Molecular detection of noroviruses in ready-to-eat foods and fruit products

In this PhD, three main goals were defined. The first goal consisted of the development and evaluation of a methodology for detection of noroviruses (NoV) in ready-to-eat (RTE) foods and soft red fruits while the second main goal included the evaluation of the murine norovirus 1 (MNV-1) as control reagent for different steps throughout the NoV detection protocols. Finally, a screening study on a selection of fruit produce products towards NoV presence was the third main goal of this PhD. To illustrate these goals, a literature study was performed in chapter 1. In this literature study, a brief overview of the most important food borne viruses was followed by a more detailed description of NoV in terms of classification, virion and genome structure. The NoV genotype most commonly identified in NoV gastroenteritis outbreaks (NoV GII.4) was described as well. The importance of NoV as a food borne pathogen was highlighted by data originating from official bodies such as CDC (Centers for Disease Control and Prevention; USA) and EFSA (European Food Safety Authority; Europe), accompanied with data gathered on own initiatives by research groups. The two main transmission routes of NoV contamination of foods (pre-harvest contamination via contact with contaminated water and (post-) harvest contamination via an infected food handler/picker) were investigated by summarizing and analyzing 59 NoV food borne outbreaks described between 2000 and 2010. Furthermore, the three main steps of NoV detection in food were portrayed in detail to prepare the development and evaluation of the NoV detection methodology in chapters 2, 3, 4 and 5. Finally, the use of adequate positive and negative controls to assure reliable detection of NoV in foods was illustrated. For the first and second goals of this PhD, a molecular assay for detection of the purified NoV genomic material was optimized and subsequently combined with protocols for extraction of (genomic material of) NoV from RTE foods and soft red fruits in chapters 4 and 5. MNV-1 was included in these protocols as control reagent. The molecular detection assay described in chapter 2 was a quantitative two-step multiplex real-time reverse transcriptase (RT-) PCR assay for simultaneous detection of NoV genogroup I (GI) and II (GII) and the murine norovirus 1 (MNV-1), the latter used as internal amplification control (IAC). For this multiplex assay, NoV GI and GII specific primers and hydrolysis probes designed by the European Committee for Standardization/ Technical Committee 275 / Working Group 6 /Task Group 4 on virus detection in foods (CEN/TC275/WG6/TAG4 working group) were combined with primers for murine norovirus 1 designed by Baert and colleagues (2008b). Evaluation of this multiplex assay showed a high concordance between the multiplex assay and the corresponding singleplex PCR assays. Specificity analysis of the multiplex assay by testing a NoV RNA reference panel and clinical GI and GII NoV samples showed that specific amplification of NoV GI and GII was possible. In addition, no cross-amplification was observed when subjecting a collection of bovine NoV and other (non-NoV) enteric viruses to the multiplex assay. Finally, MNV-1 was successfully integrated as IAC, although a sufficiently low concentration was needed to avoid interference with the possibility of the developed multiplex assay to quantitatively and simultaneously detect the presence of GI and GII NoV within one sample. During development of the multiplex real-time RT-PCR assay, contamination issues were encountered and the investigation towards the source of the positive no template controls (NTCs) was described in chapter 3. This investigation was believed to be necessary because of the need for reliable detection of 10 or less NoV genomic copies per PCR reaction, due to the low infectious dose of GI and GII NoV. In this chapter, a suspicion of well-to-well migration of positive control DNA (a short synthetic single stranded DNA (ssDNA) fragment) during real-time PCR runs was uttered as hypothetic cause of the positive NTCs. Results in this chapter showed that evaporation of water occurred during real-time PCR runs regardless of the DNA type, the reaction plate seal type and the use of mineral oil as cover layer. It was also suggested that co-evaporation of DNA took place, with an apparent negative correlation between the size of the DNA type and the extent of this co-evaporation. The use of mineral oil as cover layer and plasmid DNA as quantitative positive PCR control resulted in a complete absence of positive NTCs while only negligible effects were noticed on the performance of the real-time PCR. After development of the multiplex real-time RT-PCR assay and the resolving of the contamination issues, two protocols for extraction of (genomic material) of NoV from foods were evaluated towards robustness and sensitivity while MNV-1 was evaluated as process control in both protocols. The evaluation of a direct RNA extraction protocol for extraction of NoV genomic material (RNA) from RTE foods was described in chapter 4, while the evaluation of an elution-concentration protocol for extraction of NoV from soft red fruits was illustrated in chapter 5. For the RTE foods, the direct RNA extraction protocol made use of a guanidine isothiocyanate containing reagent 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). The protocol for extraction of NoV from soft red fruits consisted of alkaline elution of NoV particles from the food, followed by polyethylene glycol (PEG) precipitation and organic solvent purification. For both protocols the RNA was subsequently purified. This purified RNA was detected by the multiplex real-time RT-PCR assay as described in chapter 2. To evaluate both NoV extraction methods towards sensitivity and robustness, the influence of (1) the NoV inoculum level and (2) different food types on the recovery of NoV from these foods was investigated. First of all, a significant influence of the NoV inoculum level on the recovery of NoV from foods was demonstrated for both protocols. High level inocula could be recovered from penne salad, selected as typical RTE food, with higher recovery success rates compared to low level inocula. For these high inoculum levels, the TriShort and TriConc protocols resulted in mean recovery efficiencies of >1 % and 0.1 to 10 %, respectively. Recovery of these low and high level NoV inocula from frozen raspberry crumb was possible with high recovery success rates and with mean recovery efficiencies of 10 to 30 % in most cases. Secondly, a significant influence of the food type on the recovery of NoV could be shown for both protocols. For the direct RNA extraction protocol, the TriConc protocol provided better NoV recoveries for soups, while TriShort and TriConc protocols performed likewise for composite meals and deli sandwiches, although NoV recovery from the latter food type was problematic. For the elution-concentration protocol, a significant influence of the soft red fruit product type on the recovery efficiency of NoV GI and MNV-1 was noticeable, while no significant differences could be shown for GII NoV. In general, the recovery of NoV was more efficient and successful from the strawberry puree compared to a frozen forest fruit mix and fresh raspberries. Regarding the evaluation of MNV-1 as control reagent, results from chapter 4 and chapter 5 suggested that a sufficient high concentration of the MNV-1 PC was needed to allow an estimation of possible inhibition of the RT-PCR or of inefficient virus extraction. When used as reverse transcription control or internal amplification control, the concentration should be adjusted to avoid interference with the quantitative properties of the developed multiplex real-time RT-PCR assay. Chapter 6 described the screening of 75 fruit products (raspberries, strawberries, cherry tomatoes and fruit salads) for NoV presence using the virus extraction protocol described in chapter 5 combined with the multiplex real-time RT-PCR assay illustrated in chapter 2. In total, 18 samples tested positive for GI and/or GII NoV genomic material despite a good bacteriological quality. The level of detected NoV genomic copies concentrations ranged between 2.5 and 5.0 logs per 10 grams of fruit sample. NoV GI and/or GII were found in 4/10, 7/30, 6/20 and 1/15 of the tested raspberries, cherry tomatoes, strawberries and fruit salad samples, respectively. However, confirmation of the positive real-time PCR results by sequencing genotyping regions in the NoV genome was not possible. The question whether or not these unexpected high number of NoV positive results obtained should be perceived as a public health threat was raised and discussed. In conclusion, methods for detection of NoV in RTE foods and soft red fruits were developed and evaluated towards sensitivity and robustness. For detection of NoV in soft red fruits and ready-to-eat foods, an elution-precipitation protocol and a direct RNA extraction protocol were combined with an optimized multiplex real-time RT-PCR assay leading to NoV detection protocols with detection limits of ~104 genomic copies / 10g food product. Influence of the NoV inoculum level and food type on NoV recovery was shown. Additionally, MNV-1 was successfully evaluated as control reagent, and suggestions were made towards its use. However, application of the method for NoV detection in fruit products has shown that interpretation of NoV presence by molecular methods is not straightforward and raises several questions, especially towards the public health safety.

http://ift.tt/2qcuz8y