Evaluation of a norovirus detection methodology for ready-to-eat foods

Abstract

Despite recent norovirus (NoV) foodborne outbreaks related to consumption of ready-to-eat (RTE) foods, a standardized assay to detect NoV in these foods is not available yet. Therefore, the robustness of a methodology for NoV detection in RTE foods was evaluated. The NoV detection methodology consisted of direct RNA extraction with an eventual concentration step, followed by RNA purification and a multiplex RT-qPCR assay for the detection of GI and GII NoV and the murine norovirus-1 (MNV-1), the latter used as process control. The direct RNA extraction method made use of the guanidine-isothiocyanate containing reagent (Tri-reagent®, Ambion) to extract viral RNA from the food sample (basic protocol called TriShort), followed by an eventual concentration step using organic solvents (extended protocol called TriConc). To evaluate the robustness of the NoV detection method, the influence of (1) the NoV inoculum level and (2) different food types on the recovery of NoV from RTE foods was investigated. Simultaneously, the effect of two RNA purification methods (manual RNeasy minikit (Qiagen) and automated NucliSens EasyMAG (BioMérieux)) on the recovery of NoV from these foods was examined. Finally, MNV-1 was evaluated as process control. First of all, high level GI and GII NoV inocula (~ 10 6 NoV genomic copies/10 g) could be recovered from penne salad samples (10 g) in at least 4 out of 6 PCRs, while low level GI and GII NoV inocula (~ 10 4 NoV genomic copies/10 g) could be recovered from this food product in maximally 3 out 6 PCRs, showing a significant influence of the NoV inoculum level on its recovery. Secondly, low level GI and GII NoV inocula (10 4 NoV genomic copies/10 g) were spiked onto 22 ready-to-eat food samples (10 g) classified in three categories (soups, deli sandwiches and composite meals). The GI and GII NoV inocula could be recovered from 20 of the 22 samples. The TriConc protocol provided better recoveries of GI and GII NoV for soups while the TriShort protocol yielded better results for the recovery of GII NoV from composite meals. NoV recovery from deli sandwiches was problematic using either protocol. Thirdly, the simultaneous comparison of two RNA purification protocols demonstrated that automated RNA purification performed equally or better compared to manual RNA extraction. Finally, MNV-1 was successfully evaluated as process control when detecting NoV in RTE foods using this detection methodology. In conclusion, the evaluated NoV detection method was capable of detecting NoV in RTE foods, although recoveries were influenced by the inoculum level and by the food type.