Abstract
The evaluation of recovery rates by extracting milk powder and egg powder using eleven different extractants gave approximately similar results for both foods. Compared with the other extraction solutions investigated, ‘1% Tween 20® and 0.4% Triton X-100®’ and ‘4% SDS’ are the most suitable extractants to isolate proteins of hen's egg or milk. When comparing calculated protein recovery rates of egg and milk powder extracts, the results clearly indicated that the choice of a suitable extractant is of particular importance. Qualitative investigation of the extracts via LDS-PAGE followed by silver staining as well as immunoblotting confirmed the results of protein quantification. Hence, the immunoblots showed that the extraction agents had no negative influence on the antigenicity of the extracted allergenic proteins. In this study, variation of extraction temperature led neither to any benefit in extraction quality nor to degradation. Changing pH did not reveal any trends, but progressive protein hydrolysis under strong alkaline conditions. Evaluation of recovery rates as well as results of unspecific and specific staining of the extracts showed that an extraction time of 1 h is sufficient for an appropriate sample preparation. For investigations with and without food matrix different results were obtained. In summary, wheat starch did not influence the extraction quality within all examined materials and different extractants. In contrast, using fat powder and dry cake mix, respectively, led to different results in the extraction procedure. When fat powder and dry cake mix were used as food matrices, some protein recovery rates decreased and some increased depending on the allergen material. These results highlight the fact that the suitability of the extractant not only depends on the properties of the allergen but furthermore on the type of matrix containing the allergen.
Acknowledgements
This study was supported by Arbeitsgemeinschaft industrieller Forschungsvereinigungen (AiF) Project No. AiF 14805 N. Special thanks go to B. Jacobsen, K. Ramcke and S. Schwarz for technical support.