Essential Residues in the Polar Loop Region of Subunit c of

Vol. 173, No. 8

Essential Residues in the Polar Loop Region of Subunit c ofEscherichia coli F1F0 ATP Synthase Defined by Random

Oligonucleotide-Primed MutagenesisDEAN FRAGA AND ROBERT H. FILLINGAME*

Department of Physiological Chemistry, University of Wisconsin Medical School,Madison, Wisconsin 53706

Received 13 November 1990/Accepted 4 February 1991

The conserved, polar loop region of subunit c of the Escherichia coli F1F0 ATP synthase is postulated tofunction in the coupling of proton translocation through Fo to ATP synthesis in F1. We have used a randommutagenesis procedure to define the essential residues in the region. Oligonucleotide-directed mutagenesis wascarried out with a random mixture of mutant oligonucleotides, the oligonucleotide mixture being generated bychemical synthesis by using phosphoramidite nucleotide stocks that were contaminated with the other threenucleotides. Thirty mutant genes coding single-amino-acid substitutions in the region between Glu-37 andLeu-45 of subunit c were tested for function by analyzing the capacity of plasmids carrying the mutant genesto complement a Leu-4-->amber subunit c mutant. All substitutions at the conserved Arg-41 residue resulted inloss of oxidative phosphorylation, i.e., transformants could not grow on a succinate carbon source. The otherconserved residues were more tolerant to substitution, although most substitutions did result in impairedgrowth on succinate. We conclude that Arg-41 is essential in the function of the polar loop and that theensemble of other conserved residues collectively maintain an optimal environment required for that function.

The F1F0 ATP synthase in the inner membrane of Esche-richia coli catalyzes the synthesis of ATP during oxidativephosphorylation in a reaction that is driven by a protonelectrochemical potential generated by a proton-pumpingelectron transport system (23). Mitochondria and chloro-plasts contain similar ATP synthases that are responsible forthe synthesis of ATP during oxidative and photosyntheticphosphorylation. The ATP synthases are composed of twofunctionally distinct sectors termed F1 and Fo. The F1portion is easily disassociated from the membrane, and inthis soluble form, it retains the capacity to hydrolyze ATP.The Fo portion traverses the membrane, and when isolated,it promotes the passive translocation of protons across themembrane. The complete F1F0 complex reversibly couplesthe synthesis or hydrolysis of ATP to proton translocationacross the membrane (5, 23). In E. coli, F1 is composed offive types of subunits in a stoichiometry of X3P3'Y181E1 and Fois composed of three types of subunits in a stoichiometry ofa1b2c1011 (6). Each of the genes encoding the F1 and Fosubunits of E. coli is located in the unc operon, which hasbeen sequenced (25).

Subunit c of Fo is a hydrophobic protein of 79 amino acidsthat is predicted to form a hairpinlike structure with twoa-helices inserted through the membrane (2, 5, 22). The loopportion of the hairpin is relatively polar in amino acidcomposition and extends from the cytosolic side of themembrane (8, 9). Subunit c plays a key role both in protontranslocation across the membrane and in the coupling ofproton translocation to ATP synthesis. Aspartyl 61 at thecenter of the second membrane traversing helix is thought toundergo a protonation-deprotonation cycle during protontranslocation (5). Mutations changing Asp-61 to Asn or Glyabolish proton translocation (11, 12). The polar loop regionof subunit c is proposed to function in the binding of F1 to Fo

* Corresponding author.

and in the coupling of proton translocation through Fo toATP synthesis or hydrolysis in F1 (5, 7, 17, 19). Substitutionof Glu for Gln-42 results in total loss of coupled function(19). Other substitutions in the conserved Gln-42 residue(e.g., Gly, Ala, or Val) lead to less severe defects in coupling(7). Substitutions of Ala or Ser for Pro-43 also lead to minordefects in coupling (17). Since Gln-42 and Pro-43 are con-served across a wide range of species (13), we were sur-prised that function was largely retained in these mutants.

In this study, we have attempted to determine whetherany of the residues of the polar loop region were absolutelyessential for function. We used randomly mutated oligonu-cleotides to generate a series of mutations in the polar loopregion of the gene. The method is more convenient thancassette mutagenesis in that new restriction sites do not haveto be introduced. We conclude that the conserved Arg-41residue is the only residue in the region that is indispensablefor function. The other residues appear to confer an optimalstructure or environment but are not individually essentialfor that function.

MATERIALS AND METHODS

E. coli strains, phage, and plasmids. Strain TG-1 [supE thihsdD5 A(lac-pro) F' (traD36 proA+B+ lacIq lacZAM15)] wasprovided by Amersham Corporation (Chicago, Ill.). StrainCAG1688 [araD139 A(ara leu)7697 AlacX74 galU galK hsrhsm+ rpsL F' (lacIq lacZ::TnS proA+B+)] was provided byCarol Gross (University of Wisconsin-Madison). StrainMJM413 [F+ asnA+ asnB31 thi-J recA56 srl-1300::TnJOuncEJ003(Leu-4--*amber)] was constructed by Michael J.Miller in this laboratory. Strain JM83 [ara A(lac proAB) rpsL1'80(lacZAM15)], phage M13mpl8, and plasmid pUC18 aredescribed elsewhere (26). Strain DF514 was constructed bymating strain CAG1688 with strain MJM413 with selectionfor transfer of the F' containing the lacIq and lacZ::Tn5(Kan') markers by plating onto minimal agar containing

2639

JOURNAL OF BACTERIOLOGY, Apr. 1991, p. 2639-26430021-9193/91/082639-05$02.00/0Copyright © 1991, American Society for Microbiology

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

07

Janu

ary

2022

by

192.

162.

179.

117.

2640 FRAGA AND FILLINGAME

glucose, thiamine, and kanamycin (50 ,ug/ml). The M13phage F-2 (17) contains the BamHI 1561-PvuII 2376 unc

DNA fragment cloned between the BamHI and HincII sitesof M13mpl8. (The nucleotide numbering system for unc

DNA is that described by Walker et al. [25].)General techniques. M63 minimal medium (16) was sup-

plemented with 0.2% glucose or 0.6% disodium succinatehexahydrate as a carbon source and 2 ,ug of thiaminehydrochloride per ml. LB and 2YT media are described byMiller (16). Ampicillin was added when appropriate to a finalconcentration of 50 jig/ml (for minimal or LB medium) or 100p.g/ml (2YT medium). M13 phage was grown in strain TG-1.Phage RF DNA and plasmid DNA were prepared from cellpellets by a procedure utilizing alkaline lysis and purificationby polyethylene glycol precipitation (14). This procedureyielded DNA that was suitable for double-stranded DNAsequencing as previously described (14). Restriction enzymedigestions and ligation reactions were done according to thegeneral recommendations of the manufacturers. Single-stranded phage DNA template was prepared, and chaintermination DNA sequencing (21) was performed as de-scribed in the U.S. Biochemical Corporation Sequenase Kit.

Oligonucleotide-directed mutagenesis. Oligonucleotideswere synthesized by the University of Wisconsin Biotech-nology Center and purified by reverse-phase high-pressureliquid chromatography as previously described (17). Thesynthesis procedure involved the stepwise addition of di-methyltrityl nucleotides to a solid-phase resin that had thefirst nucleotide of the oligonucleotide attached to it. Theoligonucleotide primers Im and lwt (5'-TTGACGCGCTGCGCCTTC-3') corresponded to amino acid residues Glu-37 toGln-42 of subunit c. The oligonucleotides 2m and 2wt (5'-CAGATCAGGTTGACGCGC-3') corresponded to amino acidresidues Ala-40 to Leu-45 of subunit c. The primers desig-nated Im and 2m were made by using "contaminated"dimethyltrityl nucleotide stocks containing 1.85% each ofthe three non-wild-type nucleotides, in a procedure similarto that of Derbyshire et al. (3). For each round of nucleotidesynthesis, the frequency of errant nucleotide incorporationwas 1.85% x 3. At this frequency of contamination, approx-imately 38% of the oligonucleotides of mixed primers Im and2m were predicted to have the wild-type sequence, 38%were predicted to contain one base change, 18% were

predicted to contain two base changes, and 6% were pre-dicted to be all other possibilities (see footnote a, Table 1).Because of the nature of the synthesis procedure, the firstbase is fixed in its sequence. The primers lwt and 2wt were100% wild type in their sequences and were used as probesto detect mutant phage by hybridization.The uncE M13mpl8 recombinant phage F-2 (17) was used

as the template for site-directed mutagenesis by the methodof Taylor et al. (24), by using an Amersham Corporation invitro mutagenesis kit. The Amersham protocol for mutagen-esis was modified such that a template-to-primer ratio of 2:1instead of 1:2 was used. This change was introduced toprevent wild-type oligonucleotides that were present in themixed primers Im and 2m from preferentially binding thetemplate to the exclusion of the mutant oligonucleotides.Phage plaques were screened by hybridization on nitrocel-lulose membranes (20), by using 32P-labeled wild-type oligo-nucleotides (Jwt or 2wt) with washes at 25, 42, 55, and 72°C.Phage that released the hybridization probes at lower tem-peratures than the wild-type phage were further screened byDNA sequencing. The entire uncE gene of each mutant wassequenced to ensure that there were no additional mutationselsewhere in the gene.

TABLE 1. Efficiency of mutagenesis by two contaminatedoligonucleotides

Primer

im 2m ExpectedMeasurement frequencya

Fre- Fre-Total quency Total quency

(%) (%)

Total phage screened by 191 100 73 100 100dot blot hybridization

Mutant phage found 50 26 33 45 44Mutant phage with:

Single base change 50 26 21 29 26Two base changesb 0 0 7 10 12Three base changesb 0 0 0 0 4Four base changesb 0 0 1 1 1

Deletions/rearrangements 0 0 4 5 0

a The expected frequency is calculated by the following formula (10): P(x)= [N!/x!(N - x)!]CWMAl - CN-x, where P(x) is the frequency of oligonucle-otides with x base changes in the entire population, x is the number of basechanges, N equals the number of mutable bases in the oligonucleotide (17 forthe 18 base oligonucleotides synthesized here), C equals the chance of puttingin the wrong base at any mutable position during synthesis (5.55%), and Mfequals the mutagenic efficiency of the mutagenesis procedure, which Amer-sham Corporation states is a minimum of 70% for the kit used.

b Of the mutants with two base changes, one (Arg-41Trp) had both changesin the same codon, two resulted in two-amino-acid changes (Asp-44Glu,Gln-42Leu; Ala-39Glu,Gly-38Val), and four resulted in one amino acid changeand one silent base change. The mutant that had four base changes was adouble mutant (Arg-41His,Leu-45Pro) with a silent change (Pro-43Pro).

Subcloning the uncE mutant alleles from M13 into pUC18.The BamHI-PstI DNA fragment of the mutant phage con-tains the uncE gene within unc nucleotides 1727 to 2376. Formost mutants, this fragment was cloned between the BamHIand PstI sites of the plasmid pUC18. Alternatively, the uncEgene within the BamHI 1727-to-HpaI 2162 fragment of theunc DNA was cloned between the BamHI and HincIl sitesof pUC18 in the case of mutants Pro-43Ser (a mutant inwhich Ser is substituted for Pro-43) (17), Pro-43Ala (17),Pro-43Leu, Asp-44Glu/Gln-42Leu, Pro-43His, and Gln-42His and a wild-type control.

Complementation of an uncE amber mutation. StrainDF514 was transformed with the mutant uncE derivatives ofthe pUC18 plasmid. Transformant colonies were picked witha sterile toothpick and spread in patches on succinateminimal agar plates containing 0, 0.4, 4.0, or 40 ,uM isopro-pylthio-p-D-galactoside (IPTG). These patches were allowedto incubate at 37°C for 2 to 6 days and were scored forgrowth. To more quantitatively compare the growth ofmutants, liquid LB cultures were grown overnight, and afterdilution, 50 to 200 cells were spread onto succinate minimalagar plates. The time of colony appearance and colony sizewere recorded after 2 to 6 days of growth at 37°C. Cells weregrouped into three categories: "like wild type" (coloniesappearing by day 3 and reaching .1 mm by day 6), "nil" (nogrowth on succinate), and "slow growth" (colonies smallerthan wild type appearing between 3 and 6 days).

RESULTS

Efficiency of mutagenesis. The procedure described aboveusing contaminated oligonucleotide primers generated awide range of base substitutions (Fig. 1). Many of themutations observed were found more than once. The twomutagenic primers, Im and 2m, gave different mutationfrequencies (Table 1). On the basis of an index of dispersion

J. BACTERIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

07

Janu

ary

2022

by

192.

162.

179.

117.

POLAR LOOP MUTANTS OF SUBUNIT c 2641

GG

CGCACAT A

T TA C AT TA C G A AT TT AGG C A T

Mutations T TT AAG C A ATFound *CT TT TAG TT A TT

Sense GAA GGC GCA GCG CGT CAAStrand

Wdd-type Glu Gly Ala Ala Arg Gln

Amino Add Glu GAAa [ rg GlnSubstitutions Val Val Glu [ kArg

[la] Ala | Glu Gly ProGly Asp -UV [ er LeuLys Ser Thr Thr His LysGln Arg Pro Pro Pro GluAspEI Cys Ser Ser Leu HisUM UM

C CG A

G G A TMutations TAG GT AAA T

Found * AA TTA GT T AT A T

CCG

A TGC

Sense GCG CGT CAA CCT GAT CTGStrand

Wld-type AJa Arg Gln Pro Asp Leu

Amino Acid AlaSubsttutons Val

GluGlyThrProSer

Gln Pro| Asp LeuCsAg Leu Gly Pr

Ser Pro His Ala GlnGly| Lu | Arg Val Ar

is Lvs- Ser Glu |MetPro nTVal[Lou His Ala His

UM Tyr

Primer 1 Primer 2FIG. 1. Results of contaminated-oligonucleotide mutagenesis with two overlapping oligonucleotides covering the Glu-37-to-Gln-42 and

Ala-40-to-Leu-45 regions of subunit c. The distribution of base changes is shown above the wild-type sense strand. Each letter represents amutation found at that nucleotide position. The asterisk (*) above the first base indicates that the base was not mutable because it was theinitial base attached to the resin. The contaminated oligonucleotide used for the mutagenesis was complementary to the sense strandsequence. Below the base sequence is the wild-type amino acid sequence and all possible amino acid substitutions that can be generated bya single base change of the wild-type codon. The boxed amino acids are those actually found and include three double mutants (Table 2). Themutant Arg-41Trp is not shown because it resulted from two base changes in the same codon.

calculation (10), primer 2m was concluded to have yielded arandom distribution of mutants across the sequence. Thisprimer also gave close to the expected frequency of single,double, and quadruple mutants (Table 1). Mutants contain-ing up to four substitutions in the region covered by thisoligonucleotide were isolated. On the other hand, the distri-bution of mutants generated by primer Im was concluded tobe nonrandom.

Assay for function by complementation. The mutant uncEgenes were subcloned from phage RF DNA into plasmidpUC18, where transcription of the cloned uncE gene wouldbe under the control of the lac promoter. The pUC18 cloneswere then transformed into strain DF514 [uncE1003(Leu-4-+amber) lacI"], and the transformants were tested for theirability to grow on succinate. Translation of the uncE1003gene in strain DF514 is terminated by the Leu-4--+ambermutation, and the lacIq gene produces lac repressor insufficient amounts to ensure repression of the lac promoterin the multicopy pUC18 plasmid. The transformed strainswere first tested on succinate minimal medium plates withvarious concentrations of the lac inducer, IPTG. Optimalgrowth of uncE+IpUC18 (wild-type) transformants occurredat IPTG concentrations of 0 and 0.4 ,uM. Growth on succi-nate was reduced at 4 ,uM IPTG and abolished at 40 ,uMIPTG. The same general pattern was observed for themutant uncE plasmids that promoted growth on succinate.Mutants that did not grow on succinate at any of the IPTG

concentrations tested are designated "nil" in Table 2. Thosemutants that were capable of some growth on succinate, asmeasured by the patch test experiment, were grown as singlecolonies on succinate minimal agar plates lacking IPTG, andthe colony sizes were compared with those of the wild type.In Table 2, these mutants are grouped into two categories:those that grew as well as the wild type on succinate(designated "like wild type") and those that grew moreslowly than the wild type on succinate (designated "slow-er"). Transformants in the slower growth category typically

produced colonies after 6 days that were c0.5 mm, versus a1.0 mm colony size for the uncE+ transformants. A few ofthe slower growing transformants were intermediate in sizeafter 6 days (i.e., between 0.5 and 1.0 mm). In order todetermine the sensitivity of this assay, transformants con-taining pUC18 clones of the Pro-43Ser and Pro-43Ala muta-tions were plated on succinate minimal agar plates, and theircolony sizes were compared with those of the wild type.Both mutations, when present in the chromosome, result ingrowth yields on succinate of -90% that of the wild type(17). When the Pro-43Ala and Pro-43Ser mutations weretested with the pUC18 complementation system describedhere, they both resulted in colonies smaller than did theequivalent uncE+IpUC18 plasmid (i.e., .0.5 mm versus 1.0mm). The assay is thus sensitive enough to distinguishmutations that have relatively minor effects on growth andfunction.

DISCUSSION

We report here 30 new mutants with single-amino-acidsubstitutions between Glu-37 and Leu-45 in the polar loopregion of subunit c. We conclude that Arg-41 is the onlyessential residue in the conserved polar loop region ofsubunit c. Although many of the other residues are con-served (Fig. 2), they are not crucial. It is noteworthy,however, that even minor changes in this region resulted inimpaired growth (function), so there seems to be an advan-tage conferred by each of the wild-type residues. A preferredsequence has clearly evolved that most likely optimizesfunction, even though the individual residues in that se-quence are not absolutely essential for function. SinceArg-41 is clearly essential, the environment conferred by thesurrounding residues may be crucial to the function. Asingle-amino-acid change might lead to rather small effectson function because the other residues collectively maintainthe essential features of the environment conferred by the

VOL. 173, 1991

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

07

Janu

ary

2022

by

192.

162.

179.

117.

2642 FRAGA AND FILLINGAME

TABLE 2. Relative growth rates of mutants with succinateas a carbon source

Substitution(s) resulting in the followingWild-type growth categorya:residue

Like wild type Slower Nil

Glu-37 Asp Ala

Gly-38 AlaVal

Ala-39 Gly Ser GluThr Pro

Ala-40 Val GluSerPro

Arg-41 CysSerLeuTrpHisLysb

Gln-42 Leu GluHis Arg

Pro-43 His LeuThrAlabSerb

Asp-44 Asn His Tyr

Leu-45 ProArg

Gly-38, Ala-39 Val-38, Glu-39

Gln-42, Asp-44 Leu-42, Glu-44

Arg-41, Leu-45 His-41, Pro-45

a "Like wild type" indicates that the mutation results in growth like that ofthe wild type on succinate (colony size .1.0 mm after 6 days), "slower"indicates that the mutation results in growth slower than that of the wild type(colony size c0.5 mm colonies after 6 days), and "nil" indicates that themutation results in no growth on succinate after 6 days.

I These mutations were generated individually by oligonucleotide-directedmutagenesis as described by Miller et al. (17) and Fraga et al. (7a). See text.

ensemble of amino acids. Only drastic changes that, forexample, introduce a charged group (Ala-40Glu) or a signif-icant structural change (Asp-44Tyr) would appear to disruptstructure sufficiently to result in nil growth on succinate.

Substitutions that result in total loss of function can beviewed as defining the limits of what can be tolerated at eachposition. For example, in the sequence Gly-38-Ala-39-Ala-40 it appears that small, uncharged amino acids can beaccommodated but that a charged amino acid such asglutamate is disruptive. When the results from substitutionsat Gln-42 are combined with the results from our previousstudy (i.e., Gln-42-->Ala, Gly, or Val [7]), it is clear thatmany amino acids can be accommodated at this position butthat charged residues, such as glutamate or arginine, aredisruptive. Similarly, at conserved Pro-43, a surprising va-riety of substitutions are accommodated but the bulky Leu isnot.

It is of interest that the Gly-38Ala substitution in E. coli

E. coliPS-3S. faecalisM. laminosusA. caldanusSynechococusSpinach chlorop.R. rubnumA. nidulansYeast mito.N. crassa mito.Bovine mito.Plant mito.

Amino Acid Number

37 38 39 40 41 42 43 44 45

Glu Gly Ala Ala Arg Gin Pro Asp LouGlu Gly lie Ala Arg Gln Pro Glu LouGlu Gly Met Ala Arg Gln Pro Glu MetGlu Gly lie Ala Arg Gln Pro Glu AlaGlu Gly Val Ala Arg Gin Pro Glu AlaGlu Gly lie Ala Arg Gin Pro Glu AlaGlu Gly lie Ala Arg Gin Pro Glu AlaSer Thr Val Gly Arg Asn Pro Ala AlaLeu Ala Val Ser Arg Asn Pro Ala LoAsn Gly Val Ser Arg Asn Pro Ser lleAsn Gly Val Ala Arg Asn Pro Ala Leulie Gly Tyr Ala Arg Asn Pro Ser LeuHis Ser Val Ala Arg Asn Pro Ser Lou

FIG. 2. Sequence comparisons in the conserved polar loop ofsubunit c of different species. The amino acid numbering system isfor the E. coli protein. The sequences are from the followingsources: E. coli (13), PS-3 (13), Streptococcus faecalis (13),Mastigocladus laminosus (13), Acido caldarius (13), Synechococcusstrain 6301 (1), spinach chloroplast (chlorop.) (13), Rhodospirillumrubrum (13), Aspergillus nidulans (13), Saccharomyces cerevisiaemitochondria (Yeast mito.) (13), Neurospora crassa (13), bovinemitochondria (13), and plant mitochondria (Zea mays [4] andPetunia hybrida [27]).

results in slower growth on succinate despite the fact thatAla is found at this position in other species (Fig. 2). Asimilar example has been reported by Miller et al. (18). Thereplacement of Asp-61 with Glu resulted in a major decreasein function despite the fact that Glu appears at the equivalentposition in all other species of subunit c. Both examplessuggest that there may be complementary residues in theinteracting segments of protein that place restraints on thepacking around these crucial residues. Lim and Sauer (15)have studied the effects of different single, double, and triplemutations at conserved amino acid positions on the stabilityof the lambda repressor protein. They concluded that theeffects of a given mutation are determined by the identity ofthe amino acids around it and were able to correlate theindividual effects of the different mutations with the specificpacking arrangements of the amino acids in the foldedprotein.

In summary, we have examined a large number of muta-tions in a defined, highly conserved sequence in the polarloop region of subunit c. Only Arg-41 appears to be essential.Models attempting to explain the coupling of proton trans-location to ATP synthesis should include a specific role forthis arginine. Mutations at other positions in the conservedpolar loop region generally resulted in reduced growth onsuccinate. Relatively drastic changes were required to abol-ish growth totally. We suggest that these residues mayprovide a general environment for Arg-41 to fulfill its morespecific role. The observation that other species have aslightly different combination of residues in the polar loopindicates that there may be more than one combination ofresidues with the capacity of providing that general environ-ment. Finally, the contaminated oligonucleotide mutagene-sis procedure used here offers many of the advantages ofcassette mutagenesis but does not require construction ofrestriction sites within the gene prior to mutagenesis.

J. BACTERIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

07

Janu

ary

2022

by

192.

162.

179.

117.

POLAR LOOP MUTANTS OF SUBUNIT c 2643

ACKNOWLEDGMENTS

We thank N. Drinkwater for his help in the statistical analysispresented here. We thank M. J. Miller for strain MJM413 and hismany helpful comments.

This work was supported by Public Health Service grant GM-23105 from the National Institutes of Health.

REFERENCES1. Cozens, A. L., and J. E. Walker. 1987. The organization and

sequence of the genes for ATP synthase subunits in the Cyano-bacterium synechococcus 6301: support for an endosymbioticorigin of chloroplasts. J. Mol. Biol. 194:359-383.

2. Deckers-Heberstreit, G., R. Schmid, H. Kiltz, and K. Altendorf.1987. Fo portion of Escherichia coli ATP synthase: orientationof subunit c in the membrane. Biochemistry 26:5486-5492.

3. Derbyshire, K. M., J. J. Salvo, and N. D. F. Grindley. 1986. Asimple and efficient procedure for saturation mutagenesis usingmixed oligonucleotides. Gene 46:145-152.

4. Dewey, R. E., A. M. Schuster, C. S. Levings HI, and D. H.Timothy. 1985. Nucleotide sequence of Fo ATPase proteolipid(subunit 9) gene of maize mitochondria. Proc. Natl. Acad. Sci.USA 82:1015-1019.

5. Fillingame, R. H. 1990. Molecular mechanics of ATP synthesisby F1F0-type H+-transporting ATP synthases, p. 345-391. InT. A. Krulwich (ed.), The bacteria: a treatise on structure andfunction, vol. 12. Academic Press, Inc., New York.

6. Foster, D. L., and R. H. Fillingame. 1982. Stoichiometry ofsubunits in the H+-ATPase complex of Escherichia coli. J. Biol.Chem. 257:2009-2015.

7. Fraga, D., and R. H. Fillingame. 1989. Conserved polar loopregion of Escherichia coli subunit c of the F1F0 H+-ATPase:glutamine 42 is not absolutely essential, but substitutions alterbinding and coupling of F1 to Fo. J. Biol. Chem. 264:6797-6803.

7a.Fraga, D., D. J. Miller, M. Oldenberg, and R. H. Fillingame.Unpublished data.

8. Girvin, M. E., J. Hermolin, R. Pottorf, and R. H. Fillingame.1989. Organization of the Fo sector of Escherichia coli H+-ATPase: the polar loop region of subunit c extends from thecytoplasmic face of the membrane. Biochemistry 28:4340-4343.

9. Hensel, M., G. Deckers-Hebestreit, R. Schmid, and K. Altendorf.1990. Orientation of subunit c of ATP synthase of Escherichiacoli-a study with peptide-specific antibodies. Biochim.Biophys. Acta 1016:63-70.

10. Hoel, J. G. 1943. On indices of dispersion. Ann. Math. Stat.34:65-72.

11. Hoppe, J., H. U. Schairer, P. Freidi, and W. Sebald. 1982. AnAsp-Asn substitution in the proteolipid of a mutant ATPasefrom Escherichia coli leads to a non-functional proton channel.FEBS Lett. 145:21-24.

12. Hoppe, J., H. U. Schairer, and W. Sebald. 1980. The proteolipid

of a mutant ATPase from Escherichia coli defective in H+-conduction contains a glycine instead of the carbodiimide reac-tive aspartyl residue. FEBS Lett. 109:107-111.

13. Hoppe, J., and W. Sebald. 1984. The proton conducting Fo partof bacterial ATP synthases. Biochim. Biophys. Acta 768:1-27.

14. Kraft, R., J. Tardiff, K. S. Krauter, and L. A. Leinwand. 1988.Using mini-prep plasmid DNA for sequencing double strandedtemplates with sequenase. Biotechniques 6:544-546.

15. Lim, W. A., and R. T. Sauer. 1989. Alternative packing arrange-ments in the hydrophobic core of lambda repressor. Nature(London) 339:31-35.

16. Miller, J. H. 1972. Experiments in molecular genetics. ColdSpring Harbor Laboratory, Cold Spring Harbor, N.Y.

17. Miller, M. J., D. Fraga, C. R. Paule, and R. H. Fillingame. 1989.Mutations in the conserved proline 43 residue of the uncEprotein (subunit c) of Escherichia coli F1F0 alter the coupling ofF1 to Fo. J. Biol. Chem. 264:305-311.

18. Miller, M. J., M. Oldenburg, and R. H. Fillingame. 1990. Theessential carboxyl group in subunit c of the F1F0 ATP synthasecan be moved and H+-translocating function retained. Proc.Natl. Acad. Sci. USA 87:4900-4904.

19. Mosher, M. E., L. K. White, J. Hermolin, and R. H. Fillingame.1985. H+-ATPase of Escherichia coli: an uncE mutation impair-ing coupling between F1 and Fo but not Fo-mediated H+-translocation. J. Biol. Chem. 260:4807-4814.

20. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecularcloning: a laboratory manual, p. 15.68-15.71. Cold SpringHarbor Laboratory, Cold Spring Harbor, N.Y.

21. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequenc-ing with chain terminating inhibitors. Proc. Natl. Acad. Sci.USA 74:5463-5467.

22. Sebald, W., and J. Hoppe. 1981. On the structure and genetics ofthe proteolipid subunit of the ATP synthase complex. Curr.Top. Bioenerg. 12:1-64.

23. Senior, A. E. 1988. ATP synthesis by oxidative phosphoryla-tion. Physiol. Rev. 68:177-231.

24. Taylor, J., J. Ott, and F. Eckstein. 1985. The rapid generation ofoligonucleotide-directed mutations of high frequency usingphosphoorothioate-modified DNA. Nucleic Acids Res. 13:8764-8785.

25. Walker, J. E., M. Saraste, and N. J. Gay. 1984. The unc operon:nucleotide sequence, regulation and structure of ATP synthase.Biochim. Biophys. Acta 768:164-200.

26. Yanisch-Perron, C., J. Vieira, and J. Messing. 1985. ImprovedM13 phage cloning vectors and host strains: nucleotide se-quences of the M13mpl8 and pUC19 vectors. Gene 33:103-119.

27. Young, E. G., M. R. Hanson, and P. M. Dierks. 1986. Sequenceand transcription analysis of the petunia mitochondrial gene forthe ATP synthase proteolipid subunit. Nucleic Acids Res.14:7995-8005.

VOL. 173, 1991

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

07

Janu

ary

2022

by

192.

162.

179.

117.