Exploring the role of large conformational changes in the fidelity of DNA polymerase β

Abstract

The relationships between the conformational landscape, nucleotide insertion catalysis and fidelity of DNA polymerase β are explored by means of computational simulations. The simulations indicate that the transition states for incorporation of right (R) and wrong (W) nucleotides reside in substantially different protein conformations. The protein conformational changes that reproduce the experimentally observed fidelity are significantly larger than the small rearrangements that usually accompany motions from the reactant state to the transition state in common enzymatic reactions. Once substrate binding has occurred, different constraints imposed on the transition states for insertion of R and W nucleotides render it highly unlikely that both transition states can occur in the same closed structure, because the predicted fidelity would then be many orders of magnitude too large. Since the conformational changes reduce the transition state energy of W incorporation drastically they decrease fidelity rather than increase it. Overall, a better agreement with experimental data is attained when the R is incorporated through a transition state in a closed conformation and W is incorporated through a transition state in one or perhaps several partially open conformations. The generation of free energy surfaces for R and W also allow us to analyze proposals about the relationship between induced fit and fidelity.

INTRODUCTION

The reproduction of all forms of life depends on the accurate replication of the genome, which is facilitated by DNA polymerases.¹ These enzymes increase the rate of phosphodiester bond formation by many orders of magnitude when compared with the corresponding reaction in water.² DNA polymerases catalyze the reaction not for a single substrate, like most other enzymes, but for many structurally distinct substrates. A correct dNTP substrate is often selected for insertion by the polymerase depending on the identity of the templating base so that Watson-Crick base pairs are formed. The fidelity of DNA replication is controlled in part within the polymerase active site where phosphodiester bond catalysis occurs and by the binding site for the incoming nucleotide base paired opposite a template nucleotide. The rate of incorporation of an incoming wrong nucleotide (W) is drastically slower than the corresponding rate for the right nucleotide (R) for DNA polymerases that exhibit high replication fidelity (for review see Refs. ^3–6). Since the rate of W insertion is approximately the same for all DNA polymerases irrespective of their fidelities, enzyme selectivity is primarily determined by the rate of R insertion.^6,7

Despite significant experimental advances in studies of the fidelity of various DNA polymerases (e.g., Refs. ^3–6 and ^8–11), we still do not have a clear quantitative picture of the energetics of this process. Progress has been made in recent theoretical studies, which explored the contributions to fidelity from the binding and chemical steps,^12–17 as well as on the possible roles of conformational changes¹⁸ and protein motions.¹⁹ However, progress in these key areas is still at a relatively early stage.

Additional progress in studies of the fidelity of DNA polymerase has been made by modeling the free energy contributions to the binding of correct and incorrect nucleotides.^12,14,20,21 Further insight has been provided in a recent study,¹⁹ which used computer simulations to model the effect of mutations on the structure of Pol β, and examined the corresponding electrostatic effects by a qualitative analysis. Additionally, theoretical studies of related problems have been reported recently.^22–25 Molecular dynamics simulations have helped delineate a possible sequence of conformational change events after dNTP binding in the catalytic cycle of Pol β.^18,26,27 Finally, an attempt to progress in a more quantitative direction in the comparison of calculated and observed mutational effects has been reported recently.²⁸ This study also introduced a useful approach of constructing interaction matrices and using them in probing the transfer of information between the binding and catalytic sites.

One of the most intriguing aspects of the action of DNA polymerases is the role of substrate-induced conformational changes in the catalytic process and the influence of these changes on DNA replication fidelity. This aspect is related to the general issue of the role of the enzyme conformational landscape in catalysis^29–31 and to proposals that dynamical effects and coupled motions play a major role in enzyme catalysis.^32,33 Although careful studies (for review, see Ref. 34) have questioned the validity of such proposals, there are aspects that were not explored in a conclusive way and one of these aspects is the relationship between DNA polymerase conformational changes and replication fidelity, which will be explored in this work. More specifically, based on structural and kinetic studies of Pol β, it is generally believed that the first step of the nucleotide insertion pathway includes the N-subdomain’s (equivalent to the finger subdomain of right-handed DNA polymerases) stable closing triggered by correct dNTP binding, whereas binding of an incorrect dNTP hampers this closing, consistent with an “induced fit” mechanism.³⁵ It has been suggested that conformational changes that occur prior to the chemical step play a major role in establishing the fidelity of Pol β (e.g., Ref. 36) and DNA polymerases in general.^4–6 Experimental evidence indicates that the open to closed subdomain transition is too rapid to be rate determining.^37–40 Instead, we think it is considerably more likely that fidelity is determined by the binding energy plus the activation barrier of the chemical step, and that the barrier for conformational changes between the open and closed forms is not likely to determine the fidelity in Pol β.^15,16 It is likely, however, that the chemical step for correct and incorrect nucleotide insertion occurs in somewhat different enzyme/substrate conformational states.^6,16

To further clarify our perspective and working hypothesis we turn the reader to Figure 1. This conceptual figure has emerged from our study of T7 DNA polymerase^13,16 where we concluded, based on calculations of the barrier for the rate determining chemical step, that the path along the coordinate for the chemical process passes at a somewhat different value of the “orthogonal” coordinate of the protein structural changes (typically considered as the motion between the open and closed forms). We would like to emphasize that this yet untested hypothesis is fundamentally different than related assumptions¹⁸ that the barriers for the conformational changes between the open and closed form contribute to fidelity or to the reaction rate. The barriers for the motion from the open to closed form (from R_OR to R_CR in Fig. 1) is not likely to influence fidelity, or for that matter, the reaction rate for both W and R as long as they are lower than the chemical barriers. Similarly the interesting possibility that the pre-chemistry barriers are different for W and R (e.g., Ref. 18) are not likely to influence fidelity as long as they are lower than the chemical barrier. In other words we are focusing here on the path between the bound incoming dNTP to the product (R_CR to R_CP and W_CR to W_CP) and on its dependence on the protein conformation, assuming (based on experimental findings) that this chemical step is rate-determining (this point will be considered further in the Discussion section).

Figure 1.

Schematics free energy surfaces that illustrate possible coupling between the conformational and chemical coordinates for the insertion of a correct (left figure) and incorrect (right figure) dNTP by a DNA polymerase. R_o and R_r designate, respectively, the open and closed configurations. R, TS1, I, TS2, and P designate, respectively, the reactant states (ES), the first TS, the first intermediate, the second TS and the product state (adapted from Ref. 16).

With the earlier perspective in mind we try here to shed some light on the following questions: (i) the possible relationship between the chemical barrier for W and R and the orthogonal protein conformational coordinate; (ii) the general relationship between the landscape of the protein conformations and the catalytic power of enzymes; (iii) the molecular meaning of the “induced fit” effect and its relationship, if any, to fidelity. Obviously these are complex problems that cannot be resolved in one study. However, Pol β offers a reasonable way of exploring these issues by allowing one to generate protein conformations between the known open and closed structures and then to evaluate the activation barrier for the chemical step for W and R. Obviously, other configurations might well be coupled to the chemical step in a tighter way, but (see later) some findings that are only based on exploring the open to closed conformational dimension might lead to quite general conclusions. In fact, despite restricting this work to the open to closed conformational space we have obtained here a useful picture on the relationship between the N-subdomain’s conformational landscape and fidelity. In particular, it is concluded that the conformation of the N–subdomain can impact replication fidelity in a non trivial way by reducing it rather than increasing it. This and other instructive findings are considered in the subsequent sections.

COMPUTATIONAL METHODS

To simulate the chemical steps in the reaction of DNA polymerase, it is essential to have a reasonable potential energy surface that describes the bond-breaking and bond-making processes. The empirical valence bond (EVB) approach provides us a powerful tool for this task. The EVB method has been described extensively elsewhere (e.g., Refs. ^41–43) and we discuss several key points here.

The EVB method is a mixed quantum mechanics/molecular mechanics(QM/MM) method that describes reactions by mixing resonance states (or more precisely diabatic states) that correspond to classical valence-bond structures, which represent the reactant, intermediate (or intermediates) and product states. The potential energies of these diabatic states are represented by classical MM force fields of the form:

$$H_{ii}={\epsilon }_i={\alpha }_{\text{gas}}+U_{\text{intra}}(R,Q)+U_{\text{S}s}(R,Q,r,q)+U_{\text{ss}}(r,q)$$ (1)

Here, R and Q represent the atomic coordinates and charges of the diabatic states, and r and q are those of the surrounding protein and solvent. α_gas is the gas-phase energy of the ith diabatic state (where all the fragments are taken to be at infinity), U_intra(R,Q) is the intramolecular potential of the solute system (relative to its minimum), U_Ss(R,Q,r,q) represents the interaction between the solute (S) atoms and the surrounding (s) solvent and protein atoms. U_ss(r,q) represents the potential energy of the protein/solvent system (“ss” designates surrounding–surrounding). ε_i is given by Eq. (1) from the diagonal elements of the EVB Hamiltonian (H_EVB). The off-diagonal elements of this Hamiltonian, H_ij, are assumed to be constant, or they can be represented by a simple function, for example, an exponential function of the distances between the reacting atoms. These H_ij elements are assumed to be the same in the gas phase, in solutions and in the proteins. The adiabatic ground-state energy, E_g, and the corresponding eigenvector, C_g, are obtained by solving the secular equation:

$$H_{\text{EVB}}{C}_{\text{g}}=E_{\text{g}}{C}_{\text{g}}$$ (2)

The EVB treatment provides a natural picture of intersecting electronic states that is useful for exploring environmental effects on chemical reactions in condensed phases. The ground-state charge distribution of the reacting species (“solute”) polarizes the surroundings (“solvent”), and the charges of each resonance structure of the solute then interacts with the polarized solvent.

The EVB approach evaluates the free-energy function Δg that is needed to calculate the activation energy, Δg^‡, by using a free energy perturbation (FEP)/umbrella sampling (US) method, which has been described in details elsewhere.⁴³

Although the EVB mapping approach explores the conformational landscape around the starting configuration, it does not explore (as explained in the discussion of Fig. 3) regions that are separated from the starting configurations by large barriers (excluding of course those changes that are captured by the EVB mapping procedure). Since we are interested here in the effect of large conformational changes, we therefore have to generate the corresponding structures by using some working hypothesis. The best option is to have direct structural information. Since limited structural information exists for “intermediate” structures, we have chosen to use the closed and open structures for Pol β and generate intermediate N–subdomain positions. The open and closed structures were obtained from the crystal structures of Pol β in the open binary (1bpx) and closed ternary (1bpy) forms.⁴⁴ The simulation systems were prepared by adding the missing hydrogen atoms to both structures, and the missing 3′OH group to the DNA primer strand. All the crystal waters were taken away, except the one coordinated to the nucleotide-binding Mg²⁺ ion. Another water molecule was manually added to coordinate the catalytic Mg²⁺. For the open structure, the substrate dCTP was manually docked into the polymerase active site, followed by gas phase steepest-descent optimization to remove atom clashes. A rigid structural alignment for the open and closed conformers was performed using VMD.⁴⁵ This was done by keeping the closed conformer fixed and performing a rigid rotation and translation on the open conformer to find the best root-mean-squared fit between the open and closed forms. This structural alignment was necessary because experimental data has shown that the Pol β lyase domain does not change significantly after the binding of the DNA substrate, whereas the major conformation changes occur in the N–subdomain and the polymerase active site upon binding of the dNTP substrate.^46,47 Since the coordinates of both structures deposited in protein data bank (PDB) have difference in the overall orientation, we minimized the difference in a way that allowed us to use the aligned open structure as the starting point and the aligned closed structure as the target in a constrained MD simulation. After alignment, the lyase domain of both structures is almost superimposable and the root mean square deviation (RMSD) for the difference between the positions of all the atoms of the protein in the two structures is about 4 Å.

Figure 3.

A schematic representation of the free energy landscape of the chemical step. This figure presents a situation where the barrier between points 2 and 3 is higher than 5 kcal/mol, where the barrier between points 1 and 2 is smaller. The coordinates 1–4 correspond to different protein conformations, which may all lead to chemical transition states but give different activation barriers.

The aligned structures were then solvated in a 24 Å water sphere, using the surface-constrained all-atom solvent (SCAAS) model.⁴⁸ The simulation center was chosen to be within the polymerase domain, as discussed in the previous section. During the simulations, the coordinates of atoms outside the sphere were kept fixed and their interactions with atoms inside the sphere were turned off. Long-range electrostatic interactions beyond a 12-Å cutoff were treated by the local reaction field (LRF) method.⁴⁹ The nonbond van der Waals (vdW) interactions were smoothed to zero beyond the same cutoff. However, the EVB and FEP calculations were done without any cutoff for the electrostatic interactions between region I and the rest of the system and all these interactions were evaluated explicitly. All simulations were performed using the MOLARIS simulation package.⁵⁰ The nucleic acid bases were represented by AMBER charges⁵¹ and ENZYMIX vdW parameters (after validation that these parameters give reliable solvation energies and base-pairing energies and structures). A single center model with the vdW parameters r* = 1.30 and ε = 0.06 kcal/mol¹³ was used for the Mg²⁺ ions.

A targeted molecular dynamics (TMD)⁵² simulation was performed by introducing Cartesian constraint forces of the form

$$V_{\text{cons}}=K{[d_{\text{rms}}(X(t),X^{\text{target}})-d_0]}^2$$ (3)

where K = 10 kcal/mol/Å ² is the force constant and runs over all the protein heavy atoms inside the simulation sphere, and d₀ is an offset constant. The open structure was used as the starting point and the closed one as the target in the simulations. Starting from different initial conditions, we ran multiple trajectories with a time step of 1 fs, and each trajectory took 2 ns. The coordinates of the structure of the whole system were saved every 1 ps for later analysis. A few intermediate structures were chosen for the subsequent FEP/EVB calculations.

Since our focus is on the polymerase N-subdomain, the simulation center and radius were carefully chosen to enable the simulation sphere to accommodate the entire subdomain and also allow sufficient space for the atoms to move in the direction determined by the TMD. We also change the center gradually in a series of simulations. Explicit water molecules were added to solvate the system. Since it is well known that TMD can generate unphysical structures, a long time MD relaxation was carried out after each specified RMSD was reached. Furthermore, careful visual inspection was conducted to ensure that the generated structures were reasonable.

The structures with mispairs (G:A, G:G, G:T) in the nascent base pair binding pocket were constructed by manually docking the corresponding dNTP to the generated intermediate and closed conformers. Then, the same procedure as earlier was followed to solvate the system, and a regular MD relaxation of 500 ps was performed to bring the system temperature from 0 to 310 K until equilibrium was reached. These relaxed structures served as the ground states for calculating the free energy profiles of the chemical step.

The EVB calculations require information about the reaction mechanism and the potential surface for the solution reaction as well as the most likely mechanism for the reaction in the protein. In the case of the reaction catalyzed by Pol β, there is still debate about whether phosphodiester hydrolysis reaction follows an associative, dissociative or concerted mechanism. Careful EVB calculations of activation free energies calibrated on ab initio calculations¹³ and ab initio QM/MM mapping studies¹⁷ have indicated that an associative mechanism is the most likely reaction mechanism for correct nucleotide insertion. Thus, for chemistry in the active site of Pol β, we assumed the reaction takes place in the three steps depicted in Figure 2; a proton transfer from the primer 3′OH is followed by an in-line nucleophilic attack of 3′O⁻ on the dNTP α-phosphate, and the reaction is completed by the departure of the pyrophosphate leaving group. For the proton transfer step, there are also several possibilities: the proton of the primer 3′OH could either transfer to Asp256, 5′-oxygen or a nonbridging oxygen of the dNTP α-phosphate, or a surrounding water. Based on previous simulations, it seems most likely that the dominant mechanism involves a proton transfer to an acidic group (Asp256 in Pol β and Asp654 in T7)^13,17,28 or water^16,28 followed by nucleophilic attack on P_α by the deprotonated 3′OH and formation of a pentavalent intermediate in an associative mechanism. The nature of the exact surface is still an open question (it may require a more concerted and flat surface) but the overall changes between the W and R activation barriers are assumed to be similar for both surfaces. Here, we based the EVB on the stepwise mechanism described in Figure 2. This is viewed as a convenient strategy that is useful also for studies of the concerted mechanism.

Figure 2.

The assumed chemical steps of the nucleotide incorporation reaction of Pol β.

The nucleotide incorporation reaction in Pol β involves an initial proton transfer step. In principle, to have a complete reaction free energy profile, we need to calculate the proton transfer activation barrier, Δg^‡_PT. But since this step is not likely to be rate limiting, it is sufficient just to calculate the proton transfer reaction energy, ΔG_PT, which in turn requires calculating the pK_a shift of the proton donor. The pK_a shift of the ribose 3′OH group in the protein is calculated by using the FEP method and the following relationship⁵³

$$\mathrm{\Delta }\text{p}{K}_{\text{a}}=\mathrm{\Delta }\mathrm{\Delta }{G}_{\text{sol},\text{OH}\to {\text{O}}^{-}}^{\text{wat}\to \text{pro}}/2.303\text{RT}$$ (4)

ΔG_PT is then calculated as

$$\mathrm{\Delta }{G}_{\text{PT}}=2.303\text{RT\hspace{0.28em}}(\text{p}{K}_{\text{a},3^{\prime }\text{OH}}-\text{pH})$$ (5)

The pK_a of the ribose 3′OH in water is taken as 13.⁵⁴ Here pK_a,3′OH = 13 + ΔpK_a and pH = 7 are used in the calculations.

The free energy profiles of the P—O bond-forming and bond-breaking reaction were calculated by the EVB method. The reference solution reaction is chosen as the attack of CH₃O⁻ on dCTP⁴⁻ in the presence of two Mg²⁺ ions in water. The calculated EVB surface was adjusted by forcing it to give an activation barrier of 21 kcal/mol to reproduce the corresponding experimental estimate and the general features found by previous ab initio calculations.¹³ The resulted EVB parameters are summarized in Table I. The same EVB parameters were then used without any change in the simulation of the reaction in Pol β. The EVB simulations were carried out using the SCAAS model.⁴⁸ The simulation center was chosen as the geometric center of the selected EVB atoms. The electrostatic long range effects were treated without a cutoff by using the LRF method.⁴⁹ The EVB mapping was carried out in 21 windows each of a simulation time of 20 ps, with a time step of 1 fs at 310 K.

Table I.

The EVB Parameters Used in the Present Work

	Reactant state		Intermediate state		Product state
Atom no.	Charge	Atom type	Charge	Atom type	Charge	Atom type
1	1.336	P0	1.343	P0	1.370	P0
2	−0.963	O−	−0.993	O−	−1.035	O−
3	−0.963	O−	−0.993	O−	−1.035	O−
4	−0.963	O−	−0.993	O−	−1.035	O−
5	1.487	P0	1.402	P0	1.370	P0
6	−0.921	O−	−0.956	O−	−1.035	O−
7	−0.921	O−	−0.956	O−	−1.035	O−
8	0.594	Op	0.541	Op	0.530	Op
9	1.341	P0	1.548	P+	1.270	P0
10	0.901	O−	0.933	O−	0.810	O−
11	0.901	O−	0.933	O−	0.810	O−
12	0.617	Op	0.781	Op	1.035	O−
13	0.420	O0	0.448	O0	0.329	O0
14	1.000	O−	0.766	Oq	0.321	Op
Ebond = Kb(1 − e−β(r − r0))2
Atom1	Atom2	Kb	r0	β
P0	O0	95.0	1.58	2.0
P0	Op	85.0	1.70	2.0
P0	Oq	85.0	1.70	2.0
P0	O−	120.0	1.50	2.0
P+	O0	80.0	1.65	2.0
P+	Op	50.0	2.10	1.0
P+	Oq	50.0	2.10	1.0
P+	O−	100.0	1.52	2.0
Eangle = Kθ(θ − θ0)2 + D(1 − e−(θ − θ0)2/ζ)
Atom type	Ky	θ0	D	ς
O0	80.00	109.5	0.00	1.00
Op	80.00	140.0	0.00	1.00
Oq	0.00	109.5	40.00	0.48
O−	50.00	120.0	0.00	1.00
P0	140.00	119.2	0.00	1.00
P+	0.00	120.0	60.00	0.43
Etorsion = Kt (1+cos(nφ −φ0)
Atom type		Kt	n	ϕ0
P0	O0	3.0	3.0	0.0
P0	Op	2.0	3.0	0.0
P+	O0	3.0	3.0	0.0
P+	O−	3.5	3.0	0.0
P+	Op	2.0	3.0	0.0
P+	Oq	2.0	0.0	0.0
van der Waals Parameters
Atom type	A	B
O0	774.0	24.0
Op	1379.0	36.0
Oq	774.0	24.0
O−	1379.0	36.0
P0	1250.0	24.0
P+	1500.0	24.0

RESULTS AND ANALYSIS

In exploring the energetics of the transition states (TSs) in Pol β and related systems we will focus on the chemical step in Figure 1, considering a relatively wide region in the reactant state (RS). In this case one faces a situation of the type presented in Figure 3, where the activation free energy of the chemical step reflects all the possible paths in a rather complex landscape. The coordinates 1–4 correspond to different protein conformations, which may all lead to chemical TSs but give different activation barriers. Pulling the system from its RS to its TS by the FEP/US approach described in the Method section, for say 1 ns, does not allow the system to jump between regions, which are separated by more than 5 kcal/mol barriers along the protein coordinate. Thus starting at point (1) in Figure 3 and then performing a FEP/US mapping will leave the system in the region between 2′ and 1′. A more complete mapping should include starting at the (3)–(4) region and averaging the results of the different mappings. Of course, there might be additional challenging problems, such as the fact that in reality when Δg^‡<Δg_(2)–(3) there may be transitions between these states that are not sampled by the FEP/US simulation. This challenging issue will be addressed in later studies. At present we believe that most of the physics of having somewhat different landscape for W and R can be explored by starting the simulations of Δg^‡ from different initial configurations. Thus we started by generating representative protein conformations that included the open, closed, and intermediate structures according to the procedures outlined in the Method section. We will refer to the TS conformation as the conformation obtained by starting at a given conformational region of the protein and then allowing the system to reorganize during the umbrella sampling mapping to the TS.

The TMD generation of intermediate structures involved the use of the Cartesian constraint described in the Method section that allowed us to move between the closed and open conformations depicted in Figure 4. An intermediate conformation (the one marked in black in Fig. 6) as well as the closed and open conformations is depicted in Figure 4. As seen from the figure, the overall closed to open structural change involves a change in the conformation of the α-helix N (residues 275–295) of the N–subdomain. The structural changes in this region are of interest since substrate binding brings this α-helix to close contact with the nascent base pair of the template and incoming dNTP. Mutational studies^7,37,56–60 have indicated that some residues (e.g., Asp276, Lys280, and Arg283) in this helix play an important role in controlling Pol β replication fidelity. The figure depicts structural changes of the Mg²⁺ coordinated residues since this environment plays an important role in the chemical step.

Figure 4.

Open and closed conformations of key Pol β and DNA residues. The crystallographic structure of the C-subdomain of the open binary DNA (pdb accession code 1BPX; pink) and closed ternary complex (green) were superimposed. Left panel: Ribbon representation of Pol β illustrating the open and closed conformations of the carboxyl-terminal N-subdomain as highlighted by α-helix N. The incoming nucleotide (dTTP) and active site Mg²⁺ (dark green spheres) are shown, but the DNA is omitted for clarity. Right panel: The position of the N-subdomain can be monitored in the active site through a series of altered interactions between Asp192 (D192), that coordinates both active site metals, and Arg283 (R283) that is situated in α-helix N and interacts with the minor groove edge of the templating strand. The arrows represent motions associated with the closing transition. The nascent base pair (dA—dTTP) is illustrated (gray carbons), but the duplex DNA is omitted for clarity. This figure was made with Chimera.⁵⁵

Figure 6.

The representation of the configurations considered in this work, in terms of (a) the number of hydrogen bonds and the RMSD and (b) the relative contact order and RMSD. The calculations of these quantities only include the Pol β polymerase domain, not the DNA and dNTP substrate. The criterion used for identifying hydrogen bond is R(X-H…Y)∈[2.5,3.5] Å. The relative contact order is calculated as originally defined in Ref. 62.

The generated intermediate conformations are more or less half way between the closed and open structures and a typical intermediate structure is depicted in Figure 5. Another way to view the structural changes is to consider the overall conformational space in terms of the parameters used in studies of protein folding. This is done in Figure 6, where we describe the configurational space in terms of the RMSD relative to the closed structure (considering all the protein atoms) and the number of hydrogen bonds, N_hb, or alternatively the relative contact order, S_co^61,62. At present it is not clear how these parameters are likely to change during protein folding or conformational transitions (whether linearly or nonlinearly). Searching for appropriate reaction coordinates or order parameters for describing these processes in complex systems remains a challenge. However, one can use these parameters as a measure of the progress of the forced structural change in the simulations. As shown in Figure 6, N_hb and S_co for intermediate structures fall between the corresponding values of the open and closed structures. The absence of any sudden jump indicates a smooth transition during the Pol β open to closed conformational change. Additional data points will be required to derive a complete transition pathway.

Figure 5.

Intermediate and closed conformations of key Pol β and DNA residues. The C-subdomains of the calculated intermediate (yellow) and closed ternary complexes (green) were superimposed. The nascent base pair (dA—dTTP) and active site Mg²⁺ (spheres) are shown, but the DNA is omitted for clarity. The position of the N-subdomain can be assessed through the position of α-helix N and several key active site residues. The view is similar to that in Figure 4 (right panel).

Although we have used the crystal structure of a matched base pair system (G:C) in the closed conformational state as the target in the TMD simulations, it is possible that the generated configurations may not represent the actual conformation sampled by the TS of the wrong dNTP. This is the most likely case if the relevant configurations do not occur along the open-close pathway. However, as will be discussed in the next section, the present treatment is still instructive.

We also like to emphasize that we do not consider the TMD as an optimal approach for exploring conformational states and we are refining new approaches that use simplified folding models for such a task.⁶³ However, as will be shown later, we do not need to know the free energy for transitions between the different protein conformational states because we use an alternative approach to obtain the relevant information.

After generating the conformations described earlier, we performed systematic EVB calculations of the activation barriers for the chemical steps described in Figure 2 and discussed in the Method section. The calculations were done with correct and incorrect base pairs in the different protein conformations described earlier. The results of these simulations are summarized in Table II and the calculated activation barriers are compared with the corresponding barriers obtained by applying TS theory to the k_pol determined by kinetic measurements.⁶⁴ In doing so it is assumed that k_pol represents the rate constant of the chemical step. As seen from the table, we have good agreement between the calculated and observed activation barriers for the incorporation of the correct nucleotide. The situation with the incorrect nucleotide is far more complicated (see Fig. 7 for a typical behavior). Here we see that the barriers depend very strongly on the generated conformations. In general, it appears that the barriers for the mismatches are reduced significantly in conformational states that correspond to partially open structures. This trend, which was already predicted from qualitative considerations,¹⁶ seems to hold for the more quantitative analysis described here. The overall trend is illustrated in Figure 8 for G:C and G:A.

Table II.

Calculated and Observed Chemical Free Energies of the Nucleotide Incorporation Reaction of Pol β^a

	Observed		Calculated
Base pair (template:primer)	kpol (1/s)	Δg‡cat (kcal/mol)	ΔGPT (kcal/mol)	Δg‡cat, Nuc (kcal/mol)	Pentavalent intermediate (kcal/mol)	Δg‡cat, PO (kcal/mol)	Product (kcal/mol)	Δg‡cat (kcal/mol)
G:C	12.50	16.6	2.4	13.9	11.4	4.3	−10.2	18.1
			2.5	15.5	14.2	3.9	−8.2	20.6
			2.0	15.2	12.8	6.0	−9.6	20.8
			3.5	14.4	12.1	5.8	−9.2	21.4
			3.3	14.5	13.0	6.4	−6.8	22.7
G:T	0.37	18.7	2.6	26.2	25.0	4.2	−5.0	31.8
			3.7	18.4	15.3	3.8	−4.1	22.8
			3.2	19.3	17.1	2.8	−7.2	23.1
			5.3	24.3	21.0	3.2	−7.2	29.5
			3.0	17.2	14.5	6.2	−7.5	23.7
G:A	0.02	20.5	3.0	25.3	24.2	7.0	3.5	34.2
			4.2	22.4	19.8	4.5	2.1	27.5
			2.3	24.8	22.0	5.5	−6.0	29.8
			3.3	18.0	16.6	5.4	−3.7	25.3
			4.1	15.6	14.4	6.5	−5.5	25.0
G:G	0.04	20.1	5.2	18.3	17.4	5.5	−4.5	28.1
			4.3	17.5	15.2	4.6	−5.0	24.1
			3.8	18.0	16.2	5.2	−5.3	25.2
			3.8	18.7	15.3	4.4	2.0	23.5
			4.8	18.2	14.6	6.3	1.5	25.7

^aThe observed k_pol from the kinetic measurement⁶⁴ is interpreted as the rate constant of the chemical step. For each base pair we give the calculated results for the closed configuration in the first row (bold) and the four intermediate configurations in the following rows. The value of Δg^‡_cat was obtaind by combining ΔG_PT, the energy of the pentavalent intermediate, and Δg^‡_{cat, PO}.

Figure 7.

Reaction free energy profiles for G:C and G:T in the closed and intermediate configurations.

Figure 8.

Free energy profiles for (a) G:C and (b) G:A in the space defined by the protein conformational coordinate and the chemical reaction coordinate in the open, intermediate and closed conformations. The figure only considers the reaction profiles for the closed, intermediate and open conformations and connects them by a simple interpolation. This treatment does not take into account the barriers between the states (e.g., see a schematic description in Fig. 3), which were not evaluated in this work. However, the figure establishes that the TS energy for W is lower in the intermediate conformation than in the closed conformation.

The evaluation of Figure 8 should also include the estimate of the effect of the conformational changes on K_D and this would require additional extensive study (e.g., see Ref. 28). Here we decided to focus on a qualitative estimate that seems to be sufficient to provide general support for our findings. That is, we considered the thermodynamic cycle of Figure 9 and obtained the relationship

Figure 9.

A schematic description of the thermodynamic cycle used to evaluate ground state binding in different protein conformations.

$$\begin{array}{l}\mathrm{\Delta }{G}_{\text{bind}}^{\text{op}\to \text{cl}}=\mathrm{\Delta }{G}_{\text{enz}}^{\text{op}\to \text{int}}+\mathrm{\Delta }{G}_{\text{enz}}^{\text{int}\to \text{cl}}+\mathrm{\Delta }{G}_{\text{bind}}^{\text{cl}}\\ \mathrm{\Delta }{G}_{\text{bind}}^{\text{op}\to \text{int}}=\mathrm{\Delta }{G}_{\text{enz}}^{\text{op}\to \text{int}}+\mathrm{\Delta }{G}_{\text{bind}}^{\text{int}}\\ \mathrm{\Delta }\mathrm{\Delta }{G}_{\text{bind}}^{\text{int}\to \text{cl}}=\mathrm{\Delta }{G}_{\text{bind}}^{\text{op}\to \text{cl}}-\mathrm{\Delta }{G}_{\text{bind}}^{\text{op}\to \text{int}}\\ =\mathrm{\Delta }{G}_{\text{enz}}^{\text{int}\to \text{cl}}+\mathrm{\Delta }{G}_{\text{bind}}^{\text{cl}}-\mathrm{\Delta }{G}_{\text{bind}}^{\text{int}}\end{array}$$ (6)

where the different terms in the equations are defined in Figure 9. In this case, it is sufficient to show that $\mathrm{\Delta }\mathrm{\Delta }{G}_{\text{bind}}^{\text{int}\to \text{cl},\text{W}}$ (where W designates “wrong”) is small or nagative, to support the trend of Figure 8(b). Thus we calculated $\mathrm{\Delta }\mathrm{\Delta }{G}_{\text{bind}}^{\text{cl}}$ and $\mathrm{\Delta }{G}_{\text{bind}}^{int}$ by the microscopic linear response approximation (LRA) method as done in Ref. 28. The corresponding results are summarized in Table III. The calculations reproduce the general observed trend of having reduced binding (ca. 4 kcal/mol) for W, but more importantly, for the purpose of the present article, they indicate that the bound W is more stable in the intermediate conformation. We also noted that $\mathrm{\Delta }{G}_{\text{bind}}^{\text{int}\to \text{cl}}$ is probably negative (since op is more stable than cl in the absence of the “substrate” (dNTP + 2Mg²⁺) and since int is on the way from op to cl. Thus $\mathrm{\Delta }{G}_{\text{enz}}^{\text{int}\to \text{cl},\text{W}}$ is most probably negative. Adding this result to the finding of a lower barrier in the chemical step of W provides further support to the general trend of Figure 8 where the overall barrier that corresponds to k_pol/K_D for W is lower at the intermediate state than in the closed state. At any rate, since the change in the energy of the ES ternary complex for op→cl is negative for R and less negative for W with a positive trend for int→cl, this contribution appears to be smaller than the changes in the chemical barrier and thus were not incorporated in Figure 8.

Table III.

Calculated and Observed Ground State Binding Free Energies^a

	Observed		Calculated
			ΔGbind (kcal/mol)
Base pair (template:primer)	KD (uM)	ΔGbind (kcal/mol)	Closed	Intermediate
G:C	1.9	−8.1	−12.4	−7.2
G:T	650	−4.5	−7.3	−9.4
G:A	270	−5.0	−8.5	−9.1
G:G	360	−4.9	−5.4	−7.3

^aThe observed K_D’s were taken from Ref. 64. The binding energy calculation was carried out using the LRA method for the closed and one of the generated intermediate conformations. The calculations are based on the scheme shown in Figure 9. A dielectric constant ε_eff = 12 was used, according to our previous study. ²⁸

The calculated features reported in Figure 8 can depend on the protonation states of the protein ionizable residues and in principle these ionization states can be different for different protein conformations. To explore this effect we evaluated the ionization states of some key residues in the closed and partially open configuration of W by the PDLD/S-LRA approach.⁶⁵ As shown in Table IV, the ionization states are likely to stay similar in both structures. Moreover, the relatively small effect of the ionizable groups on the TS binding energy (that has been evaluated in our recent work by combining calculations and mutational analysis²⁸) indicates that even if few groups change their ionization states, we will still have lower energy for the TS of W in the intermediate state than in the closed state.

Table IV.

Calculated pK_a of Some Key Residues in the Closed and One of the Intermediate Conformations^a

Residue	Closed	Intermediate
Arg183	13.9	13.8
Asp190	−3.4	−4.7
Asp192	−3.4	−3.4
Arg254	11.2	11.4
Asp256	0.0	−0.3
Arg258	11.4	11.7
Asp276	3.6	3.5
Lys280	9.3	10.3
Arg283	11.4	12.2

^aThe calculations provide the PDLD/S_LRA results for the apparent pK_a’s for the set of dielectrics ε_p = 20 and ε_eff = 40. The use of ε_p= 20 is dictated by the explicit representation of the Mg²⁺ ion whose effect must be screened by a relatively large dielectric constant.

To further clarify the origin of the calculated conformational effects, we evaluated the electrostatic contributions of selected residues to the TS energy of the (G:A) mispair. These calculations used the relationship

$$\mathrm{\Delta }{G}_{\text{el}}=332\sum _{i,j}{Q}_i^{\text{TS}}{Q}_j^m/r_{ij}{\epsilon }_{\text{eff}}$$ (7)

where Q_i^TS and Q_j^m are, respectively, the partial charges of the ith atom of the substrate at the TS and jth atom of the mth protein residue, r_ij is the distance between the i and j atoms (given in Å) and ΔG is the electrostatic contribution. The effective dielectric constant, ε_eff, was taken as 12 and 4 for ionized and neutral residues, respectively, following previous experimentally based parameterization.²⁸ The calculations included all the protein residues. The calculated changes in the electrostatic contributions upon the closed to intermediate transition are depicted in Figure 10. As shown in the figure, we predicted significant changes for some residues, for example, Asp190, Arg258, Asp276, and the Mg²⁺ ions. Having a large effect from these groups is consistent with mutational studies and previous studies of the protein allosteric effect.²⁸ However, more refined analysis would benefit from experiments that will probe the effect of Pol β mutations on the structural changes induced in this enzyme by TS analogues.

Figure 10.

The change in the electrostatic contribution to the transition state binding free energy for the G:A mispair for the transition from the closed to an intermediate structure. ΔΔG corresponds to the difference between the values reported in the upper figure for the intermediate and closed states.

DISCUSSION

Landscape and fidelity

The fidelity of DNA polymerase reflects an allosteric competition between the binding of the base of the incoming dNTP to the base-binding site and the binding of the TS to the catalytic site. In the case of an incorrect dNTP, poor preorganization in the base-binding site and poor binding of the base lead to structural rearrangements that are propagated to the catalytic site and destroy its preorganization.²⁸ This allosteric control of fidelity is described in Figure 11. Applying this concept and previous findings²⁸ to the results presented in Figure 8 indicates that the binding of an incorrect base in the closed configuration results in a larger destruction of the preorganization of the active site than would be expected from experimentally measured Pol β catalysis (Table II). This effect appears to be reduced in some generated intermediate configurations. Thus the structural changes per se do not increase fidelity but instead reduce the very large difference in fidelity that would occur if both the W and R reactions were to involve the same closed structure. This means that the coupling to conformational changes is not the origin of the fidelity (if it was, then the conformational change would lead to larger fidelity).

Figure 11.

A schematic description of our view of the origin of the fidelity on DNA polymerases. The protein provides perfect sites for both the chemical and base-pairing parts of R. On the other hand, in the case of W, the protein has to relax in the base-pairing site and this relaxation destroys the preorganization in the chemical part. Arrows indicate the protein dipoles.

A recent study⁶⁶ used a conformationally sensitive fluorophore attached to the recognition domain of T7 DNA polymerase to probe structural changes. Kinetic analyses suggest that the conformational changes associated with the florescence kinetics are those involved in the alignment of the catalytic site for R and for the misalignment of the active site for W. It was then suggested that conformational changes that occur after the binding step and before the chemical step play a major role in controlling the fidelity of DNA polymerases. It was suggested that this step is used to align the active site for the correct nucleotide.

Although these findings are potentially important, it is important to realize that there are many possible structural changes and only some of them have an impact on the chemical step. It is likely that the fluorescence probe reflects changes in the base-binding site and does not necessarily tell us about the preorganization in the catalytic site. That is, the most important catalytic preorganization occurs upon binding of the charged part of the incoming nucleotide, and this preorganization is expected to be similar for R and W. The difference occurs in the base binding process which is used, in the case of W, to destroy the preorganization (see Fig. 11). The florescence changes observed with the current probe may reflect mainly the structural changes in the base binding site rather than the relevant part of the induced fit in the chemical site. It seems that only direct structural studies and/or simulation approaches can tell us about the nature of the preorganization effect that follows the binding step. Using fluorescence labeling can be very useful in estimating the time scale of these changes, but even this should be done with great care, with two or more probes, trying to correlate the position of the given probe with the two allosteric regions.

What is the relationship between induced fit and fidelity?

Since our work explored the role of conformational changes in Pol β, it is important to discuss the idea that the fidelity of DNA polymerase is related to the induced-fit proposal (e.g., 66). The fact that the substrate binding induces conformational changes that bring Pol β to its catalytic configuration can be described as an induced fit effect. However, this does not provide information about the nature of the catalytic effect in the preorganized active site, nor does it tell us how the structure of the ES (or in our case the E-DNA-dNTP) complex is related to the activation barrier of the chemical step. Chemical catalysis is defined by the difference between the activation barrier in the ES complex and the barrier for the reference reaction in water. It does not depend on the structure of the enzyme before it binds the substrate. Thus, the induced fit concept does not describe the origin of the ES structure. Enzyme catalysis can be considered on the same level as other parts of the folding to ES state (in general the folding energy is invested in reorganizing the active site⁶⁷).

The use of the term “induced fit” to describe the origin of the polymerase fidelity is also problematic. There are opposing views on whether or not the induced fit can lead to increased selectivity or fidelity.^68,69 The use of kinetic diagrams of the type considered by Post and Ray⁶⁸ does not provide a unique thermodynamic analysis because they do not stipulate well defined free energy surfaces. Perhaps this is the basis for the current controversies. In our view, it is most useful to describe an induced fit effect in terms of energy diagrams of the type presented in Figure 8. In this way one can define the relationship between the structural changes upon substrate binding and the corresponding energetics, including: (i) the energy change of the enzyme upon moving from r(E) (where r is the protein configuration at the designated state) to r(ES); (this energy is defined as the reorganization energy upon binding⁷⁰). (ii) the change in the enzyme-substrate interaction upon moving from r(E) to r(ES) and (iii) the change in the activation barrier for the chemical step as the result of the binding induced structural change (this is defined in terms of the change in the reorganization energy of the chemical step and the work term⁷¹). Without such estimates it is likely that the discussion will become circular.

Even the definition of the reference state, where we do not have an induced fit (this is considered usually as the “hypothetical” rigid structure^68,69) presents a logical problem since it could be any arbitrary structure (closed or open). Thus the most logical definition is to first take the actual structure of the specific system without the substrate and then to examine what is the effect of the binding process, rather than to talk about an ill defined arbitrary rigid structure. At any rate, the induced fit effect in the binding step is described in the lower part of Figure 8. If we identify the motion from the open to the closed conformation as the induced fit effect in the binding step (this is done with the knowledge that the system is in the open state without the substrate) then the induced fit effect should be larger in R than in W, since the binding free energy in the closed configurations is expected to be smaller for W than for R. A difficulty occurs, however, when we explore the induced fit effect on the binding of the TS, which determines the fidelity. In the likely situation that the chemical step is rate-limiting, we expect k_pol/K_D to determine the overall fidelity of DNA polymerases.¹⁶ Here we have to first define the reference structure (the structure of the isolated E), which is changed by the induced fit effect. Taking the open structure as the reference state, we find that the induced fit always increases fidelity since the reactivity in the open structure is similar to the reactivity in water where we have similar barriers for R and W and no selection for fidelity. This result is not meaningful since it does not tell us about the actual nature of the fidelity, that is, what are the factors that increase it. The situation becomes more problematic if we choose the closed structure with R as the reference state (note that we can consider any possible definition). Once the reference structure is taken as r(ES(R)), the induced fit effect will be described as the motion from r(ES(R)) to r(ES (W)), but here there is no way to predict a priori whether this structural change will increase or decrease the energy difference between the TSs of R and W, which in turn would either decrease or increase the fidelity. If the induced fit concept is extended to the structural changes in the TS and to the configuration that stabilizes the TS, then the reorganization energy associated with the move from r(ES(R)) to r(TS(R)) will generally be smaller than that from r(ES(W)) to r(TS(W)). However, it is not clear what the reorganization energy will be upon moving from r(ES(R)) to r(TS(W)), and so it is not clear if the fidelity will increase or decrease because of the induced fit effect.

Since the induced fit upon binding of R leads to an improved catalytic configuration and since some of the binding energy is being invested in the reorganization process, it might seem reasonable to assume that W will induce less reorganization (less induced fit) because its active site is less preorganized. Unfortunately, this argument, which is equivalent to the assumption of a direct relationship between induced fit and fidelity, is unjustified. That is, W and R are different molecules and there is no way to predict a priori which of them will involve larger reorganization upon binding. In other words, there is no way to tell (without calculations) if the binding of W will lead to smaller or larger reorganization than the binding of R since we only know that the binding of R improves the preorganization of the catalytic site. The binding of W may involve small reorganization or a very large reorganization with a very large cost. For example, the binding of W may be used to move the active site to an anticatalytic configuration.

The idea that the enzyme uses the “intrinsic” substrate binding energy to achieve its optimal catalytic effect (e.g., Ref. 72) has also been invoked in analyzing the induced fit effect.^66,73 In particular it has been implied that a portion of the binding energy must be consumed by reorganizing the active site and thus will not be available to stabilize the TS.⁶⁶ Unfortunately this line of thought is problematic. First, the whole assumption that the enzyme uses the binding energy of distant groups lead to destabilization of the ground state of the chemical part (this is the essence of Jencks’ idea⁷² about the usage of binding energy) has not been confirmed by careful computational and conceptual studies of specific cases.⁶⁷ One of the simplest ways to realize this point is to see that most mutations that reduce catalysis also reduce binding rather than increase it. Second, the binding energy in the case of induced fit is invested in part in pushing the TS to the preorganized structure and not “consumed” in destabilizing the ground state. In other words, the binding energy that leads to structural rearrangement increases rather than decrease catalysis. Now, as much as the fidelity is concerned we may try to view the base as the distant part of the substrate (the part whose energy should be used for inducing catalysis). Doing so we face the problem that a good binding of the base (a good base pairing) is used to stabilize the TS rather than destabilize the RS. This point can be seen by noting that mutations that increase k_pol significantly for R usually decrease K_D (lower K_D means stronger binding), or leave it unchanged, rather than increase it,²⁸ as would be expected if the reduction of Δg^‡_cat is because of the ground state destabilization. Interestingly the distant part is quite effective in destroying rather than increasing catalysis, as is obviously the case with W.

Landscape catalysis and dynamical effects

The present work considered in a preliminary way the effect of the complexity of the landscape in enzyme catalysis. As illustrated in Figure 3, we are faced with a complex situation in assessing the probability of reaching the TS region in the multidimensional surface of the enzyme-substrate complex. However, in contrast to some recent proposals,³⁰ we do not see any fundamental problem with this complexity, nor do we find a catalytic advantage from this complexity. The possibility that the TS region has more configurations than the RS is a simple entropic effect that does not appear to be supported by the fact that the observed values of the activation entropies in enzymes are not large.⁷⁴ The possible dynamical effects associated with transitions between different conformational subspaces during the fluctuations that lead to the TS (see Fig. 1) are not expected to be large, although simulation studies of this issue are clearly needed. Of course, it is clear that different conformations can be associated with different activation barriers, as indicated by recent single molecule studies of enzymes.²⁹ However, this point has been taken into account (at least in part) already in our approach of averaging over different initial configurations (e.g., Refs. ^{71, 75} and ⁷⁶). That is, in our simulations we frequently run series of FEP/US calculations starting with different initial conditions and take the average of the calculated activation barriers. This approach accomplishes more or less what is done in the so called replica exchange studies (see discussion in Ref. 71) and provides an estimate of the effective activation barriers.

We have assumed that k_pol is the rate constant for the chemical step, which is consistent with available kinetic evidence. These observed rate constants were converted to activation free energies to compare to the calculated values. Additional experiments may be required to resolve any shortcomings of this assumption, and a more sophisticated model may be needed for a deeper understanding of the kinetics during Pol β catalytic cycling. It may be also reasonable to question whether the measured k_pol for the incorporation of correct and incorrect dNTP represents the same microscopic reaction rate constant. Recent experimental³⁸ and theoretical results⁷⁷ indicated that this is indeed the case, and that the chemistry is generally the rate-determining step at least for Pol β.

Recent ideas that the fidelity of DNA polymerases may involve “gate keeping” dynamical effects⁷⁸ and related earlier ideas⁷⁹ are of interest and should be examined further. Time-resolved fluorescence⁴⁶ and NMR^80,81 studies have demonstrated that the motions (which have often been referred to as dynamics) of α-helix N are reduced when Watson-Crick hydrogen bonding occurs (i.e., correct nucleotide binding), but not when hydrogen bonding is absent. In contrast, protein side-chain motions are increased in the vicinity of the active site aspartates when the closed complex is formed.^80,81 However, these motional effects are not likely to represent true dynamical effects but merely equilibrium thermal fluctuations that follow the Boltzmann populations in different landscapes. In this case, we simply have well defined entropic effects; an analysis and interpretation of these types of dynamical effects is discussed in Refs. ³⁴ and ⁶⁵. We have not found convincing theoretical or experimental evidence that dynamical effects play a significant role in catalysis.³⁴

If the dynamical effects are considered as the motions between the open to closed conformations, then it is difficult to see how it would be advantageous to use these as a mechanism to control replication fidelity. Fidelity is determined by the ratio between k_pol/K_D of W and R. If the rate limiting step is the conformational transition between the open and closed conformations and the barrier for these states is much larger for W than for R, then there could, in principal, be conformational control of fidelity. However, this mechanism is in conflict with the chemical step being rate limiting (see Refs. ^{15, 16}, and ³⁸). If the chemical step is rate limiting, then it seems that the only way for conformational changes to control fidelity is that the TS for both W and R must occur in a different conformation than the RS, and the barrier along the conformational axis in Figure 8 is higher than the chemical barrier, and that this barrier will be higher in W than in R. Although we have not determined the conformational barriers (in part because it is not clear what, if any, conformational transition occurs in Pol β.), our calculations are consistent with R having both the TS and RS in the same closed conformational region. As argued earlier, the situation with W is such that a barrier for the transition to the TS will only push the fidelity above its observed value.

Finally, it may be useful to comment here on the implication of Ref. 18 that pre-chemistry steps can have a major effect on the fidelity. Apparently, as long as the barriers for the pre-chemistry steps (e.g., the open to closed conformational change) are significantly lower than the chemical barrier they cannot change the kinetics and the corresponding fidelity (except in some particular substrate saturation conditions). That is, the reaction rate is determined by the difference between the energy of the TS and the E + S state and the TS for the chemical step. Having many barriers between the open and closed configurations in the binding step of Figure 8 (or Fig. 1) is not going to change the kinetics as long as these barriers are smaller than the TS barrier. An additional insight is obtained for example from inspection of Figure 1 of Ref. 27. This figure describes the calculated barriers for the pre-chemistry steps. Since the barriers for W are smaller than that for R it is very hard to see how these barriers could account for the observed fidelity. The obvious answer is that the real difference is in the chemical step.

CONCLUDING REMARKS

We have explored the idea that the incorporation of incorrect nucleotides (dATP, dGTP, or dTTP opposite template dG) occurs in different protein conformations than incorporation of correct dCTP opposite template dG. It was found that the TS is likely to involve different conformational region of the protein. The structural changes that were consistent with the measured fidelity appear to be much larger than the relatively small rearrangements that follow common changes in the shape of the reacting system during the movement from the ground state to the TS in enzymatic reactions.

Since the conformational changes reduce the TS energy of W incorporation drastically they decrease fidelity rather than increase it. Overall, a better agreement with experimental data is attained when R is incorporated through a TS in a closed conformation and W is incorporated through a TS in one or perhaps several partially open conformations.

The semi-quantitative surfaces obtained in this work allowed us to discuss the induced fit effect in a consistent way. Previous discussions of this issue have been incomplete because they were not related to well defined structure-energy relationships. Although our discussion of the induced fit effect seems complex, we view it as an important step in illustrating that there is no way to predict the trend in fidelity by using the typical induced fit concept since it is ambiguous as to whether or not the fidelity might decrease or increase. In contrast, the calculations described in Figure 8 probe the actual effect of the structural change. These calculations can predict the consequence of the induced fit. Thus accepting that the induced fit concept does not explain fidelity, we can still investigate the effect of substrate induced structural changes on fidelity in specific cases, using computer simulation or experimental approaches.

We also considered the relationship between the landscape and dynamic effects and concluded that convincing evidence for these dynamical effects influencing the fidelity of DNA polymerases is lacking. Similarly we found that the barriers along the path from the open to closed structures in the pre-chemistry step are not likely to contribute to the overall fidelity, unless the corresponding barriers are higher than that for the chemical step.

The present study reflects a hypothetical search along the path between the open and closed N-subdomain conformations. We realize that this conformational space might not include the actual conformational changes that may occur in the reaction of W. However, it is possible that some of the conformers generated here have features in common with the actual structural changes, at least in terms of some key electrostatic interactions. Regardless of the exact nature of the structural changes, it appears from Table II and Figure 8 that the lowest activation barrier for correct base pairs occurs in the closed structure, whereas for the incorrect pairs there are lower barriers in other protein conformations.

In conclusion, our study indicates that the TS for W is likely to be at different conformational states than that of R. The relevant structural change may also involve a shift of the DNA rather than a change in the protein structure. However, the precise nature of the polymerase structural changes cannot be determined without additional X-ray data for ternary complexes containing base mispairs. The missing link would probably be structural data for mispaired TS analogues.

Abbreviations

EVB

empirical valence bond

MD

molecular dynamics

Pol β

DNA polymerase β

RMSD

root mean square deviation

TMD

targeted molecular dynamics