Metagenomics to identify viral indicators in the produce chain

Detection of human pathogenic viruses in produce or irrigation water currently relies on culture- based methods of bacterial indicators, which frequently fail to correlate with the presence of enteric viruses. Culture independent metagenomic approaches (i.e. massive sequencing) provide the highest resolution to analyze species diversity and will be applied to irrigation water, stools (which may contaminate agricultural and produce handling facilities by food handlers) and produce in order to search for new indicators. Our project goal is to identify viral specie/s which correlates with presence/abundance of pathogenic viruses in irrigation waters and produce. Our specific objectives are: 1) Optimization of sample preparation procedure for viral metagenomics from irrigation water samples 2) Determination of the viral community composition of samples previously analyzed that tested positive or negative for the presence of human pathogenic viruses 3) Identification of specific viral species or groups whose presence/abundance correlates with the occurrence of human pathogenic viruses in stools, irrigation waters and produce. The identification of meaningful viral indicator/s will allow produce industry to simplify the control of enteric viruses by an easy and rapid procedure to detect and quantify the indicator, which in a short term will be implemented in the produce chain.

Fresh produce is an important vector for the transmission of human pathogenic viruses like norovirus (NV), hepatitis A virus (HAV), and hepatitis E virus (HEV) to the human host. These viruses are transmitted through the fecal-oral route and cause diarrheic and hepatic diseases in the population. NVs, for instance, are the leading cause of foodborne outbreaks of acute gastroenteritis, with an estimate of 120 million cases each year. Surface irrigation water and stools (that may contaminate agricultural and produce handling facilities by food handlers) represent the main transmission sources of viruses to produce. Therefore, there is a clear need to identify microorganisms that can easily and reliably indicate the presence or absence of human pathogenic viruses along the produce chain. Current fecal indicator bacteria (FIB; i.e. enterococci and Escherichia coli) and emerging microbial source tracking (MST) methods may indicate the presence and even the likely source of bacterial contamination of water and food, but they are less effective at determining the potential risk to health from human pathogenic viruses. Bacteriophages (including F-specific RNA, somatic coliphages and Bacteroides spp. phages) are generally better predictors of human pathogenic virus persistence and environmental behavior than FIB because of structural similarities. However there is no consensus on the best phage to indicate human pathogenic viral contamination. Due to the lack of good reliable indicators, it has been suggested that searching for the target human pathogenic virus is the only way to check fecal contamination by viruses, however this is expensive, time consuming, and even difficult to interpret. Recently next generation sequencing (NGS) tools open a great opportunity to identify reliable viral indicators. The main objective of this proposal is to determine the viriome (presence and relative abundance of sequences from viruses, including phages and human pathogenic viruses) of irrigation waters samples, produce and stools (that may contaminate agricultural and produce handling facilities by food handlers) by using NGS. The proposed project aims at identifying a more significant indicator/s of the presence or absence of human pathogenic viruses in agricultural inputs, within the on-farm agricultural environment, in/on produce commodities and in produce handling facilities (similarly to the use of E. coli numbers associated with the presence of human bacterial pathogens). This insight will facilitate the development of a tool for a rapid detection and quantification of the viral indicator/s. This proposal benefits from having a large panel of stools and reclaimed and surface irrigation water samples positive for NV GI and GII, HAV and hepatitis E virus (HEV).

1. Optimization of sample preparation procedure for viral metagenomics from irrigation water samples. 

1.1. Optimization of the sample preparation by using spiked samples 

1.2. Validation of the optimized procedure in positive water samples 

2. Determination of the viral community composition of samples previously analyzed that tested positive or negative for the presence of human pathogenic viruses. 

2.1. Determination of the entire viriome of archived samples 

2.2. Collection of irrigation waters and produce samples 

2.3. Determination of the entire viriome of newly collected samples 

3. Identification of specific viral species or groups whose presence/abundance correlates with the occurrence of human pathogenic viruses in stools, irrigation waters and produce.

Detection of human enteric viruses would be the ideal option for determining the virological quality of irrigation waters and produce; but at present, this is neither practical nor feasible in routine testing for most of the laboratories. The rapid and extensive development of nextgeneration sequencing (NGS) has opened up more opportunities to advance understanding in virome characterization. Virome characterization via metagenomics comprises the general steps of sampling, sample processing, sequencing, and data analysis. In this study, the selected metagenomics workflow was already proven suitable for the virome characterization of irrigation water, produce and stool samples. These samples were also analyzed for the presence of the most relevant human enteric viruses (i.e., norovirus GI, GII, RV, HAV and HAstV) by viability RT-qPCR, and their presence was correlated with metagenomics data in order to select a suitable virus indicator. Overall, our data indicated that the recently proposed viral indicator, crAssphage, significantly correlated with the presence of human enteric viruses. To confirm this point, crAssphage occurrence was also determined in the same samples, confirming these results, although 100% correlation was not achieved. Given that an ideal indicator should be easily detectable by rapid and inexpensive laboratory tests and inform ideally on viral potential infectivity, the detection of crAssphage by molecular methods lacks in matching these requirements, especially because it does not infer infectivity. Metagenomics data from this study also suggested that EBPR podovirus 2, Aeromonas phage4_L372X, Pseudomonas phage NP1, EBPR siphovirus 1, Bordetella phage vB_BbrM_PHB04, and Aeromonas phage 62AhydR11PP are alternative suitable indicators for contamination by enteric viruses. However further research is needed in order to develop rapid methods for their detection and assess their correlation using other methods as qPCR or culture.