Pre-Steady-State Kinetic Characterization of an Antibiotic-Resistant Mutant of Staphylococcus aureus DNA Polymerase PolC

ABSTRACT

The emergence and spread of antibiotic resistance in bacterial pathogens are serious and ongoing threats to public health. Since chromosome replication is essential to cell growth and pathogenesis, the essential DNA polymerases in bacteria have long been targets of antimicrobial development, although none have yet advanced to the market. Here, we use transient-state kinetic methods to characterize the inhibition of the PolC replicative DNA polymerase from Staphylococcus aureus by 2-methoxyethyl-6-(3′-ethyl-4’-methylanilino)uracil (ME-EMAU), a member of the 6-anilinouracil compounds that specifically target PolC enzymes, which are found in low-GC content Gram-positive bacteria. We find that ME-EMAU binds to S. aureus PolC with a dissociation constant of 14 nM, more than 200-fold tighter than the previously reported inhibition constant, which was determined using steady-state kinetic methods. This tight binding is driven by a very slow off rate of 0.006 s⁻¹. We also characterized the kinetics of nucleotide incorporation by PolC containing a mutation of phenylalanine 1261 to leucine (F1261L). The F1261L mutation decreases ME-EMAU binding affinity by at least 3,500-fold but also decreases the maximal rate of nucleotide incorporation by 11.5-fold. This suggests that bacteria acquiring this mutation would be likely to replicate slowly and be unable to out-compete wild-type strains in the absence of inhibitors, reducing the likelihood of the resistant bacteria propagating and spreading resistance.

INTRODUCTION

The World Health Organization has placed antibiotic-resistant strains of Staphylococcus aureus on the list of priority pathogens for the research and development of new antibiotics (1). Antibiotics that directly target the replicative DNA polymerases, those that are primarily responsible for duplicating the bacterial chromosome, represent a novel therapeutic strategy. DNA polymerase III (Pol III) replicates the bulk of the bacterial genome and is essential for cell survival and pathogenesis (2). PolC is the catalytic subunit of Pol III in Gram-positive bacteria with low GC content, while DnaE serves that role in Gram-negative bacteria (3, 4). Both enzymes are members of the C family of DNA polymerases (5). C-family DNA polymerases are only found in bacteria, and structural studies (6–8) have shown that they are not related to any of the eukaryotic replicative DNA polymerases, reducing the chances that inhibitors targeting the C-family polymerases would have off-target effects in humans.

Specific inhibitors of PolC were first identified in the 1960s and are still being actively developed for the treatment of Gram-positive bacterial infections (9, 10). In fact, a phase 2 clinical trial of the PolC inhibitor ACX-362E (Ibezapolstat) is in progress for the treatment of Clostridium difficile (11). These compounds compete with deoxynucleotide triphosphate (dNTP) binding at the active site of PolC by forming hydrogen bonds with the nucleotide in the templating strand of DNA that is the next to be copied by the polymerase. The 3′-ethyl-4′-methylaniliouracils (EMAUs), an earlier class of compounds that work in the same way, are potent inhibitors of PolC and comprise two functional domains (Fig. 1A): the uracil domain forms hydrogen bonds that mimic the pairing of dGTP with dC while the aryl domain interacts with the polymerase (12). The 2-methoxyethyl derivative of EMAU (ME-EMAU) is bactericidal at a concentration of four-times the MIC of 8 to 16 μg/mL for the strains of S. aureus tested (13). Resistance to a related EMAU, N³-hydroxybutyl-6-(3′-ethyl-4′-methylanilino) uracil (HB-EMAU), arises in S. aureus at a single-step mutation frequency of 7.3 × 10⁻¹⁰ (14). PolC that was purified from resistant bacteria was >2,600-fold less sensitive to inhibition by HB-EMAU than PolC from sensitive S. aureus (14). The HB-EMAU-resistant bacteria were also resistant to other EMAU analogs (14). A single amino acid substitution, F1261L, was found when the polC gene from resistant bacteria was sequenced (14). Extensive structure-activity relationship studies indicate that changes in the substituent at position 6 of EMAU have little impact on the MIC (13). This is consistent with the model (Fig. 1A and B) showing that when EMAU is paired with a templating dC at the polymerase active site, ME and other chemical groups at this position project into the major groove side of the DNA duplex, where they would be unable to interact with the protein at residue 1261 (13).

FIG 1.

Mechanisms of ME-EMAU inhibition, resistance and DNA synthesis. (A) ME-EMAU and other anilinouracil derivatives compete with dGTP binding to cytosine in the template DNA by forming hydrogen bonds. This mimics a Watson-Crick base pair and sequesters the polymerase in an unproductive complex. (B) Active site of G. kaustophilus (Gka) PolC (PDB code 3F2B) showing the location of primer-template DNA and dNTP relative to residue F1271, which is equivalent to S. aureus (Sau) PolC residue F1261. (C) Minimal kinetic reaction pathway for DNA polymerases. Pol, DNA polymerase; DNA_n, unextended DNA; dNTP, deoxynucleotide triphosphate; DNA_n+1, DNA extended by one nucleotide; PPi, pyrophosphate.

During catalysis, enzymes adopt multiple states along a reaction pathway and each structural state represents a different drug target (15). To effectively target the different reaction intermediates, a comprehensive kinetic description of these states is essential and can be achieved using transient-state rather than steady-state kinetic measurements. We have recently determined the mechanism S. aureus PolC using transient-state kinetics (16, 17), the first detailed kinetic characterization of any C-family DNA polymerase. The structure of a thermophilic homolog of S. aureus PolC in a ternary complex that is poised for catalysis (Fig. 1B) provides detailed information about the contacts between the polymerase and its substrates, the primer-template DNA, and the incoming dNTP (8). Geobacillus kaustophilus PolC residue 1271 (equivalent to S. aureus PolC residue 1261) forms part of the dNTP-binding pocket and is universally conserved as phenylalanine in all C-family DNA polymerases (8). DNA polymerases share a two-metal-ion mechanism for phosphodiester bond formation (18), with the different activities of polymerases arising primarily because of differences in substrate binding affinity and the rates of each step in the reaction pathway (Fig. 1C).

Here, we use transient-state kinetics to characterize the interaction of ME-EMAU with S. aureus PolC and the properties of the F1261L PolC mutation that is predicted to confer resistance to ME-EMAU. We find that ME-EMAU binds to PolC much more tightly than predicted based on prior steady-state kinetic measurement and that the F1261L mutation significantly impairs the polymerase activity of PolC, thus reducing the fitness of the enzyme.

RESULTS

PolC has two catalytic domains: the polymerase domain that synthesizes DNA and a 3′- to 5′-exonuclease proofreading domain that increases the fidelity of DNA replication by removing mis-incorporated nucleotides. The exonuclease domain of both the wild-type and mutant enzymes was inactivated by four aspartate-to-alanine substitutions so that we could characterize the polymerase activity independently from any DNA degradation (17). For simplicity, we refer to these enzymes as PolC-wild type PolC-WT and PolC-F1261L, even though they both contain the exonuclease-inactivating mutations. In all assays, we have also included the S. aureus β-clamp (DnaN), a key component of the bacterial replisome that acts as a processivity factor by encircling the DNA duplex and binding to the C-terminal tail of PolC, which topologically tethers the polymerase to the circular chromosomal DNA. Even on a short, linear DNA substrate, the β-clamp increases the affinity of PolC for DNA by 4-fold (17). The DNA substrate used in this study (Fig. 2A) has a duplex length of 29 basepairs, a length that is sufficient to allow binding of both PolC and the β-clamp (8, 19), and a single-stranded template region of 39 nucleotides (Fig. 2A). This substrate differs from the one that we used in our prior study of PolC (17) by a single nucleotide; the FAM-labeled primer strand is one nucleotide shorter, such that the first templating base is a cytosine that will pair either with dGTP as the next-correct nucleotide or with the ME-EMAU inhibitor.

FIG 2.

The F1261L mutation in S. aureus PolC confers resistance to ME-EMAU. (A) The DNA substrate used in this study. The primer strand is labeled at the 5′ end with 6-carboxyfluorescein (FAM; *). (B and C) Primer extension reactions performed in the presence of increasing concentrations of ME-EMAU. Reactions in panel B contained 0.5, 5, or 50 μM ME-EMAU, 10 μM each nucleotde (dGTP, dTTP, and dATP) and were incubated for 60 s. Reactions in panel C contained 5, 50, or 500 nM ME-EMAU, 10 μM dGTP and were incubated for 1 s. All reactions contained 500 nM each PolC and DNA. Control reactions contained either DMSO or water (H₂O) in place of the inhibitor. Lanes with DNA alone show the unextended primer.

ME-EMAU binds tightly to PolC•DNA and dissociates slowly.

Initial experiments showed that PolC-WT is more sensitive to ME-EMAU than we had anticipated based on the previously reported inhibition constant (K_i) of ~3 μM (13). Various concentrations of ME-EMAU (0.5, 5, and 50 μM) were preincubated with PolC and DNA to allow the formation of the inhibited PolC•DNA•ME-EMAU complex. The first three correct nucleotides (dGTP, dTTP, and dATP) were then added to initiate DNA synthesis, and the reaction was allowed to proceed for 60 s. DNA synthesis by PolC-WT was substantially reduced in the presence of ME-EMAU, but all three concentrations tested showed comparable amounts of inhibition (Fig. 2B, left). In subsequent assays that used lower concentrations of ME-EMAU (5, 50, and 500 nM), a shorter reaction time (1 s), and a single nucleotide (dGTP), we observed concentration-dependent inhibition (Fig. 2C, top), suggesting that the K_i in our assays was in the low nanomolar range.

To determine the affinity of ME-EMAU for the PolC•DNA binary complex accurately, we turned to transient-state kinetics in which the concentrations of enzyme and substrate are comparable, thus facilitating the detection of reaction products from a single enzyme turnover (20, 21). PolC displays biphasic reaction kinetics, with an initial burst of rapid synthesis occurring in the first 10 to 20 ms of the reaction followed by a slower linear phase (16, 17). The amplitude of the burst phase corresponds to the amount of active, preassembled PolC•DNA complex that is capable of DNA synthesis when the reaction is initiated with dGTP. The slower linear phase after the burst results from subsequent catalytic cycles and indicates the presence of a slow step after catalysis that limits the rate of multiple enzyme turnovers.

The binding dissociation constant (K_D) for the primer-template DNA and dNTP substrates and the K_i for inhibitor were determined by varying the concentration of each reaction component independently of the others and measuring a time course that covers both phases of the reaction. PolC and DNA were incubated together in the presence or absence of ME-EMAU, before the addition of dGTP (Fig. 3A). In all reactions, the total concentration of PolC was 500 nM and the incubation time was varied from 2 to 80 ms. The amplitude of the fast phase of the reaction increased with increasing DNA concentration (50 to 1,600 nM; Fig. 3B) and decreased with increasing ME-EMAU concentration (2.5 to 200 nM; Fig. 3C). Both the amplitude and rate of the fast phase of the reaction increased with increasing dGTP concentration (0.5 to 75 μM; Fig. 3D). To measure the apparent rate of ME-EMAU binding, the reaction scheme was altered (Fig. 3E) such that PolC and DNA were preincubated together, and then a saturating concentration of ME-EMAU (25 μM) was added and allowed to interact with the PolC•DNA complex for various times (2 to 1,000 ms) before the addition of a saturating concentration of dGTP (250 μM) to initiate the reaction, which was allowed to proceed for 28 ms before quenching. The amount of DNA synthesis decreased as the incubation time with ME-EMAU increased (Fig. 3F). The results from these experiments were fit globally to the reaction pathway shown (Fig. 3G) using KinTek Explorer (22, 23) to determine a single set of kinetic constants that best fit all the data (Table 1). The two-dimensional confidence contour plots obtained using FitSpace Explorer (23) for the globally fit data (Fig. 4) show the dependence of the sum-square of error for each pair of parameters when all other parameters are varied to achieve the best fit. These plots illustrate the relationships between kinetic constants that are sometimes complex and may not be apparent from standard error. For example, the ratio of the binding (k₊₅) and dissociation (k₋₅) constants for ME-EMAU, which gives the K_D^EMAU, is more tightly constrained than the individual rate constants, as shown by the diagonal area of red in Fig. 4, indicating high confidence in these values (χ² > 0.93).

FIG 3.

Single-nucleotide incorporation kinetics of PolC-WT in the absence and presence of ME-EMAU. (A) Reaction scheme used in panels B to D to determine DNA binding affinity (B), ME-EMAU binding affinity (C), and nucleotide binding affinity (D). Polymerase (0.5 μM) was preincubated with DNA (50 to 1,600 nM in panel B; 50 nM in panels C and D) in the absence (B and C) or presence (D) of ME-EMAU (2.5 to 200 nM) before mixing with dGTP (100 μM in panel B and D; 0.5 to 75 μM in panel C) for various times (2 to 80 ms). (E) Reaction scheme used in (F) to determine the rate of ME-EMAU binding to the PolC•DNA complex. Polymerase (0.5 μM) was preincubated with DNA (50 nM) and mixed with ME-EMAU (25 μM) for various times (2 to 1,000 ms) before the addition of a saturating concentration of dGTP (125 μM) for 28 ms. All concentrations given are the final total concentrations in the reaction after mixing. Curves in panels B to D and F were derived by globally fitting the data to the reaction pathway shown in panel G (steps 1 to 6) using KinTek Explorer. Kinetic constants derived from the global fitting are shown in Table 1. Line colors indicate concentrations of the reaction component varied: (B) DNA (nM): red, 100; green, 200; blue, 400; yellow, 800; cyan, 1,600; (C) ME-EMAU (nM): red, 5; green, 10; blue, 20; yellow, 40; cyan, 80; magenta, 200; purple, 400; (D) dGTP (μM): red, 1; green, 2; blue, 5; yellow, 10; cyan, 20; magenta, 50; dark green, 150.

TABLE 1.

Pre-steady-state kinetic parameters for wild-type and F1261L Sau-PolC^a

	WT	F1261L	WT/F1261L
Parameters	Best fit ± SE (95% CI)	Best fit ± SE (95% CI)	Fold change
Globally fit parameters
DNA binding
k+1 (μM−1 s−1)	6.2 ± 0.1 (5.8–6.6)	3.7 ± 0.4 (2.7–5.8)	1.7 ± 0.2
k−1 (s−1)	1.1 (fixed)	0.42 ± 0.01 (0.39–0.45)	2.6 ± 0.1
Nucleotide binding
k+2 (μM−1 s−1)	100 (fixed)	100 (fixed)
k−2 (s−1)	1,440 ± 87 (1,160–1,860)	274 ± 104 (0–707)	5.2 ± 2.0
Polymerization rate
k+3 (s−1)	1,196 ± 35 (1,070–1,360)	104 ± 17 (60–168)	11.5 ± 1.9
k−3 (s−1)	122 ± 11 (75–166)	51 ± 18 (0.05–124)	2.4 ± 0.9
ME-EMAU binding
k+5 (μM−1 s−1)	0.42 ± 0.02 (0.35–0.47)	ND
k−5 (s−1)	0.006 ± 0.001 (0.005–0.007)	ND
Calculated parameters
KDDNA (μM)	0.18 ± 0.02	0.11 ± 0.01	1.6 ± 0.2
KDdGTP (μM)	14.4 ± 0.9	2.7 ± 1.0	5.2 ± 2.0
kpol (s−1)	1318 ± 37	155 ± 106	8.5 ± 5.9
kpol/KDdGTP(μM−1 s−1)	92 ± 6	57 ± 44	1.6 ± 1.3
KDME-EMAU (μM)	0.014 ± 0.003

^aBest fit values were determined by global fitting of the data to the model shown in Fig. 3 using KinTek Explorer and are shown ± standard error, with lower and upper limit for 95% confidence intervals shown in parentheses. Standard error (SE) and 95% confidence intervals (CI) were derived from one-dimensional confidence contour analysis using a χ² threshold of 0.93 for PolC-WT and 0.92 for PolC-F1261L. During the global fitting, the first-order rate constant for DNA dissociation (k₋₁) was fixed at 1.1 s⁻¹ for PolC-WT, based on previous experimental data (17); the second-order rate constant for nucleotide binding (k₊₂) was fixed at a diffusion-limited rate of 100 μM⁻¹s⁻¹ for both enzymes; and the forward and reverse rate constants for PPi release (k₊₄ and k₋₄, respectively) were fixed at 66.5 and 0.001 s⁻¹ for PolC-WT and 62.7 and 0.001 for PolC-F1261L. Calculated parameters were obtained using these equations: K_D^DNA = k₋₁/k₊₁; K_D^dGTP = k₋₂/k₊₂; k_pol = k₊₃ + k₋₃; K_D^ME-EMAU = k₋₅/k₊₅. ND, not determined.

FIG 4.

Two-dimensional confidence contour analysis for single-nucleotide incorporation by Sau-PolC-WT. The confidence contours for the global fit of the kinetic data shown in Fig. 3B, C, D, and F to the model shown in Fig. 3G (steps 1 to 5) were calculated using FitSpace Explorer (23). The heat map displays the dependence of the error for each pair of parameters when all other parameters are varied to achieve the best fit (red). During global fitting, the first-order rate constant for DNA dissociation (k₋₁) was fixed at 1.1 s⁻¹ based on previous experimental data (17), and the second-order rate constant for nucleotide binding (k₊₂) was fixed at a diffusion-limited rate of 100 μM⁻¹ s⁻¹; the forward and reverse rate constants for PPi release (k₊₄ and k₋₄, respectively) were fixed at 66.5 and 0.001 s⁻¹, based on previous experimental data (17). Confidence intervals and standard errors (shown in Table 1) were determined using a χ² threshold of 0.93, at the red/yellow boundary.

These experiments demonstrate that ME-EMAU binds to PolC-WT with a K_D^EMAU of 14 nM, >200-fold more tightly than determined from steady-state measurements (13), and dissociates very slowly, with an off rate (k_off^EMAU = k₋₅) of 0.006 s⁻¹. The substrates bind with a K_D^DNA of 180 nM and a K_D^dGTP of 14.4 μM, with ~49% of the enzyme capable of forming an active PolC•DNA complex that can catalyze product formation upon dGTP addition. Additionally, the release of pyrophosphate (PPi) is rate-limiting, with a k_off^PPi (k₄) of 66.5 s⁻¹ calculated from the global fitting, which is ~20-fold slower than the maximal polymerization rate (k_pol) of 1,318 s⁻¹, the sum of the forward and reverse reaction rates. All these parameters are comparable to those that we determined previously for the single-nucleotide incorporation of dTTP and dATP (17).

ME-EMAU-resistant mutant PolC-F1261L incorporates dNTP slowly.

PolC-F1261L was resistant to ME-EMAU, even at a concentration of 50 μM (Fig. 2B, right; Fig. 2C, bottom), indicating that the single mutation decreases inhibitor binding affinity by at least 3500-fold. To determine if the F1261L mutation altered the kinetic parameters of PolC, we used the same transient-state methods that we used for PolC-WT. In addition to measuring the binding affinities for DNA and dGTP (Fig. 5A and B), we determined the rate of DNA dissociation from PolC-F1261L using a modified reaction scheme (Fig. 5C) in which a preincubated mixture of enzyme and labeled DNA was mixed with a 500-fold excess of unlabeled trap DNA for various times (2 ms to 120 sec) before the addition of dGTP. Upon dGTP addition, only the enzyme still bound to the labeled DNA would produce visible reaction products, resulting in a decreased reaction amplitude as the mixing time increased (Fig. 5D). The data for PolC-F1261L were fit globally to the model shown in Fig. 3G, steps 1 to 3, to determine a single set of kinetic constants that best fit all the data (Table 1). The two-dimensional confidence contour analysis (Fig. 6) shows that the lower limits of dNTP dissociation (k₋₂) and the reverse rate of chemistry (k₋₃) are not well defined individually but vary proportionately with respect to each other and to the forward rate of chemistry (k₊₃).

FIG 5.

Single-nucleotide incorporation kinetics of PolC-F1261L. (A and B) The affinities of PolC-F1261L for (A) DNA and (B) dGTP were determined as for the wild-type enzyme (Fig. 3A to C). (C) Reaction scheme used in (D) to determine the rate of DNA dissociation from PolC-F1261L. PolC (0.5 μM) was preincubated with labeled DNA (50 nM), mixed first with unlabeled DNA (25 μM) for various times (0.08 to 9 s) before the addition of dGTP (1 mM) for 28 ms. All concentrations given are the final total concentrations in the reaction after mixing. Curves in A, B, and D were derived by globally fitting the data to the reaction pathway shown in Fig. 3G using KinTek Explorer. Kinetic constants derived from the global fitting are shown in Table 1. Line colors indicate concentrations of the reaction component varied: (A) DNA (nM): red, 100; green, 200; blue, 400; yellow, 800; cyan, 1600. (B) dGTP (μM): red, 0.5; blue, 1; green, 2.5; yellow, 5; cyan, 10; magenta, 25; dark green, 75.

FIG 6.

Two-dimensional confidence contour analysis for single-nucleotide incorporation by Sau-PolC-F1261L. The confidence contours for the global fit of the kinetic data shown in Fig. 4A, B, and D to the model shown in Fig. 3G (steps 1 to 3) were calculated using FitSpace Explorer (23). The heat map displays the dependence of the error for each pair of parameters when all other parameters are varied to achieve the best fit (red). The second-order rate constant for nucleotide binding (k₊₂) was fixed at a diffusion-limited rate of 100 μM⁻¹ s⁻¹ and the forward and reverse rate constants for PPi release (k₊₄ and k₋₄) were fixed at 62.7 and 0.001 s⁻¹, respectively. Confidence intervals and standard errors (shown in Table 1) were determined using a χ² threshold of 0.92, at the red/yellow boundary.

We found two substantial differences between the mutant and wild-type polymerases (Table 1). The mutant enzyme has a maximal polymerization rate that is 8.5-fold slower than wild-type (k_pol 155 s⁻¹ versus 1318 s⁻¹) but binds 5.2-fold more tightly to dGTP (K_D^dGTP 2.7 μM versus 14.4 μM). Taken together, these parameters result in a 1.6-fold decrease in nucleotide incorporation efficiency (k_pol/K_D^dGTP) by the mutant enzyme. Comparable to the wild-type enzyme, ~43% of PolC-F1261L was capable of forming an active complex with DNA, indicating that the mutation does not have a significant effect on the overall activity of the enzyme.

DISCUSSION

This transient-state kinetic analysis of S. aureus PolC demonstrates that ME-EMAU is a very potent inhibitor, binding 1,000-fold more tightly to the PolC•DNA complex than does dGTP and dissociating 220,000-fold more slowly than the maximal rate of polymerization. Steady-state assays are dominated by the rate-limiting step(s) in a reaction cycle. In the case of single nucleotide incorporation by DNA polymerases, the rate-limiting step is typically the dissociation of the product DNA from the postchemistry binary complex (Pol•DNA_n+1). However, during processive DNA synthesis in vivo, the product DNA does not dissociate after each round of nucleotide incorporation. Instead, the polymerase translocates along the DNA to catalyze the addition of subsequent nucleotides. Thus, steady-state assays interrogating the kinetics of DNA polymerases typically inform us about a catalytic step that is not physiologically relevant. Moreover, such multiple-turnover assays do not typically provide information about the dNTP binding and incorporation steps that are often of the most interest for understanding the mechanism of nucleotide incorporation and replication fidelity. This effect is clearly illustrated in the 200-fold difference between the 14 nM dissociation constant for ME-EMAU that we report here and the 2.8 μM inhibition constant reported previously (13). In the case of PolC, multiple-turnover reactions are limited by rates of both DNA and PPi dissociation (17). In the previous steady-state assays of ME-EMAU, reaction times were 10 min, which we now know is 5-fold longer than the half-life of ME-EMAU binding to PolC. Even though the dissociation rate is extremely slow (0.006 s⁻¹), the maximal rate of nucleotide incorporation (1,322 s⁻¹) is more than 5 orders of magnitude faster, causing the apparent inhibition of PolC to diminish over time, as is seen in the apparently weaker level of inhibition with a reaction time of 1 min (Fig. 2B, left, versus 1 s in Fig. 2C, top).

A key concern with any antibiotic is the potential for resistance mutations and the consequences of those mutations on the fitness of the resistant bacteria. The single amino acid substitution F1261L that confers ME-EMAU resistance on PolC decreases the maximal rate of nucleotide incorporation by 8.5-fold, but the mutation also increases the affinity of the polymerase for nucleotides by 5.3-fold. These changes partially counteract each other such that the overall enzyme efficiency only decreases by 1.6-fold. However, the concentration of dNTPs in bacterial cells is 50 to 150 μM during exponential growth (24, 25), which is near saturation for both wild-type and mutant PolC. Under these conditions, the differences in nucleotide incorporation rate would have the most impact on bacterial fitness and we would expect cells containing the PolC-F1261L mutation to be at a substantial growth disadvantage.

The power of combining detailed structural and mechanistic kinetic information has been exemplified by the progression of antiretroviral therapies used to treat HIV-1 infection and AIDS (26). The first cocrystal structure of Nevirapine bound to HIV-1 reverse transcriptase (27, 28) and the kinetic mechanisms of nucleotide incorporation and inhibition by the enzyme (29, 30) provided the foundation for developing more effective generations of nonnucleoside reverse transcriptase inhibitors, including ones that are active against mutants that are resistant to Nevirapine. While transient-state kinetic methods would not replace steady-state assays for inhibitor screening, a detailed understanding of the kinetic pathway can help in the design of high-throughput screens that target individual reaction steps and can aid in optimizing initial hits in a screen, for example by comparing the binding and dissociation rates during the analysis of structure-activity relationships.

Compared to viral and bacteriophage replicative DNA polymerases, mechanistic studies of cellular replicative polymerases have lagged behind. Among the cellular replicative polymerases, the bacterial C-family polymerases have been particularly challenging to study, largely because of weak binding to DNA and the unique catalytic cycle of the enzyme, but the available structures of several essential C-family polymerases (7, 8, 31–36) and our transient-state kinetic characterization of S. aureus PolC (16, 17, 24) provide the framework needed for a mechanistic structure-function approach to developing antibiotics targeting bacterial DNA replication.

MATERIALS AND METHODS

Materials.

DNA oligonucleotides were purchased from Integrated DNA Technologies Inc. (Coralville, IA, USA). Deoxynucleotides (dNTPs), ultrapure grade, and chromatography materials were purchased from GE Healthcare Biosciences (Schenectady, NY, USA).

Plasmids.

The polC gene from S. aureus strain COL was synthesized and codon usage was optimized for expression in Escherichia coli by GenScript USA Inc. (Piscataway, NJ, USA) and cloned into pET28a(+) using restriction sites NcoI and BamHI, The resulting plasmid contained an N-terminal deca-histidine tag followed by a TEV cleavage site and contained the following mutations: F1261L in the polymerase domain to confer resistance to ME-EMAU, together with D424A, E426A, D509A, and D568A in the exonuclease domain to prevent nucleic acid degradation. A plasmid containing the wild-type sequence at the active site was created using the Q5 site-directed mutagenesis kit from New England Biolabs (Ipswich, MA, USA), to reverse the resistance mutation at position 1261.

Protein expression and purification.

Plasmids containing the genes for wild-type and F1261L mutant PolC were transfected into Rosetta(DE3)pLysS E. coli, and cells were grown in Terrific Broth with 100 μg/mL ampicillin, 17 μg/mL chloramphenicol, 1 mL/L glycerol, 660 mM sorbitol, and 2.5 mM betaine at 37°C to an optical density at 600 nm of 0.6 to 0.7. The temperature was then decreased from 37°C to 17°C and cells were induced with 0.5 mM IPTG and harvested 18 h after induction. Protein was purified as described previously by chromatography over HisTrap HP charged with Nickel, Q-Sepharose, heparin, and Superdex 200 columns (16, 17). Purified protein was concentrated, flash-frozen in liquid nitrogen, and stored at −80°C in 50 mM Tris-Cl pH 7.5, 250 mM NaCl, 5 mM EDTA, 10% glycerol, and 5 mM DTT. PolC concentration was determined by UV absorbance at 280 nm using a theoretical extinction coefficient of 113,000 M⁻¹ cm⁻¹. β-clamp was purified as described previously (17).

DNA substrates.

Oligonucleotide sequences for the polymerase substrate are shown in Fig. 2A. The primer strand, labeled at the 5′ end with 6-FAM, was annealed with a 1.1 molar excess of the unlabeled template strand in 10 mM Tris-HCl pH 8.0 and 50 mM NaCl by denaturing at 95°C and slowly cooling to room temperature.

Polymerase assays.

Polymerase reactions contained final concentrations of 25 mM MES-Tris (pH 7.5), 25 mM NaCl, 8 mM MgCl₂, 2 mM DTT, 5% glycerol, 5 μM BSA, 0.5 μM total polymerase (PolC-WT or PolC-F161L), and 20 μM β-clamp monomer (10 μM dimer). All reactions were performed at room temperature and were terminated by the addition of an equal volume of quench buffer (250 mM EDTA and 50% formamide). Reactions with incubation times less than 1 min were performed as quench-flow assays using a KinTek RQF-3 apparatus (KinTek Corp., Austin, TX, USA). Reaction products were separated on denaturing gels containing 7 M urea, 1x TBE, and 11 or 17% acrylamide (19:1) and run at 55°C. All gels were imaged on a Typhoon RGB (GE Life Sciences) scanner with an excitation wavelength of 488 nm (blue laser) and an emission cutoff of 525 nm. Products were quantitated by measuring the intensity of the band corresponding to primer extension relative to the amount of total labeled DNA (both unextended and extended) using ImageQuant software.

For multiple-nucleotide incorporation assays (Fig. 2B), PolC (WT or F1261L) was preincubated with DNA (500 nM final concentration each) and β-clamp in the presence of 0.05 to 50 μM ME-EMAU (dissolved in 100% DMSO) for 30 min. The final amount of DMSO in the reaction mixtures was kept constant at 10%. Reactions were initiated by the addition of dGTP, dTTP, and dATP (the first three correct dNTPs for the substrate DNA) to a final concentration of 10 μM each and incubated for 60 s. For single-nucleotide incorporation assays (Fig. 2C, 3, and 4), the concentrations of primer-template DNA, dGTP, or ME-EMAU and the incubation time were varied depending on the experiment; final total concentrations during the extension reactions and reaction time are listed in the figure legends.

To determine the apparent rate of ME-EMAU binding to the PolC-WT•DNA complex (k_app), a double-mixing experiment was performed (Fig. 3E and F) in which preincubated PolC-WT, DNA and β-clamp were mixed with a saturating concentration (25 μM final) of ME-EMAU for various times (2 ms to 1 s) after which a saturating concentration (125 μM final) of dGTP was added and the reaction was allowed to proceed for 28 ms before quenching. A single exponential decay equation (equation 1) was fit to the concentration of extended product (Y) to determine k_app, where A₀ is the amount of product formed in the absence of inhibitor, t is the mixing time, and c is a constant.

$$Y\hspace{0.17em}=\hspace{0.17em}{A}_0\left(e^{-k_{\mathit{app}}•t}\right)\hspace{0.17em}+\hspace{0.17em}c$$ (1)

The rates of ME-EMAU binding and dissociation (k_on and k_off) were calculated from the apparent rate of binding (k_app) and the concentration of ME-EMAU using equations 2 and 3, where K_D,app^EMAU is the apparent dissociation constant for ME-EMAU, obtained from the experiment shown in Fig. 3C.

$$k_{\mathit{app}}^{\mathit{EMAU}}\hspace{0.17em}=\hspace{0.17em}{k}_{\mathit{on}}^{\mathit{EMAU}}\left(\left[\mathit{EMAU}\right]\hspace{0.17em}+\hspace{0.17em}{K}_{D,\mathit{app}}^{\mathit{EMAU}}\right)$$ (2)

$$K_{D,\mathit{app}}^{\mathit{EMAU}}\hspace{0.17em}=\hspace{0.17em}\frac{k_{\mathit{off}}^{\mathit{EMAU}}}{k_{\mathit{on}}^{\mathit{EMAU}}}$$ (3)

Global analysis of kinetic data.

Global fitting of the data by simulation was done using KinTek Explorer software version 10.0.200514 (22). The forward rate of dNTP binding was locked in at a diffusion-limited rate of 100 μM s⁻¹, and the reverse rate for PPi release was locked at a slow rate of 0.001 s⁻¹, since rebinding of PPi to PolCWT is negligible under the reaction conditions used. The dissociation rate of PolC-WT from DNA was fixed at 1.1 s⁻¹ based on previous kinetic characterization of the polymerase from a similar template where the primer strand was a single nucleotide longer. For PolC-F1261L, forward and reverse rates for PPi release were fixed at 62.7 s⁻¹ and 0.01 s⁻¹, respectively, according to our previous data demonstrating that PPi release is rate limiting (17). Confidence contour analyses were performed using the Fitspace function (23).

Accessibility of kinetic data.

KinTek Explorer files containing the kinetic data and confidence contour analyses are available for download from Zenodo (https://doi.org/10.5281/zenodo.7757200).