Quantitative farm-to-fork risk assessment model for norovirus and hepatitis A virus in European leafy green vegetable and berry fruit supply chains.

Fresh produce that is contaminated with viruses may lead to infection and viral gastroenteritis or hepatitis when consumed raw. It is thus important to reduce virus numbers on these foods. Prevention of virus contamination in fresh produce production and processing may be more effective than treatment, as sufficient virus removal or inactivation by post-harvest treatment requires high doses that may adversely affect food quality. To date knowledge of the contribution of various potential contamination routes is lacking. A risk assessment model was developed for human norovirus, hepatitis A virus and human adenovirus in raspberry and salad vegetable supply chains to quantify contributions of potential contamination sources to the contamination of produce at retail. These models were used to estimate public health risks. Model parameterization was based on monitoring data from European supply chains and literature data. No human pathogenic viruses were found in the soft fruit supply chains; human adenovirus (hAdV) was detected, which was additionally monitored as an indicator of fecal pollution to assess the contribution of potential contamination points. Estimated risks per serving of lettuce based on the models were 3×10(-4) (6×10(-6)-5×10(-3)) for NoV infection and 3×10(-8) (7×10(-10)-3×10(-6)) for hepatitis A jaundice. The contribution to virus contamination of hand-contact was larger as compared with the contribution of irrigation, the conveyor belt or the water used for produce rinsing. In conclusion, viral contamination in the lettuce and soft fruit supply chains occurred and estimated health risks were generally low. Nevertheless, the 97.5% upper limit for the estimated NoV contamination of lettuce suggested that infection risks up to 50% per serving might occur. Our study suggests that attention to full compliance for hand hygiene will improve fresh produce safety related to virus risks most as compared to the other examined sources, given the monitoring results. This effect will be further aided by compliance with other hygiene and water quality regulations in production and processing facilities.

Virus transfer proportions between gloved fingertips, soft berries, and lettuce, and associated health risks.

Multiple outbreaks of human norovirus (hNoV) have been associated with fresh produce, such as soft berries and lettuce. Even though food handlers are considered an important source for the introduction of hNoV into food chains, their contribution to public health risks associated with hNoV remains unknown. To assess to which extent food handlers contribute to the introduction and spread of hNoV in fresh produce chains quantitative virus transfer data are needed. We estimated transfer proportions of hNoV GI.4, GII.4, murine norovirus (MNV-1), a culturable surrogate of hNoV, and human adenovirus (hAdV-2), a human pathogen proposed as an indicator for human faecal pollution, between gloved fingertips and raspberries, strawberries, and lettuce, by quantitative RT-PCR and cell culture if applicable. Virus transfer proportions were corrected for virus-matrix specific recoveries, and variability and uncertainty of the parameters were estimated. Virus transfer from gloves to soft berries was generally lower as compared to lettuce, with mean transfer proportions ranging between 0.1 to 2.3% and 9 to 10% for infectious MNV-1 and hAdV-2, respectively. Transfer from produce to glove was mostly greater than transfer from glove to produce, adding to the likelihood of virus transfer due to cross contamination from contaminated produce via food handlers. HNoV GI.4 and hNoV GII.4 showed no significant difference between their mean transfer proportions. Using the estimated transfer proportions, we studied the impact of low and high transfer proportions on the public health risk, based on a scenario in which a food handler picked raspberries with contaminated fingertips. Given the made assumptions, we could show that for a pathogen as infectious as hNoV, low transfer proportions may pose a greater public health risk than high transfer proportions, due to a greater viral spread. We demonstrated the potential of food handlers in spreading hNoV in food chains, showing that prevention of virus contamination on food handlers' hands is crucial for food safety. Nevertheless, complete prevention of virus contamination on fresh produce cannot be achieved in reality, and reliable and effective intervention measures are consequently required. We estimated that, especially for low transfer proportions, a robust one log10-unit reduction of infectious hNoV on contaminated produce, and on food handlers' hands, could lower the public health risk substantially. Using the obtained data in quantitative risk assessment will aid in elucidating the contribution of food handlers in hNoV transmission.

Persistence of human norovirus in reconstituted pesticides--pesticide application as a possible source of viruses in fresh produce chains.

The consumption of fresh produce is frequently associated with outbreaks of human norovirus (hNoV) disease. To prevent the contamination of fresh produce with hNoV, knowledge of the possible introduction sources of the viruses, such as water, is needed to be able to implement appropriate and efficient preventive measures. Contaminated water used to reconstitute pesticides could be a relevant source of infectious hNoV, determined by the initial level of virus contamination and the persistence of these viruses in reconstituted pesticides. We studied the persistence of hNoV GI.4, hNoV GII.4 and murine norovirus (MNV-1), the only culturable norovirus, in eight different pesticides after 0 and 2h. Virus concentrations were determined by reverse transcriptase PCR, and infectivity of MNV-1 was determined by endpoint dilutions followed by maximum likelihood estimations. MNV-1 was found to remain infectious in seven of the eight tested pesticides at the highest concentration applied in practice. In the presence of the insecticide Vertimec, MNV-1 infectivity decreased rapidly with a 1.9 log(10)-unit reduction at timepoint T(0). Also, the concentration of NoV GI.4 RNA decreased considerably with a 1.7 log(10)-unit reduction; whereas the detected PCR fragment of hNoV GII.4 remained stable. Assuming a similar persistence of infectious MNV-1 and hNoV we can conclude that water containing hNoV used to dilute pesticides may be an important source of infectious hNoV in fresh produce chains. The application of pesticides may therefore not only be a chemical hazard, but also a microbiological hazard for public health. The inclusion of antiviral substances in reconstituted pesticides may be appropriate to reduce the virological health risk posed by the application of pesticides.

Persistence of human norovirus GII.4 and GI.4, murine norovirus, and human adenovirus on soft berries as compared with PBS at commonly applied storage conditions.

Human noroviruses (hNoV) have been detected on soft fruits. Especially raspberries have been found to be associated with outbreaks of gastroenteritis suggesting persistence of hNoV on these fruits. Therefore, the persistence of hNoV GII.4 and GI.4, murine norovirus (MNV-1, a culturable surrogate for hNoV), and human adenovirus (hAdV, an indicator for human fecal contamination), on raspberries, strawberries and in phosphate buffered saline (PBS) at 4°C, 10°C and 21°C, mimicking commonly applied storage conditions was studied by molecular and cell culture techniques. Monophasic, biphasic and Weibull models were fitted to virus counts with maximum likelihood estimation. The tested viruses were persistent (≤0.5 log(10)-unit reduction in viral titer) under all studied conditions in PBS, at 4°C and 10°C on raspberries, and at 4°C on strawberries. The difference in viral persistence on raspberries and strawberries was most pronounced at 21°C. Here, infectious MNV-1 and hAdV particles decayed rapidly on strawberries with TFL-values (time for the first log(10)-unit reduction) of only 1day (95% CI of 0.6-1 and 0.8-1days, respectively). On raspberries, however, the TFL-value of infectious MNV-1 was found to likely exceed the shelf life of the berries with 3days (95% CI of 2.8-3.1days); hAdV remained infectious with only 0.3 log(10)-unit reduction (95% CI of 0.2-0.4) in viral titer. For hNoV GI, a TFL-value of 2days (95% CI 1-4days) was determined based on the targeted genome fragment, whereas the TFL-value of hNoV GII exceeded the shelf life of strawberries at 21°C. The greater viral persistence on raspberries as compared to strawberries, especially at 21°C, may at least in part explain why raspberries are more frequently associated with hNoV outbreaks than strawberries. Moreover, our results show that due to the high persistence of the virus already low contamination levels of the highly infectious hNoV may be associated with an infection risk of humans after consumption of raspberries. The estimated decay parameters and uncertainties of this study serve as important input requirements in the quantitative assessment of public health risks from the consumption of soft fruits.

The tumor-selective viral protein apoptin effectively kills human biliary tract cancer cells.

Biliary tract cancer, or cholangiocarcinoma, has a poor prognosis. Resection is the only curative treatment, but only a minority of patients are eligible. Chemotherapy and gamma-irradiation are merely palliative, as they are unable to remove the malignancy completely. The chicken anemia virus-derived protein apoptin induces apoptosis in a wide range of human tumor cells and is not hindered by mutations inactivating p53 or by overexpression of Bcl-2, changes known to frustrate chemotherapy and radiation therapy. We examined whether apoptin kills human biliary tract cancer cells. Expression of apoptin by means of plasmids caused extensive cell death in three independent cholangiocarcinoma cell lines, CC-LP, CC-SW, and Mz-ChA-1, regardless of their oncogenic mutations, which included inactivated p16 and p53 and the disruption of the transforming growth factor beta signaling pathway. In vitro delivery of apoptin by an adenoviral vector completely eradicated cholangiocarcinoma cells. Moreover, coexpression of the broad-spectrum caspase inhibitor p35 with apoptin only delayed the induced cell death. Changes in nuclear morphology still occurred early after transfection, and nuclei eventually disintegrated, suggesting that apoptin-induced cell death in these cells is not blocked by mutations in either the initiation or execution phase of apoptosis. The efficient induction of cell death by apoptin in cholangiocarcinoma cell lines makes apoptin an attractive candidate for molecular therapy of biliary tract cancer.

Induction of insolubility by herpes simplex virus VP22 precludes intercellular trafficking of N-terminal Apoptin-VP22 fusion proteins.

The herpes simplex virus protein VP22 has the intriguing ability to deliver proteins from an expressing cell to neighboring cells. Fusion of VP22 to Apoptin, a protein that induces apoptosis in tumor cells but not in normal cells, might enhance the delivery of Apoptin. To analyze this hypothesis two fusion proteins of VP22 and full-length Apoptin were constructed, namely VP22-VP3 and VP3-VP22, and their apoptosis-inducing ability and intercellular spreading behavior were analyzed by transfection in tumor cells. While both of the Apoptin-VP22 fusion proteins retained the capacity to kill tumor cells, neither of them showed intercellular trafficking. To determine whether the presence of a nuclear localization signal in the C-terminus of Apoptin caused nuclear retention of the fusion protein and the subsequent lack of intercellular spreading, VP22 was fused to the biologically active N-terminal part (residues 1-69) of Apoptin (VP3n), which lacks the nuclear localization signal. However, analysis of the VP3n-VP22 fusion constructs gave no evidence of intercellular transport. A more careful inspection of different fractions of cell lysates expressing Apoptin with or without fusion to VP22 revealed that both the full-length Apoptin protein and its fusion with VP22 are insoluble. Despite the fact that VP3n was found to be soluble on its own, which could make it amenable to transport by VP22, the VP3n-VP22 fusion proteins were present exclusively in the insoluble fraction. We hypothesize that the N-terminal multimerization domain of Apoptin cooperates with VP22 to facilitate aggregation with cellular proteins, thereby inducing insolubility. From these results we conclude that, depending on the fusion partner, VP22 can have a negative effect on the solubility of fusion proteins, which consequently precludes intercellular trafficking. Such properties should be taken into account when establishing new VP22-mediated protein transduction systems.