JBC INTERFERin siRNA transfection reagent

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M308286200 on December 29, 2003

J. Biol. Chem., Vol. 279, Issue 12, 11834-11842, March 19, 2004
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
279/12/11834    most recent
M308286200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Gardner, A. F.
Right arrow Articles by Jack, W. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Gardner, A. F.
Right arrow Articles by Jack, W. E.

Comparative Kinetics of Nucleotide Analog Incorporation by Vent DNA Polymerase*

Andrew F. Gardner{ddagger}, Catherine M. Joyce§, and William E. Jack{ddagger}||

From the {ddagger}New England Biolabs Inc., Beverly, Massachusetts 01915 and the §Department of Molecular Biophysics and Biochemistry, Yale University, New Haven, Connecticut 06520

Received for publication, July 29, 2003 , and in revised form, December 22, 2003.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Comparative kinetic and structural analyses of a variety of polymerases have revealed both common and divergent elements of nucleotide discrimination. Although the parameters for dNTP incorporation by the hyperthermophilic archaeal Family B Vent DNA polymerase are similar to those previously derived for Family A and B DNA polymerases, parameters for analog incorporation reveal alternative strategies for discrimination by this enzyme. Discrimination against ribonucleotides was characterized by a decrease in the affinity of NTP binding and a lower rate of phosphoryl transfer, whereas discrimination against ddNTPs was almost exclusively due to a slower rate of phosphodiester bond formation. Unlike Family A DNA polymerases, incorporation of 9-[(2-hydroxyethoxy)methyl]X triphosphates (where X is adenine, cytosine, guanine, or thymine; acyNTPs) by Vent DNA polymerase was enhanced over ddNTPs via a 50-fold increase in phosphoryl transfer rate. Furthermore, a mutant with increased propensity for nucleotide analog incorporation (VentA488L DNA polymerase) had unaltered dNTP incorporation while displaying enhanced nucleotide analog binding affinity and rates of phosphoryl transfer. Based on kinetic data and available structural information from other DNA polymerases, we propose active site models for dNTP, ddNTP, and acyNTP selection by hyperthermophilic archaeal DNA polymerases to rationalize structural and functional differences between polymerases.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
All free living organisms encode several DNA polymerases that are jointly responsible for the replication and maintenance of their genomes, thereby ensuring accurate transmission of genetic information (1-3). The majority of identified DNA polymerases can be classified into Families A, B, C, and Y according to amino acid sequence similarities to Escherichia coli polymerases I, II, III, and IV/V, respectively (4, 5). Additional families have been identified, including the two-subunit replicative DNA polymerases from hyperthermophilic Archaea (Family D) (6) and eukaryotic DNA polymerase {beta} and terminal transferases (Family X) (4).

Structural and kinetic analyses of Family A (7-14) and Family B (15-25) DNA polymerases have increased the understanding of nucleotide selection and incorporation mechanisms. Although amino acid sequences diverge between these two families, the structures of Family A and B DNA polymerases share recognizable finger, thumb, and palm subdomains that allow comparison of structural elements important for function (3, 11). In the case of Family A DNA polymerases from bacteriophage T7, Escherichia coli (Klenow fragment, large fragment of DNA polymerase I), and Thermus aquaticus, as well as the Family B DNA polymerase from bacteriophage RB69, interpretation of the structural information is complemented by steady-state and pre-steady-state kinetic studies, allowing a detailed description of the polymerization pathway. Reaction parameters describing the discrimination against naturally occurring nucleotide analogs encountered in vivo, such as NTPs, or unnatural nucleotide analogs, such as ddNTPs and dye-labeled ddNTPs (13, 25-30), have added insights into the basis for nucleotide discrimination.

Hyperthermophilic archaeal DNA polymerases have not been scrutinized in such detail, hampering a complete characterization and comparison with other polymerases. Family B DNA polymerases from hyperthermophilic Archaea Thermococcus sp. 9°N (22), Thermococcus gorgonarius (18), and Pyrococcus kodakaraensis KOD1 (24) and mesophilic bacteriophage RB69 (23) have high sequence and structural homologies and provide a framework for analysis of active site structure and function in this enzyme family (Fig. 1). Furthermore, steady-state kinetic studies have identified hyperthermophilic DNA polymerase residues important for polymerization and exonuclease activities and for nucleotide binding (18, 29, 31-35). Nucleotide analogs have also been important in identifying dNTP recognition determinants important in the polymerase reaction (32-36) and have proven useful in a variety of molecular biology applications, such as DNA sequencing and detection of single nucleotide polymorphisms (37-41). One group of analogs, 9-[(2-hydroxyethoxy)methyl]X triphosphates (where X is adenine, cytosine, guanine, or thymine; acyNTPs),1 is particularly intriguing due to the wide spectrum of incorporation efficiency noted in different DNA polymerases, even within the same family of polymerase. For example, within Family B, the herpes simplex virus type 2 and human cytomegalovirus DNA polymerases incorporate acyNTPs more efficiently than ddNTPs, whereas human polymerase {alpha} more readily inserts ddNTPs over acyNTPs (42). Such differences have been exploited in drug therapies where infective agents encode polymerases that more readily insert acyNTP than does the host DNA polymerase (43). Hyperthermophilic archaeal DNA polymerases (Vent®, Deep VentTM, 9°NTM, and Pfu) all incorporate acyNTPs with greater efficiency than ddNTPs (33), in contrast with the behavior of Taq and Klenow fragment DNA polymerases, which prefer ddNTPs (33, 44).



View larger version (45K):
[in this window]
[in a new window]
  		FIG. 1.
Alignment of Family B DNA polymerase active sites. A, active site residues in hyperthermophilic Archaea Thermococcus sp. 9°N (green; Protein Data Bank code 1QHT [PDB] ) (22), T. gorgonarius (TGO; red; code 1TGO [PDB] ) (18), and P. kodakaraensis KOD1 (KOD; blue; code 1GCX [PDB] ) (24) are aligned with the apo-RB69 DNA polymerase (purple; code 1IH7 [PDB] ) based on conserved region III amino acids using Deep View/SwissPdbViewer Version 3.7 using default settings (available at www.expasy.org/spdbv/) and rendered with Quanta software (Accelrys Inc., San Diego, CA). Structural deviations (root mean square deviations) between backbone atoms along the entire proteins are 1.87, 2.08, and 2.02 Å, respectively, compared with RB69 DNA polymerase. Conserved active site residues (RB69 DNA polymerase numbering) Lys560, Asn564, Tyr567, Tyr416, and Asp411 are highlighted. Ala485 is shown in green on 9°N DNA polymerase; the homologous residue is mutated to leucine in VentA488L DNA polymerase. B, Family B active site residues from conserved regions II and III (4) are aligned.

 

In the course of probing the determinants of nucleotide sugar discrimination in the Family B DNA polymerase from the hyperthermophilic Archaea Thermococcus litoralis (Vent DNA polymerase), we identified a mutant (VentA488L DNA polymerase) that reduces discrimination against several altered nucleotides (32, 33). Subsequent crystal structures of closely related DNA polymerases strongly suggested that this residue makes neither direct nor indirect contacts with the reaction substrates, raising questions about the structural basis for the observed variation (Fig. 1B). The universality of the A488L phenotype was later confirmed by homologous mutations in other hyperthermophilic DNA polymerases (Pfu A486Y DNA polymerase (34), 9°N A485L DNA polymerase (33), and Tsp JDF-3 A485T DNA polymerase (36)), further emphasizing a conserved role for this residue.

Although instructive, these steady-state observations failed to address the underlying kinetic mechanisms responsible for nucleotide and nucleotide analog incorporation in hyperthermophilic DNA polymerases. Therefore, we initiated pre-steady-state kinetic studies to compare the modes of nucleotide discrimination in Vent and other DNA polymerases.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Nucleotides, Nucleotide Analogs, DNA Substrate, and Enzymes—All DNA polymerases used in this study are 3' -> 5' exonuclease-deficient as a result of mutation of catalytic aspartic and glutamic acids to alanine in the exonuclease active site (31, 32, 45). These mutations prevent exonuclease removal of newly incorporated nucleotides or terminators. Vent and VentA488L DNA polymerases were purified as described previously (31), and the concentration was determined spectroscopically at 280 nm using an extinction coefficient of 115,960 liter mol-1 cm-1. The concentration of E. coli DNA polymerase I (Klenow fragment exo-; New England Biolabs Inc., Beverly, MA) was calculated using a specific activity of 20,000 units/mg. dNTPs, ddCTP, and 9-[(2-hydroxyethoxy)methyl)]cytosine triphosphate (acyCTP) were from New England Biolabs Inc. 2'-Deoxycytidine 5'-O-(1-thiotriphosphate) (dCTP{alpha}S) and CTP were from Amersham Biosciences. 2',3'-Dideoxycytidine 5'-O-(1-thiotriphosphate) (ddCTP{alpha}S) was from TriLink BioTechnologies (San Diego, CA). 6-Carboxy-X-rhodamine (ROX)-derivatized nucleotide analogs ROX-ddCTP and ROX-acyCTP were kindly provided by Phil Buzby (PerkinElmer Life Sciences) (Fig. 2). Oligonucleotides used to measure 2'-deoxycytosine 5'-triphosphate (dCTP) and cytosine analog incorporation were synthesized and purified by the Oligonucleotide Synthesis Division at New England Biolabs with a 6-carboxyfluorescein (FAM) label on the primer strand for detection: 5'-FAM-CCCTCGCAGCCGTCCAACCAACTCA-3' (25-mer) and 3'-GGGAGCGTCGGCAGGTTGGTTGAGTGCCTCTTGTTT-5' (36-mer).



View larger version (16K):
[in this window]
[in a new window]
  		FIG. 2.
Nucleotide and nucleotide analogs used for study of Vent DNA polymerase pre-steady-state kinetic reactions. The maximum rate of nucleotide addition and the dissociation constant for binding were determined with the following nucleotides and nucleotide analogs: dCTP, dCTP{alpha}S, and CTP (A); nucleotide terminators ddCTP and acyCTP (B); and dye-substituted nucleotide terminators ROX-ddCTP and ROX-acyCTP (C).

 
FAM-duplex DNA was formed by mixing equimolar amounts of the dye-labeled 25-mer primer with the 36-mer template in annealing buffer (5 mM Tris-HCl (pH 8.0 at 20 °C), 5 mM NaCl, and 0.2 mM EDTA) and heating the solutions to 95 °C for 5 min, followed by incubation for 10 min at 60 °C and then cooling for 15 min at room temperature.

Burst Kinetics and Active Site Titration—To measure the fraction of active Vent DNA polymerase and to determine the position of the rate-limiting step within the polymerase reaction pathway, we investigated whether the reaction followed burst kinetics. Rapid quench reactions were carried out as described below with 50 nM FAM-duplex DNA; 10 or 20 nM Vent or VentA488L DNA polymerase; and 0.20 mM dCTP, ddCTP, ddCTP{alpha}S, CTP, or acyCTP (final concentrations after mixing) in 1x ThermoPol buffer (10 mM KCl, 20 mM Tris-HCl (pH 8.8 at 25 °C), 10 mM (NH4)2SO4, 2 mM MgSO4, and 0.1% Triton X-100). The steady-state rate (k2), the burst amplitude (A, which is equal to the active site concentration), and the initial rate of product formation (r, the burst rate) were extrapolated from the burst equation: [product] = A(1 - exp-rt) + k2t (45). The steady-state turnover number (kSS) was calculated by dividing k2 by A.

Measurement of DNA Polymerase Pre-steady-state Kinetic Parameters—Single turnover nucleotide incorporation reactions were initiated by mixing Vent or VentA488L DNA polymerase (0.10 µM) and FAM-duplex DNA (0.050 µM) in 1x ThermoPol buffer together with an equal volume of nucleotides or nucleotide analogs in 1x ThermoPol buffer. The reactions were allowed to proceed for the indicated times and then quenched by addition of EDTA to a final concentration of 0.35-0.40 M. Reactions in the range of 3 ms to 10 s were sampled using an RQF-3 rapid quenched-flow instrument (Kintek Corp., Austin, TX). Reactions with an initial time point >10 s were mixed and quenched manually. All Vent DNA polymerase reactions were analyzed at 60 °C. Although this temperature is lower than the optimal reaction temperature of 72 °C (31), it is the highest temperature at which the rapid quench instrument can be operated reliably. Single turnover acyCTP incorporation by Klenow fragment DNA polymerase was initiated by mixing 1.0 µM Klenow fragment DNA polymerase (exo-) and 0.10 µM FAM-duplex DNA substrate in 1x Klenow buffer (50 mM Tris-HCl (pH 7.5) and 2 mM MgCl2) with an equal volume of acyCTP in 1x Klenow buffer. Reactions were then incubated at 25 °C for various times and quenched manually with EDTA (0.1 M final concentration).

Conversion of the fluorescently labeled DNA primer-template to product was monitored by denaturing PAGE and automated fluorescence detection methods. Product DNA was denatured by mixing a 7.5 µl aliquot of quenched sample with 45 µl of formamide and 1.5 mM EDTA and heating at 90 °C for 3 min. Fluorescent 5'-FAM-labeled 25-mer oligonucleotide substrate and 5'-FAM-labeled 26-mer oligonucleotide product bands were fractionated by electrophoresis on an 8.8 M urea and 16% polyacrylamide denaturing gel using an ABI377 automated sequencer (Applied Biosystems, Foster City, CA) and quantified using GeneScan Version 2.1 software (Applied Biosystems). The first-order rate constant for polymerase-catalyzed addition at each nucleotide concentration was calculated from a plot of ln[substrate] versus time. Rate constants (kobs) were subsequently plotted as a function of nucleotide or analog concentration and fitted to the hyperbolic equation: kobs = (kpol[nucleotide])/(KD + [nucleotide]), yielding kpol, the maximum rate of nucleotide addition, and KD, the dissociation constant for nucleotide binding (46). The activation energy difference between dNTP and nucleotide analog incorporation was calculated by Equation 1 (47).

(Eq. 1)

Single turnover kinetics require saturating enzyme concentrations. We established that 0.10 µM Vent DNA polymerase was sufficient under the reaction conditions described by demonstrating that the rates of ddCTP incorporation were the same using Vent DNA polymerase concentrations of 0.10, 0.20, and 0.40 µM (data not shown).

Measurement of Pyrophosphorolysis Catalyzed by Vent DNA Polymerase—To measure the rate of DNA degradation by pyrophosphorolysis, Vent or VentA488L DNA polymerase (0.10 µM) was equilibrated with the DNA substrate (0.050 µM) in 1x ThermoPol buffer and then mixed with PPi in 1x ThermoPol buffer at 60 °C using rapid quench techniques as described above. The extent of pyrophosphorolysis at each time point was calculated by multiplying the mole fraction of each DNA species by the number of phosphodiester bonds hydrolyzed to generate that species. KD(PPi) and kpyro were derived using fitting protocols analogous to those described above for nucleotide addition.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
Analysis of dNTP Incorporation by Vent DNA Polymerase—Previous studies with Family A DNA polymerases have shown that the steady-state rate-limiting step for addition of a single correctly paired dNTP follows phosphodiester bond formation (8, 13, 46, 48). Consequently, the first round of polymerization occurs more rapidly than subsequent rounds, resulting in a rapid initial burst of product. Incorporation of dCTP by Vent DNA polymerase displayed a burst pattern similar to those seen with RB69 and AmpliTaq-CS DNA polymerases, with a rapid burst (kburst = 60 s-1) followed by slow steady-state turnover (kSS = 0.90 s-1) (Fig. 3A and Table I). As indicated above, the burst is diagnostic for a rate-limiting step following bond formation; moreover, its amplitude is equal to the concentration of active enzyme, indicating that >90% of the Vent DNA polymerase preparation was active. Under similar conditions, Vent DNA polymerase failed to show a significant burst with ddCTP (Fig. 3B) or CTP (data not shown) incorporation. These data suggest that the rate-limiting step during nucleotide analog incorporation has changed compared with dNTP. Upon substitution of ddCTP with ddCTP{alpha}S, both Vent and VentA488L DNA polymerases showed a 10- and 6-fold thio elemental effect (kburst(ddCTP)/kburst(ddCTP{alpha}S)), respectively (Table I), consistent with an altered rate-limiting step.



View larger version (11K):
[in this window]
[in a new window]
  		FIG. 3.
Pre-steady-state burst kinetics of dCTP and ddCTP incorporation by Vent DNA polymerase. Conversion of fluorescently labeled substrate (25-mer) to product (26-mer) by 20 nM Vent DNA polymerase with 200 µM dCTP (A) or ddCTP (B) was monitored as described under "Experimental Procedures." Product (nanomolar) formation is plotted versus time and fit to the burst equation: [product] = A(1 - exp-rt) + k2t. In A, the Vent DNA polymerase dCTP burst amplitude (A) was 21 nM, the burst rate (r) was equal to 85 s-1, and the steady-state rate (k2) was equal to 18 s-1. In B, the first-order initial rate of ddCTP incorporation was 0.5 s-1.

 


View this table:
[in this window]
[in a new window]
  		TABLE I
Pre-steady-state burst kinetics

The kinetic parameters for Vent and VentA488L DNA polymerases are from at least two independent determinations and are reported as the means ± S.D. ND, not determined.

 
Determinations of KD and kpol for dCTP addition by Vent DNA polymerase gave kinetic constants similar to those determined for other DNA polymerases (Fig. 4A and Tables II and III). The relatively high KD for nucleotides (KD = 70 µM) is similar to the Km for nucleotides determined in multiple turnover steady-state measurements (Km = 40 µM) (31). Kinetic constants show little dependence on nucleotide identity, as similar Vent DNA polymerase binding (KD = 58 µM) and rate (kpol = 64 s-1) constants were observed for dATP incorporation. Substitution of dCTP with dCTP{alpha}S had little effect on binding (KD) or phosphodiester bond formation (kpol); thus, the polymerase displays a minimum thio elemental effect (kpol(dCTP)/kpol(dCTP{alpha}S) = 0.80) (Table II).



View larger version (11K):
[in this window]
[in a new window]
  		FIG. 4.
Vent DNA polymerase pre-steady-state kinetics of nucleotide and nucleotide analog incorporation. The dependence of the reaction rate (kobs) on nucleotide or nucleotide analog concentration was fit to a hyperbola according to the Michaelis-Menten equation: kobs = (kpol[nucleotide])/(KD + [nucleotide]), where kobs is the observed reaction rate, kpol is the maximum rate of phosphodiester bond formation, and KD is the equilibrium dissociation constant, as described under "Experimental Procedures." A, a fit of the data for dCTP incorporation gave KD(dCTP) = 74 µM and kpol = 65 s-1. B, a fit of the data for ddCTP single turnover gave KD(ddCTP) = 37 µM and kpol = 0.13 s-1.

 


View this table:
[in this window]
[in a new window]
  		TABLE II
Pre-steady-state kinetic constants for nucleotide and nucleotide analog incorporation by Vent and VentA488L DNA polymerases

In almost all cases, the kinetic parameters for Vent and VentA488L DNA polymerases are from at least two independent determinations (except where indicated by Footnote b) and are reported as the means ± S.D.

 


View this table:
[in this window]
[in a new window]
  		TABLE III
Pre-steady-state kinetic constants for nucleotide analog incorporation by DNA polymerases

The kinetic parameters for Vent and VentA488L DNA polymerases are from at least two independent determinations and are reported as the means ± S.D. ND, not determined.

 
Analysis of Vent DNA Polymerase-catalyzed Pyrophosphorolysis—To examine Vent DNA polymerase pyrophosphorolysis activity, we monitored degradation of a FAM-labeled oligonucleotide duplex in the presence of increasing concentrations of PPi. The dependence of the rate of Vent DNA polymerase pyrophosphorolysis on PPi concentration yielded an equilibrium dissociation constant for PPi binding of KD = 340 µM and a maximum velocity of kpyro = 1.1 s-1 (Table IV).


View this table:
[in this window]
[in a new window]
  		TABLE IV
Pyrophosphorolysis

The kinetic parameters for Vent and VentA488L DNA polymerases are from at least two independent determinations and are reported as the means ± S.D.

 
Analysis of Ribonucleotide and Nucleotide Analog Incorporation by Vent DNA Polymerase—Kinetic parameters of ribonucleotide incorporation were determined to analyze the effect of the presence of a 2'-OH ribonucleotide on polymerization. Vent DNA polymerase discriminated strongly against CTP incorporation via a 16-fold reduced binding affinity (KD = 1100 µM) and a 400-fold slower rate of nucleotide addition (kpol = 0.160 s-1) (Table II). Comparison of CTP and dCTP parameters (expressed as the ratio of catalytic efficiencies: (kpol/KD)dCTP/(kpol/KD)CTP) revealed that Vent DNA polymerase preferred dCTP over CTP by 6000-fold.

In contrast to CTP, discrimination by Vent DNA polymerase against ddCTP and acyCTP was almost exclusively due to a slower rate of nucleotide addition, with KD values for dCTP, ddCTP, and acyCTP being roughly equal (Fig. 4B and Table II). Indeed, the approximate 30-fold preference for acyCTP over ddCTP incorporation can almost entirely be attributed to steps measured by kpol.

Similar experiments with Klenow fragment DNA polymerase showed a 32,000-fold higher discrimination against acyCTP, affecting steps measured by both KD and kpol. The Klenow fragment DNA polymerase equilibrium binding constant for acyCTP was increased by 20-fold compared with dCTP and ddCTP, whereas kpol for acyCTP incorporation was reduced by >1500-fold compared with dCTP (Table III).

ROX-ddCTP and ROX-acyNTP Incorporation—Previous studies found ROX-derivatized ddCTP and acyCTP to be more efficient terminators than their unmodified forms when using Vent DNA polymerase (33). Pre-steady-state kinetics revealed higher binding affinities, but slower incorporation kinetics for the ROX derivatives (Table II), resulting in only marginal alterations in incorporation selectivity.

Analysis of Enhanced Nucleotide Analog Incorporation by VentA488L DNA Polymerase—We previously reported enhanced incorporation of nucleotide analogs by VentA488L DNA polymerase (32, 33). In a burst kinetics experiment, the A488L mutant enzyme gave an initial burst of dCTP incorporation at a rate similar to that seen with the wild-type enzyme (kburst = 45 s-1; >75% active) (Table I). Moreover, the single turnover kinetic parameters for dCTP addition (KD = 77 µM and kpol = 56 s-1) were similar to values derived for the wild-type enzyme (Table II). However, following the initial turnover, the steady-state rate of the A488L mutant polymerase was 9-fold slower than that of the wild-type enzyme (kSS = 0.10 s-1) (Table I), accounting for the lower specific activity of VentA488L DNA polymerase (32). As with wild-type Vent DNA polymerase, replacement of dCTP with dCTP{alpha}S had little effect on KD or kpol (Table II). VentA488L DNA polymerase was less active in pyrophosphorolysis compared with the wild-type enzyme, the result of a 3-fold reduction in PPi binding affinity and a 2-fold decrease in kpyro (Table IV). Incorporation of CTP, ddCTP, acyCTP, and ROX-ddCTP by VentA488L DNA polymerase was more efficient (by 5-10-fold) compared with incorporation by the wild-type enzyme; in each case, this is attributable to both increased binding affinity (lower KD) and faster reaction rates (kpol) (Table II). Incorporation of ROX-acyCTP was largely unaffected by the A488L mutation (Table II).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
The fundamental information derived for Vent DNA polymerase incorporation of dCTP confirms and expands earlier steady-state data (31) and places Vent DNA polymerase in the context of other Family A and B DNA polymerases. As with these other polymerases, the steady-state rate for single nucleotide addition is limited by a slow step after phosphodiester bond formation. Previous steady-state measurements using an assay in which all four dNTPs were present gave a kcat value of 10 s-1.2 This value is higher than the steady-state rate derived here in burst experiments (0.9 s-1), most likely reflecting the higher temperature (72 °C) and, more importantly, the processive synthesis allowed in the earlier studies. In contrast, the experimental design reported here forces the DNA polymerases to act in a distributive manner, i.e. dissociating from the DNA before binding another primer-template and incorporating another nucleotide.

The single turnover parameters for Vent DNA polymerase with the normal dCTP substrate are similar to those of other Family A and B polymerases, both mesophilic and thermophilic. As shown in Table III, KD and kpol values differ by <10-fold for all polymerases tested, with no clear division between Family A and B DNA polymerases. Furthermore, the Family A Klenow fragment and Family B Vent and RB69 DNA polymerases carry out the reverse reaction of DNA polymerization, pyrophosphorolysis, with similar rates (kpyro), and Klenow fragment and Vent DNA polymerases share comparable PPi binding constants (KD(PPi)) (Table IV). Similarities in nucleotide incorporation kinetics and active site structure underscore the evolution of DNA polymerases to efficiently carry out DNA replication and repair. Significant kinetic differences between the polymerases become apparent only when examining nucleotide analog incorporation and elemental effects, as detailed below.

Ribonucleotides—Despite a similar level of selectivity against NTPs, this discrimination is supplied almost exclusively by elements measured by kpol for Klenow fragment DNA polymerase, whereas Vent DNA polymerase shows not only kpol effects, but also a 16-fold weaker ground state binding of the nucleotide. RB69 DNA polymerase also shows effects in both KD and kpol, achieving an even higher discrimination by virtue of a 230-fold weaker ground state binding. Discrimination against NTPs has, in large part, been attributed to a steric clash between the 2'-OH and a conserved side chain in the polymerase active site (26, 30, 32). The kinetic parameters suggest that the steric clash is first encountered in the ground state nucleotide binding by Vent and RB69 DNA polymerases, but does not affect Klenow fragment DNA polymerase until the transition state of the reaction. This could occur, for example, if the KD term for Klenow fragment DNA polymerase primarily measures binding prior to a conformational shift that engages the 2'-OH sensing machinery.

Dideoxynucleotides—When incorporating ddCTP, RB69 DNA polymerase discriminates at the level of both KD and kpol. Discrimination by Vent, Klenow fragment, and KlenTaq (truncated Taq DNA polymerase with a 236-amino acid N-terminal deletion (13)) DNA polymerases is almost exclusively in the steps measured by kpol and not those involved in KD, with Vent DNA polymerase showing less discrimination than the other two polymerases. This parallel behavior appears to reflect a mutual lack of 3'-OH involvement in ground state substrate binding rather than a conserved set of nucleotide contacts.

On the surface, the similarity in kpol values for dNTP incorporation by Vent and Klenow fragment DNA polymerases (10) suggests similar discriminatory mechanisms for these two enzymes, a conclusion reinforced by the absence of an elemental effect with dNTP{alpha}s using either enzyme. The simplest interpretation of the lack of an elemental effect with {alpha}-thio-substituted dNTPs with Klenow fragment and Vent DNA polymerases is that a non-chemical step(s) preceding phosphodiester bond formation is rate-limiting. Similarly, the lack of a significant elemental effect for Klenow fragment DNA polymerase incorporation of ddNTP{alpha}s (12) argues that steps preceding phosphodiester bond formation continue to be rate-limiting for that enzyme. In contrast, the elemental effect noted for Vent and VentA488L DNA polymerase incorporation of ddNTP{alpha}s is an indication that the chemistry of phosphodiester bond formation significantly influences the rate-limiting step for these polymerases.

The kpol rates with both polymerases were significantly slower for ddNTPs than for dNTPs: 5000- and 400-fold for Klenow fragment and Vent DNA polymerases, respectively. In the case of Vent DNA polymerase, this must reflect at least a slowing of the chemical rate, whereas for Klenow fragment DNA polymerase, at least the rate of pre-chemical step(s) must be slowed. Thus, the pre-chemical rate for Vent DNA polymerase ddNTP incorporation is at least 10-fold faster than comparable steps for Klenow fragment DNA polymerase.

Conserved amino acids positioned within either Family A or B DNA polymerase active sites probe for correctly base-paired substrates and concordantly align phosphates into a geometry required for phosphoryl transfer. As observed by Franklin et al. (23) and Yang et al. (25) in the RB69 DNA polymerase ternary crystal structure (and by analogy, in the Vent DNA polymerase active site) (Fig. 1), the dNTP deoxyribose moiety assumes a favorable 3'-endo-sugar conformation. This conformation is constrained by hydrogen bonds between the 3'-OH and a main chain amide (corresponding to Vent DNA polymerase position 412) and a non-bridging {beta}-phosphate oxygen (Fig. 5, A and B). Nucleotide {alpha}-, {beta}-, and {gamma}-phosphates are further stabilized by direct or water-mediated hydrogen bonds with active site residues (Fig. 5, A and B). The absence of the 3'-OH on ddNTPs disrupts hydrogen bonding with the {beta}-phosphate (and main chain amide), potentially increasing the activation energy required to orient the {alpha}-phosphate for phosphoryl transfer (Fig. 5C). Indeed, the measured energetic difference between dNTP and ddNTP incorporation (15 kJ mol-1) is equivalent to that expected for the loss of at least two hydrogen bonds in the ddNTP transition state (47).



View larger version (19K):
[in this window]
[in a new window]
  		FIG. 5.
Active site models of dNTP, ddNTP, and acyNTP interactions. A, the RB69 DNA polymerase ternary crystal structure shows active site interactions that stabilize the substrate dTTP (23). Vent DNA polymerase numbering is shown in parentheses. B, a schematic of the Vent DNA polymerase active site interactions that stabilize the dNTP transition state is presented in a two-dimensional view. W, water molecule. C, a model of the Vent DNA polymerase active site with ddNTP bound reveals loss of hydrogen bonding with the Tyr412 main chain amide, dTTP 3'-OH, and non-bridging {beta}-phosphate. D, a model for the binding of acyNTP suggests that, in the absence of a ribose ring, a molecule X (which could be a water molecule) can re-establish hydrogen bonding between the Tyr412 main chain amide and non-bridging {beta}-phosphate.

 
Although the active site bonding network differs, in the Family A Klenow fragment DNA polymerase the dNTP 3'-OH contributes 21 kJ mol-1 to transition state stabilization, accounting for inefficient ddNTP incorporation (27). This energy loss is counteracted in the closely related T7 DNA polymerase active site by a hydroxyl group at Tyr526 (Klenow fragment DNA polymerase has Phe in the analogous position) that contributes a hydrogen bond to stabilize the ddNTP {beta}-phosphate in the transition state, re-establishing a hydrogen bonding network similar to interactions formed by dNTP (12). As a result, T7 DNA polymerase selectivity between dNTP and ddNTP is greatly reduced, as is the selectivity of the analogous Phe-to-Tyr mutation in both Klenow fragment and Taq DNA polymerases (49).

Acyclonucleotides—Divergence between polymerases is also noted upon acyCTP addition, again suggesting divergent mechanisms for nucleotide recognition and incorporation between the polymerases. Similar to ddNTPs, acyNTPs lack the 3'-OH required to establish a hydrogen bonding network between the main chain amide of Tyr412 and {beta}-phosphate of the substrate (Fig. 5D). Klenow fragment DNA polymerase displays a strong discrimination in both KD and kpol terms, resulting in a selectivity value of 32,000. In this case, the efficiency of acyNTP incorporation is nearly as low as for an incorrect base pair (kpol/KD = 240 and 160 M-1 s-1, respectively) (Table III) (50). A strong bias against acyNTP incorporation has also been noted for Taq DNA polymerase (33).

In contrast, acyNTPs are incorporated by hyperthermophilic archaeal DNA polymerases with only 10-fold lower efficiency than dNTPs. By analogy with ddNTP incorporation by T7 DNA polymerase, it seems reasonable that the space normally occupied by the sugar 2'- and 3'-carbons and associated substituents would be accessible to water molecules, metals, or protein side chains that might establish interactions to compensate for those disrupted by the missing 3'-OH group. The difference in activation free energy between ddNTP and acyNTP incorporation ({Delta}{Delta}G{ddagger} = {Delta}G{ddagger}ddNTP - {Delta}G{ddagger}acyNTP = 10 kJ mol-1) is equivalent to a gain of two additional hydrogen bonds, which could be provided by hydrogen bonding between the main chain 412 amide, a putative water, and an acyNTP {beta}-phosphate non-bridging oxygen to mimic interactions that exist in the dNTP active site (Fig. 5D). At the same time, we cannot rule out stabilizing interactions arising from residues near the active site normally excluded by the ribose 2'- and 3'-carbons that are absent in acyNTP. Clearly, three-dimensional structural analysis will be necessary for a full understanding of the interactions important for Vent DNA polymerase incorporation of acyNTPs.

Dye-substituted Nucleotides—Dye-substituted nucleotides have been useful in a variety of analytical applications (40, 41). Not surprisingly, given the diversity of dye structures and charges, dye-substituted nucleotides are accepted by DNA polymerases with varying efficiencies (28, 29, 33, 36). Previous studies identified nucleotide derivatives containing the fluorescent dye ROX as being more efficiently incorporated by Vent DNA polymerase than the parental nucleotides lacking the dye (29, 33). In the current kinetic studies, the magnitude of enhanced dye-substituted terminator addition was much lower than previously estimated in semiquantitative gel titration assays, even though those same gel titration assays give good agreement with the relative incorporation efficiency of ddNTP and acyNTP substrates (33). ROX substitution of the nucleotide results in a 5-10-fold lower KD, suggesting that contacts in or adjacent to the Vent DNA polymerase active site stabilize dye binding. However, at the same time, kpol is reduced, suggesting that one consequence of the enhanced binding is to slow nucleotide addition. Thus, substrate incorporation is a balance of both binding and catalysis: a substrate bound with too high affinity requires higher activation energy for efficient turnover by the polymerase.

VentA488L DNA Polymerase Pre-steady-state Kinetics—Previous studies identified a Vent DNA polymerase variant (A488L) with enhanced nucleotide analog incorporation properties (32, 33). Correct dNTP incorporation by VentA488L DNA polymerase is characterized by similar binding affinity (KD), nucleotide transfer rate (kpol), and rate-limiting step compared with Vent DNA polymerase, presumably reflecting the conservation of residues actively involved in coordinating the incoming dNTP. In contrast, each of the nucleotide analogs tested with VentA488L DNA polymerase have higher binding affinity and faster rates of phosphoryl transfer than the unmodified polymerase. Energy differences between Vent and VentA488L DNA polymerase incorporation of ddNTP or acyNTP are modest ({Delta}{Delta}G{ddagger}ddNTP = 4.5 kJ mol-1 and {Delta}{Delta}G{ddagger}acyNTP = 4.8 kJ mol-1), suggesting that subtle hydrophobic or hydrogen bond-mediated effects could account for enhanced analog incorporation.

One hypothesis to account for these effects envisions the A488L variant as lying closer to the activated conformation, thus facilitating incorporation of analogs. The residue analogous to Ala488 in the RB69 DNA polymerase crystal structure points away from the active site and lies at the interface between the solid core of the polymerase and an {alpha}-helix that must make a 60° rotation to form the closed complex (Fig. 1). In Vent DNA polymerase, positioning a larger leucine residue at the position normally occupied by alanine in the {alpha}-helix may shift equilibrium from the open toward the closed conformation, thus reducing the activation energy for both binding and nucleotide transfer. This comes at a price: burst experiments demonstrate that subsequent turnover by the A488L variant is inhibited, perhaps reflecting hindrance of the transition from closed to open states required for release and/or binding of the template and dNTP. This proposal does not, however, easily account for the fact that pre-steady-state kinetics for the natural substrates are unaltered in this variant. Alternatively, resolution of this discrepancy may lie in the greater ability of this variant to overcome distortions in the nucleotide-binding site, distortions that are not present when the normal nucleotide is bound.

In summary, from these comparative studies, we observed that kinetics of dNTP incorporation pathways are conserved among Family A and B DNA polymerases despite diversity in primary amino acid sequence, thermostability, fidelity, and biological roles. However, differences in acyNTP and other nucleotide analog catalytic efficiencies in Klenow fragment, Vent, and other DNA polymerases illuminate fundamental differences underlying the kinetic pathway for DNA polymerization. As more DNA polymerases are studied kinetically, it is apparent that subtle structural variations in the active site influence how nucleotides are bound and positioned for catalysis.


   FOOTNOTES
 
* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

Supported by National Institutes of Health Grant GM-28550. Back

|| To whom correspondence should be addressed: New England Bio-labs Inc., 32 Tozer Rd., Beverly, MA 01915. Tel.: 978-927-5054; Fax: 978-921-1350; E-mail: jack{at}neb.com.

1 The abbreviations used are: acyNTP, 9-[(2-hydroxyethoxy)methyl]X triphosphate, where X is adenine, cytosine, guanine, or thymine; acyCTP, 9-[(2-hydroxyethoxy)methyl)]cytosine triphosphate; dCTP{alpha}S, 2'-deoxycytidine 5'-O-(1-thiotriphosphate); ddCTP{alpha}S, 2',3'-dideoxycytidine 5'-O-(1-thiotriphosphate); ROX-, 6-carboxy-X-rhodamine; FAM, 6-carboxyfluorescein. Back

2 H. Kong, H. M. R. B. Kucera, and W. E. Jack, unpublished data. Back


   ACKNOWLEDGMENTS
 
We are grateful to Charles Richardson for overseeing this research as part of the Master of Liberal Arts Program at Harvard University (to A. F. G.). We also thank Phil Buzby for providing dye-labeled compounds and for helpful discussions; John Brandis (Applied Biosystems) for useful discussions and technical advice; Nicole Nichols, Sriharsha Pradhan, and Tom Evans for critical review of this manuscript; and Chris Benoit, Lucia Greenough, and Julie Menin for providing expert technical assistance. We are also indebted to Don Comb for fostering a supportive research environment at New England Biolabs Inc.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Kornberg, A. (1980) DNA Replication, pp. 87-97, W. H. Freeman & Co., San Francisco
  2. Joyce, C. M., and Steitz, T. A. (1994) Annu. Rev. Biochem. 63, 777-822[CrossRef][Medline] [Order article via Infotrieve]
  3. Steitz, T. A. (1999) J. Biol. Chem. 274, 17395-17398[Free Full Text]
  4. Braithwaite, D. K., and Ito, J. (1993) Nucleic Acids Res. 21, 787-802[Free Full Text]
  5. Filée, J., Forterre, P., Sen-Lin, T., and Laurent, J. (2002) J. Mol. Evol. 54, 763-773[CrossRef][Medline] [Order article via Infotrieve]
  6. Cann, I. K., Komori, K., Toh, H., Kanai, S., and Ishino, Y. (1998) Proc. Natl. Acad. Sci. U. S. A. 95, 14250-14255[Abstract/Free Full Text]
  7. Polesky, A. H., Steitz, T. A., Grindley, N. D. F., and Joyce, C. M. (1990) J. Biol. Chem. 265, 14579-14591[Abstract/Free Full Text]
  8. Patel, S. S., Wong, I., and Johnson, K. A. (1991) Biochemistry 30, 511-525[CrossRef][Medline] [Order article via Infotrieve]
  9. Dahlberg, M. E., and Benkovic, S. J. (1991) Biochemistry 30, 4835-4843[CrossRef][Medline] [Order article via Infotrieve]
  10. Polesky, A. H., Dahlberg, M. E., Benkovic, S. J., Grindley, N. D. F., and Joyce, C. M. (1992) J. Biol. Chem. 267, 8417-8428[Abstract/Free Full Text]
  11. Steitz, T. A. (1993) Curr. Opin. Struct. Biol. 3, 31-38
  12. Doublié, S., Tabor, S., Long, A. M., Richardson, C. C., and Ellenberger, T. (1998) Nature 391, 251-258[CrossRef][Medline] [Order article via Infotrieve]
  13. Brandis, J. W., Edwards, S. G., and Johnson, K. A. (1996) Biochemistry 35, 2189-2200[CrossRef][Medline] [Order article via Infotrieve]
  14. Johnson, S. J., Taylor, J. S., and Beese, L. S. (2003) Proc. Natl. Acad. Sci. U. S. A. 100, 3895-3900[Abstract/Free Full Text]
  15. Gillin, F. D., and Nossal, N. G. (1975) Biochem. Biophys. Res. Commun. 64, 457-464[CrossRef][Medline] [Order article via Infotrieve]
  16. Dong, W., Copeland, W. C., and Wang, T. S.-F. (1993) J. Biol. Chem. 268, 24163-24174[Abstract/Free Full Text]
  17. Frey, M. W., Nossal, N. G., Capson, T. L., and Benkovic, S. J. (1993) Proc. Natl. Acad. Sci. U. S. A. 90, 2579-2583[Abstract/Free Full Text]
  18. Hopfner, K. P., Eichinger, A., Engh, R. A., Laue, F., Ankenbauer, W., Huber, R., and Angerer, B. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, 3600-3605[Abstract/Free Full Text]
  19. Zhao, Y., Jeruzalmi, D., Moarefi, I., Leighton, L., Lasken, R., and Kuriyan, J. (1999) Structure 7, 1189-1199[Medline] [Order article via Infotrieve]
  20. Bonnin, A., Lázaro, J. M., Blanco, L., and Salas, M. (1999) J. Mol. Biol. 290, 241-251[CrossRef][Medline] [Order article via Infotrieve]
  21. Truniger, V., Blanco, L., and Salas, M. (1999) J. Mol. Biol. 286, 57-69[CrossRef][Medline] [Order article via Infotrieve]
  22. Rodriguez, A. C., Park, H.-W., Mao, C., and Beese, L. S. (2000) J. Mol. Biol. 299, 447-462[CrossRef][Medline] [Order article via Infotrieve]
  23. Franklin, M. C., Wang, J., and Steitz, T. A. (2001) Cell 105, 657-667[CrossRef][Medline] [Order article via Infotrieve]
  24. Hashimoto, H., Nishioka, M., Fujiwara, S., Takagi, M., Imanaka, T., Inoue, T., and Kai, Y. (2001) J. Mol. Biol. 306, 469-477[CrossRef][Medline] [Order article via Infotrieve]
  25. Yang, G., Franklin, M., Li, J., Lin, T.-C., and Konigsberg, W. (2002) Biochemistry 41, 2526-2534[CrossRef][Medline] [Order article via Infotrieve]
  26. Astatke, M., Ng, K., Grindley, N. D. F., and Joyce, C. M. (1998) Proc. Natl. Acad. Sci. U. S. A. 95, 3402-3407[Abstract/Free Full Text]
  27. Astatke, M., Grindley, N. D. F., and Joyce, C. M. (1998) J. Mol. Biol. 278, 147-165[CrossRef][Medline] [Order article via Infotrieve]
  28. Brandis, J. W. (1999) Nucleic Acids Res. 27, 1912-1918[Abstract/Free Full Text]
  29. Ilsley, D. D., and Buzby, P. R. (1999) FASEB J. 13, A144 (abstr.)
  30. Yang, G., Franklin, M., Li, J., Lin, T.-C., and Konigsberg, W. (2002) Biochemistry 41, 10256-10261[CrossRef][Medline] [Order article via Infotrieve]
  31. Kong, H., Kucera, R. B., and Jack, W. E. (1993) J. Biol. Chem. 268, 1965-1975[Abstract/Free Full Text]
  32. Gardner, A. F., and Jack, W. E. (1999) Nucleic Acids Res. 27, 2545-2553[Abstract/Free Full Text]
  33. Gardner, A. F., and Jack, W. E. (2002) Nucleic Acids Res. 30, 605-613[Abstract/Free Full Text]
  34. Evans, S. J., Fogg, M. J., Mamone, A., Davis, M., Pearl, L. H., and Connolly, B. A. (2000) Nucleic Acids Res. 28, 1059-1066[Abstract/Free Full Text]
  35. Perler, F. B., Kumar, S., and Kong, H. (1996) Adv. Protein Chem. 48, 377-435[Medline] [Order article via Infotrieve]
  36. Arezi, B., Hansen, C. J., and Hogrefe, H. H. (2002) J. Mol. Biol. 322, 719-729[Medline] [Order article via Infotrieve]
  37. Sears, L. E., Moran, L. S., Kissinger, C., Creasey, T., Perry-O'Keefe, H., Roskey, M., Sutherland, E., and Slatko, B. E. (1992) BioTechniques 13, 626-633[Medline] [Order article via Infotrieve]
  38. Peterson, M. G. (1988) Nucleic Acids Res. 16, 10915[Free Full Text]
  39. Bankier, A. T. (1993) Methods Mol. Biol. 23, 83-90[Medline] [Order article via Infotrieve]
  40. Prober, J. M., Trainor, G. L., Dam, R. J., Hobbs, F. W., Robertson, C. W., Zagursky, R. J., Cocuzza, A. J., Jensen, M. A., and Baumeister, K. (1987) Science 238, 336-341[Abstract/Free Full Text]
  41. Chen, X., Levine, L., and Kwok, P. Y. (1999) Genome Res. 9, 492-498[Abstract/Free Full Text]
  42. Reid, R., Mar, E. C., Huang, E. S., and Topal, M. D. (1988) J. Biol. Chem. 263, 3898-3904[Abstract/Free Full Text]
  43. Freeman, S., and Gardiner, J. M. (1996) Mol. Biotechnol. 5, 125-137[Medline] [Order article via Infotrieve]
  44. Trainor, G. L. (February 7, 1996) U. S. Patent 5,558,991
  45. Derbyshire, V., Freemont, P. S., Sanderson, M. R., Beese, L., Friedman, J. M., Joyce, C. M., and Steitz, T. A. (1988) Science 240, 199-201[Abstract/Free Full Text]
  46. Johnson, K. A. (1995) Methods Enzymol. 249, 38-61[Medline] [Order article via Infotrieve]
  47. Fersht, A. R., Shi, J. P., Knill-Jones, J., Lowe, D. M., Wilkinson, A. J., Blow, D. M., Brick, P., Carter, P., Waye, M. M., and Winter, G. (1985) Nature 314, 235-238[CrossRef][Medline] [Order article via Infotrieve]
  48. Benkovic, S. J., and Cameron, C. E. (1995) Methods Enzymol. 262, 257-269[Medline] [Order article via Infotrieve]
  49. Tabor, S., and Richardson, C. C. (1995) Proc. Natl. Acad. Sci. U. S. A. 92, 6339-6343[Abstract/Free Full Text]
  50. Minnick, D. T., Liu, L., Grindley, N. D. F., Kunkel, T. A., and Joyce, C. M. (2002) Proc. Natl. Acad. Sci. U. S. A. 99, 1194-1199[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 Nucleic Acids ResHome page
J. K. Ichida, A. Horhota, K. Zou, L. W. McLaughlin, and J. W. Szostak
High fidelity TNA synthesis by Therminator polymerase
Nucleic Acids Res., September 12, 2005; 33(16): 5219 - 5225.
[Abstract] [Full Text] [PDF]

Home page
 Nucleic Acids ResHome page
V. Kempeneers, M. Renders, M. Froeyen, and P. Herdewijn
Investigation of the DNA-dependent cyclohexenyl nucleic acid polymerization and the cyclohexenyl nucleic acid-dependent DNA polymerization
Nucleic Acids Res., July 12, 2005; 33(12): 3828 - 3836.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
279/12/11834    most recent
M308286200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Gardner, A. F.
Right arrow Articles by Jack, W. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Gardner, A. F.
Right arrow Articles by Jack, W. E.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH