Structural and Enzymatic Studies of the T4 DNA Replication System

Vol. 264, No. 21, Issue of July 25, pp. 12717-12729,1989 Printed in U.S.A.

THEJOURNALOF BIOLOGICAL CHEMISTRY Q 1989 by The American Society for Biochemistry and Molecular Biology, Inc.

Structural and Enzymatic Studies of the T4 DNA Replication System 11. ATPasePROPERTIESOFTHEPOLYMERASE

ACCESSORY PROTEINCOMPLEX* (Received for publication, November 28, 1988)

Thale C. JarvisS, Leland S. Paul& JoelW. HockensmithlI, and PeterH. von Hippel11 From the Instituteof Molecular Biology and Department of Chemistry, University of Oregon, Eugene, Oregon97403

In this paper we report a detailed enzymatic char- DNA-binding protein) can compete with the polymercomplex forsingle-stranded acterization of the interactionof the polymerase acces- ase accessoryprotein sory proteincomplex of the T4DNA replication system DNA but not for the primer-templatejunction activawith the various nucleic acid cofactors that activate tion sites. The implications of these results for the the ATPase of the complex. We show that theATPase structure and function of the polymerase accessory DNA replication system activity of the T4coded gene 44/62 proteincomplex is protein complex within the T4 stimulated synergistically by binding of DNA and T4 are discussed. gene 45 protein and that the level of ATPase activation appears to be directly correlated with the binding of nucleic acid cofactor. Bindingof any partially orcomAccurate and efficient DNA replication of bacteriophage pletely single-strandedDNA to thecomplete accessory T 4 requires the co-ordinated action of a multienzyme repliprotein complex increasesthecatalyticactivity(as measured by V,,,,,)while decreasing the binding affin-cation complex. The ability to reconstitute functional repliity for the ATP substrate. While single-stranded DNA cation complexes i n vitro has greatly facilitated the study of is a moderately effective cofactor, we find that the detailed interactionsbetween the individual replication comoptimal nucleic acid-binding site for the complex is the ponents of the system and thenucleic acid template. The T 4 primer-template junction, rather than single-stranded genes 43,44,62,45, and32 encode the five proteins essential for leading strand synthesis from a nicked duplex template DNA ends as previously reported in the literature. Gene 45 protein playsan essential role in directing the(strand displacement synthesis) in vitro (1, 2). Of particular specificity of binding to primer-templatesites, lower- interest in this paper are the gene 44, 62, and 45 proteins, ing theK , for primer-templatesites almost 1000-fold, known collectively as the “polymerase accessory proteins.” and increasingV,,, 100-fold, compared with the anal-Gene 44 and 62 proteins are isolated as a tight complex and ogous values for gene 44/62 protein alone. The most have a DNA-dependent (d)ATPase activity (ATP+ ADP + effective primer-template site for binding and enzyP,) that is stimulatedby the gene 45 protein (3, 4). Together, matic activation has the physiologically relevant re- these proteins function to increase the rate, processivity, and cessed 3‘-OH configuration and an optimal size in ex- fidelity of the T4 DNApolymerase (gene 43 protein) (5-7). cess of 18 base pairs of duplex DNA. We find that the While numerous observations have helped to pinpoint the chemical nature of the primer terminus(Le. 3”OH or role of the accessory proteins in the replication apparatus, the 3“H) does not affect the extent of ATPase activation and that binding of the polymerase accessory protein precise nature of theirfunctionalinteractionswithinthe complex has remained elusive. It is known that theaccessory complex to DNA cofactors is salt concentration deproteins stimulatepolymerase exonuclease activity onduplex pendent but appreciably less so when the activating DNA is a primer-template junction. Finally, we show DNA, while inhibiting it on single-stranded DNA, and that that the gene 32 protein (T4 coded single-stranded these effects require ATP hydrolysis (8, 9). Furthermore, the gene 45 protein alone is capable of inhibiting the incorpora*This research was supported in part by United States Public tion of incorrect nucleotide residues by polymerase, thereby Health Service (USPHS) Research Grants GM-15792 and GM-29158 increasing its fidelity (10). The accessory proteins have been (to P. H. v. H.), by Individual Postdoctoral National ResearchService shown to increase markedly the processivity of polymerase Award GM-09353 (to J. W. H.), and by a grant from the Lucille P. during synthesis (5,7), well as as the rate at which polymerase Markey Charitable Trust. TheApplied Biosystems DNA Synthesizer moves through regions of secondary structure on singlewas purchased under an equipment grant from the Murdock Charitable Trust and National Science Foundation Equipment Grant DMB stranded templates (11).Thus, these proteins are capable of 8507352. The costs of publication of this article were defrayed in part modulating the activitiesof the DNA polymerase in both its by the payment of page charges. This article must therefore be hereby synthesis and its “editing”(3’-5’ exonuclease) modes. marked “aduertisement” in accordance with 18 U.S.C. Section 1734 The role of ATP hydrolysis in facilitating these functions solely to indicate this fact. remains tobe defined. Although Piperno andAlberts (6) have $ Predoctoral trainee supported by USPHS InstitutionalResearch five-protein system)that less than 1 ATP Service Award GM-07759. Submitted to the GraduateSchool of the shown(inthe University of Oregon in partial fulfillment of the requirements for molecule is hydrolyzed/lO nucleotides incorporated by polymerase on a single-stranded template, the comparablelevel of the Ph.D. degree in Chemistry. § Predoctoral trainees supported by USPHS InstitutionalResearch ATP hydrolysis during strand displacement synthesis has not Service Award GM-07759. Present address: AbbottLaboratories, been reported. The resultsof Newport et al. (5) indicate that Abbott Park, IL 60664. for synthesis in the absence of secondary structure (i.e. syn7 Present address: Dept. of Biochemistry, University of Virginia, thesis on a homopolymeric template), ATP hydrolysis is reCharlottesville, VA 22908. 11 American Cancer Society ResearchProfessor of Chemistry. quired for the assembly of a processive complex, but not for moving the complex along the DNA (i.e. there is no coupling Present address: Dept. of Molecular, Cellular, and Developmental Biology, University of Colorado, Boulder, CO 80309. of the rate of ATP consumption with the numberof nucleo-

12717

12718

ATPase Propertiesof the Accessory Protein Complex

tides incorporated). The role of ATP hydrolysis in the elon- of the later assays were performed in buffer Bcontaining6 mM gation phaseof synthesis on natural DNA templates, onwhich Mg(OAc)n,60 mM KOAc, 5 mM P-mercaptoethanol, 25 mM Tris-OAc, polymerase may encounter kinetic barriers due to sequence pH 7.5, and 1.0 mM ATP. Comparisons of the ATPase activities in the two buffers are given under “Results” (Table11). or secondary structure, remains somewhat unclear. singleOn Unless otherwise stated, assays were performed spectrophotometstranded fdDNA, thestimulatory effect of the accessory rically, using a coupled enzyme system (17) to monitor the oxidation replications on synthesis persistsup to 10 min after addition of NADH at 340 nm. Reactions were followed in an HP8450A UV of the nonhydrolyzable ATP analog ATP-7-S’(11); addition spectrophotometer equipped with an 89100A temperature controller. of ATP-7-S toa strand displacement synthesis reaction stopsExcept where indicated,assay temperature was 37 ”C. Reactions typically contained 10units/ml each of rabbit muscle pyruvate kinase the reactionwithin 10 s (11, 12). This suggests that the and lactic dehydrogenase (Sigma). Stock ammonium sulfate suspenrequirement for ATP hydrolysis may be considerably higher sions of these “regeneration enzymes” were spun for 5 min in an when the polymerase is synthesizing throughduplex regions. Eppendorf centrifuge. The pellet was briefly rinsed with water and The hydrolysis of ATP may provide energy for melting of resuspended in assay buffer. The initial reaction concentration of NADH was 0.10-0.13 mg/ml, and phospho(eno1)pyruvate (Sigma) helical regions in advance of the polymerase, in addition to assembling the complex; in this case there should be a direct was initially present at 3 mM. The time course of the reaction was coupling between the number of base pairs melted and the regularly assayed to determine the “linear lifetime” of the accessory protein activity. This ranged anywhere from 5 to 30 min, depending amount of ATP consumed. on enzyme batch and buffer conditions. In the preceding paper, we have examined the physical The colorimetic phosphate assay directly measures the amount of association states of the polymerase accessory proteins. To inorganic phosphate released by ATP hydrolysis and follows a modapproach the questionof the mechanism whereby the acces- ification of the procedure described by King (18). 100-pl reactions sory protein complex modulates the functional activity of the were quenched by addition of 225 pl of 10% sodium dodecyl sulfate, by 350 p1 of double-distilled water, and 180 pl of 1:l 5% replication complex from an enzymatic, structural and phys- followed ammonium molybdate, 60-72% perchloric acid. Absorbance at 720 ical chemical point-of-view, we have undertaken a study of nm was read 10 min after addition of 45 pl of aminonaphtholsulfonic the ATPase activity of the polymerase accessory proteins, acid reagent, and compared with a standard phosphate curve. initially in the absence of other replication proteins. Studyof A third assay, particularly sensitiveat low concentrations of ATP, the interactionsbetween the individual proteins, definition of utilized polyethylenimine-cellulose F thin layer chromatography the nucleic acid-binding site and the elucidationof the rela- plates (EMScience) to separate ATP andP,. The reactions contained about 0.1 pCi/pl of [Y-~’P]ATP andwere quenched by “spotting” 1 tionship of binding to ATPase activity has led to a number of pl on the thin layer chromatography plate. Plates were developed in insights into therole of the polymerase accessory proteins in 0.3 M potassiumphosphate, pH 7.0, and autoradiographed. Spots the T4DNA replication system. corresponding to ATP andPi were cut out andcounted in a Beckman MATERIALS ANDMETHODS

Protein Purification”T4 gene 45 protein, gene 32 protein, and the gene 44/62 protein complex, were purified from T4 infected cells as described in the preceding paper (13). We have shown (13) that the gene 44 and 62 proteins form a tight complex consisting of four 44 subunits andone 62 subunit with amolecular mass of 163,700daltons; gene 45 protein associates to form a trimer with a molecular mass of 74,100 daltons. Protein concentrations were obtained using calculated molar extinction coefficients, as described in the companion paper (13, 14). Thus we use, for a 4:l complex of gene 44/62 protein, = 1.23 X lo5 l/mol(complex)-cm, or 1 Azso= 1.33 mg/ml. For gene 45 protein, ezs0 = 5.72 X 10‘ l/mol(trimer)-cm, or 1AZsO= 1.30 mg/ml. Preparation of Nucleic Acids-The following oligonucleotides were synthesized by the University of Oregon Biotechnology Laboratory onan Applied Biosystems model 380B DNA Synthesizer,using standard phosphoramidite chemistry (15): 5’-GCG(A),,, 5’(T)&GC, 5’-(A),oGCG, and 5’-CGC(T),o. These oligonucleotides were purified by reverse-phase high pressure liquid chromatography (16), and trityl groups were removed by treatment with 3% acetic acid, followed byether extraction. Homopolynucleotides and all other oligonucleotides were obtained from Pharmacia LKB Biotechnology Inc. S1-treated double-stranded DNA was prepared by 30 min of incubation of 500 mgof native calf thymus DNA (Sigma) with 100 units of S1 nuclease (Bethesda Research Laboratory (BRL)) in 0.5 ml of S1 buffer (BRL) at37 “C. Dideoxythymidine (Pharmacia) was added to oligo(dT),, using terminal deoxyribonucleotidyl transferase (IBI), in a 0.3-ml reaction containing 20 pM,,, of the oligonucleotide, 0.9 mM dideoxythymidine, and 15 units of terminal transferase. The reaction was incubated for 1 h at 37 “C, followed by 10 min at 70 “C, and went >98% to completion, based on gel electrophoresis of the products. The final product was separated from unreacted ddT by Sephadex G-25-80 chromatography (Sigma). Restriction fragments of pBR322 were a gift from William Vorachek (this laboratory). ATPase Assays-Earlier work was done in buffer A containing 0.5 mM ATP, 6 mM MgCl,, 20 mM NaC1, 1 mM (3-mercaptoethanol, 20 mM HEPES, pH 7.5, and 100 pg/ml bovine serum albumin. In order to facilitate comparison with published work on T4replication, most The abbreviationsused are: ATP-7-S, riboadenosine 5’-0-(3thiotriphosphate);HPLC, high pressure liquid chromatography; HEPES, 4-(2-hydroxyethyl)-l-piperazineethanesulfonicacid; p~,,,, micromolar in nucleotide residues.

LS7000 liquid scintillation counter, using Ecoscint (National Diagnostics) as a fluor. The different ATPase assays yielded identical rates in side-by-side comparisons. RESULTS

Activation of the 44/62 ATPase Activity Operating in the absence of a nucleic acid cofactor, the gene 44/62 protein complex is able to catalyze ATP hydrolysis only at very low rates (Table I). Addition of either gene 45 protein or DNA results in a modest increase in this activity. Stimulation of the gene 44/62 protein ATPase activity by gene 45 protein in the absence of DNA is clearly not due to nucleic acid contamination of the gene 45 protein preparation, since this ATPase activation can be eliminated by heat inactivation gene 45 of the gene 45 protein. Binding of both DNA and the protein has asynergisticeffect (as observed by Mace and Alberts (4)), in which the fully activated complex hydrolyzes ATP at a much higher rate than would be expected from component additivity. Table I summarizes the maximal turnover rates we have observed in the presence of saturating amounts of each activator. In addition, distinct changes in K, for the substrate are produced upon bindingof each of the two activators. As shown in Table I, binding of DNA actually raises the K,,, for ATP relative to that measured forgene 44/62 + 45 proteins in the absence of DNA. The possiblesignificance of this will be addressed under “Discussion.” Determination of the Optimal Nucleic AcidCofactor for ATPase Stimulation We will approach the issue of the nucleic acid recognition site of the accessory proteins by initially describing the enzymatic properties of the accessory protein complex as a whole, i.e. gene 45 and 44/62 proteins together. As will be shown below, the interaction between the gene 45 and 44/62 proteins is relatively weak, so that only a small proportion of the gene 44/62 protein complex will have gene 45 protein

ATPase Properties of the Accessory Protein Complex TABLE I ATPase turnover rates and K, for substrate for various components of the polymeraseaccessory protein complex Reaction comuonents” kcat (min”)b K , for ATP‘ K , (lit.)d

12719

We have measured thelevel of ATPase activity of the gene 44/62-45 protein complex in the presence of a variety of single-stranded nucleic acid homopolymers. The rate of ATP hydrolysis, plotted asa function of polynucleotide concentration, generates curves that can be readily fit assuming MiPM chaelis-Menten type cofactor saturation (i.e. direct non-co95%, under these conditions) TABLEI1 by the characteristics of the gene 45 protein-bound species. Therefore, even though the following studies are not all per- ATPase rates for the gene 44/62 + 45 protein complex as a function of nucleotide concentrationfor homopolynucleotide cofactors formed at saturating amountsof gene 45 protein, the results Assays were performed in buffer A (see“Materials and Methods”), nevertheless reflect the nucleic acidbinding characteristicsof a t 35”C, with [44/62] = 73 nM, and [45] = 160 nM. K, values are in the complete accessory protein complex. p~ nucleotide residues. Note that many of the subsequent experiStimulation by Single-stranded Polynucleotides-Previous ments are done in buffer B, in order to facilitate comparison with studies of the ATPase activityof the accessory proteins dem- published literature onT4 in vitro DNA replication experiments. For onstrated that the ATPase activity of the complex is best reference, the same experiment in buffer B yields a K, of 4.7 p~~~~ stimulated by short (sonicated) single-stranded pieces of nat- and a Vmaxof 22 nmol/min/ml for poly(dT) (see discussion in “Salt ural DNA (3). Longer single-stranded DNA isless efficacious Dependence of ATPase Activity” section). Cofactof K”. V”..,b as a cofactor and double-stranded DNA is worse yet. These results led to the conclusion that the gene 44/62-45 protein PM nrnolfrninfrnl complex recognizes and is activated by single-stranded DNA 15 t 2 POlY(dA) 6.6 2 0.6 Poly(dT) 6.3 f 0.7 8.9 f 1.2 ends. Piperno et al. (3) found no preference for the primerPoly(d1) 12 f 1 11 f 1 template model poly(dA).oligo(dT)lo when compared with Poly(dC) 3.2 f 0.2 32 2 5 poly(dA). In contrast, Mace and Alberts (4), working under Poly(dU) 110 f 30 4.4 f 0.7 slightly different conditions, reporta moderate preferencefor Poly(rA) 6.4 f 2.0 260 2 80 partially double-stranded cofactors. Thus, there is some conPoly(rI), poly(rU), poly(rC), and poly(rcA) were tested up to 69, fusion in the literature as to the optimal cofactor for this ~ respectively, ~ ~ and ~ showed , little or no stimula114, 140, and 42 p enzymatic activity. With this in mind,we have undertaken a tion of the protein ATPaseactivity. more detailed study of the enzymaticbehavior of the accessory The errors reported indicate the standard error obtained in the protein complex using betterdefined nucleic acid cofactors. nonlinear curve fitting of the data.

*

ATPase Propertiesof the Accessory Protein Complex

12720

TABLE111 ATPase rates for the gene 44/62 + 45 protein complex as a function of nucleotide concentration for oligonucleotide cofactors Conditions as in Table 11.

Competition Assayswith Polynucleotide Cofactors-The assumptions implicit in the previous discussion are that the observed ATPase activity directly reflects the binding behavior of the DNA cofactor and that all single-stranded polyCofactor K, Vm*X nucleotides bind to a common protein-binding site on the accessory protein complex. This view is strengthened by the PM nmollminlml observation that the various polynucleotides (and long oligo( W Z O 20 f 9 11 f 2 13 f 2 nucleotides) stimulate protein ATPase activityto comparable (dA),o-m 5.8 f 0.5 >60