Inferred hepatitis C virus quasispecies diversity is influenced by choice of DNA polymerase in reverse...

The hepatitis C virus (HCV) is known to circulate in vivo as a quasispecies, a population of closely related, but genetically nonidentical virions. HCV reverse transcriptase (RT)-(nested) polymerase chain reaction (PCR) strategies are used to study quasispecies diversity at certain important viral genetic loci, predominantly at hypervariable region 1 (HVR 1) of the E2 envelope gene, and the interferon sensitivity determining region (ISDR) of the nonstructural 5a (NS5a) gene. We have found that the choice of DNA polymerase employed in viral PCR has effects on the inferred viral diversity at two distinct loci on the HCV genome. Nested HVR 1 and ISDR PCR was performed with both proofreading (Pwo) and nonproofreading (Taq) DNA polymerases on identical cDNA derived from three separate HCV-positive sera. Amplicons were cloned and sequences determined for 18-20 individual clones per sample. Quasispecies diversity determined from HVR 1 and ISDR PCR products showed that there was a marked effect on the inferred diversity depending on which DNA polymerase was employed in the PCR. The deduced amino acid sequences of the major variants within each specimen were identical for both Taq and Pwo DNA polymerase-mediated PCRs. However, a greater number of minor variants were observed in the Taq-generated amplicons, 80% of which were not observed in the Pwo-generated amplicons. Primer editing in the Pwo-generated amplicons was observed in 19% (20/104) of clones examined. Single-strand conformational polymorphism analysis of multiple replicates of each amplicon revealed good intra-PCR reproducibility in terms of genetic heterogeneity, and that as such the observations were not due to poor PCR reproducibility. The use of nonproofreading DNA polymerases to assess viral diversity can yield an incorrect quasispecies spectrum and affect RT-PCR assay performance. The contribution of Taq-induced errors and lack of adaptability of primers to potentially heterologous template-binding sites indicate that proofreading DNA polymerases should be the enzyme of choice in these systems.