Fidelity of eukaryotic and archaeal family-B DNA polymerases

Abstract

DNA polymerases are essential for replication, recombination and repair of DNA. However these enzymes have multiple applications in biotechnology. Since PCR has been developed, thermostable DNA polymerases have become important for numerous PCR based applications. Currently these enzymes are used routinely in laboratories all over the world. The fidelity of DNA polymerases is a key feature for PCR. We have developed a fidelity assay, based on a gapped plasmid template containing the lacZα gene reporter, allowing easy and straightforward measurement of the accuracy of DNA synthesis by DNA polymerases in vitro. Previous studies on the family-B DNA polymerase from Pyrococcus furiosus demonstrated that fidelity is controlled by D473, an amino acid located in the loop of the fingers domain. It was observed that the mutation D473G had a strong error-prone phenotype and Pfu-Pol D473G can be successfully used for random mutagenesis. To test if eukaryotic family-B DNA polymerases use the same aspartic acid residue to control fidelity we prepared the D799G mutant of a proteolytic fragment of polymerase epsilon from Saccharomyces cerevisiae. Unfortunately we did not observe the expected modulation of the fidelity of DNA synthesis for the D799G polymerase epsilon variant. Overexpression of the multi-subunit family-B DNA polymerases from Saccharomyces cerevisiae was found to be extremely demanding. Therefore, we decided to modify the thermostable family-B DNA polymerase from Thermococcus gorgonarius to obtain variants containing the loop region of the fingers domain from family-B DNA polymerases of Saccharomyces cerevisiae. We have observed no change in DNA synthesis accuracy when the loop region was transferred from high fidelity yeast replicative polymerase delta. However when the loop region was transferred from Saccharomyces cerevisiae error-prone family-B DNA polymerase zeta we observed a strong error-prone phenotype, in some instances, loop swapping with polymerase zeta is complicated by alignment ambiguity, so several variants were prepared. The primary sequence alignment of the fingers domain of eukaryotic polymerases zeta suggests no strong consensus within the loop region. Therefore, we decided to replace the major part of the fingers domain of the family-B polymerase from Thermococcus gorgonarius with the equivalent functional module from Saccharomyces cerevisiae polymerase zeta. The main aim of such a rearrangement was to test if the module from the error-prone DNA polymerase zeta has the potential to decrease fidelity. The chimeric polymerase variant was indeed found to be a very inaccurate DNA polymerase. To our surprise we also discovered that the polymerase variant possesses reverse transcriptase activity. Several further modifications allowed us to significantly improve reverse transcriptase activity of the chimeric polymerase variant.