Foodborne diseases induced by viruses involving plant-derived food such as berries are increasingly recognized as threat for public health. For example, a large human norovirus (hNV) gastroenteritis outbreak involving >10,000 diseased people occurred in Germany in 2012, which was traced back to contaminated frozen strawberries. To increase food safety for consumers, three areas were identified, which should be addressed in this thesis. First, different previously described virus extraction methods in combination with quantitative real time PCR (RT-qPCR) were compared using frozen strawberries artificially contaminated with hNV and the recovery rates (RRs) were calculated. Out of five published protocols, the method according to ISO/TS15216-2 revealed the best RRs. Further improvement of RRs from 2.83±2.92% to 15.28±9.73% was achieved by an additional RNA purification using Sephacryl®-based columns. The optimized protocol showed significantly better hNV detection rates by testing of berries originally involved in gastroenteritis outbreaks in 2012 and 2016. Second, novel methods for virus quantification and broad pathogen identification were tested on strawberries from the outbreak in 2012: Digital PCR (dPCR) and Next Generation Sequencing (NGS). A quantification of the hNV genomes using RT-dPCR confirmed a very low mean virus amount in the berries (185 RNA copies/25g), which was similar to the independently assessed RT-qPCR result (257 RNA copies/25g). By NGS, about 29 mio sequence reads were generated and analyzed by RIEMS 4.0 software. They mainly showed homologies to sequences from the plant matrix and from the bacterial flora. The most abundant virus sequences originated from plant-specific viruses. Two virus reads showed homologies to hNV identical to those derived from gastroenteritis patients and a strawberry sample obtained during the outbreak. Third, the thermal inactivation kinetics of hNV in strawberry puree was assessed. Due to the lack of a reliable system for hNV infectivity assessment, capsid integrity assays using RNase treatment prior to detection of protected viral RNA were applied. In addition, murine norovirus (mNV) and Tulane virus (TV) served as surrogates and were analyzed by plaque assay and RT-qPCR. Infectious mNV and TV were completely inactivated (>7 log10 reductions) after treatment at 80°C for 8 sec. However, in comparison to hNV, TV showed a markedly lower stability in capsid integrity assays. A predictive model generated from the data, which covers reduction of hNV capsid-protected RNA between 50-80°C, suggests that temperatures under 70°C are not reliable for inactivation. About 3.5 log10 reductions are achieved by heating at 80°C for 8 sec. In conclusion, methods for detection, identification and characterization of hNV in berries could be improved. The generated heat inactivation model may help developing efficient heat treatment strategies to prevent hNV infection due to consumption of berries in the future.
