Crystal Structure of Full Length Topoisomerase I from Thermotoga maritima

doi:10.1016/j.jmb.2006.03.012

J. Mol. Biol. (2006) 358, 1328–1340

Crystal Structure of Full Length Topoisomerase I from Thermotoga maritima Guido Hansen1,2, Axel Harrenga1*, Bernd Wieland1 Dietmar Schomburg2* and Peter Reinemer1* 1

Bayer HealthCare AG, Pharma R&D Europe, Enabling Technologies, D-42096 Wuppertal Germany 2 Institute of Biochemistry University of Cologne, Zu¨lpicher Strasse 47, D-50674 Ko¨ln Germany

DNA topoisomerases are a family of enzymes altering the topology of DNA by concerted breakage and rejoining of the phosphodiester backbone of DNA. Bacterial and archeal type IA topoisomerases, including topoisomerase I, topoisomerase III, and reverse gyrase, are crucial in regulation of DNA supercoiling and maintenance of genetic stability. The crystal structure of full length topoisomerase I from Thermotoga ˚ resolution and represents an intact and maritima was determined at 1.7 A fully active bacterial topoisomerase I. It reveals the torus-like structure of the conserved transesterification core domain comprising domains I–IV and a tightly associated C-terminal zinc ribbon domain (domain V) packing against domain IV of the core domain. The previously established zinc-independence of the functional activity of T. maritima topoisomerase I is further supported by its crystal structure as no zinc ion is bound to domain V. However, the structural integrity is preserved by the formation of two disulfide bridges between the four Znbinding cysteine residues. A functional role of domain V in DNA binding and recognition is suggested and discussed in the light of the structure and previous biochemical findings. In addition, implications for bacterial topoisomerases I are provided. q 2006 Elsevier Ltd. All rights reserved.

*Corresponding authors

Keywords: crystal structure; topoisomerase I; Thermotoga maritima; zinc ribbon

Introduction DNA topoisomerases are a family of ubiquitous enzymes known to manage the topological state of DNA in cells.1 They function by transiently breaking one or two strands of DNA, passing single or double-stranded DNA through the resulting gap and finally resealing the gap. Topoisomerases are crucial in a variety of cellular processes such as replication, transcription and recombination. Present addresses: P. Reinemer, Proteros Biostructures GmbH, Am Klopferspitz 19, D-82152 Martinsried, Germany; G. Hansen, Biota Structural Biology Laboratory, St. Vincent’s Institute of Medical Research, 9 Princess Street, Fitzroy, Vic. 3065, Australia; A. Harrenga, Bayer HealthCare AG, PH-R&D-DRE-PRRBCB, 42096 Wuppertal, Germany; B. Wieland, Bayer HealthCare AG, PH-R&D-DRE-TR, 42096 Wuppertal, Germany. E-mail addresses of the corresponding authors: [email protected]; d.schomburg@ unikoeln.de; [email protected]

A number of different enzymes has evolved and based on their biochemical properties, topoisomerases have been classified into type II topoisomerases, which break both strands of a DNA duplex in concert, and type I topoisomerases, which only break one strand of DNA. Type I topoisomerases can be further classified into two subfamilies, namely type IA enzymes, which form a transient covalent bond with the 5 0 -end of the cleaved DNA, and type IB enzymes forming a transient 3 0 -end covalent bond. While bacterial type II topoisomerases are well established drug targets in antibacterial therapy, bacterial type I topoisomerases have only recently emerged as potential new targets for therapy.2,3 Initial structural insight to the topoisomerase IA family was gained from crystal structures of a 67 kDa N-terminal fragment of Escherichia coli topoisomerase I,4 an E. coli topoisomerase III5 and a reverse gyrase from a hyperthermophilic archaeon6 and has been a major breakthrough in understanding of this enzyme class: the structures revealed the presence of a common core element,

0022-2836/$ - see front matter q 2006 Elsevier Ltd. All rights reserved.

1329

Structure of T. maritima Topoisomerase I

comprising four protein domains, which is responsible for DNA cleavage and strand passage. The four domains are situated around a central hole where domains I and IV, comprising the base of the structural element, are linked to domain III via an extended, arch-like domain II. The conformation of the core element as observed in the crystal structures is characterized by intimate interactions between domains I and III. The current “enzymebridged” model of the enzymatic reaction7–9 requires drastic conformational changes of the protein in the course of the catalytic cycle including the complete dissociation of domain III from domains I and IV. The active site of the enzyme is situated in domain I and III, and contains several conserved residues including an essential tyrosine. In addition, crystal structures of E. coli topoisomerases I and III in complex with short oligonucleotides have established the presence of a DNA binding cleft formed by residues of domain I and IV at the base of the protein.10,11 Despite the knowledge derived from the crystal structure analysis of E. coli topoisomerase I, the truncated protein of E. coli topoisomerase I employed for structure determination 4,11 only contained the well conserved N-terminal region of the protein, comprising the four domain central core element, but lacked the C-terminal domain harbouring putative zinc binding motifs. The C-terminal region of bacterial topoisomerase I varies in length and sequence with up to five copies of zinc binding motifs present in this region.12 Although the truncated E. coli topoisomerase I is able to cleave single-stranded DNA, the relaxation of negatively supercoiled duplex DNA is impaired.13 The zinc binding motifs have been implicated in substrate binding, strand passage or protein–protein interactions in vivo14–16 although their precise function is largely unknown. Up to now, no crystal structure of a full-length bacterial topoisomerase I comprising the core domain and the zinc binding motifs has been reported. In order to support drug discovery efforts on topoisomerase I, we have determined the crystal structure of the fully functional topoisomerase I from Thermotoga maritima. Topoisomerase I from T. maritima, a hyperthermophilic eubacterium with an optimal growth temperature of 80 8C, has been identified as a member of the type IA topoisomerase family.17,18 Both T. maritima topoisomerase I and E. coli topoisomerase I are able to bind and cleave singlestranded DNA with a common cleavage preference for a cytosine in position K4 of the cleavage point.18,19 During the catalytic cycle both enzymes alter the linkage number of DNA substrates in steps of one, indicating a common enzyme bridged, strand-passage mechanism.20 In contrast, T. maritima topoisomerase I has an exceptional high DNA relaxation activity, which is at least 100-fold higher compared to E. coli topoisomerase I.18 Moreover, the C terminus of the enzyme contains only a single zinc binding motif18 and truncation of the

whole C-terminal domain results in greatly reduced DNA relaxation activity, probably caused by inefficient substrate binding.21 However, in contrast, mutations within the zinc binding motif, which lead to a loss of the zinc binding ability, do not affect DNA cleavage, relaxation or decatenation activity of T. maritima topoisomerase I.18 In E. coli topoisomerase I, zinc depletion inactivates the enzyme22 and a point mutation within the second zinc binding motif alters the cleavage efficiency and specificity.16 In addition, truncated E. coli topoisomerase I, comprising solely the core domain, cannot relax negatively supercoiled DNA at all.13 Here we report the crystal structure of full length ˚ resolution topoisomerase I from T. maritima at 1.7 A representing an intact and fully active bacterial topoisomerase I. In particular, we describe here structure and spatial orientation of the C-terminal zinc binding repeat, a structural motif conserved in most bacterial topoisomerases I, and attempt to relate this to its functional role.

Results and Discussion Overall structure of T. maritima topoisomerase I T. maritima topoisomerase I shares the typical topoisomerase type IA fold with the known members of this family. Structures of E. coli topoisomerase I (40% sequence identity) and E. coli topoisomerase III (18% sequence identity) can be superimposed on T. maritima topoisomerase ˚ and 4.9 A ˚ using the I with RMS deviations of 1.5 A Ca atoms of domains I to IV, respectively. The overall structure of T. maritima topoisomerase I (Figure 1) comprises five different protein domains. It exhibits electron density for most of the amino acid residues except for six N-terminal residues (residues 1–6), a loop at the surface of the protein (residues 319–332) and the 32 C-terminal residues (residues 602–633). The domains I–IV (residues 1–542), constituting the protein core, share high structural similarity with cleavage/strand passage domains of other type IA topoisomerases, namely the 67 kDa N-terminal fragment of E. coli topoisomerase I (residues 1–581) and the cleavage/strand passage domain of E. coli topoisomerase III (residues 1–609). Similar to the E. coli structures domains I–IV in T. maritima topoisomerase I are arranged to form a ˚ !60 A ˚ !45 A ˚ with a hole torus of dimensions 95 A ˚ of approximately 25 A diameter in the centre. The torus is thought to accommodate single or doublestranded DNA after strand passage.4,5 Although T. maritima topoisomerase I is efficient in decatenating nicked minicircles present in kinetoplastic DNA18 it lacks a so-called “decatenation loop” in proximity to the central hole that is necessary for efficient decatenation of various substrates by E. coli topoisomerase III.5,23 The mechanism of decatenation in T. maritima topoisomerase I is therefore likely

1330

Structure of T. maritima Topoisomerase I

(a) H

II 8

9

10 5

12

6

239 375

443

L K

III K'

185

N I

IV

O

J

19

129

601

C

E

I

13

G 3

V

20

14

542

15

D 16

N

7

1

P F

2

Q C

17

A 18

R

(b)

Figure 1. Overall structure of topoisomerase I from T. maritima. Domains are colour-coded: domain I (yellow), II (green), III (red), IV (blue) and V (purple). For simplification in domains I–IV only b-strands of more than two residues length are indicated. (a) Topology cartoon of the structure. Cysteine residues 559, 561, 578 and 580 in domain V are indicated by circles. (b) Stereo representation of the molecule. To enable easy orientation secondary structure elements are labelled according to the E. coli topoisomerase I scheme.4 Helices are represented by letters A–R and b-sheets by numbers 1–20.

Structure of T. maritima Topoisomerase I

to be different from E. coli topoisomerase III and based on other structural properties. A flexible loop connects the core with the globular C-terminal domain V (residues 543–601) ˚ !20 A ˚ !15 A ˚ which includes of dimensions 25 A the zinc binding signature motif (Figure 1). It is positioned adjacent to domain IV, solely forming contacts with this domain. A loop element of ten residues length (residues 543–552) originating from helix aR of domain IV, connects the core domain of the enzyme with a four-stranded antiparallel b-sheet, which is the major secondary structure element of domain V. It comprises the strands b15 (residues 553–558), b16 (residues 564–570), b17 (residues 573–577), and b18 (residues 584–585). The strands b15/b16 and b17/b18 are connected by two loop elements (residues 559–563 and 578– 583). Each loop harbours two cysteine residues arranged in two C-X-C elements. A b-turn motif connects the strands b16 and b17 and is stabilized by two hydrogen bonds formed between the mainchain carbonyl and amino groups of residues Gly570 and Gly573. These structural elements together form a Cys-X-Cys-X16-Cys-X-Cys zincfinger motif that closely resembles the overall structure of other known zinc-finger motifs (see below). Following strand b18 a loop element (residues 586–592) reaches back towards the core domain of the protein and positions a second twostranded b-sheet (residues 593–600) in close proximity to helices aG and aP of domain IV (Figure 1). The second sheet comprises strands b19 (residues 593–595) and b20 (residues 598–600), which are connected by a type I b-hairpin. Domain V is tightly packed to the core domain of the enzyme forming a ˚ 2 with domain IV contact area of about 1300 A calculated from solvent-exposed surfaces using a ˚ . The contact surface area probe radius of 1.4 A between domains IV and V comprises about 25% of the total surface area of domain V. It is mainly characterized by hydrophobic interactions despite two prominent salt-bridges formed between Asp547/Lys501 and Arg185/Asp595 and a number of hydrogen bonds formed between residues Asn588, Gln590, Ala592, and Ile594 of domain V and Val490, Ile489, Pro487, and Arg181 of domain IV. The two-stranded b-sheet of domain V also exhibits good shape complementarity to a cavity formed by helices aG and aP of domain IV and thereby also mediates the tight association of domain V with domain IV. Taken together, these structural features suggest that domain V is fixed at its position with only limited flexibility concerning its orientation towards the core domain of the enzyme. Domain V harbours a conserved zinc ribbon motif The four-stranded b-sheet of domain V (residues 553–585) closely resembles a zinc binding motif known as zinc ribbon.24 It contains two noncanonical motifs named zinc knuckles (residues

1331 559–563 and 578–583) residing in the loops connecting the strands b15/b16 (primary hairpin) and b17/ b18 (secondary hairpin), which are arranged almost perpendicular to each other (Figure 3). As observed in many other members of the family the zinc knuckle consensus sequence CPXCG25 is not fully conserved in T. maritima topoisomerase I, exhibiting sequences CSCG and CECG for the primary and secondary hairpin, respectively. The structural element exhibits strong compliance with the zinc ribbon fold as indicated by RMSD values in the ˚ (human TFIIB,26 PDB code 1DL6) to range of 1.3 A ˚ (human TFIIS,24 PDB code 1TFI) for Ca atoms 1.8 A when compared to members of the zinc ribbon family originating from archea, eubacteria and eukaryotes (Figure 2)25. Preceding work using biochemical assays and atomic absorption spectroscopy has revealed that zinc is bound in stoichiometric amounts to T. maritima topoisomerase I.18 Replacement of the cysteine residues in the primary zinc knuckle (Cys559/Cys561) by alanine or truncation of the C-terminal part of the enzyme including the zinc binding motif both resulted in a loss of the zinc binding capability of the enzyme.18,21 These findings strongly suggest that the zinc binding site located in the C-terminal domain V is able to coordinate one zinc ion. However, the zinc ribbon domain as observed in the present crystal structure does not contain a zinc ion and both pairs of cysteine residues (Cys559/Cys578 and Cys561/ Cys580) are connected by disulfide bonds, with ˚ . This finding is sulfur–sulfur distances of 2 A clearly supported by an electron density map with 2FoKFc coefficients calculated from a “simulated annealing omit model” where the cysteine residues ˚ distance have and surrounding atoms within 5 A been removed from the model, which reveals electron density clearly accounted for by two disulfide bridges (Figure 3). In addition, an anomalous difference Fourier map based on X-ray data collected at the zinc K absorption edge did not show any significant, interpretable difference Fourier peak (data not shown). It has been previously demonstrated, that zinc finger motifs are highly sensitive to oxidation and reduction processes.27,28 Oxidation of cysteine residues within the zinc binding motifs and loss of the zinc ion has been observed in several proteins (DNA primase;29 ribonucleotide reductase;30 heat shock protein 33;31,32 anti sigma factor RsrA33). In the light of these findings, the formation of disulfide bonds in the zinc ribbon motif of T. maritima topoisomerase I may be explained by oxidation of the thiol groups by aerial oxygen. It is most likely that the zinc ion was lost during crystallization and not during purification, as it has been shown that T. maritima topoisomerase I purified under similar conditions coordinates one zinc molecule.18 Furthermore a strong fluorescence signal detected in the absorption edge scan prior to data collection at the synchrotron strongly indicates the presence of zinc ions in the crystal mother liquor.

1332

Structure of T. maritima Topoisomerase I

Figure 2. Cartoon representation of the zinc ribbon motif of domain V of T. maritima topoisomerase I and comparison with other zinc ribbon motifs: T. maritima topoisomerase I (present study), human TFIIS,26 Pyrococcus furiosus TFIIB,60 T. celer RPB961 and human TFIIB.28 Disulfide bonds in T. maritima topoisomerase I are indicated as sticks and zinc ions are depicted by spheres.

The structural integrity of domain V is crucial for full functional activity In most proteins containing zinc finger motifs zinc ions are thought to play a crucial role in

maintaining the structure of those elements.25 However, studies on small mitochondrial Tim (translocase of inner membrane) proteins, which contain two conserved CX3C motifs34 that coordinate a zinc ion,35 have shown that recombinant and

Figure 3. Stereo representation of the formation of disulfide bonds in domain V of T. maritima topoisomerase I. Juxtaposed cysteine residues 559/578 and 561/580 residing within the primary and secondary hairpin of the zinc ribbon motif form disulfide bonds. Amino acid residues are shown as sticks. A 2jFojKjFcj simulated annealing omit map contoured at 1 sigma is superimposed and confirms the formation of two disulfide bonds. For the calculation of the map, ˚ diameter have been omitted from the structural model used to calculate the cysteine residues and all atoms within a 5 A the jFcj components.

Structure of T. maritima Topoisomerase I

endogenous Tim proteins could form disulfide bonds between structurally juxtaposed cysteine residues without loss of function.36,37 The current model for the assembly of small Tim proteins assumes that the structure in vicinity of the zinc binding motif could be either stabilized by zinc coordination or disulfide bond formation.38 Likewise, although lacking a zinc ion, the disulfide bonds in T. maritima topoisomerase I connecting primary and secondary hairpins have a stabilizing effect on the motif, in particular with regard to fixing the positions of the b-hairpins in respect to each other. Furthermore, low RMS deviations for T. maritima topoisomerase I and zinc ribbon structures with coordinated zinc molecules indicate that the fold of the domain V is fully maintained even in the absence of zinc. Accordingly, since a bound zinc ion is not essential for functional DNA relaxation activity18 it is evident that structural integrity of domain V is crucial for full functional activity of the enzyme, rather than the presence of the zinc ion. Active site and DNA binding and recognition site The structure of T. maritima topoisomerase I presents the enzyme in its closed conformation, which is characterized by intimate contacts between domains I and III. The active site is located at the interface of domains I and III (Figure 4) and is inaccessible in this conformation of the enzyme. An important difference between T. maritima topoisomerase I and E. coli topoisomerase I is the presence of an eight amino acid insertion in proximity of the active site,17 resulting in an additional loop localized between helix aJ and aK 0 (residues 276–283). Although such insertions in proximity of the active site have been shown to strongly influence DNA relaxation activity,39–41 it is

1333 difficult to establish the functional significance of this loop from structural data, since the loop is oriented towards bulk solvent during the closed state of the topoisomerase. Accordingly, it is unlikely that it is directly involved in catalysis during pre-cleavage or post-ligation steps of the reaction cycle. However, due to its position at the interface of domain I and III, it is possible that it facilitates catalytic function during reaction steps that require an open conformation of the enzyme. Similar to E. coli topoisomerase I and E. coli topoisomerase III the structure of T. maritima topoisomerase I shows a deep DNA binding cleft mainly built by domains I and IV and extending from the active site across the base of the enzyme. The DNA binding cleft in T. maritima topoisomerase I is formed by helix aO of domain IV building its “upper edge”, helix aF and an adjacent loop (residues 160–165) forming the “bottom” and a large loop protruding from domain I (residues 37–62) providing the “lower edge”. The cleft is coated with a variety of positively charged amino acid residues, which are conserved throughout the subfamily, and resulting in a positively charged surface. Within this area, the residues His36, Arg141, Gln169, Arg290, Thr464 and Arg475 exhibit spatial orientations allowing for hydrogen bonding interactions with single-stranded DNA substrates. In addition, two aromatic residues (Tyr149, Trp156) located in the cleft may provide p–p-stacking interactions with DNA bases. The DNA binding area in the structure of T. maritima topoisomerase I shows orientations of the coating residues, which cannot be superimposed with structures of E. coli topoisomerase I and III determined in complex with short oligonucleotides,10,11 indicating that this area exhibits flexibility and that residues involved in substrate recognition and binding undergo re-orientation upon substrate binding. This point

Figure 4. Stereo representation of the active site of T. maritima topoisomerase I. The catalytic residue Tyr288 is surrounded by residues that are highly conserved in bacterial topoisomerases I. Hydrogen bonds between residues of domains I and III are shown as broken lines.

1334

Structure of T. maritima Topoisomerase I

of view is supported by poor electron densities for certain regions of the DNA binding cleft (residues 156–165 at the bottom and residues 37–62 at the lower edge) and alternative conformations in crystallographically independent molecules (residues 41–45 at the lower edge). Especially the loop including residues 156–165 in T. maritima topoisomerase I protrudes significantly into the DNA binding cleft of the enzyme. Modelling of the singlestranded DNA from E. coli topoisomerase I and III complexes into the binding cleft of the T. maritima enzyme shows extensive steric clashes in the loop region. In summary, this may suggest that the DNA binding cleft of T. maritima topoisomerase I is more flexible than structural data of other type IA topoisomerases indicate and that DNA binding results in structural rearrangements in the cleft, which precede dissociation of domains I and III. Domain V may function as an extended DNA binding and recognition site The functional role of zinc ribbon motifs is diverse and includes DNA binding,42,43 RNA binding,43–46 or zinc ribbon-mediated protein– protein interactions.47–48 Based on the structure of T. maritima topoisomerase I we are suggesting that domain V may function as an extended DNA binding and recognition site. Domain V shows a positively charged surface suitable to mediate DNA interactions and is positioned adjacent to the DNA binding cleft, which extends across the base of the core domain (Figure 5). The orientation of the domain as observed in the structure may enable the interaction of additional nucleotides of singlestranded DNA with domain V. Within this elongated binding surface, the DNA would stretch from the active site, crossing the interface of domains I and IV towards domain V (Figure 5). In particular, exposed aromatic and positively charged residues, namely Trp156, Arg157, Lys160, Arg548, Arg566, and Tyr575, which are located in the region may recognize and bind DNA through favourable interactions. This is further supported by the fact that the C-terminal part of helix aF and the loop region 543–552 exhibit some flexibility, suggesting that this part of the structure may rearrange upon DNA binding to optimise protein substrate interactions. Studies employing a 22mer oligonucleotide showed that T. maritima topoisomerase I preferably cleaves this substrate after a cytosine at position 17. After cleavage, both DNA fragments remain tightly associated with the protein.18 According to the catalytic mechanism, the 5 0 end of the shorter part of the oligonucleotide (nucleotides 18–22) is covalently linked to the active site tyrosine after cleavage, leaving the remaining 17 nucleotides (nucleotides 1–17) for non-covalent interactions within the substrate binding cleft. Similar studies with a substrate labelled at the first nucleotide have shown that this fragment remains tightly associated

Figure 5. Schematic representation of single-stranded DNA binding to T. maritima topoisomerase I. Molecular surface of the molecule showing the DNA binding region of the protein. The individual domains are coloured as in Figure 1. The broken line marks the postulated elongated interaction surface path for single-stranded DNA. The DNA occupies the binding cleft in vicinity of the active site, travels across the interface of domains I and IV and interacts with the positively charged surface of domain V.

with the protein after cleavage, even during native gel electrophoresis,21 indicating a strong interaction within the complex. Structural comparison of the T. maritima topoisomerase I and the E. coli topoisomerase I and III DNA binding clefts reveals that T. maritima topoisomerase I also presents space sufficient to accommodate six to eight nucleotides.10,11 Interestingly, it has been found that the truncation of domain V alters the cleavage specificity and leads to a reduced cleavage efficiency.21 Cleavage experiments have shown that with truncated T. maritima topoisomerase I lacking domain V the cleavage point of 22mer oligonucleotides is shifted by four bases (after cytidine 17 in full-length T. maritima topoisomerase I versus after adenosine 13 in truncated T. maritima topoisomerase I). This suggests that domain V in T. maritima topoisomerase I may interact with nucleotides in the region K7 to K17 located 5 0 of the substrate cleavage site, ultimately resulting in increased relaxation activity facilitated by an extended DNA interaction area. Additionally, domain V may also have a role in positioning the enzyme on the substrate DNA, thereby determining the major cleavage site. This is also supported by the finding that the cleavage specificity of truncated T. maritima topoisomerase I is significantly altered.21

1335

Structure of T. maritima Topoisomerase I

Implications for other members of the bacterial toposiomerase I protein family The structure of full-length T. maritima topoisomerase I, including core and C-terminal domain provides a detailed structural view of the intact enzyme and suggests a role of the zinc ribbon domain in DNA binding. Although the C-terminal regions of bacterial topoisomerases I are very diverse in length and sequence, the vast majority of the enzymes contain zinc ribbon motifs in this region of the protein.12 The most prominent bacterial topoisomerase I, E. coli topoisomerase I, harbours a long C-terminal region (268 residues) comprising three functional zinc ribbon motifs.22 Similar to T. maritima topoisomerase I this region has been implicated in binding of single-stranded DNA.14,22,49,50 However, in contrast to T. maritima topoisomerase I, biochemical studies revealed that truncation of the entire C-terminal region14,21 or full zinc depletion50 inactivates the DNA relaxation activity of E. coli topoisomerase I. In the absence of a structure of intact E. coli topoisomerase I, a structure-based sequence alignment of T. maritima topoisomerase I and E. coli topoisomerase I (Figure 6) reveals structural equivalence of the zinc ribbon domain in T. maritima topoisomerase I and the first zinc ribbon motif within the C-terminal region of E. coli topoisomerase I. A linker region of 14 residues length in E. coli topoisomerase I situated between helix aR and the first zinc ribbon motif allows a similar orientation of the first zinc ribbon motif as observed for domain V in T. maritima topoisomerase I. Moreover, the conserved zinc knuckle motifs and the predicted secondary structure elements within the first zinc ribbon motif align well with similar elements observed in domain V of T. maritima topoisomerase I (Figure 6). The first zinc knuckle motif in E. coli topoisomerase I, situated between strands b15 and b16, exhibits the sequence CPTCG, which is fully in line with the consensus sequence CPXCG of the zinc ribbon fold.25 The second motif, situated following strand b17, exhibits the sequence CSGYALPPKERCK and therefore reveals a larger deviation from the consensus sequence. Notwithstanding the above, the sequence may still be in good alignment with the zinc ribbon fold, since the observed variation in sequence and length of the consensus sequences is quite diverse and even larger deviations are found quite frequently.25 Therefore, we conclude that the structure of domain V and its orientation relative to the core domain as observed in T. maritima topoisomerase I are a common feature of bacterial topoisomerases I, that will also be observed for the first zinc ribbon motif following the core domain of topoisomerase I from E. coli and other bacteria. Moreover, this also suggests a common functional role of this domain similar to the role suggested for domain V of T. maritima topoisomerase I. This also hints to a general non-functional significance of the zinc ion of the first zinc ribbon domain, which in turn

further implicates that the observed zinc dependency of the functional activity of E. coli topoisomerase I16 resides within the remaining second and third zinc ribbon motif. Concurrently, this also suggests that the impaired DNA relaxation activity of E. coli topoisomerase I, observed upon truncation of the entire C-terminal region,14,21 may also reside in the functional roles of the second and/or third zinc ribbon motif of E. coli topoisomerase I rather than in the first zinc ribbon domain. The structure and orientation of the second and third zinc ribbon motif of E. coli topoisomerase I are currently unknown, however, based on the structure analysis of T. maritima topoisomerase I and in accordance with other experimental findings and mechanistic models, we postulate a common structural and functional role for the first zinc ribbon domain of bacterial topoisomerases I: the domain and residues situated in the vicinity supply an additional DNA binding interface for the cleavable DNA strand, while the remaining C-terminal zinc ribbon motifs may support passing of the intact DNA strand through the “gate” in the cleaved strand.14 Further biochemical studies, for e.g. employing partially truncated C-terminal variants, may shed further light on the functional importance of domain V and of the second and third zinc ribbon motif

Materials and Methods Purification and crystallization of topoisomerase I from T. maritima The topoisomerase I gene (topA) from T. maritima was cloned into the NcoI/NotI sites of pET-28a(C) expression vector (Novagen) using the following primers: 5 0 TACCATG GCTAAGAAAGTGAAGAAATATAT-3 0 and 5 0 -ACTTGCGGCCGCTTAAGAGCCTTTTTTACCCTTTC3 0 . This yielded a product expressing topoisomerase I with serine at position two changed to alanine under the control of T7 promoter. The plasmid was transformed into E. coli expression strain BL21 DE3, protein was expressed under conditions as described.18 To obtain protein suitable for crystallization cells were lysed by sonification in 10 volumes of buffer A (50 mM Tris–HCl (pH 7.5), 100 mM NaCl) containing 0.5 mg/ml lysozyme. After centrifugation at 50,000g for 20 min at 4 8C the supernatant was heated for 25 min at 75 8C. Denaturated E. coli proteins were removed by centrifugation at 50,000g for 40 min at 4 8C. Identical centrifugation conditions were used during all following purification steps. To remove DNA, polyethylenimine (Sigma; P-3143) was added to the supernatant up to a final concentration of 0.5%. After stirring for 1 h at 4 8C the solution was centrifuged to remove the polyethylenimine precipitate. Solid ammonium sulfate (70% saturation) was added, and the solution was stirred overnight at 4 8C before centrifugation. To remove remaining DNA associated with the protein the pellet was dissolved in 20 ml buffer B (50 mM potassium phosphate (pH 7.5) prepared from 1 M KH2PO4 and 1 M K2HPO4, 100 mM NaCl) and incubated with DNAase I (80 mg/ml; Merck) in the presence of 5 mM MgCl2 for 2 h at room temperature. DNAase

1336

Structure of T. maritima Topoisomerase I

Figure 6. Structure-based sequence alignment of T. maritima topoisomerase I and E. coli topoisomerase I. Secondary structure elements of T. maritima topoisomerase I are depicted as cartoons. E. coli topoisomerase I secondary structural

1337

Structure of T. maritima Topoisomerase I

Table 1. Data collection statistics Space group Beamline ˚) Wavelength (A ˚) Unit-cell dimensions (A ˚) Resolution (A No. of observations No. of unique reflections Completeness (%)a Overall Highest resolution shell I/sigma Overall Highest resolution shell Multiplicity Overall Highest resolution shell Rmergeb Overall Highest resolution shell

P1 PX (SLS) 0.9190 aZ44.90, bZ94.24, cZ95.63, aZ83.78,bZ86.18,gZ84.88 63–2.5 190,918 53,590

P1 ID29 (ESRF) 0.9756 aZ45.12, bZ95.42, cZ96.51, aZ83.48, bZ86.28,gZ84.98 45–1.7 1,812,308 174,953

P21 ID29 (ESRF) 1.2815 aZ45.05, bZ142.71, cZ127.64, aZgZ908, bZ90.84 45–1.95 2,165,191 116,615

˚) 92.9 (63–2.5 A ˚) 94.3 (2.59–2.5 A

˚) 95.6 (45–1.7 A ˚) 95.1 (1.79–1.7 A

˚) 99.6 (45–1.95 A ˚) 99.1 (2.06–1.95 A

˚) 10.5 (63–2.5 A ˚) 2.8 (2.59–2.5 A

˚) 10.5 (45–1.7 A ˚) 3.0 (1.79–1.7 A

˚) 8.2 (45–1.95 A ˚) 2.7 (2.06–1.95 A

˚) 1.7 (63–2.5 A ˚) 1.7 (2.59–2.5 A

˚) 3.0 (45–1.7 A ˚) 3.0 (1.79–1.7 A

˚) 4.0 (45–1.95 A ˚) 3.7 (2.06–1.95 A

˚) 0.07 (63–2.5 A ˚) 0.27 (2.59–2.5 A

˚) 0.07 (45–1.7 A ˚) 0.44 (1.79–1.7 A

˚) 0.12 (45–1.95 A ˚) 0.44 (2.06–1.95 A

a

Ratio of the of possible P number Pand P P the number of present unique reflections. Rmerge Z h i jIðh; iÞKhIðhÞij= h i ðh; iÞ, where I(h,i) is the intensity value of the ith measurement of h and hI(h)i is the corresponding mean value of h for all i measurements of h. The summation is over all measurements. b

activity was removed by a second heating step (25 min, 75 8C) followed by a centrifugation step. The supernatant was applied to a 5 ml HiTrap SP FF column (Amersham) equilibrated with buffer B. Protein was eluted using a linear gradient of 0.1 M–1 M NaCl in buffer B. Topoisomerase-containing fractions were pooled, concentrated and loaded onto a gel filtration column (Superdex S200; Amersham) equilibrated with buffer C (20 mM Tris–HCl (pH 7.5), 100 mM NaCl). Fractions containing topoisomerase I as confirmed by SDS-PAGE were pooled, concentrated to 7 mg/ml and stored at K80 8C. Activity of the enzyme was checked with the help of a DNA relaxation assay as described18 using pUK18 plasmid DNA. Homogeneity of the protein solution was confirmed by SDS-PAGE and amino acid analysis (data not shown). Crystallization experiments were performed at 18 8C using hanging-drop vapour diffusion method with O-(2aminopropyl)-O 0 -(2-methoxyethyl)polypropylene glycol 500 (Jeffamine M-600) as precipitant. Droplets were made by mixing 1 ml of protein solution (7 mg/ml protein in buffer C) and 1 ml of precipitating buffer. Crystals belonging to two different space groups, namely P1 and P21, were obtained using 33% Jeffamine M-600 (pH 7.5), 100 mM sodium citrate/HCl (pH 4.8) and 34% Jeffamine M-600 (pH 7.5), 100 mM Mes/NaOH (pH 6.0), 64 mM sodium citrate, respectively. 50% Jeffamine stock solutions were prepared from 100% Jeffamine M-600 solution (Fluka) using HCl to adjust pH to 7.5. Small crystals grew after three to seven days to a final size of 0.1 mm!0.05 mm!0.05 mm with lattice constants aZ ˚ , bZ95.42 A ˚ , cZ96.51 A ˚ , aZ83.48, bZ86.28, gZ 45.12 A ˚ ˚ , cZ127.64 A ˚ , aZ 84.98 (P1) and aZ45.05 A, bZ142.71 A gZ908, bZ90.848 (P21), respectively. Due to high Jeffamine concentrations present in the mother liquor, crystals of both space groups could be flash frozen directly in liquid nitrogen prior to diffraction experiments.

X-ray data collection, structure solution, and refinement ˚ were collected for Initial X-ray diffraction data to 2.5 A the triclinic crystal form at 100 K using the SLS undulator beamline PX (Villingen) The beamline was equipped with a horizontal Si(111) monochromator, a vertical dynamically bendable mirror and MAR225 mosaic CCD detector (MAR Research). High resolution X-ray diffraction data to ˚ and 1.95 A ˚ resolution were collected for the triclinic 1.7 A and the monoclinic crystal form at 100 K using the ESRF undualtor beamline ID29 (Grenoble), equipped with a Si(111) monochromator, a bendable mirror and a ADSC Q210 2D CCD detector (ADSC). All datasets were processed, scaled and merged using MOSFLM and SCALA;51 selected data collection statistics are given in Table 1. A first model of topoisomerase I from T. maritima was obtained using molecular replacement methods with the program AMoRE52 and a dataset taken at the SLS. Although a clear solution with good crystal packing was obtained using the 67 kDa fragment of E. coli topoisomerase I4 (PDB code 1ECL) as search model, electron density for most parts of the protein was very poor. Using the program O53 regions of the model not clearly defined by electron density were deleted and ambiguous side-chains were changed to alanine. Starting from this strongly reduced model containing 52% of the TmTopI residues an almost complete model (88%) was generated using ARP/wARP54 with the high-resolution ˚ . The model was dataset of a P1 crystal diffracting to 1.7 A completed by iterative cycles of manual model building using the program O and crystallographic refinement using the program REFMAC5. 55 During the final refinement step each monomer was defined as a rigid body and translation, liberation and screw rotation tensors (TLS) were refined for each to account for overall

elements are indicated as boxes, and disordered regions in crystal structures are indicated by grey shadows. The alignment of the C-terminal domain V (T. maritima topoisomerase I residues 543–601) is based on the sequence alone. Secondary structure elements of E. coli topoisomerase I in this region result from secondary structure prediction and are indicated as broken boxes.

1338 Table 2. Final refinement statistics Space group P1 P21 Beamline ID29 (ESRF) ID29 (ESRF) ˚) Resolution range (A 45–1.7 45–1.95 No. of unique reflections 159,417 110,740 No. of molecules per 2 2 asymmetric unit No. of non-hydrogen atoms 10,600 10,305 No. of water molecules 1054 877 R factor (%)a Overall 19.7 (23.2) 19.5 (24.6) 25.3 (32.5) 27.0 (30.1) Highest resolution shellb ˚ 2) Isotropic B-factors (A All atoms 26.53 24.69 Main-chain atoms 23.81 22.69 Side-chain atoms 27.38 25.81 Solvent 34.67 29.22 RMS deviations from target values ˚ Bonds lengths (A) 0.014 0.014 Bond angles (deg.) 1.8 2.0 ˚) RMS deviations between molecules in AU (A 0.36 0.33 Ca-atoms All atoms 0.65 0.66 Solvent content (%) 55.2 56.4 P P a RZ hkl jjFo jKjFc jj= hkl jFo j, value for Rfree given in parentheses. b ˚ for P1 and 2.0–1.95 A ˚ for P21. 1.74–1.7 A

differences in displacements between the molecules and anisotropy in the data.56 The structure of the monoclinic crystal form was solved using the structure of the P1 crystal form as search model in molecular replacement, followed by model building in O and refinement in REFMAC. The refinement procedure was carried out as described for the structure of the P1 space group. Final refinement statistics for both models are given in Table 2. Figures Figures were made with Pymol58 and Topdraw.59 Protein Data Bank accession codes The coordinates and structure factors for both crystal forms have been deposited in the RCSB Protein Data Bank57 with PDB codes ZGAI and ZGAJ, respectively.

References 1. Champoux, J. J. (2001). DNA topoisomerases: structure, function, and mechanism. Annu. Rev. Biochem. 70, 369–413. 2. Mitscher, L. A. (2005). Bacterial topoisomerase inhibitors: quinolone and pyridone antibacterial agents. Chem. Rev. 105, 559–592. 3. Nagaraja, V., Sikder, D. & Jain, P. (2003). DNA topoisomerase I from mycobacteria—a potential drug target. Curr. Pharm. Des. 8, 1995–2007. 4. Lima, C. D., Wang, J. C. & Mondragon, A. (1994). Three-dimensional structure of the 67K N-terminal fragment of E. coli DNA topoisomerase I. Nature, 367, 138–146. 5. Mondragon, A. & DiGate, R. (1999). The structure of Escherichia coli DNA topoisomerase III. Struct. Fold. Des. 7, 1373–1383.

Structure of T. maritima Topoisomerase I

6. Rodriguez, A. C. & Stock, D. (2002). Crystal structure of reverse gyrase: insights into the positive supercoiling of DNA. EMBO J. 21, 418–426. 7. Brown, P. O. & Cozzarelli, N. R. (1981). Catenation and knotting of duplex DNA by type 1 topoisomerases: a mechanistic parallel with type 2 topoisomerases. Proc. Natl Acad. Sci. USA, 78, 843–847. 8. Champoux, J. J. (1990). DNA Topology and Its Biological Effects, Cold Spring Habor Laboratory Press, Cold Spring Habor, New York. 9. Tse, Y. & Wang, J. C. (1980). E. coli and M. luteus DNA topoisomerase I can catalyze catenation of decatenation of double-stranded DNA rings. Cell, 22, 269–276. 10. Changela, A., DiGate, R. J. & Mondragon, A. (2001). Crystal structure of a complex of a type IA DNA topoisomerase with a single-stranded DNA molecule. Nature, 411, 1077–1081. 11. Perry, K. & Mondragon, A. (2003). Structure of a complex between E. coli DNA topoisomerase I and single-stranded DNA. Structure, 11, 1349–1358. 12. Grishin, N. V. (2000). C-terminal domains of Escherichia coli topoisomerase I belong to the zinc-ribbon superfamily. J. Mol. Biol. 299, 1165–1177. 13. Lima, C. D., Wang, J. C. & Mondragon, A. (1993). Crystallization of a 67 kDa fragment of Escherichia coli DNA topoisomerase I. J. Mol. Biol. 232, 1213–1216. 14. Ahumada, A. & Tse-Dinh, Y. C. (2002). The role of the Zn(II) binding domain in the mechanism of E. coli DNA topoisomerase I. BMC Biochem. 3, 13. 15. Cheng, B., Zhu, C. X., Ji, C., Ahumada, A. & Tse-Dinh, Y. C. (2003). Direct interaction between Escherichia coli RNA polymerase and the zinc ribbon domains of DNA topoisomerase I. J. Biol. Chem. 278, 30705–30710. 16. Zhu, C. X., Qi, H. Y. & Tse-Dinh, Y. C. (1995). Mutation in Cys662 of Escherichia coli DNA topoisomerase I confers temperature sensitivity and change in DNA cleavage selectivity. J. Mol. Biol. 250, 609–616. 17. Bouthier de la Tour, C., Kaltoum, H., Portemer, C., Confalonieri, F., Huber, R. & Duguet, M. (1995). Cloning and sequencing of the gene coding for topoisomerase I from the extremely thermophilic eubacterium, Thermotoga maritima. Biochim. Biophys. Acta, 1264, 279–283. 18. Viard, T., Lamour, V., Duguet, M. & Bouthier de la Tour, C. (2001). Hyperthermophilic topoisomerase I from Thermotoga maritima. A very efficient enzyme that functions independently of zinc binding. J. Biol. Chem. 276, 46495–46503. 19. Dean, F. B. & Cozzarelli, N. R. (1985). Mechanism of strand passage by Escherichia coli topoisomerase I. The role of the required nick in catenation and knotting of duplex DNA. J. Biol. Chem. 260, 4984–4994. 20. Dekker, N. H., Rybenkov, V. V., Duguet, M., Crisona, N. J., Cozzarelli, N. R., Bensimon, D. & Croquette, V. (2002). The mechanism of type IA topoisomerases. Proc. Natl Acad. Sci. USA, 99, 12126–12131. 21. Viard, T., Cossard, R., Duguet, M. & Bouthier de La Tour, C. B. (2004). Thermotoga maritima–Escherichia coli chimeric topoisomerases. Answers about involvement of the carboxyl-terminal domain in DNA topoisomerase I-mediated catalysis. J. Biol. Chem. 279, 30073–30080. 22. Tse-Dinh, Y. C. & Beran-Steed, R. K. (1988). Escherichia coli DNA topoisomerase I is a zinc metalloprotein with three repetitive zinc-binding domains. J. Biol. Chem. 263, 15857–15859. 23. Li, Z., Mondragon, A., Hiasa, H., Marians, K. J. & DiGate, R. J. (2000). Identification of a unique domain

Structure of T. maritima Topoisomerase I

24.

25. 26.

27. 28.

29. 30. 31. 32. 33.

34. 35. 36.

37.

38. 39.

40.

41.

essential for Escherichia coli DNA topoisomerase IIIcatalysed decatenation of replication intermediates. Mol. Microbiol. 35, 888–895. Qian, X., Gozani, S. N., Yoon, H., Jeon, C. J., Agarwal, K. & Weiss, M. A. (1993). Novel zinc finger motif in the basal transcriptional machinery: three-dimensional NMR studies of the nucleic acid binding domain of transcriptional elongation factor TFIIS. Biochemistry, 32, 9944–9959. Krishna, S. S., Majumdar, I. & Grishin, N. V. (2003). Structural classification of zinc fingers: survey and summary. Nucl. Acids Res. 31, 532–550. Chen, H. T., Legault, P., Glushka, J., Omichinski, J. G. & Scott, R. A. (2000). Structure of a (Cys3His) zinc ribbon, a ubiquitous motif in archaeal and eucaryal transcription. Protein Sci. 9, 1743–1752. Baldwin, M. A. & Benz, C. C. (2002). Redox control of zinc finger proteins. Methods Enzymol. 353, 54–69. Maret, W. (2002). Optical methods for measuring zinc binding and release, zinc coordination environments in zinc finger proteins, and redox sensitivity and activity of zinc-bound thiols. Methods Enzymol. 348, 230–237. Griep, M. A. & Lokey, E. R. (1996). The role of zinc and the reactivity of cysteines in Escherichia coli primase. Biochemistry, 35, 8260–8267. Padovani, D., Mulliez, E. & Fontecave, M. (2001). Activation of class III ribonucleotide reductase by thioredoxin. J. Biol. Chem. 276, 9587–9589. Aslund, F. & Beckwith, J. (1999). Bridge over troubled waters: sensing stress by disulfide bond formation. Cell, 96, 751–753. Jakob, U., Muse, W., Eser, M. & Bardwell, J. C. (1999). Chaperone activity with a redox switch. Cell, 96, 341–352. Paget, M. S., Bae, J. B., Hahn, M. Y., Li, W., Kleanthous, C., Roe, J. H. & Buttner, M. J. (2001). Mutational analysis of RsrA, a zinc-binding anti-sigma factor with a thiol–disulphide redox switch. Mol. Microbiol. 39, 1036–1047. Koehler, C. M., Merchant, S. & Schatz, G. (1999). How membrane proteins travel across the mitochondrial intermembrane space. Trends Biochem. Sci. 24, 428–432. Lutz, T., Neupert, W. & Herrmann, J. M. (2003). Import of small Tim proteins into the mitochondrial intermembrane space. EMBO J. 22, 4400–4408. Curran, S. P., Leuenberger, D., Oppliger, W. & Koehler, C. M. (2002). The Tim9p–Tim10p complex binds to the transmembrane domains of the ADP/ATP carrier. EMBO J. 21, 942–953. Curran, S. P., Leuenberger, D., Schmidt, E. & Koehler, C. M. (2002). The role of the Tim8p–Tim13p complex in a conserved import pathway for mitochondrial polytopic inner membrane proteins. J. Cell Biol. 158, 1017–1027. Koehler, C. M. (2004). The small Tim proteins and the twin Cx3C motif. Trends Biochem. Sci. 29, 1–4. Perry, K. & Mondragon, A. (2002). Biochemical characterization of an invariant histidine involved in Escherichia coli DNA topoisomerase I catalysis. J. Biol. Chem. 277, 13237–13245. Zhu, C. X. & Tse-Dinh, Y. C. (2000). The acidic triad conserved in type IA DNA topoisomerases is required for binding of Mg(II) and subsequent conformational change. J. Biol. Chem. 275, 5318–5322. Zumstein, L. & Wang, J. C. (1986). Probing the structural domains and function in vivo of Escherichia coli DNA topoisomerase I by mutagenesis. J. Mol. Biol. 191, 333–340.

1339 42. Agarwal, K., Baek, K. H., Jeon, C. J., Miyamoto, K., Ueno, A. & Yoon, H. S. (1991). Stimulation of transcript elongation requires both the zinc finger and RNA polymerase II binding domains of human TFIIS. Biochemistry, 30, 7842–7851. 43. Qian, X., Jeon, C., Yoon, H., Agarwal, K. & Weiss, M. A. (1993). Structure of a new nucleic-acid-binding motif in eukaryotic transcriptional elongation factor TFIIS. Nature, 365, 277–279. 44. Hard, T., Rak, A., Allard, P., Kloo, L. & Garber, M. (2000). The solution structure of ribosomal protein L36 from Thermus thermophilus reveals a zinc-ribbonlike fold. J. Mol. Biol. 296, 169–180. 45. Harms, J., Schluenzen, F., Zarivach, R., Bashan, A., Gat, S., Agmon, I. et al. (2001). High resolution structure of the large ribosomal subunit from a mesophilic eubacterium. Cell, 107, 679–688. 46. Klein, D. J., Schmeing, T. M., Moore, P. B. & Steitz, T. A. (2001). The kink-turn: a new RNA secondary structure motif. EMBO J. 20, 4214–4221. 47. Archer, M., Huber, R., Tavares, P., Moura, I., Moura, J. J., Carrondo, M. A. et al. (1995). Crystal structure of desulforedoxin from Desulfovibrio gigas determined at ˚ resolution: a novel non-heme iron protein 1.8 A structure. J. Mol. Biol. 251, 690–702. 48. Macol, C. P., Tsuruta, H., Stec, B. & Kantrowitz, E. R. (2001). Direct structural evidence for a concerted allosteric transition in Escherichia coli aspartate transcarbamoylase. Nature Struct. Biol. 8, 423–426. 49. Kirkegaard, K. & Wang, J. C. (1985). Bacterial DNA topoisomerase I can relax positively supercoiled DNA containing a single-stranded loop. J. Mol. Biol. 185, 625–637. 50. Tse-Dinh, Y. C. (1991). Zinc(II) coordination in Escherichia coli DNA topoisomerase I is required for cleavable complex formation with DNA. J. Biol. Chem. 266, 14317–14320. 51. Collaborative Computational Project, Number 4. (1994). The CCP4 suite: programs for protein crystallography. Acta Crystallog. sect. D, 50, 760–763. 52. Navaza, J. (1994). AMoRe: an automated package for molecular replacement. Acta Crystallog. sect. A, 50, 157–163. 53. Jones, T. A., Zou, J. Y., Cowan, S. W. & Kjeldgaard, M. (1991). Improved methods for building protein models in electron density maps and the location of errors in these models. Acta Crystallog. sect. A, 47, 110–119. 54. Perrakis, A., Harkiolaki, M., Wilson, K. S. & Lamzin, V. S. (2001). ARP/wARP and molecular replacement. Acta Crystallog. sect. D, 57, 1445–1450. 55. Murshudov, G. N., Vagin, A. A. & Dodson, E. J. (1997). Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallog. sect. D, 53, 240–255. 56. Winn, M. D., Isupov, M. N. & Murshudov, G. N. (2001). Use of TLS parameters to model anisotropic displacements in macromolecular refinement. Acta Crystallog. sect. D, 57, 122–133. 57. Berman, H. M., Bhat, T. N., Bourne, P. E., Feng, Z., Gilliland, G., Weissig, H. & Westbrook, J. (2000). The Protein Data Bank and the challenge of structural genomics. Nature Struct. Biol. 7, 957–959. 58. DeLano, W. L. (2002). The PyMOL Molecular Graphics System, DeLano Scientific, San Carlos, CA, USA. 59. Bond, C. S. (2003). TopDraw: a sketchpad for protein structure topology cartoons. Bioinformatics, 19, 311–312.

1340

Structure of T. maritima Topoisomerase I

60. Zhu, W., Zeng, Q., Colangelo, C. M., Lewis, M., Summers, M. F. & Scott, R. A. (1996). The N-terminal domain of TFIIB from Pyrococcus furiosus forms a zinc ribbon. Nature Struct. Biol. 3, 122–124.

61. Wang, B., Jones, D. N., Kaine, B. P. & Weiss, M. A. (1998). High-resolution structure of an archaeal zinc ribbon defines a general architectural motif in eukaryotic RNA polymerases. Structure, 6, 555–569.

Edited by K. Morikawa (Received 18 December 2005; received in revised form 5 March 2006; accepted 6 March 2006) Available online 23 March 2006