Spectroscopic analysis of polymerization and exonuclease proofreading by a high-fidelity DNA polymerase during translesion DNA synthesis

Abstract

High fidelity DNA polymerases maintain genomic fidelity through a series of kinetic steps that include nucleotide binding, conformational changes, phosphoryl transfer, polymerase translocation, and nucleotide excision. Developing a comprehensive understanding of how these steps are coordinated during correct and pro-mutagenic DNA synthesis has been hindered due to lack of spectroscopic nucleotides that function as efficient polymerase substrates. This report describes the application of a non-natural nucleotide designated 5-naphthyl-indole-2′-deoxyribose triphosphate which behaves as a fluorogenic substrate to monitor nucleotide incorporation and excision during the replication of normal DNA versus two distinct DNA lesions (cyclobutane thymine dimer and an abasic site). Transient fluorescence and rapid-chemical quench experiments demonstrate that the rate constants for nucleotide incorporation vary as a function of DNA lesion. These differences indicate that the non-natural nucleotide can function as a spectroscopic probe to distinguish between normal versus translesion DNA synthesis. Studies using wild-type DNA polymerase reveal the presence of a fluorescence recovery phase that corresponds to the formation of a pre-excision complex that precedes hydrolytic excision of the non-natural nucleotide. Rate constants for the formation of this pre-excision complex are dependent upon the DNA lesion, and this suggests that the mechanism of exonuclease proofreading is regulated by the nature of the formed mispair. Finally, spectroscopic evidence confirms that exonuclease proofreading competes with polymerase translocation. Collectively, this work provides the first demonstration for a non-natural nucleotide that functions as a spectroscopic probe to study the coordinated efforts of polymerization and exonuclease proofreading during correct and translesion DNA synthesis.

Introduction

DNA polymerases play essential roles in accurately replicating genomic material. One of the most faithful DNA polymerases studied to date is the bacteriophage T4 DNA polymerase (gp43) [1]. When replicating undamaged DNA, gp43 displays remarkably high fidelity by allowing only one dNMP misincorporation event per 10⁸ turnovers [2]. Base substitution frequencies for other DNA polymerases range significantly. For example, replicative DNA polymerases such as mammalian DNA polymerase delta and the Escherichia coli DNA polymerase III holoenzyme show low base-substitution frequencies of one misincorporation event per 10⁶ and ~ 10⁷ turnovers, respectively [3], [4]. However, other polymerases such as HIV-1 reverse transcriptase [5] and mouse myeloma DNA polymerase beta [6] are far less faithful when replicating DNA, displaying higher mutational frequencies of ~ 1 error every 5000 polymerization events.

Gp43, like most replicative polymerases, utilizes the multiplicative process outlined in Fig. 1A to maintain genomic fidelity [7]. This kinetic scheme encompasses three major control points for achieving fidelity: nucleobase selection and insertion (step A), removal of a misinserted nucleotide via exonuclease processing (step B), and enzyme dissociation from primer/templates that are misaligned due to mispairing (step C). In the case of gp43, the most important step for maintaining fidelity is the DNA polymerization step which encompasses several microscopic kinetic events including binding of dNTP to the Pol:DNA complex, a conformational change in the Pol:DNA complex, and phosphoryl transfer (Fig. 1B). Discrimination against misinsertion is proposed to be achieved through a combination of improper binding (step 2), inappropriate orientation (step 3), and/or diminished rate of phosphoryl transfer (step 4). These microscopic events collectively function to provide an error frequency of 1 misincorporation event per 10⁵–10⁶ turnovers [8].

Despite these kinetic barriers, genomic mistakes can occur if a nucleotide is misincorporated opposite a non-complementary partner. In these instances, the proofreading capability of the associated exonuclease domain reduces the error frequency by an additional 1000-fold [9]. While the activities of polymerization and exonuclease activities have been extensively studied during the replication of normal DNA [10], [11], [12], [13], [14], there is relatively little known on how these activities are coordinated during translesion DNA synthesis. The misreplication of damaged DNA is an important biomedical concern as translesion DNA synthesis can increase the probability of generating mutations. Indeed, defects in polymerase and exonuclease activities are associated with various genetic diseases including cancer [15], [16], [17]. As such, elucidating the biochemical mechanisms regulating polymerization and exonuclease activities is an important challenge in understanding how genomic fidelity is maintained and how defects in these activities give rise to various pathological conditions.

Accurately studying the concerted efforts of polymerase and exonuclease activities has been hindered due to the lack of chemical entities that can accurately probe these reactions. For example, most natural nucleotides display poor kinetic behavior when incorporated opposite miscoding and non-coding DNA lesions. In general, the catalytic efficiencies for incorporating natural nucleotides opposite most DNA lesions are > 1000-fold lower than those measured during the replication of undamaged DNA [18], [19], [20], [21]. In addition, natural nucleotides are spectroscopically inactive. There are, however, several examples of fluorescent analogs that can be used as surrogates for natural dNTPs [22], [23], [24], [25], [26], [27], [28]. In particular, Tor and co-workers have developed pyrrolo-dC as a fluorescent pyrimidine analog [22] whereas Saito and colleagues have developed fluorescent triazole deoxycytidine analogs [23], [24]. In addition, several pteridine analogs of guanine and adenine have been reported [25], [26]. Finally, Wilhelmsom and co-workers recently described a novel fluorescent triazole-adenine analog that was developed using “click” chemistry [27]. However, perhaps the most widely used analog for enzymological studies of DNA synthesis is 2-aminopurine deoxyribose triphosphate (2-APTP), a highly fluorescent analog of dATP [28], [29], [30]. Most biochemical studies using this nucleotide analog place the fluorogenic analog into single-stranded DNA [28], [29] or double-stranded DNA [31] as a way to measure conformational changes in nucleic acid. Since 2-APTP has a significantly higher intrinsic fluorescence while free in solution, experiments designed to directly measure its incorporation into DNA are technically challenging as the concentration of 2-APTP must be maintained low (< 5 μM). This low concentration is typically below the K_m value for insertion of the nucleotide opposite damaged DNA, and this kinetic feature hinders its use to monitor the kinetics of nucleotide incorporation and excision, particularly during translesion DNA synthesis.

These deficiencies mandate that discontinuous assays be used to study nucleotide incorporation and excision independently. This creates a serious complication as the inability to study both reactions simultaneously provides indirect evidence for kinetic events associated with the interaction between each active site. To circumvent these complications, this report describes the use of a non-natural nucleotide designated 5-naphthyl-indole-2′-deoxyribose triphosphate (5-NapITP) (Fig. 2A) as a spectroscopic probe to study both polymerization and exonuclease proofreading activities during translesion DNA synthesis. Since this non-natural nucleotide is highly conjugated, it is highly fluorescent while free in solution and this fluorescence becomes quenched upon incorporation into DNA [31]. In addition, the binding affinity of 5-NapITP during the replication of various DNA lesions including abasic sites and cyclobutane thymine dimers is relatively high (K_d ~ 10 μM) [32]. These features collectively suggest that 5-NapITP could function as a common and spectroscopically-active analog to study the coordination of nucleotide incorporation and excision during translesion DNA synthesis. The analyses provided here demonstrate the existence of distinct nuances in the mechanism by which structurally diverse DNA lesions are replicated. In addition, our studies reveal the presence of a fluorescence recovery phase that corresponds to the formation of a pre-excision complex that precedes hydrolytic excision of the non-natural nucleotide. The rate constant for forming this complex is lesion-dependent, indicating that exonuclease proofreading is regulated by the physical composition and nature of the DNA lesion. Taken together, this work provides the first demonstration of a unique non-natural nucleotide that can be used for real-time spectroscopic analyses to probe the coordinated efforts of polymerization and exonuclease proofreading.