ment of plasmid YIP5-B2 (a gift from Ira Herskowitz) that contains an LEU2-marked disruption of the HO gene. In strain SH124, the. PMA2 coding sequence (20) ...
Vol. 266. No. 36, Issue of December 25, PP. 24439-24445.1991 Printed in U.S.A.
THEJOURNALOF BIOLOGICAL CHEMISTRY d 1991 by The American Society for Biochemistry and Molecular Biology, Inc.
Evidence forCoupling Between Membrane and Cytoplasmic Domains of the Yeast Plasma Membrane H’-ATPase AN ANALYSIS OF INTRAGENICREVERTANTS
OF pmal-105* (Received for publication, July 19, 1991)
Sandra L. Harris$, David S. Perling, Donna Seto-Young§, and JamesE. HaberSlI From the SRosenstiel Basic Medical Sciences Research Center and Departmentof Biology, Brandeis University, Waltham, Massachusetts 02254 and the §Public Health Research Institute, New York, New York 10016
A genetic approach wasused to identify interacting the CaZ+-ATPase, Na+,K+-ATPase, and the H+,K+-ATPase from animal cells (10-12) have been cloned and sequenced. portions of the plasma membrane H+-ATPase from Saccharomyces cerevisiae. The cellular sensitivity of The predicted transmembrane topology of these enzymes is the pmal-105 strain (S368F) to low external pH and similar and there is a high degree of homology found in the to NH: was used to select intragenic revertants of two ATPase catalytic domain. The gene encoding the S. cerevisiae classes: phenotypically wild-type full revertants and H+-ATPase, P M A l , has been cloned and sequenced (13), and partial revertants that were low pH-resistant but re- site-directed mutagenesis has been used to identify some of tained resistance to hygromycin B. All 10 full rever- the functional domainsof the enzyme (14). tants had 5368 restored. Among five partial revertants We previously reported apositiveselectionschemefor mapping to the original site within the phosphorylation isolating PMAl mutants that is based on growth resistance domain, S368L and S368V were each found twice. One to theaminoglycoside antibiotic hygromycin B (Hyg B)’ (15). revertant contained an E367V substitution adjacent to the original S368F alteration. Four of 13 independ- Resistance to Hyg B results from a deficiency in the maintenance of the normalinside-negative membrane potential(16). ently isolated second-site revertants mapped to one B-resistant cellsgenerally site, V289F, in theproposed phosphatase domain.Mu- In addition, enzymesfromHyg show changes in the kinetics for ATP hydrolysis (17). Mutant tationswithinthe proposedphosphatase and phosstrainpmal-105 exhibits pronounced cellular and biochemical phorylationdomainsresultedin enzymes withincreased vanadate sensitivity relative to the vanadate- properties that make itespecially interesting. The pmal-105 insensitive S368F enzyme. These results suggest that strain is sensitive togrowth at low pH (pH 3.0) and to NH:. sites S368, E367, and V289 contribute to a vanadate Via a poorly understood mechanism, thegrowth of pmal-105 (Pi) binding domainor are able to interact with such a strainson media containing NH: canbe rescued by the site within the catalytic domain. The remaining nine addition of 0.1 M KC1 to the media (15).Thepmal-105 partial second-site revertants mapped tosixsites enzyme, which contains an amino acid substitution (S368F) within the putative transmembrane regions. Mutations 10 residues away from the D378 residue that becomes phoswithin the transmembrane region had less of an effect phorylated during catalysis, ishighly insensitive to vanadate onvanadatesensitivity. Most revertant enzymes and exhibits a K,,, and V,,, that are strongly pH-dependent showed small but significant increases in the rate of below pH 6.5 (17). Additionally, patch-clampanalysis reATP hydrolysis relative to S368F the enzyme. Several vealed that thepmal-105 mutant membranes exhibit an ATPenzymes no longer displayed the acid-sensitive pH- activated K+-channel conductance that responds in a physdependenceseen in the S368F enzyme.These data iological voltage range (60-120 mV) (18). This work raised provide novel evidence for an interaction between puthe possibility that the pmal-105 enzyme, directly or inditative transmembrane helices 1-3 and 7 and the ATP rectly, increases the influx of K’, thereby reducing the memhydrolytic portion of the enzyme. brane potential. Theelectrogenic properties of the pmal-105encoded H+-ATPase were also investigated by examining the activity of the enzyme in. vitro as a function of applied voltage; The plasma membrane H+-ATPase from Saccharomyces the pmal-105 enzyme is insensitive to voltage over a range cerevisiae is an electrogenic proton pump that is essential for that inhibits wild-type enzyme by nearly 50%. This strongly nutrient uptake and intracellular pH regulation (1, 2). The suggests that the electrogenicity of the H+-ATPase hasbeen ATPase belongs to the P-type family of ion-translocating modified (19). The substitution of Phe for Ser at position 368 has farATPases that form an acyl-phosphate intermediate during hydrolytic (as seen by catalysis (3-5) and are sensitive to inhibition by vanadate (6). reaching effects on both the ATP The genes encoding several members of the P-type family alterations of the K,,,, V,,,, and vanadate sensitivity) and including the fungal and plant H+-ATPases (7-9) as well as transport properties (as seen by the induction of K+ channel activity and the loss of electrosensitivity) of the enzyme. *This research was supported by National Institutes of Health These effects suggest a coupling between the charge transfer Grant GM39739 (to J. E. H.) and Grant GM38225 (to D. S. P.) and and ATPhydrolysis domains of the protein. In this study, the Office of Naval Research Grant N00014-89-5-1792 (to D. S. P.). S. distinct cellular and biochemical properties of the pmal-105 L. H. was supported as a Trainee of National Institutes of Health Genetics Training Grant 07122. The costs of publication of this article mutant were exploited in a detailedrevertantanalysisto were defrayed in part by the payment of page charges. This article must thereforebe hereby marked “aduertisement” in accordance with ’ The abbreviations used are: HygB, hygromycin B; kb, kilobase(s); 18 U.S.C. Section 1734 solely to indicate this fact. Com, complete defined medium; MES, 4-morpholineethanesulfonic 7 To whom correspondence shouldbe addressed. Tel: 617-736-2462. acid.
24439
24440
Intragenic Revertants of a Plasma MembraneH+-ATPase Mutant
identify interacting portions of the plasma membrane H+ATPase. Partial revertantsselected on the basisof resistance t o low external pH or NH: were found t o have mutations within the proposed phosphatase and phosphorylation domains as well as in transmembrane portionsof the enzyme. The biochemical properties of the partial revertants of pmal105 have been used to identify protein structural domains participating in ATPhydrolysis and coupling to electrogenic H+ transport.
polymerase chain reaction (25) using Amplitaqe DNApolymerase according to thespecifications of the enzyme supplier (Perkin-Elmer/ Cetus, Nonvalk, CT). DNA Analysis-Cloned plasmid DNA was sequenced by the dideoxy method (26) with modifications for sequencingdouble-stranded template with Sequenase, described as bythe enzyme supplier (United States Biochemical). The double-stranded DNA fragments produced by the polymerase chain reaction were purifiedfromagarose gels using Gene-Cleanaccording to theprotocol supplied by the manufacturer (Bio-101, Inc., La Jolla,CA). The amplified DNAwas sequenced with Sequenase using the dimethylsulphoxide method of Winship (27). MATERIALS ANDMETHODS Site-directedMutagenesis-Plasmid pSH29 was constructed by Yeast Strains andCell Culture-All yeast strainsused were isogenic inserting a 6.1-kb PMAl URA3-containing fragment from plasmid derivatives of Y55 (HOgaB MALISUCI). All revertants were derived p S H l l (17) into the HindIII site of pBR322. The PMAl gene in from two haploid strains, SH91 (ho::LEU2, urd-I, arg4-I, trp5-1, plasmid p S H l l contains a 1.1-kb URA3 fragment inserted into the pmal-105,leu2-IMAT a) andSH124 (ho::LEU2, u r d - I , arg4-1, BglII site in the 3'-noncoding region of PMAI. Plasmid pSH29 was trpl-5, lys5-2, pmal-105, pma2::TRPI MAT a). Strains SH91 and digested with Asp718 and religated resulting in the deletionof a 2.1SH124 were made haploid by integration of the 5.5-kb BamHI frag- kb portion of the PMAl coding sequence to create plasmid pSH30. ment of plasmid YIP5-B2 (a gift from Ira Herskowitz) that contains To create a plasmid containing a mutant copy of PMAI, the 2.1-kb a n LEU2-marked disruption of the HO gene. In strain SH124, the Asp718 fragment of PMAl was insertedinto M13mp19 and was PMA2 coding sequence (20) was deleted from the chromosome by mutagenized by the method of Kunkel (28). After mutagenesis, the transforming an isogenic yeast strain with the HindIII fragment of complete Asp718 fragment was sequenced to ensure that there were plasmid pPSPZT (kindly provided by P. Supply and A. Goffeau, no mutations in the fragment other than the desired mutation. The UniversitL de Louvain) which contains the TRPl gene embedded in Asp718 fragment containing the S368A mutation was subsequently the PMAP noncoding region. Yeast strains were transformed using subcloned back into pSH3O. Plasmid pSH30-S368A encodes an ATPthe lithium acetateprocedure of Ito et al. (21). ase protein that contains Ala at amino acid position 368 in place of Cells were grown in YEPD (1% w/v yeast extract, 2% w/v Bacto- Ser. The pmal-S368A mutant gene was introduced into yeast using Peptone, 2%w/v dextrose, pH 5.5). The ability of cells to grow under a modification of the single-step gene replacement technique (23) conditions of low external pH was tested in YEPDmedium that had (Fig. 1).Strain SH122 (pmaIA::LEU2/PMAI) was transformed with been adjusted to pH3.0 with 1 N HC1 (YEPD-pH 3.0). Sensitivity to the pSH30-S368A HindIIIfragmentcontainingthepmaI-S368A NH: was tested on standardcomplete defined medium (Com) [0.67% allele linked to URA3. Recombination occurred only within regions w/v yeast nitrogen base (without amino acids) containing 38 mM of homology on either chromosome. Transformants that were Leu' (NH4)2S04,2% w/v dextrose, plus all nutritional supplements]. Hyg Ura' may not have obtained the pmal mutation, as recombination B resistance was scored using YEPD containing 300 pg/ml hygrocould occur on the PMAI-containing chromosome between URA3 mycin B (YEPD-Hyg). For biochemical analysis, cells were grown in and the mutation. On the other hand, all Ura' Leu- transformants 1 liter batches of YEPD a t 22 "C until mid-log phase and harvested must have obtained the mutation. The Ura+ Leu- transformants were bycentrifugationas previouslydescribed (17). All culture media dissected and the phenotypesof the Ura+spore colonies were tested. supplies were from Difco. Generation of Secondary Mutant Strains-To create strains that Reuertant Selection-All revertants were derived from pmal-105- contained only the secondary reversion mutation in the absence of containing strains SH91 and SH124 which cannot grow on YEPD the S368F mutation, yeast strains with pmal alleles containing both single mutations were transformed with a 4.2-kb PuuII-Hind111 fragment (pH 3.0) oron Com plates. T o obtainrevertants,isolated colonies from strains SH91 and SH124 were patched to YEPD and of plasmid p S H l l (17) that includeda URA3 marker in the 3'grown overnight a t 30 "C. The patches on YEPD were replicaplated nontranslated portion of PMAI. When this fragmentwas integrated to YEPD (pH 3.0) and Com plates. After 2 days, revertants arose into the chromosome by homologous recombination, the mutation at spontaneously as papillae growing on a background of dead cells. nucleotide position 1691 (S368F)was replaced by wild-type sequence. Papillae (no more than one papillus from each isolated patch) were Because the transforming fragments did not include the regions of then streaked for single colonies on the same typeof selective plate. DNA encoding the N-terminal mutations, the secondary mutations T h e set of independent revertants was then patched onto YEPD to make a master platewhich was replicaplated to YEPD (pH 3.0), Com, were not altered. Membrane Isolation and ATPase Analysis-Plasma membraneand YEPD-Hyg. Mapping and Cloning of pmal-I05 Reuertant Alleles-The pmal- enriched fractions were prepared in duplicate from revertant strains grown in 1-liter batchesas previously described (17).ATP hydrolysis 105 mutation creates an EcoRI restriction endonuclease recognition sitewithin nucleotides 1687-1692 (17).(The nucleotide numbers measurements as a function of pH and vanadate concentrationwere correspond toRef. 13.) To determine if the reversion occurred at the also performed aspreviously described (17) except that ATPhydrolsite of the original mutation, DNA samples from the revertant strainsysis was performed in a 1-ml volume containing 5 mM ATP, 5 mM 5.5-7.5 with were restricted with EcoRI, separated by gel electrophoresis, and MgSO,, 10 mM NH4C1, and 50 mM MES adjusted to pH blotted onto nylon filters (22). The filters were hybridized to a 32P- Tris base. labeled DNA fragment that contained the entire PMAl coding seRESULTS quence. This analysis divided the revertants into two groups, those that had a reversion within nucleotides 1687-1692 and those that had Two Phenotypic Classes of Revertants-The pmal-105 rea second mutation elsewhere in the genome. To determine if the full. secondary mutations that did not alter the EcoRI site were within vertants from strain SH91were of two types, partial and PMAI, the revertant strainswere crossed to a Y55 PMAI wild-type Of 197 revertants obtained, 17 were partial revertants. Phestrain, and the resultant diploids were dissected. The segregation of notypically wild-typerevertants thatgrew on Com and YEPD mutant phenotypes revealed that in allcases the reverting mutation (pH 3.0) but were sensitive to Hyg B were termed full reverwas tightly linked to PMAI. To further map the position of the tants. Cells that grew at low pH or in the presence ofNH: secondary mutations within PMAI, markerrescue [a modification of but retained resistance to Hyg B were termed partial reverthe single-step gene replacement scheme (23)]was used. This method was previously used to map mutation sites within PMAl(17). How- tants. All revertants that were selected as resistant to low ever, to map the revertant mutations, strains were transformed with external pH were also resistant to NH: and vice versa. DNA DNA fragments containing portions of the mutant pmal-105 gene isolated from a total of 25 full revertants was analyzed by rather than thewild-type gene, and transformants were screened for Southern analysis to map the position of the reversions. The the reappearanceof the pmal-105 mutant phenotype. The pmal-105pmal-105 mutation fortuitously createsa new EcoRI restricrevertant alleles were cloned bygap repair(24) as describedpreviously for the cloning of pmal-105 (17). Forseveral strains with mutations tion endonuclease recognition site. Every full revertant anafull mapping within nucleotides 1687-1692, the 1.35-kb regions (nucleo- lyzed had lost the EcoRI site, thus indicating that all the tides 950-2301) spanningthemutantsite were amplified by the revertants had a mutation at or adjacent to the site of the
24441
Intragenic Revertants of a P l a s m a M e m b r a n e H + - A T P a s e M u t a n t URA3
Hjndlll fragment
pmal A ::LEU2
SH122 Leu+ Ura-
chromosome 7 chromosome 7 PMA 1
FIG. 1. Introduction of mutant pmal sequences. PMAl was replaced with mutant pmal genes using a modification of the one-step gene replacement techniaue (23). Because recombination must take place within regions of homology, all Ura’ Leu- transformants should contain the desired mutation. ~~
~
URA3
~~~I
pmsl A::LEUZ
pmal-mulsnl
llIIL1llll
URA3
Leu+ Ura+
Leu- Ura+
I
Sporulate Dissect tetrads
URA3
original mutation. DNAsequence analysis was used to determine the natureof the reversions in 10 full revertant strains. S368 was restored in every full revertant sequenced. In three revertants, the restorationof Ser at position368 occurred by base substitution. However, in the remainingseven full revertants, replacement of the ser at 368 resulted from a gene conversion event between PMAl and the 85% homologous PMA2 gene (20). These ectopic gene conversion events will be discussed elsewhere.’ T o avoid such geneconversion events, the coding sequence of PMA2 was removed and replaced with TRPl in strainSH124. Of 57 revertants isolated from strain SH124, all were partial revertants. Mapping of Partial Revertants-Among a total of 33 partial revertants of strains SH91 and SH124 that were analyzed, nine had lost the EcoRI site. DNA from five strains with mutations mapping to the EcoRI site was amplified by the polymerase chain reaction and sequenced. Two of the genes encoded proteins with a substitution of Val at position 368, and two alleles contained Leu at position 368. In one revertant, the S368F mutation was retained and a second substitution, E367V, was present. The positions of the secondary mutations in 13 of the remaining partial revertants were mapped by genetic crosses and marker rescue (see “Materials and Methods”). The mutant genes were cloned by gap repair and sequenced. The positions of the secondary mutations are shownin Fig. 2. Eight of the 13 second-site mutationsoccurred at two sites, (2148s and V289F. Each mutationwas independently isolated four times. In three cases, a G to C transition at nucleotide 1031 resulted in a Ser codon a t 1 4 8 a fourth C148S substitution was the result of a T to A transition at nucleotide position 1030. Because thesameamino acid substitution resulted from two distinct nucleotide changes, it is unlikely that the high number of reversions atthissite occurred because thesite is a mutationalhotspot.Onerevertant contained two substitutions, I133F and A135S. The I133F, A135S mutations aswell as theI147M, C148S, V289F, A828S, and S836C residues lie within the membrane-spanning domains based on the predicted topology of the protein (13). T h e L298F substitution lies within theproposed phosphatase domain ( 14).
‘S. L. Harris, K. Stern,
andJ. E. Haber, manuscript in preparation.
Generation and Phenotypic Characterization of Strains Containing the Secondary Mutation Only-To further characterize the effect of the secondary mutations on the function of the enzyme, five pmal alleles were created that contained only the secondary mutations in the absence of the S368F substitution. This was accomplished by replacing the 3’ portion (containing the S368F mutation) of the pmal revertant allele with wild-type sequence (see“Materials and Methods”). Strains with each mutation (I133F, A135S; 1147M; C148S; V289F; and L298F) were able to grow a t low pH and in the presence ofNH:, and allwere resistant toHyg B. In order to betterunderstandthe role of amino acid 368, an S368A mutation was generated by site-directedmutagenesis (see “Materials and Methods”). The S368A strain is phenotypically wild type. The phenotypes of all the mutants in this analysis are shown in Table I. A T P Hydrolysis and pH Dependency-Table I summarizes the catalyticbehavior of mutant enzymes under conditionsof ATP hydrolysis that are optimal for the wild-type enzyme (pH 6.5). All revertant enzymes, except for one second-site mutant (S368F, L298F), were found to increase the rate of ATP hydrolysis at pH 6.5 relative to the S368F (pmal-105) enzyme. A characteristic property of the S368F mutant enzyme is a precipitous decline in the rate of ATP hydrolysis below pH 6.5 which is believed to contribute to the acidsensitive growth phenotype of the pmal-105 strain (17). It was of interest to determine whether the enzymes from revertant strains isolated on the basis of resistance to low external pH or NH: were altered in their pH-dependent rates of ATP hydrolysis. Therefore, ATP hydrolysis was examined over the pH range of 5.5 to 7.5. Fig. 3A indicates that wildtype enzyme(S368) and an S368A substitution showed a symmetric ATP hydrolysis profile over this pH range with a broad p H optimum. Two other substitutionsat thisposition, S368L and S368V, resembled wild-type enzyme on the acid side but were less active above pH 6.5. In contrast, theS368F enzyme showedan asymmetricprofile with a clearly displaced and more distinct pH optimum (at pH 7.0) relative to wildtype. The pH-dependent behavior of second-site revertant enzymes containing the S368F mutation is most striking since several types of pH profiles were observed (Fig. 3 B ) . An
24442
Intragenic Revertants of a Plasma Membrane H+-ATPase Mutant
pmal-105 partial revertants
CI48s A135s I133F
I
1
~828s
L298F
I
FIG. 2. Schematic view of H+ATPase structural topography. A
schematic view of the structural topography of the H+-ATPase as determined from hydropathy analysis (13). Previously assigned functional domains (14) are indicated as well as the position of the originalpmal-l05mutation (S368F). The positions of the amino acid substitutions found in the partialrevertant enzymes are shown.
TABLEI Effect of pmal mutations onATP hydrolysis ATP hydrolysis"
Growth phenotypes Mutations Hyg B (300 I I M )
S368 (WT) S368F S368A S368V S368L S368F E367V S368F V289F S368F I133F, A135S S368F I147M S368F C148S S368F L298F S368F A828S S368F S836C V289F I133F, A135S I147M C148S L298F a
-
++ + + + + + + + + + + + + + +
YEPD (pH 3)
++ + + + + + + + + + + + + + + +
Corn
++ + + + + + + + + + + + + + + +
Wild-type activity (PH 6.5) % 100 f 1.9
64.5 f 0.9 99.1 f 1.0 115 f 4.7 79.5 f 6.7 77.1 f 8.2 76.0 f 0.8 76.2 f 22.4 ( n = 4) 83.3 & 2.2 87.4 f 15.6 ( n = 3) 66.6 f 4.4 76.9 f 8.3 73.2 f 0.6 103 f 3.6 114 f 1.0 97.6 f 1.1 104 f 1.0 92.7 f 0.7
Vanadate inhibition,
K, IIM
0.7 400 0.9 8.5 160 13 17 35 35 70 900 320 195 8 0.4 2 0.4 1
Determined on two independent sets of membranes isolated from cells grown under identical conditions.
asymmetric pH profile with a clearly displaced pH optimum at pH 6.0was observed for the enzymes with second-site changes lying within the catalytic (E367V) and phosphatase (V289F) domains. The C148S enzyme appeared essentially wild-type. Several enzymes with second-site mutations lying within the membrane sector [I133F, A135S;I147M (not shown); A828S (not shown); and S836C (not shown)] appeared to be less acid-sensitive than the pmal-105 mutant enzyme but were somewhat inhibited above pH 6.5. The S368F, L298F mutant enzyme was not acid stabilized. The properties of enzymes with second-site changes in the absence of S368F were also examined and are presented in Fig. 3C. Enzymes with only the second-site mutations produced pH profiles that were either symmetric and similar to wild-type (I147M; L298F) or asymmetric with an acid-dis-
placed pH optimum (pH 6.0) (1133F, A135S; C148S; V289F). Vanadate Sensitivity of Revertant Enzymes-The S368F mutation renders the ATPase highly resistant to inhibition by vanadate (17). The sensitivity of revertant enzymes to vanadate is shown in Fig. 4 and theki for vanadate inhibition for each enzyme is shown in Table I. Alterations at amino acid 368 caused variations in vanadate sensitivity spanning three orders of magnitude (Fig. 4A). The presence of Phe or Leu at position 368 resulted in vanadate-insensitive behavior, whereas the effect ofValwas less pronounced. Vanadate insensitivity appeared to correlate with the size of the sidegroup. The substitution of a small Ala side-group at position 368 resulted in an enzyme with wild-type sensitivity to vanadate. Theseresults suggest that steric effects, involving
Intragenic Revertants of a Plasma Membrane H+-ATPase Mutant
4
I
#
4
-
-t
5368F.1133F.Al355 5368F.Cl485
-0-
S368F. V289F 5368FL298F 5368F.E367V
b -
24443
effects, whereas revertant enzymes with second-site mutations expected to lie within the membrane sector (1133F, A135S; I147M; C148S; L298F; A828S; S836C) had a significant but smaller influence on vanadate insensitivity. Separated second-site mutations lying within the first two putative transmembrane segments had a small effect onvanadate sensitivity (Fig. 4C). However, a separated second-site mutation within the phosphatase domain (V289F) showed marked vanadate insensitivity. These results suggest that sites S368, E367, and V289 contribute to a vanadate (and presumably Pi) binding domain or are able to interact with such a site within the catalytic domain. DISCUSSION
-0-t
5.0
55
6.0
6.5
70
7.5
V289F WBF
80
pH
FIG.3. Effect of pH on revertant enzyme kinetics. ATP hydrolysis was determined over the pH range 5.5-7.5 on membrane bound enzymes, as described under "Materials and Methods." Enzymes with reversions at the site of the pmal-105 mutation (residue 368) are shown in A. Enzymes which retain the original mutation and also have a second-site mutation are shown in B. Enzymes that contain only the second-site mutation are shown in C.
-+-
5368FC1485 S368kE367V
-c
5368F.VZ89F
-Q-
S368F:1133Ffil35S
I S368F;L298F
-
" t
" t
". 01
.
.
.
1
1
10
100
S368F1147M 5368FA8285 5368F:Y136C
1147M M 8 F C1485 1133F;A135S V289F
lob0
Vanadate (pM)
FIG.4. Inhibition of ATP hydrolysis by vanadate. The effect of vanadate on ATP hydrolysis was analyzed by adding vanadate to the ATP hydrolysis reaction medium. ATP hydrolysis in the absence of inhibitor was taken as the control value. Enzymes with reversions at position 368 are shown in A. Enzymes which retain the original mutation and also have a second-site mutation are shown in B. Enzymes that contain only the second-site mutation are shown in C. residue 368, play a role in the vanadate sensitivity of the enzyme. Second-site mutations also influenced the vanadate sensitivity of the S368F mutant enzyme but the variations in sensitivity were less extreme (Fig. 4B). Ingeneral, mutations expected to lie within the catalytic domain (E367V) and putative phosphatasedomain (V289F) resulted in the greatest
The pmal-105 mutant was chosen for revertant analysis because of its unique biochemical and transport properties. The highly vanadate-insensitive enzyme exhibits a sharp decline in the rate of ATP hydrolysis below pH 6.5, and pmal105 cells have a depolarized membranepotential (16, 17). Previous studies suggest that thepmal-IO5 enzyme may have a defect in electrogenic proton transport that includes the capacity to leak K' into the cell while H' is being actively exported (18).To understandhow a mutation near the active site of the enzyme could result in such diverse functional alterations, a genetic approach has been used to define the role of amino acid 368 and to locate interacting domains within the yeast plasma membrane H'-ATPase. The isolated revertants furtherdefine the phosphate binding site and suggest a role for the first hydrophilic cytoplasmic loop domain and transmembranehelices 1-3 and 7 inthe coupling of ATP hydrolysis to electrogenic H' transport. This powerful approach to studyingstructure-functionrelationshipsin the ATPase did not require assumptions to be made about the essentiality of the specific residues. As a result, it revealed interactions between domains of the protein that were not expected. The fact that every full revertant we obtained had Ser restored at residue 368 argues that there is a very limited subset of amino acid substitutions that can provide wild-type function. This is not surprising given the conservation of this residue as well as the immediate region surrounding the site of phosphorylation in four different fungal H'-ATPases (29). The substitutionof other amino acids at position 368 resulted in a partial revertantphenotype. The bulkiness of the amino acid side chain a t position 368 appeared to correlate with vanadate sensitivity. This suggests that steric effects in this region are important in defining the vanadate sensitivity of the enzyme. It was expected that compensating secondary-site mutations would be found within the proposed phosphatase and phosphorylation domains. This expectation arose because the primary site mutation (S368F) lies 10 residues away from the site of phosphorylation (D378), in the large central cytoplasmic domain which has been proposed to interact with a "phosphatase" domain contributed by the first hydrophilic loop region (14). Vanadate insensitivity, which is believed to result from modification of the phosphate binding site (3032), has also resulted from mutations within the highly conserved region of the phosphatase domain (G268D) in the H+ATPase of S. cereuisiae (33). Surprisingly, only one pmal-105 revertant (V289F) which was independently isolated four times, fits the expected paradigm. Vanadate-insensitive mutants mapping within and near the proposed ATP binding domain [residues 534-666 (29)] have also been isolated in S. cereuisiae (17) and Schizosaccharomycespombe (34), yet none of the pmal-105 revertants were found in the ATP binding
24444
Intragenic Revertants of a Plasma Membrane H+-ATPase Mutant
domain. The identification of second-site revertants of S368F parently inseparable properties, as selection for resistance to that map to within the phosphatase domain supports the one invariably selected the other. However, reversion of these notion that thefirst two hydrophilic loops interact duringthe phenotypes often left the cells still resistant to Hyg B, suggesting that a wild-type membrane potential had not been hydrolysis of the aspartyl-phosphate intermediate. The fact that several different partial revertant mutations restored. Indeed, none of the second-site revertants were full were each isolated more than one time indicates a limitation revertants, suggesting that the S368F mutation induces a to the type of second-site mutations that will enable thepmal- fundamental change in the structure or function of the en105 strain to grow in the presence oflow pH or NH:. To zyme that cannot be compensated for by a secondary change understand the role of the mutations within the transmem- in the protein. Models can be proposed to account for these brane domains, we analyzed the kinetic behavior of the rev- observations by invoking changes in the H'-ATPase that ertant enzymes and studied the effect of the secondary mu- affect either the kinetics of ATP hydrolysis or ion translocatations in isolation. In some cases, the secondary mutation tion. Both models require an assumption of tight conformaconferred on the enzyme an increased ability to function at tional coupling between the membrane-embedded transport low pH. We do not know ifintracellular domains of thepmal- domain and ATP hydrolytic domain. In the kinetic model, we assume that the inability of the 105 enzyme are sensitive to low internal pH or if the pH sensitivity is due to an effect of low extracellular pH. Never- pmal-105 strain to grow at low pH or on Com is related to theless, a trend exists in that most of the revertant enzymes the sensitivity of the catalytic site to low internal pH which were less inhibited bylow pH than the pmal-105 enzyme. results in a pronounced decrease in V,,,. The ability of 0.1 M Because several isolated second-site mutants showed either KC1 to suppress the no-growth phenotype of p m l - 1 0 5 cells normal pH dependence of ATP hydrolysis (I147M, L298F) or on Com can be explained if the effect ofK' is to collapse residual membrane potential. If the membrane potential is showedmoreacidic pH optimums (1133F,A135S;C148S; V289F) but were still resistant to Hyg B, it is clear that the collapsed, there will be a reduced driving force for the uptake Hyg B-resistance phenomenon is not related to the decrease of NH: or reabsorption of H', both of which would have the in Vmaxof the enzyme below pH 6.5. When ATP hydrolysis capacity to lower internal pH. The enhanced rates of ATP was measured at pH 6.5 (the optimal pH for the wild-type hydrolysis at low pH in many revertant enzymes compared to enzyme) most revertant enzymes showed slight increases in the pmal-105enzyme support this idea. In the transport model, the pmal-105 enzyme is assumed the rate of ATP hydrolysis that may in part account for the ability of the revertant strainsto grow on Com or YEPD (pH to exist more often in a conformation that facilitates a K+ 3.0). However,oneenzyme (S368F, L298F) showed only a leak into the cell. This K' leak results in depolarization of marginal increase in the rate of ATP hydrolysis. The small the cellular membrane potential and cellular resistance to but distinct effects of second-site mutations within the mem- Hyg B. The K' conductance pathway is assumed to result brane sector on vanadate-insensitive behavior make it likely from either the same translocation pathway as protons or a that the first hydrophilic loop region is highly conformation- parallel pathway. The inhibitory effect of NH: on the growth of pmal-105 cells is explained by the capacity of NH: to act ally active and participates in the coupling process. and K' move in opposite Role of Conserved Amino Acids-A high degree of conser- asa K' analogue (36). IfH' vation at a particularamino acid position generally correlates directions within the same translocation pathway (or in a with the functional importance of the residue. The Glu at linked parallel pathway) then NH: may bind and block at a position 367 is conserved in all fungal H'-ATPases and every K+ site which could in turn block proton transport. Thiseffect known member of the P-typeATPase family (29). The E367V, would deleteriously alter both ATP hydrolysis and proton enzyme since these functions are S368F revertant however is viable. Therefore, at least in the transport in the mutant context of the S368F mutation, the Glu residue at position tightly coupled.High K' concentrations couldrescue the 367does not appear to be essential. The Cys residue at mutants from NH: inhibition by competing for the K+ transposition 148 is found in the plasma membrane H+-ATPases port binding site. Whether a transport or catalytic defect is invoked to explain from S. pombe and Neurospora crassa as well as S. cerevisiae. The C148S mutation, both in conjunction with S368F and by the sensitivity of the pmal-105 mutantcells to low external itself, is not particularly deleterious. Cells with the mutation pH and NH:, it is apparent that mutations in the transmemare not only viablebut areable to grow at only slightly reduced brane regions of the proteinare having an effect on the rates relative to ~ i l d - t y p eRao . ~ and Scarborough (35) recently cytoplasmic domain of the ATPase andvice versa.The analydetected a single disulfide linkage in the Neurospora plasma sis of revertants of pmal-105 has implicated transmembrane membrane ATPase between C148 and either (2840 or (2869. segments 1-3 and 7 as playing a central role in electrogenic In the S. cerevisiae enzyme,the cysteines at position 148 and transport aswell as in coupling to theATP hydrolytic domain. 869 are conserved. This suggests that transmembrane segAcknowledgments-We thank P. Supply and A. Goffeau for the ments 2 and 8 are linked by a disulfide bridge in the S. gift of plasmid pPSPZT. We thank Songqing Na, Shalini Anand, cerevisiae enzyme. Rao and Scarborough (35) suggested that Brian Monk, and Dan Oprianfor helpful discussion and critical this disulfide bridge might play an important role in main- reading of this manuscript. taining the structure of the membrane-embedded portion of REFERENCES the enzyme. If C148 is also involved in a disulfide bridge in the S. cerevisiae enzyme,then theresults of our study suggest 1. Serrano, R. (1984) Curr. Top. Cell. Regul. 23,87-126 that the disulfide bond does not play an essential role. 2. Goffeau, A.. and Slavman. Acta C. W. (1981) Biochim. Bi0ph.y~. . . ~. 639, i97-223 Implications for the Mechanism of the Plasma Membrane 3. Amorv. A.. and Goffeau., A. (1982) . . J.Biol. Chem. 257,4723-4730 H+-ATPase-The distribution of second-site revertants sug4. Malpahida, F., and Serrano, R. (1981) Eur. J. Biochem. 116, gests several possible explanations for the phenotypes of the 413-417 pmal-105 mutant and its revertants. Any explanation must 5. Dame, J. B., and Scarborough, G.A. (1980) Biochemistry 19, account for several observations, including the following. 2931-2937 First, the low pH and NH: sensitivity of pmal-105 are ap6. Bowman. B. J.. and Slavman, . C. W. (1979) J. Biol. Chem. 254, 2928-2934 7. Hager, K. M., Mandala, S. M., Davenport, J. W., Speicher, D. '
S. L. Harris and J. E. Haber, unpublished results.
Intragenic Revertants of a Plasma Membrane H+-ATPase Mutant W., Benz, E. J., and Slayman, C.W. (1986) Proc. Natl. Acad. Sci. U.S. A . 83, 7693-7697 8. Harper, J. F., Surowy, T. K., and Sussman, M. R. (1989) Proc. Natl. Acad. Sci. U.S. A. 86, 1234-1238 9. Pardo, J. M., and Serrano, R. (1989) J. Bwl. Chem. 2 6 4 , 85578562 10. MacLennan, D.H., Brandl, C. J., Korczak, B., and Green, N. M. (1985) Nature 316,696-700 11. Shull, G. E., Schwartz, A., and Lingrel, J. B. (1985) Nature 316, 691-695 12. Shull, G. E., and Lingrel, J. B. (1986) J. Biol. Chem. 261, 1678816791 13. Serrano, R., Kielland-Brandt, M.C., and Fink, G.R. (1986) Nature 319,689-693 14. Portillo, F., and Serrano, R. (1988) EMBO J. 7, 1793-1798 15. McCusker, J. H., Perlin, D. S., and Haber, J. E. (1987)Mol. Cell. Bid. 7,4082-4088 16. Perlin, D. S., Brown, C. L., and Haber, J. E. (1988)J. Biol. Chem. 263,18118-18122 17. Perlin, D. S., Harris, S. L., Seto-Young, D., and Haber, J. E. (1989) J. Biol. Chem. 264, 21857-21864 18. Ramirez, J. A., Vacata, V., McCusker, J. H., Haber, J. E., Mortimer, R.K., Owen, W. G., and Lecar, H. (1989) Proc. Natl. Acad. Sci. U.S. A. 86. 7866-7870 19. %to-Young, D., and Perlin, D. -S. (1991) J. Biol.Chem. 266, 1383-1389 20. Schlesser, A., Ulaszewski, S., Ghislain, M., and Goffeau, A. (1988) J. Biol. Chem. 263,19480-19487
24445
21. Ito, H., Fukuda, Y., Murata, K., and Kimura, A. (1983) J. Bacteriol. 163,163-168 22. Sambrook, J., Fritsch, E. F., and Maniatis, T.(1989) Molecular Cloning: A Laboratory Manual, 2nd Ed., Cold Spring Harbor Laboratory, Cold Spring Harbor, NY 23. Rothstein, R. (1983) Methods Enzymol. 101,202-211 24. Orr-Weaver, T.L., Szostak, J. W., and Rothstein, R. J. (1981) Proc. Natl. Acad. Sci. U. S. A. 78,6354-6358 25. Saike, R. K., Gelfand, D. H., Stoffel, S., Scarf, S. J., Higuchi, R., Horn, G. T., Mullis, K.B., and Erlich, H. A. (1988) Science 239,487-491 26. Sanger, F., Niklen, S., and Coulson, A. R. (1977)Proc. Natl. Acad. Sci. U.S. A. 74,5463-5467 27. Winship, P. R. (1989) Nucleic Acids Res. 17, 1266 28. Kunkel, T. A (1985) Proc. Natl. Acad. Sci. U.S. A. 82,488-492 29. Serrano, R., and Portillo, F. (1990)Biochim. Biophys. Acta1 0 1 8 , 195-199 30. Cantley, L. C., Cantley, L. G., and Josephson, L. (1978) J. Biol. Chem. 253,7361-7368 31. Pick, U. (1982) J. Bwl. Chem. 257,6111-6119 32. Huang, W.-H., and Askari, A. (1984) J. BioZ. Chem. 269,13287132291 33. Ghislain, M., Schlesser, A., and Goffeau, A. (1987) J. Bwl. Chern. 262,17549-17555 34.Van Dyck, L., Petretski, J. H., Wolosker, H., Rodregues, G., Schlesser, A., Ghislain, M., and Goffeau, A. (1991) Eur. J. Biochem. 194, 785-790 35. Rao, U. S., and Scarborough, G. A. (1990) J. Biol. Chem. 265, 7227-7235 36. Yool, A. J., and Schwartz, T.L. (1991) Nature 349, 700-704
