A novel PCR-based method to enumerate Salmonella in animal feed

Animal feed can serve as a reservoir for Salmonella in the food production chain. Therefore, it is important to have rapid and sensitive methods for detection and quantification. In this study, a novel approach for quantification of low numbers of Salmonella in feed samples was developed. The protocol included a PCR based method combined with an optimised most probable number (MPN) scheme. The PCR method included an enrichment step in buffered peptone water (BPW) at 37ºC for 18 ± 2 h, followed by centrifugation of a withdrawn 1-ml BPW aliquot. DNA was extracted by an automated procedure from the pellet and subjected to real-time PCR. The qualitative PCR method was compared to a reference culture method using modified semisolid Rappaport-Vassilades (MSRV) agar plates (ISO 6579, Amd D, 2007). Of 81 naturally or artificially contaminated samples tested (soya meal, rape seed meal, rape seed cake and pellets) only three gave results that differed between the PCR and MSRV methods. Ct values for naturally contaminated samples were higher compared to samples artificially contaminated with low numbers (approx. 2 CFU/25 g feed) of stressed Salmonella. To allow quantification of low numbers of Salmonella in feed the developed PCR method was combined with an MPN approach. The traditional MPN scheme was modified in order to make the procedure less laborious, time consuming and costly, as well as being better adjusted to enumerate Salmonella in feed samples. This was achieved using two different approaches: (i) the dilution scheme was adjusted to better enumerate the low numbers presumably found in feed and (ii) the selective enrichment steps were replaced by the qualitative PCR method. In conclusion, the developed PCR method can be used as an alternative method for detecting low numbers of Salmonella in feed samples. In combination with the novel MPN scheme, it can also be employed to generate quantitative data. Studies are in progress to further validate the performance on a larger number of naturally contaminated feed samples and to generate quantitative data on naturally contaminated feed.