Site-directed mutagenesis and time-resolved fluorescence spectroscopy were used to evaluate the contributions of individual amino acid side chains to the binding of DNA primer-templates to the 3'-5' exonuclease site of the large proteolytic fragment (Klenow fragment) of DNA polymerase I. Mutations were introduced into side chains that have been shown crystallographically to be in close proximity to a DNA 3' terminus bound at the 3'-5' exonuclease site. The wild-type residues were replaced by alanine in each case. To assess the effects of the mutations on DNA binding, time-resolved fluorescence anisotropy measurements were performed on dansyl-labeled primer-templates bound to the mutant enzymes. In contrast to techniques that simply monitor the overall binding of proteins to DNA, the time-resolved fluorescence anisotropy technique was used to determine the fractional occupancies of the polymerase and 3'-5' exonuclease active sites of Klenow fragment. Equilibrium constants describing the partitioning of DNA between the two active sites were obtained for nine different mutant enzymes bound to both matched and mismatched DNA sequences. Mutations of Leu361 and Phe473 caused the largest effects, significantly destabilizing the binding of mismatched DNA substrates to the 3'-5' exonuclease site relative to DNA bound at the polymerase site, consistent with structural data showing that the side chains of these residues are involved in intimate hydrophobic interactions with the 3' terminal and penultimate bases of the primer strand [Beese, L., and Steitz, T. A. (1991) EMBO J. 10, 25-33]. Mutations of the His660 and Glu357 side chains also resulted in significant effects on the binding of mismatched DNA to the 3'-5' exonuclease site. Surprisingly, mutation of Tyr497 increased the partitioning of mismatched DNA into the 3'-5' exonuclease site, suggesting that the tyrosine side chain in the wildtype enzyme destabilizes substrate binding, despite crystallographic data showing that Tyr497 is H-bonded to the DNA substrate. The effects of mutating the amino acid side chains that serve as ligands to two divalent metal ions bound at the 3'-5' exonuclease site, designated A and B, indicated that metal A also helps to bind DNA to the 3'-5' exonuclease site. These results demonstrate that the time-resolved fluorescence anisotropy technique can be used to quantify the energetic contributions associated with each of the crystallographically defined DNA-protein contacts at the 3'-5' exonuclease site.

The Klenow fragment of Escherichia coli DNA polymerase I catalyzes template-directed synthesis of DNA and uses a separate 3'-5' exonuclease activity to edit misincorporated bases. The polymerase and exonuclease activities are contained in separate structural domains. In this study, nine Klenow fragment derivatives containing mutations within the polymerase domain were examined for their interaction with model primer-template duplexes. The partitioning of the DNA primer terminus between the polymerase and 3'-5' exonuclease active sites of the mutant proteins was assessed by time-resolved fluorescence anisotropy, utilizing a dansyl fluorophore attached to the DNA. Mutation of N845 or R668 disrupted favorable interactions between the Klenow fragment and a duplex containing a matched terminal base pair but had little effect when the terminus was mismatched. Thus, N845 and R668 are required for recognition of correct terminal base pairs in the DNA substrate. Mutation of N675, R835, R836, or R841 resulted in tighter polymerase site binding of DNA, suggesting that the side chains of these residues induce strain in the DNA and/or protein backbone. A double mutant (N675A/R841A) showed an even greater polymerase site partitioning than was displayed by either single mutation, indicating that such strain is additive. In both groups of mutant proteins, the ability to discriminate between duplexes containing matched or mismatched base pairs was impaired. In contrast, mutation of K758 or Q849 had no effect on partitioning relative to wild type, regardless of DNA mismatch character. These results demonstrate that DNA mismatch recognition is dependent on specific amino acid residues within the polymerase domain and is not governed solely by thermodynamic differences between correct and mismatched base pairs. Moreover, this study suggests a mechanism whereby the Klenow fragment is able to recognize polymerase errors following a misincorporation event, leading to their eventual removal by the 3'-5' exonuclease activity.

Frameshift mutagenesis occurs through the misalignment of primer and template strands during DNA synthesis and involves DNA intermediates that contain one or more extrahelical bases in either strand of the DNA substrate. To investigate whether these DNA structures are recognized by the proofreading apparatus of DNA polymerases, time-resolved fluorescence spectroscopy was used to examine the interaction between the Klenow fragment of DNA polymerase I and synthetic DNA primer-templates containing extrahelical bases at defined positions within the template strand. A dansyl probe attached to the DNA was used to measure the fractional occupancies of the polymerase and 3'-5' exonuclease sites of the enzyme for DNA substrates with and without the extrahelical bases. The presence of an extrahelical base at the first position from the primer 3' terminus increased the level of partitioning of the DNA substrates into the 3'-5' exonuclease site by 3-7-fold, relative to the perfectly base-paired primer-template, depending on the identity of the extrahelical base. The ability of different extrahelical bases to promote partitioning of DNA into the 3'-5' exonuclease site decreased in the following order: G > A approximately T > C. The results of partitioning measurements for DNA substrates containing a bulged adenine base at different positions within the template showed that an extrahelical base is recognized up to five bases from the primer 3' terminus. The largest effects were observed for the extrahelical base at the third or fourth positions from the primer terminus, which increased the level of partitioning of DNA into the 3'-5' exonuclease site by 8- and 18-fold, respectively, relative to that of the perfectly base-paired substrate. Steady-state fluorescence measurements of analogous primer-templates containing 2-aminopurine (AP) at the primer 3' terminus indicate that extrahelical bases increase the degree of terminus unwinding, especially when close to the terminus. In addition, steady-state kinetic measurements of removal of AP from the primer-templates indicate that the exonucleolytic cleavage activity of Klenow fragment is correlated with the increased level of partitioning of bulged DNA substrates to the 3'-5' exonuclease site relative to that of properly base-paired DNA. The results of this study indicate that misalignment of primer and template strands to generate an extrahelical base strongly promotes transfer of a DNA substrate to the 3'-5' exonuclease site, suggesting that the premutational intermediates in frameshift mutagenesis are subject to proofreading by the polymerase.