HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 25, 22662-22669, June 21, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/25/22662    most recent M200202200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Weiss, K. K. Articles by Kim, B. Search for Related Content PubMed PubMed Citation Articles by Weiss, K. K. Articles by Kim, B. Social Bookmarking                      What's this?

Mechanistic Role of Residue Gln¹⁵¹ in Error Prone DNA Synthesis by Human Immunodeficiency Virus Type 1 (HIV-1) Reverse Transcriptase (RT)

PRE-STEADY STATE KINETIC STUDY OF THE Q151N HIV-1 RT MUTANT WITH INCREASED FIDELITY^*

Kellie K. Weiss, Robert A. Bambara^§, and Baek Kim^¶

From the Departments of  Microbiology and Immunology and ^§ Biochemistry and Biophysics, University of Rochester Medical Center, Rochester, New York 14642

Received for publication, January 8, 2002, and in revised form, March 28, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

It has previously been reported that mutations in the Gln¹⁵¹ residue of human immunodeficiency virus type 1 (HIV-1) reverse transcriptase (RT) greatly enhance RT fidelity. In this study, we employed pre-steady state kinetic assays to elucidate the mechanistic role of residue Gln¹⁵¹ in highly error prone DNA synthesis by HIV-1 RT. Using our Q151N high fidelity mutant, which is structurally altered in its ability to interact with the 3'-OH on the sugar moiety of the incoming deoxynucleotide triphosphate (dNTP), we examined how this change in RT-dNTP interaction affects HIV-1 RT fidelity. First, we found the binding affinity (K_D) of wild type and Q151N RT proteins to different template/primers to be similar. These results indicate that the Gln¹⁵¹ residue is not involved in the formation of the binary complex (RT·template/primer) during DNA polymerization. We also found that by changing residue 151 from a GlnAsn, the maximum rate of dNTP incorporation (k_pol) for both correct and incorrect dNTPs was not affected. In contrast, the ability of the Q151N mutant to bind both correct and incorrect dNTPs (K_d) was diminished. The Q151N mutant was 120-fold less efficient at binding correct dNTP than wild type RT, and its decrease in binding was such that we were unable to measure the actual binding affinity of Q151N for incorrect dNTPs. Presumably, the fidelity increase observed during the steady state is explained by this defect in Q151N binding to incorrect dNTP. In wild type RT, residue Gln¹⁵¹ is important for tight binding of incorrect dNTPs and may contribute to the low fidelity nature of HIV-1 RT. Since the Q151N mutation also alters RT binding to correct dNTPs, the wild type Gln¹⁵¹ residue may play an important role in efficient binding of RT to correct dNTPs. Our findings suggest that residue Gln¹⁵¹ is an important element for the execution of both highly error prone and efficient DNA synthesis by HIV-1 RT.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

It is becoming more apparent that many organisms employ multiple DNA polymerases to replicate their genomes. Some of these DNA polymerases are specifically involved in error prone DNA synthesis required for either spontaneous mutagenesis or bypassing DNA damage (1-3). Recent biochemical studies of DNA polymerases  and  and their bacterial UmuC/DinB homologs (4-6) show that these are actually very low fidelity polymerases. It is possible that these organisms have evolved functionally diverse DNA polymerases for the different activities of genomic replication and mutagenesis. As demonstrated in a series of kinetic experiments with various DNA polymerases, it is clear that the ability to incorporate incorrect dNTPs¹ affects the efficiency of DNA synthesis (7). In other words, low fidelity and poor ability to discriminate between correct and incorrect dNTPs are detrimental to efficient DNA polymerization, which is likely essential for chromosomal DNA replication. The fact that efficient DNA synthesis and error prone DNA synthesis are kinetically at odds with one another may explain why possession of separate DNA polymerases specific for either replication or mutagenesis is beneficial.

Human immunodeficiency virus type 1 (HIV-1) has a single DNA polymerase called reverse transcriptase (RT). HIV-1 RT is one of the most error prone DNA polymerases involved in DNA replication (8-10). This low replication fidelity of HIV-1 RT and the resultant error prone DNA synthesis are a presumptive source of HIV-1 genomic hypervariation (9, 10). Although HIV-1 RT must efficiently synthesize DNA during viral genomic replication, it concomitantly produces genomic mutations in order to evolve and escape host immune selection. In contrast to organisms that employ multiple polymerases for the activities of genomic replication and mutagenesis, HIV-1 RT has been adapted to execute both of these functions. The fact that HIV-1 RT is able to resolve the kinetic issues associated with efficient but error prone DNA synthesis makes it a unique model to study in understanding the mechanistic and structural elements involved in replication fidelity.

DNA polymerization is an ordered reaction that consists of a series of sequential steps. First the polymerase must bind T/P to form a binary polymerase·T/P complex. Binding of the dNTP follows and the ternary polymerase·T/P·dNTP complex then undergoes conformational change and catalysis (7, 11). Pre-steady state kinetics examines the ability of the polymerase to bind and then incorporate dNTP. Using a rapid quench instrument, one can measure 1) the binding of dNTPs to the DNA polymerase (K_d) and 2) the maximum rate of dNTP incorporation (k_pol) (7, 11). Incorporation of incorrect dNTPs under pre-steady state conditions has been studied in order to understand the nature of mutation synthesis by DNA polymerases (12-14). These types of studies examining wild type HIV-1 RT fidelity demonstrate that HIV-1 RT differentiates between correct and incorrect dNTPs in both the binding (K_d) and incorporation (k_pol) steps by 250-fold and 9-80-fold, respectively (15). This suggests that events that occur during both dNTP binding and dNTP incorporation affect the accuracy of DNA synthesis by HIV-1 RT.

Studies involving mutants with altered fidelity have been invaluable in delineating the structural and biochemical determinants of replication accuracy. Examples are the pre-steady state kinetic studies performed with the Klenow fragment from Escherichia coli DNA polymerase I (14, 16). Structural analyses of Klenow fragment and results from random mutagenesis of Taq polymerase illustrate that interactions between the incoming dNTP and the dNTP binding domain, called the O helix, are important molecular determinants in overall polymerase fidelity (17-20). Like these studies, our analysis of high fidelity HIV-1 RT mutants will elucidate the mechanistic and structural role of wild type residues during DNA polymerization and mutation synthesis.

Recently, we reported that the Q151N mutant has 12.5 times higher fidelity than wild type HIV-1 RT based on the M13 lacZ forward mutation assay (21). Data from single nucleotide steady state kinetic assays using the Q151N mutant also substantiated our findings (22). Structural examination showed that the wild type Gln¹⁵¹ residue interacts with the 3'-OH on the sugar moiety of the incoming dNTP (21). It has been suggested that this interaction between the Gln¹⁵¹ residue and the 3'-OH of the incorrect dNTP stabilizes RT binding to the incorrect dNTP. This would allow for chemical DNA polymerization to occur even in the absence of base-pairing between the incoming incorrect dNTP and the template nucleotide. When residue 151 is altered from a GlnAsn, there is a loss in the interaction of Gln¹⁵¹ with the incoming incorrect dNTP as well as in the base-pairing between the template nucleotide and the incorrect dNTP. Failure of stable binding precludes incorporation of the incorrect dNTP and results in an increase in RT fidelity.

Interestingly, the Q151N mutation, like the Q151M viral mutation, has increased resistance to nucleoside RT inhibitors such as azidothymidine (23, 24). This finding suggests that the wild type Gln¹⁵¹ residue may interact with the azido group of the incoming AZTTP, assisting in the AZTTP incorporation reaction. Presumably, the Q151N and Q151M mutations might prevent binding of AZTTP to the active site of HIV-1 RT, conferring AZTTP resistance (21, 22). However, unlike the Q151N mutant, the fidelity of Q151M is the same as that of wild type HIV-1 RT (22). Even though Q151M has evolved a mechanism to discriminate against AZTTP, it is still able to bind incorrect dNTPs like wild type RT.

In this study, we investigated the mechanistic steps of DNA polymerization specifically affected by the Q151N HIV-1 RT high fidelity mutation. Specifically, we decided to examine the kinetics of DNA polymerization from an RNA template, which is an activity unique to reverse transcriptases. To do this we employed pre-steady state kinetic assays to identify alterations in dNTP binding or dNTP incorporation. We also used a double filter dot blot assay to assess changes in RT binding to the T/P. Our study reveals that residue Gln¹⁵¹ is a molecular element in HIV-1 RT infidelity. The role of residue Gln¹⁵¹ is to promote mutation synthesis by increasing RT binding to incorrect dNTPs.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Chemicals and HIV-1 RT Proteins-- N-terminal end His tagged wild type and Q151N HIV-1 RT proteins were purified as previously described (21, 25). With our protocol, we were able to purify 2 mg of RT proteins with >95% purity from a 1-liter culture. All dNTPs were purchased from Amersham Biosciences. DNA oligonucleotides were purchased from Invitrogen, and RNA template was synthesized by Dharmacon. Primers were labeled with [-³²P]ATP (Amersham Biosciences) and T4 nucleotide kinase (New England BioLabs).

Pre-steady State Kinetic Assay-- Pre-steady state burst and single turnover experiments were employed to examine the transient kinetics associated with a single nucleotide incorporating onto the 3' end of a ³²P-labeled 17-mer A primer (5'-CGCGCCGAATTCCCGCT-3') annealed to a 40-mer RNA template (5'-AAGCUUGGCUGCAGAAUAUUGCUAGCGGGAAUUCGGCGCG-3') (11, 15). For both sets of reactions, we used 20 µl of T/P preincubated with purified HIV-1 RT protein and reaction buffer (25 mM Tris-HCl, pH 8.0, 40 mM KCl, 2 mM dithiothreitol, 5 mM MgCl₂, and 0.1 mg/ml bovine serum albumin). This mixture was injected into one sample tube of the rapid quench machine (Kintek). An equal volume of dNTP preincubated with Mg²⁺ (10 mM) was injected into the other sample tube. The polymerization reaction was initiated by rapidly mixing the two reactants and terminated by adding 0.25 M EDTA at different time points.

In the pre-steady state burst experiments, T/P (150 nM) was present in excess of RT (~50 nM) and the reaction was initiated by the addition of 400 µM dNTP. These experiments were used to determine the active site concentrations of the RT proteins (see data analysis; Refs. 7 and 11). The pre-steady state single turnover experiments were used to determine the dNTP concentration dependence of the purified HIV-1 RT proteins. In the presence of varying dNTP concentrations (in the range of 600 nM to 2.5 mM), RT (100 nM) was used in slight excess of T/P (90 nM). In reactions involving incorrect dNTPs, the experiments were performed manually at longer periods of time and used a higher concentration of RT (700 nM) (15, 26).

Product Analysis-- The reactions were analyzed by 14% denaturing sequencing gel electrophoresis. The extended product in each reaction was quantified with the Cyclone phosphorimager (PerkinElmer Life Sciences).

Data Analysis-- Pre-steady state kinetic data were analyzed using nonlinear regression. Equations were generated with the KaleidaGraph program, version 3.51 (Synergy Software). Data points obtained during the burst experiment were fitted to the burst equation shown below (7, 11).

(Eq. 1)

The value A is the amplitude of the burst, which reflects the actual concentration of enzyme that is in active form. k_obs is the observed first-order rate constant for dNTP incorporation, whereas k_ss is the observed steady state rate constant (11, 15, 26). Data from single-turnover experiments were fit to a single exponential equation that measures the rate of dNTP incorporation (k_obs) per given dNTP concentration ([dNTP]). These results can then be used to determine K_d, the dissociation constant for dNTP binding to the RT·T/P binary complex. This was done by fitting the data to the following hyperbolic equation.

(Eq. 2)

From this equation, we could then identify the kinetic constants for each RT during pre-steady state kinetics: k_pol is the maximum rate of dNTP incorporation, and K_d is equilibrium dissociation constant for the interaction of dNTP with the E·DNA complex (11, 12).

Double Filter Dot Blot Assay for K_D-- We employed a standard assay protocol previously established to determine the K_D of wild type and mutant RTs on three different T/Ps: 17-mer/40-mer RNA template, 17-mer/18-mer DNA template, and 18-mer/18-mer blunt end T/P. A protein binding filter (top filter, nitrocellulose; Schleicher & Schuell), nucleic acid binding filter (bottom filter, DEAE; Schleicher & Schuell), and dot blot system (Schleicher & Schuell) were prepared as described (27, 28). RT proteins (50 nM active site concentration) were incubated with different concentrations (10-800 nM, 20 µl) of 5' ³²P-labeled T/Ps at 37 °C for 3 min. The RT and T/P mixtures were applied to each well of the dot blot system and washed twice with 100 µl of reaction buffer. After drying both filters, the filters were analyzed using a phosphorimager, and the percentage binding at each T/P concentration was quantitated. The data were then fitted to the binding curve equation previously described (see Eq. 3 below; Ref. 28).

(Eq. 3)

The variables RT-T/P, K_D, RT_t, and T/P reflect productive RT-template concentration, equilibrium dissociation constant for RT binding to T/P, active RT concentration (see Fig. 1), and total T/P concentration, respectively. From Eq. 3, the K_D values of wild type and mutant RTs to these three T/Ps were determined (28).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Initial Burst and Active Site Concentration of HIV-1 RT Proteins-- We determined the active site concentrations of the wild type and Q151N HIV-1 RT proteins (Fig. 1). For this, the single nucleotide (dATP) incorporation reaction was performed using a T/P concentration in molar excess over the enzyme concentration. We observed both single turnover events associated with the pre-steady state burst and multiple rounds of DNA polymerization occurring during steady state kinetics. Reactions ranging from 5 ms to 2 s were performed using the rapid quench machine.

View larger version (10K): [in this window] [in a new window]   Fig. 1.   Pre-steady state and steady state kinetics of HIV-1 RT wild type and Q151N proteins. Pre-steady state kinetics of incorporation of dATP with wild type (A) and Q151N (B) HIV-1 RT proteins. Pre-steady state kinetics of incorporation of dAMP onto a 32P-labeled 17-mer primer annealed to the 40-mer RNA template by wild type and Q151N HIV-1 RTs were measured. Reactions were initiated by mixing a preincubated solution of RT and T/P (150 nM) with dATP (400 µM) and Mg2+ (10 mM) under rapid quench conditions. The reactions were quenched at the indicated time (5 ms to 2 s) and analyzed by 14% denaturing gel. The RT·T/P concentrations that were equal to the burst heights, indicating active site concentrations, were determined by extrapolation to the y axis (7, 12). The solid line represents the best fit of the data to a burst equation with an amplitude A (active site concentration) = 63 ± 4 nM (wild type), and 46 ± 5 nM (Q151N). The observed first-order rate constant for the burst phase kobs = 17.3 × 103 ± 2 × 103 ms1 (wild type) and 9.0 × 103 ± 2 × 103 ms1 (Q151N), and the observed rate constant for the linear phase kss = 0.09 × 103 ± 0.05 × 103 ms1 (wild type) and 0.6 × 103 ± 0.1 × 103 ms1 (Q151N).

From Eq. 1, we can determine the active site concentration, the rate of dNTP incorporation under pre-steady state conditions (k_obs), and the rate of dNTP incorporation during the steady state (k_ss) for both the wild type and Q151N RT proteins at 400 µM dATP. In these initial burst experiments, we used 100 nM wild type and 200 nM Q151N protein. Our results indicate that only 63 nM (63%) wild type RT and 46 nM (23%) Q151N were actually active (Fig. 1). For the wild type RT, k_obs was 17.3 × 10³ ms¹, and k_ss was 0.09 × 10³ ms¹. For the Q151N protein, the values for k_obs and k_ss were 9.0 × 10³ and 0.6 × 10³ ms¹, respectively (Fig. 1). These differences in k_obs and k_ss suggest that reaction rates for both proteins are faster under pre-steady state kinetic conditions (k_obs) than steady state kinetic conditions (k_ss), which has been previously demonstrated (7, 15). Presumably, the lower rate in the steady state is due to the fact that in this time scale, reactions include multiple rounds of dNTP incorporation and involve various complex steps that may be rate determining (i.e. T/P binding and product release). Surprisingly, the steady state rate of catalysis (k_ss) for the Q151N protein is 6-fold higher than that of wild type RT. It is possible that the Q151N mutation may overcome kinetic barriers that restrict wild type DNA polymerization during the steady state.

Pre-steady State Incorporation of Correct dATP by HIV-1 RT Proteins-- Next, we determined the binding affinity (K_d) of both the wild type and Q151N mutant HIV-1 RT proteins for correct dATP on our T/P (Table I). Concurrently, we measured the maximum rate at which wild type and mutant RTs incorporated correct dATP (k_pol). By analyzing the dependence of reaction rate (k_obs) on dNTP concentration (Fig. 2), we are able to calculate K_d and k_pol (Eq. 2). Our wild type HIV-1 RT pre-steady state kinetic values when wild type is using dATP are similar to those previously published with dCTP as the correct dNTP (12, 15). When we examined the Q151N protein, we observed a 120-fold decrease in binding affinity (increase in K_d) compared to wild type HIV-1 RT, whereas we found only a slight increase (1.6-fold) in the maximum rate of correct dNTP incorporation (k_pol; Table I). This suggests that the Q151N mutation specifically reduces the initial binding affinity of RT for the incoming correct dATP. A consequence of this reduction in initial binding affinity (increase in K_d) for dATP is a 75-fold reduction in the pre-steady state incorporation efficiency (k_pol/K_d) of Q151N relative to wild type RT (Table I). It is apparent that the rate of correct dATP incorporation for the Q151N mutant differs with dNTP concentration. Compared to wild type RT, which maximally incorporates correct dATP (k_pol) at 25 µM, the rate of dATP incorporation (k_pol) for the Q151N mutant at 25 µM is much less that that of wild type (Fig. 2). It is possible that at low dNTP concentrations, the step of DNA polymerization involving the Gln¹⁵¹ residue, likely initial dNTP binding, becomes rate limiting. At high correct dATP concentration, initial dNTP binding is not the rate-limiting step. Q151N actually has a faster rate of dNTP incorporation than wild type RT (k_obs in Fig. 1 and k_pol in Fig. 2).

View larger version (9K): [in this window] [in a new window]   Fig. 2.   dATP dependence of wild type and Q151N HIV-1 RT proteins. kobs values of wild type (A) and Q151N (B) proteins were determined at different dATP concentrations. Higher dATP concentrations were used for the Q151N protein reactions due to its Kd increase. The kpol and Kd values were calculated by fitting these results to Eq. 2 (see "Materials and Methods").

Pre-steady State Incorporation of Incorrect dNTPs by HIV-1 RT Proteins-- Next, we measured the kinetic parameters associated with the incorporation of the three incorrect dNTPs (dGTP, dCTP, and TTP) (Table I). Since incorporation of incorrect dNTPs is very slow and inefficient, we performed the reactions manually and at high concentrations of RT protein and incorrect dNTP (see "Materials and Methods" and Ref. 12). Analysis of the Q151N mutant is complicated because the Q151N reaction rate (k_obs) increases in a linear fashion with up to 2.5 mM incorrect dNTPs. Although saturating concentrations of dNTP are >2.5 mM, higher dNTP concentrations (>2.5 mM) cannot be used because they inhibit RT activity (reducing k_obs; Ref. 13). Because of this technical difficulty, we could only estimate minimum k_pol values for Q151N (i.e. k_obs at 2.5 mM incorrect dNTP). Additionally, we can conclude that the K_d for Q151N with all incorrect dNTPs is apparently higher than 2.5 mM.

In the scenario with incorrect dGTP, the binding affinity of wild type HIV-1 RT is at least 75-fold lower than its binding affinity for correct dATP (increase in K_d; Table I). The maximum rate at which wild type RT incorporates incorrect dGTP is 4 × 10⁵ times slower than when it incorporates correct dATP (k_pol). Similar changes in K_d and k_pol for wild type HIV-1 RT have also been observed in previous experiments using correct dCTP and incorrect dNTPs (12, 15). The efficiency of incorrect dGTP incorporation (k_pol/K_d) by wild type RT on our T/P, also termed misincorporation efficiency, was 4.4 × 10⁵ µM¹ s¹ (Table I). In contrast, the Q151N mutant is diminished in its ability to bind incorrect dGTP a minimum of 14-fold (increase in K_d), with only a slight increase in its rate of dGTP incorporation (k_pol) when compared with wild type RT. Similar to the case with correct dNTP, the Gln¹⁵¹ residue appears to be involved mainly in the initial binding of RT to incorrect dNTP, and not in the incorporation (k_pol) of incorrect dNTPs.

Experimentally, we cannot obtain a true value for Q151N misincorporation efficiency (k_pol/K_d). The values for k_pol and K_d are determined by assessing the dependence of reaction rate on dNTP concentration (Eq. 2). However, Q151N incorporation of incorrect dNTPs at every dNTP concentration tested was very low. The reaction rates with incorrect dNTPs (up to 2.5 mM) were linear with respect to dNTP concentration (data not shown), and as such, we cannot measure the values of k_pol and K_d for Q151N. In the case of incorrect dGTP (Table I), if we estimate the k_pol of the mutant RT to be similar to that of wild type RT and the K_d of the mutant to be higher than 2.5 mM (Table I), then the misincorporation frequency should be an order of magnitude lower than that of wild type RT. For incorrect dCTP the K_d value is likely so high that the misincorporation efficiency of Q151N is lower than wild type RT. Further supporting the high fidelity nature of Q151N, incorrect TTP incorporation was negligible at our sensitivity of measurement (Fig. 3).

View larger version (55K): [in this window] [in a new window]   Fig. 3.   Pre-steady state incorporation of incorrect TTP by wild type and Q151N HIV-1 RT proteins. The 32P-labeled 17-mer primer (S) annealed to the 40-mer RNA template was used in this assay. The T/P (100 nM) bound to RT proteins (700 nM) was incubated, and the reaction was initiated by the addition of incorrect TTP (750 µM for wild type and 2.5 mM for Q151N). Aliquots were taken at different time points (30 s to 10 min), and the reactions were analyzed by 14% denaturing gel. TTP incorporation (E) by wild type HIV-1 RT was used to determine the pre-steady state parameters of wild type RT using incorrect TTP (Table I).

Keeping in mind that the actual Q151N misincorporation efficiency is likely even lower than our estimated value, we see that these differences are translated into a diminished rate of misincorporation by Q151N during steady state DNA synthesis. In the M13 lacZ forward mutation assay, which measures polymerase fidelity under multiple rounds of primer extension, we observe a fidelity change of 12.5 for the Q151N mutant relative to wild type RT (21). Additionally, single nucleotide steady state kinetic analyses show that the Q151N mutant is 8-26-fold less efficient at incorporating incorrect dNTP than wild type RT (22). In other words, pre-steady state kinetic changes in K_d for the Q151N mutant with incorrect dNTPs appear to directly affect RT fidelity as determined under steady state conditions. This finding suggests that initial binding to the incorrect dNTP is a rate limiting step during mutation synthesis.

When using incorrect dCTP, wild type HIV-1 RT has at least a 150-fold lower binding affinity (higher K_d) and a 3 × 10⁵-fold slower rate of dNTP incorporation (k_pol) than when using correct dATP (Table I). When we examine Q151N incorrect dCTP pre-steady state kinetic parameters, we observe a minimum 7-fold decrease in binding affinity (K_d) and a 7-fold increase in maximum rate of dNTP incorporation (k_pol) as compared to wild type. It is clear that residue Gln¹⁵¹ has a role in the initial binding step of HIV-1 RT to incorrect dNTPs. Because diminished binding to incorrect dCTP is somewhat offset by the increased rate at which the Q151N mutant incorporates incorrect dCTP, the misincorporation efficiency of Q151N for incorrect dCTP is similar to that seen in wild type HIV-1 RT (3.0-3.1 × 10⁵ µM¹ s¹). It is likely that Q151N is still less efficient at incorporating incorrect dCTP than wild type. Since this is actually a high end estimate of misincorporation frequency; however, the reduction in misincorporation of dCTP is probably less than that seen for dGTP due to the increase in rate of dCTP incorporation (k_pol) by the Q151N mutant.

Data for the last incorrect dNTP, TTP, could not be obtained. As shown in Fig. 3, even at the highest TTP concentration (2.5 mM) and the longest incubation time (10 min) there was no detectable extension of incorrect TTP by the Q151N mutant. This suggests that the Q151N misincorporation efficiency for TTP must be extremely low, which corresponds with the high fidelity nature of the Q151N mutant.

Binding Constants of RT Proteins to T/Ps-- We also tested whether the Q151N mutation affects RT binding to T/P (Table II). For this we determined the binding affinity of wild type and Q151N RT proteins to three different T/Ps: 1) the 18-mer/40-mer RNA T/P used in our pre-steady state experiments, 2) a 17-mer/18-mer DNA T/P, and 3) an 18-bp blunt end DNA T/P, as measured by the double filter dot blot assay (27, 28). Binding curves and binding constants of the RT proteins were obtained using the same active site concentrations as determined in the pre-steady state kinetic assay described previously. In this experiment, we found that the wild type and Q151N RT proteins have very similar binding affinity to all three T/Ps (Table II). However, the K_Ds of these two proteins to the blunt end T/P are higher than those to the 18-mer/40-mer RNA T/P. This indicates that both the wild type and Q151N RT proteins bind a 3' recessed T/P better than a blunt ended T/P. These findings support the likelihood that residue Gln¹⁵¹ is not involved in the DNA polymerization step of RT binding to T/P (i.e. formation of the RT·T/P binary complex).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Pre-steady state kinetic studies have shown that the ability of DNA polymerases to distinguish between correct and incorrect dNTPs affects the efficiency of DNA polymerization (7). More specifically, binding and incorporating incorrect dNTPs are inhibitory to DNA synthesis. HIV-1 RT is a viral polymerase that executes the tasks of viral mutagenesis and genomic replication. It is able to incorporate incorrect nucleotides so as to introduce mutations into the viral genome, but it also readily incorporates correct nucleotides to efficiently synthesize DNA. The fact that HIV-1 RT simultaneously performs both these functions makes it a unique model to examine for understanding the mechanistic and structural elements involved in replication fidelity. Our study suggests that HIV-1 RT has evolved residue Gln¹⁵¹ as a mechanism to resolve this kinetic issue.

We and others have demonstrated that mutations in the Gln¹⁵¹ residue increase HIV-1 RT fidelity as measured by steady state fidelity assays (21, 22). Structural modeling suggests that the Gln¹⁵¹ residue of HIV-1 RT directly interacts with the 3'-OH on the sugar moiety of the incoming dNTP (21, 29). Since residue 151 interacts with the incoming dNTP, it is likely that changes in the steps of DNA polymerization involving RT-dNTP interactions account for the increased fidelity in the Q151N HIV-1 RT mutant. To identify mechanistic steps affected by the Q151N HIV-1 RT high fidelity mutation, we employed pre-steady state kinetic assays that specifically examine the steps of dNTP binding and dNTP incorporation (7). By comparing the pre-steady state kinetic parameters of the wild type and Q151N HIV-1 RT proteins, we found that residue Gln¹⁵¹ may contribute to HIV-1 RT infidelity by improving RT binding to incorrect dNTPs. Our filter binding assay results also substantiated the idea that Gln¹⁵¹ is specifically important for RT interaction with the incoming dNTP. We measured the binding affinity (K_D) of the wild type and mutant RT proteins to three different types of T/P and found that the Q151N mutation does not affect RT-T/P binding.

Binding affinity as a mechanism of fidelity is founded on the premise that there exists a difference in polymerase binding to correct and incorrect dNTPs. For example, the initial binding affinities (K_d) of the E. coli Klenow fragment to correct and incorrect dNTPs are similar. As such, the polymerase is not able to distinguish between correct and incorrect dNTPs. In the Klenow fragment, initial dNTP binding is not a mechanistic determinant of fidelity (14, 16). In contrast, initial binding of the T7 DNA polymerase to dNTPs does contribute to its low misincorporation efficiency. The binding affinity of the T7 DNA polymerase for incorrect dNTP is much lower than that for correct dNTP (30). Like the T7 DNA polymerase, HIV-1 RT differs in its ability to bind correct and incorrect dNTPs (Table I; Ref. 15); however, the difference is that HIV-1 RT (micromolar K_d) binds incorrect dNTPs better than the T7 DNA polymerase (millimolar K_d; Ref. 30). As a result, HIV-1 RT has lower fidelity than the T7 DNA polymerase.

When residue 151 is altered from a GlnAsn, there is a significant decrease in the mutant binding affinity (increase in K_d) for incorrect dNTPs. As described under "Results," the K_d values for Q151N with incorrect dGTP and dCTP are at least 2.5 mM. Since higher dNTP concentrations actually inhibit DNA polymerization, we can only make a high end estimate for binding affinity (minimum K_d) to provide a reference point in our discussions (13). True K_d values are higher than 2.5 mM, meaning that the binding affinity of the Q151N mutant for incorrect dGTP and dCTP is actually lower than what we describe. The binding affinity (K_d) for Q151N with TTP could not be determined because no detectable primer extension was observed even at high concentrations of TTP and with longer incubation periods (Fig. 3). It is likely that the K_d value for Q151N with incorrect TTP is much higher than those of Q151N binding to incorrect dGTP and dCTP.

Not surprisingly, since structural data of HIV-1 RT complexed with T/P and TTP show that the side chain of residue 151 interacts with the 3'-OH on the sugar moiety of correct TTP, we also observe a change in Q151N binding affinity for correct dNTP (29). Under pre-steady state conditions, binding of the Q151N RT to correct dATP is dramatically reduced (120-fold increase in K_d). Like binding to incorrect dNTPs, the dramatic reduction in binding affinity (increase in K_d) observed in the Q151N mutant suggests that wild type residue Gln¹⁵¹ is a major determinant for correct dNTP binding. Although binding of Q151N to correct dNTP is diminished, we must keep in mind the magnitude of this change. Whereas we estimate that the K_d for incorrect dNTPs is a minimum of 2.5 mM, the K_d for Q151N with correct dATP on our T/P is 293 µM. In actuality, Q151N binds incorrect dNTPs much less efficiently than it binds correct dNTPs.

As mentioned earlier, efficient enzymatic DNA polymerase activity requires efficient discrimination between correct and incorrect dNTPs. Tight binding to incorrect dNTPs or poor discrimination against incorrect dNTPs may actually reduce the efficiency of DNA polymerization (7, 13). Even though wild type residue Gln¹⁵¹ may bind incorrect dNTPs and promote highly error prone DNA synthesis, a consequence could be inefficient viral replication. On the other hand, HIV-1 RT is still able to discriminate between correct and incorrect dNTPs because of the large difference in Gln¹⁵¹ binding affinity for these nucleotides. This relatively large binding difference between correct and incorrect dNTPs is likely a key element that maintains the balance between highly error prone DNA polymerization and highly efficient viral replication by HIV-1 RT.

The kinetic fidelity of a DNA polymerase is usually measured by the ratio of k_pol/K_d between incorrect and correct dNTPs. This defines the actual capability of the polymerase to discriminate between correct and incorrect dNTPs during DNA polymerization (12, 15). In this study, we could not determine the actual discrimination efficiency of the Q151N mutant because the binding affinity (K_d) of the mutant for incorrect dNTP could not be experimentally determined. We were able to make a high end estimate for the efficiency of incorrect dGTP incorporation (k_pol/K_d), or the misincorporation efficiency of dGTP. If we compare the misincorporation efficiency of incorrect dGTP (k_pol/K_d) between wild type and mutant RT, we see that the Q151N mutant is at least 6-fold less efficient at incorporating incorrect dGTP than wild type RT. Since Q151N is diminished in its ability to bind dCTP and TTP, it is likely that Q151N is also less efficient at misincorporating these nucleotides. Interestingly, a reduction in the error rate of the Q151N mutant was also observed in steady state fidelity assays (21, 22). It is possible that reduction in RT binding to incorrect dNTPs affects both the pre-steady state rate of misincorporation and the overall steady state rate of mutation synthesis.

The structure of a transient HIV-1 RT ternary complex with an incorrect dNTP (RT·incorrect dNTP·T/P) is not currently available. To better understand the mechanism of residue Gln¹⁵¹ in RT fidelity, we previously generated a working model based on structural data of HIV-1 RT complexed with a dNTP (29). As seen in Fig. 4, correct positioning of the incoming TTP (black) in the dNTP binding pocket of HIV-1 RT is likely engineered by various interactions between the dNTP binding residues and the incoming dNTP. Wild type Gln¹⁵¹ residue (light green) is positioned such that it makes a nonspecific interaction with the 3'-OH on the ribose of the incoming TTP. The nearby Arg⁷³ residue (gray) assists in Gln¹⁵¹ binding with the incoming TTP by stabilizing the placement of Gln¹⁵¹ in the active site. It is possible that improper base-pairing between the template nucleotide (blue-green) and the incoming incorrect dNTP distorts the geometric orientation of the incorrect dNTP in the RT active site. If the side chain of the wild type Gln¹⁵¹ residue is flexible enough, the interaction between the Gln¹⁵¹ residue and the 3'-OH of the incoming incorrect dNTP may still be made, as suggested by the low K_d (micromolar range) of wild type to incorrect dNTPs. This interaction alone may be able to secure binding of the incorrect dNTP to the dNTP binding pocket, allowing the open ternary complex (RT·dNTP·T/P) to undergo conformational change and catalysis. In contrast, when the Q151N mutant (orange) incorporates incorrect dNTPs, there is an absence in 1) interaction between RT and the incoming incorrect dNTP and 2) base-pairing between the template nucleotide and the incoming incorrect dNTP. Incorrect dNTPs may not stably bind to the dNTP binding pocket of the Q151N RT, which would account for the extremely high K_d (unmeasurable micromolar range) of Q151N to incorrect dNTPs. Consequently, HIV-1 RT fidelity would increase.

View larger version (172K): [in this window] [in a new window]   Fig. 4.   Interaction of HIV-1 RT residue 151 with the incoming dNTP. This figure shows the interaction between wild type residue Gln151 (light green) or mutant residue N151 (orange) and the incoming TTP (black; Ref. 29). This model of HIV-1 RT complexed with a template (yellow), primer (purple), and incoming dNTP was generated previously with the programs Molscript (32) and Raster3D (33) using coordinate set 1rtd.pdb from the Protein Data Bank (21, 34). Three interactions, shown as dotted black lines, are depicted: 1) base pairing between the template nucleotide and the incoming TTP, 2) nonspecific interaction between wild type Gln151 and the 3'-OH on the sugar moiety of the TTP, and 3) nonspecific interaction between wild type Gln151 and the nearby Arg73 (gray) residue. Mutant residue N151 has a shortened R side chain and cannot interact with the sugar on the incoming TTP. Other residues that make up the HIV-1 RT active site are drawn in space-filling representation. The finger domain is shown in dark gray, and the palm domain is present as light gray space-filling residues.

We recently isolated another high fidelity RT mutant obtained from an in vivo simian immunodeficiency virus molecular clone (SIVMNE170; Ref. 31). We found² that a mutation in residue 148, ValIle, is responsible for the increased fidelity of this simian immunodeficiency virus RT mutant. Structural examination reveals that residue Val¹⁴⁸ lies near residue Gln¹⁵¹ on the 8 region of RT. Presumably, the mechanism by which V148I increases RT fidelity is via disrupting the interaction between residue Gln¹⁵¹ and the 3'-OH of the incoming dNTP. In wild type RT, the side chain of residue 148 makes contact with an opposing peptide backbone between residues 117 and 118. The V148I mutant, which has a longer side chain than the wild type (Val) residue, may push the 8 region (including residue Gln¹⁵¹) away from the RT active site. If V148I modulates the positioning of Gln¹⁵¹, we can predict that V148I will have altered enzymatic activity similar to that observed for the Q151N mutant (i.e. fidelity and incorporation of nucleotide analogs; Refs. 21 and 22).

Another interesting facet to the Q151N mutant is that at high dNTP concentrations during both the pre-steady state and steady state, Q151N is more efficient at incorporating correct nucleotides than wild type RT (k_pol in Table I; k_ss in Fig. 1). These results indicate that in wild type RT, the interaction between residue Gln¹⁵¹ and the incoming dNTP is inhibitory to the steps of DNA polymerization. One possible explanation for this is the energetic cost associated with breaking the interaction between Gln¹⁵¹ and the incoming dNTP. After binding the dNTP, RT must undergo a conformational change before it catalyzes the incorporation of the dNTP. During the conformational change from the open ternary complex to the chemically competent closed complex (RT·T/P·dNTP  RT*·T/P·dNTP), the interaction between residue Gln¹⁵¹ and the dNTP may be disrupted. In the Q151N mutant, there is no energetic cost associated with breaking this nonspecific interaction. Consequently, the conformational change from an open to a closed ternary complex is accelerated, resulting in an increase in k_pol. Therefore, despite the fact that residue Gln¹⁵¹ is essential for error prone DNA synthesis by HIV-1 RT, the presence of Gln¹⁵¹ actually decreases the activity of wild type RT. It is possible that this reduction in activity is an expense that wild type HIV-1 RT has to pay to be a highly error prone DNA polymerase. Because we also observed increases in k_pol for the Q151N mutant when using incorrect dNTPs (Table I), a disruption in the interaction between residue Gln¹⁵¹ and the 3'-OH of the incorrect dNTP may also occur during the conformational change of dNTP misincorporation.

Other studies that employed pre-steady state kinetic analysis to assess HIV-1 RT activity examined the M184V HIV-1 RT mutant, which is a 3TC resistant mutant with a slight increase in fidelity (12). This kinetic study showed that the M184V mutation slightly lowers HIV-1 RT misincorporation efficiency. Due to the fact that the fidelity difference between the mutant and wild type is minimal, it is not clear whether the M184V mutation affects the ability of RT to bind or incorporate incorrect dNTPs. What is unique about the M184V mutation is that it specifically reduces RT binding to 3TCTP, but not to natural dNTPs. The decrease in binding to 3TCTP is what renders the M184V resistant to 3TC. Mutations in the Gln¹⁵¹ residue, which is a residue that interacts with the 3'-OH on the sugar of the incoming dNTP, makes HIV-1 RT resistant to AZTTP (21, 22). Gln¹⁵¹ HIV-1 RT mutants may lose the interaction with the 3' azido group on AZT, and the result is a reduction in RT binding to the incoming AZTTP. Like the evolution of the M184V mutant, it is apparent that HIV-1 RT is able to alter residue Gln¹⁵¹ in such a way that the overall fidelity of the polymerase is not altered, as seen in the low fidelity AZT-resistant Q151M mutant (22).

In summary, the Q151N mutation significantly affects the K_d of both correct and incorrect dNTPs. These changes in K_d for Q151N with both correct and incorrect dNTPs greatly reduce the efficiency at which Q151N incorporates these nucleotides (k_pol/K_d). This suggests that wild type residue Gln¹⁵¹ promotes RT binding to correct and incorrect dNTPs, which precedes the incorporation of these nucleotides. Although we were unable to measure the initial binding affinity of our Q151N mutant with incorrect dNTPs due to the large increase in K_d, we can comfortably assume that residue Gln¹⁵¹ has to be an important molecular element in HIV-1 RT infidelity. It is also likely that this ability of residue Gln¹⁵¹ to facilitate the binding of RT to incorrect dNTPs contributes to the rapid evolution of HIV. Furthermore, it appears that the interaction between residue Gln¹⁵¹ and correct dNTPs has a role in efficient DNA synthesis by HIV-1 RT. The relatively large binding differences between correct and incorrect dNTPs may allow Gln¹⁵¹ to discriminate against incorrect dNTPs. This ensures that the DNA polymerase active site is functional and allows for viral genomic replication. Our study suggests that HIV employs residue Gln¹⁵¹ to resolve the paradoxical issue of rapid viral evolution versus efficient viral replication.

	FOOTNOTES

^* This work was supported by Grants GM55500 (to B. K.) and GM29573 (to R. A. B.) and Training Grant AI07362-12 (to K. K. W.) from the National Institutes of Health.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^¶ To whom correspondence should be addressed: Dept. of Microbiology and Immunology, University of Rochester Medical Center, 601 Elmwood Ave., Box 672, Rochester, NY 14642. Tel.: 585-275-6916; Fax: 585-473-9573; E-mail: baek_kim@urmc.rochester.edu.

Published, JBC Papers in Press, April 1, 2002, DOI 10.1074/jbc.M200202200

² T. L. Diamond, K. Y. Lee, J. Kimata, and B. Kim, unpublished data.

	ABBREVIATIONS

The abbreviations used are: dNTP, deoxynucleotide triphosphate; HIV-1, human immunodeficiency virus type 1; RT, reverse transcriptase; T/P, template/primer; AZT, azidothymidine; AZTTP, azidothymidine triphosphate; 3TC, ()-2',3'-dideoxy-3'-thiacytidine; 3TCTP, ()-2',3'-dideoxy-3'-thiacytidine triphosphate.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				V. K Jamburuthugoda, J. M. Santos-Velazquez, M. Skasko, D. J. Operario, V. Purohit, P. Chugh, E. A. Szymanski, J. E. Wedekind, R. A. Bambara, and B. Kim Reduced dNTP Binding Affinity of 3TC-resistant M184I HIV-1 Reverse Transcriptase Variants Responsible for Viral Infection Failure in Macrophage J. Biol. Chem., April 4, 2008; 283(14): 9206 - 9216. [Abstract] [Full Text] [PDF]

				C. A. Howell, C. M. Kondratick, and M. T. Washington Substitution of a residue contacting the triphosphate moiety of the incoming nucleotide increases the fidelity of yeast DNA polymerase {zeta} Nucleic Acids Res., March 1, 2008; 36(5): 1731 - 1740. [Abstract] [Full Text] [PDF]