THE JOURNAL OF BIOLOGICAL CHEMISTRY d 1991 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 266, No. 19, Issue of July 5, pp. 12168-12172,1991 Printed in U.S.A.

Yeast PPAZ Gene Encodes a Mitochondrial Inorganic Pyrophosphatase That Is Essential for Mitochondrial Function*

(Received for publication, November 29, 1990)

Maria LundinSB, Herrick BaltscheffskyS, and Hans Ronnen From the $Department of Biochemistry, Arrhenius Laboratories for Natural Sciences, University of Stockholm, S-106 91 Stockholm. Sweden and the lLudwip Institute for Cancer Research, Uppsala Branch, Biomedical Center, Box 595, S-751 23 Uppsala, Sweden

”

We have cloned a gene encoding a mitochondrial inorganic pyrophosphatase (PPase) in the yeast Sac- charomyces cerevisiae by low stringency hybridiza- tion to P P A l , the yeast gene for cytoplasmic PPase. The new gene, PPAZ, is located on chromosome 13 and encodes a protein whose sequence is 49% identical to the cytoplasmic enzyme. The protein differs from cy- toplasmic PPase in that it has a leader sequence en- riched in basic and hydroxylated residues, which is typically found in mitochondrial proteins. Yeast cells overproducing PPA2 had a 47-fold increase in mito- chondrial PPase activity. This activity was further stimulated %fold by the uncoupler carbonyl cyanidep- trifluoromethoxyphenylhydrazone, which suggests that PPAQ is part of an energy-linked enzyme. Using gene disruptions, we found that P P A l is required for cell growth. In contrast, cells disrupted for PPAZ are viable, but unable to grow on respiratory carbon sources. Fluorescence microscopy revealed that these cells have lost their mitochondrial DNA. We conclude that the mitochondrial PPase encoded by PPAZ is es- sential for mitochondrial function and maintenance of the mitochondrial genome.

Inorganic pyrophosphatase (PPase,’ EC 3.6.1.1) catalyzes the hydrolysis of inorganic pyrophosphate, PPi, which is formed as a by-product in a number of metabolic reactions. Hydrolysis of PPi is necessary to maintain the forward direc- tion of these reactions. This essential function is performed by a soluble cytoplasmic PPase which is present in all cells (Cooperman, 1982; Lahti, 1983). The gene for soluble PPase has been cloned in Escherichia coli (Lahti et al., 1988), in the two budding yeasts Saccharomyces cerevisiae and Kluyvero- myces lactis (Kolakowski et al., 1988; Stark and Milner, 1989), and in the fission yeast Schizosaccharomyces pombe (Kawa- saki et al., 1990). The amino acid sequence of the protein has also been determined directly in Saccharomyces and in the termophilic bacterium PS-3 (Cohen et al., 1978; Ichiba et al., 1990), and the crystal structure of the Saccharomyces enzyme

* This work was supported by grants from the Nordic Yeast Re- search Program (to M. L. and H. B.) and the Swedish Natural Science Research Council (to H. B.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed. Present address: Ludwig Institute for Cancer Research, Uppsala Branch, Biomedical Center, Box 595, S-751 23 Uppsala, Sweden. Tel.: 46-18-174239; Fax:

’ The abbreviations used are: PPase, inorganic pyrophosphatase; DAPI, 4’,6-diamidino-2-phenylindole; FCCP, carbonyl cyanide p - trifluoromethoxyphenylhydrazone; SDS, sodium dodecyl sulfate.

46-18-506867,

has been published (Terzyan et al., 1984; Kuranova, 1987). However, both prokaryotes and eukaryotes also possess

energy-linked membrane-bound PPases. Such a PPase was first found in the chromatophore membrane of Rhodospirillum rubrum (Baltscheffsky, 1967), where it catalyzes the light- induced formation of PPI (Baltscheffsky et al., 1966). This PPase is an intrinsic membrane protein that functions as a coupling factor (Baltscheffsky and Nyrkn, 1984). The PPase and proton-pumping activities are present in a single poly- peptide (Nyrkn et al., 1991). It has been suggested that this PPase could be a relic from an early stage in evolution, when PP, was used instead of ATP to store light energy (Baltscheff- sky, 1981; Baltscheffsky et al., 1986). A proton-pumping PPase has also been found in the vacuoles of several eukar- yotes. Like the Rhodospirillum enzyme, this vacuolar PPase is an intrinsic membrane protein with both PPase and proton- pumping activities residing in the same polypeptide (Mas- lowski and Maslowska, 1987; Maeshima and Yoshida, 1989; Britten et al., 1989; Sarafian and Poole, 1989). The amino- terminal sequence of mung bean vacuolar PPase has been determined, and it shows no similarity to the soluble PPases (Maeshima and Yoshida, 1989).

A third membrane-bound PPase is found in mitochondria (reviewed by Mansurova, 1989). This PPase has a more com- plex structure than the bacterial and vacuolar PPases. The PP, hydrolyzing activity resides in a soluble 28-30-kDa cata- lytic subunit which binds noncovalently to a protein complex in the mitochondrial membrane (Volk et al., 1983; Volk and Baykov, 1984). In this respect, the PPase resembles the mi- tochondrial ATPase, in which the catalytic F, complex binds to the proton-pumping F, complex (Walker et al., 1985). In addition to functions in PPi hydrolysis and energy coupling, it has also been suggested that mitochondrial PPase could be involved in the calcium-mediated response to certain hor- mones (Halestrap, 1989). We describe below a new Saccha- romyces gene, PPAB, which encodes the catalytic subunit of an uncoupler-stimulated mitochondrial PPase. The PPAZ protein is essential for mitochondrial function, and has exten- sive sequence similarity to the soluble cytoplasmic PPases.

EXPERIMENTAL PROCEDURES

atives of W303-1A (Thomas and Rothstein, 1989). To disrupt the Yeast Strains and Plasmids-All yeast strains were isogenic deriv-

PPAl gene, a HpaI-Sal1 fragment carrying the LEU2 gene was inserted into the NcoI site. To disrupt PPAZ, the URA3 Hind111 fragment was cloned between the two XbaI sites (Fig. 1). Both disruptions were introduced into W303-1A or its isogenic diploid D67 using the one-step method of Rothstein (1983). The disruptions were verified by Southern analysis. Plasmid pHR81 and the yeast genomic library made in this vector have been described (Nehlin et al., 1989). Plasmid pMLG1 is pHR81 with a 2.4-kilobase ClaI fragment carrying the PPAZ gene (Fig. 1) cloned into the BamHI site.

Screening for PPase Actiuity-Our screening method was a modi-

12168

Yeast PPAB Gene Encodes a Mitochondrial Pyrophosphatase 12169

fication of the PPase plate assay developed for E. coli (Kukko et al., 1982). The genomic yeast library was first transformed into yeast strain W303-1A. Approximately 20,000 transformants were selected on uracil-less medium and then replicated to PPase assay plates. These plates are YPGal media containing 5 mM NaPPi, 1% Triton X-100, and 20 mM MgCI,. NaPPi and Triton X-100 were filter sterilized separately and then added to the autoclaved medium. An opalescent MgPP, precipitate is formed in the plates, but clear zones develop around yeast colonies that overproduce PPase.

Low Stringency Hybridization-The probe used to clone PPAS was a 675-base pair EcoRV fragment of PPAl (Fig. 1). The filters were first pretreated for 2 h at 65 "C, in 5 X Denhardt's solution, 0.5% SDS, to remove bacterial debris. Hybridization was carried out a t 42 "C in 5 X SSPE (180 mM NaC1, 1 mM EDTA, 10 mM NaH2P0, (pH 7.4)), 20% formamide, 5 X Denhardt's solution, 10 mM NaH,PO,, 0.1% SDS, and 100 pg/ml of single-stranded salmon sperm DNA. The filters were prehybridized overnight, and then hybridized over- night to the :"P-laheled probe. Washing was carried out for 2 X 30 min a t 42 "C in 2 X SSPE, 0.1% SDS.

Preparation of Mitochondria-Mitochondria were prepared as pre- viously described (Lundin et al., 1987), but with 4 mM MgC12 and 2 mM dithiothreitol in the mannitol buffer. Phenylmethylsulfonyl flu- oride (1 mM) and bovine serum albumin (0.1 mg/ml) were included during the osmotic shock and in the first of the two washes. Yeast cells containing pHR81 or pML61 were grown to early stationary phase in leucine-less media, to select for a high plasmid copy number (Nehlin and Ronne, 1990). PP, hydrolyzing activity was determined according to Shatton et al. (1982), in 20 mM Tris-HC1 buffer, pH 7.2, containing 0.6 M mannitol, 2 mM dithiothreitol, 4 mM MgCl,, and 2 mM PP,.

Other Methods-The methods used for yeast genetics and molec- ular cloning have been described (Nehlin et al., 1989; Nehlin and Ronne, 1990). The entire PPAB sequence was determined on both strands after subcloning in pUC118 and pUC119 (Vieira and Messing, 1987). DAPI staining of ethanol-fixed cells was carried out as de- scribed by Pringle et al. (1989). The cells were photographed in a Leitz Ortholux 2 microscope equipped with a Ploemopak 2 fluores- cence illuminator, using filter system A and a 63/1.30-0.60 objective. Phylogenetic trees were computed from distance data using the Fitch- Margoliash method (1967) as modified by Felsenstein (1985), or the neighbor-joining method of Saitou and Nei (1987). The sequences were first aligned using the MULTISEQ program.2 The proportion of amino acid replacements, p, was determined in pairwise compari- sons of the aligned sequences. The p scores were converted to evolu- tionary distances using the formula d = -ln(l-p-0.2 p') to correct for multiple and parallel mutations (Kimura, 1983). These distances were then used as input data for the FITCH and NJTREE programs (Felsenstein, 1985; Saitou and Nei, 1987).

RESULTS

Cloning of PPAl and PPA2"In a search for gene(s) encod- ing mitochondrial PPase, we first screened a yeast high copy number library for plasmids conferring increased PPase ac- tivity (see "Experimental Procedures"). We expected to clone the previously known PPAI gene, encoding cytoplasmic PPase (Kolakowski et al., 1988), but also genes coding for other PPases, in particular mitochondrial PPase. However, we only found plasmids containing PPAl. This is probably due to the fact that only PPAl has a clear overexpression phenotype in the PPase plate assay (see below). We proceeded on the assumption that any other PPase genes might be structurally related to PPAl. We therefore screened the li- brary again, by low stringency hybridization to a PPAl probe. One plasmid was found that contained the 5' end of a new PPase gene. A fragment from this plasmid was used as a probe to isolate the entire gene, which we call PPA2. Southern hybridizations with PPAl and PPA2 probes suggest that no other closely related genes are present in yeast. PPA2 Encodes a Mitochondrial PPase-The PPA2 gene

was located to chromosome 13 in a Southern blot of yeast DNA separated on a CHEF gel (Chu et al., 1986). A 4.8-

' H. Ronne, unpublished data.

kilobase Sac1 fragment carrying PPAB was subcloned and restriction mapped (Fig. l), and the sequence of the gene was determined (Fig. 2). PPA2 encodes a polypeptide of 310 resi- dues, which is a typical soluble protein lacking long hydro- phobic regions. A TATA box is found at position -245, but no other sites for known regulatory proteins (Verdier, 1990) are present in the promoter. The sequence of the PPA2 protein is 49% identical to the cytoplasmic PPase encoded by PPAl (Fig. 3). However, PPA2 has an amino-terminal leader sequence of 30 residues not found in PPA1, which is rich in basic and hydroxylated amino acids and has a high propensity to a-helix formation. This suggested that it could be a mito- chondrial targeting signal (von Heijne et al., 1989).

To confirm that PPA2 encodes a mitochondrial protein, we isolated mitochondria from cells carrying the PPA2 gene on the high copy number plasmid pHR81 (Nehlin et al., 1989). These cells had a 47-fold increase in mitochondrial PPase activity as compared to cells harboring pHR81 without the PPAB insert (Table I). In contrast, the cytoplasmic PPase activities were comparable (not shown). We conclude that PPA2 encodes a PPase which is located in the mitochondria. While cells overexpressing PPAI produce a clear zone in the PPase plate assay, cells overexpressing PPAB fail to do so. This explains why only PPAl was recovered in the PPase screen. The difference may be due to the fact that the cyto- plasmic PPase is expressed a t a much higher level. Alterna- tively, the substrate in the assay plates is inaccessible to the mitochondrial PPase.

Energy-linked enzymes are usually stimulated by uncou- plers. To determine whether PPAB is part of an energy-linked enzyme, we therefore tested the effect of the uncoupler FCCP on mitochondrial PPase activity. We found that the PPase in

V N V C

PPA I

S NVC X H X B H C AHV us 0 PPAP

1- pML6I

FIG. 1. Restriction maps of PPAl and PPA2. Open reading frames are indicated by arrows. The insert of the PPAS overexpres- sion plasmid pML61 is shown as a bar between the map. Abbrevia- tions: A, ApaI; B, SnaBI; C, ClaI; H, HindIII; N, NcoI; S, S a d ; V, EcoRV; X , XbaI.

l y T T W W ~ ~ ~ ~ A ~ K l y l A ~ M ~ M ~ ~ ~ ~ ~ ~ A ~ ~ M ~ A l ~ U ~ ~ 1 IkVI G T G l A l l l C l l C M a l l I476

FIG. 2. Sequence of the PPA2 gene and its predicted gene product. Nucleotide and amino acid numbers start at the first methionine codon in the PPAQ open reading frame. The leader peptide in the PPA2 protein is underlined. Also underlined is the TATA box in the PPA2 upstream region.

12170

P P A l SCE PPA1 KLA P P A I SPO PPAZ SCE

CONSERVED PPA ECO

PPAl SCE PPAl KLA PPAl SPO PPA2 SCE

CONSERVED PPA ECO

P P A l SCE P P A l KLA P P A l SPO PPA2 SCE

CONSERVED PPA ECO

P P A I SCE P P A I KLA P P A l SPD PPA2 SCE

CONSERVED PPA ECO

P P A l SCE P P A l KLA PPAI SPO PPAZ SCE CONSERVED

Yeast PPAB Gene Encodes a Mitochondrial Pyrophosphatase 20

MT-YTTRQIGAKNTLEYKVYIEK-DGKPVSAFHDIPLYAOKENNIFNMW;IP-RWTNA~

4 0

MS-YTTROVGAKNSLOYKVYIEK-DGKPISAFHDIPLYADEANGIFNMWEIP-RWTNAK MSEYTTREVGALNTLOYOVYVEK-NGTPISSWHDIPLYANAEKTIL~EIP-RWTOAK HRQFSTIMXISKYTLGFKKYLTLLNGEVGSFFHOVPLOLNEHEKTVNNIVEVP-RWTTGK

T G L Y G s H D P L MSLLNVPAGKOLPEDIY-WIEIPANADPIK

NN V E RWT 5

80 80 100

L;IfKEETLNPI IWTKKGKL~F'hNCFPHHGYlHNYGAFPOTWEDPNVSHPETK----A LElTKEEPLNPlIODTKKGKLRFVRNCFPHHGYIHNYGAFPOTWEDPNESHPETK----A LEITKEATLNPIKODTKKGKLRFVRNCFPHHGIlWNYGAFPOTYEDP~HPETK----A FElSKELRFNPlVODTKNGKLRFVNNIFPYHGYlHNYGAIPQTWEOPTIEHKLGKCDVAL YEIDKE-----------SGALFVDRFMST~FYPCNYGYINHT----------------L

. .

" EI KE NPI ODTK GKCRFV N FP H G ~ I N ~ G A POI EDP n K

1 2 a 140 Inn " . . .

V G ~ N ~ P I ~ V L E I G E T I A Y T O Q V K O V K A i G l M A L L ~ ~ G ~ T ~ ~ l A l D ~ N D P L A P K L N D l ~ VGDNDPLDVLEIGEOVAYTGOVKOVKVLG~ALLDEGETOWKVIAIDINDPLAPKLNDIE KGDSDPLDVCEIGEARGYTGOVKOVKVLG~ALLOEGETOWKVIVIDVNDPLAPKLNOIE KGDNDPLDCCEIGSDVLEMGSIKKVKVLGSLALIDDGELDWKVIVIOVNDPLSSKIDDLE SLDGDPVOVLVPTPYPLOPGSVIRCRPVGVLKMTDEAGEDAKLVA-VPHSKLSKEYDHIK GO Op 0 ElG 4 K VK L B AL 0 GE PWSVI ID NOPC K D E

."

180 .. ZOO

DVEKYFPGLLRA-TNEWFRIYK-IPDGKPENOFAFSGEAKNKKYALDI IUETHDSWKOLI

2 2 0

DVEKHLPGLLRA-TNEWFRIYK-IPDGKPENOFAFSGEAKNKKYTLOVIRECNEAWKKLI DVER~PGLIRA-TNEWFRIYK-IPDGKPENSFAFSGECKNRKYAEEWRECNEAWERLI KIEEYFPGILDT-TREWFRKYK-VPAGKPLNSFAFHEOYONSNKTlOTlKKCHNSWKNLl DV-NDLPELLKAOIAHFFEHYKDLEKGKWVK-----VE~ENAEAAKAElVASFERAKNK

E PG T EWER E P Gyp N FAF E W LI

2 4 0 2 8 0

AGKSSDSKGiDLTNVTLPDTPTYSKAASD~lPPASLKADAPlDKSlD~~FlSGSV SGKSAOAKKIDLTNTTLSDTATYSAEAASAVPAANVLPDEPIDKSID~FFISGSA TGKTDAKSDFSLVNVSVTGSVANOPSVSSTIPPAOELAPAPVDPSVH~FYISGSPL SGSLOEKYD-NLPNTERAGNGV--TLEOSVKPPSO------IPPEVOKWYYV

2 8 0

G L N P KW

FIG. 3. Alignment of PPase protein sequences. Consensus residues that are conserved in all eukaryotic PPases are shown below the alignment, with those conserved also in the E. coli PPase being underlined. Putative active site residues (Lahti et al., 1990a) are marked by asterisks. The numbering of the residues is that of the mature Saccharomyces PPAl protein. Abbreviations: SCE, S. cerevis- iae; KLA, K. lactis; SPO, S. pombe; ECO, E. coli.

TABLE I Mitochondrial PPase activities

Mitochondria were isolated from W303-1A cells containing either pHR81 or pML61, and the PPase activity was determined. Plasmid pML61 is pHR81 with a PPA2 insert. FCCP was used at a final concentration of 10 UM.

Plasmid -FCCP +FCCP nmol/PJmin/p~ protein

pHR81 0.16 0.37 D M L ~ ~ 7.5 22.0

mitochondria from both control and PPAB overproducing cells is stimulated 2-3-fold by FCCP (Table I). Since the PPase activity in the PPA2 overproducing cells was 47 times higher than in the control cells, we conclude that most of this activity is due to PPA2. The fact that the PPase activity in these cells is further stimulated 3-fold by FCCP therefore suggests that PPA2 is part of an energy-linked enzyme.

PPAl Is Essential for Growth-It is thought that hydrolysis of PP,, catalyzed by cytoplasmic PPase, is essential to main- tain the forward direction of a large number of metabolic reactions (Cooperman, 1982; Lahti, 1983). However, it has not been shown whether PPAI, which encodes the cyto- plasmic PPase, is an essential gene. To test whether this is the case, we disrupted PPAl in both haploid and diploid strains, using the LEU2 marker. No haploid transformants were obtained, which suggests that the PPAl gene is essential. The transformed diploids were allowed to sporulate and tet- rads were dissected. In each tetrad, two spores survived and two died. Surviving spores were all leu-, confirming that the PPAl disruption is lethal. PPAB on the high copy number plasmid pHR81 did not rescue the spores from this lethality.

PPAB Is Essential for Maintenance of the Mitochondrial Genome-To investigate the function of the PPA2 gene, we disrupted it with the URA3 marker. The disruption strain was viable, but could no longer grow on respiratory carbon sources such as glycerol, acetate, or ethanol (Fig. 4). Thus, it has the phenotype of a nuclear petite mutation, which is

GLUCOSE

ACETATE

ETHANOL

GLYCEROL

FIG. 4. Growth of wild type and ppa2- cells on various carbon sources. The ppa2- disruption strain H304 and its isogenic wild type W303-1A were grown on glucose, and then replicated to plates containing different carbon sources, as indicated in the figure.

FIG. 5. Absence of mitochondrial DNA in the ppa2- strain. The ppa2- strain H304 and its isogenic wild type W303-1A were grown in glucose, fixed in ethanol, and then stained with DAPI to visualize DNA.

consistent with a role for PPAB in mitochondrial metabolism. T o confirm that the respiratory phenotype is due to the PPA2 disruption, we crossed the strain to an isogenic wild type strain and dissected tetrads. The respiratory deficiency seg- regated 2:2, and was completely linked to the URA3 marker. Thus, the PPA2 gene is required for respiratory growth. TO further investigate the nature of the lesion, we used fluores- cence microscopy with the DNA-specific stain DAPI (Fig. 5). In the wild type strain W303-1A, the intense nuclear stain is surrounded by a granulated cytoplasmic stain caused by mi- tochondrial DNA (Pringle et d., 1989). In contrast, there is no cytoplasmic fluorescence in the ppa2-deficient strain H304 (Fig. 5). Thus, it appears to be a p" petite, i.e. a strain lacking mitochondrial DNA (Dujon, 1981). We conclude that PPA2 is essential for maintenance of the mitochondrial genome. This would be consistent with a role for PPAB in hydrolyzing PPi formed during mitochondrial DNA synthesis.

Evolution of PPases-To clarify the evolutionary position of the mitochondrial PPase, we compared its sequence to the soluble PPases. We first made painvise comparisons of five PPase sequences, and then used the resulting similarity scores to construct an evolutionary tree (see "Experimental Proce- dures"). The comparison was limited to those parts of the proteins where all five sequences could be unambiguously aligned. It is evident that the four eukaryotic PPases, includ- ing PPA2, are much more similar to each other than to the E. coli enzyme (Fig. 6). Among the eukaryotic PPases, the three cytoplasmic PPases stand closer to each other than to PPA2. We conclude that the PPAl and PPA2 genes were generated by a duplication event that took place long after the divergence of prokaryotes from eukaryotes, but well before the separation of budding yeasts from fission yeasts. The mitochondrial PPase seems to have undergone a period of accelerated evolution, since its branch length is considerably longer than those of the cytoplasmic PPases (Fig. 6).

Conserved Amino Acid Residues-Functionally important residues are well conserved in PPAZ. The crystal structure of PPAl (Terzyan et al., 1984; Kuranova, 1987) has identified 17 residues thought to be involved in binding M$+ and PPi (Lahti et al., 1990a). Evidence that some of these residues are

Yeast PPA2 Gene Encodes a Mitochondrial Pyrophosphatase 12171

Hi Saccharomyces PPAl

Khyveromyces PPAl

Schizosaccharomyces PPAI

Saccharomyces PPA2

Escherichia PPA

FIG. 6. Phylogenetic tree for PPases. Similarity scores were calculated from pairwise comparisons of PPase sequences. For this, we used three regions where the five sequences in Fig. 3 could be aligned unambiguously (residues 32-62, 74-99, and 112-193). The similarity scores were converted to evolutionary distances (Kimura, 1983) which were used to compute a tree by the Fitch-Margoliash procedure (1967). The neighbor-joining method of Saitou and Nei (1987) produced a tree with identical topology and almost the same branch lengths. The bur shows an evolutionary distance of 10%.

essential for enzyme function has also been obtained by modification studies and site-directed mutagenesis (Gonzalez and Cooperman, 1986; Lahti et al., 1990b). Of these 17 puta- tive active site residues, 16 are conserved in PPAB (Fig. 3). The only exception is Glu-148 in PPA1, where PPA2 has a chemically similar aspartic acid residue.

However, there is one conspicuous difference between PPA2 and the cytoplasmic PPases. It involves residues 114- 122 in PPAl (Fig. 3). This region contains several putative active site residues, and has the highest local similarity score in a comparison to bacterial PPase (Lahti et al., 1990a). Furthermore, a sequence motif similar to this region is found in the F,-ATPases (Baltscheffsky et al., 1987). The region is conserved in PPA2, but it is flanked by 3 cysteine residues not found in PPA1. Moreover, PPAB has an insertion of 4 residues adjacent to this region (Fig. 3). One of the 3 cysteines is also found in the cytoplasmic PPase of fission yeast (Ka- wasaki et al., 1990), bu the other 2 cysteines and the insertion are unique to PPA2. If two of the cysteines can form a disulfide bond, then the polypeptide would have a short loop spanning the active site region. Interestingly, such a cysteine loop is found in the ligand-gated ion channel family of neu- rotransmitter receptors (Schofield et al., 1987). In that case, the loop contains a conserved P-h-D motif, where h is a hydrophobic residue. Remarkably, the putative loop in PPA2 also contains a P-h-D motif. I t is conceivable that this could reflect some common function. However, there is no signifi- cant sequence similarity between PPA2 and the ligand-gated ion channels. The presence of these similar motifs must thus await a detailed interpretation.

DISCUSSION

We have cloned a new yeast gene, PPA2, which encodes a mitochondrial PPase. Like bacterial chromatophores (Balts- cheffsky and Nyrbn, 1984), mitochondria contain a membrane bound PPase (reviewed by Mansurova, 1989). The subunit structure of this mitochondrial PPase has been determined for the mammalian enzyme (Volk et al., 1983; Volk and Baykov, 1984). The catalytic subunit, a dimer of a 28-30-kDa polypeptide, is a soluble protein. It binds noncovalently to a protein complex in the inner mitochondrial membrane, con- ceivably a proton channel similar to the Fo-ATPase. The subunit structure of the yeast mitochondrial PPase is not known. However, the quartenary structures of other mito- chondrial proteins, such as the ATPase (Walker et al., 1985), are rigorously conserved between animals and fungi. It is therefore a reasonable assumption that the yeast enzyme is similarly organized to its mammalian counterpart. The PPA2 polypeptide has a molecular mass of 32 kDa if the targeting peptide is excluded. It is thus similar in size to the catalytic subunit of mammalian mitochondrial PPase, and is likely to

be a homologue of the latter. The sequence similarity of PPAB to the soluble PPases supports the notion that it is the catalytic subunit of a mitochondrial PPase. Further evidence that PPA2 is part of an energy-linked PPase comes from our finding that the elevated mitochondrial PPase activity in cells overproducing PPA2 is stimulated %fold by the uncoupler FCCP (Table I).

The PPA2 protein is much more similar to eukaryotic cytoplasmic PPases than to the E. coli PPase (Fig. 6). This is surprising, considering what is known about mitochondrial evolution. Thus, when homologous proteins are found in both cytoplasm and mitochondria, the mitochondrial protein is usually more similar to its bacterial counterpart than to the cytoplasmic protein. I t is thought that this reflects an evolu- tionary origin of mitochondria from purple bacteria (Woese, 1987). Even mitochondrial proteins that are encoded in the nucleus, such as the F1-ATPase subunits and elongation factor Tu, are frequently more similar to bacterial proteins (Iwabe et al., 1989). In such cases, it is thought that the genes moved from the organelle to the nucleus during evolution (Fox, 1983). I t is possible that the mitochondrion originally had a soluble PPase that was inherited from its bacterial ancestor. How- ever, the similarity of PPA2 to cytoplasmic PPases shows that the PPA2 gene is not of mitochondrial origin. Instead, PPAl and PPAB were created by duplication of a nuclear eukaryotic gene encoding a soluble PPase.

A comparison of cytoplasmic PPases can provide informa- tion on when this duplication occurred. The evolutionary distance between fission yeast and budding yeast equals that between mouse and carp (Qu et al., 1988). The ancestors of the latter two species became separated during the Silurian period, 420 million years ago (Nelson, 1969). In our compari- son, the distance between PPAl and PPA2 is 3 times that between the cytoplasmic PPases of fission and budding yeasts. However, this long distance is in part due to an accelerated evolution of PPA2. Thus, the tree in Fig. 4 suggests that the divergence of PPAl from PPA2 is in fact 2.1 times more ancient than the split between fission and budding yeasts. This would put the PPAl/PPA2 duplication approximately 880 million years ago. Interestingly, the endosymbiosis that generated mitochondria is thought to have occurred between 800 and 1500 million years ago, with an estimate based on bacterial sequence divergence suggesting a figure of 900 mil- lion years (Ochman and Wilson, 1987). Thus, it would appear that a separate nuclear gene encoding mitochondrial PPase was created soon after the endosymbiotic origin of mitochon- dria. The accelerated evolution of PPAB could be due to selective adaptation to the mitochondrial environment and to a function in energy coupling (see below). This would be similar to the rapid evolution that took place in vertebrate hemoglobin during its adaptation to a new function in the blood (Goodman et al., 1975).

Since PPA2 is closely related to the soluble cytoplasmic PPases, it is conceivable that it also has a similar function, i.e. to facilitate biosynthetic reactions that generate PPi. The fact that PPA2 is required for maintaining the mitochondrial genome is consistent with such a function, as DNA synthesis is one reaction that generates PP;. It is possible that PPA2 was originally imported into the mitochondrion to perform this essential function, thus replacing a soluble PPase present in the organelle’s bacterial ancestor. However, the unique role of the mitochondrion in energy metabolism may have pro- vided an opportunity for further developments. Thus, cou- pling of the PPase to a proton channel would have made it possible to save some of the energy released, for use in energy- requiring reactions. It is conceivable that a selective pressure

12172 Yeast PPAB Gene Encodes a Mitochondrial Pyrophosphatase

for this development existed in the early eukaryote, and that the apparent accelerated evolution of PPAB reflects an ad- aptation to this new function.

Clearly, further studies, including the cloning of the puta- tive proton channel and studies on the structural and dynamic properties of soluble and membrane-bound PPases, could provide answers contributing to the understanding of both biological energy conversion and phosphate metabolism.

Acknowledgments-We thank Monika Carlberg for excellent tech- nical assistance, Reijo Lahti and Jan Olof Nehlin for helpful discus- sions, and Lena Welsh for advice on the low stringency protocol.

REFERENCES

Baltscheffsky, M. (1967) in Regulatory Functions of Biological Mem- branes (Jarnefelt, J., ed) BBA Library Vol. 11, pp. 277-286, Elsevier Scientific Publishing Co., Amsterdam

Baltscheffsky, H. (1981) BioSystems 14, 49-56 Baltscheffsky, M., and NyrLn, P. (1984) in New Comprehensive

Biochemistry: Bioenergetics (Ernster, L., ed) Vol. 9, pp. 187-206, Elsevier Scientific Publishing Co., Amsterdam

Baltscheffsky, H., von Stedingk, L.-V., Heldt, H.-W., and Klingen- berg, M. (1966) Science 133,1120-1122

Baltscheffsky, H., Lundin, M., Luxemburg, C., NyrLn, P., and Balts- cheffsky, M. (1986) Chem. Scr. 26B, 259-262

Baltscheffsky, H., Alauddin, M., Falk, G., and Lundin, M. (1987) Acta Chem. Scand. B41, 106-107

Britten, C. J., Turner, J . C., and Rea, P. A. (1989) FEBS Lett. 256, 200-206

Chu, G., Vollrath, D., and Davis, R. W. (1986) Science 234, 1582- 1585

Cohen, S. A., Sterner, R., Keim, P. S., and Henrikson, R. L. (1978) J . Biol. Chem. 253,889-897

Cooperman, B. S. (1982) Methods Enzymol. 87, 526-548 Dujon, B. (1981) in T h e Molecular Biology of the Yeast Saccharomyces:

Life Cycle and Inheritance (Strathern, J., Jones, E. W., and Broach, J. R., eds) pp. 505-635, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY

Felsenstein, J. (1985) Euolution 39, 783-791 Fox, T. D. (1983) Nature 301, 371-372 Fitch, W. M., and Margoliash, E. (1967) Science 155, 279-284 Gonzalez, M. A., and Cooperman, B. S. (1986) Biochemistry 25,

Goodman, M., Moore, G. W., and Matsuda, G. (1975) Nature 253,

Halestrap, A. P. (1989) Biochim. Biophys. Acta 973, 355-382 Ichiba, T., Takenaka, P., Samejima, T., and Hachimori, A. (1990) J.

Iwabe, N., Kuma, K., Hasegawa, M., Osawa, S., and Miyata, T. (1989)

Kawasaki, I., Adachi, N., and Ikeda, H. (1990) Nucleic Acids Res. 1 8 ,

Kimura, M. (1983) The Neutral Theory of Molecular Euolution, Cam-

Kolakowski, L. F., Jr., Schloesser, M., and Cooperman, S. (1988)

7179-7185

603-608

Biochem. (Tokyo) 108,572-578

Proc. Natl. Acad. Sci. U. S. A. 86, 9355-9359

5888

bridge University Press, Cambridge

Nucleic Acids Res. 22, 10441-10452

Microbiol. Lett. 15, 309-311 Kukko, E., Tammi, K., Sundell, L., and Heinonen, J . (1982) F E M S

Kuranova, I. P. (1987) Biokhimiya 53, 1821-1827 Lahti, R. (1983) Microbiol. Reu. 47, 169-179 Lahti, R., Pitkaranta, T., Valve, E., Ilta, I., Kukko-Kalske, E., and

Heinonen, J. (1988) J. Bacteriol. 170, 5901-5907 Lahti, R., Kolakowski, L. F., Jr., Heinonen, J., Vihinen, M., Pohja-

noksa, K., and Cooperman, B. S. (1990a) Biochim. Biophys. Acta 1038,338-345

Lahti, R., Pohjanoksa, K., Pitkaranta, T., Heikinheimo, P., Salminen, T., Meyer, P., and Heinonen, J. (1990b) Biochemistry 29, 5761- 5766

Lundin, M., Pereira da Silva, L., and Baltscheffsky, H. (1987) Biochim. Biophys. Acta 890, 279-285

Maeshima, M., and Yoshida, S. (1989) J. Biol. Chem. 264, 20068- 20073

Mansurova, S. E. (1989) Biochim. Biophys. Acta 977, 237-247 Maslowski, P., and Maslowska, H. (1987) Biochem. Physiol. Pflanz.

Nehlin, J. O., and Ronne, H. (1990) E M B O J. 9, 2891-2898 Nehlin, J . O., Carlberg, M., and Ronne, H. (1989) Gene (Amst.) 85,

Nelson, G. J. (1969) Ann. N . Y. Acgd. Sci. 167, 18-30 Nyrbn, P., Nore, B. F., and Strid, A. (1991) Biochemistry, 30, 2883-

Ochman, H., and Wilson, A. C. (1987) J. Mol. Euol. 26, 74-86 Pringle, J. R., Preston, R. A., Adams, A. E. M., Steams, T., Drubin,

D. G., Haarer, B. K., and Jones, E. W. (1989) Methods Cell Biol. 3 1,357-435

Qu, L.-H., Nicoloso, M., and Bachellerie, J.-P. (1988) J. Mol. Euol.

Rothstein, R. J. (1983) Methods Enzymol. 101, 202-211 Saitou, N., and Nei, M. (1987) Mol. Biol. Euol. 4, 406-425 Sarafian, V., and Poole, R. J. (1989) Plant Physiol. 91, 34-38 Schofield, P. R., Darlison, M. G., Fujita, N., Burt, D. R., Stephenson,

F. A., Rodriguez, H., Rhee, L. M., Ramachandran, J., Reale, V., Glencorse. T. A.. Seeburp. P. H., and Barnard, E. A. (1987) Nature

182,73-84

313-319

2887

28,113-124

-. 328,221-227 '

Shatton. J. B.. Ward. C.. Williams. A.. and Weinhouse. S. (1982) , , . . Anal. biochek. 130, 114-119

Stark, M. J . R., and Milner, J. S. (1989) Yeast 5,35-50 Terzyan, S. S., Voronova, A. A., Smirnova, E., Kuranova, I. P.,

Nekrasov, Y. V., Arutyunyan, E. G., Vainshtein, B. K., Hohne, V., and Hansen, G. (1984) Bioorg. Khim. 10, 1469-1482

Thomas, B. J., and Rothstein, R. (1989) Cell 56,619-630 Verdier, J.-M. (1990) Yeast 6, 271-297 Vieira, J., and Messing, J . (1987) Methods Enzymol. 153, 3-11 Volk, S. E., and Baykov, A. A. (1984) Biochim. Biophys. Acta 791,

Volk, S. E., Baykov, A. A,, Kostenko, E. B., and Avaeva, S. M. (1983) Biochim. Biophys. Acta 744, 127-134

von Heijne, G., Steppuhn, J., and Herrmann, R. G. (1989) Eur. J . Biochem. 180,535-545

Walker, J. E., Fearnley, I. M., Gay, N. J., Gibson, B. W., Northrop, F. D., Powell, S. J., Runswick, M. J., Sarastre, M., and Tybulewicz,

198-204

V. L. J. (1985) J. Mol. Biol. 184, 677-701 Woese, C. R. (1987) Microbiol. Reu. 5 1 , 221-271