sign in sign up
<<
Home , Methanopyrus, Type I topoisomerase

Proc. Natl. Acad. Sci. USA Vol. 93, pp. 106-110, January 1996 Biochemistry

A two-subunit type I DNA (reverse gyrase) from an extreme

REGIS KRAH*, SERGEI A. KOZYAVKIN*t, ALEXEI I. SLESAREVt, AND MARTIN GELLERT*§ *Laboratory of Molecular Biology, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD 20892; and tMolecular Biology Institute and Department of Biology, University of California, Los Angeles, CA 90024 Contributed by Martin Gellert, September 6, 1995

ABSTRACT A recently described reverse gyrase from the helicases (15-17). These helicases are nucleic acid-dependent hyperthermophilic kandleri is the ATPases, as is reverse gyrase (18, 19). It was proposed that the only known example of a heterodimeric . helicase-like ATPase domain uses ATP to translocate along The is made up of a 42-kDa subunit which covalently the DNA, generating a transient wave of positive and negative interacts with DNA (RgyA) and a 138-kDa subunit which supercoiling (15, 20, 21), while the topoisomerase domain binds ATP (RgyB). We have now cloned and sequenced the selectively removes negative supercoils, producing a super- for both subunits of this enzyme. Surprisingly, the coiled molecule. It is important to note that helicase activity universally conserved type I topoisomerase domain [Lima, has not yet been demonstrated for reverse gyrase and that ATP C. D., Wang, J. C. & Mondragon, A. (1994) Nature (London) hydrolysis may also be more directly linked to the mechanism 367, 138-146] which has been found as a contiguous polypep- of DNA strand passage. tide in the and is shared between the Recently, a reverse gyrase was described from the hyper- protomers. The subdomain with the active-site is thermophilic methanogen Methanopyrus kandleri (18). Unlike entirely within RgyA, whereas the subdomain implicated in other type 1-5' , this enzyme was found to be a noncovalent binding of the cleaved DNA strand is contained heterodimer composed of a smaller subunit with an apparent entirely in RgyB. The appearance of this unique structure in molecular mass of 50 kDa (RgyA) that formed a covalent a highly conserved enzyme family supports the hypothesis that complex with DNA, and a larger subunit with an apparent the branched from other prokaryotes and eu- molecular weight of 150 kDa (RgyB) that was involved in the karyotes very early in evolution. hydrolysis ofATP. To further study this , we cloned and sequenced the genes encoding the subunits of this enzyme.1 A DNA topoisomerases are found in all organisms. They are remarkable feature of the sequence is that the domain which required for replication, decatenation, and unknotting of DNA characterizes all type I-5' topoisomerases, and is found as a in cells (1-3), and by controlling the torsional stress of DNA single polypeptide chain in all other of this class, is they can also affect the efficiency of expression and here divided between the two protomers. various recombination processes. All topoisomerases can be classified into three general subgroups that are distinguished MATERIALS AND METHODS by common biochemical properties and similarity in primary structure. These are the type 1-5' topoisomerases, the type I-3' Amino-Terminal Sequencing of and Development topoisomerases, and the type II topoisomerases. of DNA Probes. Microsequencing of the amino termini of topoisomerase I, the first topoisomerase RgyA and RgyB was performed on an Applied Biosystems discovered (4), is the prototype of the type 1-5' topoisomer- 471A protein sequencer. Five picomoles ofM. kandleri reverse ases. Enzymes of this class have been found in eubacteria, gyrase (18) was separated into the A and B subunits by , and eukaryotes (4-9). They share a number of SDS/PAGE. The proteins were blotted to Immobilon_psq common properties: they change the linking number of DNA transfer membrane (Millipore), and subjected to 20 cycles of by steps of 1, require a single-stranded region in the DNA Edman degradation (22). The amino-terminal amino acid substrate, are active only in the presence of a divalent metal sequence of RgyA was determined to be MNATLRIRNRP- cation, and form a covalent bond to DNA through a 5' VAE. Determination of the amino-terminal sequence of RgyB phosphoryl group during the cleavage reaction. required the comparison of two separate sequencing reactions, In the mid 1980s, a novel type I-5' topoisomerase, reverse and the sequence was found to be VLXRAXXMVPKGF (X gyrase, was isolated from extremely hyperthermophilic ar- indicates unreadable positions). The valine in the first position chaea (6, 7, 10, 11). This enzyme was able to positively indicates that the amino terminus was modified by proteolytic supercoil DNA in the presence ofATP or dATP. Although the cleavage. The observed signal for each cycle during the physiological role of this enzyme remains obscure, positive microsequencing was much weaker than expected, suggesting supercoiling would help to stabilize the DNA duplex against that the amino terminus in a significant fraction of the protein denaturation at the extreme temperatures at which these molecules is blocked. A blocked amino terminus has also been organisms grow (12, 13). Reverse gyrase has been found in all found in reverse gyrase from S. acidocaldarius (15). major branches of hyperthermophilic archaea as well as in PCR amplification with degenerate oligonucleotide primers hyperthermophilic eubacteria and is thus considered a molec- was used to generate DNA probes for rgyA and rgyB. Knowing ular marker for hyperthermophilic prokaryotes (12, 14). that the amino acid sequence of S. acidocaldarius RgyS had The cloning and sequencing of a reverse gyrase from Sulfolobus acidocaldarius (RgyS) revealed a monomeric pro- Abbreviation: ORF, open reading frame. tein which fused the universally conserved type 1-5' topo- tPresent address: Third Wave Technologies, Madison, WI 53711-5399. domain to a domain reminiscent of nucleic acid §To whom reprint requests should be addressed at: National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, Laboratory of Molecular Biology, Building 5, Room 241, The publication costs of this article were defrayed in part by page charge 9000 Rockville Pike, Bethesda, MD 20892-0540. payment. This article must therefore be hereby marked "advertisement" in 1The sequences reported in this paper have been deposited in the accordance with 18 U.S.C. §1734 solely to indicate this fact. GenBank database (accession nos. U41058 and U41059). 106 Downloaded by guest on September 30, 2021 Biochemistry: Krah et al. Proc. Natl. Acad. Sci. USA 93 (1996) 107 revealed a protein that fused a helicase-like ATPase domain to RESULTS AND DISCUSSION the type 1-5' topoisomerase domain (15), we used alignments of helicases and topoisomerases with RgyS to locate conserved Cloning and Sequence Analysis of rgyB. Sequence analysis amino acid sequences and direct the design of these primers. revealed an open reading frame (ORF) encoding a protein of For rgyB, a probe complementary to the amino terminus 1221 aa, with a predicted molecular mass of 138 kDa. This is (probe 1) and a second probe (probe 2) based on highly in good agreement with the molecular mass of 150 kDa conserved motifs found in the type 1-5' DNA topoisomerases determined by SDS/PAGE (18). The coding sequence of the were developed. For probe 1, a series of degenerate PCR ORF agrees completely with the protein microsequencing primers based on the amino acid sequence MVPKG taken data: there is an ATG codon for methionine directly 5' of the from the amino terminus of RgyB were paired with a series of codon for the first valine. Strong similarity with S. acidocal- darius RgyS near aa 52 of degenerate PCR primers based on the sequence flanking and begins RgyB (Fig. 1) and includes a conserved zinc near including a conserved motif, the DEAH box of nucleic acid tetracysteine finger motif the amino termini of both enzymes. helicases (16, 17) (see Fig. 1). The second probe for rgyB was RgyB has the series of amino acid motifs found in the nucleic developed by using PCR primers based on the highly con- acid helicases which have also been identified in RgyS (Fig. 1) served motifs TEGEKI and VQTPVL found in the type I-5' (16, 17). Throughout the ATPase domain, RgyB and RgyS are DNA topoisomerases (see Fig. 2A). PCR yielded amplification 30% identical and 57% similar. Protein motifs common to products from both sets of that were cloned and primer pairs helicases are overlined and underlined in Fig. 1, and within and The usefulness of 1 for was sequenced. probe rgyB confirmed near these regions RgyB and RgyS share the greatest identity by the presence of a phenylalanine codon adjacent to the 5' (64%). However, between these regions RgyB has 14 inser- primer, in agreement with the sequence analysis of the amino terminus. Adjacent to the 3' primer, amino acid sequences with similarity to RgyS were found (15). For probe 2, amino acid 1 MVLKRAADMVPKGFRDLVEPILDDCADLEELADRVVETEMEPDEVRRRDVG 51 rgyB sequences conserved in type I-5' topoisomerases were found. 52 NTDSNEPVAIFGSSCVLCGGDCSSVRLTSRIGICERCLPVDTE..TLREV 99 rgyB Probe 1 was used to screen 5000 plaques from an amplified M. ..1: 1..1: 1111.1 11 :1 1111 .:: kandleri A library. The DNA from 6 plaques that hybridized to 2 QSLSDIPPSIYLFSCPNCGRSISTYRLL . LGSVCNICLEEDKEYKNIGDL 50 rgyS this probe was purified and probed with probe 2. All 6 A clones 100 LKEARKRHGYVGEALLMFILVERYSPDRVEEFFRRYVWPELFTEIVDRVF 149 rgyB were found to hybridize to probe 2 and were found to share 51 IKDIEKQGN ...... LIKLKDIQRVLDDYESFVS ...... VF 80 rgyS common restriction fragments that hybridized to both probes. 150 DRATGFRLYSAQRVWTRRLVKGCSFSILAPTGTGKTSWGSLVAAVFGHAG 199 rgyB For rgyA, a single probe was developed on the basis of 81 RRLLGFPPFGPQKSWIYRLLSGESFAIIAPPGLGKTTFGLISSIYLYLRG 130 rgyS sequence taken from the amino terminus (MNATLRI) and 200 the sequence of conserved motifs near the active-site tyrosine RRVYYLVPTTTLVRQVENR.... IKGFARDAELDVDVVAYHAAMPTQAKR 245 rgyB of the type 1-5' topoisomerases (ITYHRTD; see Fig. 2B). PCR 131 KKSILVFPTKSLVROAIDKLSSYIQNLAEIKENPPKVIY5SASERK 180 rgyS amplification yielded a product that was cloned and se- 246 EALERISSGDFDVLITTAQFLVHRVEDLEKLNFDLILVDDVDAIIRGTGR 295 rgyB quenced. The probe was identified as targeting rgyA since it 181 EADEGLQSKTFD1 .RFLIDNIDQISSTSYQFL DVTALKS.SK 229 rgyS encoded the other 7 aa found at the amino terminus, as well 296 NVDRVLRVAGLEQEEIDSAYRLATLRRRYYSLRDWLRSLEDRGDKRAERV 345 rgyB as amino acid sequences strongly conserved near the active- .::1:: 1:...: .:::.1:...:..:. site tyrosine of type 1-5' topoisomerases. The rgyA probe was 230 SAQAILKLLGFTPSDQD ...... KIKESLKKYRENTQKNEQN. 265 rgyS used to screen 5000 plaques of an amplified M. kandleri A 346 REELREVEREIEELEELLKRVKKERDLARIVFM. SATGAAAPSRRLAVVRE 395 rgyB 1: :.:::1:I .:.1:: 111 .1.::. library, and 5 separate A clones were isolated. All 5 clones 266 ... EYIFEEIDKIRKDRLASKTYVFS .SA.... LNRSNPILTS 301 rgyS shared similar restriction fragments which hybridized to the 396 LFDFEVGAGGEGLRNIQDIAVISEPSPEA ... VERIVRKAGVKGGLIFV 441 rgyB Probes to not to A 1.1...:1.: 1: ..:. .:::I .111111 rgyA probe. rgyA did cross-hybridize clones 302 carrying rgyB, and probes to rgyB did not cross-hybridize to A LVGFKPGSSVIYIRKVYDMYVKQPDKEQETFNLIKSLLHRLG.DGGLIFV 350 rgyS 442 PQRLPGEKKAREIVEELAEHLRSSG IEARAIHAGTPAEEREEAIDGFSEG 491 rgyB clones carrying rgyA. .1 .1.:. 1..1 1.:::. 11..:. 111: Recombinant DNA Methods and DNA Sequencing. For 351 P....VDKKQEYIKRLQSEL.SNEFNVAAITSTSATK ..... IDDFANG 389 rgyS subcloning, competent E. coli strain DH5a (GIBCO/BRL, 492 DVDVLVAVASPYGVIVRGLDLPQAARYAVFYGVPRQRIRLTPREEDLKDP 541 rgyB Gaithersburg, MD) was transformed as described by the 390 EIDVLIGSATHYGI_LVRGLDLLPWRVKYSIFVGIPKFKFRLGEKVN ..... 434 rgyS supplier. E. coli LE392 carrying bacteriophage P1 was used as 542 TYVASALSNLARLLDDRRARSRLEGVAGRLWRIIRRGTWIRERLEEAVEP 591 rgyB a host for of A Restriction T4 .1.111. 1:: .: :1 :1. 1:.:: . growth phage (23). enzymes, 435 ...... LLTLSRLLS ...... YISRRVKDKIRR DNA , T4 polynucleotide kinase, and calf alkaline phos- LIALITKDQEVI 467 rgyS phatase (New England Biolabs) were used as described by the 592 LSLNTLMKLAKRDPEDIAEQLDVDRWLARHVQTLAEGVRELTRLLGDPDR 641 rgyB vendor. Genomic M. kandleri DNA was isolated as described 468 LSPAALTMLSVQAKEGKLED ...... SILLKAYDLLNKYLSNQNV 506 rgyS (24). A partial Sau3A1 digest and pre isolated arms of A 642 VKALAEEATTVAVYEEGEEAYLEVPDLRTYIQASGRVSRLFAGGVTFGLS 691 rgyB EMBL3 (Stratagene) were used to construct a A library (23). 507 LKRISE.... IGDFVLSPDNDILIPDYLTYVOASGRTSRIYAGDVTTGLS 552 rgyS Radiolabeled probes were prepared by random labeling using 692 FVLCPEDERELRTLNGLIRRMSYTYGSEFEWRSYPKSLDMKEIGLELKEI 741 rgyB 1 = [a-32P]dCTP (3000 Ci/mmol, NEN; Ci 37 GBq). rgyB was 553 ILLVD .... DFNLFRLLNKKLQYILD.DIQWR ...... ELDVEKWTA 588 initially cloned on three DNA fragments: a 2.7-kb EcoRI- 742 SDEELEELVRKVDEDRERVRKVL.AGELKPE 771 rgyB BamHI fragment, a 3-kb BamHI fragment, and a 2.25-kb :1 I:.:I:.I::1:I:S:.1: .1:: 1. BamHI fragment. rgyA was cloned on two separate DNA 589 GDVEIKNLISKINEERNEISKLKNEGNVAPA 619 rgyS fragments-a 1-kb EcoRI-BamHI fragment that would encode FIG. 1. BESTFIT (25) alignment of the helicase-like ATPase do- the amino terminus and upstream sequences and a 1.8-kb mains of RgyB and S. acidocaldarius RgyS indicates that these proteins BamHI fragment that was expected to carry most of the gene. are 30% identical and 57% similar. Identical residues are indicated by pGEM-3 or pGEM-3Zf(-) (Promega) pBluescript II SK(-) vertical lines. Similar residues are indicated by dots. Cysteines that are (Stratagene) was used to subclone smaller DNA restriction part of the conserved tetracysteine motif are indicated with stars. Motifs common to nucleic acid helicases are overlined and underlined. fragments which were then used as templates for DNA se- Amino acid sequences used during the design of degenerate synthetic quencing. The Wisconsin Genetics Computer Group program oligodeoxynucleotides for PCR amplification of probes have arrow- was used in the assembly and analysis of the DNA sequence heads above (based on the amino terminus) or below (based on (25). conserved motifs) the sequence to indicate orientation of the primer. Downloaded by guest on September 30, 2021 108 Biochemistry: Krah et al. Proc. Natl. Acad. Sci. USA 93 (1996) tions, varying from 4 to 14 aa in length, not found in RgyS, as a methionine start codon in E. coli. Upstream of the rgyA while RgyS has two insertions of 4 aa not found in RgyB, TTG start codon is an A+G-rich sequence which has strong indicating a considerable divergence in primary structure similarity to ribosome binding sites. Collectively, these data between these two thermostable enzymes (26, 27). indicate that RgyA is encoded by a separate gene and is not the As observed with RgyS, the ATPase helicase-like domain of product of the proteolytic cleavage of a larger precursor protein. the protein is located in the amino half of the molecule, with RgyA has considerable sequence identity and similarity with E. the topoisomerase domain making up the second half of the coli topoisomerase I (36% identity, 56% similarity) and S. aci- molecule (15). Beginning at aa 777 of RgyB, there are strong docaldarius RgyS (35% identity, 57% similarity), particularly in similarity and identity to the sequences of type I-5' topo- the region flanking the putative active-site tyrosine (Fig. 2B). , of which E. coli topoisomerase I is a prototype The Topoisomerase Domain Is Divided Between the Sub- (Fig. 2A; see Fig. 3). The similarity between RgyB and units. The conserved type I-5' topoisomerase domain has been topoisomerase I extends through approximately the first quar- found as a contiguous polypeptide in eubacteria (30, 31), ter of topoisomerase I; similarity abruptly ends after residue archaea (6, 8), and eukaryotes (9). By analogy to the domain 210 (Fig. 2A). structure of reverse gyrase from S. acidocaldarius, we had Cloning and Sequence Analysis of rgyA. Sequence analysis expected the division in the domain structure of M kandleri revealed an ORF encoding a protein of 358 aa and 41.7 kDa, reverse gyrase to occur between the conserved helicase-like in reasonable agreement with the molecular mass of 50 kDa domain and the topoisomerase domain. Instead we find the estimated from gel electrophoresis (18). The coding sequence topoisomerase domain of this enzyme to be shared between of the ORF agrees completely with the microsequencing the two protomers. Comparison of strongly conserved se- analysis of the amino terminus of RgyA, except that the codon quence motifs in the type 1-5' topoisomerase domain with for the first amino acid is TTG (normally coding for lysine), RgyB and RgyA demonstrates that all motifs of this enzyme instead of the expected ATG codon for methionine. Micro- family are present, and there is no duplication of the motifs on sequencing analysis of RgyA very clearly showed a methionine the two protomers (Fig. 3). Amino acids 780-1116 of RgyB in the first cycle of the Edman degradation (data not shown). correspond to aa 4-211 of topoisomerase I (Fig. 2A). Amino Directly preceding the putative TTG start codon of rgyA is a acids 42-358 of RgyA correspond to aa 283-582 of topo- TAA stop codon, indicating that TTG is the first codon of this isomerase I. Amino acids 1-41 of RgyA have similarity to aa ORF. In a recent study of 1055 translational starts for E. coli, 212-282 of topoisomerase I, but the exact point where simi- 13 use TTG (29). Thus, there is precedent for TTG being used larity begins is difficult to determine since the sequence of this A Ft. ulli.lt 1. 1 +1 .xvB '"O ALMIVESPNKAirmI.ILaSI FSt. RPSrRt 1 ic-g. YE.--taa (iGIhLt7VvAI, {GHVi DL. EpGVhGV1i.i 4 ' LLVVESPNKAKTI';S-S .RSt?.r 'YE. 1 -LVt A.. 4HV,DLt:. Tt,; DiGI', GV. i AL'IVESPaKAKTII.;.Y L ...... GGHir L:T.: ...... -

P.jyf(1.yF-*t.... 'WVP rnY'.11'C, G .vGS E!';.PNC;--.j.evE.Q;1:IIL*SI ...RALAsEA D,.I7LIGITDPD o RyS I. FIP i.YnsikCEi nhelFt- ff -S x' PCuIt- KV -lLKS n.vLRrtLAvF.A Dc:VLIGTDPD T~A81'.t<. ;.t Oh'kS. -.i -- t ;-' t -l k'0;,IkzDE i 5aL .^r.Lt5irC D': thEt-;T:t} ;' 1.eEVII,jk,K, ,Vi 'Lk: IIA- kA Dho'&TD-JD I I H Ux F I lu1]in in. 4. + .,i B TEGEKIgjWDvi Fr.-bLGvt t AIf VyRtEFHEVT RLrIWs-:ALkee.s (;kNVdag? V,-iAQilRRVa DRWIGFsLS. . PRgy. TEGEKIAWDL YaLI .21 ItRaEFHEVT RkAIIQAiNqPr- EFNV. .ni VIJsQIvPRI& DRWIGF.`LS.- 2 -EGEa-IAWhL z-.G liA LrR.'FnEIT :..AIi,QAFN;:Pj GEL-NI.. . iR VnAQqaRRfni DR7VWGYir,OZSr J.

F3'j" f L,.ijl ifk1i -:ik.j ii1. I ri(-3i'; ,-I 2.2 , iIf 1. i14;

gtyll J:''."i;' .tii .1. -1if r: rkl1q}iii i 4ajvRPEI *iT-.'.1 Svrip' ..i' lNj., epitI LSAGRVQ TPVLcgWIIDRaREY 111( Fyy.. .: .WPEyk.:r.s;. sSn; _:t Ps-X Nk:l.t.E. LSAGRVQ TPVLsWIVDR,EY27N' ...... '' ...... LSAGRVQ Is;'tav iVVEReRE iI i _ 1

EqYB riIy"i,t -<*e.iddvtA+i.+ akvr'n1 ie I aite}cleat igrf.te : f 1186

P.e,%, l<;y 1; g??k< <*yiglmnlL.l-la (Tv Y.; E%11-

B FtgyA 1 mrnratl RIrr.rpVAes TylSlrgaKa EVVR ...... VEREErE_.. 37

RgyS &7'i qrnks RVY...YgkIdql qdiviyvpKq DgVRkNsKiv vvfNe ...... InqIEeEf 924 TopA 21.2 kafvp eeFWevdAst TtpSgealal qVthqNdKPf rpvNkeq'ca avsllekary svlEREDkpt 276 IDomain 3 +I + + ++* I+++ FIG. 2. Comparison of the topoisomerase I RgyA hPKP. IPFeT gTMLQAAtrR LrLSservMq LAQDLFEgGL :TYHRTDSTR VSeEGkrVrR dYIraNFc.. 103 domain of RgyB (A) and RgyA (B) to similar RgyS gP_P.. IPYtT dTfLLsdsnrf FGLSapeTMr iAQDLFE1GL ITYHRTDSnR ISntGlsVae nY1kDvLGdK 99 TopA tsKPgaIPF4iT sTLqQAAstR LGFgvkkTMm MAQrLYEaGy ITYmRTDSTn lSqDaVnmvR gY:sDNFGkK 346 domains from S. acidocaldarius RgyS and E. coli topoisomerase I. Identical and similar amino RgyA .peDYnPRtW epeaEhveGA HEcIRPTrP. aDAEeLRtMV rEGaiqttvt LTshHLrLYD LVFRRFVASQ 171 acids are in uppercase type and shaded. As RgyS YtniFkPRsW gd.....gGA HEgIRPTkP. IDvEqLRlLI eEGelelakr LTnnHFKvYD iIFRRFsSQ 1056 revealed by the crystal structure of E. coli topoi- . TcpA Y:pE.sPnqY ask.EnsqeA HEaIRPsdvn VrmAEsLkdMe aDaq ...... EKYq LIWRqFVAcQ 402 somerase I, conserved amino acids that are di- ID}main 2B rectly involved in coordinating the active-site RgyA MkPAKV...... LYqEav LEVevkgvpV ... aElelsg vleiVEPgFt KvLtEydLPA ygirEtpeLe 230 of subdomain 3 or that form a tgyS iiPlKVrkeilvkieLYgEnk kEkinsnql .. . iEVitgi tLpgbDteis K.Fayvpvrn VsrsvAerLk 1122 tyrosine ( ) TopA MtPAK...... Ydstt LtVgagdfrl kargrIlrfd gwtkVn.PaLr KgdeDriLPA VnkgDAltL. 461 second "shell" around the coordinating amino acids (+) are indicated (28). Amino acid se- IDomain 4B + + RgyA EgdR ..... L ElgavevLer heeypYdlpsE LyeDMreRG1 GRPSy.zY.tV eklERRGVy EvpqrRWif P 295 quences used for the design of degenerate syn- RgyS ElgRs+iptdF sIeisnsFiK stvn]Yt Vm EMknkkI3U gulrt1. YV1 EslktkkLiP 1192 thetic oligodeoxynucleotides for PCR amplifica- TopA ...... v EltpaqhFtK ppa. rFsIeKs I3cVELekRGI 6+lV't-YslT sT.Iq;- YV. .rvenRRFYa 519 tion of probes have arrows above (based on the amino terminus) or below (based on conserved pEP . RgyA TtrGEaVyEY LsthYeRF.VS EET.RdLEer MDa,VAlGkAXk Yq!emEkIk1l tLer7vEpD 358 sequence motifs) the sequence to indicate orien- RgyS TrLGveVnkY DnP.NgREVS 3rXRklql fibm9eaGqek e kqfieE Eineir .... . 1248 TopA ekMOGEiVtDr LeEFireLmn ybfTaqMEns LqVAnheRZ ka DhPFas DFtqqllDkaE kDPeeggmrp 589 tation of the primer. Downloaded by guest on September 30, 2021 Biochemistry: Krah et al. Proc. Natl. Acad. Sci. USA 93 (1996) 109

1 4A 2A 3 2B 4B Iw rzz_#_ZZZZZ Top I I. 1 WZZZZZrA y RgyS M--l rA I EMEM y

RgyB --=-|

RgyA m MINE I y FIG. 3. Schematic alignment of type 1-5' topoisomerases. The top line shows which parts of the linear polypeptide chain are found in the subdomains of the E. coli: topoisomerase I (Top I) crystal structure. Amino acids forming subdomains 4 and 2 are not contiguous and are labeled 4A/4B and 2A/2B, respectively. Conserved sequences indicate that the order of this subdomain structure is maintained in RgyS, RgyB, and RgyA. Subdomain 4A for RgyS and RgyB has an insertion when compared with topoisomerase I (indicated with a solid line). FIG. 4. Regions ofE. coli topoisomerase I that are similar to RgyB Subdomain 2A for RgyS and RgyA is smaller when compared with and RgyA. The protein structure of topoisomerase I was depicted with topoisomerase I. The nonconserved amino acids at the carboxyl the RIBBONS computer program. RgyB has similarity with aa 4-211 of terminus of RgyB are indicated with a question mark and cross- topoisomerase I. This region is indicated in cyan. Amino acids 182-190 hatching. Y, active-site tyrosine. of topoisomerase I are indicated in white and form a loop on the surface of the molecule. This loop corresponds to a large insertion of region is poorly conserved among type I-5' topoisomerases amino acids in RgyB, 977-1096 (see Fig. 2A), and lies between helices (Fig. 2B) (28). F and G of topoisomerase I. RgyA has similarity with aa 212-582 of The crystal structure of a large part ofE. coli topoisomerase E. coli topoisomerase I. This region is indicated in yellow, with the I, representing this conserved region of the type 1-5' topo- active-site tyrosine and two flanking amino acids on either side shown isomerases, has been solved by Lima et al. (28). The striking in magenta. feature of this structure is the arrangement of four distinct subdomains which enclose a hole large enough to accommo- matographic steps during purification from M. kandleri (18). date a double-stranded DNA molecule, as well as a smaller The C-terminal amino acids of RgyA, 253-358, and the channel which is thought to bind single-stranded DNA (Fig. 4). C-terminal amino acids of RgyB, 949-1116, contain motifs The polypeptide chain follows a unique path, with two of the found in subdomain 4 (Fig. 3). In the E. coli topoisomerase I subdomains formed by noncontiguous regions of the polypep- structure, helices F and G (motifs found on RgyB) form an tide (subdomain 2 and subdomain 4). On the basis of amino "elbow" around which the other amino acids of domain 4 acid sequence similarity, this subdomain structure is probably (motifs found on RgyA) wrap themselves to form a base for maintained in M. kandleri reverse gyrase (Figs. 2 and 3). Using domains 1 and 3 (Fig. 5). RgyB has an additional 220 aa not the crystal structure ofE. coli topoisomerase I as a model (28), found in other topoisomerases interspersed with conserved we indicate in Fig. 4 which analogous regions of the structure motifs of subdomain 4 (see Fig. 3). Based on the proximity of are located on RgyA and RgyB. Regions encoded by rgyA are the nonconserved sequences to the conserved motifs of sub- yellow, and those encoded by rgyB are cyan. domain 4, we speculate that these amino acids are involved in The strongest identities between topoisomerase I and RgyB stabilizing the heterodimer. These residues may also help form and RgyA are found in sequences making up subdomain 1 and some of subdomain 2, which is the least well-conserved subdomain 3. The interface between these domains in the subdomain of the type I topoisomerases (28). The spacing of topoisomerase I crystal structure is extensive, with 24 aa conserved motifs for subdomain 4 in RgyA, in particular the directly or indirectly involved in coordinating the active-site amino acids directly interacting with the active-site tyrosine in tyrosine. When amino acid alignments of RgyA, RgyB, RgyS, the crystal structure (T496, A498, and E547), suggests a similar and topoisomerase I are compared, all but two of the amino spatial organization with the helix F and G-like motif of RgyB acids that make up this network are strictly conserved (Fig. 2A; (Fig. 5). A367 and Y312 of topoisomerase I are not conserved). We The reverse gyrases of M. kandleri and S. acidocaldarius interpret this strong conservation to indicate that for all three differ not only in subunit structure but also in large portions enzymes the interactions between subdomains 1 and 3, as well of their sequence. For example, compared with the S. acido- their interactions with DNA, are very similar. Similar substrate caldarius enzyme, the M. kandleri reverse gyrase has many requirements and cleavage specificities with respect to the insertions in its ATPase domain. As these two organisms topoisomerase reaction support this hypothesis (6, 7, 18, 32, represent the two major phylogenetic branches of the Archaea, 33). the and the (35), it will be Adaptation of Domain 4 as a Dimerization Domain. Al- interesting to see whether these differences are maintained in though subdomains 1 and 3 are on separate protomers, the other members of each branch. Another question is whether amino acid sequence within these subdomains shows no ob- other methanogens that are not retain a vious alterations, such as insertions or deletions, that are heterodimeric structure in other type 1-5' topoisomerases. unique to this enzyme. Thus, any specialized interactions Reverse gyrase-like enzymes might be present in eukaryotes unique for the maintenance of the heterodimer must be found and nonhyperthermophilic prokaryotes. For example, a sec- elsewhere in the molecule. The RgyA/RgyB heterodimer is ond-site suppressor (SGS1) (36) for a Saccharomyces cerevisiae very stable, being maintained through four different chro- strain deficient in topoisomerase III (9) is a protein with strong Downloaded by guest on September 30, 2021 110 Biochemistry: Krah et al. Proc. Natl. Acad. Sci. USA 93 (1996) 9. Wallis, J. W., Chrebet, G., Brodsky, G., Rolfe, M. & Rothstein, R. (1989) Cell 58, 409-419. 10. Kikuchi, A. & Asai, K. (1984) Nature (London) 309, 677-681. 11. Slesarev, A. I. (1988) Eur. J. Biochem. 173, 395-399. 12. Bouthier de la Tour, C. B., Portemer, C., Huber, R., Forterre, P. & Duguet, M. (1991) J. Bacteriol. 173, 3921-3923. 13. Marguet, E. & Forterre, P. (1994) Nucleic Acids Res. 22, 1681- 1686. 14. Bouthier de la Tour, C. B., Portemer, C., Nadal, M., Stetter, K. O., Forterre, P. & Duguet, M. (1990) J. Bacteriol. 172, 6803-6808. 15. Confalonieri, F., Elie, C., Nadal, M., De La Tour, C. B., Forterre, P. & Duguet, M. (1993) Proc. Natl. Acad. Sci. USA 90,4753-4757. 16. Gorbalenya, A. E., Koonin, E. V., Donchenko, A. P. & Blinov, V. M. (1989) Nucleic Acids Res. 17, 4713-4730. 17. Schmid, S. R. & Linder, P. (1992) Mol. Microbiol. 6, 283-292. 18. Kozyavkin, S. A., Krah, R., Gellert, M., Stetter, K. O., Lake, J. A. & Slesarev, A. I. (1994) J. Biol. Chem. 269, 11081-11089. 19. Shibata, T., Nakasu, S., Kumiko, Y. & Kikuchi, A. (1987) J. Biol. FIG. 5. A view of subdomain 4 in the crystal structure of E. coli Chem. 262, 10419-10421. topoisomerase I generated with the MOLSCRIPT computer program 20. Mizuuchi, K., O'Dea, M. & Gellert, M. (1978) Proc. Natl. Acad. (34). Motifs found on RgyB are black; motifs found on RgyA are white. Sci. USA 75, 5960-5963. Helices F and G are labeled, with the nonconserved amino acids 21. Liu, L. F. & Wang, J. C. (1987) Proc. Natl. Acad. Sci. USA 84, between these helices shown as a dotted line. 7024-7027. 22. Findlay, J. B. C. & Geisow, M. J. (1989) Protein Sequencing (Oxford Univ. Press, New York). similarity to E. coli RecQ, a DNA helicase (37). More recently, 23. Silhavy, T. J., Berman, M. L. & Enquist, L. W. (1984) Experi- use of the yeast two-hybrid system demonstrated an interaction ments with Gene Fusions (Cold Spring Harbor Lab. Press, Plain- between S. cerevisiae topoisomerase II and SGS1 (38). Thus, view, NY). the direct interaction of topoisomerases with helicases may be 24. Rinker, A. G. J. & Evans, D. R. (1991) BioTechniques 11, 612- a more general phenomenon, with functional consequences 613. that are not yet clear. 25. Genetics Computer Group (September 1994) Program Manual of the Wisconsin Package, V (Genetics Computer Group, Mad- We thank George Poi and Dan Camerini-Otero for assistance and ison, WI). 26. Hoffman, M. (1992) Science 257, 32. access to automated DNA-sequencing equipment, and members of our laboratory for helpful and stimulating conversations. We Al- 27. Rivera, M. C. & Lake, J. A. (1992) Science 257, 74-76. thank 28. Lima, C. D., Wang, J. C. & Mondragon, A. (1994) Nature (Lon- fonso Mondragon and his laboratory for providing coordinates of the don) 367, 138-146. E. coli topoisomerase I crystal structure prior to their submission to the 29. Barrick, D., Villanueba, K., Childs, J., Kalil, R., Schneider, T. D., Brookhaven Protein Data Bank, and Fred Dyda and Alison B. Lawrence, C. E., Gold, L. & Stormo, G. D. (1994) Nucleic Acids Hickman for assistance and instruction in the use of computer Res. 22, 1287-1295. programs to manipulate the Protein Data Bank file. A.I.S. wishes to 30. Tse-Dinh, Y. C. & Wang, J. C. (1986) J. Bio. Mol. 191, 321-331. acknowledge the support of James Lake during this work (National 31. DiGate, R. J. & Marians, K. J. (1989)J. Biol. Chem. 264, 17924- Science Foundation Grant DEB 9119801). 17930. 32. Kirkegaard, K. & Wang, J. C. (1985) J. Mol. Bio. 185, 625-637. 1. Gellert, M. (1981) Annu. Rev. Biochem. 50, 879-910. 33. Kovalsky, 0. I., Kozyavkin, S. A. & Slesarev, A. I. (1990) Nucleic 2. Wang, J. C. (1985) Annu. Rev. Biochem. 54, 665-697. Acids Res. 18, 2801-2805. 3. Luttinger, A. (1995) Mol. Microbiol. 15, 601-606. 34. Kraulis, P. J. (1991) J. Appl. Crystallogr. 24, 946-950. 4. Wang, J. C. (1971) J. Mol. Biol. 55, 523-533. 35. Olsen, G. J., Woese, C. R. & Overbeek, R. (1994) J. Bacteriol. 5. DiGate, R. J. & Marians, K. J. (1988) J. Biol. Chem. 263, 13366- 176, 1-6. 13373. 36. Gangloff, S., McDonald, J. P., Bendixen, C., Arthur, L. & Roth- 6. Forterre, P., Mirambeau, G., Jaxel, C., Nadal, M. & Duguet, M. stein, R. (1994) Mol. Cell. Biol. 14, 8391-8398. (1985) EMBO J. 4, 2123-2128. 37. Umezu, K., Nakayama, K. & Nakayama, H. (1990) Proc. Natl. 7. Nakasu, S. & Kikuchi, A. (1985) EMBO J. 4, 2705-2710. Acad. Sci. USA 87, 5363-5367. 8. Slesarev, A. I., Zaitzev, D. A., Kopylov, V. M., Stetter, K. 0. & 38. Watt, P. M., Louis, E. J., Rhona, H. B. & Hickson, I. D. (1995) Kozyavkin, S. A. (1991) J. Biol. Chem. 266, 12321-12328. Cell 81, 253-260. Downloaded by guest on September 30, 2021