Again, sequence variation may also have a significant impact upon the accuracy of viral qPCR assays. Recently, we examined the impact of sequence mismatches in the primer target sites on qPCR by constructing a number of primers incorporating various mismatches with a target sequence on the BKV T-antigen gene. Essentially these primers were modifications of the primers used in the BKV-Tag-qPCR assay described previously (see Protocol 14.3). The results showed that as few as two mismatches in the 3' end of a single primer could introduce considerable error, underestimating viral load by up to 3 logs (Whiley and Sloots, 2005b). The overall impact this has on qPCR results is dependent on the conditions used in the assay and may hinge on the nucleotide composition of the primers, the PCR annealing temperature, the reaction mix composition and the specific nucleotides that are the subject of mismatch.
The concern of sequence variation and its impact on qPCR led us to examine a second BKV qPCR assay (BK-qLC) previously developed in our laboratory. This assay was based on the original polyomavirus hybridization probe assay, described above, using consensus primers and probes to detect both human polyomaviruses, JCV and BKV, followed by melting curve differentiation (Whiley et al., 2001). To make the assay specific for BKV, the consensus forward primer (PoL1s; Whiley et al., 2001) was replaced with a BKV specific forward primer. This new assay methodology (BK-qLC assay) was then validated against the BKV-Tag-qPCR assay (see Protocol 14.3) for the detection of BKV in urine samples obtained from transplant patients. The results showed good agreement between the two assays, with the detection of 19 BKV positive samples by both methods. In addition, there was good agreement between the Ct values and estimated viral loads for all but one of these 19 positive specimens. For 18 of these the average difference in Ct values between the two assays was only 1.0 cycles and differences in viral load were within one log. However, one positive specimen gave a Ct value of 30.5 in the BK-qLC and 20.7 in the BKV-Tag-qPCR assay. This difference of 9.8 cycles represented a difference in viral load of three logs and therefore was a significant discrepancy between the two assays. Sequencing the PCR product obtained in the BK-qLC from this specimen revealed a single base mismatch of the nucleotide targeted by the extreme 3' end of the forward primer (see Troubleshooting guide 14.4).
Overall, these results show that sequence variation in the primer binding sites can be the source of significant error in the estimation of viral load using qPCR even when other sources of error are standardized. This has considerable implications for viral quantification studies, particularly those targeting viruses that exhibit extensive sequence variation between strains and subtypes, such as the RNA viruses. Human metapneumovirus, respiratory syncytial virus and adenoviruses are just a few examples where sequence heterogeneity has been demonstrated. In fact, by performing realtime PCR assays targeting different genes in parallel for each of these viruses, we have found significant discrepancies between Ct values for positive specimens (data not published), suggesting that the impact of sequence variation on qPCR may be significant.
Overall, these results show that the impact of sequence variation is quantitative as well as qualitative. Interestingly, the BK-qLC performed well as a qualitative assay in the initial evaluation, demonstrating that a good qualitative PCR assay may not necessarily be suitable for use as a quantitative assay. Clearly, a different procedure must be followed to validate a quantitative assay to that employed for a qualitative assay. Such validation may involve sequencing the primer and probe target sites and determining if mismatches occur. Nevertheless, before accepting a viral qPCR protocol, we recommend performing two separate assays in parallel, validating one against the other, and comparing the Ct values to identify any discrepancies in performance. Not withstanding the above issues, we have found the BKV-Tag-qPCR assay to be reliable for quantifying BKV DNA and to date, have not found sequence variation in the primer or probe target sites of any local BKV strains.
Was this article helpful?