Hexameric RSF1010 helicase RepA - Nucleic Acids Research

Hexameric RSF1010 helicase RepA: the structuraland functional importance of single amino acidresiduesGuÈnter Ziegelin1,2, Timo Niedenzu2, Rudi Lurz1, Wolfram Saenger2 and Erich Lanka1,*

1Max-Planck-Institut fuÈr Molekulare Genetik, Ihnestrasse 73, Dahlem, D-14195 Berlin, Germany and 2Institut fuÈrKristallographie, Freie UniversitaÈ t Berlin, D-14195 Berlin, Germany

Received July 8, 2003; Revised August 15, 2003; Accepted August 26, 2003

ABSTRACT

In the known monoclinic crystals the 3-dimensionalstructure of the hexameric, replicative helicaseRepA encoded by plasmid RSF1010 shows 6-foldrotational symmetry. In contrast, in the cubic crystalform at 2.55 AÊ resolution described here RepA has3-fold symmetry and consists of a trimer of dimers.To study structure±function relationships, a seriesof repA deletion mutants and mutations yieldingsingle amino acid exchanges were constructed andthe respective gene products were analyzed in vivoand in vitro. Hexamerization of RepA occurs via theN-terminus and is required for NTP hydrolysis. TheC-terminus is essential both for the interaction withthe replication machinery and for the helicaseactivity. Functional analyses of RepA variants withsingle amino acid exchanges con®rmed most of thepredictions that were based on the published 3-dimensional structure. Of the ®ve motifs conservedin family 4 helicases, all residues conserved inRepA and T7 gp4 helicases participate in DNAunwinding. Residues K42, E76, D77, D139 and H178,proposed to play key roles in catalyzing the hydroly-sis of NTPs, are essential for RepA activity. ResidueH178 of motif H3 couples nucleotide consumptionto DNA strand separation.

INTRODUCTION

DNA helicases are ubiquitous motor proteins that utilize theenergy obtained by the hydrolysis of nucleoside triphosphates(NTPs) to unwind double-stranded nucleic acids. The proteinsplay key roles in a variety of biological processes like DNAreplication, recombination, repair and transcription. Theenzymes have received increasing attention since it becameknown that at least six hereditary diseases, like xerodermapigmentosum and Cockayne syndrome, are caused by variantsof DNA helicases or of putative helicases (1). There are twogroups of structurally known DNA helicases, one forminghexameric rings that operate at the DNA replication fork toseparate both strands of the duplex DNA ahead of the DNA

polymerase complexes, whereas the other group includesmonomeric or dimeric enzymes.

Plasmid RSF1010 encodes its own replication initiationsystem, making its replication independent of the hostinitiation machinery. Proteins RepC and RepB are requiredfor origin recognition and primer synthesis, respectively,whereas RepA is the hexameric replicative helicase essentialfor RSF1010 replication (2,3). RepA has 5¢®3¢ polarity andrequires a forked DNA substrate for optimal activity. Amongribonucleoside triphosphates ATP is the preferred lowmolecular weight substrate. The pH optimum of the helicaseactivity is at pH 5.5±6, which corresponds with optimalbinding to single-stranded (ss)DNA (3). In contrast to otherhexameric helicases RepA assembles into stable hexamers inthe absence of any nucleotide or metal cofactor, as demon-strated by chemical cross-linking, gel ®ltration and electronmicroscopy (3).

In monoclinic RepA crystals grown at pH 6.0, dimers ofhexamers in head-to-head orientation are observed (5). Bothimage reconstruction of electron microscopy data and the highresolution 3-dimensional crystal structures (2.4 and 1.95 AÊ )revealed a 6-fold rotational symmetry (3,5,6). The ®vehelicase motifs H1, H1a and H2±H4 conserved in DnaB-likeenzymes (7) are present in RepA and spatially clusteredaround the NTP binding pocket. The proposed catalytic sitefor ATP hydrolysis is located at the interface of neighboringmonomers, with the adenine base being sandwiched betweenR85 of the NTP binding monomer and Y242 of the adjacentsubunit.

Besides RepA, the 3-dimensional structures of several otherhelicases have been determined, e.g. the monomeric homologsBacillus stearothermophilus PcrA and Escherichia coli Rephelicases, (8,9), the T7 helicase domain [amino acids 272±566(10) and 241±566 (11)] and the 130 amino acid RNA bindingdomain of the E.coli Rho RNA helicase (12). The latter twoenzymes are hexamers. A low resolution structure for thehexameric prototype E.coli DnaB has been determined byelectron microscopy and 3-dimensional image reconstruction(13), but high resolution data of hexameric replicative DnaB-type helicases are still not available except for a short stretchof an N-terminal domain (14).

For translocation of helicases along DNA a variety ofmodels have been proposed (15; reviewed in 16). For themonomeric or dimeric helicases the `inchworm' model, the

*To whom correspondence should be addressed. Tel: +49 30 8413 1696; Fax: +49 30 8413 1130; Email: [email protected]

Nucleic Acids Research, 2003, Vol. 31, No. 20 5917±5929DOI: 10.1093/nar/gkg790

Nucleic Acids Research, Vol. 31 No. 20 ã Oxford University Press 2003; all rights reserved

Dow

nloaded from https://academ

ic.oup.com/nar/article/31/20/5917/1039484 by guest on 26 D

ecember 2021

principles of which were proposed by Gefter in 1979 (17), hasemerged as the most plausible in the light of recent work(reviewed in 18). Although the information on hexamerichelicases is lagging behind, a modi®ed version of this modelmight be applicable to these ring-shaped DNA unwindingenzymes (5,10,11).

Here we report on the high resolution structure of RepAcrystallized at pH 7.5 in a cubic space group. Predictions madeon the function of de®ned residues based on this 3-dimensional structure were evaluated experimentally. Thevariant proteins used for this purpose were generated by singleamino acid exchange and deletion mutagenesis. The in¯uenceof speci®c residues proposed to play key roles in ATPhydrolysis and DNA unwinding were analyzed in vivo andin vitro. We discuss the implications for enzyme function ofconformational differences of residues that probably bind theATP between adjacent monomers.

MATERIALS AND METHODS

Bacterial strains and plasmids

Escherichia coli SCS1 [recA1, endA1, gyrA96, thi-1,hsdR17(rK± mK+), supE44, relA1; Stratagene] was used ashost for repA overexpression plasmids, for construction of theRSF1010KDrepA plasmid and for complementation experi-ments. Recombinant plasmids used in this study are given inTable 1. Media were as described previously (19). Whenappropriate, antibiotics were added at the following concen-trations: 100 mg/ml ampicillin (sodium salt) or 30 mg/mlkanamycin sulfate.

Construction of plasmids

For DNA manipulation standard techniques were used (20). InRSF1010, repA is ¯anked by repB¢ and repC, encoding theprimase and the origin binding protein, respectively. repA wasdeleted leaving a few 5¢ and 3¢ codons in frame to avoid anypolar effects. To construct RSF1010KDrepA, the kanamycin-resistant derivative RSF1010K (Table 1) was used. The2924 bp A¯III±BstEII fragment carrying most of repB¢, repAand the 5¢ portion of repC was replaced by two fragmentsgenerated by PCR using RSF1010K as template to restorerepB¢ and repC. An A¯II site that served to join the PCRfragments was created by changing the leucine codon repAL259 from CTC to CTT. Small 5¢ and 3¢ repA remnants wereconnected in frame encoding a 29 amino acid peptide thatconsists of RepA residues A1±L9 and L259±A278. Thenucleotide sequence of this arrangement was veri®ed bysequencing.

To construct pGZ18-20D1±5, pGZ18-20D1±7 and pGZ18-20D1±9, which lack RepA amino acid residues 1±5, 1±7 and1±9, respectively, NdeI±HindIII fragments of 970, 964 and958 bp were generated by PCR using pGZ18-20 as templateand inserted into pMS470D8 prepared with NdeI and HindIII.To construct pGZ18-20D276±278, pGZ18-20D274±278 andpGZ18-20D269±278, which lack the very C-terminal RepAamino acid residues 276±278, 274±278 and 269±278, respect-ively, EcoRI±HindIII fragments of 901, 895 and 880 bp weregenerated by PCR using pGZ18-20 as template and insertedinto pMS119EH prepared with EcoRI and HindIII. Thenucleotide sequence of each ampli®ed DNA fragment was

determined and veri®ed according to the published sequence(GenBank accession no. M28829; 21).

Point mutations and a single codon deletion (RepA DF12)were introduced into repA directly using a PCR-based method(22) with pGZ18-20 as template. The primers used weredesigned by changing as few bases as possible (Table 2).Following mutagenesis the nucleotide sequence of each repAallele was determined. Since formylmethionine is cleaved offpost-transcriptionally, alanine occupies RepA amino acidposition 1 (21). Therefore, our numbering differs from thatused by Niedenzu et al. (5) and Xu et al. (6), who assignedmethionine as position 1.

In vivo complementation and interference assays

Complementation studies were done by transforming SCS1harboring a pGZ18-20 derivative that carried a repA mutantallele with RSF1010DrepA. An overnight culture was dilutedwith medium containing ampicillin (see above) and IPTG(5 mM) to an A600 of 0.1 and grown at 37°C with shaking. Atan A600 of 0.4, the cells were harvested and made competentusing the CaCl2 method (23). Aliquots of 20 ml of culturewere centrifuged and resuspended to result in 1 ml ofcompetent cells. To 100 ml of these cells was added 0.1 mgRSF1010KDrepA (Table 1). Following incubation for 3 h onice and 2 min at 40°C, 1 ml of medium was added. Afterincubation for 1 h at 37°C, the cells were diluted appropri-ately, plated on selective medium containing 5 mM IPTG andincubated overnight at 37°C. The colonies obtained werecounted. Interference tests were done accordingly, except thatRSF1010K was used instead of RSF1010KDrepA.

Overexpression and puri®cation of proteins

RepA mutant proteins were overproduced and puri®ed usingthe procedure described previously (3) with the followingmodi®cation. Chromatography on ATP±agarose as the lastpuri®cation step was replaced by phenyl Sepharose. A columnof phenyl Sepharose was equilibrated with 20 mM Tris±HCl(pH 8.0), 20 mM NaCl and 1 mM DTT (buffer A) containing1 M ammonium sulfate. The DEAE Sephacel RepA fraction(3) was dialyzed against buffer A, adjusted to 1 M ammoniumsulfate and loaded onto the column. The column was washedwith buffer A containing 1 M ammonium sulfate and then withbuffer A. Proteins were eluted with a gradient of 0±70%ethylene glycol in buffer A. RepA eluted at ~60% ethyleneglycol.

Helicase and ATPase assays

A forked helicase substrate based on that of Crute et al. (24)was used. The unpaired 3¢-portion of the 53mer oligonucleo-tide comprised 22 nt, whereas 31 nt were paired to viralM13mp18 DNA. The 5¢-end was 32P-labeled. RepA-catalyzedunwinding of double-stranded (ds)DNA was assayed essen-tially as described (3) at 30°C for 15 min in 20 ml of buffer B[40 mM MES/NaOH pH 5.6, 10 mM MgCl2, 1 mM DTT and50 mg/ml bovine serum albumin (BSA)] containing 1 mMATP. The products were separated by electrophoresis on 10%polyacrylamide gels. The radioactivity of the substrate and thedisplaced oligonucleotide was visualized using phosphorstorage technology and quanti®ed by the use of ImageQuantsoftware version 5.0 (Amersham/Pharmacia).

5918 Nucleic Acids Research, 2003, Vol. 31, No. 20

Dow



ecember 2021

RepA-catalyzed ATP hydrolysis reactions were run for 15min at 30°C as described (3) in 20 ml of buffer B containing0.5 mM ATP, 100 nCi [g-32P]ATP and 1 mg viral M13mp18DNA. Products were separated by thin layer chromatographyand quanti®ed as described above.

Crystallization and X-ray data collection

For crystallization, RSF1010 RepA was puri®ed using amodi®cation of the protocol described in RoÈleke et al. (4). Toobtain crystals of the highest quality, the purity of the ®nal

product was further improved by inserting an additional anionexchange chromatography (Q-Sepharose) before the ®nal gel®ltration. For crystallization using the sitting drop vapordiffusion method equal volumes of 5 ml of RepA stocksolution (26 mg/ml RepA, 10 mM Tris±HCl pH 8.0, 150 mMNaCl, 0.1 mM EDTA) and reservoir solution (100 mMTris±HCl pH 7.5, 26% PEG400, 100 mM MgSO4) were mixedand equilibrated against reservoir solution. Diffraction datafrom a native crystal were collected at 100 K using a mar345Image Plate detector mounted on a Nonius FR571 rotating

Table 1. Plasmids and bacteriophage used in this study

Plasmid/bacteriophage Description Reference or origin

pMS119EH Cloning vector; pMB1 replicon, Ptac/lacI; Apr (40)pMS470D8 Cloning vector; pMB1 replicon, Ptac/lacI; T7 gene 10 Shine±Dalgarno sequence; Apr

RSF1010 IncQ replicon, Smr, Sur (41)RSF1010K RSF1010 D[HpaI±EcoRV, RSF1010 nt 1±1704] W[Tn903 StuI±StuI, Tn903 nt 878±2217]; Sms, Snr, Kmr E. ScherzingerRSF1010KDrepA RSF1010K D[RSF1010 nt 5920±6696] Kmr, replication de®cient This workmOT18-20 M13mp18 W[RSF10101 MaeIII±MaeIII ®lled in, repA; RSF1010 nt 5841±6840] E. ScherzingerpGZ18-20 pMS119EH D[EcoRI±HindIII] W[mOT-18-20 EcoRI±HindIII; RSF1010 nt 5841±6840] This workpGZ18-20X000Y pGZ18-20 carrying the RepA amino acid exchange at the position indicated This workpGZ18-20D1±5 pMS470D8 D[NdeI±HindIII] W[PCR pGZ18-20 NdeI±HindIII; RSF1010 nt 5908±6840] This workpGZ18-20D1±7 pMS470D8 D[NdeI-HindIII] W[PCR pGZ18-20 NdeI±HindIII; RSF1010 nt 5914±6840] This workpGZ18-20D1±9 pMS470D8 D[NdeI-HindIII] W[PCR pGZ18-20 NdeI±HindIII; RSF1010 nt 5920±6840] This workpGZ18-20DF12 pGZ18-20 lacking repA codon F12 This workpGZ18-20D276±278 pMS119EH D[EcoRI±HindIII] W[PCR pGZ18-20 EcoRI±HindIII, RSF1010 nt 5841±6717] This workpGZ18-20D274±278 pMS119EH D[EcoRI-HindIII] W[ PCR pGZ18-20 EcoRI±HindIII, RSF1010 nt 5841±6711] This workpGZ18-20D269±278 pMS119EH D[EcoRI-HindIII] W[ PCR pGZ18-20 EcoRI±HindIII, RSF1010 nt 5841±6687] This workM13mp18 phage M13 cloning vector (42)

Sur, Smr, sulfonamide- and streptomycin-resistant; Sms, streptomycin-sensitive.

Table 2. Oligonucleotides used for the molecular cloning of RSF1010KDrepA and repA mutagenesis

Plasmid Oligonucleotide sequence

pGZ18-20D1±5 atatttgacatatgATCAATATTCTGGAGGCGTTCGCAGCAGpGZ18-20D1±7 atatttgacatatgATTCTGGAGGCGTTCGCAGCAGCGCCpGZ18-20D1±9 atatttgacatatgGAGGCGTTCGCAGCAGCGCCGCCACpGZ18-20DF12 CAATATTCTGGAGGCGGCAGCAGCGCCGpGZ18-20D276±278 tatataagcttcTTAACGGGGCACCCCCTTGCTCpGZ18-20D274±278 tatataagcttcTTACACCCCCTTGCTCTTGCGCTGpGZ18-20D269±278 tatataagcttcTTAGCGCTGCCTCTCCAGCACGpGZ18-20K42A GGTGGTGCCGGTGCATCCATGCTGGCCpGZ18-20E76A CTGCCCGCCGCAGACCCGCCCpGZ18-20D77A CCCGCCGAAGCCCCGCCCACCpGZ18-20R85A CGCCATTCATCACGCCCTGCACGCCCTTGpGZ18-20D139A CCTGATGGTGCTGGCCACGCTGCGCCGGpGZ18-20E147Q GCCGGTTCCACATCCAGGAAGAAAACGCCAGpGZ18-20E148Q GGTTCCACATCGAGCAGGAAAACGCCAGCGGpGZ18-20E149Q GTTCCACATCGAGGAACAGAACGCCAGCGGCpGZ18-20H178A CGTGTTCCTGCACGCTGCCAGCAAGGGCGpGZ18-20Q191E GGCGCAGGCGACGAACAGCAGGCCAGCpGZ18-20Q192E GCAGGCGACCAGGAACAGGCCAGCCCGpGZ18-20Q193E CAGGCGACCAGCAGGAAGCCAGCCCGGGpGZ18-20R206A CTGGTCGATAACATCGCCTGGCAGTCCTACCTGpGZ18-20Y242A GTGAGCAAGGCCAACGCTGGCGCACCGTTCpGZ18-20Y242T GTGAGCAAGGCCAACACTGGCGCACCGTTCpGZ18-20R253A GATCGGTGGTTCAGGGCGCATGACGGCGGPCR-RSF1010K-1 GAACACGCCATGGCTCAAGCGGPCR-RSF1010K-2 atatatCTtAAGCAGAATATTGATAGGCTTATGGGTAGCCATTGPCR-RSF1010K-3 atatatCTtAAGGCCCGCCGTGCTGGAGAGPCR-RSF1010K-4 CCTTGGGCCGGGTGATGTCG

For the N-terminal and C-terminal deletion derivatives, only the relevant primer is given. For the pointmutations, the sequence of the sense strand is shown and the codon changed is printed in bold. Non-repAsequences are in lower case letters. Hexanucleotides printed in bold represent restriction sites.

Nucleic Acids Research, 2003, Vol. 31, No. 20 5919

Dow



ecember 2021

anode X-ray generator equipped with Osmic focusing mirrors.A mercury derivative was prepared by soaking crystals inreservoir solution with 1 mM o-chloromercurinitrophenol asdescribed (5). MAD data were collected up to 3.0 AÊ resolutionat beamline BW7A, EMBL outstation, DESY/Hamburg, usinga mar CCD detector.

Phasing and re®nement

All diffraction data were processed with DENZO/SCALEPACK (25). Programs from the CCP4 suite (26)were used for subsequent calculations and model re®nement.The single mercury site in the derivative was located byinspection of an anomalous difference Patterson map. A ®rstelectron density map was calculated, the starting model takenfrom the previously published RepA crystal structure (5) wasplaced manually into the map and the orientation of theindividual monomers was ®tted. Further model building usingthe program O (27) and re®nement employed the higherresolution native data set (Table 3). The coordinates of thecubic RepA form have been deposited in the RCSB ProteinData Bank under the accession code 1olo.

RESULTS

A new crystal form of RepA

The improvement in the puri®cation scheme for RepA forcrystallization was visible in the PAGE analysis, where evenin heavily overloaded lanes nothing but RepA was detectable(not shown). In addition to the described monoclinic RepAcrystals (space group P21) grown at pH 6.0 (4,5), previouslyunobserved crystals of perfect octahedral shape appeared inone of the crystallization screen trials. Following optimizationof crystallization conditions, crystals suitable for X-raydiffraction data collection were grown as described inMaterials and Methods. The crystals belong to the cubicspace group P4332, with two of the RepA subunits in theasymmetric unit (Matthews coef®cient 4.8 AÊ 3/Da, 74.4%solvent). The structure was determined by MAD using amercury derivative (see Materials and Methods) and re®nedagainst a native data set at 2.55 AÊ resolution. The ®nal modelconsists of 499 residues (249 and 250 for monomers A and B,respectively), 361 water molecules in the solvation shells ofthe protomers and 60 additional waters. A sulfate ion wasbound to the p-loop of each monomer as described (6).

In contrast to the published RepA crystal structures (5,6)comprising pairs of hexameric rings with 6-fold rotationsymmetry in head-to-head orientation in the asymmetric unit,

the RepA oligomers in the cubic crystal system occur as singlehexameric rings. Since the crystallographic C3 axis passesright through the molecule, the hexamers have 3-fold rotationsymmetry and the crystallographic asymmetric unit containsonly two monomers. Although a few more of the C-terminalamino acids (A262±S270) could be modeled here that werenot seen in the monoclinic RepA (5), the loop formed byresidues 180±199 containing half of the helicase motif H4 stillhad a poorly de®ned electron density and could not bemodeled (6). A view from the top of the hexamer is shown inFigure 1a. In cubic RepA crystals, two different conformationsof the monomer±monomer interface are found that featuredifferent orientations of the side chain of Y242 (Fig. 2). Aclosed conformation is observed for the A±B interface,whereas the B±A interface shows a more open conformation(Fig. 2a and b). ATP was modeled manually into the RepAstructure with a similar procedure to that described inNiedenzu et al. (5). The overlay of the open and the closedconformation shows that both conformations can be convertedinto each other by a rotation of the left protomer (monomer Aat the A±B interface; see Fig. 2a) around the axis of helix F

Table 3. Diffraction data collection and re®nement statistics

Data collectionWavelength (AÊ ) 1.5418Resolution limit (AÊ ) 2.52Total observations 305969Unique observations 40046Rsym

a 0.106Rsym (2.58±2.52 AÊ ) 0.620I/s 13.6I/s (2.58±2.55 AÊ ) 3.3Space group P4332Unit cell: a (AÊ ) 190.29Model re®nementResolution range (AÊ ) 25±2.55Re¯ections 37753Protein atoms re®ned 3864Ligand atoms re®ned 10Waters O re®ned 421Rcryst/Rfree

b 0.176/0.211r.m.s.d. from stereochemical target valuesr.m.s.d. bond lengths (AÊ ) 0.021r.m.s.d. bond angles (°) 1.838

aRsym = å|I ± <I>|/åI, where <I> is the average intensity over symmetryequivalent re¯ections.bRcryst (Rfree) = å||Fo| ± |Fc||/å|Fo|, where Fo and Fc are the observed andcalculated structure factor amplitudes. For the calculation of Rfree, a randomsubset of 991 re¯ections (2.6%) was chosen after data processing andexcluded from re®nement.

Figure 1. Localization of mutations in the RepA structure. (a) View of the RepA hexamer from the C-terminal side (top) of the ring. Monomer A is shown ingreen, monomer B in orange. (b) 3-Dimensional structure of the RepA monomer viewed from inside the ring. Positions of mutated amino acids arehighlighted, N-terminal deletions are marked. The conserved family 4 helicase motifs are colored: H1, green; H1a, orange; H2, blue; H3, red; H4, violet.Dotted lines represent regions of poorly de®ned electron density of segment 180±199. (c) The bar represents RepA consisting of 278 amino acid residues. Thepositions of exchanged residues are given. Single dots (monomer A) and double dots (monomer B) mark the proposed ATPase catalytic residues at themonomer±monomer interface. Single and double colons highlight the residues of monomers A and B, respectively, that may sandwich the adenine in theclosed conformation (see also Fig. 2a). Within the bar, regions of poorly de®ned electron densities are marked in black. (d) Sequence comparison of residuesstructurally equivalent according to the output of MAPS (http://bioinfo1.mbfys.lu.se/TOP/maps.html) in RepA, T7 gp4, E.coli RecA, F1 ATPase subunit B,E.coli Rep and B.stearothermophilus PcrA. Above the RepA sequence the corresponding a-helices A±G and b-strands 1±9 are shown by boxes and arrows,respectively. Dotted lines represent the disordered segments. The conserved motifs H1, H1a, H2, H3 and H4 are boxed and colored according to (a). Aminoacids exchanged are highlighted in cyan. Deleted N-terminal and C-terminal residues are in cyan boxes.


Dow



ecember 2021

(Fig. 2c). This movement opens and closes the cleft betweenthe monomers, which is the proposed binding site for thepurine moiety of ATP. The position of the N-terminal hook

(see below and Fig. 3) to the neighboring protomer (monomerB at the A±B interface; see Fig. 2a) is not affected by thismovement.


Dow



ecember 2021


Dow



ecember 2021

The RepA N-terminus is essential for hexamerformation

The remarkable stability of the RepA hexamers might resultfrom interactions of the hook-like N-terminus of one monomerwith a central portion of the neighboring molecule (5) (Fig. 3).To test this hypothesis, deletion derivatives were constructedlacking ®ve, seven or nine N-terminal amino acid residues(Figs 1 and 3). Since F12 is situated at the end of helix A,which is the central part of the N-terminal hook (Fig. 3), wepredict that the angle would widen if this residue was deleted.Then the stability of the hexamer might be altered. Therefore,the derivative RepA DF12 lacking only residue F12 was made.To study the oligomeric state of these RepA variants, theywere puri®ed and subjected to velocity sedimentation in aglycerol gradient. Subsequently aliquots of the collectedfractions were analyzed by gel electrophoresis (Fig. 4). Bothwild-type RepA and RepA DF12 behaved as hexamers of~180 kDa. However, compared to the wild-type protein, RepADF12 was less stable, since protein sedimenting with reduced

velocity was observed in signi®cant amounts. Obviously thepresence of F12 stabilizes the RepA hexamer signi®cantlythrough hydrophobic interactions (Fig. 3).

RepA D1±7 sedimented with a coef®cient below that ofBSA (68 kDa), suggesting the presence of monomeric ordimeric RepA subunits (Fig. 4). In the electron microscope,RepA DF12 and RepA D1±5 had a hexameric ring-likestructure comparable to wild-type RepA, whereas for RepAD1±7 and for RepA D1±9 no hexamers were observed (Table 4and data not shown). These data show that hexamerizationoccurs via the very N-terminus of each monomer with positionI6 and/or N7 being essential, whereas residues A1±P5 aredispensable. Consequently, at least part of the oligomerizationdomain is assigned to the N-terminal end of the molecule A.Its N-terminal segment A1±N7 interacts with the neighboringmolecule B through a-helices C and D and b-strand 3, whichform four hydrogen bonds, one water-mediated (Ile107N-H´´´W´´ÓIle6) and the other three direct (Ile6N-H´´ÓIle107;Gln108Ne-H´´ÓHis3; His3Ne-H´´ÓIle111). In addition, I6

Figure 2. Subunit interface: the catalytic center of the NTPase. Stereo views of sections of the two different interfaces of the monomer±monomer interface inthe cubic crystal structure are shown. ATP was modeled into the structure as described (5). Monomer A is shown in green, monomer B in orange. The sidechains of amino acids mutated for this study are shown as stick models. The dotted line indicates the disordered segment 180±199. (a) The closedconformation at the A±B interface. (b) The open conformation at the B±A interface. (c) Overlay of the closed conformation (dark colors) and the openconformation (light colors). Monomer A of the open interface was superposed on monomer B of the closed interface. Helix F is marked.

Figure 3. Linkage between adjacent monomers that results in stable interaction. The proposed dimerization regions of two adjacent monomers are given ingreen and orange (monomers A and B, respectively). The dotted lines represent hydrogen bonds. The interactions are supported/enhanced by van der Waalscontacts (not indicated). Oxygen and nitrogen atoms are in red and blue, respectively. The red dot represents a water molecule that ®ts in between the I6 back-bone oxygen of monomer A and the I107 backbone nitrogen of monomer B.


Dow



ecember 2021

and I8 of monomer A form hydrophobic interactions with I82,I107, L86 and L105 of monomer B, respectively (Fig. 3).

The replication-de®cient plasmid RSF1010KDrepAserved as a tool for the phenotypic evaluation of repAmutations in vivo

The repA deletion derivative RSF1010KDrepA was con-structed as described in Materials and Methods. Since repA isessential for RSF1010 replication, such a construct is onlyviable in the presence of repA in trans. Consequently theability of the repA variants to support RSF1010 replicationin vivo could be tested by complementation. IntroducingRSF1010KDrepA DNA into E.coli strain SCS1(pGZ18-20)

which provides repA in trans restores the replication ability ofRSF1010KDrepA. Accordingly, RSF1010KDrepA was intro-duced into each strain that carried a repA allele. This in vivosystem (Fig. 5) was used to analyze the replication propertiesof all repA mutations generated and served as a tool to selectreplication-de®cient repA variants. The mutated repA geneswere overexpressed and the encoded products were puri®edfor in vitro studies.

A stable hexamer is a prerequisite for RepA activity

Puri®ed RepA DF12 hydrolyzed ATP in the presence ofM18mp18 ssDNA less ef®ciently than wild-type protein,whereas RepA D1±7 showed rather minor activity, althoughclearly above background. To separate minor impurities fromthese preparations, the ATPase activity of different fractionsof the glycerol gradient centrifugation (Fig. 4) was measured.For RepA DF12 the peak of maximum activity correspondedto the protein peak, whereas for the monomeric/dimeric RepAD1±7 (see above), the peak of the residual activity did notcoincide with the protein peak, indicating that the observedminor ATP hydrolysis was mainly due to an impurity (Fig. 4).RepA DF12 is less stable than the wild-type protein. TheATPase activity of this derivative was not stimulated byssDNA. In the absence of ssDNA, the hydrolyzing activity ofRepA DF12 was reduced 3-fold compared to the wild-typeprotein (Fig. 4). Since for the RepA DF12 variant no helicaseactivity was detectable (not shown), a stable RepA hexamer isrequired for stimulation of the ATPase activity by ssDNA andfor unwinding of dsDNA.

Of the 5¢-terminal deletions in repA (Fig. 1b and d), only theD1±5 derivative was able to complement RSF1010KDrepA,whereas the D1±7 and D1±9 deletions were not (Table 4).Also, repADF12 failed to restore replication ofRSF1010KDrepA. Hence, interactions with the replisome didnot stabilize the labile hexamers of RepA DF12 in vivo toful®ll its role in replication, although this protein stillpossesses ATPase activity (Fig. 4). These data coincide withthose for the helicase activity of the respective puri®ed protein(Table 4) and demonstrate that a stable hexamer is not onlyessential in vitro for helicase activity but also in vivo forRSF1010 replication.

Alanine scan of single amino acid residues proposed toplay key roles

Nine residues proposed to be essential for RepA function (5)were substituted by alanine (Fig. 1c). These substitutionsinclude amino acids within all ®ve motifs conserved inhelicase family 4 (Fig. 1d). Y242 of one monomer and R85 ofthe neighboring monomer were proposed to bind the purinemoiety of ATP at the interface between RepA monomers(Fig. 2a and b). For steric reasons R253 might serve as analternative for R85 (Fig. 2a). R206 points towards theg-phosphate of the nucleotide and forms an `arginine ®nger'analogous to that observed for GTPase-activating proteins andalso discussed for T7 gp4 helicase (10). K42, E76 and D139are conserved in RepA, RecA and T7 gp4 and proposed tofunction as essential residues of the catalytic center. H178 islocated in the vicinity of the g-phosphate of ATP andconserved in RepA and T7 gp4. D77 was changed to alaninesince its carboxyl group was likely to interact with N6 of ATP(Fig. 2b).

Figure 4. Glycerol gradient centrifugation of RepA proteins. A volume of150 ml of puri®ed protein (RepA wild-type and RepA DF12, 0.4 mg each;RepA D1±7, 0.12 mg) was layered onto a 3.7 ml, 15±35% linear glycerolgradient in buffer containing 20 mM Tris±HCl pH 7.6, 20 mM NaCl, 2 mMDTT, 0.1% Brij-58 and 0.1 mM EDTA and centrifuged at 270 000 g for15 h at 4°C. After the run, the bottoms of the tubes were pricked with aneedle and fractions were collected. Aliquots were taken and analyzed forprotein content by gel electrophoresis and for ATPase activity in thepresence and absence of ssDNA as described in Materials and Methods.Open and ®lled circles, wild-type RepA without and with ssDNA; ®lledsquares, RepA DF12 both with and without ssDNA; ®lled triangles, RepAD1±7, both with and without ssDNA. Note the logarithmic scale of theordinate. Aliquots of the fractions were electrophoresed in 15% polyacryla-mide gels containing 0.1% SDS. The gels were stained with Serva Blue Rand scanned. Fraction numbers of the gradient correspond to slot numbersof the gel. The direction of sedimentation is from right to left. 1, catalase(1250 kDa, s20,w = 11.3 S), 2, aldolase (2158 kDa, s20,w = 7.8 S) and 3,BSA (368 kDa, s20,w = 4.4 S) were run in parallel as standards, the peakpositions are marked by arrows.


Dow



ecember 2021

All these nine repA mutations failed to complementreplication of RSF1010KDrepA (Table 4). Accordingly, allnine puri®ed RepA variants were inactive in hydrolyzing ATPexcept for RepA H178A (see below and Table 4). RepAY242T, which in contrast to RepA Y242A still provides ahydroxyl residue, was inactive, as was the Y242A variant.Therefore the aromatic side chain of tyrosine proposed to bindthe purine residue of ATP is essential for enzymatic function.

Decoupling ATPase and helicase activity

Residue H178 is part of the conserved helicase motif H3 and islocated at the end of b-strand 5 (Fig. 1b and d; 5). H178 wassuggested to be part of the catalytic centers of RepA helicase.Since it is in close proximity to the g-phosphate of ATP(Fig. 2a and b) it might act as a sensor for this phosphate todistinguish ATP from ADP (10). While the ATPase activity ofRepA H178A was similar to that of the wild-type protein, itshelicase activity was only marginally above backgroundlevels: at a monomer concentration of 50 pM, less than 5% ofthe wild-type activity was observed (Fig. 6). Helicase activityof RepA H178A was neither increased notably by largeramounts of protein nor by higher ATP concentrations in thereaction mixture. Since the ATPase activity of RepA H178Awas stimulated by the addition of ssDNA, the RepA H178Ahexamer is suggested to still be able to move along ssDNAcomparable to the wild-type protein. Hence, the resultsdemonstrate that H178 is dispensable for catalyzing the

hydrolysis of ATP, but plays an essential role in DNA stranddisplacement.

Hexamerization is independent of proposed catalyticamino acids

Hexamerization is required for RepA to function as a motorprotein. Since each of the six ATPase catalytic centers of theRepA hexamer is constituted by two adjacent monomers, localalteration of the 3-dimensional structure at the contact surfacemay in¯uence oligomerization and the orientation of subunits,besides inhibiting catalysis. Therefore the oligomerizationstate of the replication-de®cient RepA variants was studied byelectron microscopy. They all maintained a hexameric ring-like structure comparable to wild-type RepA irrespective ofthe position or the proposed task of the residue exchanged(summarized in Table 4). Consequently, none of the aminoacids suggested to participate in catalysis are essential tomaintain the hexameric state.

One of the negative charges of the glutamate triplet inloop L2 is essential for RepA enzymatic activity

The triplet E147±E148±E149, located in the loop L2 formedbetween b-strand b4 and a-helix aF, con®nes the narrowestpart of the central channel of the hexamer. To answer thequestion as to whether the negative charge of the sidechain is important, each glutamate was changed to theuncharged glutamine. Using the in vivo complementation of

Table 4. In vivo and in vitro properties of RepA mutant proteins

repA allele Complementation ofRSF1010DrepA

Transdominance Hexamerization ATPase activity Helicase activity

wt + ± + + +D1±5 + ± + + +D1±7 ± ± ± (±) ±D1±9 ± ± nt nt ntDF12 ± ± + + ±D276±278 ± ± + + +D274±278 ± ± + + +D269±278 ± ± + (+) (±)K42A ± +a + ± ±E76A ± (+)b + ± ±D77A ± +a + ± ±R85A ± ± + ± ±D139A ± +c + ± ±E147Q + ± nt nt ntE148Q +d ± nt nt ntE149Q ± +e + ± ±H178A ± (+)b + + (±)Q191E + ± nt nt ntQ192E ± ± + + +Q193E (+)f ± nt nt ntR206A ± +a + ± ±Y242A ± ± + ± ±Y242T ± ± nt nt ntR253A ± +e + ± ±

Hexamerization was studied by electron microscopy as described (43). (±) represents a negligible activity above background in the presence of high proteinconcentrations.aIn the transdominance test, the number of colonies is reduced 100-fold.bIn the transdominance test, the number of colonies is reduced 3-fold.cIn the transdominance test, the number of colonies is reduced 1000-fold.dCompared to complementation with wild-type repA, the number of colonies is reduced 3-fold.eIn the transdominance test, the number of colonies is reduced 10-fold.fCompared to complementation with wild-type repA, the number of colonies is reduced 10-fold.


Dow



ecember 2021

RSF1010KDrepA assay outlined in Figure 5, only repA E149Qresulted in a negative phenotype, whereas E147Q andE148Q behaved similarly to wild-type repA (see above).Since E149Q still forms hexamers but lacks ATPase andhelicase activity, the carboxyl group exerts an importantfunction in the protein. In the cubic structure, this carboxylgroup chain forms a hydrogen bridge with H177, thereby®xing the position of helix F relative to the central b-sheet(Fig. 1b).

Do mutant RepA monomers form heterooligomers withwild-type protein?

Transdominance was found for hexameric replicative heli-cases like DnaB, T7 gene 4 protein or SV40 large T-antigen(28±30). If mutant and wild-type protein monomers assembledinto an inactive heterohexamer that inhibits the replicationprocess, a dominant lethal phenotype was observed. Thesystem used to study the behavior of the mutations wasidentical to that outlined in Figure 6 except that RSF1010Kwas used. Each strain containing a repA allele was trans-formed with the replication-pro®cient and selectableRSF1010K as described in Materials and Methods. None ofthe repA mutations that were functional in vivo showed anyinterference with the wild-type (Table 4). Even for RepAE148Q and RepA Q193E, which restored replication ofRSF1010KDrepA less ef®ciently than the wild-type gene, notransdominant effect was observed. Therefore, in the case thatmixed complexes of mutant and wild-type RepA exist, theseheterooligomers are both enzymatically active and able tofunctionally interact with the replication machinery. For themajority of the replication-de®cient repA mutations, includingall N-terminal and C-terminal deletion derivatives, no or only

slight transdominance was observed (Table 4). However, themutations repA K42A, repA D77A, repA D139A and repAR206A in helicase motifs H1, H1a, H2 and H4, respectively(Fig. 1d), strongly interfered with wild-type repA. Comparedto the wild-type, the number of colonies obtained with repAK42A, repA D77A and repA R206A was reduced 100-fold,with the D139A mutation even 1000-fold (Table 4). Hence,in vivo the enzymatically inactive variants RepA K42A, RepAD77A, RepA D139A and RepA R206A assemble intocomplexes with wild-type protein, which are unsuitable topromote RSF1010K replication.

The amide of glutamine Q192 is essential for RepAfunction in vivo, but irrelevant for enzymatic activity

To explore the possible roles of the glutamines of the tripletQ191±Q192±Q193 located in the disordered segment of 20residues (S180±S199) between b5 and b6 each of the residueswas changed to glutamic acid. Only the Q192E mutationshowed a negative phenotype in repA complementation. repAQ191E behaved like the wild-type, but repA Q193E was10-fold less ef®cient in restoring the replication ability ofRSF1010KDrepA (see below). Since the in vitro activities ofpuri®ed RepA Q192E proved to be comparable to wild-typeprotein (Fig. 6), we speculate that this residue might beinvolved in the interaction with other replication proteins.

The C-terminus is required for RepA enzymatic activityand interacts with the replication machinery

To test whether the ¯exibility of the RepA C-terminus is offunctional importance for motoring, derivatives lacking three,®ve and ten C-terminal residues were constructed. Since allthree C-terminally truncated RepA proteins assembleinto hexamers, the C-terminal 10 residues are dispensablefor hexamerization (Table 4). RepA D276±278 and RepAD274±278 lacking three and ®ve amino acids, respectively,

Figure 5. Scheme of the in vivo repA complementation assay. SCS1 cellsharboring a plasmid that carries a repA allele were transformed withRSF1010KDrepA. Then the cells were plated onto solid medium containingampicillin and kanamycin. If the mutation leaves RepA protein active,RSF1010KDrepA replicates, confers resistance to kanamycin and the cellsform colonies. If the mutation renders RepA inactive, RSF1010KDrepAcannot replicate and no colonies are observed.

Figure 6. Helicase activity of wild-type and mutant RepA proteins. Puri®edRepA proteins were assayed for the ability to displace a radioactivelylabeled oligonucleotide bound to viral M13mp18 DNA as described inMaterials and Methods. Circles, RepA wt; squares, RepA H178A; triangles,RepA Q192A; dots, RepA D269±278. The percentage of the oligonucleotidedisplaced was determined. For each protein, the values plotted wereaveraged from at least two independent experiments.


Dow



ecember 2021

behaved similarly to the wild-type protein concerning ATPhydrolysis and DNA unwinding (data not shown). AlthoughRepA D269±278 lacking 10 residues still hydrolyzes ATP, theDNA unwinding capability was only marginally abovebackground (Fig. 6). Addition of higher amounts of thisprotein did not result in a signi®cantly higher helicase activity.Therefore, the RepA helicase domain extends into theC-terminus except for the last ®ve C-terminal residues thatdo not contribute to helicase activity. In contrast to theenzymatic activities, none of the three C-terminal variants wasfunctional in vivo in restoring the replication ability ofRSF1010KDrepA (Table 4). This demonstrates that even thethree very C-terminal residues are essential for RSF1010replication and suggests that the C-terminus interacts withRSF1010-encoded and/or host-encoded replication proteins.

DISCUSSION

Domain structure of RepA

The combination of the data of this study yielded a partialdomain structure of RepA that includes the residues forhexamerization, for interaction with the low molecular weightsubstrate ATP and regions possibly involved in the interactionwith replication proteins of RSF1010 and the host (Fig. 7). Weconsider the unstructured ¯exible 20 residue region (180±199)of RepA as the most interesting portion of the molecule sinceit is likely to constitute the moving part(s) of the motorprotein. This stretch of the protein points to the central hole ofthe hexamer, which is proposed to enclose or embrace theDNA single strand on which the relative movement takesplace, driving the strand separation process (31). To determinethe structure of the ¯exible region in RepA additional work isrequired that would arrest the residues in one conformation bymeans of mutations or by the addition of tightly bindingeffector molecules. A comparison with the correspondingregion of the T7 gp4 structure shows that in the monomer asimilarly unstructured region is present at an almost identicalposition in motifs H3 and H4 (10), whereas in the hexamericstructure the same region is clearly de®ned (11). Since inRecA the equivalent loop proposed to be involved in DNAbinding is also unstructured (32), we speculate that thissegment in RepA may only be arrested and of de®ned structurein the presence of a suitable ssDNA substrate. Since thedistance between the Ca atoms of the ¯anking amino acidspositions 179 and 200 is 11.2 AÊ , there is no space for a singlea-helical connection of ~30 AÊ corresponding to these 20disordered residues. Otherwise, in this region an extensiveconformational change of RepA would have to take place.

Stability of the hexameric ring structure of RepA

Since no cofactors like divalent cations or nucleotides areneeded to sustain the hexameric ring structure of RepA, itproves to be robust under our experimental conditions in vitro,contrasting with other hexameric helicases that requirecofactors to stabilize the ring structures. The high stabilityof the ring architecture of RepA is obviously associated withhooking up the N-terminus of one subunit to an eye of theadjacent subunit by hydrogen bonding and several hydro-phobic interactions. If the inter-linkage of subunits isabolished by deletion of the N-terminal seven amino acids,

the ring structure and in parallel the enzymatic activity arelost. This demonstrates that the ring structure is essential forRepA function in vivo and in vitro. In other systems similarobservations were made: hexamerization is also a prerequisitefor DNA unwinding activity of the T7 helicase, since deletionsat the N-terminus of the helicase domain prevent hexameriza-tion and abolish helicase activity (33).

Mechanistic implications of the two conformationsobserved at the subunit interfaces

There are additional interaction surfaces between adjacentsubunits that contribute to ring stability and/or enzymaticactivity. The hook±eye interaction to hold the ring structuretogether appears to be identical in the A±B and B±Ainterfaces, whereas differences are obvious in the proposedbinding site for the purine moiety of ATP (Fig. 2). Althoughthe major part of the ATP binding site resides in one subunit,the adjacent subunit seems to contribute considerably to ATPinteraction mainly by residues Y242 and R206 at theinteraction interface. An open and a more closed conformationare arranged in alternating positions in the cubic structure ofRepA, demonstrating the existence of three equivalent dimersin the hexameric ring structure associated with two types ofATP binding sites. This ®nding is also in agreement withbiochemical studies which showed that the ATP binding sitesin RepA are not identical but three are of low and three of highaf®nity for ATP (34). Mechanistically this is probablyimportant because the conversion of the open into the closedtype adenine binding site may lead to some kind of a pulsationof the molecule that could periodically change the diameter orthe shape of the central hole. Such conformational changesmay induce the relative movement between RepA and DNA.

Motif H4 was proven to be part of the ssDNA bindingdomain of the T7 gp4 helicase (35). In RepA, R206, which islocated in a position suitable to serve as an `arginine ®nger', ispart of H4. This suggests that binding of ssDNA alters theconformation of H4, pushing R206 towards the ATPase activesite, thereby stimulating ATP hydrolysis. Since RepA R206Ais ATPase-de®cient, the `arginine ®nger' appears to berequired for ATP hydrolysis.

Coupling of ATPase and translocation viaconformational switching

Of particular importance seems to be residue H178 as part ofthe ATP binding pocket that may act as a sensor for the g-phosphate. Replacement of H178 by alanine renders RepA

Figure 7. RepA domains for protein±protein interaction. Above and belowthe bar, which represents RepA, boxes mark the domains for hexamerizationand interaction with host and/or plasmid-encoded replication proteins. TheN-terminal hexamerization domain around positions 1±12 of one monomerinteracts with the counterpart of the adjacent monomer around positions107±111 given below the bar (for details see Fig. 3). K42 marks the positionof the Walker A motif.


Dow



ecember 2021

completely inactive in vivo. In addition, the in vitro propertiesare interesting because RepA H178A is the only variant thatstill hydrolyses ATP in an ssDNA-dependent manner but lackshelicase activity. This variant still binds to viral M13 DNA asdoes the wild-type protein (data not shown). In helicases thehydrolysis of NTPs is generally thought to be coupled to themovement along a single DNA strand (reviewed in, forexample, 16,18,36). One intriguing explanation for the effectobserved for RepA H178A is that the released energy cannotbe utilized for DNA unwinding although the translocation onssDNA apparently takes place, as indicated by the stimulationof ATPase activity. However, it is conceivable that the forkmimicked by the non-base paired tail of the substrate acts as aninsurmountable obstacle for the H178 variant.

Alternatively, the coupling between the ATPase and theconcerted binding/release of ssDNA required for translocationof the single strand is lost. The histidyl residue at the end ofmotif H3 (H178) is positionally conserved in RepA and T7gp4 and in close proximity to the bound g-phosphate of theNTP (Fig. 1d; 5,10,11). In analogy to the suggestion made forT7 gp4 (10), H178 may act as a conformational switch orsensor for the g-phosphate of NTP. Such conformationalswitching has already been proposed for the strand exchangeactivity of RecA (37) and the DNA unwinding activity of themonomeric helicase PcrA (8) and the RNA helicase NS3 (38).In DnaB proteins, PcrA and RecA, the positional equivalent toH178, is a conserved glutamine. For Bacillus subtilis DnaB,Soultanas and Wigley demonstrated by site-directed mutagen-esis that substitution of this glutamine results in a proteinpro®cient in ATPase activity but mostly de®cient in DNAunwinding (39). Depending on the presence or absence of a g-phosphate, the ssDNA binding site, part of which is located inloop L2 that connects motifs H3 and H4, and motif H4(Fig. 1d) are altered from high to low af®nity and vice versa. Ifthis switch is lost in RepA H178A, the af®nity for ssDNAcannot be modulated. The enzyme is left in a non-functionalstatic state with at least one catalytic ATPase active siteactivated by eventually bound ssDNA.

Interaction of RepA with other proteins

Two segments of RepA play vital roles in the interaction withthe replication machinery: (i) the very C-terminus, sincedeletion of three C-terminal residues abolishes RSF1010replication without affecting the enzymatic activities (Table 4and Fig. 7); (ii) the ¯exible loop (positions 180±199)containing one half of motif H4. Q192 is hypothesized to bean essential residue in protein±protein interaction, since thein vitro activities of RepA Q192E are comparable to wild-typeprotein, however, in vivo the variant is unable to complementrepA (Table 4). SSB protein and the polymerase IIIholoenzyme have to be taken into account as interactionpartners. Since in other replication systems primase andhelicase are interacting (E.coli DnaG±DnaB, phage T4 gp61±gp41) or even located on a single polypeptide chain (phagesT7 gp4, P4 gpa), the most likely interaction partner for RepAhelicase is the RSF1010-encoded RepB primase, which couldnot be substituted by the host primase DnaG. Anotherpromising candidate is the RSF1010-speci®c RepC originbinding protein. We will purify RepB and RepC to experi-mentally prove these predictions.

ACKNOWLEDGEMENTS

We thank Hans Lehrach for generous support. The experttechnical assistance of Marianne Schlicht is greatly appreci-ated. This work was supported by European Commission grantQLRT-1999-30634 to E.L. and W.S.

REFERENCES

1. van Brabant,A.J., Stan,R. and Ellis,N.A. (2000) DNA helicases, genomicinstability and human genetic disease. Annu. Rev. Genomics Hum.Genet., 1, 409±459.

2. Scherzinger,E., Haring,V., Lurz,R. and Otto,S. (1991) Plasmid RSF1010DNA replication in vitro promoted by puri®ed RSF1010 RepA, RepBand RepC proteins. Nucleic Acids Res., 19, 1203±1211.

3. Scherzinger,E., Ziegelin,G., BaÂrcena,M., Carazo,J.M., Lurz,R. andLanka,E. (1997) The RepA protein of plasmid RSF1010 is a replicativeDNA helicase. J. Biol. Chem., 272, 30228±30236.

4. RoÈleke,D., Hoier,H., Bartsch,C., Umbach,P., Scherzinger,E., Lurz,R. andSaenger,W. (1997) Crystallization and preliminary X-raycrystallographic and electron microscopic study of a bacterial DNAhelicase (RSF1010 RepA). Acta Crystallogr., D53, 213±216.

5. Niedenzu,T., RoÈleke,D., Bains,G., Scherzinger,E. and Saenger,W. (2001)Crystal structure of the hexameric replicative helicase RepA of plasmidRSF1010. J. Mol. Biol., 306, 479±487.

6. Xu,H., StraÈter,N., SchroÈder,W., BoÈttcher,C., Ludwig,K. and Saenger,W.(2003) Structure of DNA helicase RepA in complex with sulfate at 1.95AÊ resolution implicates structural changes to an `open' form. ActaCrystallogr., D59, 815±822.

7. Ilyina,T.V., Gorbalenya,A.E. and Koonin,E.V. (1992) Organization andevolution of bacterial and bacteriophage primase-helicase systems.J. Mol. Evol., 34, 351±357.

8. Subramanya,H.S., Bird,L.E., Brannigan,J.A. and Wigley,D.B. (1996)Crystal structure of a DExx box DNA helicase. Nature, 384, 379±383.

9. Korolev,S., Hsieh,J., Gauss,G.H., Lohman,T.M. and Waksman,G. (1997)Major domain swiveling revealed by the crystal structures of complexesof E. coli Rep helicase bound to single-stranded DNA and ADP. Cell, 90,635±647.

10. Sawaya,M.R., Guo,S., Tabor,S., Richardson,C.C. and Ellenberger,T.(1999) Crystal structure of the helicase domain from the replicativehelicase-primase of bacteriophage T7. Cell, 99, 167±177.

11. Singleton,M.R., Sawaya,M.R., Ellenberger,T. and Wigley,D.B. (2000)Crystal structure of T7 gene 4 ring helicase indicates a mechanism forsequential hydrolysis of nucleotides. Cell, 101, 589±600.

12. Allison,T.J., Wood,T.C., Briercheck,D.M., Rastinejad,F., Richardson,J.P.and Rule,G.S. (1998) Crystal structure of the RNA-binding domain fromtranscription termination factor rho. Nature Struct. Biol., 5, 352±356.

13. San Martin,M.C., Stamford,N.P., Dammerova,N., Dixon,N.E. andCarazo,J.M. (1995) A structural model for the Escherichia coli DnaBhelicase based on electron microscopy data. J. Struct. Biol., 114,167±176.

14. Fass,D., Bogden,C.E. and Berger,J.M. (1999) Crystal structure of theN-terminal domain of the DnaB hexameric helicase. Structure, 7,691±698.

15. Ali,J.A. and Lohman,T.M. (1997) Kinetic measurement of the step sizeof DNA unwinding by Escherichia coli UvrD helicase. Science, 275,377±380.

16. Patel,S.S. and Picha,K.M. (2000) Structure and function of hexamerichelicases. Annu. Rev. Biochem., 69, 651±697.

17. Yarranton,G.T. and Gefter,M.L. (1979) Enzyme-catalyzed DNAunwinding: studies on Escherichia coli rep protein. Proc. Natl Acad. Sci.USA, 76, 1658±1662.

18. Marians,K.J. (2000) Crawling and wiggling on DNA: structural insightsto the mechanism of DNA unwinding by helicases. Structure, 8,R227±R235.

19. Ziegelin,G., Linderoth,N.A., Calendar,R. and Lanka,E. (1995) Domainstructure of phage P4 a protein deduced by mutational analysis.J. Bacteriol., 177, 4333±4341.

20. Sambrook,J., Fritsch,E.T. and Maniatis,T. (1989) Molecular Cloning: ALaboratory Manual, 2nd Edn. Cold Spring Harbor Laboratory Press,Cold Spring Harbor, NY.


Dow



ecember 2021

21. Scholz,P., Haring,V., Wittmann-Liebold,B., Ashman,K., Bagdasarian,M.and Scherzinger,E. (1989) Complete nucleotide sequence and geneorganization of the broad-host-range plasmid RSF1010. Gene, 75,271±288.

22. Weiner,M.P., Costa,L.G., Schoettlin,W., Cline,J., Mathur,E. andBauer,J.T. (1994) Site-directed mutagenesis of double-stranded DNA bypolymerase chain reaction. Gene, 151, 119±123.

23. Hanahan,D. (1983) Studies on transformation of Escherichia coli withplasmids. J. Mol. Biol., 166, 557±580.

24. Crute,J.J., Mocarski,E.S. and Lehman,I.R. (1988) A DNA helicaseinduced by herpes simplex virus type 1. Nucleic Acids Res., 16,6585±6596.

25. Otwinowski,Z. and Minor,W. (1997) Processing of X-ray diffraction datacollected in oscillation mode. Methods Enzymol., 276, 461±472.

26. Bailey,S. (1994) The CCP4 suite-programs for protein crystallography.Acta Crystallogr., D50, 760±763.

27. Jones,T.A., Zou,J.Y., Cowan,S.W. and Kjeldgaard,M. (1991) Improvedmethods for binding protein models in electron density maps and thelocation of errors in these models. Acta Crystallogr., D47, 110±119.

28. Maurer,R. and Wong,A. (1988) Dominant lethal mutations in the dnaBhelicase gene of Salmonella typhimurium. J. Bacteriol., 170, 3682±3688.

29. Notarnicola,S.M. and Richardson,C.C. (1993) The nucleotide-bindingsite of the helicase-primase of bacteriophage T7: interaction of mutantand wild-type proteins. J. Biol. Chem., 268, 27198±27207.

30. Farber,J.M., Peden,K.W. and Nathans,D. (1987) Trans-dominantdefective mutants of simian virus 40 T antigen. J. Virol., 61, 436±445.

31. Egelman,E.H., Yu,X., Wild,R., Hingorani,M.M. and Patel,S.S. (1995)Bacteriophage T7 primase/helicase proteins form rings around single-stranded DNA that suggest a general structure for hexameric helicases.Proc. Natl Acad. Sci. USA, 92, 3869±3873.

32. Story,R.M., Weber,I.T. and Steitz,T.A. (1992) The structure of the E. coliRecA protein monomer and polymer. Nature, 355, 318±325.

33. Guo,S., Tabor,S. and Richardson,C.C. (1999) The linker region betweenthe helicase and primase domains of the bacteriophage T7 gene 4 proteinis critical for hexamer formation. J. Biol. Chem., 274, 30303±30309.

34. Xu,H., Frank,J., Niedenzu,T. and Saenger,W. (2000) DNA helicaseRepA: cooperative ATPase activity and binding of nucleotides.Biochemistry, 39, 12225±12233.

35. Washington,M.T., Rosenberg,A.H., Grif®n,K., Studier,F.W. andPatel,S.S. (1996) Biochemical analysis of mutant T7 primase/helicaseproteins defective in DNA binding, nucleotide hydrolysis and thecoupling of hydrolysis with DNA unwinding. J. Biol. Chem., 271,26825±26834.

36. Singleton,M.R. and Wigley,D.B. (2002) Modularity and specialization insuperfamily 1 and 2 helicases. J. Bacteriol., 184, 1819±1826.

37. Story,R.M. and Steitz,T.A. (1992) Structure of the recA protein±ADPcomplex. Nature, 355, 374±376.

38. Yao,N., Hesson,T., Cable,M., Hong,Z., Kwong,A.D., Le,H.V. andWeber,P.C. (1997) Structure of the hepatitis C virus RNA helicasedomain. Nature Struct. Biol., 4, 463±467.

39. Soultanas,P. and Wigley,D.B. (2002) Site-directed mutagenesis revealsroles for conserved amino acid residues in the hexameric DNA helicaseDnaB from Bacillus stearothermophilus. Nucleic Acids Res., 30,4051±4060.

40. Balzer,D., Ziegelin,G., Pansegrau,W., Kruft,V. and Lanka,E. (1992)KorB protein of promiscuous plasmid RP4 recognizes inverted sequencerepetitions in regions essential for conjugative plasmid transfer. NucleicAcids Res., 20, 1851±1858.

41. Guerry,P., van Embden,J. and Falkow,S. (1974) Molecular nature of twononconjugative plasmids carrying drug resistance genes. J. Bacteriol.,117, 619±639.

42. Yanisch-Perron,C., Vieira,J. and Messing,J. (1985) Improved M13 phagecloning vectors and host strains: nucleotide sequences of the M13mp18and pUC19 vectors. Gene, 33, 103±119.

43. Krause,S., BaÂrcena,M., Pansegrau,W., Lurz,R., Carazo,J.M. andLanka,E. (2000) Sequence-related protein export NTPases encoded bythe conjugative transfer region of RP4 and by the cag pathogenicityisland of Helicobacter pylori share similar hexameric ring structures.Proc. Natl Acad. Sci. USA, 97, 3067±3072.


Dow



ecember 2021

Hexameric RSF1010 helicase RepA - Nucleic Acids Research

Documents

Transcript of Hexameric RSF1010 helicase RepA - Nucleic Acids Research