Anal Bioanal Chem (2005) 383: 282–290 DOI 10.1007/s00216-005-0013-x

ORIGINA L PA PER

Philippe Corbisier . Stefanie Trapmann . David Gancberg . Liesbeth Hannes . Pierre Van Iwaarden . Gilbert Berben . Heinz Schimmel . Hendrik Emons

Quantitative determination of Roundup Ready soybean (Glycine max) extracted from highly processed flour Received: 19 May 2005 / Revised: 1 July 2005 / Accepted: 5 July 2005 / Published online: 10 August 2005 # Springer-Verlag 2005

Abstract Roundup Ready soybean powder has been subjected to different amounts of DNA fragmentation to assess the accuracy of real-time PCR on processed food. Certified reference material (CRM) containing 10 g kg−1 of Roundup Ready soybean (ERM-BF410d) prepared by a dry-mixing processing method was exposed to water at two temperatures, using three different mixing devices, or to baking temperature (250°C) for 30 min. The amount of DNA extracted from the different samples was quantified by fluorimetry. The amount of fragmentation of the extracted DNA was characterised by gel and capillary electrophoresis and the percentage of genetically modified (GM) soybean was determined by a double quantitative real-time PCR method. Measurement of the event GTS 40-3-2 (RUR) was possible in all the treated materials, because small amplicons were amplified. Correct RUR percentages could be measured for intact powders with little or no DNA fragmentation. For samples with a high level of DNA degradation, however, the accuracy of the measurement was found to depend on the method used for DNA extraction. Genomic DNA isolated by use of silica resin resulted in statistically significant overestimation of the amount of GM. Keywords Glycine max . Genetically modified soybean (GMS) . Polymerase chain reaction (PCR) . Processed food . DNA degradation P. Corbisier (*) . S. Trapmann . D. Gancberg . L. Hannes . P. Van Iwaarden . H. Schimmel . H. Emons Institute for Reference Materials and Measurements, European Commission, Joint Research Centre, Retieseweg 111, 2440 Geel, Belgium e-mail: [email protected] Tel.: +32-14-571890 Fax: +32-14-571548 G. Berben Quality Department of Agro-food Products, Walloon Agricultural Research Centre, Chaussée de Namur 24, 5030 Gembloux, Belgium

Introduction Genetically modified organisms (GMOs) and their derived products are analysed to detect, screen, identify or quantify GMOs in a given matrix [1]. Polymerase chain reaction (PCR) has become a common method of quantifying GM content. Quantitative PCR can be performed in real-time enabling simultaneous amplification and quantification of the target DNA. During that amplification process, unique nucleic acid sequences specific for the genetic modification and nucleic acid sequences specific for the plant taxon or species are detected in DNA extracted from the analysed sample. The relative amount of those sequences is calculated and expressed as the GM percentage. Certified reference materials (CRMs) are essential tools to ensure correct performance of those analytical measurements. Indeed, reference materials certified for their GM content can be used as positive or negative control samples or as calibrants for real-time PCR. The first generation of GMO CRMs was produced at IRMM in 1998–1999 by wet-mixing of the powders in water using a Turrax mixing device to obtain highly homogeneous material. This method led to elevated temperatures during the production phase resulting in highly fragmented DNA. The temperature increase was reduced by cooling the wet-mixed powder during production of the second-generation of GMO CRMs (1999– 2000). Third generation GMO CRMs (since 2001) have been produced by means of a dry-mixing method. Experience showed that the DNA of CRMs produced in this way was less fragmented and, presumably, more stable over long period of time. Fragmented DNA is extracted from highly- processed food, however. The effects of processing have been shown to affect analytical detection of GM ingredients in different highly processed food such as multigrain bread, “poppyseed bakers” or toasted bread [2], industrially processed soybean oil fractions [3], wheat bread [4] or artificial food matrices containing Roundup Ready soybeans (RRS) [5]. Most results indicated that with small amplicon sizes (