Introduction: We aimed to design a real-time reverse-transcriptase-PCR (rRT-PCR), high-resolution melting (HRM) assay to detect the H275Y mutation that confers oseltamivir resistance in influenza A/H1N1 2009 viruses.
Findings: A novel strategy of amplifying a single base pair, the relevant SNP at position 823 of the neuraminidase gene, was chosen to maintain specificity of the assay. Wildtype and mutant virus were differentiated when using known reference samples of cell-cultured virus. However, when dilutions of these reference samples were assayed, amplification of non-specific primer-dimer was evident and affected the overall melting temperature (T(m)) of the amplified products. Due to primer-dimer appearance at >30 cycles we found that if the cycle threshold (C(T)) for a dilution was >30, the HRM assay did not consistently discriminate mutant from wildtype. Where the C(T) was <30 we noted an inverse relationship between C(T) and T(m) and fitted quadratic curves allowed the discrimination of wildtype, mutant and 30∶70 mutant∶wildtype virus mixtures. We compared the C(T) values for a TaqMan H1N1 09 detection assay with those for the HRM assay using 59 clinical samples and demonstrated that samples with a TaqMan detection assay C(T)>32.98 would have an H275Y assay C(T)>30. Analysis of the TaqMan C(T) values for 609 consecutive clinical samples predicted that 207 (34%) of the samples would result in an HRM assay C(T)>30 and therefore not be amenable to the HRM assay.
Conclusions: The use of single base pair PCR and HRM can be useful for specifically interrogating SNPs. When applied to H1N1 09, the constraints this placed on primer design resulted in amplification of primer-dimer products. The impact primer-dimer had on HRM curves was adjusted for by plotting T(m) against C(T). Although less sensitive than TaqMan assays, the HRM assay can rapidly, and at low cost, screen samples with moderate viral concentrations.