Development of a Recombinase Polymerase Amplification Assay for Detection of Epidemic Human Noroviruses

Moore, Matthew D.; Jaykus, Lee-Ann

Each 50 μ l reaction contained 2.1 pmol forward and reverse primer, 0.6 pmol probe, 29.5 μ l proprietary rehydration buffer, 10 μ l template, and nuclease-free water to 47.5 μ l. The reaction mixture was then added to rehydrate TwistAmp exo RT lyophilized enzyme pellets, and 2.5 μ l of 280 mM magnesium acetate was added to the tops of reaction tube lids. The magnesium acetate droplets were then spun down using a minifuge and a timer immediately started. Reactions were quickly transferred to a Bio-Rad CFX Thermal Cycler (Bio-Rad, Hercules, CA) set to 40 °C with cycles read every 30 sec. Five min after the reactions were started, the cycler was paused, reactions mixed, spun down, and placed back on the thermal cycler to continue cycling. The total time to produce signal was calculated by the following equation (1) Evaluation of purified RNA and heated stool samples using RT-RPA and RT-qPCR. Twenty percent suspensions of norovirus GII.4 New Orleans-confirmed clinical samples from 12 patients were serially diluted and the RNA released by heating in a Bio-Rad T100 Thermal Cycler (Bio-Rad, Hercules, CA) at 99 °C for 5 min. Tubes were cooled on ice, mixed and briefly centrifuged. These heated stool dilutions were used directly as a template in RT-RPA reactions as described above, and in RT-qPCR reactions using the JJV2F-COG2R-Ring2-TP Scientific RepoRts | 7:40244 | DOI: 10.1038/srep40244 primer probe set (Table 1 ). In addition, NucliSENS ® easyMAG RNA extracts from each clinical sample were serially diluted and used in RT-RPA and RT-qPCR. The degree of correlation between RT-RPA time to detection (in min) and log 10 genomic copy input (by standard curve) was determined by linear regression using Microsoft Excel 2013 ( Figure S2 ). For evaluating the RT-RPA assay analytical sensitivity, serial dilutions of purified RNA from three selected stool samples (samples 29, 47 , and 58) were tested using as described above for a total of 8 replicates and a probit regression was performed for each using the JMP 12 software package (SAS, Cary, NC). The limit of detection and standard deviation was calculated by averaging the three values produced by the regressions. Each value reflected the predicted LGC for which 95% of tests would produce a positive result. All reaction sets included no template controls to confirm lack of contamination. Exclusivity testing and sequence analysis. The specificity was evaluated by analyzing multiple relevant enteric virus and bacterial strains for cross reactivity with the RT-RPA assay (Table 3) . Purified genomic nucleic acid was serially diluted, and 10 −2 and 10 −3 dilutions used directly as templates in RT-RPA. To analyze the degree of sequence difference between GII.4 New Orleans, for which the primers were designed, and two other human norovirus strains (GII.3 and GII.4 Sydney), sequences were aligned by performing Clustal W analysis using the MEGA 6 Software suite 39 .