Annu. Rev. Biochem. 1993. 62:685-713
`
`
`
`
`
`Copyright © 1993 by Annual Reviews Inc. All rights reserved
`
`CONFORMATIONAL
`COUPLING IN DNA
`POL YMERASE FIDELITY
`
`Kenneth A. Johnson
`106 Althouse Laboratory,
`
`
`
`
`16802
`
`
`University, University Park, Pennsylvania
`
`
`
`Biochemistry and Molecular Biology, Pennsylvania State
`
`
`
`KEYWORDS: DNA replication, DNA error correction, transient kinetics, induced-fit
`
`
`
`
`
`
`
`CONTENTS
`
`685
`PERSPECTIVES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`. 686
`ENZYMOLOGY OF DNA REPLICATION . . . . . . . . . . . . . . . . . . . . . . . . .
`686
`
`
`
`Base-Pairing Free Energies and the Fidelity Problem . . . . . . . . . . . . . . . . . .
`
`Properties of Simple DNA Polymerases
`688
`. . . . . . . . . . . . . . . . . . . . . . . . . . .
`690
`STRUcruRES OF DNA POL YMERASES . . . . . . . . . . . . . . . . . . . . . . . . . .
`693
`DNA POLYMERASE MECHANISM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`694
`DNA Binding and Processivity
`. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`
`E-DNA-dNTP Complex. . . . . . . . . . 696
`
`
`Nucleotide Binding: Kdfor the Productive
`Change ...... . ........ .... .. ...... 697
`Rate-Limiting Conformational
`in the Duplex DNA . . . . . . . . . . . . . . . . . . . . . . . . . .
`700
`Effect of Mismatches
`...... . .. ....... ... .... .... . . ....... .. 700
`
`Induced-Fit Selectivity
`Errors. 701
`
`
`
`
`Nucleotide Sequence Dependence, Mutational Hot Spots, and Frameshift
`Exchange . . . . . • . . . . . • • . . . . . . 702
`
`Pyrophosphorolysis versus Pyrophosphate
`EXONUCLEASE PROOFREADING .............................. 703
`.................. 704
`
`
`Partitioning in Selectivity of the Exonuclease
`Kinetic
`of the DNA to the Exonuclease
`706
`Site and Back . . . . . . . . . .
`Sliding
`Bidirectional
`HIV REVERSE TRANSCRIPTASE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`709
`FUTURE DIRECTIONS: REPLICATION AT A FORK .................. 710
`710
`SUMMARY .............................................
`
`PERSPECTIVES
`
`Most enzymes have evolved to possess either a broad substrate specificity in
`
`
`
`order to accommodate a range of structurally similar substrates or a strict
`
`
`substrate specificity allowing selection of a single substrate against a pool of
`
`
`close homoiogs. DNA polymerases are remarkable in that they do both by
`685
`
`0066-4154/93/0701-0685$02.00
`
`Annu. Rev. Biochem. 1993.62:685-713. Downloaded from www.annualreviews.org
`
` Access provided by Yeshiva University - Albert Einstein College of Medicine on 03/29/21. For personal use only.
`
`Columbia Ex. 2078
`Illumina, Inc. v. The Trustees
`of Columbia University in the
`City of New York
`IPR2020-00988, -01065,
`-01177, -01125, -01323
`
`Further
`
`ANNUAL
`REVIEWS
`Quick links to online content
`
`

`

`Annu. Rev. Biochem. 1993. 62:685-713
`
`
`
`
`
`Copyright © 1993 by Annual Reviews Inc. All rights reserved
`
`CONFORMATIONAL
`COUPLING IN DNA
`POL YMERASE FIDELITY
`
`Kenneth A. Johnson
`106 Althouse Laboratory,
`
`
`
`
`16802
`
`
`University, University Park, Pennsylvania
`
`
`
`Biochemistry and Molecular Biology, Pennsylvania State
`
`
`
`KEYWORDS: DNA replication, DNA error correction, transient kinetics, induced-fit
`
`
`
`
`
`
`
`CONTENTS
`
`685
`PERSPECTIVES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`. 686
`ENZYMOLOGY OF DNA REPLICATION . . . . . . . . . . . . . . . . . . . . . . . . .
`686
`
`
`
`Base-Pairing Free Energies and the Fidelity Problem . . . . . . . . . . . . . . . . . .
`
`Properties of Simple DNA Polymerases
`688
`. . . . . . . . . . . . . . . . . . . . . . . . . . .
`690
`STRUcruRES OF DNA POL YMERASES . . . . . . . . . . . . . . . . . . . . . . . . . .
`693
`DNA POLYMERASE MECHANISM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`694
`DNA Binding and Processivity
`. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`
`E-DNA-dNTP Complex. . . . . . . . . . 696
`
`
`Nucleotide Binding: Kdfor the Productive
`Change ...... . ........ .... .. ...... 697
`Rate-Limiting Conformational
`in the Duplex DNA . . . . . . . . . . . . . . . . . . . . . . . . . .
`700
`Effect of Mismatches
`...... . .. ....... ... .... .... . . ....... .. 700
`
`Induced-Fit Selectivity
`Errors. 701
`
`
`
`
`Nucleotide Sequence Dependence, Mutational Hot Spots, and Frameshift
`Exchange . . . . . • . . . . . • • . . . . . . 702
`
`Pyrophosphorolysis versus Pyrophosphate
`EXONUCLEASE PROOFREADING .............................. 703
`.................. 704
`
`
`Partitioning in Selectivity of the Exonuclease
`Kinetic
`of the DNA to the Exonuclease
`706
`Site and Back . . . . . . . . . .
`Sliding
`Bidirectional
`HIV REVERSE TRANSCRIPTASE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`709
`FUTURE DIRECTIONS: REPLICATION AT A FORK .................. 710
`710
`SUMMARY .............................................
`
`PERSPECTIVES
`
`Most enzymes have evolved to possess either a broad substrate specificity in
`
`
`
`order to accommodate a range of structurally similar substrates or a strict
`
`
`substrate specificity allowing selection of a single substrate against a pool of
`
`
`close homoiogs. DNA polymerases are remarkable in that they do both by
`685
`
`0066-4154/93/0701-0685$02.00
`
`Annu. Rev. Biochem. 1993.62:685-713. Downloaded from www.annualreviews.org
`
` Access provided by Yeshiva University - Albert Einstein College of Medicine on 03/29/21. For personal use only.
`
`Further
`
`ANNUAL
`REVIEWS
`Quick links to online content
`
`

`

`686 JOHNSON
`
`altering their substrate specificity during each catalytic cycle according to the
`
`
`
`
`
`
`
`DNA template. It is even more remarkable that polymerases can accomplish
`
`
`this task with such extraordinary speed and efficiency. For example, T7 DNA
`
`
`
`polymerase catalyzes correct base pair insertion at rates of 300-500 bases per
`
`second with error frequencies of approximately one in 105 to 106 bases (1,
`
`
`2). When this polymerase does make a mistake, it stops to correct the error
`by removing the mismatch for all but one out of 103 to 104 bases, such that
`one in 108 to Hio base pairs (3). To
`the overall error frequency approaches
`
`
`
`determine how the polymerase can achieve such extraordinary fidelity, we
`
`
`must study the structures of the base pairs and the enzymes and understand
`
`
`
`
`the kinetics and thermodynamics of the reactions occurring at the enzyme
`active site.
`Three distinct reactions contribute to the overall fidelity of DNA replication
`
`
`
`
`
`
`in vivo: the polymerization reaction per se, the proofreading exonuclease
`
`
`reaction (4, 5), and the postreplication error repair system (6). This review
`
`
`focuses on the first two reactions, which are properties of the DNA polymerase
`
`
`
`holoenzyme, by summarizing our current understanding of the structural,
`
`
`
`
`kinetic, and thermodynamic basis of DNA polymerase fidelity. This analysis
`
`
`has led to a new paradigm for understanding DNA polymerization based upon
`
`two conformational states of the E-DNA complex required for each round of
`
`processive synthesis.
`Two excellent reviews describing eukaryotic DNA polymerases and the
`
`
`
`
`
`fidelity of DNA replication were published in this series in 1991 (5, 7), and
`
`a recent review summarized the relevant data on T4 polymerase (8). Since
`
`
`
`these reviews have adequately covered the literature prior to 1991, this review
`
`focuses on the significant advances made in the past two years in our
`
`
`
`
`understanding of the mechanistic basis of DNA polymerase fidelity. We are
`
`
`
`selective rather than exhaustive in our coverage of the polymerases, so that
`
`
`appropriate focus can be given to those systems for which sufficiently detailed
`
`
`
`
`mechanistic information is available. Comparisons are made to other poly­
`
`
`
`merases when possible to illustrate the generality of the conclusions.
`
`ENZYMOLOGY OF DNA REPLICATION
`
`
`
`Base-Pairing Free Energies and the Fidelity Problem
`
`
`
`The difference in free energy of binding for correct versus incorrect base pairs
`
`
`
`in solution is not sufficient to account for the selectivity observed for the
`
`
`
`enzyme-catalyzed polymerase reaction. The free energy difference is defined
`
`aaG = RTln(KclKi), where Kc and Ki
`by the thermodynamic relationship,
`
`
`
`are the binding constants for correct and incorrect nucleotides, respectively.
`The term aaG has been estimated
`
`to be in the range of 1-3 kcallmole, based
`
`Annu. Rev. Biochem. 1993.62:685-713. Downloaded from www.annualreviews.org
`
` Access provided by Yeshiva University - Albert Einstein College of Medicine on 03/29/21. For personal use only.
`
`

`

`DNA POLYMERASE MECHANISM 687
`
`on measurements of stability of polynucleotide helices containing variable
`
`
`
`
`
`
`numbers of mismatched bases (4). Even lower estimates resulted when the
`
`
`analysis was based on terminal mispairs (9). The 1 -3 kcallmole free energy
`
`difference would lead to an error frequency of one out of 5-150 bases if the
`
`polymerase simply zippered together those base pairs that formed a stable
`
`
`
`complex in solution. This is in contrast to the observed error frequencies for
`
`most polymerases, which are in the range of 10-3 to 10-5 (4, 5).
`(10-13)
`
`Analysis of the structures of mispairs by crystallography and
`
`nuclear magnetic resonance (NMR) (14-17) has been quite revealing by
`
`
`pointing to ways in which enzymes might selectively destabilize mismatched
`
`
`
`base pairs relative to the structures observed in aqueous solution. The most
`remarkable property of the G:C and A:T base pairs is their geometric
`
`equivalence and the pseudo twofold axis of symmetry in the plane of the base
`
`pairs. As illustrated by the structure of the G:T mismatch shown in Figure 1
`
`( 1 0), the Watson-Crick base pair geometry is violated in forming the
`
`mismatched base pair, with the thymine projecting into the major groove and
`
`the guanine projecting into the minor groove. However, there is very little
`
`local perturbation of the helix and, more importantly, the global conformation
`
`of the duplex is unaffected. Similar results have been reported for the A:C
`mispair (12). Secondly, the structures of mismatched bases reveal water
`
`molecules that hydrogen bond to any unsatisfied hydrogen bond donors or
`
`acceptors in the mismatched bases.
`
`At the outset, we are not restricted in our analysis of fidelity by the
`
`Me
`
`,
`,2·9
`---S
`2·7
`
`Figure 1 Guanine-Thymine
`mismatch structure. This structure shows the wobble of the base
`
`
`
`
`
`pairs and the locations of hydrogen-bonded water molecules in the G:T mismatch. The thymine
`
`
`
`
`
`
`
`projects into the major groove while guanine projects into the minor groove of the DNA helix.
`Atoms involved in hydrogen bonds are labeled.
`
`The numbers give the lengths of the hydrogen
`
`bonds. Reproduced with permission (10).
`
`Annu. Rev. Biochem. 1993.62:685-713. Downloaded from www.annualreviews.org
`
` Access provided by Yeshiva University - Albert Einstein College of Medicine on 03/29/21. For personal use only.
`
`

`

`688 JOHNSON
`
`as it appears in solution.
`magnitude of the selectivity
`
`Rather, we are faced
`
`
`with the task of specifying the mechanisms by which the free energy
`
`
`
`differences between right and wrong bases are enhanced by the polymerase
`
`
`
`
`
`and utilized in translating nucleotide binding energy into catalytic efficiency.
`
`Thus, it is conceivable that the enzyme could reduce the free energy of
`
`
`mismatch formation by excluding water from the active site; this would
`
`
`
`effectively subtract the thermodynamic contribution of the hydrogen-bonded
`
`
`
`
`water molecules seen in the mispairs in solution. In addition, the enzyme
`
`
`
`
`
`could select against mispairs by maintaining an active-site surface contour
`
`
`
`
`energy from The intrinsic binding that restricts the geometry of the base pairs.
`
`
`
`interactions with the ribose and triphosphate portions of the deoxynucleoside
`
`
`
`triphosphate (dNTP) could be utilized to assist in the desolvation and
`
`
`
`alignment of the bases in the proper orientation. The details of how this· is
`
`
`
`
`
`accomplished must be addressed by direct measurement of the individual steps
`
`
`
`of the polymerization reaction. However, conceptually, one is faced with the
`
`
`
`problem of rapidly binding dNTP from solution while maintaining a highly
`
`
`restricted active-site topology.
`kcaJKm for the
`Selectivity is defined by the enzymatic specificity constant,
`
`
`
`
`
`
`correct versus incorrect base pairs. These steady-state kinetic parameters
`
`
`
`
`define the relative rates of catalysis for incorporation of correct versus
`
`
`
`
`
`incorrect base pairs, weighted by the concentrations of the corresponding base
`pairs.
`x [correct]
`rates (I'cat Km)correct Relative
`
`X [incorrect]
`(kcat KnJincorrect
`
`1 .
`
`kcatlKm, defines
`Although the steady-state kinetic parameter,
`
`
`
`the selectivity,
`of kcat and Km towards fidelity
`
`
`contributions resolution of the relative
`does
`
`
`
`
`not provide sufficient information to establish the mechanistic basis for
`
`
`
`
`fidelity. Rather, resolution of the elementary steps in the reaction sequence
`
`
`
`
`and their contributions towards the observed selectivity is necessary to
`
`
`
`understand the mechanisms by which the enzymes achieve such extraordinary
`fidelity.
`
`Properties of Simple DNA Polymerases
`
`
`
`Although quite a large number of polymerases have now been purified and
`
`
`
`
`of the characterized from various sources (7, 8, 1 8-23), only the simplest
`
`
`
`
`
`polymerases have lent themselves to the detailed investigations necessary to
`
`
`
`
`examine fidelity at the mechanistic level. Polymerases that are involved in
`
`
`
`
`replication of genomic DNA consist of a complex of several proteins that are
`
`
`
`assembled in a highly controlled manner as a necessary part of cellular
`
`
`
`of the and regulation regulation (7, 18, 22, 24--27). Although the assembly
`
`Annu. Rev. Biochem. 1993.62:685-713. Downloaded from www.annualreviews.org
`
` Access provided by Yeshiva University - Albert Einstein College of Medicine on 03/29/21. For personal use only.
`
`

`

`DNA POLYMERASE MECHANISM 689
`
`replication complex is an important problem, we restrict our attention to those
`
`
`
`
`
`
`
`
`
`polymerases with the simplest subunit compositions so as to focus on the
`
`mechanism of polymerization.
`
`
`Most of the early work on DNA polymerase mechanisms was performed
`
`I (Pol I) from Escherichia using DNA polymerase coli and its Klenow
`
`has the major advantage
`(18, 28-38).
`fragment
`This simple polymerase
`that
`
`
`it consists of a single polypeptide containing three domains. One domain
`active site, a second the 3'-5' proofreading
`
`contains the polymerase
`ex­
`(KF)
`and the third the 5'-3' exonuclease.
`onuclease,
`The Klenow fragment
`Because KF
`and 3'-5' exonuclease
`
`contains only the polymerase
`domains.
`
`
`
`
`
`provided the simplest system, it afforded the initial characterization of the
`(32). However, several
`
`
`minimal kinetic mechanism for polymerization
`
`
`
`properties of the enzyme differ from the characteristics now shown for other
`(1,8,36). These unusual
`
`
`well-characterized enzymes involved in replication
`may reflect the role of Pol I as a repair properties enzyme, or they could be
`
`
`
`
`
`
`
`
`a function of the problems inherent in studying a protein fragment. Nonethe­
`
`
`less, we can now tum to simple enzyme systems for viral replication as more
`0, 8, 23)
`
`
`
`appropriate prototypes for analysis of DNA replication
`
`
`
`DNA polymerase from phage T4 has also provided an excellent model for
`enzyme system (8). The T4 polymerase
`
`analysis of a more complicated
`
`
`
`contains both polymerase and exonuclease domains on a single polypeptide.
`
`
`
`
`Optimal rate, processivity, and fidelity require the addition of four accessory
`(8, 40, 41). Two other
`
`
`
`proteins to form the complete five-protein holoenzyme
`
`
`
`
`
`proteins, a helicase and a primase, are required for polymerization at a
`
`
`
`
`
`replication fork. Detailed kinetic studies of the T4 polymerase holoenzyme
`
`
`
`have not yet succeeded because of the requirement to pre-assemble the
`
`
`
`
`complete set of proteins in the proper stoichiometry. Recent studies from the
`(40, 41, 48, 49), and
`laboratories
`of Nancy Nossal (42-47),
`Bruce Alberts
`Peter von Hippel (8, 50-54) have defined
`
`the stoichiometry and conditions
`(8). Given this background,
`
`
`necessary to reassemble the polymerase
`a
`
`
`
`
`complete kinetic analysis of T4 polymerase should be possible.
`
`
`
`
`The laboratories of Richardson and Studier have established T7 DNA
`polymerase
`
`as one of the best understood of the enzyme systems responsible
`(23, 5�82, 84, 8�90). Only three
`for DNA replication
`
`proteins are required
`T7 DNA polymerase;
`T7 RNA polymerase;
`
`for replication of the viral genome:
`
`
`
`and a helicase/primase. T7 DNA polymerase consists of a 1:1 complex
`and E. coli thioredoxin between T7 gene 5 protein
`(58). The gene 5 protein
`
`and 3' -5' exonuclease
`
`contains both the polymerase
`sites on a single
`
`
`
`
`
`polypeptide. The thioredoxin contributes no catalytic activity to the polymer­
`
`
`
`
`
`
`
`ase reaction; rather, it appears to constitute a structural accessory protein that
`
`
`
`the T7 phage has recruited from its host. The thioredoxin increases the
`
`
`
`processivity (74) and fidelity of the polymerase. The T7 system is ideal
`
`Annu. Rev. Biochem. 1993.62:685-713. Downloaded from www.annualreviews.org
`
` Access provided by Yeshiva University - Albert Einstein College of Medicine on 03/29/21. For personal use only.
`
`

`

`690 JOHNSON
`
`because it is the most simple of the enzymes involved in DNA replication
`
`
`
`
`and is easily reconstituted from purified proteins. For this reason, and because
`
`of recent complete kinetic analysis of the T7 DNA polymerase (1-3), we
`
`as revealed
`
`focus in this review on the pathway of polymerization
`by studies
`using the T7 enzymes.
`This approach allows us to refer to a single set of rate
`and equilibrium constants. Available evidence
`supports the conclusion that
`
`
`the model and kinetic parameters observed for T7 polymerase most accur ately
`(1, 8, 38).
`
`represents enzymes involved in DNA replication
`
`
`
`defining the mechanism We also highlight recent results of HIV reverse
`
`transcriptase (RT), which has attracted a great deal of attention due to its
`(An)
`important medical interest
`
`as the target of 3' -azido-3' -deoxy-thymidine
`
`
`three reactions and other anti-AIDS drugs (83, 9 1 -96). mv R T catalyzes
`necessary for viral replication:
`
`
`RNA-dependent DNA polymerase, DNA-de­
`
`pendent DNA polymerase, and RNase H activities. There is no proofreading
`
`low (91,
`
`exonuclease, and the overall
`
`is relatively fidelity of replication
`97-99),
`allowing the virus to mutate rapidly to evade the immune system or
`
`
`antiviral drugs (l00, 101). HIV RT consists of a heterodimer
`of 51-and
`66-kDa subunits.
`
`Both subunits are derived by proteolytic processing of the
`product such that both contain a polymerase domain,
`pol gene translation
`
`while only the 66-kDa subunit retains
`the RNase H domain. Recent structural
`
`
`and mechanistic studies on HIV RT are compared with the re sults on T7
`polymerase.
`
`STRUCTURES OF DNA POL YMERASES
`
`of the structures We begin our analysis by an examination
`
`
`of DNA
`polymerases, since these serve to define the problem and give us a picture to
`
`
`
`
`
`use in correlating structure with the kinetics. Structures have been published
`
`
`the Klenow fragment and mv reverse transcriptase
`for only two polymerases,
`the (3 subunit of Pol ill of E. coli (102, 103), and for one accessory protein,
`
`
`
`(104). There is considerable three-dimensional structural homology betwee n
`
`primary structure homology in
`
`
`the two polymerases in addition to significant
`
`
`It has been argued that the residues thought to participate directly in catalysis.
`
`
`homology at both levels
`
`may well extend to nearly all polymerases (103), so
`
`
`
`
`the conclusions derived by examination of these structures may be generally
`to other systems.
`applicable
`Figure 2 shows the structure of the Klenow fragment (102), showing the
`
`domains (l05). DNA
`
`relationship between the polymerase and exonuclease
`is thought to bind in the large cleft of the polymerase, although the absolute
`orientation
`
`of the DNA in the cleft is not known with certainty (103). The
`
`
`
`exonuclease site has been identified by localization of dTMP bound to the
`
`Annu. Rev. Biochem. 1993.62:685-713. Downloaded from www.annualreviews.org
`
` Access provided by Yeshiva University - Albert Einstein College of Medicine on 03/29/21. For personal use only.
`
`

`

`DNA POLYMERASE MECHANISM 691
`
`Figure 2 Structure
`of Klenow Fragment. This diagram summarizes the crystal structure of KF
`
`
`(102).
`
`
`showing polymerase and exonuclease domains. Reproduced with permission
`
`crystals, and two of the active-site residues involved in metal ion binding
`
`
`by mutagenesis (106).
`(D355, E357) have been identified
`
`
`
`The most unexpected observation from the crystal structure of Klenow was
`active sites (102).
`
`
`the large distance between the polymerase and exonuclease
`30 A
`
`
`The exonuclease site is located on a separate domain approximately
`from the polymerase site. Steitz and coworkers (102) proposed that the DNA
`8-9 base pairs to move the 3' end of the
`must slide and melt approximately
`site (see Figure 2).
`
`primer strand from the polymerase site to the exonuclease
`
`This intramolecular transfer has now been shown to occur for T7 and T4
`(3, 53), and all of the rates have been measured using the T7
`polymerases
`
`
`
`
`DNA polymerase as described below (3). The intramolecular transfer pathway
`
`
`is insignificant with Klenow because of the fast rate of dissociation of the
`(32, lO7).
`
`
`DNA from the enzyme relative to an exceedingly slow exonuclease
`
`
`Figure 3 compares the structures of the polymerase domains of Klenow and
`
`
`RT (103). Each can be described as a right hand with the fingers and thumb
`
`wrapping around a cleft of sufficient size to accommodate duplex DNA. Three
`acidic residues lie in the palm of the hand and are thought to be involved in
`
`
`
`catalyzing the polymerization reaction, since mutation of these residues
`
`Annu. Rev. Biochem. 1993.62:685-713. Downloaded from www.annualreviews.org
`
` Access provided by Yeshiva University - Albert Einstein College of Medicine on 03/29/21. For personal use only.
`
`

`

`692 JOHNSON
`
`A
`
`p66 POL
`
`•
`
`p51 POL
`
`.,
`
`Figure 3 Structure of DNA polymerase
`domains. Ribbon diagrams of the polymerase domains
`of (A) the p66 subunit of mv RT; (8) the pSI subunit of mv RT; and (C) the Klenow fragment
`of RT and KF are compared.
`
`
`of Pol I. (D) The locations of subdomains in the linear sequences
`
`Reproduced with permission from (103).
`
`(0882, E883, and 0705 in Klenow or 0185,0186, and D llO in RT) markedly
`
`
`reduces polymerase activity with either enzyme (10 1, 108-11 0). It has been
`
`
`
`suggested that these residues may bind two divalent metal ions that are
`
`
`
`
`
`
`involved in catalysis of the nucleoside phosphoryl transfer reaction (103, Ill ).
`
`There is also a high degree of homology between T7 gene 5 protein and
`
`
`
`Klenow fragment, especially in the residues that line the active sites of each
`
`
`
`domain. For example, homology between T7 gene 5 protein and Klenow is
`
`
`
`
`sufficiently high to allow identification of key residues in the polymerase and
`
`
`
`
`the exonuclease domains. Mutagenesis of the residues in the exonuclease
`
`
`
`of an construction active site based upon homology to Klenow has allowed
`
`
`exonuclease-deficient mutant of T7 DNA polymerase (1). The double mutant
`
`
`
`D5A,E7A ofT? DNA polymerase has an exonuclease activity that is reduced
`
`
`
`
`
`by a factor of 106 relative to wild type, similar in activity to a deletion mutant
`
`
`of T7 DNA polymerase (80) and the exonuclease-deficient mutant of Klenow
`
`constructed by Joyce (106).
`
`Annu. Rev. Biochem. 1993.62:685-713. Downloaded from www.annualreviews.org
`
` Access provided by Yeshiva University - Albert Einstein College of Medicine on 03/29/21. For personal use only.
`
`

`

`DNA POLYMERASE MECHANISM 693
`
`The structures of the polymerases have not yet been established with DNA
`
`
`
`
`
`
`
`bound. It is likely that the fingers and thumb of the hand-like structure wrap
`
`
`around the DNA upon binding; moreover, as we shall see, conformational
`
`
`
`changes after dNTP binding are necessary to bring the protein to the active
`
`catalytic state.
`
`DNA POLYMERASE MECHANISM
`
`The study of DNA polymerase mechanisms has been revolutionized by the
`
`
`
`
`
`
`
`
`application of transient kinetic methods. In particular, the use of chemical
`
`
`
`quench flow methods (113-115) to measure single-enzyme turnover reactions
`
`
`
`
`has allowed the identification of individual steps along the reaction sequence,
`
`
`
`
`and quantitation of each step's contributions to the overall fidelity (1-3).
`
`
`Because of the tight binding of polymerases to DNA, the small volumes of
`
`
`
`
`material consumed with modem chemical quench flow instruments [15 IJ-L
`
`
`
`per sample, (115)], and the sensitivity afforded by using radiolabelled
`
`
`
`
`substrates, these experiments do not require large quantities of enzyme. The
`
`power of these methods is that they are based upon direct examination of
`
`
`
`single-enzyme turnovers to give the stoichiometry and rate of each reaction.
`
`
`
`
`
`Moreover, the kinetic and equilibrium constants pertain to the productive
`
`
`
`
`enzyme-substrate complexes, that is, those poised to proceed through catal­
`ysis.
`Analysis of T7 DNA polymerase has led to the reaction pathway shown in
`
`
`
`
`
`
`Figure 4 (1-3). A similar pathway was proposed for Klenow fragment (37),
`
`
`
`
`although the reactions governing the movement of the DNA to the exonuclease
`
`
`
`site have not been defined for Klenow because the exonuclease is too slow
`
`
`
`to contribute significantly to fidelity (35).
`DNA can bind either to the polymerase site (Ep) or to the exonuclease site
`
`
`
`
`
`(Ex), or it can move between the two sites without dissociating from the
`
`
`
`
`enzyme (3). Binding to the polymerase site is favored such that only a small
`
`
`
`fraction of DNA molecules occupy the exonuclease site during polymeriza­
`
`
`
`
`
`tion. The sequence of reactions leading to processive polymerization of DNA
`
`
`
`can be understood in terms of two conformational states or modes of enzyme
`
`
`
`
`binding to the DNA. The binding of a correct dNTP induces a rate-limiting
`
`
`from the open to the closed state. In the closed
`change in protein conformation
`
`
`
`
`
`
`state, the reactants reach the critical transition state for catalysis leading to
`
`
`
`
`
`rapid chemical reaction involving the nucleophilic displacement of pyrophos­
`
`
`
`
`phate (PP;) by the 3' -hydroxyl. Following the chemical reaction, the enzyme
`
`
`
`
`returns to the open state, allowing the release of PP; and translocation to allow
`the next round of reaction.
`Translocation is not given in this pathway as a distinct step because it does
`
`
`
`
`
`
`not limit the rate of the reaction and is therefore too fast to measure. We view
`
`Annu. Rev. Biochem. 1993.62:685-713. Downloaded from www.annualreviews.org
`
` Access provided by Yeshiva University - Albert Einstein College of Medicine on 03/29/21. For personal use only.
`
`

`

`694 JOHNSON
`
`Ep.DNAn·dNTP
`
`/,spA!
`�mM
`18,-1 .::;:::===�I
`
`E;'DNAn+1'PPI
`
`Ex·DNA.
`
`j 900,·1
`
`
`
`Ex.DNA.'1 + dNMP
`
`EXONUCLEASE
`Figure 4 Kinetic mechanism of DNA polymerization. The pathway of the reaction
`as determined
`
`for T7 DNA polymerase is shown. The kinetic constants for
`
`each step are given above or below
`
`
`the arrows. Steps that are too fast to be rate limiting are assigned only an equilibrium constant.
`
`Reproduced with permission from (1).
`
`
`translocation as a free diffusion of the polymerase along the DNA while in
`
`
`
`
`
`
`
`
`the open state. This free linear diffusion allows rapid translocation as an
`
`
`
`
`equilibration between the nand n + 1 sites rather than as a distinct rate-limiting
`
`
`
`forward step of the polymerase along the DNA. As the enzyme rapidly slides
`
`
`
`
`between the two sites, the binding of either dNTP or PPi shifts the equilibrium
`
`
`to bring the enzyme to the corresponding site to either allow polymerization
`or pyrophosphorolysis.
`In this section, we describe the evidence in support of the two-state model
`
`
`
`
`
`
`
`
`and attempt to define the characteristics of the closed conformation based
`
`
`
`upon observable parameters. We first examine the reactions leading to single
`
`
`
`
`correct nucleotide incorporation, then describe the kinetics of misincorporation
`
`
`
`
`
`and finally, discuss the reactions governing the selectivity of error correction
`3'-5' exonuclease.
`by the proofreading
`
`DNA Binding and Processivity
`The rate of incorporation of a nucleotide can be measured directly in a
`
`
`
`
`
`
`
`
`single-turnover experiment using a synthetic random sequence oligonucleotide
`such as the one shown below:
`
`6' -GCCTCGCAGCCGTCCAACCAACTCA
`
`
`CGGAGCGTCGGCAGGTTGGTTGAGTAGGTCTTGTTT-6 '
`
`Annu. Rev. Biochem. 1993.62:685-713. Downloaded from www.annualreviews.org
`
` Access provided by Yeshiva University - Albert Einstein College of Medicine on 03/29/21. For personal use only.
`
`

`

`DNA POLYMERASE MECHANISM 695
`
`Using a 1:1 complex of enzyme and DNA, the rate of incorporation of the
`
`
`
`
`correct nucleotide (dTIP in this case) can be measured directly using a rapid
`
`
`chemical quench flow instrument in what is termed a single-turnover
`
`
`
`experiment. The rapid chemical quench flow instrument simply allows
`
`
`
`measurement of reactions on a faster time scale than can be done manually.
`
`
`
`The important feature of the experiment is the use of a 1: 1 E-DNA complex
`
`
`
`(often with enzyme in slight excess), because this allows direct observation
`
`and measurement of single reactions.
`,
`
`
`
`is an active-site with any polymerase The fIrst experiment to be performed
`
`
`
`titration to measure the I«I for forming a productive E-DNA complex, based
`
`
`
`upon the amplitude of the fast incorporation of a single nucleotide. This
`
`
`
`
`experiment provides a direct measurement of the concentration of E-DNA
`
`
`
`complexes as a function of DNA concentration because the incorporation
`
`
`
`
`reaction is much faster than dissociation of the E-DNA complex. Moreover,
`
`
`
`because the assay is a function of the polymerization reaction, the measure­
`
`ment provides the K.J for the binding of DNA to enzyme in a productive
`
`
`
`complex. These experiments show that the binding of DNA to the enzyme is
`
`
`
`tight, exhibiting a K.I of approximately 18 nM and a slow dissociation rate
`
`
`
`
`equal to 0.2 sec -I. Similar experiments have given dissociation constants of
`
`5 nM for KF (32), 4 nM for HIV RT (99), and approximately an order of
`
`
`
`magnitude higher for T4 polymerase (116). One is forced to conclude that
`
`
`
`the weak binding of the T4 polymerase must represent a nonphysiological
`
`
`state due to the absence of the accessory proteins. A low I«I has become
`
`
`
`diagnostic for establishing conditions for the formation of an E-DNA complex
`prior to beginning extensive mechanistic studies.
`
`
`Processivity of the polymerase, defIned as the average number of bases
`can
`
`
`polymerized per encounter of the enzyme with a DNA primer/template,
`
`
`be calculated simply from the ratio of the rate of polymerization (300 sec-I)
`
`
`to the rate of dissociation of the E-DNA complex (0.2 sec-I). Thus, under
`
`
`
`the conditions of this assay, the processivity of T7 DNA polymerase is 1500
`
`
`base pairs. Corresponding values for KF and T4 polymerase are 250 and 800,
`
`
`
`respectively, as summarized in Table 1. While the low value for KF may
`
`reflect its role as a repair enzyme, the modest processivities of T4 and T7
`
`
`
`must increase upon the assembly of a complex at a replication fork to fully
`
`
`account for the replication of the viral genome in vivo (8). The simple systems
`
`
`
`
`involving the polymerase alone define the basic mechanism and kinetics of
`
`
`
`
`reaction. The pertinent rates of reaction can then be altered by the addition
`
`
`
`
`of accessory proteins or other enzymes necessary for replication through a
`
`
`
`suggests that the interaction fork (see below), For example, some evidence
`
`
`
`
`of the leading-strand polymerase with the helicase may increase its processiv­
`ity (84).
`
`Annu. Rev. Biochem. 1993.62:685-713. Downloaded from www.annualreviews.org
`
` Access provided by Yeshiva University - Albert Einstein College of Medicine on 03/29/21. For personal use only.
`
`

`

`696 JOHNSON
`
`Table 1 Processivity
`of DNA polymerases'
`Polymerase kPOl kerr kpolikorr
`
`300 0.2 1500
`
`1'7
`T4
`KF
`50 0.2
`250
`HIV RT (DNA) 30 0.16 200
`HIV RT (RNA) 70 0.06 1200
`
`250 0.3
`
`800
`
`a Processivity estimates are calculated from the rates
`
`
`(kpol) div