Crystal structure of the N-terminal domain of the DnaB hexameric helicase

Research Article

691

Crystal structure of the N-terminal domain of the DnaB hexameric helicase Deborah Fass1†, Cynthia E Bogden1 and James M Berger2* Background: The hexameric helicase DnaB unwinds the DNA duplex at the Escherichia coli chromosome replication fork. Although the mechanism by which DnaB both couples ATP hydrolysis to translocation along DNA and denatures the duplex is unknown, a change in the quaternary structure of the protein involving dimerization of the N-terminal domain has been observed and may occur during the enzymatic cycle. This N-terminal domain is required both for interaction with other proteins in the primosome and for DnaB helicase activity. Knowledge of the structure of this domain may contribute to an understanding of its role in DnaB function.

Addresses: 1Whitehead Institute, Cambridge, Massachusetts 02142, USA and 2Department of Molecular and Cell Biology, Stanley Hall, University of California, Berkeley, California 94720, USA.

Results: We have determined the structure of the N-terminal domain of DnaB crystallographically. The structure is globular, highly helical and lacks a close structural relative in the database of known protein folds. Conserved residues and sites of dominant-negative mutations have structurally significant roles. Each asymmetric unit in the crystal contains two independent and identical copies of a dimer of the DnaB N-terminal domain.

Key words: DnaB, DNA helicase, DNA replication, domain

†Present Address: Department of Structural Biology, Weizmann Institute of Science, Rehovot 76100 Israel.

*Corresponding author. E-mail: [email protected]

Received: 1 February 1999 Revisions requested: 17 February 1999 Revisions received: 4 March 1999 Accepted: 11 March 1999 Published: 1 June 1999

Conclusions: The large-scale domain or subunit reorientation that is seen in DnaB by electron microscopy might result from the formation of a true twofold symmetric dimer of N-terminal domains, while maintaining a head-to-tail arrangement of C-terminal domains. The N-terminal domain of DnaB is the first region of a hexameric DNA replicative helicase to be visualized at high resolution. Comparison of this structure to the analogous region of the Rho RNA/DNA helicase indicates that the N-terminal domains of these hexameric helicases are structurally dissimilar.

Introduction The DnaB helicase is an essential protein that is involved in both the initiation and elongation stages of Escherichia coli chromosome replication [1]. It is a hexameric ATPase that uses the hydrolysis of ATP to unwind double-stranded DNA [2]. DnaB also contains binding sites for other proteins making it a central component in assembly of the primosome, movement of the replication apparatus and termination of replication [1]. For example, DnaB interacts with its DNA-loading factor DnaC [3,4], with the primase DnaG [5] and with the replication termination protein, RTP [6]. Furthermore, translocation of DnaB is coordinated with the movement of the replication fork through direct contact to the τ-subunit of the pol III holoenzyme; this complex couples the helicase and polymerase activities and appears to stimulate DnaB action as well [7]. The domain structure of DnaB has been examined to locate various functional regions (Figure 1a). Proteolysis of intact DnaB removes the N-terminal 14 residues to yield the 50 kDa Fragment I [8]. Fragment I can be further cleaved into two species: a 33 kDa C-terminal region termed Fragment II and a smaller, N-terminal domain of

Structure June 1999, 7:691–698 http://biomednet.com/elecref/0969212600700691 © Elsevier Science Ltd ISSN 0969-2126

12 kDa called Fragment III [8]. Fragment II contains signature sequences of nucleotide-binding proteins [9] and retains hexamerization, single-stranded DNA-binding and ATPase activities [8,10]. The predicted secondary structure around the nucleotide-binding signature sequences matches the consensus topography predicted for other hexameric helicase ATP-binding domains [11], and is likely to comprise a Walker-type ATP-binding fold. Fragment III has no known activity alone but this domain is essential for DnaB helicase activity, general priming and interaction with DnaC [8,10]. On the basis of negative-stain electron microscopy studies, low-resolution structural models for the DnaB hexamer reveal six bi-lobed protomers arranged in a ring around a channel [12,13]. Fragment II has been proposed to correspond to the larger, central lobe, because it is the larger of the two protease-resistant regions and contains the primary multimerization activity [10,12,13]. The smaller domain farther from the sixfold axis was assumed to be Fragment III [12,13]. DnaB can display either threefold or sixfold symmetry [13] and dimerization of the Fragment III domain is associated with the switch

692

Structure 1999, Vol 7 No 6

Figure 1 (a)

Fragment III Regulatory T

Fragment II ssDNA binding, ATPase T

12 kDa

T

33 kDa

N

C 2º oligomerization 'Hinge'

(b)

1º oligomerization

α1 α2 PPHSIEAEQSVLGGLMLDNERWDDVA • • • 30

40

50

3101 α3 ERVVADDFYTRPHRHIFTEMARLQES • • 60

70

α4 α5 GSPIDLITLAESLERQGQLDSVGGFA • • • 80

90

100

α6 α7 YLAELSKNTPSAANISAYADIVRER • • 110

120

Structure

Organization of DnaB. (a) The primary domain structure of DnaB. The polypeptide chain has been divided into distinct regions by proteolytic mapping [8]. Nomenclature, sizes and positions of these domain boundaries are indicated. (b) Secondary structure of DnaB-III. The sequence of the DnaB N-terminal domain is shown with helices boxed and secondary structure labels assigned. These secondary structural assignments agree well with previous assignments from NMR [41].

[12,13]. A similar distribution of symmetry states has recently been observed for the papillomavirus E1 hexameric helicase [14]. Fragments II and III of DnaB are connected by a 45 amino acid linker, which may facilitate the quaternary structural changes of the enzyme. The rate of protease cleavage at this linker decreases dramatically in the presence of a non-hydrolyzable ATP analog, as compared with ATP or ADP [8], indicating that ATP hydrolysis results in conformational changes affecting accessibility of the hinge. Interestingly, protease treatment to uncouple Fragments II and III leads to a fivefold increase in the ATPase activity of Fragment II without affecting ATP binding ([8], but also see [10]), suggesting that the presence of the N-terminal domain of DnaB can regulate ATP hydrolysis. Despite having enhanced ATPase activity, the hexameric Fragment II shows no helicase activity [10], which further indicates that the Fragment III domain is critical for coupling ATPase activity to enzyme function. Other members of the replication machinery that require the DnaB N-terminal domain for binding may either bind this domain directly or may recognize a quaternary structure of the enzyme that is assumed only when the Fragment III domains are present. Temperature-sensitive mutations in the hinge region of DnaB have been proposed to alter the position of the Fragment III domains or the ability of these domains to change orientation [15]. In a recent investigation of the interaction between DnaB and DnaC by three-dimensional cryoelectron microscopy,

the Fragment III domains in DnaB alone do not appear to protrude radially but instead may lie closer to the oligomer axis stacked above the Fragment II domains [16]. This arrangement is consistent with fluorescence energy transfer experiments that suggest that the Fragment III domains face the 5′ end of the replication fork arm [17]. Positioning of the N-terminal domains atop the ATPase regions is also consistent with the known low-resolution structures of phage T7 hexameric helicase/primase protein [18] and the RuvB branch migration protein [19]; these enzymes also have a pronounced polarity, with both domains of the bilobed protomers circling the hexamer axis. The similarities between DnaB and other hexameric helicase assemblies suggest that these enzymes may share certain mechanistic features [20]. Hexameric ATPases with Walker-type nucleotide-binding regions, however, have been found to play diverse roles in numerous cellular events, such as ATP synthesis via a proton gradient [21], termination of transcription [22] and facilitation of vesicle fusion [23,24]. The molecular motors responsible for each of these functions must be adapted to their particular tasks through structural differences between regions outside the nucleotide-binding motifs. A high-resolution view of DnaB, in particular of regions that distinguish it from other hexameric ATPases, is ultimately necessary to understand the details of its reaction mechanism. The N-terminal domain (Fragment III) is critical to the function of DnaB, regulating the conformational consequences of ATP binding and hydrolysis for both priming of replication and unwinding of DNA. To begin to address how the hexameric ATPase DnaB is customized for its cellular role in DNA replication, we have determined the structure of Fragment III, which we will refer to as DnaB-III, using X-ray crystallography.

Results Structure of the DnaB N-terminal domain

The DnaB N-terminal domain structure was solved by multiple isomorphous replacement and has been refined to 2.3 Å resolution (Figure 2; Table 1). The working R factor is 26.3% and the free R factor is 29.3% (Table 2). The overall dimensions of the DnaB-III domain are 25 Å × 25 Å × 35 Å, which is roughly consistent with the size of the globules at the vertices in the electron microscope (EM) reconstruction of intact DnaB after correction for typical losses in measured volume because of artifacts of negative staining [12]. The DnaB-III structure is largely helical (Figures 2b,c) as previously reported from circular dichroism studies [25]. The domain consists of six α helices, five of which (α2–α6) are wrapped around a central helix (α1). Although the crystallized fragment contains amino acids 15–128 of the DnaB sequence, the first residue for which electron density could be seen was Pro26; continuous density was evident, however, from this residue to the end of the expressed fragment. Similar domain limits have

Research Article The N-terminal domain of the DnaB hexameric helicase Fass, Bogden and Berger

693

Figure 2 DnaB-III monomer structure. (a) Stereogram showing experimental electron density (green) for DnaB-III (Fobs, experimental phases). Contouring is at 1σ. The refined model is shown in ball-and-stick representation (yellow). (b) The structure of the DnaB-III protomer is shown as a stereogram stick representation [42]. Every fifth Cα position is marked by a sphere and every tenth position is labeled. (c) Same view as in (b), but the molecule is shown as a ribbon representation [43] with the secondary structure elements labeled.

been identified previously by proteolysis (15–128) [10] and by NMR (24–136) [25]. A search through the protein fold database using the Dali server [26] revealed no appreciable similarity to any other proteins. The amino acid sequence conservation in the DnaB Fragment III domain is less extensive than in the Fragment II domain. The residues in DnaB-III that are conserved among bacterial and bacteriophage helicases are generally hydrophobic core residues in the structure. Interesting exceptions, however, include Glu33, Tyr60, His64, Asp82,

and Thr85. Glu33 is both buried and invariant among all Fragment III homologs; this residue forms hydrogen bonds to the backbone amide of Tyr60, to the Nε of His28 and to the Nδ of the conserved His64. The aromatic ring of Tyr60 stacks against His28 but the opposite face of the Tyr60 ring is exposed to solvent. The invariant Asp82 is partially buried and uses its sidechain carboxylate to initiate helix α4 by hydrogen bonding to Oγ of Thr85, to its own backbone amide and to the backbone amides of Ile84 and Thr85. Two additional residues, Pro27 and Glu128, are also conserved but lie at the extreme N terminus and

694

Structure 1999, Vol 7 No 6

Table 1

Table 2

Data collection.

Refinement. Native

MeHgCl

Resolution (Å) 3.0 3.5 Rsym* (%) 6.1 12.1 Completeness (%) 96.0 92.7 Riso† (%) – 19.1 Number of sites – 6 Phasing power‡ – 2.08 Cullis R§ – 0.67 Figure of merit (to 3.1 Å) 0.442

TmAc 3.5 11.9 94.6 21.4 4 0.95 0.90

–SeMet Native 2 3.5 7.9 95.2 15.5 8 1.67 0.73

2.3 6.9 98.7 – – – –

*Rsym = Σj|Ij–|/ΣIj, where Ij is the intensity measurement for reflection j and is the mean intensity for multiply recorded reflections. †R iso = Σ||Fph|–|Fp||/ Σ |Fp|, where Fph and Fp are the derivative and native structure factors, respectively. ‡Phasing power = /E, where is the root mean square heavy-atom structure factor and E is the residual lack of closure error. §Cullis R = Σ ||Fph ± Fp|–|Fh,c||/Σ |Fph ± Fp|, where Fh,c is the calculated heavy-atom structure factor.

C terminus of this domain. The reason for their conservation is not readily apparent from the DnaB-III structure. Structure of a DnaB-III dimer

Crystals of DnaB-III grew in the space group P21 with four molecules in the asymmetric unit arranged as a pair of dimers (Figure 3). The two dimer interfaces are identical, with each composed of two antiparallel α helices (α4 and α6) abutting their symmetry-mates such that the twofold rotation axis relating the protomers is parallel to the helical axes. This arrangement forms a four-helix bundle at the dimer interface comprising the following hydrophobic residues and their symmetry-mates: Leu83, Ile84, Ala87, Phe102 and Ala106. A small α helix, α5, bridges α4 and α6 and places Leu96 into the hydrophobic region of this interface. In addition, an ion-bridge is observed across the dimer between the sidechains of Glu88 and Lys′110 (where the prime denotes the dimer-related residue). The sidechain of Glu92 is also hydrogen bonded to the backbone amides of Phe′102 and Ala′103. The surface area buried per dimer is 465 Å2, or 10.5% of the protomer surface area. Though somewhat small, the size of this interface is significantly larger than other packing interfaces in the crystal, with the next largest being only 241 Å2, or 5.5% of the available surface area. Other dimeric proteins are also known to contain small interfaces, such as the 434 cro protein [27], which buries 260 Å2 or 7.4% of its protomer surface area. It can be argued that proteins with such small interfaces dimerize only in response to an inducer such as DNA binding. It is possible, however, that the DnaB-III domain does likewise, dimerizing only at high effective concentrations such as on the surface of the DnaB hexamer.

Discussion Although the ATPase, oligomerization, and singlestranded DNA-binding regions of DnaB are located in

Resolution (Å) Number of reflections working free Rwork/Rfree* (%) Number of atoms protein water rmsd bond lengths (Å) rmsd bond angles (°)

20.0–2.3 17,850 1277 0.263/0.293 3198 43 0.007 1.433

*Rwork, free = Σ ||Fobs|–|Fcalc||/Σ |Fobs|, where the working and free R factors are calculated using the working and free reflection sets, respectively. The free reflections were held aside throughout refinement.

Fragment II, Fragment III is required for DnaB function [8,10]. The activities of the N-terminal domain appear to include regulation of conformational changes important for helicase activity, the regulation of ATP turnover, and coordination of helicase function to that of other replication proteins through protein–protein interactions. The N-terminal domain of DnaB therefore appears to play a specific role in chromosomal replication; this region is structurally unrelated to the analogous portion of the Rho RNA/DNA hexameric helicase and transcription termination factor, which contains a five-stranded β barrel housing an RNA-binding cleft [28–30]. The requirement for the integrity of the DnaB-III domain in vivo derives from studies on the highly homologous DnaB enzyme of Salmonella typhimurium. This protein was screened for dominant-negative mutations [31], which are changes that are likely to generate proteins sufficiently structurally sound to form mixed oligomers with wild-type protein, yet lack key functions and thereby poison the hexameric assembly. Lethal mutations that fall in the DnaB-III region correspond in the E. coli sequence to Asp82→Asn, Ala106→Val and Asn117→Ser. Although these mutations are fairly conservative, they eliminate DnaB function in the cell. Asp82 provides a cap for helix α4 and Asn117 helps initiate helix α6; changes in these two residues appear to affect formation of stable domain secondary structure. Ala106, on the other hand, is partially surface-exposed and does not appear to have a key structural role in the DnaB-III monomer, yet this residue is part of the observed dimer interface, packing against both Phe102 of its own protomer and Ile84 of the dimer-related molecule. Rather than inhibit the folding of the protomer structure, mutations at this position may affect the interaction between DnaB-III protomers. Modeling the Ala106→Val mutation into the DnaB-III dimer structure does not result in gross steric clashes but in fact provides additional hydrophobic interactions across the interface. Such a mutation may exert its effects by altering the equilibrium of DnaB-III domain dimerization, but not necessarily by inhibiting dimerization.

Research Article The N-terminal domain of the DnaB hexameric helicase Fass, Bogden and Berger

695

Figure 3 DnaB-III dimer structure. (a) Structure of the dimer viewed perpendicular to the twofold axis. One protomer is colored blue and the other gold. (b) Structure of the dimer viewed along the twofold axis. Coloring is as in (a). (c) Stereo representation of the dimer interface from the view shown in (b). Segments from one protomer are colored gold and the other blue. Individual residues are labeled. This figure was generated by Ribbons [43].

Analytical ultracentrifugation has shown the isolated DnaB N-terminal domain to be monomeric at a concentration of 0.2 mM [25]. In the intact DnaB hexamer, however, these domains have been observed to dimerize

[12,13]. Biochemical studies have further suggested that the presence of the DnaB-III domain affects the overall stability or conformation of the hexamer [10]. Thus, in the context of the hexamer, the effective concentration of

696

Structure 1999, Vol 7 No 6

Figure 4 Illustration of one possible rotation to transform the DnaB-III domain from a C6-symmetrical to a C3-symmetrical state containing the observed DnaB-III dimer. (a) Approximate positions for a C6-symmetrical configuration of DnaB-III domains. Each domain is rotated 60° with respect to its neighbor. (b) A combined rotation of 120° per N-terminal domain pair would generate a C3-symmetrical state in which the N-terminal domains oligomerize about a local twofold axis while the C-terminal domains retain their head-to-tail packing arrangement. The rotation of the DnaB-III domains need not occur around an axis parallel to the hexamer axis, and other rotations leading to alternate orientations for the local twofold axis are possible. Purple and orange ovals represent relative locations of the ATP-binding sites, based on other hexameric ATPases [21,23,24]. Neighboring sites in DnaB are non-equivalent for ATP binding and hydrolysis, as indicated by the two colors. Biochemical studies on DnaB [44] and other hexameric helicases [45,46] have demonstrated that only three high-affinity ATP-binding sites exist per hexamer.

DnaB-III domains may be sufficiently high to promote dimerization. Indeed, it is possible that the affinity of this interface is in the range where its formation both affects and is affected by the quaternary state of the enzyme. Interestingly, the homologous DnaB-III domain from a thermophilic bacterium forms a dimer at room temperature in solution (D Wigley, personal communication), suggesting that the monomer–dimer equilibrium is shifted in this organism to compensate for the conditions under which replication of its DNA occurs. According to this model, mutations that alter the affinity of dimerization such as perhaps the Ala106→Val alteration described above, would be expected to have significant effects on enzyme activity. If the quaternary arrangement of DnaB oscillates between C6-symmetric and C3-symmetric states via dimerization of the N-terminal domains according to the true twofold symmetry that we observe in the crystals, dimerization must occur through large-scale conformational changes. Such structural variations have been observed by electron microscopy [13]. In a C6-symmetrical state each N-terminal domain would be rotated 60° from its neighbor in the hexamer. Given that DnaB-III dimers are related by a rotation of 180°, the N-terminal domains must undergo rotations that total at least 120° per pair relative to the

C-terminal domains (Figure 4). It is important to note that no evidence for D3 symmetry can be seen in electron micrographs of DnaB [12,13]. The resolution of the electron microscopy is insufficient to resolve the precise orientation of individual N-terminal domains. Thus, the C3 symmetry evident in the projections probably arises from the maintenance of head-to-tail packing by the ATPase regions whereas the N-terminal domains may form a true dimer about a local twofold axis. The linker between the DnaB-III domain and the ATPase domain of DnaB would be expected to accommodate these motions [12,13]. Large-scale domain rearrangements have a precedent in helicase structures: for the E. coli Rep helicase, two copies of the protomer in the asymmetric unit of the crystals differ by a swiveling of one of the Rep domains by 130° about a hinge region [32]. It remains to be seen whether the conformational variability observed in dimeric helicases such as Rep and hexameric helicases such as DnaB reflect any common features in their mechanism of translocation along the DNA or their duplex-unwinding activity.

Biological implications The DnaB protein is the hexameric helicase responsible for unwinding the DNA duplex during replication of the Escherichia coli chromosome. The enzyme is a molecular

Research Article The N-terminal domain of the DnaB hexameric helicase Fass, Bogden and Berger

motor that can undergo large-scale quaternary structural changes as visualized by electron microscopy. Although several critical aspects of DnaB function have been addressed biochemically and via electron microscopy, a thorough understanding of the mechanism by which DnaB acts on DNA will also rely on atomic-resolution structural information. We report an X-ray crystallographic study of the N-terminal domain of the DnaB helicase. This domain is essential for helicase activity, priming of replication, and interaction with other factors in the replisome. The structure of this domain is different from that of the N-terminal domain of the Rho hexameric DNA/RNA helicase, illustrating that the enzymatic activity of hexameric ATPases is customized to particular cellular tasks through structural differences in these auxiliary domains. The DnaB N-terminal domain plays a central role in the conformational changes of the enzyme: adjacent N-terminal domains appear to coalesce into dimers when the helicase switches from a sixfold to a threefold symmetrical form. We observe two copies of a DnaB N-terminal domain dimer in our crystals, providing a model for how these domains may pack in the threefold symmetric state of the intact enzyme.

Materials and methods Protein production and crystallization The DnaB-III coding sequence (spanning residues 15–118 of the wildtype E. coli DnaB protein) was amplified by PCR from a plasmid containing the intact DnaB gene (CEB and JMB, unpublished observations) and cloned into pET28b (Novagen). Five residues, MASHM (in single-letter amino acid code), were inserted at the N terminus during cloning. Protein expression was performed in BL21(DE3) pLysS cells by inducing at an A600 of 0.3 with 1 mM IPTG for 2.5 h at 37°C. DnaB-III containing selenomethionine was expressed in BL21(DE3) cells as described [33] except that cells were induced at an A600 of 0.3 and harvested 3 h post-induction. In both cases, harvested cells were resuspended at 1 g/ml in buffer A (50 mM Tris-HCl, pH 7.5; 1 mM EDTA; 1 mM EGTA; 10% glycerol; 1 mM DTT; 1 mM phenyl-methylsulfonylfluoride (PMSF); 1 µM pepstatin-A; 1 µM leupeptin) and flashfrozen by dropwise pipetting into liquid nitrogen. DnaB-III was purified as follows: cells were thawed in a room temperature water bath, sonicated and centrifuged. The clarified lysate was ammonium sulfate precipitated (45–65%), centrifuged, and the pellet resuspended in buffer A to a conductance equivalent to buffer A + 50 mM KCl. The protein was passed over a 10 ml HQ50 POROS column (Perseptive BioSystems), washed with buffer A + 50 mM KCl, and eluted with a 50–500 mM KCl gradient in buffer A. A significant fraction of the DnaB-III protein was present in the HQ50 flow-through and wash; these fractions were reapplied and eluted from the column as above. Peak fractions were determined by SDS–PAGE, pooled, and concentrated in a Centriprep-3 (Amicon) to 1.5 ml. The solution was passed over a Sephacryl S300 gel-filtration column (Pharmacia) equilibrated in buffer A + 300 mM KCl. Peaks fractions were pooled, the purified protein concentrated to 18 mg/ml (as assayed by UV absorbance [34]), and then dialyzed against 10 mM HEPES-KOH, pH 7.5; 50 mM NaCl. Crystals were grown using the hanging-drop method by mixing protein 1:1 with a crystallization solution consisting of 20% PEG 8000; 50 mM cacodylate, pH 6.9; 25 mM MgCl2; 10% ethanol, and equilibrating at 20°C against a reservoir solution of 16% PEG 8000; 40 mM

697

cacodylate, pH 6.9; 20 mM MgCl2, 8% ethanol. Thin plates (~150 × 150 × 20 µm) of space group P21 and unit cell a = 35.3 Å, b = 66.4 Å, c = 107.0 Å, α = γ = 90°, β = 93.8° grew in 2–3 weeks.

Data collection and phasing Crystals were equilibrated in reservoir solution plus increasing amounts of glycerol (5%, 10 min; 10%, 10 min; 15%, 5 min) before flash-freezing in a gaseous N2 stream at –155°C. Initial data were collected on Brookhaven beamline X4A at wavelengths of 0.9793, 0.9789 and 0.9667 Å. An attempt was made during this time to use these data for multiwavelength anomalous dispersion (MAD) experiments, but the data showed no strong dispersive signal as evidenced from Patterson analysis. We have attributed this to severe anisotropy readily apparent in the diffraction patterns and to the fact that the diffraction decayed visibly during the extended course of data collection to obtain multiple P21 data sets. Derivatives were therefore produced by soaking heavy metals into the crystals (methyl mercuric chloride, 1 mM, 3 h; thulium acetate, 0.1 mM, 48 h). Derivative data, including a data set from a crystal containing selenomethionine protein, were collected at 1.54 Å on a Rigaku rotating anode RU-300 equipped with an Raxis IV detector and an X-stream cryo-cooler (MSC). Data indexing and reduction were carried out using the program DENZO/SCALEPACK [35]. Data truncation and scaling between data sets were performed with the CCP4 program suite [36]. Electron-density maps were generated by multiple isomorphous replacement (MIR) using the program SHARP [37]. The positions of selenium atoms were determined from difference Fourier maps using MIR phases from the other two derivatives as well as from using difference Patterson maps between native and selenomethionine data. Anomalous difference Patterson maps from the MAD data did show peaks at a subset of the appropriate positions, as was subsequently verified using MIR methods. Selenium and thulium heavy-atom positions were used to determine the noncrystallographic symmetry (NCS) operators relating the four molecules in the asymmetric unit, and NCS operators were refined using IMP [38]. Electron-density maps were improved by solvent flattening and NCS-averaging with the program dm [39]. X-PLOR [40] was used for refinement, and anisotropic and bulk-solvent corrections were applied.

Accession numbers The coordinates have been deposited in the Protein Data Bank with the accession code 1b79.

Acknowledgements The authors thank S Gamblin for advice on data collection and anisotropic corrections, MD Nichols, M Botchan, J Keck and S Lynch for thoughtful discussions and critical reading of the manuscript, C Ogata for assistance at Brookhaven beamline X4A, and the Harold G and Leila Y Mathers Charitable Foundation, Whitehead Institute and Berkeley MCB department for funding assistance.

References 1. Kornberg, A. & Baker, T.A. (1992). In DNA Replication, second edn, Freeman, San Francisco, California, USA. 2. LeBowitz, J.H. & McMacken, R. (1986). The Escherichia coli DnaB replication protein is a DNA helicase. J. Biol. Chem. 261, 4738-4748. 3. Wahle, E., Lasken, R.S. & Kornberg, A. (1989). The dnaB-dnaC replication protein complex of Escherichia coli, I. Formation and properties. J. Biol. Chem. 264, 2463-2468. 4. Wickner, S. & Hurwitz, J. (1975). Interaction of Escherichia coli dnaB and dnaC(D) gene products in vitro. Proc. Natl Acad. Sci. USA 72, 921-925. 5. Lu, Y-B., Ratnakar, P.V.A.L., Mohanty, B.K. & Bastia, D. (1996). Direct physical interaction between DnaG primase and DnaB helicase of Escherichia coli is necessary for optimal synthesis of primer RNA. Proc. Natl Acad. Sci. USA 93, 12902-12907. 6. Manna, A.C., Pai, K.S., Bussiere, D.E., Davies, C., White, S.W. & Bastia, D. (1996). Helicase-contrahelicase interaction and the mechanism of termination of DNA replication. Cell 87, 881-891. 7. Kim, S., Dallman, H.C., McHenry, C.S. & Marians, K.J. (1996). Coupling of a replicative polymerase and helicase: a tau-DnaB interaction mediates rapid replication fork movement. Cell 84, 643-650.

698

Structure 1999, Vol 7 No 6

8. Nakayama, N., Arai, N., Kaziro, Y. & Arai, K-I. (1984). Structural and functional studies of the Dnab protein using limited proteolysis. J. Biol. Chem. 259, 88-96. 9. Biswas, S.B. & Biswas, E.E. (1987). Regulation of dnaB function in DNA replication in Escherichia coli by DnaC and lambda P gene products. J. Biol. Chem. 262, 7831-7838. 10. Biswas, S.B., Chen, P.-H. & Biswas, E.E. (1994). Structure and function of Escherichia coli DnaB protein: role of the N-terminal domain in helicase activity. Biochemistry 33, 11307-11314. 11. Bradley, M.K., Smith, T.F., Lathrop, R.H., Livingston, D.M. & Webster, T.A. (1987). Consensus topography in the ATP binding site of the simian virus 40 and polyomavirus large tumor antigens. Proc. Natl Acad. Sci. USA 84, 4026-4030. 12. San Martin, C., Stamford, N.P.J., Dammerova, N. & Dixon, N.E. (1995). A structural model for the Escherichia coli DnaB helicase based on electron microscopy data. J. Struct. Biol. 114, 167-176. 13. Yu, X., Jezewska, M.J., Bujalowski, W. & Egelman, E.H. (1996). The hexameric E. coli DnaB helicase can exist in different quaternary states. J. Mol. Biol. 259, 7-14. 14. Fouts, E., Yu, X., Egelman, E.H. & Botchan, M.R. (1999). Biochemical and electron microscopic image analysis of the hexameric E1 helicase. J. Biol. Chem., 274, 4447-4458. 15. Stordal, L. & Maurer, R. (1996). Defect in general priming conferred by linker region mutants of Escherichia coli DnaB. J. Bacteriol. 178, 4620-4627. 16. San Martin, C., Radermacher, M., Wolpensinger, B., Engel, A., Miles, C.S., Dixon, N.E. & Carazo, J-M. (1998). Three-dimensional reconstructions from cryoelectron microscopy images reveal an intimate complex between helicase DnaB and its loading partner DnaC. Structure 6, 501-509. 17. Jezewska, M.J., Rajendran, S. & Bujalowski, W. (1998). Complex of Escherichia coli primary replicative helicase DnaB protein with a replication fork: recognition and structure. Biochemistry 37, 3116-3136. 18. Egelman, E.H., Yu, X., Wild, R., Hingorami, M.H. & Patel, S.S. (1995). Bacteriophage T7 helicase/primase proteins form rings around singlestranded DNA that suggest a general structure for hexameric helicases. Proc. Natl Acad. Sci. USA 92, 3869-3873. 19. Stasiak, A., Tsaneva, I.R., West, S.C., Benson, C.J.B., Yu, X. & Egelman, E.H. (1994). The Escherichia coli RuvB branch migration protein forms double hexameric rings around DNA. Proc. Natl Acad. Sci. USA 91, 7618-7622. 20. West, S.C. (1996). DNA helicases: new breeds of translocating motors and molecular pumps. Cell 86, 177-180. 21. Abrahams, J.P., Leslie, A.G.W., Lutter, R. & Walker, J.E. (1994). Structure at 2.8 Å resolution of F1-ATPase from bovine heart mitochondria. Nature 370, 621-628. 22. Richardson, J.P. (1996). Structural organization of transcription termination factor Rho. J. Biol. Chem. 271, 1251-1254. 23. Lenzen, C.U., Steinmann, D., Whiteheart, S.W. & Weis, W.I. (1998). Crystal structure of the hexamerization domain of N-ethylmaleimidesensitive fusion protein. Cell 94, 525-536. 24. Yu, R.C., Hanson, P.I., Jahn, R., & Brünger, A.T. (1998). Structure of the ATP-dependent oligomerization domain of N-ethylmaleimide sensitive factor complexed with ATP. Nat. Struct. Biol. 5, 803-811. 25. Miles, C.S., Weigelt, J., Stamford, N.P.J., Dammerova, N., Otting, G. & Dixon, N.E. (1997). Precise limits of the N-terminal domain of DnaB helicase determined by NMR spectroscopy. Biochem. Biophys. Res. Commun. 231, 126-130. 26. Holm, L.L. & Sander, C. (1996). Mapping the protein universe. Science 273, 595-602. 27. Mondragón, A. & Harrison, S.C. (1991). The phage 434 Cro/OR1 complex at 2.5 Å resolution. J. Mol. Biol. 219, 321-334. 28. Briercheck, D.M., Wood, T.C., Allison, T.J., Richardson, J.P. & Rule, G.S. (1998). The NMR structure of the RNA binding domain of E. coli Rho factor suggests possible RNA-protein interactions. Nat. Struct. Biol. 5, 393-399. 29. Allison, T.J., Wood, T.C., Briercheck, D.M., Rastinejad, F., Richardson, J.P. & Rule, G.S. (1998). Crystal structure of the RNA-binding domain from transcription termination factor Rho. Nat. Struct. Biol. 5, 352-356. 30. Bogden, C.E., Fass, D., Bergman, N., Nichols, M. & Berger, J.M. (1999). Structural basis for terminator recognition by the Rho transcription termination factor. Molecular Cell, 3, 487-493. 31. Maurer, R. & Wong, A. (1988). Dominant lethal mutations in the dnab helicase gene of Salmonella typhimurium. J. Bacteriol. 170, 3682-3688.

32. Korolev, S., Hsieh, J., Gauss, G.H., Lohman, T.M. & Waksman, G. (1997). Major domain swiveling revealed by the crystal structures of complexes of E. coli rep helicase bound to single-stranded DNA and ADP. Cell 90, 635-647. 33. Van Duyne, G.D., Standaert, R.F., Karplus, A.P., Schreiber, S.L., & Clardy, J. (1993). Atomic structures of the human immunophilin FKBP12 complexes with FK506 and rapamycin. J. Mol. Biol. 229, 105-124. 34. Edelhoch, H. (1967). Spectroscopic determination of tryptophan and tyrosine in proteins. Biochemistry 6, 1948-1954. 35. Otwinowski, Z. & Minor, W. (1997). Processing of X-ray diffraction data collected in oscillation mode. In Methods in Enzymology, Macromolecular Crystallography, A, (C.W.J. Carter, and Sweet, R.M., eds), pp.472-494, Academic Press, Boston, Massachusetts, USA. 36. CCP4 (1994). Collaborative computational project number 4. The CCP4 suite programs for protein crystallography. Acta Crystallogr. D 50, 760-763. 37. de la Fortelle, E. & Bricogne, G. (1997). Maximum-likelihood heavyatom parameter refinement for multiple isomorphous replacement and multiwavelength anomalous diffraction methods. In Methods in Enzymology, Macromolecular Crystallography, (C.W.J. Carter, and Sweet, R.M., eds), pp472-494, Academic Press, Boston, Massachusetts, USA. 38. Jones, T.A. (1992). A, yaap, asap, @#*? A set of averaging programs. In Molecular Replacement, (E.J. Dodson, Gover, S., and Wolf, W., eds) pp. 91-105, Warrington: SERC Daresbury Laboratory. 39. Cowtan, K. (1994). A CCP4 density modification package. Joint CCP4 and ESF-EACBM Newsletter on Protein Crystallography. 31, 34-38. 40. Brünger, A.T. (1993). XPLOR (Version 3.1): A System for X-ray Crystallography and NMR. Yale University Press, USA. 41. Weigelt, J., Miles, C.S., Dixon, N.E., & Otting, G. (1998). Backbone NMR assignments and secondary structure of the N-terminal domain of DnaB helicase from E. coli. J. Biomol. NMR 11, 233-234. 42. Kraulis, P. (1991). MOLSCRIPT: a program to produce both detailed and schematic plots of protein structures. J. Appl. Crystallogr. 24, 946-950. 43. Carson, M. (1991). Ribbons 2.0. J. Appl. Crystallogr. 24, 958-961. 44. Biswas, E.E., Biswas, S.B. & Bishop, J.E. (1986). The dnaB protein of Escherichia coli: mechanism of nucleotide binding, hydrolysis, and modulation by dnaC protein. Biochemistry 25, 7368-7374. 45. Hingorani, M.M. & Patel, S.S. (1996). Cooperative interaction of nucleotide ligands are linked to oligomerization and DNA binding in bacteriophage T7 gene 4 helicases. Biochemistry 35, 2218-2228. 46. Stitt, B.L. (1988). Escherichia coli transcription termination protein Rho has three hydrolytic sites for ATP. J. Biol. Chem. 263, 11130-11137.

Because Structure with Folding & Design operates a ‘Continuous Publication System’ for Research Papers, this paper has been published on the internet before being printed (accessed from http://biomednet.com/cbiology/str). For further information, see the explanation on the contents page.