Comparison of two extraction methods for the detection of hepatitis A virus in lettuces using the murine norovirus as a process control.

Enteric viruses are important agents of foodborne diseases. In recent years, raw fruits and vegetables have frequently been involved in foodborne transmission of enteric viruses to humans, particularly noroviruses and hepatitis A virus (HAV). Although viral contamination can occur at any stage of food processing, primary production is a critical stage in which prevention measures are essential to minimise the risk of infection to consumers. Due to the low infectious doses and low concentrations of enteric viruses in food samples, an efficient and rapid virus concentration method is required for routine control and risk assessment. In this study, the virus concentration reference method proposed by the CEN/TC275/WG6/TAG4 working group for samples of soft fruits and salad vegetables was compared with a method including a filtration step in order to recover hepatitis A virus (HAV) on lettuces. Murine norovirus (MNV-1) was used as a process control and detected simultaneously with HAV in a one-step duplex RT-qPCR following both procedures. The HAV LOD ranged from 10 to 100 PFU/25g of lettuce in the presence or absence of MNV-1, regardless of method used. In conclusion, MNV-1 offers a very reliable and simple way to monitor the quality of the detection procedures. Although it has been found that both methods achieved an identical limit of detection, the method including a filtration step requires less processing and could be proposed as an alternative method.

Adhesion of human pathogenic enteric viruses and surrogate viruses to inert and vegetal food surfaces.

Enteric viruses, particularly human Noroviruses (NoV) and hepatitis A virus (HAV), are key food-borne pathogens. The attachment of these pathogens to foodstuff and food-contact surfaces is an important mechanism in the human contamination process. Studies were done to investigate the nature of the physicochemical forces, such as hydrophobic and electrostatic ones, involved in the interaction virus/matrix but, at this day, only few data are available concerning surface properties of viruses and prediction of the adhesion capacity of one specific virus onto matrices is still very difficult. The purpose of this study was to propose a reference system, including a representative virus surrogate, able to predict as close as possible behaviour of pathogenic viruses in term of adhesion on inert (stainless steel and polypropylene) and food surfaces (lettuce leaves, strawberries and raspberries). The adhesion of human pathogenic enteric viruses, cultivable strain of HAV and non-cultivable strains of human NoV (genogroups I and II), have been quantified and compared to these of human enteric viruses surrogates, included the MNV-1 and three F-specific RNA bacteriophages (MS2, GA and Qß). A standardized approach was developed to assess and quantify viral adhesion on tested matrices after a contact time with each virus using real-time RT-PCR. Methods used for virus recovery were in accordance with the CEN recommendations, including a bovine Enterovirus type 1 as control to monitor the efficiency of the extraction process and amplification procedure from directly extracted or eluted samples. The adhesion of human pathogenic viruses, ranging from 0.1 to 2%, could be comparable for all matrices studied, except for NoV GII on soft fruits. Adhesion percentages obtained for the studied surrogate virus and phages were shown to be comparable to those of HAV and NoV on inert and lettuce surfaces. The MNV-1 appeared as the best candidate to simulate adhesion phenomena of all human pathogenic enteric viruses on all studied surfaces, while MS2 and GA bacteriophages could be a good alternative as model of viral adhesion on inert and lettuce surfaces. These results will be usable to design relevant experimental systems integrating adhesion behaviour of enteric viruses in the assessment of the efficiency of a technological or hygienic industrial process.

A predictive microbiology approach for thermal inactivation of Hepatitis A virus in acidified berries.

Hepatitis A virus (HAV) is a food-borne enteric virus responsible for outbreaks of hepatitis associated with consumption of raw vegetables. Soft fruits, such as red berries, exposed to faecal contamination are increasingly responsible for collective food-borne illnesses associated with HAV, when eaten raw or used in unprocessed foods. Heat is the most effective measure for the inactivation of HAV. Thermal treatments are used on fruits as a decontamination method, but they have to be adapted to product characteristics; indeed, factors such as sugar or pH may have an impact on the viral sensitivity to thermal treatments. A model was developed for the inactivation of HAV in red berries without supplemented sugar and with different pH values. Nonlinear inactivation curves in acidified raspberries were modelled using an integrated model, with a single equation nesting secondary models of temperature and pH in the primary model. Model predictions were then confronted to experimental results obtained in another laboratory on other berries with different pH values. Excellent predictions were obtained in most cases, while failed predictions provided safe results, with the model predicting higher residual virus titres than what was observed.

Use of a robotic RNA purification protocol based on the NucliSens easyMAG for real-time RT-PCR detection of hepatitis A virus in bottled water.

Hepatitis A virus (HAV) infection is the leading cause of acute viral hepatitis throughout the world. An important part of viral control is rapid detection of HAV in drinking water contaminated with feces. One critical step in HAV detection methods is sample preparation. The objective of this study was to test the efficacy of different approaches to extracting RNA from HAV-inoculated bottled water. The optimal method is based on viral concentration by filtration on membrane filters and elution of adsorbed viruses from filters before RNA extraction and RT-PCR amplification. In this approach, the commercially available NucliSens easyMAG bio-robot (Biomérieux) performs viral RNA purification with silica magnetic beads, which mediate purification of nucleic acids by binding them and allowing other substances to be washed away. A new rapid simplified NucliSens easyMAG-based approach is described and compared with the classical NucliSens easyMAG approach and with manual silica-based spin column purifications (Qiagen). The limit of detection (LOD) with the new rapid simplified NucliSens easyMAG approach was about 1PFU/1.5L against about 100PFU/1.5L using conventional sample treatments that included a concentration step using ultra-filtration.

Detection and quantification by real-time RT-PCR of hepatitis A virus from inoculated tap waters, salad vegetables, and soft fruits: characterization of the method performances.

Water, salad vegetables and fruits exposed to fecal contamination may cause outbreaks of hepatitis A. A protocol of viral concentration by filtration on electronegative membrane filter and a protocol based on a viral elution in Tris-glycine buffer, pH 9.5 with concentration by polyethylene glycol precipitation were associated with real-time, reverse transcriptase-PCR to detect hepatitis A virus (HAV) artificially inoculated in 2 l of tap water, or on 25 g of fruits or salad vegetables. These methods were characterized by an intra-laboratory study using the international standard ISO 16140 on five types of tap water, six types of fruit and five types of salad vegetable. Linear regression models describing the quantitative reactions were good fits to data, and the variances of results were constant in the whole range of viral concentrations tested, which was from about 1.7 to 5.7 log plaque-forming units (PFU) per 2 l of tap water, from about 2.0 to 4.5 log PFU/25 g of fruits, and from 1.5 to 3.5 log PFU/25 g of salad vegetables. Fractions of inoculated viruses recovered were estimated to be about 20% for tap water, about 16% for salad vegetables, and about 7% for fruits. The probability of detecting positive samples was 50% (the critical level of detection) when 2 l samples of tap water were inoculated with 0.7 log PFU of HAV, 25 g samples of iceberg lettuce were inoculated with 1.0 log PFU of HAV, and 25 g samples of fresh and frozen raspberries were inoculated with 1.0 and 1.5 log PFU of HAV, respectively.

Intra-laboratory validation of a concentration method adapted for the enumeration of infectious F-specific RNA coliphage, enterovirus, and hepatitis A virus from inoculated leaves of salad vegetables.

Salad vegetables exposed to fecal contamination may cause outbreaks of hepatitis or gastro-enteritis if they are eaten raw. A procedure, based on elution with phosphate-buffered saline and concentration by filtration through membrane filters, was developed for the recovery of enteric viruses from salad leaves. The method was evaluated using lettuce leaves inoculated with hepatitis A virus (HAV), poliovirus, and MS2 bacteriophage. In addition, this method was validated by an intra-laboratory study using leaves of various salad vegetables inoculated with MS2 phage. The French standard NF V 03-110 was used to establish the general principle and the technical protocol of the validation procedure. Linear regression models describing the quantitative reactions were good fits to data in the whole range of viral concentrations tested, which was from about 1 to 4 log plaque-forming units (PFU) per 25 g of lettuce. The fractions of inoculated viruses recovered were estimated to be about 64% for HAV, 18% for poliovirus, and 29% for MS2. No significant effect of the food matrix was found using various types of salad vegetable (butter lettuce, iceberg lettuce, romaine lettuce, witloof chicory, curly endive, corn salad, rocket and watercress). Moreover, the variance of the results was constant for all levels of virus contamination within the experimental range. Intermediate reproducibility experiments were also performed to allow calculation of the uncertainty factor, which was found to be 0.58 log PFU/25 g. When used in association with phage enumeration, this validated procedure is rapid enough to be used for screening salad vegetables for evaluation of the efficacy of processes for control of pathogenic microorganisms on such foods.

Diversity of enterovirus sequences detected in oysters by RT-heminested PCR.

Oysters harvested in western France, from five sites associated with outbreaks of food-borne norovirus gastroenteritis between February 2000 and March 2001, were assayed for enterovirus RNA by reverse transcriptase-heminested polymerase chain reaction (RT-heminested PCR). Forty percent (21/52) of shellfish samples (pool of seven oysters) were contaminated by enteroviruses. Infectious coxsackieviruses serotype A21 were isolated from three of these positive samples. Amplicons corresponding to 65 base sequences in the 5' untranslated region of the enteroviral genome were analyzed by direct sequencing. Interpretable results were obtained from 18 amplicons, but mixtures of sequences confused the results from 3 samples. Sequences isolated from samples from the different sites were different but similarities were observed between sequences detected in shellfish from two sites at different dates. Sequences were also compared to sequences of human, bovine and porcine enteroviruses. Both human and animal origins of enterovirus contamination of shellfish seemed likely.

Modified concentration method for the detection of enteric viruses on fruits and vegetables by reverse transcriptase-polymerase chain reaction or cell culture.

Fruits and vegetables may act as a vehicle of human enteric virus if they are irrigated with sewage-contaminated water or prepared by infected food handlers. An elution-concentration method was modified to efficiently detect, by reverse transcriptase-polymerase chain reaction (RT-PCR) or by cell culture, contamination by poliovirus, hepatitis A virus (HAV), and Norwalk-like virus (NLV) of various fresh and frozen berries and fresh vegetables. The protocol included washing the fruit or vegetable surface with 100 mM Tris-HCl, 50 mM glycine, and 3% beef extract, pH 9.5 buffer, which favors viral elution from acid-releasing berries, supplemented with 50 mM MgCl2 to reduce the decrease in viral infectivity during the process. The viral concentration method was based on polyethylene glycol precipitation. Copurified RT-PCR inhibitors and cytotoxic compounds were removed from viral concentrates by chloroform-butanol extraction. Viruses from 100 g of vegetal products could be recovered in volumes of 3 to 5 ml. Viral RNAs were isolated by a spin column method before molecular detection or concentrates were filtered (0.22-microm porosity) and inoculated on cell culture for infectious virus detection. About 15% of infectious poliovirus and 20% of infectious HAV were recovered from frozen raspberry surfaces. The percentage of viral RNA recovery was estimated by RT-PCR to be about 13% for NLV, 17% for HAV, and 45 to 100% for poliovirus. By this method, poliovirus and HAV RNA were detected on products inoculated with a titer of about 5 x 10(1) 50% tissue culture infectious dose per 100 g. NLV RNA was detected at an initial inoculum of 1.2 x 10(3) RT-PCR amplifiable units. This method would be useful for the viral analysis of fruits or vegetables during an epidemiological investigation of foodborne diseases.