Detection of naturally occurring enteroviruses in waters using direct RT-PCR and integrated cell culture-RT-PCR.

Viruses detected by rapid molecular assays are not always infectious. In this study we compared enterovirus levels in natural waters using culture and reverse transcription-polymerase chain reaction (RT-PCR) techniques to determine whether molecular units of naturally occurring enteroviruses can be utilized to predict viral infectivity. Viruses were concentrated from 12 river water and effluent samples using 1 MDS filter-filtration and beef extract-elution. An integrated cell culture-RT-PCR (ICC-RT-PCR) was applied to the concentrates; and these waters contained up to 1.9 MPN of culturable (on BGM cells) viruses per litre (0.57 MPN/300 ml). Sample concentrates were also subjected to a direct 'molecular' approach using solvent-extraction, PEG-precipitation, and RNA-extraction before RT-PCR detection. The detection sensitivity of the direct RT-PCR was equivalent to 0.46 estimated (culturable) MPN/reaction, per 300 ml water. Two-thirds of the samples demonstrated consistent presence or absence of viruses by ICC-RT-PCR and direct RT-PCR. The direct RT-PCR approach resulted in over-estimation of naturally occurring infectious viruses as high as 91-fold in waters. Increased RT-PCR units may not reflect higher levels of culturable viruses in natural waters. The differences in virus levels detected by molecular and culture assays could be attributed to factors of volume of sample analyzed, different concentration schemes utilized that may affect the presence of residual inhibitors, and different stability exhibited by enterovirus strains/groups.