Proc. NatL Acad. Sci. USAVol. 79, pp. 6903-6907, November 1982Biochemistry

Location and nucleotide sequence of the gene for the proton-translocating subunit of wheat chloroplast ATP synthase

(chloroplast DNA/dicyclohexylcarbodiimide-binding proteolipid/in vitro transcription-translation)

C. J. HOWE*, A. D. AUFFRETt, A. DOHERTY*t, C. M. BOWMAN§, T. A. DYER§, AND J. C. GRAY**Botany School, University of Cambridge, Downing Street, Cambridge CB2 3EA, United Kingdom; tProtein Sequencing Unit, Department of Biochemistry,University of Leeds, Leeds LS2 9JI, United Kingdom; and Plant Breeding Institute, Mans Lane, Trumpington, Cambridge CB2 2LQ, United Kingdom

Communicated by F. Sanger, August 23, 1982

ABSTRACT The proton-translocating subunit of wheat chlo-roplast ATP synthase is encoded by a chloroplast gene that hasbeen accurately mapped and whose nucleotide sequence has beendetermined. The predicted sequence of 81 amino acids has beenconfirmed in part by determination of the sequence of the first 40amino acids from the NH2 terminus of the protein, and it shows100% homology with the known amino acid sequence of the spin-ach protein but no more than 35% homology with the amino acidsequences of bacterial and mitochondrial proteins. The geneshows no deviation from the "universal" genetic code and is notsplit. A potential ribosome binding site is located 12 nucleotidesupstream from the initiation codon, but sequences homologous toprokaryotic promotors and transcription terminators are notapparent.

Chloroplasts, mitochondria, and bacteria contain a membrane-polypeptide complex that is capable of synthesizing ATP cou-pled to proton translocation across the membrane. The struc-ture of the ATP synthase complex from all these sources is sim-ilar, being composed of a peripheral component (CF1, F1, andBF1, respectively), showing latent ATPase activity, and an in-trinsic membrane component (CFO, FO, and BFO) that translo-cates protons across the membrane (see refs. 1-3 for reviews).This translocation, and consequent ATP synthesis, is inhibitedin all cases by dicyclohexylcarbodiimide (DCCD), a lipid-sol-uble compound that binds covalently to an 8-kilodalton (kDal)polypeptide of the membrane component. This low molecularweight polypeptide is highly hydrophobic and therefore issometimes called the DCCD-binding proteolipid (see ref. 4 fora review). Isolation of this polypeptide from chloroplasts andits incorporation into liposomes have been used to demonstrateits proton-translocating activity (5), although at least in Esche-richia coli other polypeptides may also be involved (6).The similarity of the size and function of this polypeptide in

chloroplasts, mitochondria, and bacteria strongly suggests acommon evolutionary origin. The location of the gene for thispolypeptide is therefore a matter of considerable evolutionaryinterest. In Neurospora, the mitochondrial polypeptide is en-coded by a nuclear gene (7), whereas in Saccharomyces the geneis located in the mitochondrial genome (8) and is transcribedand translated by the organellar protein synthetic apparatus (9).The chloroplast polypeptide has been shown to be synthesizedby isolated pea chloroplasts (10). This suggested that the codingsequence may be located within the organelle.

Using a cell-free transcription-translation system pro-grammed with wheat chloroplast DNA, we have found that thegene for the proton-translocating subunit is indeed located inchloroplast DNA. The gene has been accurately mapped on the

circular chloroplast chromosome and its nucleotide sequencehas been determined, allowing comparisons with the sequencesof the genes for the corresponding mitochondrial and bacterialpolypeptides.

MATERIALS AND METHODSChemicals. Restriction endonucleases and M13 strains mp7,

mp8, and mp9 were obtained from Bethesda Research Labo-ratories (Cambridge, U.K.) and the enzymes were used ac-cording to the manufacturer's instructions. L-[35S]Methionine(13 mCi/ml; 1,300 Ci/mmol; 1 Ci = 3.7 X 101° becquerels) waspurchased from Amersham. DCCD was obtained from BDH,calf intestinal phosphatase was from Boehringer Mannheim,and Quadrol was from Fluka.

In Vitro Transcription-Translation of DNA and Immuno-precipitation. Intact plasmid DNA (2 jig) or restriction frag-ments prepared as described (11) were incubated in an E. colicell-free transcription-translation system with 3 ,ul of L-[35S]methionine in a total volume of 50 k1., as described (11).A 2-,ul sample was taken for analysis of the total products andthe remainder was analyzed with a preimmune serum and thenwith antibodies raised against pea DCCD-binding proteolipid(10) as described by Sebald et al (12). Protein A-Sepharose wasused to precipitate antibodies, and polypeptides were analyzedby electrophoresis in 10-20% gradient polyacrylamide gels inthe presence of NaDodSO4 (13) and fluorography. The effectofDCCD on the mobility of the immunoprecipitated polypep-tide in NaDodSO4/polyacrylamide gels was shown by incuba-tion of the products of transcription and translation with 2 ,ulof 5 mM DCCD in ethanol for 16 hr at 25°C prior to immuno-precipitation.DNA Sequence Determination. Restriction fragments from

plasmid DNA were isolated from polyacrylamide or agarose gelsby the method of Dretzen et al. (14). EcoRI-Pst I fragmentswere cloned into M13 strains mp8 and mp9 by using the EcoRIand Pst I sites of those strains. Other fragments were clonedinto M13 strain mp7 by using the HincII sites. When necessary,fragments having "staggered" ends were rendered "flush" asdescribed by Sanger et al. (15). Cloning and production ofsingle-stranded DNA was essentially as described by Sanger et al. ex-cept that the digested mp7 replicative form DNA was treatedwith calf intestinal phosphatase according to the manufacturer'sinstructions, to prevent self-ligation, and it was not necessaryto use linkers. Sequence determination was carried out as de-scribed by Sanger et al. and the data were analyzed on a Hew-lett-Packard HP-85 computer with programs compiled by A.L. Phillips.

Abbreviations: DCCD, dicyclohexylcarbodiimide; kDal, kilodalton;kbp, thousand base pairs.t Present address: Beecham Pharmaceuticals, Clarendon Rd., Wor-thing, West Sussex BN14 8QH, U.K.

6903

The publication costs ofthis 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.

Dow

nloa

ded

by g

uest

on

July

11,

202

1

6Proc.Nati. Acad. Sci. USA 79 (1982)

Protein Sequence Determination. Amino. acid sequenceanalysis was performed by using a Beckman model 890C spin-ning-cup sequencer with program 030176 [a Beckman modifi-cation ofthe procedure ofBrauer et aL (16)] and 0.25 M Quadrolbuffer in the presence of Polybrene (17). Phenylthiohydantoin-amino acids were identified by reversed-phase HPLC with Zor-bax C18 [by a method modified from Zimmerman et al. (18)] andby TLC on silica gel (19).The DCCD-binding protein was treated with 1 M hydro-

chloric acid in methanol (20) for 25 min at 50-550C in the re-action cup in order to remove a putative formyl group at theamino terminus. Fluorescamine was used to decrease the back-ground contamination (21) when proline residues 24 and 43were exposed.

RESULTS AND DISCUSSIONLocalization of the Gene for the Proton-Translocating Sub-

unit. The chloroplast DNA of wheat (Triticum aestivum cv.Mardler) has been mapped with the restriction endonucleasesPst I and SalGI, and a large part of the DNA has been clonedas Pst I, SalGI, or BamHI fragments inserted into the corre-sponding sites of the vector pBR322 (22). Hybrid plasmid DNAwas used to program a cell-free coupled transcription-transla-tion system from E. coli. The [3S]methionine-labeled productswere analyzed by immunoprecipitation with monospecific an-tibodies (raised in rabbits against the pea proteolipid) and pro-tein A-Sepharose, followed by gradient NaDodSO4/polyacryl-amide gel electrophoresis and detection by fluorography. Twoplasmids were found that directed the synthesis of an 8-kDalpolypeptide immunoprecipitable with the antibodies. Diges-tion of the plasmid DNA with BamHI showed that both con-tained a 7.0-kilobase-pair (kbp) fragment of wheat chloroplastDNA, the fifth largest produced on digestion with BamHI, pre-viously designated B5. One plasmid, pTac60, was selected forfurther study.To confirm the identity of the polypeptide, the products of

transcription-translation ofpTac60 were incubated with DCCDfor 16 hr at 25°C. The binding ofDCCD under these conditionshas been shown to cause a significant decrease in the mobilityof the proton-translocating subunit on NaDodSO4/polyacryl-amide gel electrophoresis (10), and this was also found for theimmunoprecipitated transcription-translation product of pTac6O(Fig. 1). The mobility of the untreated immunoprecipitatedpolypeptide was the same as that of authentic wheat DCCD-binding proteolipid (data not shown), showing that the poly-peptide was not synthesized as a higher molecular weight formin this system. This contrasts with cytochromefand the 32-kDal"photogene" product, both membrane polypeptides that appearto be synthesized as higher molecular weight precursors priorto insertion in the thylakoid membrane (23-25).To locate the position of the gene more accurately, pTac6O

DNA was digested to completion with a number of restrictionenzymes and the linear fragments were used to program the E.coli cell-free transcription-translation system. The efficiency ofthis linear DNA as a template in this system was about 50% ofthat of undigested DNA and was adequate to determinewhether digestion of the DNA had abolished its ability to directthe synthesis of full-length DCCD-binding protein. Digestionwith Taq I, Sau3A, and EcoRI did not affect the production ofa full-length product, but digestion with Pst I completely abol-ished the synthesis of polypeptide immunoprecipitable withantibodies to the proteolipid. This indicates that there is at leastone Pst I site in or very close to the gene.The position of the 7.0-kbp BamHI fragment has been

-mapped on wheat chloroplast DNA in relation to the Pst I andSalGI sites (unpublished data). It contains only one Pst I site,

a b d

- mm-.9 kDaI

8 kL)aI

FIG. 1. Identification of the proton-translocating subunit as aproduct of coupled transcription-translation of pTac6O. Lanes: a, totalproducts of transcription-translation; b, result of immunoprecipitationwith a preimmune serum; c, result of immunoprecipitation with an-tibodies to pea DOOD-binding protein; d, as c except that the samplewas treated with DCCD prior to immunoprecipitation.

between the third and fourth largest fragments P3 and P4 (22),and thus allows precise localization of the gene. This is shownin Fig. 2, together with the positions of genes for the large sub-unit ofribulosebisphosphate carboxylase and those for the ,B andE subunits of ATP synthase (11, 22). The separation of theDCCD-binding proteolipid gene from these two other ATPsynthase genes contrasts sharply with the organization of thegenes in E. coli, in which the genes for all the subunits ofATPsynthase are arranged in a single operon (26, 27). In yeast mi-tochondria, the gene for the proton-translocating subunit (ATP-ase subunit IX) is also located some distance from the onlyother ATP synthase subunit gene (ATPase subunit VI) encodedby mitochondrial DNA (28).

Determination of Nucleotide Sequence. This was carried outby the dideoxy chain termination method of Sanger et al. (15)with the fragments indicated in Fig. 3. The nucleotide sequenceand predicted amino acid sequence of the gene are shown inFig. 4. The first 40 predicted amino acid residues have beenverified by protein sequence analysis, although it was not pos-sible by this means to differentiate unambiguously betweenglutamate and glutamine at amino acid positions 28 and 34. Theamino terminus of the protein is blocked, probably by formyl-ation, because treatment with 1 M HC1 in methanol for 1 hr at50-55°C, which is known to deformylate amino groups, re-sulted in partial unblocking ofit. The presence ofaformyl groupon the amino-terminal methionine is typical of prokaryotic pro-tein synthesis and also confirms the assertion above that thepolypeptide is not synthesized as a larger precursor.

Comparison of the amino acid sequence with that publishedfor the spinach DCCD-binding proteolipid (29) shows the strik-ing fact that they are identical over the entire 81 amino acidsof the protein. The corresponding spinach nucleotide sequence

6904 Biochemistry: Howe et aL

Dow

nloa

ded

by g

uest

on

July

11,

202

1

Proc. Natl. Acad. Sci. USA 79 (1982) 6905

5 1 5 0

5 -GAATTCCCTTCTATGTAGTTCGGACAATTCACATTATCATTTCAATTTGATTTCAATTTG

FIG. 2. Localization of the gene for the proton-translocating sub-unit. The map shows the position of the B5 fragment (unpublisheddata) and the proton-translocating subunit gene within it relative tothe positions of the fragments produced by SalGI (S1-S11) and Pst I(P1-P12) digestion. Also indicated are the positions of the 16S and 23SrRNA genes and the genes for the large subunit of ribulosebisphos-phate carboxylase (LS) and the ,B and E subunits of ATP synthase, con-tained within the second largest BamHI fragment (B2) (10, 21).

is not known, so it is not possible to say whether "silent" nu-cleotide changes have occurred between the two species. Thehomology is even greater than the 90% homology found be-tween the predicted amino acid sequences of the maize andspinach ribulosebisphosphate carboxylase large subunit poly-peptides (30). In contrast, there is only 20-30% homology be-tween the wheat or spinach amino acid sequence and those ofNeurospora crassa, Saccharomyces cerevisiae, and bovine mi-tochondrial subunits (which show only 50-60% homology withone another). There is also only 25-35% homology to the cor-responding proteins from E. coli or the thermophilic bacteriumPS-3 (29).

Aligning the nucleotide sequences in such a way as to max-imize amino acid homologies (29) indicates that the wheat chlo-roplast nucleotide sequence is 45% homologous to the E. colisequence and 41% homologous to the yeast mitochondrial one.As might be expected, the homology is greater in those regionsof the protein where the amino acid sequence is more con-served. The homology in the third codon position in both com-

I-

0

I9 100 2C

N

D

C) *- ""

ILII1

TACTTTTTAGTTACTTTACTTCTCCCCAATAGAGCTTAGAAGTAAGAATTTATTGGTTGATTG

TATCCTTAACCATTTCTTTTTTTTGACACGAGGAACTACTCACCATG MT CCA CTA ATT

Met Asn Pro Leu Ile

Met Asn Pro Leu Ile

50

GCT GCT GCT TCT GTT ATT GCT GCT GGA TTG GCC GTA GGG CTT GCT TCT

Ala Ala Ala Ser Val Ile Ala Ala Gly Leu Ala Val Gly Leu Ala Ser

Ala Ala Ala Ser Val Ile Ala Ala Gly Leu Ala Val Gly Leu Ala Ser

100

ATT GGA CCT GGA GTT GGT CM GGT ACT GCT GCA GGA CM GCT GTA GMA

Ile Gly Pro Gly Val Gly Gln Gly Thr Ala Ala Gly Gln Ala Val Glu

Ile Gly Pro Gly Val Gly Glx Gly Thr Ala Ala Gly Glx Ala Val Glu1 50

GGT ATT GCG AGA CAG CCA GM GCA GAA GGT AM ATA CGA GGT ACT TTA

Gly Ile Ala Arg Gln Pro Glu Ala Glu Gly Lys Ile Arg Gly Thr Leu

Gly Ile Ala - - Pro - Ala

200

TTG CTT AGT CTA GCT TTT ATG GM GCT TTA ACA ATT TAT GGA CTA GTT

Leu Leu Ser Leu Ala Phe Met Glu Ala Leu Thr Ile Tyr Gly Leu Val

GTG GCA CTA GCG CTT TTA TTT GCG AAC CCT TTT GTT TMTCTTAMAAAAA

Val Ala Leu Ala Leu Leu Phe Ala Asn Pro Phe Val *

3 0 0

ATTCTTTCGATTTCGATTAGATACTTTTTTCTTTTTTTAGTAMTTGGTATTTGCTTCCGCAA

TTCCMTTATATC-3

FIG. 4. Nucleotide sequence of the gene for the proton-translocat-ing subunit, together with the flanking regions and predicted and de-termined amino acid sequences. The upper amino acid sequence is pre-dicted from the nucleotide sequence by the "universal" genetic code.The lower sequence was derived by amino acid analysis of the wheatprotein. The recognition site for Pst I is underlined and the putativeribosome binding sites are doubly underlined. The nucleotide sequencerepresents residues 0-500 on Fig. 3.

parisons is less than the average homology (39% for yeast and25% for E. coli) as would be expected if some selection for con-servation ofamino acid sequence were acting. The fact that the

300 400 50,0 600 700

P"

0u

L12

120C

FIG. 3. Strategy for sequence determination of the gene for the proton-translocating subunit. The gene is contained within a 1.2-kbp EcoRIfragment, and only the restriction sites used for cloning are shown. The position of the gene is indicated and the letters N and C indicate the regionscoding for the amino and carboxyltermini, respectively, of the polypeptide.

Biochemistry: Howe et aL

)O

Dow

nloa

ded

by g

uest

on

July

11,

202

1

Proc. Natl. Acad. Sci. USA 79 (1982)

wheat sequence is more similar to the yeast sequence in thethird position but is closer to E. coli in the first and second po-sitions may be accounted for by a tendency for organelle ge-nomes to acquire adenine or thymidine residues in noncon-served positions (see below). It will be interesting to determineif other plants have chloroplast genes for this polypeptide andwhether some or all show the exceptionally high conservationof amino acid sequence we have observed between wheat andspinach. This will be particularly interesting for species suchas pea and Vicia, whose chloroplast DNA has been extensivelyrearranged in comparison with that of wheat (31, 32).

Translation of the wheat chloroplast gene starts at AUG andterminates at a single UAA stop codon. This is also the case withthe maize chloroplast gene for the large subunit of ribulosebis-phosphate carboxylase (33), whereas the spinach LS gene startswith AUG and terminates with UAG (30). A potential ribosomebinding site, formed by the overlapping sequences -G-A-G-G-or -A-G-G-A-, is located 12 nucleotides upstream from the startof translation. These tetranucleotides are complementary to asequence at the 3' end of 16S rRNA of maize (34) and tobacco(35), presumably allowing association of the ribosome with themessage prior to translation (36). A corresponding sequence isalso found close to the initiation codon in the E. coli proteolipidgene (37) but not in the mitochondrial one (8).The chloroplast gene, like the mitochondrial (8) and bacterial

(37, 38) genes, is not split, and there is no evidence of any de-viation from the universal genetic code. This is not surprisingin view of the efficiency of chloroplast DNA as a template fortranscription and translation by E. coli both in vivo (39) and invitro (40). Not all codons are used, however, because there isa marked preference for adenine or thymidine in the third po-sition of the codon (Fig. 5). Some 82% of codons have adenineor thymidine in the third position, compared with 82% in theyeast mitochondrial gene (8) but only 41% in the E. coli gene(37, 38). Other chloroplast genes (LS from maize and spinach)also show a high A+T preference in the third position (68% and71%, respectively). This A+T richness is also found in the flank-ing regions of both the chloroplast and mitochondrial (8) pro-teolipid genes, with long runs of adenine and thymidine whichare shown to a rather lesser extent by the flanking regions of

U C A GPhe 3 Ser 2 Tyr 1 Cys 0 U

U Phe 0 Ser 0 Tyr 0 Cys 0 CLeu 3 Ser 0 End 1 End 0 ALeu 2 Ser 0 End 0 Trp 0 GLeu 3 Pro 2 His 0 Arg 0 U

C Leu 0 Pro 0 His 0 Arg 0 CLeu 4 Pro 2 Gin 2 Arg 1 ALeu 0 Pro 0 Gin 1 Arg 0 GIle 5 Thr 2 Asn 1 Ser 1 U

A Ile 0 Thr 0 Asn 1 Ser 0 CIle 1 Thr 1 Lys 1 Arg 1 AMet 2 Thr 0 Lys 0 Arg 0 GVal 4 Ala 10 Asp 0 Gly 5 U

G Val 0 Ala 1 Asp 0 Gly 0 CVal 2 Ala 3 Glu 4 Gly 5 AVal 1 Ala 3 Glu 0 Gly 1 G

FIG. 5. Codon usage of the predicted mRNA sequence for the pro-ton-translocating subunit gene.

the E. coli gene (37). Therefore there would seem to be a strongtendency in organelle DNA for the acquisition of adenine andthymidine residues in nonconserved positions.

Although there are several sequences upstream from the cod-ing sequence that show some similarity to parts of the prokary-otic promotor, there is no clear homology with both the Pribnowbox and the -35 sequence (41). A similar difficulty in distin-guishing a promotor region has been reported for the maize LSgene (33), although sequences homologous to prokaryotic pro-motors precede the genes for the spinach LS (30), tobacco 16SrRNA (42), and some tRNA genes (43). Prokaryotic genes alsotypically show a sequence close to the point of termination oftranscription which is capable of forming a stem-and-loop struc-ture (41). Such sequences have been found in the maize (33) andspinach (40) large subunit genes, but none is present at the 3'end of the chloroplast proteolipid gene. This suggests eitherthat the chloroplast proteolipid gene uses transcription initia-tion and termination signals that are different from those ob-served to date or, alternatively, that the gene is part of a muchlarger transcript and would therefore not necessarily show ad-jacent promotor or terminator sequences. The latter is sup-ported by preliminary hybridization analysis (unpublished data)with nick-translated pTac6O DNA, which suggests that tran-scripts of the proteolipid gene of a similar size to the gene itselfare not found in wheat chloroplasts.

We thank A. Coulson, F. Sanger, and R. Thompson for help withDNA sequence analysis, D. Bringloe for purifying wheat chloroplastproteolipid, and A. Phillips and I. Small for help and advice. C.J.H.is the recipient of a Research Councils Co-operative Award from theScience and Engineering Research Council and the Agricultural Re-search Council.

1. Nelson, N. (1981) Curr. Top. Bioenerg. 11, 1-33.2. Criddle, R. S., Johnston, R. F. & Stack, R. J. (1979) Curr. Top.

Bioenerg. 9, 90-145.3. Fillingame, R. H. (1981) Curr. Top. Bioenerg. 11, 35-106.4. Sebald, W. & Hoppe, J. (1981) Curr. Top. Bioenerg. 12, 2-64.5. Nelson, N., Eytan, E., Notsani, B.-E., Sigrist, H., Sigrist-Nel-

son, K. & Gitler, C. (1977) Proc. NatL Acad. Sci. USA 74, 2375-2378.

6. Friedl, P., Bienhaus, G., Hoppe, J. & Schairer, H. U. (1981)Proc. NatL Acad. Sci. USA 78, 6643-6646.

7. Sebald, W., Sebald-Althaus, M. & Wachter, E. (1977) in Ge-netics and Biogenesis of Mitochondria, eds. Bandlow, W.,Schweyen, R. J., Wolf, K. & Kaudewitz, F. (de Gruyter, Berlin),pp. 433-440.

8. Hensgens, L. A. M., Grivell, L. A., Borst, P. & Bos, J. L. (1979)Proc. Nati Acad. Sci. USA 76, 1663-1667.

9. Sierra, M. F. & Tzagoloff, A. (1973) Proc. NatL Acad. Sci. USA 70,3155-3159.

10. Doherty, A. & Gray, J. C. (1980) Eur. J. Biochem. 108, 131-136.11. Howe, C. J., Bowman, C. M., Dyer, T. A. & Gray, J. C. (1982)

MoL. Gen. Genet. 186, 525-530.12. Sebald, W., Graf, T. & Lukins, H. B. (1979) Eur. J. Biochem. 93,

587-599.13. Chua, N.-H. & Bennoun, P. (1975) Proc. Nati Acad. Sci. USA 72,

2175-2179.14. Dretzen, G., Bellard, M., Sassone-Corsi, P. & Chambon, P.

(1981) AnaL Biochem. 112, 295-298.15. Sanger, F., Coulson, A. R., Barrell, B. G., Smith, A. J. H. &

Roe, B. A. (1980)J. MoL BioL 143, 161-178.16. Brauer, A. W., Margolies, M. N. & Harber, E. (1975) Biochem-

istry, 14, 3029-3035.17. Tarr, G. E., Beecher, J. F., Bell, M. & McKean, D. J. (1978)

Anal. Biochem. 84, 622-627.18. Zimmerman, C. L., Apella, E. & Pisana, J. J. (1977) AnaL

Biochem. 77, 569-573.19. Bridgen, J., Graffeo, A. P., Karger, B. L. & Waterfield, M. D.

(1975) in Instrumentation in Amino Acid Sequence Analysis, ed.Perham, R. N. (Academic, New York), pp. 111-145.

6906 Biochemistry: Howe et al.

Dow

nloa

ded

by g

uest

on

July

11,

202

1

Biochemistry: Howe et al.

20. Sheehan, J. C. & Yang, D.-D. H. (1958) J. Am. Chem. Soc. 80,1154-1158.

21. Brown, A. S., Bennett, J. C., Morgan, P. H. & Mole, J. E. (1981)AnaL Biochem. 112, 158-162.

22. Bowman, C. M., Koller, B., Delius, H. & Dyer, T. A. (1981) Mol.Gen. Genet. 183, 93-101.

23. Willey, D. L., Huttly, A. K., Phillips, A. L. & Gray, J. C. (1983)Mol Gen. Genet., in press.

24. Grebanier, A. E., Coen, D. M., Rich, A. & Bogorad, L. (1978)J. Cell Biol 78, 734-746.

25. Reisfield, A., Mattoo, A. K. & Edelman, M. (1982) Eur. J.Biochem. 124, 125-129.

26. Gibson, F., Downie, J. A., Cox, G. B. & Radik, J. (1978)J. Bac-teriot 134, 728-736.

27. Downie, J. A., Senior, A. E., Gibson, F. & Cox, G. B. (1978)J.Bacteriol 137, 711-718.

28. Borst, P. & Grivell, L. A. (1981) Nature (London) 290, 443-444.29. Sebald, W., Hoppe, J. & Wachter, E. (1979) in Functional and

Molecular Aspects of Biomembrane Transport, eds. Quagli-ariello, E., Palmieri, F., Papa, S. & Klingenberg, M. (Elsevier/North-Holland, Amsterdam), pp. 63-74.

30. Zurawski, G., Perrot, B., Bottomley, W. & Whitfeld, P. R. (1981)Nucleic Acids Res. 9, 3251-3270.

Proc. Nati. Acad. Sci. USA 79 (1982) 6907

31. Palmer, J. D. & Thompson, W. F. (1981) Proc. Natl Acad. Sci.USA 78, 5533-5537.

32. Koller, B. & Delius, H. (1980) Mol Gen. Genet. 178, 261-269.33. McIntosh, L., Poulsen, C. & Bogorad, L. (1980) Nature (London)

288, 556-560.34. Schwarz, Zs. & Kossel, H. (1980) Nature (London) 283, 739-742.35. Tohdoh, N. & Sugiura, M. (1982) Gene 17, 213-218.36. Shine, J. & Dalgarno, L. (1974) Proc. Natl Acad. Sci. USA 71,

1342-1346.37. Gay, N. J. & Walker, J. E. (1981) Nucleic Acids Res. 9, 3919-

3926.38. Kanazawa, H., Mabuchi, K., Kayano, T., Tamura, F. & Futai,

M. (1981) Biochem. Biophys. Res. Commun. 100, 219-225.39. Gatenby, A. A., Castleton, J. A. & Saul, M. W. (1981) Nature

(London) 291, 117-121.40. Bottomley, W. & Whitfeld, P. R. (1979) Eur.J. Biochem. 93, 31-

39.41. Rosenberg, M. & Court, D. (1979) Annu. Rev. Genet. 13, 319-

353.42. Tohdoh, N., Shinozaki, K. & Sugiura, M. (1981) Nucleic Acids

Res. 9, 5399-5406.43. Takaiwa, F. & Sugiura, M. (1980) Mol Gen. Genet. 180, 1-4.

Dow

nloa

ded

by g

uest

on

July

11,

202

1