φ29 DNA Polymerase–Terminal Protein Interaction. Involvement of Residues Specifically Conserved Among Protein-primed DNA Polymerases

f29 DNA Polymerase–Terminal Protein Interaction.Involvement of Residues Specifically ConservedAmong Protein-primed DNA Polymerases

Irene Rodrıguez, Jose M. Lazaro, Margarita Salas* and Miguel de Vega

Instituto de Biologıa Molecular“Eladio Vinuela” (CSIC)Centro de Biologıa Molecular“Severo Ochoa” (CSIC-UAM)Universidad Autonoma, CantoBlanco, 28049 Madrid, Spain

By multiple sequence alignments of DNA polymerases from the eukary-otic-type (family B) subgroup of protein-primed DNA polymerases wehave identified five positively charged amino acids, specifically con-served, located N-terminally to the (S/T)Lx2h motif. Here, we havestudied, by site-directed mutagenesis, the functional role of f29 DNApolymerase residues Arg96, Lys110, Lys112, Arg113 and Lys114 in specificreactions dependent on a protein-priming event. Mutations introduced atresidues Arg96, Arg113 and Lys114 and to a lower extent Lys110 andLys112, showed a defective protein-primed initiation step. Analysis ofthe interaction with double-stranded DNA and terminal protein (TP) dis-played by mutant derivatives R96A, K110A, K112A, R113A and K114Aallows us to conclude that f29 DNA polymerase residue Arg96 is animportant DNA/TP-ligand residue, essential to form stable DNA poly-merase/DNA(TP) complexes, while residues Lys110, Lys112 and Arg113could be playing a role in establishing contacts with the TP-DNA templateduring the first step of DNA replication. The importance of residue Lys114to make a functionally active DNA polymerase/TP complex is also dis-cussed. These results, together with the high degree of conservation ofthose residues among protein-primed DNA polymerases, strongly suggesta functional role of those amino acids in establishing the appropriateinteractions with DNA polymerase substrates, DNA and TP, to success-fully accomplish the first steps of TP-DNA replication.

q 2004 Elsevier Ltd. All rights reserved.

Keywords: DNA polymerase; site-directed mutagenesis; linear DNAreplication; protein-priming; terminal protein*Corresponding author

Introduction

To achieve a successful duplication of the gen-ome, the first steps of DNA replication are crucial,and a large number of interactions between dif-ferent molecules such as protein–protein and pro-tein–DNA interactions are needed.1,2 ReplicatingDNA polymerases are the central enzymes toaccomplish such a complex process.

One of the mechanisms used to initiate replica-tion of linear double-stranded genomes is the useof a terminal protein (TP) as primer, which pro-

vides the OH group needed to initiate DNA syn-thesis. This kind of initiation process is widelyused by organisms such as eukaryotic virus, bac-teriophages and linear plasmids, the most studiedones being Bacillus subtilis bacteriophage f29 andadenovirus.3,4

Bacteriophage f29 genome is a linear double-stranded DNA (dsDNA) of 19,285 base-pairs (bp),with a 6 bp inverted terminal repeat (30-TTTCAT-50)and a TP covalently linked to each 50 end.2–4 f29DNA replication requires the formation of a hetero-dimer between the f29 DNA polymerase and afree TP molecule, which recognises and interactswith the replication origin at both ends of the gen-ome. On the other hand, a histone-like viral protein(p6) forms a nucleoprotein complex at the replica-tion origins that probably contributes to theunwinding of the double helix at the DNA ends.5

During initiation, f29 DNA polymerase catalyses,

0022-2836/$ - see front matter q 2004 Elsevier Ltd. All rights reserved.

E-mail address of the corresponding author:[email protected]

Abbreviations used: bp, base-pair; ss-, and dsDNA,single-stranded and double-stranded DNA, respectively;TP, terminal protein; TP-DNA, f29 terminal protein-containing genome; BSA, bovine serum albumin.

doi:10.1016/j.jmb.2004.02.018 J. Mol. Biol. (2004) 337, 829–841

at both DNA ends, the addition of dAMP to theOH group of Ser232 in the TP, using as directorthe second dTMP of the template strand.6 – 8 Sub-sequently, the complex slides back one nucleotideto recover the information of the 30-terminaldTMP, sequence reiteration being a requisite forthe so-called sliding-back mechanism. End term-inal sequence reiteration also occurs in other lineargenomes that contain a TP covalently linked totheir DNA ends, such as the B. subtilis, Escherichiacoli and Streptococcus pneumoniae phages GA-1,9

PRD110 and Cp-1,11 respectively, linear plasmids,3

and the eukaryotic adenovirus.12 Indeed, thereplication initiation site in GA-1, PRD1, Cp-1,and adenovirus corresponds to an internal nucleo-tide close to the 30-terminal end, and a sliding-back type of mechanism has been shown to occurin these cases to recover the information of theterminal nucleotide(s). For f29 DNA replication, ithas been demonstrated that the same DNA poly-merase that catalyses the initiation reaction, carriesout the synthesis of a short elongation product(6–9 nt long), while still complexed with the TP(transition step).13 When the tenth nucleotide isinserted onto the nascent DNA chain, the DNApolymerase dissociates from the primer TP,13 andelongation proceeds processively from both DNAends coupled to strand displacement without theneed for either processivity factors or helicase-likeproteins.14 Once the two replication forks meet,the two partially replicated parental strands separ-ate, and full-length DNA synthesis is completed.14

Bacteriophage f29 DNA polymerase is a singlepolypeptide of 66 kDa, which belongs to theB-type superfamily of DNA-dependent DNA poly-merases (also referred to as eukaryotic or a-likepolymerases) that includes a vast group of DNApolymerases from eukaryotic and prokaryoticorigins.2,4,15 It displays both exonuclease and poly-merisation activities that are located in two struc-turally independent domains.16,17 The f29 DNApolymerase has served as model for the study ofthe structure–function relationships of DNA poly-merases by means of an exhaustive mutationalanalysis of the enzyme.17 Thus, the N-terminalregion of the f29 DNA polymerase, that containsthe 30-50exonuclease activity, has the five catalyticamino acids belonging to the three motifs, Exo I,Exo II and Exo III, absolutely conserved in allproofreading DNA polymerases.18 The C-terminalregion of f29 DNA polymerase contains the syn-thetic activities. Five motifs present in this region,Dx2SLYP (motif A or 1), Kx3NSxYG (motif B or2a), Tx2G/AR (motif 2b), YxDTDS (motif C or 3)and KxY (motif 4), evolutionarily conserved invarious families of DNA-dependent DNA poly-merases, allowed us to identify active-site residueswith a specific role in catalysis and interactionwith the three substrates, DNA, TP and dNTPs,used by the f29 DNA polymerase.16,17

Taking into account the bimodular structure ofDNA polymerases and the fact that those that use aprotein-priming mechanism have to accommodate

a TP (31 kDa in the case of bacteriophage f29) inthe same cleft occupied by the DNA during furtherreplication steps,19 it is expected that residues fromboth domains will contribute to the stabilisation ofthe DNA polymerase/TP interaction. Thus, in theC-terminal polymerisation domain of the f29DNA polymerase, residues belonging to motifsTx2GR,18 YxGG/A20 and TPR-121 were identified asbeing involved in the interaction with TP. On theother hand, isolated expression of the C-terminaldomain of f29 DNA polymerase led to a dramati-cally reduced interaction with the TP and DNApolymerase,22 indicating a role of the N-terminaldomain in the TP/DNA polymerase interaction.In fact, it has been possible to identify residues inthis domain, belonging to motifs (S/T)Lx2h,19,23,24

and Exo II,19,23,25 that establish contacts with the TP.Here, we investigate by means of mutational

analysis the role of f29 DNA polymerase residuesArg96, Lys110, Lys112, Arg113 and Lys114, locatedN-terminally to the “(S/T)Lx2h” motif of theN-terminal domain of protein-priming DNA poly-merases. Based on our results, we propose a func-tional role of these residues in making propercontacts with the TP.

Results

Conservation of amino acid residues at the30-50 exonuclease domain of protein-primedDNA polymerases

Figure 1 shows a multiple alignment of theregion containing the consensus sequence S/TLx2hin the subgroup of eukaryotic-type DNA poly-merases that use a protein-priming mechanism.This region is located between the Exo II and ExoIII motifs, present in all DNA-dependent DNApolymerases endowed with a 30-50 exonucleaseactivity.19,23,24

As shown in Figure 1, a positively charged resi-due, specifically conserved in 22 out of 25 DNApolymerases, is located 24–28 residues N-termin-ally (with the exception of PAI-2 DNA polymerase,placed 32 residues N-terminally) from the Ser resi-due of the S/TLx2h motif. Between this positivelycharged amino acid residue and the S/TLx2hmotif, there is a positively charged region, as canbe observed in Figure 1. The first residue of thisregion is specifically conserved in 19 out of 25 pro-tein-priming DNA polymerases. The second posi-tive residue, located two positions C-terminally, isconserved in 13 polymerases. Following this, thereis another positively charged residue conserved in15 DNA polymerases. Finally, the fourth positivelycharged residue, which follows the latter, is con-served in 16 DNA polymerases. Thus, the highlevel of conservation of these five positivelycharged amino acids prompted us to study thephysiological role of these residues by makingsite-directed mutants at the corresponding Arg96,Lys110, Lys112, Arg113 and Lys114 residues of f29

830 f29 DNA Polymerase–Terminal Protein Interaction

DNA polymerase. All the residues were changedto alanine, to eliminate the positive charge. Takinginto account secondary structure predictions26,27

and general suggestions for conservative substi-tutions,28 all the mutations introduced were pre-dicted to maintain the overall structure in thatregion of the polymerase. The mutant DNA poly-merases were overexpressed and purified asdescribed elsewhere29 and analysed using a varietyof in vitro assays corresponding to the differentstages of TP-primed f29 DNA replication. Thehigh level of conservation in the protein-primingsubgroup of eukaryotic DNA polymerasescould suggest a role in specific functions intrinsicto the mechanism used by this kind of polymerases.

Mutation of residues Arg96, Lys110, Lys112,Arg113 or Lys114 of f29 DNA polymerase donot affect the Pol/Exo balance on a primer/template structure

By using a 50-labelled oligonucleotide (15-mer)hybridised to a larger template oligonucleotide(21-mer), it is possible to analyse the equilibriumbetween the 30-50 exonucleolysis and 50-30 poly-merisation activities, since the products of both

of them on this primer-template structure can besimultaneously detected. The balance betweensuch activities (Pol/Exo balance) will depend onthe velocities of the two reactions, and on the rela-tive affinity of both active sites for the primer-strand. As it can be seen in Figure 2, the onlyproducts that can be detected in the absence ofnucleotides for both the wild-type and mutant f29DNA polymerases, are those yielded by the exo-nucleolytic digestion of the primer strand (Figure 2,lane 0 dNTP). In these conditions, the pattern andextent of primer degradation will reflect the levelof the 30-50 exonuclease activity of the mutantderivatives with respect to the wild-type enzyme.Although the degradation pattern shown by allthe mutant enzymes was similar to that of thewild-type DNA polymerase, the extent of degra-dation by mutant polymerase R96A was lower. Infact, when the 30-50 exonuclease activity of mutantDNA polymerases on such primer/template sub-strate was studied using limiting amounts of theenzyme, mutant R96A showed a 2.5-fold reductionof this activity compared with the wild-typeenzyme (see Table 1). As the dNTP concentrationincreases, the 30-50 exonuclease activity is progres-sively competed out by the polymerisation, and it

Figure 1. Multiple amino acid sequence alignment of the amino acid region located N-terminally to the S/TLx2hmotif of protein-primed DNA polymerases. DNA polymerases nomenclature and sequences are compiled byBraithwaite & Ito45 with the exception of linear mitochondrial plasmid DNA polymerases from Gelasinospora sp. (acces-sion number S62752), Brassica napus (accession number NP862323), Flammolives velutipes (accession number BAB13499),Pichia kluyveri (accession number T28426) and Porphyra purpurea (accession number NP049297); and DNA polymerasesfrom bacteriophages NF,46 PZA (accession number PO6950), B103 (accession number CAA67649) and CP1 (accessionnumber Q37989). Numbers indicate the position of the first aligned amino acid with respect to the N terminus of therespective DNA polymerase. Highly conserved residues are shown in white letters over a black background. Con-served residues are shown in boldface over a grey background. Residues studied here are indicated with asterisks(Arg96, Lys110, Lys112, Arg113 and Lys114 residues of f29 DNA polymerase) and those studied in previous workswith a black circle (Ser122 and Leu123 residues of f29 DNA polymerase19,23).

f29 DNA Polymerase–Terminal Protein Interaction 831

is possible to find a particular dNTP concentrationrequired to obtain an efficient elongation that canbe used to define the productive Pol/Exo ratio,which is independent of the amount of enzyme/DNA complexes formed (see Materials andMethods). As shown in Figure 2, the mutant poly-merases displayed a similar dNTP requirement(100 nM) as the wild-type enzyme for a net poly-merisation balance (see Table 1). This resultsuggested that the mutations introduced at thoseresidues did not affect the polymerase activity ofthe enzymes, as can be expected, considering thatthey are located at the N-terminal domain of theDNA polymerase. Although the mutant polymer-ase R96A showed a defect in its degradativeactivity, its wild-type Pol/Exo balance suggeststhat the mutation is affecting the general inter-action with the dsDNA (notice that a certainamount of substrate was not used by the R96Amutant). On the other hand, when the fidelity ofthe polymerisation process was analysed, byperforming a misincorporation assay (see Materialsand Methods) in which the same sp1/sp1c þ 6

substrate molecule and increasing amounts ofdATP as the sole deoxynucleotide were used, allthe mutant DNA polymerases only extended theprimer until the 16 and 17-mer position, as wasthe case with the wild-type enzyme, indicatingthat misincorporation of dATP at non-complemen-tary positions was not produced (not shown).

Interaction of mutant f29 DNA polymeraseswith DNA

The interaction of the different f29 DNA poly-merase mutants at residues Arg96, Lys110, Lys112,Arg113 and Lys114 with a primer/template DNA(sp1/sp1c þ 6) was analysed by using gel shiftassays, as described in Materials and Methods.Under these conditions, in the presence of Mg2þ,the wild-type DNA polymerase gives rise to asingle retardation band whose intensity dependson the amount of enzyme added, and interpretedas an enzyme/DNA complex competent forpolymerisation30 (see Figure 3). As also shown inFigure 3, f29 DNA polymerase mutants K110A,

Figure 2. Pol/Exo balance by wild-type or mutant f29 DNA polymerases. The assay was carried out as described inMaterials and Methods using 32P-labelled hybrid molecule 15/21-mer as primer/template DNA and the indicatedconcentration of dNTPs. Polymerisation or 30-50exonuclease activity is detected as an increase or decrease, respectively,in the size (15-mer) of the 50-labelled primer. The productive Pol/Exo ratio and the relative polymerisation for eachmutant polymerase are indicated in Table 1.

Table 1. Enzymatic activities of wild-type and mutant derivatives of f29 DNA polymerase

f29 DNA polymerasea

Parameter assayed Substrate Wild-type R96A K110A K112A R113A K114A

30-50 Exonuclease dsDNA (15/21-mer) 100 40 105 95 94 50ssDNA (15-mer) 100 30 78 79 88 78

Enzyme/DNA bindingb dsDNA (15-mer/21-mer) 100 20 87 76 86 57Pol/Exo ratioc 15/21-mer, dNTPs 20/100 20/100 20/100 20/100 20/100 20/100f29 TP-DNA replication f29 TP-DNA, TP, dNTPs 100 2 45 54 16 3f29 TP-DNA amplification f29 TP-DNA, TP, SSB, DBP, dNTPs 100 2 26 40 11 4M13 DNA replication Primed M13 DNA, dNTPs 100 148 97 96 110 104f29 TP-DNA initiation f29 TP-DNA, TP, dATP 100 4 30 45 25 4TP-deoxynucleotidylation TP, dATP 100 3 100 120 110 10

a Numbers indicate the average percentage of activity relative to the wild-type enzyme obtained from several experiments.b Analysed by gel retardation assay on low-ionic strength polyacrylamide gels.c Numbers indicate the dNTP concentration (in nM) required to efficiently elongate the 15-mer primer until the 20-mer position.


K112A, R113A and, to a lesser extent, K114Ashowed a level of interaction close to that of thewild-type enzyme for the primer/template struc-ture, while mutant R96A showed a reduced bind-ing efficiency producing a faint retardation band.This result indicates that the removal of the posi-tive charge of residue Arg96 and, to a lesser extent,Lys114, affect the interaction of the enzyme with adsDNA substrate, suggesting a secondary role ofboth residues in establishing the required contactsbetween the enzyme and the DNA for the for-mation of a stable complex.

Protein-primed TP-DNA replication with themutant DNA polymerases

f29 DNA replication is a process that requireshigh processivity and strand displacement activityduring elongation. Such a process involves TP-primed initiation at both origins of the TP-DNAmolecule, this being a specific activity displayedby f29 DNA polymerase. During initiation, f29DNA polymerase catalyses the template-directedformation of a covalent complex between the viralTP and 50-dAMP (initiation step), with the furtherelongation of such a complex coupled to stranddisplacement to yield full-length f29 TP-DNA.2 – 4

Considering that these properties are intrinsic tothe f29 DNA polymerase, efficient in vitro syn-thesis of full-length f29 TP-DNA (19,285 bp) canbe achieved by using a minimal replication systembased on f29 TP-DNA, f29 DNA polymerase,and f29 TP that acts as the initiation primer.14 Asit can be seen in Figure 4A, mutant DNA poly-merases K110A and K112A and to a lesser extentR113A, were able to synthesise full-length f29TP-DNA, which allows us to rule out specificdefects in processivity and strand displacement.However, mutants R96A and K114A displayed ahighly reduced activity (50 and 35-fold, respectively)relative to the wild-type enzyme (see Table 1). Basedon the f29 DNA replication machinery, one of themost efficient in vitro isothermal DNA amplificationsystems was developed several years ago.31 By

using appropriate amounts of the four phage f29DNA replication proteins, TP, DNA polymerase,f29 dsDNA binding protein and f29 single-stranded DNA (ssDNA) binding protein, limited

Figure 3. Gel retardation ofdsDNA by wild-type or mutantf29 DNA polymerases. The assaywas carried out as described inMaterials and Methods, using50-labelled 15/21-mer synthetichybrid, as substrate, in the presenceof 1.25 ng of either wild-type ormutant f29 DNA polymerases.Samples were analysed by poly-acrylamide gel electrophoresis.Bands correspond to free DNA andto the DNA polymerase/DNAsubstrate complex detected byautoradiography.

Figure 4. A, f29 TP-DNA replication by wild-type andmutant f29 DNA polymerases. The assay was carriedout as described in Materials and Methods in thepresence of 20 ng of wild-type (wt) or mutant f29 DNApolymerases. After incubation for five minutes and tenminutes at 30 8C, relative activity values were calculated(see Table 1) and the length of the synthesised DNA wasanalysed by alkaline agarose gel electrophoresis. Themigration position of unit-length f29 TP-DNA is indi-cated. B, f29 TP-DNA amplification. The assay wascarried out as described in Materials and Methods, inthe presence of 5 ng of wild-type or mutant f29 DNApolymerases, 5 ng of f29 TP and 10 mg each of f29 DBPand f29 SSB. After incubation for 90 minutes at 30 8C,samples were processed and the amplified DNA was ana-lysed by alkaline agarose gel electrophoresis as describedin Materials and Methods. The migration position ofunit-length f29 TP-DNA (19285 bases) is indicated.


amounts of TP-DNA can be amplified over 1000-fold. The infectivity of the synthetic (amplified)f29 DNA is similar to that of the natural f29DNA obtained from virions.31 When such anamplification assay was performed, similar resultsto those described above were obtained, althoughthe activity of mutants K110A, K112A and R113Awas lower than in the f29 TP-DNA replicationassay (see Figure 4B and Table 1).

Strand displacement capacity of mutant f29DNA polymerases

As mentioned above, f29 DNA polymerase hasto couple processive DNA synthesis to strand dis-placement in the absence of accessory proteins orhelicases, to replicate the genome efficiently.14 Toanalyse whether the defects displayed by mutantDNA polymerases in replicating TP-DNA werethe consequence of specific impairment in such acoupling, we carried out a primed-M13 DNA repli-

cation assay in which f29 DNA polymerase startspolymerisation from the 30 end of a hybridisedoligonucleotide. The first replication round does notrequire strand displacement but, once completed,DNA polymerase encounters the 50 end of theoligonucleotide, the next rounds of replicationrequiring an efficient coupling of polymerisationto strand displacement. As shown in Figure 5 andTable 1, the global catalytic efficiency and size ofthe replication products (several-fold the length ofthe M13 template) obtained with mutant DNApolymerases indicates that they are not affected inprocessive strand displacement. Curiously, in thecase of mutant R96A, the activity was even higherthan that of the wild-type enzyme, in spite of thefact that it shows a reduced dsDNA bindingcapacity to oligonucleotide molecules (as describedabove). Such a phenotype could be explained as aconsequence of a recovery of the DNA poly-merase/DNA complex stability due to other inter-actions established between the much longertemplate M13 DNA and the DNA polymerase.

Initiation of f29 DNA replication by the mutantDNA polymerases

As described, mutant DNA polymerases R96A,K114A and, to a lesser extent, R113A were severelyaffected in f29 TP-DNA replication (see above) butthey were still able to perform processive DNA-primed replication. To analyse the activity of themutant derivatives in the initiation of f29TP-DNA replication, we tested their ability to cata-lyse the TP-dAMP linkage (initiation reaction). Asit can be observed in Figure 6 and Table 1, mutantsR96A and K114A displayed a drastically reduced(25-fold) TP-primed initiation activity, the rest ofthe mutants being affected two- to fourfold. Dueto the fact that the initiation of f29 DNAreplication is a template-directed event,8 the defectdisplayed by the mutants could be the conse-quence of a poor affinity for the template DNA.Such a possibility can be tested, since f29 DNApolymerase can catalyse the deoxynucleotidylationof TP in the absence of template DNA.32 Undersuch conditions, the activity of mutant R96Awas similar or even lower to the corresponding

Figure 5. Strand-displacement coupled to M13 DNAreplication by wild-type and mutant DNA polymerases.Replication of primed-M13 DNA was carried out asdescribed in Materials and Methods using 40 mMdNTPs and 100 ng of either wild-type (wt) or mutantDNA polymerases. After incubation for the indicatedtimes at 30 8C, relative activity values were calculatedfrom dNMP incorporation (see Table 1). The position ofunit-length M13 DNA is shown at the right.

Figure 6. Formation of theTP-dAMP complex catalysed bywild-type and mutant f29 DNApolymerases in the presence orabsence of f29 TP-DNA. The reac-tions were carried out as describedin Materials and Methods. Thetemplate-dependent reaction (top)was carried out in the presence of

10 mM MgCl2, 0.5 mg of TP-DNA, 15 ng of TP and 30 ng of wild-type (wt) or mutant f29 DNA polymerases. Incu-bation was for 1.5 minutes at 30 8C. The template-independent reaction (bottom) was performed in the presence of1 mM MnCl2, 15 ng of TP and 30 ng of the wild-type or mutant f29 DNA polymerases. Incubation was for threehours at 30 8C. Samples were analysed by SDS-PAGE and autoradiography. The band corresponds to the TP-dAMPinitiation complex. Quantification was by densitometric analysis of the band corresponding to the labelled TP-dAMPcomplex, detected by autoradiography.


templated TP-dAMP formation and that of mutantK114A was still low, although it was 2.5-fold higherthan in the presence of template (Figure 6 andTable 1). In the case of mutants K110A, K112A andR113A the recovery of the activity was total in theabsence of template (Figure 6 and Table 1). The lat-ter result indicates that these mutants are defectivein the interaction with the template TP-DNA. Onthe other hand, the low values obtained withmutants R96A and K114A in both assays, withand without TP-DNA, are pointing to more specificdefects in the interaction with the TP.

The ability of mutant DNA polymerases to inter-act with the TP was tested by using an interferenceassay (see Materials and Methods). Figure 7 showsthe inhibition profile of the template-independentdeoxynucleotidylation of the TP carried out bythe wild-type DNA polymerase, obtained by theaddition of increasing amounts of the previouslycharacterised mutant D249E, which is catalyticallyinactive but able to interact with the TP;33 the inhi-bition profile of mutant D249E parallels the theor-etical one, indicating a normal interaction. Sincemutants K110A, K112A, R113A and K114A hadtotal or partial TP-deoxynucleotidylation activity

(see Table 1), their interaction with TP was com-peted with increasing amounts of mutant D249E,yielding an interference pattern very similar tothat of the wild-type enzyme (Figure 7). Theseresults indicate that, at least under such conditions,these mutant derivatives retained their capacity tointeract with TP. However, the wild-type enzymewas poorly competed by mutant R96A, probablyreflecting a defective interaction with TP. Thisresult could suggest an involvement of residueArg96 of f29 DNA polymerase in making contactswith the TP.

To confirm these results, the interaction ofmutant polymerases with TP was directly analysedby ultracentrifugation, using glycerol gradients asdescribed in Materials and Methods. Wild-type ormutant K110A, K112A, R113A and K114A DNApolymerases (66 kDa) formed a heterodimer of96 kDa with TP (31 kDa), and the two proteins co-sedimented in the same fractions (Figure 8 and notshown). In contrast, the two proteins sedimentedseparately as monomers when mutant DNA poly-merase R96A was used, indicating impairment inits interaction with TP (Figure 8).

Taking the above results into account, the furthertransition steps after formation of TP-dAMP werestudied in a truncated replication assay (replicationof TP-DNA in the absence of dCTP), forcingthe wild-type DNA polymerase to stop afterTP-(dNMP)8 (replication from the left origin of thegenome) or TP-(dNMP)11 (from the right origin).In all cases the proportion between the partiallyelongated molecules [TP-(dNMP)2 – 8] with respectto the initiated products [TP-dAMP þ TP-(dNMP)2 – 8] yielded by all mutant derivatives wasessentially the same as that obtained with thewild-type enzyme (not shown). This resultsuggests that the defects displayed by mutantsR96A and K114A are restricted to the initial stepof f29 TP-DNA replication.

Discussion

f29 DNA polymerase belongs to the eukaryoticsubgroup of DNA-dependent DNA polymerasesthat use a TP as primer.2 – 4 Multiple amino acidsequence alignments in the subgroup of protein-priming DNA polymerases23 – 25 have allowed us toidentify some of the TP binding residues. Here,we have analysed the role of five positivelycharged amino acids corresponding to f29 DNApolymerase residues Arg96, Lys110, Lys112,Arg113 and Lys114 for their degradative (30-50

exonuclease) and synthetic (protein-primed andpolymerisation) activities. These residues arespecifically conserved in the protein-priming sub-group of DNA polymerases, located N-terminallyto the S/TLx2h motif whose Ser122 and Phe128residues have also been involved in TP-inter-actions.19,23

The role of these residues in the syntheticactivities of f29 DNA polymerase was addressed

Figure 7. Interference assay for TP binding by thewild-type and mutant DNA polymerases. Reactionswere carried out as described for the template-indepen-dent formation of TP-dAMP by the wild-type or mutantK110A (filled squares), K112A (filled circles), R113A(open circles) and K114A (open triangles) f29 DNApolymerases, a limited amount of TP and increasingamounts of mutant DNA polymerase D249E (opensquares) as described in Materials and Methods. Theincubation of the wild-type f29 DNA polymerase withmutant D249E was used as control of 100% competitionthat paralleled the theoretical. To analyse the interferencecapacity of mutant R96A (filled triangles), wild-typeDNA polymerase was incubated with increasingamounts of this mutant DNA polymerase. The TP-dAMP formed in the different competition conditionsrelative to that formed in the absence of competition(100%) is indicated.


by using different in vitro assays for DNA poly-merisation. None of the mutant DNA polymeraseshad a reduced efficiency in their capacity to incor-porate dNMPs to a DNA primer terminus, indicat-ing that the corresponding polymerisation activesite was not affected by the introduction of themutation. On the other hand, mutant derivativesdisplayed a wild-type-like Pol/Exo balance; thatis, the relative strength between 30-50 exonucleaseand 50-30 polymerisation activities was not altered.This result suggests that the changes introducedin the f29 DNA polymerase were not specificallyimpairing the above activities. This, together withthe fact that mutants R96A and K114A showed adiminished interaction with dsDNA substrates,suggested that these residues could be playing arole in the general stabilisation of DNA polymer-ase/DNA interactions, without affecting specifi-cally to the binding of the primer terminus ateither 30-50 exonuclease or 50-30 polymerisationactive site.

f29 DNA replication is a multistep process thatstarts with the formation of a heterodimer betweena DNA polymerase and a free TP molecule, withthe further recognition of the replication origins,placed at both DNA termini. Recently, it has beendemonstrated the involvement of a putativecoiled-coil region at the N terminus of the TPsequence in the recognition of the origins throughan interaction between the parental TP covalentlyattached at the 50-end of the genome and the TP ofthe heterodimer.34 The TP of such a complex is

used as a primer by the DNA polymerase, cata-lysing the covalent linkage of dAMP to thehydroxyl group of Ser232 of TP, a reaction guidedby the second nucleotide at the 30-end of the tem-plate strand (initiation; 2–4). Once the TP-dAMPformation has taken place, it slides-back one pos-ition by means of a sliding-back mechanism torecover the first 30 nucleotide of the templatestrand. Then, DNA polymerase catalyses a normaltemplated incorporation of dNMPs, remainingbound to the TP until a TP-(dNMP)9 product issynthesised.13 Afterwards, dissociation of TP andDNA polymerase occurs and elongation progressescoupled to strand displacement, giving rise to fullyreplicated f29 DNA molecules.

As indicated by the M13 DNA replication assays,the mutation introduced at residues Arg96, Lys110,Lys112, Arg113 and Lys114 did not affect thecapacity of DNA polymerase to accomplish pro-cessive DNA-primed polymerisation coupled tostrand displacement. Interestingly, M13 DNAallowed binding stabilisation of mutant R96A, con-trary to what happens when short substrates areused. This improvement could be a consequenceof non-specific interactions established betweenthe DNA polymerase and the longer protrudingtemplate strand, as it occurs with mutants at resi-dues playing an auxiliary role in making contactswith the dsDNA substrate, as Phe128 of the S/TLx2h motif.24

When the TP-DNA replication and amplificationcapacity of the mutant derivatives were tested,

Figure 8. Interaction betweenDNA polymerase and TP analysedby glycerol gradient centrifugation.Wild-type or mutant DNA poly-merase was incubated with TPas described in Materials andMethods. After incubation for 30minutes at 4 8C, samples wereloaded on top of a continuous15%–30% glycerol gradient (4 ml)and centrifuged as described inMaterials and Methods. Gradientswere fractionated and subjected toSDS-12% PAGE. Gels were stainedafter electrophoresis with Coomassieblue. Densitometric quantificationsin arbitrary units of both f29 DNApolymerase (filled circles) and f29TP (open circles) are represented.


mutants R96A, R113A and K114A and, to a lesserextent, mutants K110A and K112A, showed areduced efficiency in comparison with the wild-type enzyme. Since they were able to performpolymerisation coupled to strand displacement ina processive way, as deduced from M13 DNAreplication assays, we were prompted to study thefirst phases of TP-DNA replication. The analysisof the protein-primed initiation step revealed thatmutant polymerases R96A, K114A and, to a lesserextent K110A, K112A and R113A were impaired insuch an activity. However, in the absence of tem-plate DNA (TP-deoxynucleotidylation), mutantsK110A, K112A and R113A completely recoveredtheir activity. This result suggests that Lys110,Lys112 and Arg113 could be playing some role inestablishing contacts with the TP-DNA templateduring the first step of DNA replication. The lackof interaction with TP displayed by mutant R96A,as deduced from interference and glycerol gradientassays, allows us to conclude that Arg96 is animportant TP-ligand residue, essential to form astable DNA polymerase/TP heterodimer. In thecase of mutant K114A, although it was still able tointeract stably with TP, its poor initiation capacity

could be reflecting the importance of residueLys114 to make a functionally active DNA poly-merase/TP complex, probably ensuring the properorientation of TP in the DNA polymerisation activesite. This is the case with other f29 DNA poly-merase TP ligand residues as Lys305 and Asp332from the TPR-1 region, specific of the protein-primed subgroup of DNA polymerases.21,35

In contrast to other identified TP ligand residuesof f29 DNA polymerase, as Phe65, Tyr59, His61,Phe69, Ser122 and Phe128,19,23 – 25 mutations intro-duced at residues Arg96 and Lys114 did not alterthe transition step between both priming modes(protein and DNA-priming), being apparentlytheir role restricted to the first steps of DNAreplication.

In spite of the fact that crystal data are notavailable for any protein-primed DNA polymerase,our knowledge about the regions of the f29 DNApolymerase that make contacts with TP is growing.Both C and N-terminal domains contain residuesthat act as TP ligands. To date, the identifiedTP-ligand residues placed at the C-terminaldomain of f29 DNA polymerase are Thr434 andArg438 of the conserved motif Tx2GR,8 proposed

Figure 9. A, Structural representation of the 30-50 exonuclease active site of bacteriophage RB69 DNA polymerase. Inthis case co-ordinates correspond to the crystal structure of an editing complex of RB69 DNA polymerase (PDB1CLQ37), but only the five 30-terminal nucleotides of the primer strand (30-GCGCC) and the four 50-terminal nucleotides ofthe template strand (50-CGCG) are depicted. Residues represented belong to three categories: (1) directly involved incatalysis (Asp114, Glu116, Asp222, and Asp327); (2) auxiliary role in catalysis (Lys302 and Tyr323); and ssDNA ligands(Asn217, Phe221 Ser289 and Leu290). In addition, RB69 DNA polymerase residues Arg260 and Lys279, homologous tothe f29 DNA polymerase residues Arg96 and Lys114, respectively, are represented. B, Location in the three-dimen-sional structure of RB69 DNA polymerase of the homologues residues playing a dual role in DNA and TP binding inprotein-primed DNA polymerases. The subdomains of RB69 DNA polymerase are coloured: 30-50 exonuclease domainin blue, thumb in green, palm in magenta and fingers in dark blue. The specific RB69 DNA polymerase N-terminalregion is coloured white. Represented residues are highlighted and coloured in the structural representation as fol-lows: residues Trp216, Phe221, Arg260, Lys279, Ser289 and Phe296 belonging to the 30-50 exonuclease domain areshown in red. Residues Tyr391 and Ala394 from the YxGG motif, located between the exonuclease andpolymerization domains are coloured in yellow. Residues Thr587 and Gln591 from the Tx2GR motif, placed at thepalm subdomain, are represented in dark blue.


to form part of the palm subdomain of thepolymerisation domain in f29 DNA polymeraseby sequence comparison with crystal data of RB69polymerase,36 Tyr226 and Gly229 (YxGG motif19),and Ile305 and Asp332 (from the TPR-1 region,specific for protein-primed DNA polymerases21,38).In addition, we have identified the followingTP-ligand residues located at the N-terminaldomain of f29 DNA polymerase: Tyr59, His61,Phe65 and Phe69 of the Exo II motif,19,23,25 Ser122and Phe128 from the (S/T)Lx2h motif,19,23,24 andArg96 and Lys114 (I.R. et al., this study).

Interestingly, with the exception of residues fromthe TPR-1 region, exclusively involved in makingcontacts with the TP, the other ones have evolved toplay a dual role, as they are performing a DNA-bind-ing function, as their counterparts in RB69, and onthe other hand, they act as TP-ligands during thefirst steps of f29 TP-DNA replication.

It has not been possible to identify unambigu-ously the f29 DNA polymerase residues homolo-gous to Arg96 and Lys114 in the rest of eukaryoticDNA polymerases. However, the alignment of thisregion in the DNA polymerases from bacterio-phages f29 and RB69, suggests that the homolo-gous to f29 DNA polymerase residue Arg96could be RB69 DNA polymerase residue Arg260.The latter forms part of a hairpin structure, pro-jected between the 30-50 exonuclease and templateclefts, constituting an effective barrier, partitioningboth active sites.37 In fact, Arg260 apparentlyblocks the template strand in the editing mode byinteracting with the penultimate nucleotide at the30 end of the primer strand,37 stabilising the meltedstructure of the DNA.

On the other hand, the homologous to f29 DNApolymerase residue Lys114 could be RB69 DNApolymerase residue Lys279, also forming a part inthe tertiary structure of the wall between the twoactive sites and, although not described as a DNAligand, the orientation towards the template cleftof its side-chain could make this residue interactwith the template strand of the DNA (see Figure9A). As depicted in Figure 9B, all the TP-bindingresidues from the N-terminal domain in which acounterpart has been found in RB69 DNA poly-merase are facing the same side of the exonucleasedomain, and opposite to the dsDNA-bindinggroove, in a similar way to the rest of TP-ligandslocated in the C-terminal domain, as residuesfrom the Tx2GR motif, placed at the palmsubdomain,30 or those belonging to the YxGGmotif (located just between the N and C-terminaldomains20). As demonstrated, such a bindinggroove is used by the TP during the initial phasesof TP-DNA replication first, being sequentiallyoccupied by the DNA as replication progresses.23

As a conclusion, one can envision the interactionbetween DNA polymerase and TP as a rather com-plex process that involves several residues of atleast three regions in the DNA polymerase thatcould wrap the TP once placed at the bindinggroove ensuring the stability of the interaction.

Materials and Methods

Nucleotides and proteins

Unlabelled nucleotides were purchased from Pharma-cia Biochemicals. [a-32P]dATP (3000 Ci/mmol) wasobtained from Amersham International plc. f29 TP waspurified as described.38 The wild-type f29 DNA poly-merase was purified from E. coli NF2690 cells harbouringplasmid pJLw2, as described.29 f29 DNA polymerasesite-directed mutants, R96A, K110A, K112A, R113A andK114A, were purified as described29 to a purity of <99%and a yield of <1.4 mg of mutant DNA polymerase/gof protein in the soluble extract. f29 ssDNA-bindingprotein (SSB) and f29 dsDNA-binding protein (p6),obtained from f29-infected B. subtilis cells, were purifiedas described.39,40

DNA templates and substrates

M13mp18 ssDNA was hybridised to the universal pri-mer in the presence of 0.2 M NaCl and 60 mM Tris–HCl(pH 7.5), and the resulting molecule was used as a pri-mer/template to analyse processive DNA polymeri-sation coupled to strand displacement by f29 DNApolymerase. Terminal protein-containing f29 DNA (f29TP-DNA) was obtained as described.41

30-50 Exonuclease assays

The incubation mixture contained, in 12.5 ml, 50 mMTris–HCl (pH 7.5), 10 mM MgCl2, 1 mM dithiothreitol,4% (v/v) glycerol and 0.1 mg/ml of bovine serum albu-min (BSA). 1.2 nM of Either 50-labelled sp1 (1.2 nM) orthe hybrid molecule sp1/sp1c þ 6 (1.2 nM) was used asssDNA and dsDNA substrate, respectively. The amountof DNA polymerase added (6 nM) was adjusted toobtain linear conditions. Samples were incubated at25 8C for the indicated times and quenched by adding3 ml of gel loading buffer. Reactions were analysed byelectrophoresis in 20% (w/v) polyacrylamide gels in thepresence of 8 M urea, and by densitometry of the autora-diographs. Total degradation was obtained by calculat-ing the number of catalytic events giving rise to eachdegradation product. From these data, the catalytic effi-ciency of each mutant derivative (indicated in Table 1)was calculated relative to wild-type f29 DNApolymerase.

Polymerase/exonuclease coupled assay

The hybrid molecule 15/21-mer contains a 6 nt long50-protruding end, and therefore, the primer strand(15-mer) can be used both as substrate for the 30-50 exo-nuclease activity and for DNA-dependent DNApolymerisation. The incubation mixture contained, in12.5 ml, 50 mM Tris–HCl (pH 7.5), 10 mM MgCl2, 1 mMdithiothreitol, 4% (v/v) glycerol, 0.1 mg/ml of BSA,0.075 ng of 50-labelled 15/21-mer, 20 ng of either wild-type or mutant f29 DNA polymerase and the indicatedconcentration of the four dNTPs. After incubation forfive minutes at 25 8C, the reaction was stopped by add-ing EDTA to 10 mM. Samples were analysed by 8 Murea-20% PAGE and autoradiography. Polymerisation or30-50 exonuclease are detected as an increase or decrease,respectively, in the size (15-mer) of the 50-labelled primer.


DNA gel retardation assays

The 50-labelled oligonucleotides sp1/sp1c þ 6 wereused as dsDNA substrate to analyse the interaction ofeither the wild-type or mutant f29 DNA polymerase.The incubation mixture contained, in a final volume of20 ml, 12 mM Tris–HCl (pH 7.5), 1 mM EDTA, 20 mMammonium sulphate, 0.1 mg/ml of BSA, 1.2 nM sp1/sp1c þ 6 and 1.25 ng of either wild-type or mutant f29DNA polymerase, in the absence or presence of 10 mMMgCl2. After incubation for five minutes at 4 8C, thesamples were subjected to electrophoresis in 4% (w/v)polyacrylamide (acrylamide to bis-acrylamide, 4:1, w/w)) gels, containing 12 mM Tris–acetate (pH 7.5) and1 mM EDTA, and run at 4 8C in the same buffer at 8 V/cm, essentially as described.42 After autoradiography,f29 DNA polymerase–dsDNA complexes were detectedas a mobility shift (retardation) in the migrating positionof the labelled DNA. Quantification of the DNA-bindingcapacity of the wild-type f29 DNA polymerase and itsmutant derivatives was carried out by densitometry ofthe retarded band.

Replication assay (protein-primed initiation pluselongation) using f29 TP-DNA as template

The incubation mixture contained, in 25 ml, 50 mMTris–HCl (pH 7.5), 10 mM MgCl2, 20 mM ammoniumsulphate, 1 mM dithiothreitol, 4% (v/v) glycerol,0.1 mg/ml of BSA, 20 mM each dCTP, dGTP, dTTP and[a-32P] dATP (2 mCi), 0.5 mg of f29 TP-DNA, 10 ng ofpurified TP and 20 ng of either wild-type or mutant f29DNA polymerase. After incubation for the indicatedtimes at 30 8C, the reaction was stopped by adding10 mM EDTA/0.1% SDS, and the samples were filteredthrough Sephadex G-50 spin columns. Relative activitywas calculated from the Cerenkov radiation correspond-ing to the excluded volume. For size analysis, thelabelled DNA was denatured by treatment with 0.7 MNaOH and subjected to electrophoresis in alkaline 0.7%agarose gels as described.43 After electrophoresis, theposition of unit-length f29 DNA (19,285 bp) wasdetected by ethidium bromide staining, and then the gelswere dried and autoradiographed. For analysis of the tran-sition products, the indicated dNTPs (20 mM), 125 ng ofpurified TP and 100 ng of either wild-type or mutant f29DNA polymerase were used. Samples were subjected toSDS-12% PAGE gel (360 mm £ 280 mm £ 0.5 mm) elec-trophoresis to obtain enough resolution to distinguishTP bound to the first elongation products.

f29 TP-DNA amplification assay

The incubation mixture contained, in 25 ml, 50 mMTris–HCl (pH 7.5), 10 mM MgCl2, 20 mM ammoniumsulphate, 1 mM dithiothreitol, 4% (v/v) glycerol,0.1 mg/ml of BSA, 80 mM each dCTP, dGTP, dTTP, and[a-32P] dATP (1 mCi), 5 ng of f29 TP-DNA, 5 ng of eitherwild-type or mutant f29 DNA polymerase, 5 ng of TP,10 mg of f29 SSB and 10 mg of f29 p6. After incubationfor 90 minutes at 30 8C, the samples were processed andthe amplified DNA was analysed by electrophoresis inalkaline agarose gels, as described.31 After electro-phoresis, the position of unit-length f29 DNA wasdetected by ethidium bromide staining.

Replication of primed M13 DNA

The incubation mixture contained, in 25 ml, 50 mMTris–HCl (pH 7.5), 10 mM MgCl2, 1 mM dithithreitol,4% (v/v) glycerol, 0.1 mg/ml of BSA, 40 mM each dCTP,dGTP, dTTP and [a-32P] dATP (1 mCi), 0.25 mg of primedM13mp8 ssDNA, and 100 ng of either wild-type ormutant f29 DNA polymerase. After incubation for theindicated times at 30 8C, the samples were processedand the synthesised DNA was quantified and analysedas described above for the f29 TP-DNA amplificationassay. After electrophoresis, unit-length M13mp8ssDNA was detected by ethidium bromide staining, andthen, gels were dried and autoradiographed.

TP-dAMP formation (protein-primed initiation assay)

The incubation mixture contained, in 25 ml, 50 mMTris–HCl (pH 7.5), 10 mM MgCl2, 20 mM ammoniumsulphate, 1 mM dithiothreitol, 4% (v/v) glycerol,0.1 mg/ml of BSA, 0.2 mM dATP [a-32dATP] (2 mCi),0.5 mg of f29 TP-DNA, 15 ng of purified TP and 30 ngof either wild-type or mutant f29 DNA polymerase,and incubated for 1.5 minutes at 30 8C. For template-independent initiation assay, f29 TP-DNA was omitted,15 ng of TP and 30 ng of either wild-type or mutant f29DNA polymerase were added, 1 mM MnCl2 was usedinstead of MgCl2, and the incubation was for 180 min-utes at 30 8C. The reactions were stopped by adding10 mM EDTA and 0.1% SDS, filtered through SephadexG-50 spin columns, and further analysed by SDS-PAGEas described.44 Quantification was done by densitometricanalysis of the labelled band corresponding to theTP-dAMP complex, detected by autoradiography.

Interference assay for TP binding

Reactions were carried out as described for the tem-plate-independent initiation assay, using a limitingamount of TP. f29 DNA polymerase mutant D249E (cat-alytically inactive but displaying a normal interactionwith the TP) was used as a positive control for the inter-ference assay, as described.33 The amounts of proteinsused were as follows: 8 ng of TP, 20 ng of either wild-type or mutant K110A, K112A, R113A and K114A f29DNA polymerases, and increasing amounts (20, 40, 80and 160 ng) of mutant DNA polymerase D249E. Inaddition, 20 ng of wild-type DNA polymerase was incu-bated with increasing amounts (20, 40, 80 and 160 ng) ofmutant R96A DNA polymerase. In all cases, the incu-bation was for four hours at 4 8C. After incubation, reac-tions were stopped and analysed as indicated for theprotein-primed initiation assay.

Analysis of the interaction between TP and DNApolymerase mutants by glycerolgradient centrifugation

The incubation mixture contained, in 200 ml, 50 mMTris–HCl (pH 7.5), 1 mM dithithreitol, 20 mM ammo-nium sulphate, 3 mg of TP and 6 mg of either wild-typeor mutant DNA polymerase. After incubation for 30minutes at 4 8C, samples were loaded on top of a con-tinuous 15%–30% (v/v) glycerol gradient (4 ml) in thepresence of 50 mM Tris–HCl (pH 7.5), 20 mM ammo-nium sulphate, 0.4 M NaCl, 1 mM EDTA and 7 mM 2-mercaptoethanol, and centrifuged at 4 8C for 24 hours at58,000 rpm in a Beckman TST 60.4 rotor. Gradients were


fractionated and subjected to SDS-12% PAGE. Gels werestained after electrophoresis with silver staining forproteins.

Acknowledgements

This investigation was aided by research grant5R01 GM27242-24 from the National Institutesof Health, by grant BMC2002-03818 from theDireccion General de Investigacion Cientıfica yTecnica, and by an institutional grant from Funda-cion Ramon Areces to the Centro de BiologıaMolecular “Severo Ochoa”. I.R. was a pre-doctoralfellow of the Consejo Superior de InvestigacionesCientıficas.

References

1. Kornberg, A. & Baker, T. (1992). DNA Replication, 2ndedit., Freeman, San Francisco.

2. Salas, M. (1999). Mechanisms of initiation of linearDNA replication in prokaryotes. In Genetic Engi-neering (Setlow, J. K., ed.), vol. 21, pp. 159–171,Kluwer Academic/Plenum Press, New York.

3. Salas, M. (1991). Protein-priming of DNA replication.Annu. Rev. Biochem. 60, 39–71.

4. Salas, M., Miller, J. T., Leis, J, & DePamphilis, M. L.(1996). Mechanisms for priming DNA synthesis. InDNA Replication in Eukaryotic Cells (DePamphilis,M. L., ed.), pp. 131–176, Cold Spring Harbor Labora-tory Press, Cold Spring Harbor, NY.

5. Serrano, M., Gutierrez, C., Freire, R., Bravo, A. &Hermoso, J. M. (1994). Phase f29 protein p6: a viralhistone-like protein. Biochimie, 76, 981–991.

6. Hermoso, J. M., Mendez, E., Soriano, F. & Salas, M.(1985). Location of the serine residue involved in thelinkage between the terminal protein and the DNAof f29. Nucl. Acids Res. 13, 7715–7728.

7. Blanco, L. & Salas, M. (1984). Characterization andpurification of a phage f29-encoded DNA polymer-ase required for the initiation of replication. Proc.Natl Acad. Sci. USA, 81, 5325–5329.

8. Mendez, J., Blanco, L., Esteban, J. A., Bernad, A. &Salas, M. (1992). Initiation of f29 DNA replicationoccurs at the second 30 nucleotide of the linear tem-plate: a sliding-back mechanism for protein-primedDNA replication. Proc. Natl Acad. Sci. USA, 89,9579–9583.

9. Illana, B., Blanco, L. & Salas, M. (1996). Functionalcharacterization of the genes coding for the terminalprotein and DNA polymerase from bacteriophageGA-1. Evidence for a sliding-back mechanism duringprotein-primed GA-1 DNA replication. J. Mol. Biol.264, 453–464.

10. Caldentey, J., Blanco, L., Bamford, D. H. & Salas, M.(1993). In vitro replication of bacteriophage PRD1DNA. Characterization of the protein-primedinitiation site. Nucl. Acids Res. 21, 3725–3730.

11. Martın, A. C., Blanco, L., Garcıa, P., Salas, M. &Mendez, J. (1996). In vitro protein primed initiationof pneumococcal phage Cp-1 DNA replication occursat the third 30 nucleotide of the linear template: astepwise sliding-back mechanism. J. Mol. Biol. 260,369–377.

12. King, A. J. & Van der Vliet, P. C. (1994). A precursorterminal protein trinucleotide intermediate duringinitiation of adenovirus DNA-replication. Regener-ation of molecular ends in vitro by a jumping backmechanism. EMBO J. 13, 5786–5792.

13. Mendez, J., Blanco, L. & Salas, M. (1997). Protein-primed DNA replication: a transition between twomodes of priming by a unique DNA polymerase.EMBO J. 16, 2519–2527.

14. Blanco, L., Bernad, A., Lazaro, J. M., Martın, G.,Garmendia, C. & Salas, M. (1989). Highly efficientDNA synthesis by the phage f29 DNA polymerase.Symmetrical mode of DNA replication. J. Biol. Chem.264, 8935–8940.

15. Van der Vliet, P. C. (1996). Roles of transcriptionfactors in DNA replication. In DNA Replication inEukaryotic Cells (DePamphilis, M. L., ed.), pp.87–118, Cold Spring Harbor Laboratory Press, ColdSpring Harbor, NY.

16. Blanco, L. & Salas, M. (1995). Mutational analysisof bacteriophage f29 DNA polymerase. MethodsEnzymol. 262, 283–294.

17. Blanco, L. & Salas, M. (1996). Relating structure tofunction in f29 DNA polymerase. J. Biol. Chem. 271,8509–8512.

18. Bernad, A., Blanco, L., Lazaro, J. M., Martın, G. &Salas, M. (1989). A conserved 30-50 exonuclease activesite in prokaryotic and eukaryotic DNA polymer-ases. Cell, 59, 219–228.

19. de Vega, M., Blanco, L. & Salas, M. (1998). f29 DNApolymerase residue Ser122, a single-stranded DNAligand for 30-50 exonucleolysis, is required to interactwith the terminal protein. J. Biol. Chem. 273,28966–28977.

20. Truniger, V., Blanco, L. & Salas, M. (1999). Role of the“YxGG/A” motif of f29 DNA polymerase in pro-tein-primed replication. J. Mol. Biol. 286, 57–69.

21. Dufour, E., Mendez, J., Lazaro, J. M., de Vega, M.,Blanco, L. & Salas, M. (2000). An aspartic acid resi-due in TRP-1, a specific region of protein-primingDNA polymerases, is required for the functionalinteraction with primer terminal protein. J. Mol. Biol.304, 289–300.

22. Truniger, V., Lazaro, J. M., Salas, M. & Blanco, L.(1998). f29 DNA polymerase requires the N-terminaldomain to bind terminal protein and DNA primersubstrates. J. Mol. Biol. 278, 741–755.

23. de Vega, M., Lazaro, J. M., Salas, M. & Blanco, L.(1998). Mutational analysis of f29 DNA polymeraseresidues acting as ssDNA ligands for 30-50 exo-nucleolysis. J. Mol. Biol. 279, 807–822.

24. Rodrıguez, I., Lazaro, J. M., Salas, M. & de Vega, M.(2003). f29 DNA polymerase residue Phe128 of thehighly conserved (S/T)Lx2h motif is required for astable and functional interaction with the terminalprotein. J. Mol. Biol. 325, 85–97.

25. Eisenbrandt, R., Lazaro, J. M., Salas, M. & de Vega,M. (2002). f29 DNA polymerase residues Tyr59,His61 and Phe69 of the highly conserved Exo IImotif are essential for interaction with the terminalprotein. Nucl. Acids Res. 30, 1379–1386.

26. Chou, P. Y. & Fasman, G. D. (1978). Prediction of thesecondary structures of proteins from their aminoacid sequence. Advan. Enzymol. 47, 45–48.

27. Garnier, J., Osguthorpe, D. J. & Robson, B. (1978).Analysis of the accuracy and implications of simplemethods for predicting the secondary structure ofglobular proteins. J. Mol. Biol. 120, 97–120.

28. Bordo, D. & Argos, P. (1991). Suggestions for “safe”


residue substitutions in sitedirected mutagenesis.J. Biol. Chem. 217, 721–729.

29. Lazaro, J. M., Blanco, L. & Salas, M. (1995). Purifi-cation of f29 DNA polymerase. Methods Enzymol.262, 42–49.

30. Mendez, J., Blanco, L., Lazaro, J. M. & Salas, M.(1994). Primer-terminus stabilization at the f29DNA polymerase active site. Mutational analysisof conserved motif Tx2GR. J. Biol. Chem. 269,30030–30038.

31. Blanco, L., Lazaro, J. M., de Vega, M., Bonnin, A. &Salas, M. (1994). Terminal protein-primed DNAamplification. Proc. Natl Acad. Sci. USA, 91,12198–12202.

32. Blanco, L., Bernad, A., Esteban, J. A. & Salas, M.(1992). DNA-independent deoxynucleotidylation ofthe f29 terminal protein by the f29 DNA polymer-ase. J. Biol. Chem. 267, 1225–1230.

33. Blasco, M. A., Lazaro, J. M., Blanco, L. & Salas, M.(1993). f29 DNA polymerase active site. ResidueAsp249 of conserved amino acid motif dx2SLYP iscritical for synthetic activities. J. Biol. Chem. 268,24106–24113.

34. Serna-Rico, A., Illana, B., Salas, M. & Meijer, W. J. J.(2000). The putative coiled coil domain of the f29terminal protein is a major determinant involved inrecognition of the origin of replication. J. Biol. Chem.275, 40529–40538.

35. Dufour, E., Rodrıguez, I., Lazaro, J. M., de Vega, M. &Salas, M. (2003). A conserved insertion in protein-primed DNA polymerases is involved in primer ter-minus stabilisation. J. Mol. Biol. 331, 781–794.

36. Wang, J., Sattar, A. K., Wang, C. C., Karam, J. D.,Konigsberg, W. H. & Steitz, T. A. (1997). Crystalstructure of a Pol a family replication DNA polymer-ase from bacteriophage RB69. Cell, 89, 1087–1099.

37. Shamoo, Y. & Steitz, T. A. (1999). Building a repli-some from interacting pieces: sliding clamp com-

plexed to a peptide from DNA polymerase and apolymerase editing complex. Cell, 99, 155–166.

38. Zaballos, A., Lazaro, J. M., Mendez, E., Mellado, R. P.& Salas, M. (1989). Effects of internal deletions on thepriming activity of the phage f29 terminal protein.Gene, 83, 187–195.

39. Martın, G., Lazaro, J. M., Mendez, E. & Salas, M.(1989). Characterization of phage f29 protein p5 asa single-stranded DNA binding protein. Function inf29 DNA–protein p3 replication. Nucl. Acids Res.17, 3663–3672.

40. Pastrana, R., Lazaro, J. M., Blanco, L., Garcıa, J. A.,Mendez, E. & Salas, M. (1985). Overproduction andpurification of protein p6 of Bacillus subtilis phagef29: role in the initiation of DNA replication. Nucl.Acids Res. 13, 3083–3100.

41. Inciarte, M. R., Lazaro, J. M., Salas, M. & Vinuela, E.(1976). Physical map of bacteriophage f29 DNA.Virology, 74, 314–323.

42. Carthew, R. W., Chodosch, L. A. & Sharp, P. A.(1985). A RNA polymerase II transcription factorbinds to an upstream element in the adenovirusmajor late promoter. Cell, 43, 439–448.

43. McDonnell, M. W., Simon, M. N. & Studier, F. W.(1994). Analysis of restriction fragments of T7 DNAand determination of molecular weights by electro-phoresis in neutral and alkaline agarose gels.J. Membr. Biol. 110, 119–146.

44. Penalva, M. A. & Salas, M. (1982). Initiation of phagef29 DNA replication in vitro: formation of a covalentcomplex between the terminal protein, p3, and50-dAMP. Proc. Natl Acad. Sci. USA, 79, 5522–5526.

45. Braithwaite, D. K. & Ito, J. (1993). Compilation,alignment, and phylogenetic relationships of DNApolymerases. Nucl. Acids Res. 21, 787–802.

46. Gonzalez-Huici, V. (2001) Reconocimiento del origende replicacion del DNA del bacteriofago f29. PhDthesis, Autonomous University of Madrid.

Edited by J. Karn

(Received 19 November 2003; received in revised form 4 February 2004; accepted 6 February 2004)


Documents

φ29 DNA Polymerase–Terminal Protein Interaction. Involvement of Residues Specifically Conserved Among Protein-primed DNA Polymerases