Proc. Nati. Acad. Sci. USAVol. 80, pp. 6192-6196, October 1983Biochemistry

Molecular cloning and sequence determination of the nuclear genecoding for mitochondrial elongation factor Tu ofSaccharomyces cerevisiae

(protein synthesis/tuf gene probe/Southern hybridization/sequence conservation)

SHIGEKAZU NAGATA, YASUKO TSUNETSUGU-YOKOTA, AYAKO NAITO, AND YOSHITo KAZIROInstitute of Medical Science, University of Tokyo, Minatoku, Takanawa, Tokyo 108, Japan

Communicated by Severo Ochoa, July 8, 1983

ABSTRACT A 3. 1-kilobase Bgl H fragment of Saccharomycescerevisiae carrying the nuclear gene encoding the mitochondrialpolypeptide chain elongation factor (EF) Tu has been cloned onpBR327 to yield a chimeric plasmid pYYB. The identification ofthe gene designated as tufM was based on the cross-hybridizationwith the Escherichia coli tufB gene, under low stringency con-ditions. The complete nucleotide sequence of the yeast tufM genewas established together with its 5'- and 3'-flankdng regions. Thesequence contained 1,311 nucleotides coding for a protein of 437amino acids with a calculated Mr of 47,980. The nucleotide se-quence and the deduced amino acid sequence of tufM were 60%and 66% homologous, respectively, to the corresponding se-quences of E. coli tufA, when aligned to obtain the maximal ho-mology. Plasmid YRpYB was then constructed by cloning the 2.5-kilobase EcoRI fragment of pYYB carrying tufM into a yeast clon-ing vector YRp-7. A mRNA hybridizable with tufM was isolatedfrom the total mRNA of S. cerevisiae D13-IA transformed withYRpYB and translated in the reticulocyte lysate. The mRNA coulddirect the synthesis of a protein with Mr 48,000, which was im-munoprecipitated with an anti-E. coli EF-Tu antibody but not withan antibody against yeast cytoplasmic EF-la. The results indicatethat the tufM gene is a nuclear gene coding for the yeast mito-chondrial EF-Tu.

The polypeptide chain elongation factor Tu (EF-Tu) promotesa GTP-dependent binding of an aminoacyl-tRNA to the A siteof ribosomes (1). EF-Tu from Escherichia coli consists of a sin-gle polypeptide chain with Mr 43,000, and the primary struc-ture comprised of 393 amino acid residues has been deter-mined (2). The protein is encoded by two nearly identical geneson the E. coli chromosome (3), tufA at 73 min and tufB at 89min (4). Both tufA (5) and tufB (6) have been cloned and theirnucleotide sequences determined. The sequences of tufA (7)and tufB (8) are nearly homologous and differ only in 13 po-sitions but the gene products, EF-TuA and EF-TuB, are iden-tical except for the COOH-terminal amino acid (2).

In eukaryotes, the counterpart of prokaryotic EF-Tu, des-ignated as EF-la, has been purified from various sources, in-cluding pig liver (9), rabbit reticulocytes (10), Artemia salina(11), wheat germ (12), and yeast (13), and was shown to consistof a single polypeptide chain, Mr 47,000-53,000. The partialamino acid sequence of EF-la from rabbit reticulocytes (14)and A. salina (15) was determined, and sequence conservationbetween A. salina EF-la and E. coli EF-Tu has been reported(15).

In addition to EF-la, which functions in the cytoplasmicfraction in conjunction with 80S ribosomes, eukaryotic cellspossess mitochondrial EF-Tu (designated as mEF-Tu) that

functions in the mitochondrial translational apparatus (16, 17).Translational factors as well as ribosomal proteins in the mi-tochondria are encoded by nuclear genes, synthesized in cy-toplasmic fractions, and transported into the mitochondria (18).The translational machineries of mitochondria have been thoughtto be closer to the prokaryotic ones than to the machinery pres-ent in eukaryotic cytoplasm (17, 19). However, the organizationof the nuclear genes for the mitochondrial translational appa-ratus, their expression, and the incorporation of the productsinto the mitochondria are not well understood.

In this report, we describe the molecular cloning of the yeastmitochondrial EF-Tu gene (tufM) utiljzing hybridization withthe E. coli tufB gene under low stringency conditions. The nu-cleotide sequence as well as the deduced amino acid sequenceof the tufM gene indicate that the structures of E. coli EF-Tuand yeast mitochondrial EF-Tu are remarkably conserved dur-ing evolution. The mRNA for tufM can be translated in the cell-free system from rabbit reticulocytes and the product has beenidentified as mEF-Tu by immunoprecipitation with an anti-body against E. coli EF-Tu.

MATERIALS AND METHODSEnzymes. Endonuclease EcoRI and T4 polynucleotide ligase

were gifts from Takashi Yokota. Other restriction enzymes andT4 polynucleotide kinase were purchased from Takara Shuzo(Kyoto, Japan) or Bethesda Research Laboratories and were usedessentially as recommended by the supplier (except for the useof 200 ,g of gelatin per ml instead of bovine serum albumin).E. coli DNA polymerase I (large fragment) was purchased fromBoehringer Mannheim.

Southern Hybridization. DNA was prepared from varioussources including Saccharomyces cerevisiae 106A (mating typea, arginine requiring) as described by Cryer et al. (20) and com-pletely digested with several restriction enzymes. Fragmentswere separated by electrophoresis on 1% agarose gel (21) andtransferred to a nitrocellulose filter (Sartorius, 0.45-gm pore)as described by Southern (22). The 1.5-kilobase (kb) Hpa I frag-ment of pTUB1 (6) was subcloned at the Pvu II site of pBR322to yield a hybrid plasmid pYT-1 which was used as a probe. Thisfragment covers the entire coding sequence of E. coli EF-TuBexcept for the NH2-terminal 12 amino acids (see ref. 8). pYT-1 was labeled by nick-translation (23) with [a-32P]dCTP (3,000Ci/mmol; 1 Ci = 3.7 x 1010 Bq; Amersham Japan) to yield 1x 10' cpm/,ug and was hybridized with yeast DNA fragmentsessentially as described by Wahl et al. (24), except that the hy-bridization temperature was lowered to 28°C, and the filter waswashed at 44°C with 15 mM NaCl/1.5 mM sodium citrate, pH7.0/0.1% NaDodSO4.

Abbreviations: EF, polypeptide chain elongation factor; bp, base pairs;kb, kilobase(s).

6192

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertise-ment" in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Proc. Natl. Acad. Sci. USA 80 (1983) 6193

1 2 3 4 5 6 7 8 9 10

* 0 0 0 *

.i-

0 m

* 0.** O

-5.45

-3.16

-2.25-2.0- 1.5

-0.75

FIG. 1. Southern hybridization analysis ofE. coli and yeast DNA.E. coli DNA (100 ng) (lanes 1-4) or yeast DNA (10 j.g) (lanes 5-8) wasdigested with several restriction enzymes and electrophoresed on a 1%agarose gel. DNA was transferred to a nitrocellulose membrane filterand hybridized with nick-translated pYT-1 (1 x 108 cpm/jug). Restric-tion enzymes used were EcoRI (lanes 1 and 5), Bgl II (lanes 2 and 6),EcoRl and Bgl I (lanes 3 and 7), and Pst I (lanes 4 and 8). A mixtureof the Sma I single-digested and the Sma I and EcoRI double-digestedpYT-l (lane 9) and the EcoRI, Pvu I, and Pst I triple-digested pBR322(lane 10) were run as size markers. Sizes ofmarkerDNA fragments aregiven in kb.

Restriction Mapping and DNA Sequence Determination.Restriction sites were determined by a single or double diges-tion of the cloned DNA. In some instances the arrangement ofthe restriction fragments was determined by Smith-Birnstielmapping (25) of the labeled fragments. DNA sequence analysiswas carried out by the chemical degradation method (26) withrestriction fragments labeled either at the 5' end with [y-32p]-ATP (prepared as described in ref. 27) and T4 polynucleotidekinase or at the 3' end with [a-32P]dCTP and E. coli DNA poly-merase I (large fragment).

Hybridization-Translation. S. cerevisiae D13-IA (a his3-532trpl gal2) was transformed with YRpYB. This was constructedby cloning of the 2.5-kb EcoRI fragment containing the yeastnuclear gene for mitochondrial EF-Tu [cloned in pYYB (see Re-sults)] into the yeast vector YRp7 (28). A transformant was grownon a minimal essential medium as described, (29), except that

H-4

pBR -

H H0; HO :3u>

w IH1H4

Hto

0m

2% glucose was replaced by 2% lactate. Total RNA was pre-pared according to Struhl and Davis (30) and poly(A)-containingmRNA was selected by ofigo(dT)-cellulose column chromatog-raphy (31).The 2.5-kb EcoRI fragment of pYYB or EcoRI-digested

pBR327 DNA was bound to a nitrocellulose filter (Schleicher& Schifil, 7 x 7 mm, 0.45-,um pore) and the filters were hy-bridized with 20 ,.g of mRNA prepared as above (32). The fil-ter-hybridized mRNA was eluted and translated in the retic-ulocyte lysate system (33). Immunoprecipitation of the trans-lational products with anti-sera and protein A adsorbent wascarried out as described (34). The anti E. coli EF-Tu antibody(35) was a gift from A. Miyajima. The antibody against yeastcytoplasmic EF-la was a generous gift from M. Miyazaki (Na-goya University).

RESULTSPresence of a Sequence Homologous to the Gene for E. coli

EF-Tu in Eukaryotes. First, we searched for a sequence ho-mologous to E. coli tufB in DNA from various eukaryotic sources,including yeast, mouse liver, and human leukocytes utilizinglow stringency conditions (36). Because the original plasmidpTUB1 containing tufB possessed other E. coli sequences, suchas a part of rrnB, genes for four tRNAs (thrT, glyT, tyrU, andthrU), a gene for an unidentified protein "U, " and a part of rplK(6), we constructed plasmid pYT-l by subcloning the 1.5-kb HpaI fragment of pTUB1 into pBR322 and used this as a probe. Thefragment covered almost the entire coding sequence of tufB aswell as about 300 nucleotides of the 3'-flanking region.

As shown in Fig. 1, E. coli DNA (lanes 1-4) gave two bandshybridizing with the tufB probe due to the presence of two genesfor EF-Tu (tufA and tufB) in E. coli (3). On the other hand, asingle DNA band was detected with yeast DNA (lanes 5-8),indicating that at least one sequence homologous to the E. coliEF-Tu gene is present in yeast. Mouse liver and human leu-kocyte DNA did not give any detectable band hybridizing withpYT-l under the same conditions (data not shown), suggestingthat the sequence of the gene in mouse or human cells is moredivergent.

Isolation of the Yeast Gene Homologous to the E. coli EF-Tu Gene. Two hundred micrograms of DNA from S. cerevisiae106A was digested with Bgl II, and fragments (3.1 kb in length)hybridizing with pYT-l were isolated by agarose gel electro-phoresis. The fragments were ligated to pBR327 at the BamHI

HH

:j

HM0u

H

a

1 pBR

HH If HH ,* H H H1H * ' H HO.'

_ --I 1>

I I

HH

H HMo '-n0.~4 'SoU 5014

I I

H H

IS

I

H H H "H H H H H H * H

-4 P; 44 H 0X qt ' 44

H U H. > U H -H4: = P = =

I I I I I I I **

4 *

it.-oD4 0'p--Ik~~~~

200 bp

FIG. 2. Restriction map of the yeast DNA fragment cloned on pYYB and the strategy for nucleotide sequence determination. Restriction mapof the yeast Bgl II fragment cloned at the BamHI site of pBR327 is shown in the upper part of the figure with major restriction sites. The lowerpart shows the expanded structure of the yeast tufM gene. The sequence corresponding to the coding region is indicated by the shadowed box; thedirection of transcription is left to right. The circles-represent labeled termini at 5' (e) or 3' (o) ends, and the solid arrows show the sequence readoff of the labeled fragments. bp, base pairs.

A- b p 0 p m a I I p I I I I M m .aV E 6 I

Biochemistry: Nagata et al.

-t

4.0

.i .......

6194 Biochemistry: Nagata et al. Proc. Nati. Acad. Sci. USA 80 (1983)

-220 -200(tufM) TrCCATA¶1T1TAACCAI7TrICT1TrAGC~XA¶1CCA7

-180 -160 -140 -120 -100 -80 -60

-40 -20 -1 1 30 60m~rrCC~'I~rATCTP~c'CG~cAAACTA~c~A AATAI IA OCT 7TA 7TA OCA AGA 7TA Cit ACA AGA AGA OCT TT AAA OCT ItT 000 AAA CIT CMC AO Cit It itA

Met Ser Ala Leu Leu Pro Arg Leu Leu Thr Arg Thr Ala Phe Lys Ala Ser Gly Lys Leu Leu Arg L-eu Ser Ser1 10 20

90 I*I I120...1....0...WGTA AlT ItT AGG AME 7Tr' I CAA ACT AC ACT OA CA OCT OCT 17rGA CGT ICC AMA CCC CAT GTA MAT ATA OG1r AltMCGO CAT CIT CAT CAT GCC AAK

va1 le ser Arg Thr Phe Ser Gin Thr T1hr Thr Ser Tyr Ala Ala Ala IPhei Asp ~Ar Ser Lys Pro His ValAs. Ile C~ly Thrt fiG1~_ly-fH!isval As His- MC tysi30 40 I50' 61 1 10 20 6

Ser Lys Giu Lys PheCGu AqThr, Pois Val Msn.Va1 Oly Thr Ile GIy His Val Asp His Gly Lys(tufA) GM ItT MAA AM MAA f GAA IM c A c ~ ,Gr Gr ~ MGCCCG A A G A___ MA CCC CN2CIT~~~~~~... .. CIT ... ... A.t ... CIgorC~

210 240 270....AlA AH TiA MM OCA OX ATr AOGO ;AM AG ITA OX GcA MAAACT GGT 0oCC AAG TrC TIC TcAT O5cT ccC MM Gi<AT ''AG 6&r 6M G'Ak._AGIU A &C'.C& Wr& AiTtw Thr Leu Th Ala Ala Ile TrLys Thr Leu Ala Lys Gly Clyl Ala Msn Phe Leu Asp Tyr Ala Ala. 'Ile Aspf Lys'Ala Pro Glu GlIj ArgPAlaArg Gly Ile

30 j70 40 80 50 90 60TtwrhrrLuTt laAaujle Th Thr Val Ile,1A Lys Thr Tyr nil Gly Ala Ala Arg Ala Phe Asp Gin IIl s Asn:Ala Pro Glu Glu Lys Ala Arg Gly le.grWTIG IOCOCT(?C7C.AIO CIAGMTG AAAMANC TAlC CC OGT OCT OCT CCT GCA TIC GAC CAAMI CAT AAMICOCG GA GA AAA OCT CGT GCr'A7t

300 330 360 390ATT=I 'TAIO6A CA.IYCEGAT.AC GMA A.CO OX A.AG MA& CAT TAT 'I&TCPC CCC GACMI 'I&TC-AG101C .AC ILCT C-AT T-AG Alt A.-AG.AT AIGC kAT I E OCTGOCT-GOCT

Thr I Ser Al i a l y Glu Ala Lys Arg His A~jSer His Val Asp Cys Pro Cly His Ala Asp Tyle Lys Msn Met Ile Thr Cly Ala Ala'11001 7001 110 I 80 120 90 130.

7h IAsn Ser HisCalGu ly Asp Pro ThrlArq His.Tyr~ AlaiHis Val Asp Cys Pro Ciy His Ala As yrak ~Lys Msn Met Ile Thr Gly Ala Ala

CAAMGGTCTGCTA. ATT' CIT CTA OCT ME-ACI OIAT OGIA CIANAATCG CCC CMA AT AGA GM CAT 7TA CIT TI OX AGA CMA CIT GOT CCC CMA CAT AUT! GICCIGin Met Asp Oly Ala'I Ile Vl Val Ala Ala Ttw Asp C Gin Met Pro Gin 7hr -Arg Ciu .Hi Lku LuL Ala GIn Val Gly V Gin His Val a

100 I140 10 150 120 1 60131GlnHot GlyAla ILAU Va.Al *Ala.Th As Pro Met Pro Gin Thr AgGlu:Hs Ile, Le Gly AqGin Val GlvVal Pro Tr Ile al~

Ho. I I___I ..

540 *570 600TIT! (*1T M AAG CIT OAT:Ac AT!' GA &z cA GMA AlE T!A'A'G CCGC GMAu I CMGA AlE MA GMAA Cl' T!'-A GMrTATiifaG TI'!' GAcGOCT GAT AT WHMT

Val Ls Val Tw Ilie ProHG Met Lou Glu Leu Val Glu Met 01 Met AgGuLeu Leu Asn Giu~l~Gly Asp G Ann AlaPr17 ISCOAlEC90C 200~

Leu CS metValGu1IluIu euGlu Leu Val GluMet GlValAmGlu Lem-Leu Ser Gin Asp Pro 01 Asp Th ro

a20I Ito M I GM C CM A4XWACO ~

630 660 690 720

l ie Met ySer Aa CysAl Leu 0lu 011 Arq Gin Pro 01 Ile. Gly 1Gin Ala IiMet Lys J Leu Asp Ala Vl Glu Tyr Ilie Thr

I170 2 0 I d 19017Va1 Arg 1 Ser Ala Lys Ala Leu G Asp 220 G Trp iuAla Lys ILeu 190 l lyPeLu e lOJT! OCGT9 f~9 AAA q9qC.OFVCf9Al GAA OG AA

AAN

C G~AAE OCT9CCG TIC CIV 9A ITt T.AIT& GA2O WAA750 780 810

GM~MA& (AT TiE AMI AMO &T W~ TM¶ AMi &%C GI'! .aU dAT At W'I I*T W4 i&C% OCT" AM o6f M!' (lh alt AMT Gd? 6f. MM G'M A& 06f AAT ITA -AMG AMGiF Asp Leu Asn Lys Pro Phela Met Va1 Ilie Phe Ser Ilie Ser Gly Arg Gly Thr Va1 Va1 Ttw Gly Arg Va1 Glu Arg 01 Asn Leu LLYS

I ~~~~~~~~~~21025 1 1 1 220 260 230 270 IGu Ala Ile Asp Ly mPeLe Leu Ilie Va h Ser Ilie Ser Gly Arg Gly Thr V.1 V.1 Thr Gly Arg Va1 Glu Arq I le Ile I!dVa1

9(r AT!' GK A T CFW AV afA y.MVT,; M.0T% .c91 7 -c-M W 0c % mA Aa

840 870' 900 930 .

01Gu Leulu IeV.10G His Amn Ser Pro Leu J 7thr V.1 G Ie u hArgFGlu~7 S£Aa21tM Asp240 28 I I2 290 II I 300 ' l erAaMt70

GyGluGiVl IeV1 l Ilie Lys Glu Gin. Ser. Cys ~~~I Va1 lu met Le Glu Gly Arg ~~GluI

720 ~~~~~~~~~~~~~~~~~~750810

Ala Gly Val Leu LauArg Gy I Arg Asp Gin Leu Lys [j'I Met Va Lou AaLysProGy V.~j l shLleLuAla Ser Lou Tyr320 12900 13401

V.1 GyyVl e Leu ArgGlyIl Lys Glu Glu Tie Giu G Gin Va euAa y Pro G I~jlie I58jPro 7t Pt. Glu Ser Glu V.I1T

19~~~~~~~~~0.1099.if~~~~~~~~~~~~~~~~~~~~10.. ....1140 .

AI'! ITA ILE AMA GAG GMl CGT GOCT MAG CAT It'! 000 l'IT OCT GMA AM6 TAG MA LECA CM. AlE TIT ATA MAh ACA OCT GAT CIT AlA 01'! OlE AlE MA TI'! AMlie Lou Ser Ly l Glu GyGyArg Ser Gly Gly Glu Amn Arg Pro 0 Met Pt Ilie AriIThr Alap V.1 V.1 Val Met Arq Pt.hec LYS310 360 331 3 0 I 380Il e e y Asp 01u 01 H -thr Pro Pt. Lys Gly Ar r Pt. Tyr Pt. "Wq¶'~ Thrw Val2 Gly Thr Ile'Glu -Imu1 Glu

1170 1200 1230 1?60Glu Val Gl Asp His Ser f~Gin Va e mGl s s Val Glu ---Glu Cys Asp ~iI is~ Thr Pro Lou Glu Val Gin [rgi Asn Ile Arg

390 I 13 5014000 I 360 4140 3701 4II20Gl a l a eroGlyAp Il y Val Val Tw Loulie His Irjle Ala Met Asp.-Asp _~~Lu Alal. r

.9 ' 1680 I Illo01290 1320 1350 1380

dIM(1'! &al AlA A&T WliIl!MV t CIA Ait AIA CC!GTI. AT IM i'AG+AGITTEICA'EA!AAII!AAAI!TtT'tC~'CTTTT.RIIIEGlu Gly ly ArgTtwValGlyThG1.Leu IlieTtw Arg Ilie Ilie Glu

380 43 390Glu GlyGArg Thr Val Gly Ala 1G Val Val Ala Lys Val Lou Gly9m.09CW-jim-uffCam GIT O~CT MACAG7 CIE GOC TAA

1410 1440 1470

AMATI'IT l~i-U!'AAA-hAM'IECCATAAATItMCI'I!'A-lI'I'~!'M~UIECTJCAGMATI

FIG. 3. Comparison of the nucleotide sequences of yeast tufM and E. coli tufA gene and the deduced amino acid sequences. The sequence of theE. coli tufA gene is from Yokota et al. (7). The sequences of tufM and tufA are aligned to give maximal homology by introducing several gaps intufA. Identical amino acids are framed and identical nucleotides are marked by a dot.

Proc. Natl. Acad. Sci. USA 80 (1983) 6195

site, and the hybrid plasmids were used to transform E. coliSK1592 (F gal thi Ti endA sbcl5 hsdR4 hsdM) by the methodof Hoeijmakers et al. (37), to yield 768 ampicillin-resistant andtetracycline-sensitive clones.

Clones containing the yeast tufM gene were identified bySouthern hybridization of plasmid DNA with a tufB probe un-der low stringency conditions. Plasmid DNA was prepared (38)from pools of-24 bacterial clones, and 0.5-1.0 ,ug of DNA fromeach pool was electropharesed on a 1% agarose gel,. to separatehost DNA, and transferred to a nitrocellulose membrane filter.As a probe, we used the 1.5-kb Hpa I fragment containing E.coli tufB prepared from plasmid pTUB311 (unpublished data).Because pTUB311 is an R-derivative plasmid containing tufB,any vector DNA that might contaminate the Hpa I fragmentwould not hybridize with the sequence of pBR327 (a colEl de-rivative) (39) in the hybrid plasmids.One of 32 groups of 24 clones gave a positive hybridization

with the 32P-labeled Hpa I fragment, whereas DNA from othergroups or pBR327 gave no hybadization, even at the low strin-gency conditions used forthe present experiments. Plasmid DNAfrom each individual clone of this group was then prepared andanalyzed as above, and a clone, that gave a positive result wasdesignated as pYYB.

Nucleotide Sequence Analysis. Fig. 2 shows- the physical re-striction map of pYYB constructed as described in Materialsand Methods and the strategy for the DNA sequence analysis.Southern hybridization analysis of pYYB with E. coli tufB showedthat most of the yeast sequence homologous to tufB resideswithin the 1.0-kb Pvu II fragment (data not shown). Thereforewe determined the sequence of this region as well as the re-gions flanking it, as indicated in Fig. 2. The nucleotide se-quence (1,712 bp) thus obtained is shown in Fig. 3 togetherwith the sequence of the E. coli tufA gene (7). The translationinitiation site for yeast tufM was assigned to the methioninecodon AUG at nucleotide-positions 1-3, because this was thefirst AUG triplet downstream of the nonsense codon UAA (po-sitions -12 to -10) found in the open reading frame of the nu-cleotide sequence aligned for maximal homology with E. colitufA (e.g., positions 124-149). The termination codon for trans-lation was assigned to UAG at positions 1,318-1,320 in the samereading frame. Because we could not find any intron in the cod-ing sequence of tufM (unpublished data), the gene codes for aprotein of 437 amino acids including the NH2-terminal methi-onine with a calculated Mr of 47,800, in a single- open readingframe of 1,311 nucleotides.A comparison of the sequence of yeast tufM with that of E.

coli tufA reveals that the homology is 60% and 66% for the nu-cleotide and amino acid sequences, respectively. To obtainmaximum homology, the initiator methionine for tufA was alignedto Ser-37 of tufM, and a limited number of gaps were intro-duced into the sequence of tufA. As seen in Fig. 3, several dis-tinct regions are highly conserved between tufM and tufA. Amost remarkable homology was found in amino acid residues74-118 of tufA and 111-155 of tufM, where 41 of 44 amino acidresidues were identical (91% homology). Because this region issupposed to be an active site for interaction with aminoacyl-tRNAs(reviewed in ref. 1; see also ref. 40), the sequence conservationof this region might be due to a functional requirement. Otherhomologies were. found in amino acid sequences 5-32, 208-234,and 371-386 of tufA with the corresponding sequences of tufM,where 25 of 28, 24 of 27, and 13 of 16 amino acid residues areidentical, respectively. Whether or not these conserved regionsconstitute functional domains of the protein remains to be de-termined.The Yeast tufM Gene Codes for the Mitochondrial Factor.

To determine whether the yeast tufM gene homologous to bac-terial tufA and tufB codes for the cytoplasmic or the mito-

chondrial factor, the mRNA specified by tufM was isolated byhybridization to the cloned tufM gene and translated in the re-ticulocyte cell-free system (33). As shown in Fig. 4, the tufMmRNA markedly stimulated the incorporation of [3S]methio-nine into a protein of Mr 48,000 (lane 3), whereas little increaseof incorporation was observed with the mRNA hybridized topBR322 (lane 2) as compared to the incubation without addedmRNA (lane 1). The product synthesized under the directionof the. tufM mRNA was. immunoprecipitable with antibodyagainst E. coi EF-Tu (lane 5) but not with.antibody against yeastcytoplasmic EF-la (lane 4).. Furthermore, as will be describedelsewhere, the M, 48,000 protein was transported into yeastmitochondria on incubation in a cell-free system (unpublisheddata). From these results, it was concluded that the yeast tufMgene codes for the mitochondrial EF-Tu.

DISCUSSIONIn the present study, a hybrid plasmid containing the yeast mi-tochondrial EF-Tu gene was isolated by-,subeloning and iden-tified by Southern hybridization by using the E. coli tufB se-quence as a probe. This procedure was found to be relativelysimple and very effective. It may be applicable for isolation ofother eukaryotic genes having a homology with E. coli genes-for.example, the eukaryotic genes encoding for. the mitochon-drial translation factors and ribosomes.

Southern hybridization analysis of the yeast restriction frag-ments with labeled DNA containing the E. coli tufB gene re-vealed only a single band that hybridized with the probe, sug-gesting that the mitochondrial EF-Tu is coded by a single nuclear

M 1 2 3 4 5

200-

92-

69-

46 -

30 -

14.3-

FIG. 4. In vitro translation of the hybridization-selected mRNA andimmunoprecipitation ofthe product. The 2.5-kbEcoRl fragment ofpYYBcontaining tufM gene or EcoRI-digested pBR322 was bound to nitro-cellulose membrane filters and hybridized with mRNA obtained fromyeast transformed with YRpYB. The hybridized mRNA was eluted andtranslated in 20 #1 of reticulocyte lysate in the presence of 12 ACi of[32S]methionine (1,490 Ci/mmol, Amersham Japan). One third of theproduct was electrophoresed on 15% polyacrylamide gels in the pres-ence of NaDodSO4 (41) and visualized by fluorography (42). Transla-tion products without added mRNA (lane 1) or with mRNA hybridizedto pBR322 (lane 2) ortufM (lane 3). One third each of the product usedfor the experiment of lane 3 was immunoprecipitated- by. antibodiesagainst either yeast cytoplasmic EF-lca (lane 4) or E. coli EF-Tu (lane5), and the precipitates were analyzed as above. Lane M, '4C-labeledmolecular weight standards (Amersham Japan) shown as Mr x10-3(from.top to bottom): myosin, phosphorylusmb, bovine serum albumin,ovalbumin, carbonic anhydrase, and lysozyme.

Biochemistry: Nagata et A

f t

..f

iI I

6196 Biochemistry: Nagata et al.

peroxidase ;TThr-Thr1l Val-ArgLu-Leu-Prc-Ser -----GlyAg-Thr-AHlIsisLysArg-Ser-Leu-TyrLenEF-Tu Met ----Ser-Al------Leu-Leu-Pr -Arg-LeLeu-Thr -Arg-Thr-Al Phe-LysAlaSe Gly-Lys Le

FIG. 5. Comparison of the NH2-terminal sequence of yeast mitochondrial EF-Tu and the signal sequence of yeast cytochrome c peroxidase. Theamino acid sequence of yeast mitochondrial EF-Tu is from Fig. 3 and the signal sequence of yeast cytochrome c peroxidase is from Kaput et al. (45).The boxed amino acids indicate identical amino acids; dashes show gaps introduced to obtain the maximal homology.

gene and that the gene for the cytoplasmic factor may be moredivergent. This agrees with the previous finding that yeast mi-tochondrial EF-Tu is functionally interchangeable with E. coliEF-Tu but not with yeast cytoplasmic EF-la (16). More re-cently, Piechulla and Kuntzel (17) have shown that an antibodyagainst yeast mitochondrial EF-Tu crossreacts with E. coli EF-Tu. As we have shown, the deduced amino acid sequence ofyeast tufM is remarkably homologous to that of E. coli EF-Tu,especially in some regions. Amons et al. (15) have recently de-termined the partial amino acid sequence of EF-la from A. sa-lina and pointed out the homology with E. coli EF-Tu. How-ever, the homology of E, coli EF-Tu with A. salina EF-l1a is farless pronounced than that with yeast mitochondrial EF-Tu(compare figure 4 of ref. 15 with Fig. 3 of the present paper).

In general, nuclear-coded mitochondrial proteins are syn-thesized as precursors in the cytoplasm and transported intomitochondria (43). Although it is not known whether yeast mi-tochondrial EF-Tu is synthesized as a precursor that is pro-cessed to the mature protein during transport into the mito-chondria, it is noteworthy that the tufM gene codes for a protein37 amino acids longer than E. coli EF-Tu at the NH2-terminalend. The 37-amino acid peptide is strongly basic, having 4 ar-ginine and 2 lysine residues but no acidic amino acids. It.is alsorich in threonine (six residues) and serine (seven residues).

Recently, the. structures of, the- proteolipid subunit of mi-tochondrial ATP synthase (44) and cytochrome c peroxidase (45)have been published. The signal sequences of these proteinsare 68 and 66 amino acids long, respectively, and possess prop-erties similar to the above NH2-terminal sequences. Becausethe precursor of cytochrome c peroxidase is transported intothe inner membrane space of the mitochondria, with the first18 amino acid residues of the signal sequence facing the mi-tochondrial matrix (45), we have compared the structure of thesignal sequence of cytochrome c peroxidase with that of themitochondrial EF-Tu. As shown in Fig. 5, the first 20 aminoacid residues of mitochondrial EF-Tu have a distinct homologywith that of cytochrome c peroxidase, which may suggest thata homologous sequence is important for proteins made in thecytoplasm that are to be imported into the mitochondria.The sequences that are conserved in many eukaryotic pro-

moters are found at -225 nucleotides upstream from the firstATG codon of the tufM gene. A consensus sequence for poly(A)addition T-A-A-A-T-A-A-A-t in yeast is also present at 96-103nucleotides downstream from the TAG termination codon. Moredetailed characteristics of the sequence as well as the analysisof the transcriptional and translational products will be dis-cussed elsewhere.

We thank Dr. M. Miyazaki (Nagoya University) for his generous giftof, the antibody against yeast cytoplasmic EF-la. Thanks are also dueto Dr. R. W. Davis for providing us with vector plasmid YRp7 and strainD13-lA.

1. Kaziro, Y. (1978) Biochim. Biophys. Acta 505, 95-127.2. Arai, K., Clark, B. F. C., Duffy, L., Jones, M. D., Kaziro, Y.,

Laursen, R. A., L'Italien, J., Miller, D. L., Nagarkatti, S., Nak-amura, S., Nielsen, K. M., Petersen, T. E., Takahasi, K. & Wade,M. (1980) Proc. Natl. Acad. Sci. USA 77, 1326-1330.

3. Jaskunas, S. R., Lindahl, L., Nomura, M. & Burgess, R. R. (1975)Nature (London) 257, 458-462.

4. Bachmann, B. J. & Low, K. B. (1980) Microbiol. Rev. 44, 1-56.5. Shibuya, M., Nashimoto; H. & Kaziro, Y. (1979) Mol. Gen. Genet.

170, 231-234.6. Miyajima, A., Shibuya, M. & Kaziro, Y. (1979) FEBS Lett. 102,

207-210.7. Yokota, T., Sugisaki, H., Takanami, M. & Kaziro, Y. (1980) Gene

12, 25-31.8. An, G. & Friesen, J. D. (1980) Gene 12, 33-39.9. Nagata, S., Iwasaki, K. & Kaziro, Y. (1977)J. Biochem. (Tokyo) 82,

1633-1646.10. Slobin, L. I. (1980) Eur. J. Biochem. 110, 555-563.11. Slobin, L. I. & M6ller, W. (1976) Eur. J. Biochem. 69, 351-366.12. Bollini, R., Soffientini, A., Bertani, A. & Lanzani, G. A. (1974)

Biochemistry 13, 5421-5425.13. Dasmahapatra, B., Skogerson, L. & Chakraburtty, K. (1981)J. Biol.

Chem. 256, 10005-10011.14. Slobin, L. I., Clark, R. V. & Olson, M. 0. J. (1981) Biochemistry

20, 5761-5767.15. Amons, R., Pluijms, W., Roobol, K. & Moller, W. (1983) FEBS

Lett. 153, 37-42.16. Richter, D. & Lipmann, F. (1970) Biochemistry 9, 5065-5070.17. Piechulla, B. & Kuntzel, H. (1983) Eur.J. Biochem. 132, 235-240.18. Richter, D. (1971) Biochemistry 10, 4422-4425.19. Borst, P. (1972) Annu. Rev. Biochem. 41, 333-376.20. Cryer, D. R., Eccleshall, R. & Marmur, J. (1975) Methods Cell

Biol. 12, 39-44.21. Nagata, S., Mantei, N. & Weissmann, C. (1980) Nature (London)

287, 1-7.22. Southern, E. M. (1975)J. Mol. Biol. 98, 503-517.23. Jeffreys, A. J. & Flavell, R. A. (1977) Cell 12, 429-439.24. Wahl, G. M., Stern, M. & Stark, G. R. (1979) Proc. Natl. Acad.

Sci. USA 76, 3683-3687.25. Smith, H. D. & Birnstiel, M. L. (1976) Nucleic Acids Res. 3, 2387-

2398.26. Maxam, A. M. & Gilbert, W. (1980) Methods Enzymol. 65, 499-

559.27. Walseth, T. F. & Johnson, R. A. (1979) Biochim. Biophys. Acta

562, 11-31.28. Struhl, K., Stinchcomb, D. T., Scherer, S. & Davis, R. W. (1979)

Proc. Natl. Acad. Sci. USA 76, 1035-1039.29. Hitzeman, R. A., Hagie, F. E., Hayflick, J. S., Chen, C. Y., See-

burg, P. H. & Derynck, R. (1982) Nucleic Acids Res. 10, 7791-7808.30. Struhl, K. & Davis, R. W. (1981)J. Mol. Biol. 152, 535-552.31. Aviv, H. & Leder, P. (1972) Proc. Natl. Acad. Sci. USA 69, 1408-

1412.32. Harpold, M. M., Dobner, P. R., Evans, R. M. & Bancroft, F. C.

(1978) Nucleic Acids Res. 5, 2039-2053.33. Pelham, H. R. B.- & Jackson, R. J. (1976) Eur. J. Biochem. 67, 247-

256.34. Kessler, S. W. (1975)J.Immunol. 115, 1617-1624.35. Miyajima, A. & Kaziro, Y. (1978)J. Biochem. (Tokyo) 83, 453-462.36. Shaw, G. D., Boll, W., Taira, H., Mantei, N., Lengyel, P. &

Weissmann, C. (1983) Nucleic Acids Res. 11, 555-573.37. Hoeijmakers, J. H. J., Borst, P., Van Den Burg, J., Weissmann,

C. & Cross, G. A. M. (1980) Gene 8, 391-417.38. Wilkie, N. M., Clements, J. B., Boll, W., Mantei, N., Lonsdale,

D. & Weissmann, C. (1979) Nucleic Acids Res. 7, 859-877.39. Soberon, X., Covarreubias, L. & Bolivar, F. (1980) Gene 9, 287-

305.40. Nakamura, S. & Kaziro, Y. (1981) J. Biochem. (Tokyo) 90, 1117-

1124.41. Laemmli, U. K. (1970) Nature (London) 227, 680-685.42. Bonner, W. M. & Laskey, R. A. (1974) Eur. 1. Biochem. 46, 83-

88.43. Schatz, G. & Butow, R. A. (1983) Cell 32, 316-318.44. Viebrock, A., Perz, A. & Sebald, W. (1982) EMBOJ. 1, 565-571.45. Kaput, J., Goltz, S. & Blobel, G. (1982)1. Biol. Chem. 257, 15054-

15058.

Proc. Natl. Acad. Sci.'USA 80 (1983)