Novel approach for interlaboratory transfer of real-time PCR methods: detecting bovine meat and bone meal in feed.

The availability of robust methods for the species-specific detection of meat and bone meal (MBM) in compound feedingstuffs is an important prerequisite to enforce current and upcoming European legislation on the use of processed animal proteins in animal nutrition. Among possible methods, those based on DNA turned out to be a reliable tool for this aim, since DNA is a quite thermostable molecule able to resist severe heat treatments applied in the manufacturing of animal meals. The application of such methods by control laboratories implies that the method has been validated including an assessment of its robustness. Successful transferability between laboratories is considered an important robustness criterion of the method. However, corresponding guidelines regarding the design of such a study relevant to this field are missing. Here, we demonstrate the feasibility of an alternative concept that was applied to check for the transferability of a qualitative assay for the detection of banned MBM in feedingstuffs at trace level based on real-time PCR. The concept was based on an experimental nested design applying analysis of variance (ANOVA) that was conducted independently in two laboratories and which allows for establishing major factors influencing the result of analysis. Statistical assessment of the results confirmed the importance of the DNA extraction/purification step utilised, whereas the PCR step turned out to be a minor factor regarding the overall variability of the results. Furthermore, blind samples comprised of compound feed adulterated with MBM at 0.1% and blank compound feed were correctly classified as "positive" or "negative" samples, thus confirming fitness of purpose for the method. This approach can be of interest for other research groups working in the development of real-time PCR methods and in their use by control laboratories.