese and isolated by continuous liquid–liquid extraction. Both were analysed by high resolution gas chromatography (HRGC), HRGC–mass spectrometry and HRGC–olfactometry. A total of 63 components were identified in the neutral extract of the natural type (21 esters, 13 ketones, 14 alcohols, 5 aldehydes, 1 sulphur compound, 7 aromatic compounds and 2 terpenes) and 52 in the creamy type (17 esters, 12 ketones, 10 alcohols, 5 aldehydes, 1 sulphur compound, 5 aromatic compounds and 2 terpenes). Ketones, whose major components were 2-nonanone and 2-heptanone, were the predominant constituents of the neutral fraction. By olfactometric analysis of the neutral extracts, 23 odour-impact compounds were found in the natural and 21 in the creamy Gorgonzola cheese. 1-Octen-3-ol, ethyl hexanoate, 2-nonanone, 2-heptanone, 2-heptanol, ethyl butanoate, 2-nonanol and 4-methylanisole were the key odorants of the natural cheese, whereas 2-heptanone, 2-heptanol, ethyl butanoate, 3-methyl thiopropanal and an unidentified constituent with a fruity odour were characteristic of the creamy Gorgonzola cheese. On the basis of high odour unity values, 2-nonanone, 1-octen-3-ol, 2-heptanol, ethyl hexanoate, methylanisole and 2-heptanone were the most important odorants of natural and creamy Gorgonzola cheese aroma.

In the Iberian Peninsula, the proteinases present in the flowers of members of the Cynara genus, C. cardunculus, C. humilis and C. scolymus, have for many years been successfully used in the manufacture of traditional cheeses from ovine and/or caprine milk on individual farms (Vieira de Sá & Barbosa, 1972; Trujillo et al. 1994). In Portugal, C. cardunculus is the species most frequently employed. Although commercial thistle was tentatively assumed to be pure in taxonomic terms, accurate analyses have shown that the flowers of C. cardunculus are often mixed with flowers of C. humilis (Pires et al. 1994). The clotting activity of C. humilis is due to an aspartic proteinase, currently designated cardosin A and similar to another enzyme obtained from C. cardunculus. This enzyme is similar in specificity and activity to chymosin (Pires et al. 1994).

The action of cardosin A from C. cardunculus upon ovine and caprine caseins has been reported recently (Ramalho-Santos et al. 1996; Simo4es, 1998; Sousa & Malcata, 1998), but as yet there is no information on the proteolytic activity of the proteinase from C. humilis upon caseins from milks other than bovine. Caseins from small ruminants' milks are the usual substrates of cardosin during milk coagulation and cheese ripening, and sodium caseinate represents an intermediate system between isolated caseins and the cheese matrix that is free from interference by fat. Thus ovine and caprine caseinates may be useful substrates for investigating the proteolytic activity of cardosin.

The aim of the present study was to compare the action of pure cardosin A, obtained from C. humilis, on caprine and ovine caseinates, and to assess the in vitro contribution of this enzyme to the overall proteolytic action of thistle rennet.

Two common genetic variants of β-lactoglobulin (β-lg), A and B, exist as co- dominant alleles in dairy cattle (Aschaffenburg, 1968). Numerous studies have shown that cows homozygous for β-lg A have more β-lg and less α-lactalbumin (α-la) and casein in their milk than cows expressing only the B variant of β-lg (Ng-Kwai-Hang et al. 1987; Graml et al. 1989; Hill, 1993; Hill et al. 1995, 1997). These differences have a significant impact on the processing characteristics of the milk. For instance, the moisture-adjusted yield of Cheddar cheese is up to 10% higher using milk from cows of the β-lg BB phenotype compared with milk from cows expressing only the A variant (Hill et al. 1997). All these studies, however, describe compositional differences associated with β-lg phenotype in established lactation only. No information is available on the first few weeks of lactation, when there are marked changes in the concentrations of β-lg and α-la (Pérez et al. 1990).

The need to improve the shelf life and quality of foods, combined with the urgency to reduce waste generated by synthetic non-biodegradable packaging, have resulted in an increasing interest in biodegradable or edible materials (Chen, 1995). The challenge in the development of such materials lies in achieving controlled lifetime together with compatibility with foods and the environment. Natural polymers or polymers derived from natural monomers such as food proteins offer the greatest opportunities, since their biodegradability and environmental compatibility are assured (Krochta & de Mulder-Johnston, 1997).

Milk proteins such as caseins have been formulated into coatings and films for improving food quality (Baker et al. 1994). Unfortunately, the highly hydrophilic nature of these proteins limits the possibility of producing films with the required properties. Gamma irradiation, which induces the formation of crosslinks in the protein structure, is known to improve the mechanical properties and water resistance of such materials (Brault et al. 1997; Mezgheni et al. 1998a; Ressouany et al. 1998), and can also sterilize materials for possible biomedical applications (Kaetsu, 1995).

Gamma irradiation is slowly becoming accepted in the food industry as a means of safely improving the shelf life of various fruits and vegetables and eliminating bacterial contamination in meats (Pszczola, 1997). All these features (sterility, increased mechanical strength, controlled life time) make gamma irradiation a promising technique for the production of commercial biodegradable films.

In the present study, we have investigated the microbial resistance of calcium caseinate films irradiated at 4 or 64 kGy. A minimum dose of 4 kGy was required to produce complete sterility in the films without inducing many crosslinks in the protein. A dose of 64 kGy was used to induce maximum crosslinking density (Ressouany et al. 1998). The rate of modification of the films was evaluated using three methods. Firstly, microbiological counts were performed on films incubated with a strain of Pseudomonas aeruginosa. This strain was chosen because it produces a wide range of proteinases and is often implicated in the process of food spoilage. These results were compared with those obtained for soluble N. Since conversion to CO2 is one of the standard test procedures to assess biodegradability (El-Din Sharabi & Bartha, 1993), we also analysed the CO2 produced by the biodegradation of the crosslinked films using a modified version of a standard test procedure (American Society for Testing Materials, 1992). This measures the aerobic degradation of plastic materials inoculated with municipal sewage sludge. We have replaced the sludge by a concentrated bacterial strain (Ps. aeruginosa) and compared these results with the two previous methods. This last method was used to measure the resistance of the crosslinked films to bacterial degradation.

The timely ingestion and absorption of colostral immunoglobulin is a critical determinant of neonatal calf health. Calves are born without appreciable concentrations of the serum immunoglobulins needed to protect against pathogenic bacteria, viruses and protozoa (Tyler & Parish, 1995). The beneficial effect of passive transfer of colostral immunoglobulin also extends beyond the neonatal period and persists into juvenile and adult life (Robison et al. 1988; Tyler et al. 1998; DeNise et al. 1989). Calves with failure of passive transfer, defined as serum protein < 50 g/l or serum IgG < 10 g/l, have increased mortality risks that persist until 10 weeks of age (Tyler et al. 1998).

Several diseases are potentially spread by the ingestion of colostrum, including bovine leukosis and Johne's disease (Perrin & Polack, 1988; Streeter et al. 1995). In one study 22% of latently infected cows were demonstrated to shed Mycobacterium paratuberculosis in their colostrum (Streeter et al. 1995). Optimal programmes to prevent and eradicate these diseases generally include the provision that calves are given colostrum derived from cows of known negative disease status. Pasteurization or heat treatment of colostrum may provide a mechanism whereby calves are provided with protection against neonatal disease without creating undue potential for infection by chronic, economically relevant diseases.

Attempts to heat disinfect colostrum are common in goat herds (MacKenzie et al. 1987). Pasteurization has been demonstrated to be effective against the caprine arthritis–encephalomyelitis virus (Adams et al. 1983; MacKenzie et al. 1987). Although Myco. paratuberculosis appears to resist pasteurization, this form of processing has been demonstrated to decrease the likelihood of positive colostral cultures for Myco. paratuberculosis under experimental conditions (Meylan et al. 1996). Sterilizing the milk and colostrum given to calves is a logical and reasonable strategy to prevent transmission of infectious microorganisms. The potential disadvantage of heat treating colostrum is that the immunoglobulins in colostrum may become denatured (Smith & Sherman, 1994). Pasteurization causes only a slight decrease in the colostral concentration of IgG in cattle (Meylan et al. 1996); however, the biological behaviour of these pasteurized immunoglobulins has not been critically examined. Therefore, we cannot be completely confident that immunoglobulin absorption, persistence in serum and biological activity are unchanged by this processing.

The goal of this study was to determine the effect of pasteurization at 76 and 63 °C on the absorption of IgG from colostrum. Should these procedures decrease immunoglobulin absorption, the use of pasteurization in disease eradication programmes would require increased efforts to optimize the passive transfer of immunoglobulin.

Bacillus cereus is a common contaminant in raw milk. The spores survive pasteurization and psychrotrophic strains of B. cereus often limit the keeping quality of pasteurized milk stored at > 6 °C (Griffiths, 1992). High numbers of B. cereus in pasteurized milk are most frequent when the cows are grazing (Slaghuis et al. 1997), mainly owing to increased levels of spores in raw milk resulting from teat contamination by soil (Christiansson et al. 1999). However, high numbers can also be found in pasteurized milk while the cows are housed indoors, and this is probably caused by additional contamination at the dairy plant (te Giffel et al. 1996; Larsen & Jørgensen, 1997; Lin et al. 1998). There is little information available about the sites of recontamination in the dairy. The use of typing techniques capable of discrimination below the species level, such as fatty acid profiles and random amplification of polymorphic DNA–polymerase chain reaction (RAPD–PCR), could be helpful in demonstrating contamination routes (Lin et al. 1998; Nilsson et al. 1998).

Spores of B. cereus are very hydrophobic and readily adhere to surfaces of steel, glass and rubber (Rönner et al. 1990), and short cleaning-in-place programmes do not always eliminate all the spores (Rönner & Husmark, 1992). Spores adhering to surfaces are more difficult to eliminate by disinfectants than spores in solution (te Giffel et al. 1995). Many B. cereus spores germinate rapidly in milk upon heat activation and, if allowed to propagate undisturbed on surfaces, may form biofilms that are extremely difficult to eliminate (Mosteller & Bishop, 1993; Wirtanen et al. 1996; Kumar & Anand, 1998).

This paper describes how we demonstrated the involvement of a pasteurizer in the contamination of pasteurized milk by B. cereus in a commercial dairy plant using a combination of classic microbiological analyses and typing of strains by RAPD–PCR.

Psychrotrophic bacteria, capable of growing in milk during refrigerated storage before processing, produce thermostable proteinases that strongly influence the keeping quality of long life dairy products (Cousin, 1982). This is the case with UHT milk, where the activity of residual or reactivated bacterial proteolytic enzymes is linked to flavour deterioration and gelation. Furthermore, since most of these enzymes are acid proteinases with a broad specificity and particularly active against κ-casein (Fairbairn & Law, 1986), their hydrolysis products can interfere with detecting the fraudulent addition of rennet whey (López-Fandiño et al. 1993a, b; Recio et al. 1996, 2000). This is usually based upon detecting, by reversed-phase (RP-) HPLC (Olieman & van Riel, 1989; Van Riel & Olieman, 1995a) or capillary electrophoresis (CE) (Van Riel & Olieman, 1995b), caseinmacropeptide (CMP), the 106–169 fragment of κ-casein released by chymosin during the initial stages of cheesemaking (for review, see López-Fandiño & Olano, 1999). Nevertheless, it is not yet clear whether degradation of κ-casein by psychrotrophic enzymes gives rise to hydrolysis products that are identical to those produced by chymosin or just indistinguishable by CE (Van Riel & Olieman, 1995b; Recio et al. 1996).

In the present investigation, we studied CMP-like degradation products produced by the action of extracellular psychrotrophic proteinases on κ-casein by RP-HPLC and CE. Electrospray ionization–mass spectrometry (ESI–MS) was used for peptide identification. Characterization of breakdown products specific to psychrotrophic enzymes may help the investigation of their residual activity remaining in milk after processing, as well as provide an indication of the suitability of CMP as an indicator of adulteration of UHT milk with whey solids.

Caseins bind strongly to Fe by their phosphoseryl residues (Hegenauer et al. 1979; Brulé & Fauquant, 1982; Bouhallab et al. 1991; Emery, 1992) and keep it soluble at the alkaline pH of the duodenum (Manson & Annan, 1971; Bouhallab et al. 1991). It has been suggested that this strong binding prevents the release of free Fe during digestion and impairs its absorption (West, 1986; Hurrell, 1997), but in human studies hydrolysis by digestive enzymes to give low molecular mass peptides improves Fe absorption (Hurrell et al. 1988, 1989).

β-Casein peptide 1–25 (β-CN(1–25)) is the phosphorylated N-terminal fragment of β-CN. Its molecular mass is 3124 Da and it contains four phosphoseryl residues that bind four Fe atoms and keep them soluble (Bouhallab et al. 1991).

Preliminary results showed that binding Fe to β-CN(1–25) enhances its bioavailability in the rat (Aît-Oukhatar et al. 1997) and its absorption by the duodenal rat loop model (Pérès et al. 1997). We recently showed that the β- CN(1–25)–Fe complex is hydrolysed to a lesser extent than free β-CN(1–25) during duodenal digestion, and that the phosphorylated region of the peptide to which Fe is bound, β-CN(15–24), cannot be detected in the digestive lumen (our unpublished results).

These previous studies did not examine the effect of luminal digestion that could release free inorganic Fe. We considered the possibility that Fe bound to β-CN(1–25) is not released in the digestive lumen, and that it reaches the absorbing surface of the small intestine brush border membrane as a complex. To test this hypothesis we examined in vitro the influence of pH and digestive enzymes on the dialysability of the β-CN(1–25)–Fe complex.

Caseins constitute the main part of milk proteins. The genes for the four major types of casein (CN), αs1-CN, αs2-CN, β-CN and κ-CN, reside on < 220 kb of the DNA on bovine chromosome 6 (Threadgill & Womack, 1990). Two casein genes (Cn), β-Cn and κ-Cn, have been found to be associated with differences in milk production, cheese yield, protein content, quality and fat yield (Ng-Kwai-Hang et al. 1984). Lin et al. (1986) have estimated that the combined contributions of αs1-Cn, β-Cn, κ-Cn and β-lactoglobulin loci accounted for 8·9% of the total phenotypic variance in milk, 8·6% of that in proteins and 5% of that in fat yield. Six alleles of κ-Cn gene have so far been identified: κ-CnA, κ-CnB, κ-CnC, κ-CnE, κ-CnF and κ-CnG. The κ-CN protein is 169 amino acids long with varying regions at codons 136 and 148. The A variant has threonine (ACC) at codon 136 and aspartic acid (GAT) at codon 148, whereas the B variant has isoleucine (ATC) and alanine (GCT) at codons 136 and 148 respectively (Mercier et al. 1973). Milk containing κ-CN B was reported to yield more cheese than that with κ-CN A because it contains more casein and has lower fat loss. Several studies have shown that cows with the κ-CnBB genotype produced milk with a higher protein content than cows with κ-CnAA (Ng-Kwai-Hang et al. 1984; Ron et al. 1994).