Reproducibility, fidelity, and discriminant validity of linear RNA amplification for microarray-based identification of major human enteric viruses.

Human enteric viruses are inherently a group of viruses that confer similar or overlapping clinical symptoms and pose a challenge for correct etiological diagnosis. DNA microarray technology has emerged to be of major interest to detect broad range of viral pathogens including enteric viruses. However, this approach requires a relative large amount of target nucleic acid for hybridization analysis. This feature limits its further applicability. To address this challenge, we evaluated a novel single primer linear isothermal amplification (Ribo-SPIA) procedure for preparation of single-stranded cDNA (sscDNA) from minute amount of starting RNA for microarray-based simultaneous detection and identification of three major human enteric viruses including hepatitis A virus, norovirus, and coxsachievirus B2. We performed a series of tests using different amounts of input RNA ranging from 30 ng to 55 pg to assess amplification yield, reproducibility, analytical sensitivity, and fidelity. We demonstrated that as little as 55 pg of viral RNA could produce adequate material by Ribo-SPIA to enable successful identification by microarray analysis without compromising detection specificity. Pairwise comparison of technical replicates hybridized to the microarrays by regression analysis showed excellent reproducibility in the appropriate sensitivity range. We also showed that the use of sscDNA as labeled targets offered increased microarray detection accuracy over complementary RNA generated by traditional T7 in vitro transcription amplification method.