Long PCR Amplification of Large Fragments of Viral Genomes: A Technical Overview

The polymerase chain reaction (PCR) has become an essential and ubiquitous tool for biological research and laboratory diagnostic applications. Until recently, reliable and sensitive amplification of large templates (several kb) was difficult to achieve. However, in 1994, an important breakthrough was reported by Barnes (1 ). He hypothesized that a major obstacle to long PCR was the Taq DNA polymerase error rate, which causes mismatches that make elongation very inefficient. Many other thermostable DNA polymerases have a 3′ to 5′ exonuclease “proofreading” activity and a higher fidelity. However, the use of these polymerases alone does not reliably achieve long PCR, presumably because of excessive degradation of primers by the exonuclease activity (1 ). The processivity of the enzyme may also be a factor. Of note, the 3′ to 5′ exonuclease activity alone is not a guarantee of high fidelity: Fidelity also depends on the degree of discrimination against misinsertion, the mismatch extension rate, and the rate of shuttling between polymerizing and proofreading modes (2 ). The breakthrough reported by Barnes consisted in performing PCR with a mixture of two DNA polymerases: a major component consisting of a highly processive DNA polymerase and a minor component consisting of a DNA polymerase with a 3′ to 5′ exonuclease “proofreading” activity. With such enzyme mixes, reliable amplification of templates up to 35 kb in length was achieved (1 ). The greater fidelity of long PCR enzyme mixes, relative to Taq , has been demonstrated (1 ,2 ). Other modifications contribute to making long PCR possible, including optimization of the buffer and the thermal cycling conditions.