Estimation of human norovirus infectivity from environmental water Samples by In Situ capture RT-qPCR method.

Authors: Tian, Peng; Yang, David; LEI, SHAN - FORMER ARS EMPLOYEE; LI, QIANQIAN - SHANGHAI INSTITUTE OF TECHNOLOGY; LIU, DANLEI - SHANGHAI JIAOTONG UNIVERSITY; WANG, DAPENG - SHANGHAI JIAOTONG UNIVERSITY

Submitted to: Food and Environmental Virology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 8/12/2017
Publication Date: 8/30/2017
Citation: Tian, P., Yang, D., Lei, S., Li, Q., Liu, D., Wang, D. 2017. Estimation of human norovirus infectivity from environmental water Samples by In Situ capture RT-qPCR method.. Food and Environmental Virology. 10(1):29-38. https://doi.org/10.1007/s12560-017-9317-1.
DOI: https://doi.org/10.1007/s12560-017-9317-1

Interpretive Summary: Human noroviruses (HuNoV) are the most common causes of foodborne illness, accounting for 50% of outbreaks and 46% of illnesses. Human noroviruses are highly-contagious pathogens and is the main cause of acute gastroenteritis. HuNoV outbreaks associated with leafy green vegetables have been substantially increasing in proportion in recent years. Being that leafy green vegetables can be contaminated by pathogens via exposure to contaminated environmental water from public waterways, we sampled environmental waters from several public watersheds in the vicinity of Salinas, CA, and quantitated for HuNoV using two molecular-based procedures of varying relevance to viral infectivity. Collected environmental water samples were spiked with Tulane virus (TV), a HuNoV surrogate, for indication of known inhibitory effects upon RT-qPCR. Samples were quantified for TV and HuNoV using RT-qPCR, RNase protection RT-qPCR, and ISC-RT-qPCR, a variant of the former that screens for receptor-associated (and necessarily encapsidated) viral RNA. From 250 RNA extracts, a total of 72 and 16 exhibited genotype I (GI) and GII HuNoV genomic signal, respectively. Being that the molecular-based RT-qPCR method can only indicate the presence of viral genome, the actual infectivity of the RT- qPCR-positive samples remains unknown. We used RNase protection and ISC-RT-qPCR assay to estimate infectivity of the virus. The RNase protection RT-qPCR assay excludes viral debris from contributing to the RNA signal by obliterating the signal-producing ability of viral debris prior to the extraction of the sample for viral RNA that had been protected from digestion by viable capsids. The RNase protection assay indicated that only 20% of RNA signal positive for GI HuNoV by RT-qPCR could be confirmed as not being sourced from viral debris and 80% were likely to be infectious. The ISC-qRT-PCR method also excludes the quantitation of viral debris, but instead achieves this by sequestering viral particles via capsid receptor activity, and quantitating the sequestered viruses. This effect was confirmed using a TV as a surrogate for HuNoV, and demonstrating that heat-lysed TV would not produce detectable RNA signal via ISC-RT-qPCR, while viable TV would. Much like the RNase protection assay, ISC-RT-qPCR found that only 20 (27.8%) out of 72 GI HuNoV positives from RT-qPCR could be confirmed as being sourced from functional viral receptors (and by extension, viral capsids). Similarly, only 1 out of 16 GII HuNoV positives from RT-qPCR could be confirmed as being sourced from functional viral receptors. Somewhat unexpected were the 5 samples testing positive for GII HuNoV by ISC-qRT-PCR when they had tested a prior negative by RT-qPCR. Both ISC-RT-qPCR and RNase protection assays suggest that the majority of samples testing positive for HuNoV by RT-qPCR do not actually contain viable viruses, and instead contain free (and presumably damaged) viral RNA.

Technical Abstract: Human noroviruses (HuNoV) are the most common causes of foodborne illness, accounting for 50% of outbreaks and 46% of illnesses. In recent years, there has been a substantial increase in the proportion of HuNoV outbreaks associated with leafy vegetables. Produce can be contaminated by viral pathogens via exposure to contaminated environmental water from public waterways. We initiated a survey to determine the prevalence of HuNoV within several public watersheds of Coastal Central California in the vicinity of Salinas, CA, a major leafy green production region of the nation. The HuNoV genomic signal from environmental water samples was extracted, then quantitated using real-time (RT-qPCR). Tulane virus (TV) was incorporated as an internal control for the indication and quantitation of the RT-qPCR inhibition manifested by extraction impurities from the environmental water samples. From 250 RNA extracts, a total of 72 and 16 exhibited genotype I (GI) and GII HuNoV genomic signal, respectively. Being that the molecular-based RT-qPCR method can only indicate the presence of viral genome, the actual infectivity of the RT- qPCR-positive samples remains unknown. In this study, we tested these RT-qPCR-positive samples for capsid integrity (one indicator of potential HuNoV infectivity) by using an RNase protection assay, and for viral receptor binding ability (another indicator of potential HuNoV infectivity) by using an In-Situ Capture RT-qPCR (ISC-qRT-qPCR). TV was used as a surrogate for HuNoV to validate the method. We demonstrated that the ISC-qRT-PCR method did not capture viral genomic RNA prior-released by a heat-inactivation procedure. From 72 samples that tested positive for GI HuNoV genomic signal by RT-qPCR of RNA extracts, 20 (27.8%) samples tested positive for receptor binding ability when retested using the ISC-RT-qPCR method, suggesting that at least 72.2% of RT-qPCR positives samples were unlikely to be infectious. Similar results were observed with the RNase protection assay, as only 20.1% of the samples were RNase resistant indicating 80% RT-qPCR positives samples were unlikely to be infectious. From 16 samples that tested positive for GII HuNoV genomic signal by RT-PCR, only one tested positive for receptor binding ability when retested using the ISC-RT-qPCR method. However, five samples that had initially tested negative for GII HuNoV genomic signal by RT-PCR, tested positive for receptor binding ability and exhibited a genomic signal when retested using the ISC-RT-qPCR method. Overall, both assays suggest that the majority of samples that may test positive for HuNoV genomic signal by RT-qPCR of RNA extracts, were unlikely to be actually infectious, as these signals derive from free and presumably damaged and/or fragmented viral RNA, rather than intact and viable viral particles.