Clam (Corbicula fluminea) as a potential sentinel of human norovirus contamination in freshwater
The purpose of this study was to evaluate and validate the use of the clam Corbicula fluminea as a sentinel of human noroviruses (HuNoV) contamination in freshwater. The first specific aim was to develop a new method to extract HuNoV RNA from contaminated bivalves (e.g. oysters, clams) that would be much faster than existing methods. The procedure developed includes an initial total RNA extraction using TRI Reagent, followed by HuNoV RNA concentration and purification using biotinylated probe-capture technology. HuNoV RNA is finally detected by real-time RT-PCR. Using bivalve homogenates spiked with HuNoV, 100 PCR detection units of the virus was detectable. Compared to published methods that require an initial virus purification step, the new method is much faster to complete. Approximately 3 h are needed to purify NoV RNA using the new method compared to at least 8 h using conventional methods. Coupled with real-time RT-PCR, the new method can detect HuNoV in contaminated bivalves within 8 h. The detection limit of the method was 10 -100 PDU of HuNoV. In addition, the method was successfully applied for HuNoV detection in live artificially-contaminated oysters, wild oysters, and also for murine norovius (MNV-1) and HuNoV detection in clams. The second specific aim was to evaluate the ability of C. fluminea to bioaccumulate and depurate HuNoV using MNV-1 as a surrogate of HuNoV. Clams were exposed to MNV-1 in 10 L artificial pond water for 6 h, 1, 2, and 3 d in an environmental chamber. Depuration experiments were carried out in 80 L artificial pond water for 0, 1, 4, 7, 10, and 15 d at 10°C and 20°C. MNV-1 was detectable after 6 h and 1 d exposure in clams exposed to virus concentrations of 10 6 PFU·L-1 and 104 PFU·L-1 , respectively. The amount of bioaccumulated MNV-1 increased as the exposure period increased from 6 h to 3 d. The lowest virus concentration at which exposed clams were PCR-positive was 102 PFU·L-1 after 2 d exposure at 20°C. Clams bioaccumulated MNV-1 more quickly at 20°C than at 10°C (p < 0.05). The virus was persistently detected in contaminated clams during depuration at both 10°C and 20°C. Depuration occurred significantly more quickly at 20°C than at 10°C (p < 0.05). The results indicate that the clam is likely to be useful as a sentinel for detecting NoV contamination in freshwater. The third specific aim was to determine whether C. fluminea is effective as a sentinel of HuNoV contamination in natural freshwater. Clams were collected from Lake Serene in Hattiesburg, Mississippi where HuNoV has never been detected from Oct 2010 to Jul 2011, and translocated to 9 sites at 4 freshwater creeks in Gulfport and Long Beach where the creek water flowed into Mississippi Sound. HuNoV RNA was isolated from clams (n = 588) using the biotinylated probe hybridization method mentioned above and detected by qRT-PCR. Correct identity of the virus was accomplished by sequencing some of the amplified RT-PCR products (HuNoV capsid N-terminal/shell domain). qRT-PCR results showed that HuNoV GI and GII were detectable in the translocated C. fluminea mainly during the warmer months (Apr to Jul and Oct), but not during the colder months (Dec to Mar). Based on sequence comparisons, the HuNoV detected in translocated clams were classified into GI/17 and GII/4, respectively. Statistical analysis using binary logistic regression showed that water temperature and turbidity (p = 0.026 and p = 0.038, respectively), but not pH, salinity, or current velocity (p = 0.476, p = 0.425, and p = 0.174, respectively), were significant factors affecting HuNoV presence/absence in clams in freshwater creeks. In conclusion, it was found that the freshwater clam C. fluminea can be translocated and serve as an effective sentinel of HuNoV contamination in freshwater of low turbidity during warm months.