No abstract

Analytical results are given for whey powders prepared on a commercial or semi-commercial scale by three companies. Altogether, five preparations enriched in ␤-lactoglobulin, four whey protein isolates and a fraction enriched in ␣-lactalbumin were analyzed for protein composition, including % ␤-lactoglobulin, ␣-lactalbumin, bovine serum albumin, casein (glyco) macropeptide and the main triglycerides. Protein composition was determined by high pressure gel permeation and reversed phase liquid chromatography and by capillary zone electrophoresis. The extent of modification of the native ␤-lactoglobulin structure was also measured through the degree of lactosylation and the fraction of accessible free sulphydryl groups. One significant finding was that the calculated recovery of protein following quantitation of the chromatogram or electropherogram was seldom above 90% and occasionally below 60% of that loaded onto the column or capillary, raising doubts as to the reliability of the analytical results. Extrapolation by linear regression to 100% recovery allowed estimates to be made of the true ␤-lactoglobulin composition of the samples. The nine samples could be placed into three distinct groups with estimat-ed true ␤-lactoglobulin weight % of 70.9 Ϯ 1.1, 62.0 Ϯ 3.4 and 39.5 Ϯ 4.9. Physico-chemical properties of the group of samples are reported elsewhere (Holt et al., 1999).The inter-laboratory comparisons involved the Royal Veterinary and Agricultural University (KVL), two laboratories of BDI (Lab1 and Lab2), NIZO food research (NIZO) and the Composition of whey protein isolates and fractions C. Holt et al. Figure 2 RP-HPLC analysis of commercial whey protein samples using the LRTL method. A. MDFwpi-1, B. MDFwpi-1a, C. MDFwpi-1b.Figure 3 Capillary zone electropherograms of selected samples as obtained by the KVL method. Within the b-Lg region, lactosylated forms have longer migration times than the native forms. on the Molecular Basis of the Aggregation, Denaturation, Gelation and Surface Activity of Whey Proteins (MADGELAS), CT96-1202. All other members of the MADGELAS group are thanked for their co-operation in the sample survey, of which this paper forms a part. The HRI work was supported by the Scottish Office Composition of whey protein isolates and fractions C. Holt et al.

Background: The nisin-controlled gene expression system NICE of Lactococcus lactis is one of the most widely used expression systems in Gram-positive bacteria. Despite its widespread use, no optimization of the culture conditions and nisin induction has been carried out to obtain maximum yields. As a model system induced production of lysostaphin, an antibacterial protein (mainly against Staphylococcus aureus) produced by S. simulans biovar. Staphylolyticus, was used. Three main areas need optimization for maximum yields: cell density, nisin-controlled induction and protein production, and parameters specific for the target-protein.

The low prices of some nonmilk proteins make them attractive as potential adulterants in dairy products. An optical biosensor (BIACORE 3000) was used to develop a direct and combined biosensor immunoassay (BIA) for the simultaneous detection of soy, pea, and soluble wheat proteins in milk powders. Affinity-purified polyclonal antibodies raised against the three protein sources were immobilized in different flow channels (Fcs) on the biosensor chip (CM5). Dissolved milk powders were injected (20 microL injections at 20 microL min(-1)) through the serially connected Fcs, and the antibody-bound plant proteins were detected directly. The total run time between samples, including a regeneration step with 5 microL of 10 mM HCl, was 5 min. The limits of detection in milk powder were below 0.1% of plant protein in the total milk protein content. The antibodies also recognized some proteins from other plant sources, which made this BIA even more suitable as a broad screening assay for nonmilk proteins.

The physico-chemical properties are reported for a group of whey protein powders prepared on a commercial or semi-commercial scale by three companies and chemically characterized as described elsewhere . The dependence of the apparent ␤-lactoglobulin % on the recovered % showed that the nine samples could be placed in three distinct groups with ␤-lactoglobulin weight % of 70.9 Ϯ 1.1 (Group 1), 62.0 Ϯ 3.4 (Group 2) and 39.5 Ϯ 4.9 (Group 3). Measurements by 1 H-NMR spectroscopy, on 3 of the samples confirmed that the native fold still predominated in the ␤-lactoglobulin. ␤-lactoglobulin could be crystallized from all the powders and the normal space group and cell dimensions were determined for the 8 samples that gave crystals of good enough quality for X-ray studies. Differential scanning microcalorimetry of samples dispersed in a phosphate buffer showed a clear difference between Goups 1 and 2 with a more prominent peak due to ␣-lactalbumin in the Group 2 samples. Light scattering and size exclusion chromatography showed that two types of aggregates were present to a variable extent in all the samples and after a heat treatment, the larger aggregates tended to predominate in