Rapid Genotyping of Swine Influenza Viruses

The emergence of pandemic (H1N1) 2009 virus highlighted the need for enhanced surveillance of swine influenza viruses. We used real-time reverse–transcription PCR–based genotyping and found that this rapid and simple genotyping method may identify reassortants derived from viruses of Eurasian avian-like, triple reassortant-like, and pandemic (H1N1) 2009 virus lineages.

C o-infection of infl uenza A viruses enables viral gene reassortments, thereby generating progeny viruses with novel genotypes. Such reassortants may pose a serious public health threat, as exemplifi ed by the emergence of pandemic infl uenza (H1N1) in 2009 (1). Transmission of pandemic (H1N1) 2009 virus from humans to pigs has been reported (2)(3)(4)(5). We recently identifi ed a reassortment between pandemic (H1N1) 2009 virus and swine infl uenza viruses in pigs (6). These results emphasize the potential role of pigs as a mixing vessel for infl uenza viruses and the need for screening tests that can identify major reassortment events in pigs.
We previously developed 8 monoplex SYBR greenbased quantitative reverse transcription-PCRs to detect all 8 gene segments derived from the pandemic (H1N1) 2009 (5). Using these PCRs, we identifi ed swine viruses of atypical genotypes. However, with the exception of the HA-specifi c assay, the meltingcurve signals of pandemic (H1N1) 2009 virus may be indistinguishable from the positive signals generated from its sister clade as indicated above. To differentiate between these closely related groups of viruses, we further optimized these assays by adding sequence-specifi c hydrolysis probes in the SYBR green assays.

The Study
For this study, all SYBR green assays were modifi ed from the previously described assays (5), with the exception of the reverse primers for the newly designed PB1 and NS segments ( Table 1). The subtype H1N1 swine infl uenza viruses isolated in Hong Kong during the past few years were mainly derived from the Eurasian avian-like swine lineage (6,7). To generate more precise genotyping data for our ongoing surveillance, the NA segment-specifi c assay was specifi cally designed to react with the pandemic (H1N1) 2009 virus and a portion of Eurasian avian-like swine viruses that are circulating in southeastern China (online Technical Appendix Figure 1, www.cdc.gov/EID/ content/17/4/691-Techapp.pdf). To avoid overlapping the emission spectrum of SYBR green, we labeled all pandemic (H1N1) 2009 virus-specifi c hydrolysis probes (Integrated DNA Technologies, Inc., Coralville, IA, USA) with cyanine 5 (Cy5) and Black Hole Quencher-2 dyes at their 5′ and 3′ ends, respectively (Table 1). To enable use of short oligonucleotide sequences without compromising the annealing temperature of these probes, we modifi ed the probes with locked nucleic acids (7). RNA extraction and complimentary DNA synthesis were identical to the protocols described (5,8). One microliter of 10-fold diluted complimentary DNA sample was amplifi ed in a 20-μL reaction containing 10 μL of Fast SYBR Green Master Mix (Applied Biosystems, Foster City, CA, USA) and the corresponding primer probe set (0.5 μmol/L each). All reactions were optimized and performed simultaneously in a 7500 Sequence Detection System (Applied Biosystems) with the following conditions: 20 s at 95°C, followed by 30 cycles of 95°C for 3 s and 62°C for 30 s. SYBR green and Cy5 signals from the same reaction were captured simultaneously at the end of each amplifi cation cycle. The expected PCR results of virus segment derived from different swine viral lineages are shown in Table 2.
The dissociation kinetics of PCR amplicons were studied by a melting curve analysis at the end of the PCR (60°C-95°C; temperature increment 0.1°C/s). We also tested various probe and SYBR green concentrations under different PCR conditions. The condition described above gave the most robust and consistent DNA amplifi cation (data not shown). We tested 31 human pandemic (H1N1) 2009 and 63 human seasonal infl uenza viruses (33 subtype H1N1, 30 subtype H3N2) as controls. As expected, all human pandemic infl uenza viruses were double positive (i.e., positive with SYBR green and Cy5) and all seasonal infl uenza samples were double negative in all 8 assays.
To evaluate the sensitivity of the assays, we tested serial diluted plasmid DNA of the corresponding segments of infl uenza A/California/4/2009 virus as a standard. The fl uorescent signals generated from the SYBR green reporter dye in all assays were highly similar to those previously reported (5), and the modifi ed assays had a linear dynamic detection range from 10 2 to 10 8 copies/reaction (online Technical Appendix Figure 2). As expected, the threshold cycle values deduced from the Cy5 reporter signal were generally higher than those from the SYBR green reporter (online Technical Appendix Figure 2) (9). This fi nding can be partly explained by the nature of these 2 kinds of real-time PCR chemistries: a single Cy5 fl uorophore of the hydrolysis probe was released from quenching for each amplicon synthesized while multiple SYBR green dyes bound to a single amplicon (10). After 35 PCR amplifi cation cycles, the linear dynamic detection range of Cy5 signals generated from these reactions was 10 2 to 10 8 copies/reaction (data not shown). However, to avoid nonspecifi c SYBR green signals, we purposely limited the number of amplifi cation cycles to 30.
Using these assays, we tested 41 swine virus isolates collected during January 2009-January 2010. In all 8 reactions, 10 pandemic (H1N1) 2009 virus samples transmitted from humans to pigs (6) were double positive (online Technical Appendix Figure 1, pink). In these assays, gene segments of another 31 swine isolates were either SYBR green positive/Cy5 negative (online Technical Appendix Figure 1, yellow) or double negative (online Technical Appendix Figure 1, green) Cy5-CGCTACCTTTCTGACAT/BHQ2/ *PB, polymerase basic protein; PA, polymerase acidic protein; HA, hemagglutinin; NP, nucleocapsid protein; NA, neuraminidase; M, matrix protein; NS, nonstructural protein; Cy5, cyanine 5; BHQ2, Black Hole Quencher 2. †Number represents nucleotide position of the first base in the target sequence (cRNA sense). ‡Locked nucleic acid-modified bases are underlined. §Primers adapted from the assays as previously described (5,6).   Figure; other data not shown). All genotyping results of the studied viruses were consistent with results of previous phylogenetic analyses (5,6), indicating that our modifi ed probes and SYBR green assays can provide more accurate genotyping results. With these genotyping data, viruses with atypical positive signal patterns might suggest a novel viral reassortment event and can be highlighted for investigation with sequencing-based methods.
To demonstrate the potential use of these assays in studying swine viruses circulating in other geographic locations, we tested 7 recent swine isolates (1 pandemic infl uenza subtype H1N1, 4 subtype H1N2, and 2 subtype H3N2) collected in the United States. Genotyping results agreed 100% with data deduced from sequence analyses (online Technical Appendix Table 1). We also analyzed all 436 contemporary (2008-2010) US swine virus segments available from the National Center for Biotechnology Information infl uenza virus sequence database. On the basis of the in silico analysis of sequences targeted by our primers and probes, 95% of the sequences (n = 413) are predicted to yield the expected results (online Technical Appendix Table 2).

Conclusions
The emergence of pandemic (H1N1) 2009 has highlighted the need for global systematic infl uenza surveillance in swine. Our results demonstrated that the addition of locked nucleic acid hydrolysis probes specifi c for pandemic (H1N1) 2009 virus into previously established SYBR green assays can help differentiate segments of pandemic (H1N1) 2009, Eurasian avianlike, and triple reassortant virus lineages. These assays might provide a rapid and simple genotyping method for identifying viruses that need to be fully genetically sequenced and characterized. They may also help provide better understanding of the viral reassortment events and viral dynamics in pigs. Although at present, genes derived from human seasonal viruses cannot be characterized with our modifi ed assays, the performance of our assays warrants similar investigations for genotyping human infl uenza viruses.