FEBS 15051 FEBS Letters 359 (1995) 1 5

Hypothesis A common topology of catalyzing ATP-triggered reactions Masasuke Yoshida*, Toyoki Amano Research Laboratory of Resources Utilization, R-l, Tokyo Institute of Technology, Nagatsuta 4259, Yokohama 227, Japan Received 11 November 1994

[5]. Indeed, the recently unraveled crystal structure of mito- Abstract A fold, six parallel i~ strands surrounding the chondrial Fi-ATPase (MF0 [6] clearly showed that the folding central a helix, is likely to be a common structure in protein topology of the ATP binding domain of MFI-fl subunit is al- families known to have a typical set of binding consen- most the same as that of the recA protein, and that the essential sus sequence motifs A and B and to catalyze ATP-triggered reactions. According to this ATP-triggered protein fold, the con- Glu (Glu j88) of MF~-]3 subunit above described is really located served Glu (or Asp), which acts as a general base to activate a at the right position to act as a general base. Here, we propose water molecule for an in-line attack of the -/-phosphate, is at the that proteins containing a typical set of Walker's motifs A and exit of the second/3 strand. The fifth i~ strand may be involved B have the conserved Glu (or Asp) between the two motifs and in propagation of conformational change triggered by ATP hy- that their topologies of the ATP binding domains are probably drolysis. all common.

Key words." ATP binding domain; ABC transporter; MDR folding; SecA folding; Lon protease; Proteasome 2. Conserved 'catalytic carboxylate' in various protein families

When the sequences of the region covering motifs A and B are aligned for several protein families which have a 1. Introduction set of typical motifs A and B (Fig. 1), it becomes clear that occurrence of a Glu or Asp (Fig. 1, catalytic carboxylate, E) at Since Walker and his colleagues first reported two consensus 24 + 2 residues from the Lys of motif A (Fig. 1, motif A, K) is sequence motifs A (GXXXXGKT/S; X can be varied) and B a common feature (Fig. 1, distance of E-K). These Glu's (or (ZZZZD; Z is a hydrophobic residue) associated with nucleo- Asp's) are well conserved in each of the protein families. Fig. 1 tide binding [1], a large number of proteins have been shown includes a variety of proteins which are supposed to catalyze to have these motifs. In all cases, motif A resides on the amino- ATP-triggered reactions and are expected to undergo confor- terminal side relative to motif B and the distance in the primary mational change upon hydrolysis (or binding) of ATE There- structure between two motifs is 50-130 residues. During the fore, it is highly likely that the Glu's (or Asp's) indicated in Fig. study of essential residues of the FI-ATPase fl subunit, which I play the same functional role, activation of a water molecule has a typical set of motifs A and B, we found that a conserved to attack y-phosphate of ATP in each of proteins as is proposed Glu located at 26 residues from the Lys in motif A is essential for the recA protein and F~-ATPase fl subunit. Interestingly, for ATPase activity; the elimination of the carboxyl group from F~-ATPase a subunit, the overall amino acid sequence of which this Glu by specific chemical modification with dicyclo- is similar to the fl subunit, has Gin instead of Glu at this hexylcarbodiimide or by mutational analysis always resulted in position. This is quite reasonable since it has been well estab- almost complete loss of ATPase activity [2,3]. The modified or lished that the ATP-binding site of the ~ subunit does not mutated retained the ability to bind substrate but were participate in ATP hydrolysis [7]. Similarly, TBP1, in which Glu unable to catalyze even a single catalytic turnover. On the other is replaced by Gln, can probably bind ATP but hydrolyze it hand, based on crystallography of the Escherichia coli recA only very slowly, if at all. Examination of the primary and protein which also has a typical set of motifs A and B, Story tertiary structure of nucleotide binding proteins which do not and Steitz proposed that Glu 96 of the recA protein is in a have a typical set of motifs A and B, such as ras p21 [8], EF-Tu position to serve as a general base to activate a water molecule [9], -~ [10], Gi~l [11], adenylate [12,13], Hsc70 for an in-line attack of the y-phosphate during ATP hydrolysis ATP binding domain [14], actin [15], and [16] revealed [4]. This Glu 96 is located at 24 residues from the Lys in motif that such a catalytic Glu (or Asp) is not present at the corre- A. Although there is no overall amino acid sequence similarity sponding position. Although Glu 2°3 of transducin-c~ is assigned between the F~-ATPase fl subunit and the recA protein, it as a catalytic residue, its position in the primary structure is appeared more than coincidental that the catalytic Glu's are different from the Glu (or Asp) described here and the architec- located at almost the same position in both proteins and a ture of its ATP binding domain [10] is not similar to those of similar topology of their ATP-binding domains was indicated the recA protein and MF~-fl subunit.

3. Putative 'catalytic carboxylate' in ABC transporters *Corresponding author. Fax: (81) (45) 924-5277. In members of the ABC (ATP-binding cassette) transporter Abbreviations: MF~, FI-ATPase from a bovine heart mitochondria. family [17] which have a typical set of motifs A and B, the

0014-5793/95/$9.50 © 1995 Federation of European Biochemical Societies. All rights reserved. SSDI 0014-5793(94)0 1438-8 motif A Catalytic cIrboxyloto motif B K E D :'osition Hstance Function tee.A f E D E-K D-E E.mll RE VII Y::~::~,M i::~I, T L T L ~VI A A A ~ R Ii'I, T C A FID ,HH A LD ~ilY AR g L~::V 2 N 144. u Ulmd~eof ssDNA by ds DNA. B.sub~lis R I IE V Y~::~::~M M:.~TT V A LH A I AE V~Q-RT M AF ID ~I~ A LD~::V Y A~K Li~::V &LVRSC3AVD I VV% VAALV • ~ 14.1 es Medhms DNA r~ombinafion. M. tul~/'xuloeis R V I E I y::~::~M ~::'~T T V A L H A V AM A ~ A Ai~!~::VA A F I D ~ A LD~iD Y AKK LI~iV (L IRM~iALD IVV VAALV I+.ATPmm %TPsymbase catalyticsubun~ z,-~ ~dm ~z [::~iL Fi~::~iA::~iVi~l~v L I MS L IMM V AK All~::~iY I V F A~i::V::~I~ T R E~:.M D L YHI '~RD ~I~!QD V L L F I FRFT 1ms M N q~Ih ~:x V~::L ~i~::~iX:~::v::~ v L ~ ~" L I.. Z A~ ~"~::~::V" V F~:~i~vi~l~ = ~"~:.x D ~. ~ ~" '~RD VN.I~D V L L F IFRFV 7~ 2o4 5 N E.coM ~ ~::L ~i~::~i^::%::V~il~ vx ~" L I*~. I A Z" "~::~iY S V F ~i:V::~l~ = ~.":~::X ~ ~ ~'"" "•DE -~iRD .V L L F' I YR YT aa 181 M BaclllusPS~ ~:." Ii~::L F~::~I^::~i::::V~I~ V L I 0" L I "W I AO" "~i~iI " V F ~::~VI~ • X,." ~!~ ~ L Y" • "•DE Q~:.Q D~iL L F ] I FR FT ms Ires 8a Wl-a ATPsymlmsenon-catdy~ mamit BIdi~ PS~ 0]1.• L I Ii~iD~ QT]~T a VA IDT I I M'~]C -D~N M I C I Y VA I]~0KR 8T VAT VVRT fFM IM~iXH VLVV I~ LMX~A 7I ~o~ ~31 Vx 6~k ibunlt Vacuolsrtype ATPase cstalylicsubunit Carrot ~iT C A I:.~:.::~:.A F::~i:.Ci~IT V I I 0 A L S K Y I! li B D T V V Y V~iC ~{Ri~IM E M A u V L M D F~IO rFRD M~iYM VS M M AM8 T "~ WA 5 71 T.thermophilus ~iT h A li~ii~i~::F~i::Si~TV T 0 0 S L A K W S MAD V V V Y V~iC~]~iX • M T D V L V • F~iX ~'F~D ~::F~ V A L M,m~ ZTIRWA u S~ddi~akiarlus fFRDQ~iYD VL LV@ BTMRWA ms ~1I family FIII ~R l::~ilF Ai~:li~:Vi~a T L Mi~iiM AK~TIAD LM I.AL V~R:.~R~ VR.E F IEKD L~ ~F~DK~i0~ v~ F MM~S VT~VA 1'3 IH 287 31 ~eiium-s~c~c c~x_L~rt - . sp~ ~R M::~iI P AIADCi~I~ F L MMM LI I X H Bi~iAD.... I Y W VT~TV~YLKW "FR I E~iHK VA L F I~S LTR Y A (18 1ms m --~----~umacevresenlation of Ipt Antigen. ~utaflve ATPa~ CDC48 (N) Ri~ V L M Y::~i~:~:::~:T::~TL M A R A V A M • T -::~:.A F F F L I M ~::~I V M • K M A~:• I • 8 M L • |g AEKM A~IA I I F I [A ~ 315 31 Cell cycle end cell divinion eontml. CDC4S (C) Ai~] V L M ¥::~i~::~::i~iTi~r~TL L A • A V A T • V - IA N F I S V • ~ [] L L S N W Y~ • S • $ M I • )K AR A A A~%iT VVF I [ A 31 pY7 (N) • ~::I L L Y::~::~ii~:::~::TI~TL I A R A V A N • T - ~:iA F F P L I M ~i~ [] I M • K L A.~::EI • I M L It |I AIKM A~iA,.... I I F ] [ A 31 Not Imown. Shows ATPIse amivity. p~7 (C) ~i~::V L F Yi~i~ii~ii~:iC~T L L A • A I AM • C -'Q AM F I I I K ~:~ ILL T M W F~• I • A M V • )g A R ~ A A~iC V L F P [ A ~ 4rFf NSF (N) i~ I L L Y ~i~::~::~C::~T L L A R ~ I ::~I M L M A I% • ~::i V V N ~:~ [] I L N K Y ~!m • • A M I • t L~]AX i~::LH I I I ~ [C 74 "~'e7 3~ .~ ~I Vesicle-medlated mmspo~ NSF (C) V'I' V L L • ~::~::~ ::u I ::~lT A L A A g I 'A'• • I - M F ~::iF I • I C ~~ I K M I ~:: F BE T A • C Q A M )D A YK !~..L.S C v V ~ LL ;87 ~ 611 P~I (N) ~iiA I I L D::%IKQI~::II~T R L L•• L IN• VE KDHM I F V• Y ~[~CETLH • TIM LDKT~ I F c Y W Y~i~!9 L I V I L F 4ms 8~ rJ u Involved in l~xi~e biosynth~i$. Put (C) ::~iI L L Y::~Y~::!~::C ~IT L L A I A V A {~ Q C - ~::L M F I I V g ~ II L M g F I~:.A I • ~ M I • IRA~ VK~::IC I L F P [ A '44 ~ 7117 seen8 (N) ~:.L L L r:~i~::~i~I~ '. ~ A~ • Ii~i= ~ L, A~ • ~::• ~ V,~ I I L S ~ Y V~::S S • •, ~ IK~IEII LH I I I ~ IF IT 310 s~ts (c) V d '. L I ,i~:.~:-::A:.~:S~ A L A A • ~A L Z S ::~::i~i:.Fi ~ L I s ~:.~ [] L S ~: ~ s • '~ A • ~ A Y I tD~.¥Ki~::LN I LV 2 Lv I 12 FIsH K]~::V L M VI~::~::~::~T~IT L L AK A I A::~I• A-KV~!]F F T I 8~::81 F VEM F V~:. ~::A M R VR N 22O 2~1 " 31 Cell division protein. SUGI x :.~iv i L ~i~i~i~:~::::~~I • ~. L A*~ A V A~" • - D C X"Z I *~ V S ~::~ IL V O • Y ~~Z~I" *~ S V 317 31 Sui~teu a lrenm~vmor mutant MSSI • ::~iV L L F::~i~::]~::~::T~iT LC AR AV A•aT -D A¢ F IRV I~:'8 I~IL V~K Y V~:.R~:.Ai M V• L M Air =•K AC L I F ~ I'q ~ HIV Tm-medinmd mLmmCfivafio~ TBP! %LAKIK A~iS I I F ] [~ N ~ :M1 31 Inu=ac~s wKh HIV Tat trensa~iv au~. ~fll family Nlnl :M~::A YOD LD F VS YIv..L~i.D.V v II 43 12I u Nitrogen relation. MInD I I V VT ~=:~•i~i~::V~T T • • AA I ATI~iL AQ•~•T VV I'DF~I~]LRM LD L I M~I=• D Lx A~D FZ F z vcls~: :It ms Cell division inldbitor. ArIA (N) ~:i Y L F F Ti~::X ~::~ V::~I~T I I I C A T A IR L A • ~ ~ R V L L V I Tara 8 • V~i0 V F I Q T I l L I~:.AC TTI I AAF~E FT~iL L 11 46 131 ~- rs ATP driven As pump. ArsA (C) ~i L I M L M~iX ~::~WIT T M A A A I A V R L A D M:.~F D V • L T T B I~:: A A H L'I'M T L M ~iI L ~io ~4 447 FrxC £ T L A V ¥!~:'g ~:'~:" I ~l I T T i C M I I T A L A K ~::~ X • V L ~ I ~ CI~:.• U D I T F T L T ~ F L : LM AFDI YD V I L FIV L~:D V V I| ~ 124 ms Involved in chlorop~ll synthesis. lncC T L VT AMQK~i~ V~IT I T L VII L AFD F F• R::~iL • V A V I DLI~I~!N AI YT L •DF A C A L AN (~iF D V C L II~T A~iT L~ 144 231 .7 Pla~nid mainUmence and replication. SecA proton Prolein export ~ d~ membnme. E.coli • RC I AI M•Ti~T LT AT Li~i::AY LM A L T~I~!~iV M V VT VNIY L A~RD A•MM R:~i 78 P.lutherii D I K I A • M • T ~ I~J'~T L V A I Li~iiA Y L M A L I I~I~'IV H I V T V N[V~ y L A K R D I L I V~IR IIK V~M~::FI F A I I~VD VL 04 12~I ~os 78 I~o Protein Tnmsaiption taminmtion f~or. E.coli Ri~iL I V Ai~i~iX A~T M L L ~ N I A Q I I A Y N H ~::]) C V L M V L L I~E R ~i• • V T • M ~ R tLVImKEDVI I LLII ITRLA 31o I M u DuaA protein Binds 1o lhe origin of replication. g.eoH M:.~!iL F L Y::~i~:T i~:: L::~IT I! L L 11 A v::d:~li~i I M A R K::~:M A K V V Y M IIIa ~R F V Q D M V • A L "K, Y YRI VD A L L Ii~ I~ FFA 78 2O4 B.Iubtili8 Ni~:.i L F I Yi~i~:: V::~IL~I~T I! L M H A I i~::ll Y V I D H M i~::i A • V V Y L i S ~ F T • • P I • i I M .'RMRYRNVD VL L I~FLA 17 1~1 2'14

DuaB protein ...... Chrmnommmereplicatk)n factor. F..coli [ ARI m~!~iI~iL IM LQLMR ~.i 81 u~ S.typhhnuMum [ F R I m~::~:.LI L IM LQLMR HI 81 l.,on Preten~ ATP-depend,~ pro~,se L*. E.eoli i~i:: "r : c L V::~::~::~I::~::V ~" I,::~i~0 S ~ A • A ~ i~ia • Y V ~ ~ A L ~::~::iV RI~ A • Z ~.~i" n n ~ '~ "r 4 A K V~i:V K N~::L F L I~ I n • M I m ~ 422 S.cerev~ae g I I C F Vi~:.~::~::~::V~T M I::~i]'ItB I A • A LUIt K F F • F M V~::~i::MTIV h • I ](~iH R R T Y I Hjapleni 40 44O 478 B.br~vls ~ilI L C L Vi~.~i~i~iViW~T 8 L A R I V A R A Li~t • P V R I I L ~::~iiVR.~ A • I ~:H R R T Y V 4X ~ A~::T IU~iV F L : IDXLA M. Yoshida, T Amano/FEBS Letters 359 (1995) 1-5

motif A Catalytic cerboxylato Catalytic cerboxylate K o~n~dmo I (E~) :ndlde~ 2 ( F.~ MDRI (N) T v A L V::~:~. S !~C ~S T T V 0 ~' "0 R L Y D ~:T Zi~ M v S V D:~IQB I R T I N Q lg:~iV LF AT T I A~N IR Y MDRI (C) T L A L "/!.~!!S S ~ C ~S T V V O L 1, E R F Y D ~:: L AI~I:K V L L O!:~!!KI'~II K R L N QE::~iI LF:DC:S Ig~ I AY Cl~rR 0%[) T.I., A V AI;::S TI~iA~T S L T. M M I ~::~iE L w. !~S Xii~K I K I~ Si~iR I S F C S 0 Q F S W I M]P~T I ] I I F CFTR(C) i:~ORV~LL~:~RT~$1~$T LLS AFLR'LLN- TE!.~::EIO ID~VS W~S IT QKVF IFSI~::TFRKN LD~ HisP :.~D V IS I I~iS Si~::$!:~[~IST F LRC IN F LEK~ SEi~::S I VVN:~ilQ T INLVR H FN L W.:S:>:.II MT V L~N VME Ma~K :~ r v v F V~IS~ ei~S • ,. ,.n n I ~::~::,.~• I • S:~i~::D'. v I~X ~ ~ V ~i s Y ~ L ~i::a LS v~l~ ~s F VtsE DEll...... L LMDR T VY~N VAI RbsA ~R V M A L V~!W- N:~iA!~B$ T M MK V L T ~i I Y TRD ~T L L W L ~KI~T T F TI.~ • L N L I i~ii::!Q L T I A~ I F L HIyB ~::IgVIi~IV~RS~S~ST LTK L 10RFY I N!GI:0VL ID~LALAD QDNVL.LNRS I I~N IS L Nodl ~g C F ~::.LL ~i!~N ~ Ai~S T I ARM L L ~ M I S~ D R ~K I T V L D~'~E~:!!VI:~IIiSR A FDN L E:P::i:.EF T VR~N L L V

motif B Position Distance Function D K El F.2 D ~.t¢ ~t¢ ~-m

MDR1 (N) R~ ~ ~. S~i~ X Q R ~ ~ I AR A L V X~N~iX I L L 5,St, I A T S A L D T 433 45"7 486 SS5 24 S3 98 U Muhidrug resistance proteinl. MDRI(C) K~:~TQ LS:.~:Ii~QKQR IAIARALVRQ~" I L LIL, BATS.ALDT 1078 1100 1129 1200 24 S3100 71 (P-glycoproteinl) c~rR(I~ i~::~iIT L$~:~i~::QRAR I$ LAI~AVYKDAD LYLL, $i~iiFiG::!Y LD V 464 504 S~ • 40 - U Cysticfibmsistransmcmbrane CV'rR(C) :~C v ~.S.i~:,X~ ~.~C ~,~aS v~.SX~X ~ ~. ,.,., I~:~ii~:S AB LD~ 1~ 1276 1302 1371 26 S2 9S 69 CoIKlUclancelEglllator. HisP Y ~!iV~I L S ~:i~:i~~ (~ R V S IARALAME~EVLLF' I[I~!!T S A L D~ - 111 177 - e7 - ~ Histidin~ pcrmcas¢ component. Ma~K x [:~ixa ~ Si~i~ a ~ ~ v a Ii~R • ~ v ~iS v ~ '. ~, I~'i~:iiLS" LD~ 42 64 94 158 22 S2 04 . MaltoscpcrmcasccomponenL PstB S ~i Y S L S !:~ii~0 O O R L C I"A]R!:~:!I A I R::~::]g V L L L' g!~i!C S A L D~ U 71 109 ~ 22 e0 107 S0 Phosphate transport system. VlSE F~!I I Q L Si!~ii~iE0 Q R VigilI AR A v v N K:!~I:A V L L A , IC~:~IT~:~iST LDi) 41 ~ U 162 24 S7 07 64 Involvcdincclldivision. RbsA L V!~{IDL$I~!DQQ MVEI AK V LS FESKVI I , gi~iiTDALTD 43 67 u lU 24 SS 0g SO Riboscpcm-e,asccomponcnL HIyB Q ~IA:!~::LSi~i:~:iQ R Q R I A I A R A L V N N!:~iK I L I F, EAT S A LD Y ~ ~ ,~1 ~0 24 s3 s8 6o ExpoxtofhemolysinA. NodI R VS L L$!~::~MKRR LT L AR A L I ND~:" L L V e~iTT~ILD~: ~ 66 ~ lS~ 23 , ~7 ~ Acts in the nodulation process.

Fig. 2. Comparison of amino acid sequences of proteins of the ABC transporter family. In addition to the first candidate for catalytic carboxylate (El), the region containing another candidate of catalytic carboxylate (E2) is also shown. In CFTR(N) and HisP sequences, candidate Glu (or Asp) is not found in the first candidate region. CFTR(C) has a replaced residue (Lys) at the second candidate position.

distance between motif A and B in the primary structure is protein, which is made up of six repetitive fl-a structures, might 133-166 residues, much longer than other proteins listed in Fig. be a prototype fold*. 1. Although we can find Glu (or Asp) in a region 24 + 2 (Fig. 2, catalytic carboxylate candidate 1, El) from the Lys of motif A in most cases, higher conservation was observed for a Glu (or Asp) in a region 59 + 8 residues from the Lys (Fig. 2, The folding topology of the ATP binding domain of the catalytic carboxylate candidate 2, E2). There is a possibility MFI-fl subunit is the same as that of the recA protein except that, in the case of the ABC transporter family, some additional that there are two ~ helices between the f14 and f15 strands [6]. structure comprising 3241 residues is inserted in the loop re- The essential feature of the structure is that the six parallel fl gion between the ~c helix and the 82 strand (see next para- strands and the ac helix form the ATP binding domain (Fig. graph). The question of which one of the two conserved Glu's 3A). This folding motif is likely to be common to proteins listed (or Asp's) acts as an essential residue will be easily answered in Figs. 1 and 2. Although there is no obvious overall sequence by mutagenesis studies. similarity among the protein families, distribution of amino acid residues in the sequences shown in Figs. 1 and 2 is consis- 4. A common topology of ATP-triggered proteins tent with the assumption of this protein fold. The first thirteen amino acid sequences of the proteins listed in Fig. 1 are consis- Conservation of three functional elements (motifs A, B, and tent with afl strand-loop structure; several hydrophobic resi- a conserved Glu or Asp) among various proteins which catalyze dues are followed by a cluster of Pro and Gly (shaded letters). ATP-triggered reactions might be an indication for the presence of the common in their ATP-binding domains. *Numbering of the a helices or fl strands is corresponding to Esche- The following structure of the ATP-binding domain of the recA richia coli recA protein.

<__ Fig. 1. Comparison of amino acid sequences* of ATP binding domain in proteins which catalyze ATP-triggered reactions containing a set of typical nucleotide-binding motifs A and B. The sequence covering motif A and putative catalytic Glu (or Asp) and the sequence around motif B are shown. Lys in motif A, catalytic Glu (or Asp), and Asp in motif B are shown as K, E, and D (letters in black boxes), respectively. Pro and Gly, residues which tend to interrupt secondary structure, are shown by shaded letters. Hydrophilic residues are indicated by bold letters. Others are hydrophobic residues. In the case when a protein has duplicate sets of motifs A and B, a set at the amino-terminal side and at the carboxyl-terminal side are defined as (N) and (C), respectively. Secondary structures of the recA protein and MF:fl subunits determined from X-ray crystallography are shown (column, helix; arrow, ,8 strand). Rho protein and Lon protease have another nearby conserved Glu which can also be candidates of catalytic carboxylate. The description of function for each protein is not complete; for example, SUG1, MSS1, and TBP1 probably play a regulatory role in the 26 S proteasome complex. *References from which sequence data were obtained are not cited here to avoid too many references. 4 M. Yoshida, T. Amano/FEBS Letters 359 (1995) 1-5

B ~C d. Q194 (RecA) Y311 (F,-~ Swlt~ reelmndlng - 0 ATPhydr~-~*

/a, mo.. 0.Kr2~R,~A~ ~..÷ F~...--~"'°'.e "-~ (-rnTo} K162 (F,-~) ,--s, ..... ( p} 0 mow A ", ~ b. E96 (RecA) .~ E188 (F,-~) ..-~79 ~ .... ~. c.,-~ ~**~,

l I'I=C'~OH C. D144 (RecA)

_ ~L a.1"/3 (RecA) / ...... T163 (F,-~ motifA .::i;::::: i. :i: II:I:;::~': ATP ...... :::!:::-: Fig. 3. (A) A common topology of the ATP binding domain of proteins which catalyze ATP-triggered reactions. Circle a, b, e, and d indicate the positions of motif A, catalytic carboxylate, motif B, and a 'switch' region, respectively. (B) Schematic illustration of residues interacting with a bound ATP molecule in this fold. Numbering of the residues in the recA protein and F,-fl is according to E. coli recA protein and bovine mitochondrial F~-ATPase 8 subunit, respectively.

The ~c helix starts from the Thr in motif A and shows am- transmission of conformational changes; Gin194' when interact- phiphilic characteristics in recA protein and MFI-fl subunit ing with the ~'-phosphate of ATP, stabilizes the crystallograph- [4,6]. It might also be the case in most proteins listed; hydro- ically disordered loop which connects f15 strand and ~z~ helix, philic residues (bold letters) appear periodically, 2nd, 5th (or and hydrolysis of ATP destroys this interaction resulting in a 6th), and 9th (or 10th) positions counting from the Lys of motif change in conformation of the loop [4]. Also in the structure A. As are observed for the recA protein and MF]-fl subunit, of MF I, conformation of the carboxyl end region of the f15 the hydrophilic side of the helix is facing to the outside. There strand in each of three copies offl subunits in a MF~ molecule is no Pro in putative ~c helix region except secA protein**. is substantially different from each other depending on the After the ~c helix, a loop region, which tends to contain Gly presence of bound nucleotide and on the interaction with the and Pro, and then the relatively hydrophobic f12 strand follow. y subunit [6]. Modification of Tyr 31L in the f15 strand by The catalytic carboxylate, Glu (or Asp), is located immediately 7-chloro-4-nitrobenzofrazan in a single copy offl subunit in a after the f12 strand [4,6]. MF1 molecule is sufficient to inactivate catalytic turnover, Except for the region from thefl~ strand to theft2 strand, the probably by interfering with the inter-subunit conformational length (or even number) of 0~ helices and loops connecting each interaction [18,19]. Thus, it is reasonable to propose that the fl strand might be varied from one protein family to another, region at the carboxyl end of theft5 strand of this fold is a switch thus accounting for the variable distances in the primary struc- region involved in propagating conformational change trig- tures between conserved Glu (or Asp) and motif B (Fig. 1, gered by ATP hydrolysis. Successive conformational change distance of D-E). The motifB sequence in the recA protein and occurs to the loop connecting the f15 strand and the czG helix MFI-fl subunit belongs to the f14 strand which is located in the which are interacting with another molecule, subunit, or func- center of the ATP binding domain. The side chains of the tional domain of the protein. The f16 strand provides the anchor residues in the f14 strand are all hydrophobic and half of them for the ~ helix and two flanking loops, and stabilizes the fl] are facing the hydrophobic side of the 0~c helix. As demon- strand. strated by high resolution X-ray crystallography for the Asp 57 in the non-typical motif B in ras P21 protein [8], the Asp in 6. Discussion motif B (Fig. 1, motif B, D) is likely to be involved in binding of active-site Mg 2+ through a water molecule and to also be The occurrence of a similar folding topology among proteins hydrogen-bonded to the Thr of motif A (Fig. 3B). with non-homologous amino acid sequences has recently been recognized for hexokinase [20], Hsc70 ATP binding domain 5. 'Switch region' [14], and actin [15]. Here, we point out that the proteins, the activities of which are coupled to or triggered by ATP hydrol- In the recA protein structure, Gln ~94 is located at the carboxyl ysis, may have a common topology of the ATP-binding do- end of the f15 strand and is proposed to play a crucial role in main. The arrangement of catalytically important residues in the right positions in their tertiary structures may have been a consequence of convergent evolution. These residues include **SecA protein is different from others in that it has a Pro residue in the middle of putative ac helix and, therefore, the ac helix of secA Lys and Thr in motif A, the catalytic Glu (or Asp), Asp in motif protein is likely to be interrupted and bent at this point. B, and the residue responsible for conformational transmission M. Yoshida, Z Amano/FEBS Letters 359 (1995) 1 5 5

(Fig. 3B). The validity of this postulation can be easily tested. [10] Noel, J.P., Hamm, H.E. and Sigler, EB. (1993) Nature 366, 654- For example, the mutagenesis of the putative catalytic Glu (or 663. Asp) in these proteins would be easy to test. [11] Coleman, D.E., Berghuis, A.M., Lee, E., Linder, M.E., Gilman A.G. and Sprang S.R., (1994) Science 265, 1405 1412. [12] Fry, D.C., Kuby, S.A. and Mildvan, A.S. (1985) Biochemistry 24, References 468(L4694. [13] Mueller, C.W. and Schulz, G.E. (1992) J. Mol. Biol. 224, 159-177. [1] Walker, J.E., Saraste, M., Runswick, M.J. and Gay, N.J. (1982) [14] Flaherty, K.M., DeLuca-Flaherty, C. and McKay, D.B. (1990) EMBO J. 1,945-951. Nature 346, 623 628. [2] Amano, T., Tozawa, K., Yoshida, M. and Murakami, H. (1994) [15] Kabsch, W., Mannherz, H.G., Suck, D., Pai, E.F. and Holmes, FEBS Lett. 348, 93-98. K.C. (1990) Nature 347, 37~,4. [3] Ohtsubo, M., Yoshida, M., Ohta, S., Kagawa, Y., Yohda, M. and [16] Rayment, I., Rypniewski, W.R., Schmidt-B/~se, K., Smith, R., Date, T. (1987) Biochem. Biophys. Res. Commun. 146, 705-710. Tomchick, D.R., Benning, M.M., Winkelmann, D.A., Wesenberg, [4] Story, R.M. and Steiz, T.A. (1992) Nature 355, 374-376. G. and Holden, H.M. (1993) Science 261, 50 58. [5] Amano, T., Yoshida, M., Matsuo, Y. and Nishikawa, K. (1994) [17] Hyde, S.C. et al. (1990) Nature 364, 362 365. FEBS Lett. 351, 1-5. [18] Andrews, W.W., Hill, F.C. and Allison, W.S. (1984) J. Biol. Chem. [6] Abrahams, J.E, Leslie, A.G.W., Lutter, R. and Walker, J.E. 259, 8219-8225. (1994) Nature 370, 621-628. [19] Ferguson, S.J., Lloyd, W.J. and Radda, G.K. (1975) Eur. J. Bio- [7] Boyer, P.D. (1993) Biochim. Biophys. Acta 1140, 215 250. chem. 54, 127 133. [8] Pai, E.F., Krengel, U., Petsko, G.A., Goody, R.S., Kabsch, W. [20] Anderson, C.M., Stenkamp, R.E. and Steitz, T.A. (1978) J. Mol. and Wittinghofer, A. (1990) EMBO J. 9, 2351-2359. Biol. 123, 15-33. [9] Berchtold, H., Reshetnikova, L., Reiser, C.O., Schirmer, N.K., Sprinzl, M. and Hilgenfield, R. (1993) Nature 365, 12(~132.