Incomplete Arrest in the Outer Membrane Sorts NADH ... - Cell Press

Cell, Vol. 79, 829-839,

December

2, 1994, Copyright

0 1994 by Cell Press

Incomplete Arrest in the Outer Membrane Sorts NADH-Cytochrome b5 Reductase to Two Different Submitochondrial Compartments KersUn Hahne, Volker and Gottfried Schatz Biozentrum University of Base1 Klingelbergstrasse 70 CH-4056 Base1 Switzerland

Haucke,

Lynn Ramage,

Summary The S. cerevisiae gene MCRI encodes two mitochondrial isoforms of NADH-cytochrome b5 reductase. The primary translation product has an amino-terminal matrix-targeting signal, followed by a stretch of 21 uncharged amino acids. This precursor protein is inserted into the outer membrane, but only about one-third of the molecules become firmly anchored to the outer face of that membrane. The remaining molecules pass through the outer membrane into the inner membrane, are cleaved by inner membrane protease 1, and are released into the intermembrane space. Incomplete translocation arrest in the outer membrane is a novel mechanism by which the product of a single gene is sorted into different compartments of the same organelle. Introduction Proteins with similar functions are often localized in more than one compartment of the eukaryotic cell. Such a topological diversity can be generated by several distinct mechanisms. In many cases, the isoproteins are encoded by a family of related genes that may be derived from a common ancestral gene. A well-studied example is the malate dehydrogenase family, members of which are found in the cytosol as well as in mitochondria, chloroplasts, and glyoxysomes of plant cells (Gietl, 1992). Alternatively, a single gene can yield different mRNAs that are translated into protein products that either contain or lack a given targeting signal. Such mRNA diversity can be achieved by the use of different transcription start sites (Perlman et al., 1984) or by differential splicing of aprecursor mRNA (Heim et al., 1992). Yet another mechanism is the selective use of alternative translational start codons on a single mRNA. (Boguta et al., 1994). Furthermore, targeting signals can be only partially effective (Garcia et al., 1988) or can be generated or inactivated by posttranslational modifications (Mall et al., 1991; Caplan et al., 1992). Finally, there exist as yet unknown mechanisms for targeting the protein product of a single gene to different subdomains of the mammalian plasma membrane (Simons and Fuller, 1985; Wessels et al., 1990). Here, we report that topological diversity of isozymes can also be achieved by inefficient translocation arrest in a biological membrane. We show that yeast mitochondria contain two NADH-cytochrome bs reductase isozymes

that are encoded by the same nuclear gene but differ in their amino termini and their intramitochondrial locations. One form, the primary translation product, is anchored to the mitochondrial outer membrane by an amino-terminal anchor sequence. However, translocation arrest in the outer membrane is leaky; more than half of the molecules reach the inner membrane and are then processed to a smaller form that is released into the intermembrane space. “Leaky stop-transfer” in a membrane is thus a novel mechanism for targeting the product of a single gene to different compartments of the same organelle. Results Identification of Yeast Mitochondrial NADH-Cytochrome b5 Reductase In an attempt to characterize the major proteins of the yeast mitochondrial outer membrane, we isolated a 34 kDa outer membrane protein, raised antisera against it, and determined its amino-terminal sequence as well as the sequences of two tryptic peptides. Both peptides were homologous to conserved regions of NADH-cytochrome bg reductase from different mammalian sources (Figure 1A). The amino-terminal sequence (MFSRLSRSHSKA) did not resemble any known protein sequence, but it did exhibit features typical of a mitochondrial matrix-targeting signal: it was rich in basic and hydroxylated residues, lacked acidic residues, and could potentially form an amphipathic helix whose first three helical turns would generate a hydrophobic moment of 0.64. MCR1, the Nuclear Gene for Yeast Mitochondrial NADH-Cytochrome b, Reductase Using the partial amino acid sequence of the 34 kDa protein, we cloned the corresponding yeast gene from a genomic yeast library in an yeast-Escherichia coli shuttle vector. The complete nucleotide sequence revealed an open reading frame encoding a protein of 302 amino acids with a calculated molecular weight of 34,337 (Figure 1 B). This value agrees with the mobility of the protein on SDS-polyacrylamide gel electrophoresis (SDS-PAGE). Immunoblotting experiments confirmed that yeast transformants carrying the multicopy shuttle vector with the intact gene overexpressed the 34 kDa protein (data not shown). Unexpectedly, these transformants also overexpressed acrossreacting 32 kDa protein. As discussed below, this smaller protein was not a degradation product but a genuine isoform of NADH-cytochrome b5 reductase. The sequence of the predicted 34 kDa protein showed 59.2% similarity and 35.9% identity to that of rat liver microsomal NADH-cytochrome bs reductase (Pietrini et al., 1992) and comparable similarities to NADH-cytochrome bs reductases from cattle (0~01s et al., 1985), humans (Yubisui et al., 1986, 1987; Tomatsu et al., 1989), and yeast (Csukai et al., 1994). The homology was found mainly in the presumed catalytic regions, particularly those implicated in the binding of NADH and FAD. We also detected

Cell 830

Wild-Type Yeast Cells Have a Single MCR7 Gene Southern blot analysis of yeast genomic DNA confirmed that yeast cells contain only a single MCR7 gene (Figure 2A). To determine the function of the MC/?7 gene, we created a gene disruption by inserting the wild-type URA3 gene into 1 of the 2 copies of the MCR7 gene in a diploid yeast strain homozygous for the ura3 mutation; then we sporulated the diploid and analyzed the haploid spores. The four spores originating from an ascus yielded the expected 2:2 segregation of uracil prototrophy and of the wild-type and the disrupted MCR7 gene, as determined by Southern blotting. All four spores grew normally on all media and at all temperatures tested (data not shown). Disruption of the MCR7 gene thus has no obvious phenotypic effects. We cannot exclude that the disruptants still produce an active fragment of the reductase, but we consider this possibility unlikely.

;: !,/,. 8’ “C.

Figure (A) Two ttyptic peptides from a 34 kDa mitochondrial outer membrane protein resemble separate regions in human NADH-cytochrome bs reductase. The numbers flanking the human sequences identify the residues in the human reductase (Yubisui et al., 1987). The underlined sequences were used to design degenerate oligonucleotide primers for PCR. (B) Nucleotide sequence of the MCR7 gene and deduced amino acid sequence of the protein product. The upper numbers in the margins indicate the nucleotide number in the open reading frame, and the numbers below indicate the number of the amino acid in the deduced sequence. Putative TATATAA and CAAT boxes are underlined.

similiarity to the flavin-binding domain of nitrate reductase from spinach (Prosser and Lazarus, 1990) and to other members of the flavoenzyme family, such as ferredoxinNADP reductase and sulfite reductase (data not shown). These comparisons suggest that the 34 kDa protein is a mitochondrial NADH-cytochrome bs reductase. We term its gene MCRl. The sequence of the 49 amino-terminal residues of the yeast mitochondrial NADH-cytochrome bs reductase did not resemble that of other flavoenzymes, suggesting that these residues mediate intracellular targeting.

Disruption of the MCR7 Gene Abolishes Synthesis of Two Proteins To confirm that the disruption of the MCR7 gene abolished synthesis of the corresponding protein product, mitochondriafrom wild-type yeast and from the haploidmcrl::URA3 disruptant were analyzed by immunoblotting with antiserum raised against the 34 kDa outer membrane protein. In contrast with the results obtained with purified outer membranes, the serum reacted with two different proteins in whole mitochondria: the 34 kDa protein and a 32 kDa protein. Mitochondria from the haploid disruptant lacked both proteins (Figure 28). Conversely, yeast transformants containing multiple copies of MCR7 expressed 5 to IO-fold higher levels of both proteins (data not shown). We conclude that the MCR7 gene encodes two isoforms of NADH-cytochrome bs reductase, a 34 kDa form (Mcrlp%) and a 32 kDa form (Mcrl p”‘). Both lsoforms of NADH-Cytochrome bs Reductase Are Mitochondrial Proteins To determine the subcellular location of the two isoforms of Mcrl p, equal amounts of microsomes, crude mitochondria, gradient-purified mitochondria, and purified outer mitochondrial membranes were analyzed by SDS-PAGE and immunoblotting with antisera raised against either Mcrlp%, porin (an outer membrane marker), or KaRp (an endoplasmic reticulum marker). McrlpU was found in outer membranes, crude mitochondria, and purified mitochondria, but not in microsomes; Mcrlp3* was found in crude or purified mitochondria, but not in outer membranes or microsomes. No KarPp was detected in purified mitochondria (Figure 3A). These results suggest that both NADH-cytochrome bs reductase isoforms are mitochondrial proteins, but that they are located in different intramitochondrial compartments. Mcrl~~~ Is Located at the Mitochondrial Surface, Whereas Mcrlps2 Resides in the Intermembrane Space Mcrl pU has a putative amino-terminal matrix-targeting signal, followed by21 uncharged, mostly hydrophobicresidues. A similar amino-terminal motif is present in the mito-

One Gene 331

Encodes

Two Mitochondrial

A

lsoproteins

Figure 2. Disruption of the Single MCRl Loss of Two Proteins

kb

rc 4, +

5

-c -

15

45 _ -

Causes

(A) Wild-type yeast cells have only a single gene for the mitochondrial NADH-cytochrome 9 reductase. Chromosomal DNA from two uracilprototrophic diploid transformants (lanes 1 and 2) and from wild-type cells (lane labeled WT) was cut with Hincll and was analyzed by Southern blotting. The numbers on the right indicate the size of DNA molecular weight standards in kilobases. (B) In haploid wild-type cells, two proteins of 34 kDa and 32 kDa react with antiserum against the 34 kDa protein; both proteins are missing in the mcrl:tURA3 disruptant. The diploid strain carrying one disrupted copy of the MCR7 gene was sporulated and the ascospores were dissected. The indicated amounts of mitochondria from a ura+ spore (lanes labeled dMCR7) and from a ura- spore (lanes labeled WT) were subjected to SDS-PAGE and were analyzed by immunoblotting with antisera raised against the 34 kDa outer membrane protein (34 kDa, 32 kDa) and HspSCI.

5.1 4.2 3.5

5

Gene of S. cerevisiae

15

45

pg mitochondrial

+

hsp60

++

34 kDa 32 kDa

chondrial import receptor Mas70p (Hase et al., 1984). Mcrl pa4, like Mas70p, is degraded by low levels of proteinase K in intact mitochondria (Figure 38; see also Hase et al., 1984; Ramage et al., 1993). Mcrlp* is not extracted from mitochondrial membranes at pH 11.5 (Figure 3C); thus, it is probably firmly anchored to the outer membrane by the hydrophobic sequence near the amino terminus, with its catalytic carboxy-terminal domain exposed to the cytosol. NADH-cytochrome bs reductase of the mammalian endoplasmic reticulum has the same topology (Borgese and Pietrini, 1986). In contrast, Mcrl pZ of intact mitochondria is inaccessible even to high concentrations of protease (Figure 36) and is extractable at pH 11.5 (Figure 3C). Mcrlp= is thus located inside the mitochondria and is either a soluble or a peripheral membrane protein. To determine the intramitochondrial location of the smaller isoform, we selectively disrupted the mitochondrial outer membrane by hype-osmotic shock (Figure 4A)

protein

or by digitonin treatment (Figure 48) and then we measured the protease accessibility (Figure 4A) or the solubilization (Figure 48) of Mcrl p3* and of marker proteins for the different submitochondrial compartments. In both experiments, Mcrlp3* behaved exactly like cytochrome b2, a soluble protein of the intermembrane space. All other markers behaved as expected, attesting to the validity of these subfractionation experiments. We conclude that the smaller NADH-cytochrome bg reductase isoform of yeast mitochondria is a soluble protein of the intermembrane space. The Smaller lsoform of NADH-Cytochrome bs Reductase Is Generated by Inner Membrane Protease 1 Amino-terminal sequencing of the smaller iaoform (Mcrlp3*) yielded the sequence ESNKVFKGDDK; the amino-terminus of the smaller form thus corresponds to residue 42

Cdl 832

B

+ TX100

1 -TX100

4

porin

+

Kar2p

+ +

Mcrlps4 Mcrlp32

of the larger form (Figure 5A), suggesting that Mcrlp3> arises by cleavage of the N41-E42 bond in the full-length precursor. Asimilarcleavageoccursduring thematuration of cytochrome b) (cleavage of an N-E bond) and of cytochrome oxidase subunit II (cleavage of an N-D bond) by the action of inner membrane protease 1, which faces the outer surface of the inner membrane (Behrens et al., 1991; Schneider et al., 1991). We therefore suspected that this protease might be responsible for processing the primary translation product of MCR7 to Mcrl~~~. Indeed, a yeast mutant in which this inner membrane protease 1 is temperature-sensitive @et rs2858; Pratje and Guiard, 1986) contained only Mcrl p% when grown at the nonpermissive temperature, but it contained both Mcrl p% and Mcrlp3* when grown at the permissive temperature (Figure 58). Therefore, the smaller isoform of NADH-cytochrome bg reductase is generated from the primary translation product by inner membrane protease 1.

1

- I= Mcrlp34 Mcrl

rll-

p32

cyt b

C

1 TI

PIS Mcrl p34 Mcr 1~32 hsp70 porin

Figure 3. Both NADH-Cytochrome chondrial Proteins

b5 Reductase

lsoforms

Are Mito-

(A) Cofractionation with mitochondria. Fractions (100 @g each) of crude mitochondria (M), microsomes (ER), purified mitochondria (N), and mitochondrial outer membranes (0) were analyzed by SDS-PAGE and immunoblotting for porin (mitochondrial marker), Kar2p (ER marker), and the 34 kDa outer membrane protein; immune complexes were visualized with ‘251-labeIed protein A and autoradiography. (B) The larger isoform is protease accessible in intact mitochondria, whereas the smaller isoform is protease protected. Intact mitochondria were treated with the indicated concentrations of proteinase K in the presence (plus sign) or absence (minus sign) of Triton X-100 (TXlOO) and were analyzed by SDS-PAGE and immunoblotting with antisera against the intermembrane space marker cytochrome bz (cyt bl) and

The Two lsoforms Are Produced by Related, but Distinct Pathways When wild-type cells were pulse-labeled for 5 min with [%]methionine, both forms of NADH-cytochrome bs reductase became labeled to approximately the same extent (ratio of Mcrl pYMcrlpz, 1.22). During a subsequent chase in the presence of unlabeled methionine, about half of the larger form was processed to the smaller one within 10 min, reducing the Mcrl pYMcrlpz ratio to 0.42. (Figure 6A). This ratio resembles the steady-state levels detected by immunoblotting (see Figures 3-5). Conversion was inhibited by collapsing the electrochemical potential across the inner membrane (Figure 6B), suggesting that it required interaction of the matrix-targeting signal of Mcrl pa with the import machinery of the inner membrane (Martin et al., 1991; Schatz, 1993). Taken together, these results indicate that the intermembrane space isoform is generated from the primary translation product by inner membrane protease 1. The fact that the two isoformsare present in comparable amounts makes it unlikely that the larger one is merely a short-lived precursor of the smaller one. Further support for this conclusion was obtained from synthesizing Mcrl p in vitro and importing it into isolated mitochondria. As expected, import into fully energized mitochondria yielded both Mcrl pa and Mcrl pa2, although generation of Mcrl pa2 was less efficient than it is in vivo (Figure 6C, upper panel, Import). However, when Mcrl pU was first accumulated in the outer membrane by import in the absence of a membrane potential, none of it could be chased to Mcrl p3’ upon

the 34 kDa outer membrane protein. lmmunocomplexes were visualized by 1251-labeled protein A and autoradiography. (C) Only the larger isoform is an integral membrane protein. Mitochondria were extracted at pH 11.5. Equal aliquots of supernatant (S), pellet (P), and total mitochondria (T) were analyzed by SDS-PAGE and by immunoblotting with antisera against porin (marker for an integral membrane protein), Hsp70 (marker for a soluble protein), and the 34 kDa outer membrane protein. Blots were developed with the alkaline phosphatase color reaction as recommended by the manufacturer (Promega).

One Gene 833

Encodes

Two Mitochondrial

lsoproteins

Figure 4. The Smaller lsoform of Mitochondrial NADH-Cytochrome bs Reductase Is Localized in the Intermembrane Space

A

w

aJlllaluuND---

.

.._ .-.

+- KDH

(A) Corelease of Mcrlp with intermembrane space proteins by hypotonic shock. Mitochondria were resuspended in import buffer (Glick et al., 1992) and were diluted IO-fold into 20 mM K+-HEPES (pH 7.4) containing 100 vglml proteinase K and the indicated concentrations of sorbitol. After 20 min on ice, PMSF was added to 1 mM, and the organelles were reisolated and treated with trichloroacetic acid. As a control, one sample was not treated with proteinase K. The samples were analyzed by SDS-PAGE and by immunoblotting with antisera against the matrix protein a-ketoglutarate dehydrogenase (KDH), the intermembrane

-

rC Mcrlpa 4 .06 .075

.l

.2

.3

.6

.6

++++++-

B

Mcrlp32 M sorbiiol

space

Proteinase K

I-

O .8 .12 .16 .20 .28

% digitonin

c

cyt b2

c

lSP45P

C

porin

C c

Mcrlps4 Mcrlp32

.-

MPPp

reestablishing the potential (Figure 6C, Chase); instead, some of the accumulated Mcrl p% was degraded after 10 min. In the same experiment, the matrix-targeted precursor pSU9-DHFR (a fusion protein consisting of the presequence of ATPase subunit 9 of Neurospora crassa fused to mouse dihydrofolate reductase; Pfanner et al., 1987) could be bound to the outer membrane of deenergized mitochondria and could then be efficiently chased into the mitochondria upon restoring the membrane potential (Figure 6C, lower panel). We also tested the possibility that the observed steadystate levels of the two forms in vivo reflect rapid processing of Mcrl p%, followed by rapid degradation of Mcrl pa2. To this end, we measured the intracellular levels of both isoforms in wild-type yeast cells at various times after protein synthesis had been inhibited in midlogarithmic phase by cycloheximide (Figure 7). If Mcrl pW was mainly a precursor for Mcrl pa2, Mcrlp% should decrease and Mcrlp32 should increase during the chase. However, neither the

protein

cytochrome

b, (cyl bp), and the

34 kDa outer membrane protein. Immune complexes were visualized by ‘Wabeled protein A and autoradiography. (8) Corelease of McrlpY with intermembrane space proteins by digitonin treatment. Mitochondria were treated with digitonin (see Experimental Procedures), and the subfractions were analyzed by SDS-PAGE and by immunoblotting with antisera against cytochrome b2

(intermembrane space of the matrix processing

marker), the p subunit peptidase MPPg (ma-

trix marker), porin (outer membrane marker), lsp45p (inner membrane marker), and the 34 kDa outer membrane protein.

absolute nor the relative amounts of the two isoforms changed significantly for as long as 4 hr. One-third to one-half of the Mcrlp molecules are thus stably inserted into the outer membrane, whereas the others are transported further to the inner membrane and are processed to the soluble form in the intermembrane space. Discussion The Puzzle of Multiple NADH-Cytochrome bs Reductases At least three different NADH-cytochrome bs reductases are present in mammalian cells. They are located on the surface of the endoplasmic reticulum, on the surface of mitochondria, and in the soluble lumen of erythrocytes (Passon and Hultquist, 1972; Borgese and Pietrini, 1986; Pietrini

et al., 1992).

The

microsomal

electron transfer chain that participates

enzyme

is part

of an

in the desaturation

A MFS+RL&HS~ALPIALGT”AlAAATAFYFAN;NQHSF”FN

Figure

5. The Smaller

lsoform

is Generated

from the Larger

lsoform

by Inner

Membrane

Protease

-

hsp60

2 -

Mcrlpa Mcrl pa*

1

(A) Amino-terminal sequences of Mcrlp” and Mcrlp”. The predicted cleavage site for inner membrane protease 1 between N41 and E42 of Mcrl p” is indicated by an arrow. The experimentally determined amino-terminal sequence of Mcrl py is underlined. The hydrophobic sequence stretch is in bold; basic and acidic amino acids are marked by plus and minus signs, respectively. (6) A yeast mutant with a temperature-sensitive inner membrane protease 1 fails to generate Mcrlp” at the nonpermissive temperature, The yeast mutant pet rs2858 (Pratje and Guiard, 1966) was grown at 23OC or nonpermissive 36OC in rich medium containing 2% galactose (YPGal). Total cell extracts were prepared (Yaffe and Schatz, 1984) and analyzed by SDS-PAGE and by immunoblotting with antiserum against Hsp60 (mitochondrial marker) and the 34 kDa outer membrane protein. The blot was developed with ‘2SI-labeled protein A.

Figure 6. Conversion of the Mcrlp Protein to Mcrlp” and Mcrl~~

A

Precursor

(A) Pulse-chase labeling of yeast cells shows that about 50% of the newly imported Mcrlp precursor molecules are converted to Mcrlp”. The relative amounts of the two labeled isoforms reach a similar value after 10 min as that

B

Mcrlp34 Mcrl p3*

C

Mcrl

p34

Mcrlp3*

labeled with[JSS]methioninesfor 5 min at 30°C and were chased for 0 min or 10 min at 30°C. Both isoforms of the reductase were immunoprecipitated from total cell extracts and were analyzed by SDS-PAGE and fluorography. (6) Generation of the smaller isoform requires a potential across the inner membrane. Yeast cells were labeled for 5 min at 3O’C with [%S]methionine in the presence (plus sign) or absence (minus sign) of 40 PM CCCP and were chased for IO min at 3O“C. The cells were then extracted, and both isoforms of NADH-cytochrome bs reductase were immunoprecipitated and analyzed by SDS-PAGE and fluorography. (C) Mcrlpa accumulated in the outer membrane cannot be chased to Mcrl p3’. The Mcrl p precursor was synthesized in vitro and was incubated for 5 min at O°C with isolated purified mitochondria that had been uncoupled with CCCP. An aliquot was then treated with proteinase K (Prot. 0: the remainder of the mitochondriawas reisolated, recoupled by destroying CCCP with dithiothreitol, and chased for the indicated times (Chase). Mcrlp was also directly imported intofullyenergized mitochondria (Import). In a parallel experiment using the same batch of mitochondria, in vitro synthesized pSlJ9-DHFR was first bound to the surface of uncoupled mitochondria and was then chased into the matrix upon reestablishing the membrane potential (lower panel). Abbreviation: STD, standard.

One Gene

Encodes

Two Mitochondrial

lsoproteins

835

30

60

Chase (min)

Figure 7. The Relative Amounts of the Two lsoforms upon Inhibition of Protein Synthesis

Do Not Change

Yeast cells overexpressing the MCR7 gene were grown at 30°C to an ODsoo of 0.5 in semisynthetic medium containing 2% galactose; were inhibited by 100 pg cycloheximidelml; and were analyzed at the indicated time points by extraction, SDS-PAGE, and immunoblotting with an antiserum against Mcrlp”. Blots were developed with ‘2SI-labeled protein A, radioautographed, and quantified by densitometry of three different amounts of samples for each point.

and elongation of fatty acid8 (Oshino et al., 1971; Reddy et al., 1977) and in drug metabolism (Hildebrandt and Estabrook, 1971); the ery-throcyte enzyme mediates the regeneration of hemoglobin from methemoglobin (Scott and Griffith, 1959) and the function of the mitochondrial isozyme remains unclear (Nishino and Ito, 1988). In mammals, NADH-cytochrome bs reductases of the endoplasmic reticulum and the mitochondrial outer membrane are encoded by the same gene (Pietrini et al., 1992) and are anchored to their respective membranes by their aminoterminal domains; it is not known how these two isozymes are sorted to their different locations. Both proteins appear to be myristoylated at their amino-terminal glycine residue (Borgese and Longhi, 1990), but it is not obvious how this modification could contribute to differential sorting. The two yeast NADH-cytochrome bs reductases described here are not myristoylated, as their amino-terminal sequences could be readily determined by Edman degradation. Yeast contains additional NADH-cytochrome bs reductases. One is probably associated with the endoplasmic reticulum (Kubota, 1977) whereas a recently discovered novobiocin-binding yeast protein of unknown location may be a soluble NADH-cytochrome bs reductase (Csukai et al., 1994). The functions of the two mitochondrial NADH-cytochrome bs reductases remain unknown. By analogy to mammalian cells, the outer membrane form may mediate the reduction of outer membrane cytochrome bg (Borgese and Longhi, 1990). The soluble intermembrane space form may transfer electrons from external NADH to cytochrome c, thereby mediating an antimycin-insensitive, energy-coupled oxidation of external NADH by yeast mitochondria (De Mantis and Melandri, 1984). This reaction has usually been attributed to a cytochrome c shuttle between NADH-cytochrome bs reductase in the outer membrane and the cytochrome clcytochrome oxidase system in the

fl Figure 8. Model for the Import lsoforms of NADH-Cytochrome

and Sorting of the Yeast bs Reductase

intermediate

Mitochondrial

See text for further explanation. Abbreviations: OMM, outer mitochondrial membrane; IMS, intermembrane space; IMM, inner mitochondrial membrane; MA, matrix; lmpl, inner membrane protease 1; fs2858, yeast mutant with temperaturesensitive inner membrane protease 1; and Y, membrane potential.

inner one, but the soluble NADH-cytochrome bs reductase isozyme described here is a much more plausible mediator of this reaction. The Sorting Mechanism The two NADH-cytochrome bg reductases of yeast mitochondria are the first known isozymes that are encoded by the same gene but transported to different locations within the same organelle. We propose that this sorting is achieved by a leaky stop-transfer of the primary translation product in the mitochondrial outer membrane. According to this model (Figure B), the gene MCR7 encodes a precursor protein (Mcrl p) with an amino-terminal signalanchor sequence for the mitochondrial outer membrane. This precursor interacts with the mitochondrial protein import machinery in the outer membrane and then follows 1 of 2 divergent pathways. About one-third of the precursor molecules are released from the import site into the outer membrane and become integral proteins of this membrane (Mcrl~~~). This pathway, like that of other outer membrane proteins, requires neither an electrochemical potential across the inner membrane nor proteolytic modification of the precursor. The other two-thirds of the precursor molecules pass through the outer membrane, interact

Cell 636

with the inner membrane import machinery, are cleaved at the N41-E42 bond by inner membrane protease 1 on the outer face of the inner membrane, and are released into the soluble intermembrane space (Mcrlp=). This pathway, like that of several other intermembrane space proteins, requires a potential across the inner membrane. This model is consistent with all of the results reported here, including the sequence of the MCRl gene, the microsequence data with the two isoproteins, the results of mitochondrial subfractionation, the specific effect of the &2858 mutation, the stable steady-state levels of both isoforms, and the pulse-chase experiments in the absence and presence of uncouplers. Our present results do not exclude the possibility that the precursor passes through the matrix before being cleaved to Mcrlp3* on the outer face of the inner membrane. Such a pathway has been proposed for several intermembrane space proteins (Hart1 and Neupert, 1990). However, as our own studies strongly argue against the existence of such a pathway in yeast mitochondria (Glick et al., 1992), the model shown in Figure 8 does not consider it. Several aspects of the model are still tentative, however. We have not directly shown that the first 11 residues of the precursor function as a matrix-targeting signal or that the hydrophobic stretch starting at residue 12 functions as a stop-transfer sequence for the outer membrane. However, the extensive information on matrix-targeting signals (Roise and Schatz, 1988) and the similarity between the amino-terminal region of Mcrl p% and the corresponding regions of Mas70p (Hase et al., 1984) and cytochromeb*(Guiard, 1985; Beasleyet al., 1993) make these assumptions reasonable. It also remains to be shown directly that the molecules escaping stop-transfer in the outer membrane interact with the inner membrane. However, the obsenrations that an inner membrane potential is required for formation of Mcrlp3* strongly implies that the targeting signal of the precursor inserts into the inner membrane (Martin et al., 1991; Schatz, 1993). Why are only a fraction of the precursor molecules arrested in the outer membrane? It seems likely that partitioning between the two divergent sorting pathways reflects a balance between the stop-transfer function of the hydrophobic stretch starting at residue 12 and the interaction of the amino-terminal amphipathic matrix-targeting signal with the inner membrane. Mas70p and the Mcrlp% precursor have similar stop-transfer sequences but quite different amino-terminal matrix-targeting signals. The matrix-targeting signal of Mas70p is relatively weak (hydrophobic moment, 0.37), and all of the Mas7Op molecules are firmly arrested in the outer membrane. In contrast, the matrix-targeting signal of the Mcrlp precursor is much stronger (hydrophobic moment, 0.84) and thus might effectively compete with the stop-transfer sequence. There is ample evidence that the amino-proximal hydrophobic sequence of Mas70p functions as a stop-transfer signal (e.g., Hase et al., 1984, 1986; Nakai et al., 1989; McBride et al., 1992) and that the stop-transfer function is very sensitive to subtle modifications of that sequence. Even a deletion of both isoleucine(l2) and leucine(l3) can inacti-

vate this function (Hase et al., 1984), suggesting that stoptransfer does not merely reflect partitioning into a hydrophobic lipid phase but requires specific protein-protein interactions. Indeed, a mutation of the outer membrane protein Mspl p appears to weaken the stop-transfer signal of MasirOp, converting it into a sorting signal for the intermembrane space (Nakai et al., 1993). Stop-transfer signals specific for the mitochondrial inner membrane also appear to operate by means of protein-protein interactions (Jensen et al., 1992; Beasley et al., 1993; Schwarz et al., 1993). As these signals presumably have to compete with the pull exerted on the amino-terminal domain of the precursor by mitochondrial Hsp70 in the matrix, they may occasionally be leaky, causing a small fraction of precursor molecules to accumulate as dead-end sorting products in the matrix (Glick et al., 1992). In vitro, the effectiveness of a stop-transfer sequence is influenced by the experimental conditions (Spiess et al., 1989), further suggesting that the translocation arrest caused by these sequences may not always be absolute. The model shown in Figure 8 predicts that mutations weakening the matrix-targeting signal of Mcrlp should increase the relative amount of the larger isoform in vivo, whereas mutations weakening the stop-transfer signal should have the opposite effect. We are currently testing these predictions. Further study of this experimental system should enhance our understanding of the mechanism by which stop-transfer sequences are decoded and our understanding of protein translocation into mitochondria. ExperImental Yeast

Procedures

Strains

Mitochondriafor isolation of outer membranes and for subfractionation studies were isolated from the wild-type Saccharomyces cerevisiae strain D273-106 (MATa; American type Culture Collection 25657) (Daum 81 al., 1962). Transformation and genetic experiments were performed with the yeast strains JKRlOl (MATa we3 leu2 his4 ade2; Bibus et al., 1988) and YKB5 (MATa/a ura3/ura3 /eu2//eu2 his4/his4 ADE2/ade2 LySZWys2) (Baker et al., 1990); plasmids YEpl3 (Broach et al., 1979) and YEplacl61 (Gietz and Sugino, 1986) have been described. Separation of Mtochondrlal Outer Membrane Proteins by Anlon Exchange Chromatography Outer membranes were solubilized at 2-3 mg/ml in buffer A (50 mM Na+-HEPES buffer [pH 8.01, 20 mM NaCI, 2% octyl-polyoxyethylene) for 30 min at 4OC. After centrifugation at 100,000 x g for 30 min at 4OC, the supernatant was chromatographed on a Mono Q HR5/5 anion exchange column with the following: 10 ml of 0%-30% buffer B (same as buffer A, but containing 1 M NaCI); 2.5 ml of 30%-50% buffer B; and 2.5 ml of 100% buffer B (at a flow rate of 0.5 mllmin). Fractions were analyzed by SDS-PAGE.

Preparatlon of Mcrlp” for Antibody Productlon and Mlcrosequencing Outer membranes were separated by preparative SDS-PAGE, and the 34 kDa band was electroeluted. It was then either injected into rabbits (Daum et al., 1962), sequenced directly by Edman degradation (Hewick et al., 1961), or converted to tryptic fragments for partial sequencing (Horst et al., 1993). Clonlng of the MCRl Gene Two degenerate oligonucleotide primers were designed based on the sequence of the two tryptic peptides of Mcrlp” shown in Figure 1A: bgr forward, sense, 5’-TTC/~ACVTTTAAJGJIICC#ACo/TGAIIGGA-3’,

One Gene 837

Encodes

Two Mitochondrial

lsoproteins

corresponding to nucleotides 210-235 in MCRI); bsr reverse, antisense, 5’-GGc/^cTclATAJ”GGJ~~AACd*AC-3’, corresponding to nucleotides 303-322 in MCR7. From each primer, 100 pmol was used to amplify the expected 1 IO bp fragment from IO ng of yeast genomic DNA by the polymerase chain reaction (PCR), under the following conditions: 1.25 mM dNTP; 2.5 U of Taq potymerase (Boehringer Mannheim); and 3.5 cycles of 30 sat 94OC, 2 min at 5VC, and 25 sat 72’C. Analysis by agarose gel electrophoresis showed one major band of about 110 bp, which was cloned into the pCRTHll cloning vector (lnvitrogen). Of 11 clones that were sequenced, only 1 contained the correct insert. This insert was purified by agarose gel electrophoresis, labeled with *P, and used to screen a libraryof large random fragments of yeast genomic DNA in the yeast-E. coli shuttle vector pFL1 (Chevalier et al., 1980). The positive clone pKH21 caused overexpression of a Mcrlp”cross-reactive protein in yeast, but it contained only a truncated gene that, for unknown reasons, encoded a protein lacking the amino-terminal 53 amino acids. A 359 bp fragment from this partial MCRI clone was amplified by PCR with the oligonucleotides bgr sequence 10 (5’-GCTCTAGAGCCATAGGATCGACTTG-3’, corresponding to nucleotides 157-l 87 in LfCR I), and bgr sequence 7 (S- GGTTGATACCGGTACCGGCA9’, corresponding to nucleotides 497-518 in MCRi). The sense oligonucleotide bsr.seql Ocontained an Xbal restriction site; the antisense oligonucleotide bsr.seq7 contained a Kpnl restriction site that is unique in the MCR7 gene. The 359 bp PCR product was labeled using ECL random primers (Amersham) and was used to screen a library of genomic DNA in the vector YEpI (Nasmyth and Tatchell, 1980). Three positive clones (pKH22, pKH23, pKH24) were tested for their ability to cause overexpression of Mcrlp” in yeast. The yeast transformant YKH3 (carrying clone pKH24) overexpressed two proteins of 34 kDa and 32 kDa and was selected for further study. The MCRl gene carried by this plasmid was sequenced (Sanger et al., 1977) step wise by using nucleotide sequence information to design additional oligonucleotide primers. Disruption of the MCRI Gene A 359 bp fragment, obtained from the partial MCR7 clone pKH21 by PCR with bsr.seq10 and bsr.seq7 (see above), was subcloned into the Xbal-Kpnl sites of pUCl9 (Yanisch-Perron et al., 1985) yielding plasmid pKHl0. A blunt-ended 1 .l kb fragment carrying the yeast URA3 gene was inserted into the unique, blunted BamHI siteof pKHl0, yielding plasmid pKHl1. A 1.48 kb Xbal-Kpnl fragment was purified and was used to transform the diploid yeast strain YKBB to uracil prototrophy. To verify disruption of the MCR7 gene, DNA from the uracil-prototrophicdiploid integrantwascutwith Hincll and was probed (Southern, 1975) with the 359 bp fragment from clone pKH21, which had been labeled using ECL random primers. Protease Treatment and Carbonate Extraction For extraction of proteins at alkaline pH (Fujiki et al., 1982) purified mitochondria (Glick et al., 1992) were incubated at 1 mglml in 100 mM N&CO3 (pH 11.5) for 30 min on ice and were spun at 100,000 x g at 4°C for 15 min. The pellet was resuspended in an equal volume of 100 mM N&CO3 (pH 11.5); proteins from supernatant and pellet were precipitated with 10% trichloroacetic acid and were analyzed by SDS-PAGE and immunoblotting. For protease treatment, purified mitochondria (50 ug) were treated for 25 min on ice with 0,50, or 500 &ml proteinase K, with or without 2% Triton X-100. Phenylmethylsulfonylfluoride (PMSF) was then added to 1 mM, and the mitochondria were reisolated and resuspended in 0.8 M sorbitol, 20 mM HEPES-KOH (pH 7.4) and 1 mM PMSF. Samples treated with 2% Triton X-100 were centrifuged for 10 min at 100,000 x g. All samples were treated with 5% trichloroacetic acid and 20% acetone for 5 min at 80°C and were incubated for 15 min on ice (Glick, 1991) reisolated by centrifugation, and analyzed by SDS-PAGE and immunoblotting. Digitonin Treatment Purified mitochondria 7.41) were treated with the supernatant was

of Mitochondria (5 mglml in 0.6 M sorbitol, 20 mM HEPES [pH digitonin (Glicket al., 1992) and centrifuged, and analyzed by SDS-PAGE and immunoblotting.

Purification of McrlpY for Amino-Terminal Sequencing Purified mitochondria from the yeast transformant YKH3 were washed

twice in 0.6 M sorbitol, 0.5 M KCI, and 20 mM K+-HEPES (pH 7.4) (high salt-BE). were resuspended in 0.8 M sorbitol and 20 mM K+HEPES (pH 7.4), and were diluted 18fold into 20 mM K+-HEPES (pH 7.4) and kept on ice for 30 min. Mitoplasts and the supernatant containing intermembrane space proteins were treated with 10% trichloroacetic acid for 5 min on ice and were subjected to SDS-PAGE. The gels were soaked for 5 min in transfer buffer (10 mM 3-cyclohexylamino-1-propane-sulfonic acid [pH 1 l] and 10% methanol); the Immobilon P membrane(Millipore)wassoakedfor 10s in 100% methanol and was then immersed in transfer buffer. After blotting for 1 hr at 1.3 A in a semidry blotting chamber with transfer buffer, the membranes were briefly rinsed with water and then with 100% methanol; were stained for 30-60 s with Coomassie blue R-250, 1% acetic acid, and 40% methanol; and were destained for l-2 min in 50% methanol by changing the solution several times. Finally, they were rinsed with water, and the 32 kDa band was excised and stored at -20°C until used. All solutions used for the purification of the 32 kDa band for microsequencing were HPLC grade.

Chase and Immunopreclpltatlon A 2.4 kb Hindlll fragment containing the MCRl gene was subcloned into the shuttle vector YEplaclEl, yielding plasmid pKH36. The yeast strain JKRlOl was transformed with pKH36 and was tested for overexpression of Mcrl p” and Mcrl p” by immunoblotting (Yaffe and Schatz, 1984). For the chase experiment with unlabeled cells, the transformed cells were grown at 30°C to the midlogarithmic phase (ODma, 0.5) (Daum et al., 1982) and were inhibited with 100 ug/ml cycloheximide; at the indicated time points, aliquots were analyzed for Mcrlp% and Mcrl p” by immunoblotting (Figure 7). For the pulse-chase experiment, the transformed cells were grown at 30°C to early logarithmic phase (ODeoo, 0.3) in synthetic medium containing 2% lactate, 0.1% glucose, 20 pglml adenine, and 20 ugl ml histidine (SLac-ade-his medium); were collected by centrifugation; and were resuspended to 0.1 g/ml in fresh SLacladelhis medium supplemented with 40 mM KPi (pH 6.0) in the presence or absence of 40 pLM carbonyl cyanide m-chlorophenyl hydrazone (CCCP) (Brandt, 1991). After 4 min at 30°C [%S]methionine was added to 50-100 uCi/ ml (1009 Cilmmol), and cells were labeled for 5 min. Labeling was stopped by 10 mM unlabeled methionine and 100 pglml cycloheximide and was followed by a chase at 30°C for the times indicated in Figure 6.Cellextracts(YaffeandSchatz, 1984)weresubjected toimmunoprecipitation with rabbit anti-Mcrlp” antiserum essentially as described (Manning-Krieg et al., 1991), and the immunoprecipitate was analyzed by SDS-PAGE and fluorography.

import into isolated Mitochondria A Pstl-EcoRI fragment encoding Mcrlp from pKH26 was cloned into pSP64 (Promega), yielding pKH32. Mcrlp as well as pSU9-DHFR were synthesized in vitro in the presence of [“Sjmethionine (Hurt et al.. 1984) and were imported into purified mitochondria (Glick et al., 1992). Translocation intermediates bound to the outer membrane were generated as described (Hines and Schatz, 1993) for 5 min at O°C with 200 ug of mitochondria in 200 ul of buffer A (0.6 M sorbitol, 25 mM HEPES-KOH [pH 7.41, 12 mM KCI, 5 mM MgC&, 0.5 mM EDTA) that contained 5 uM CCCP. The mitochondria were then chased for 1 O-30 min at 25’C in 200 ul of buffer A containing 10 mM dithiothreitol (to inactivate CCCP), 1 mM ATP, and 2 mM NADH. Valinomycin was then added to 0.1 uglml, and the mitochondria were reisolated and analyzed by SDS-PAGE and fluorography.

Miscellaneous Published methods were used for the transformation of yeast cells (Gietz et al., 1992) standard DNA procedures (Maniatis et al., 1982) SDS-PAGE (Daum et al., 1982) isolation of mitochondria (Daum et al., 1982) or outer membranes (Ramage et al., 1993) subcellular fractionation of yeast cells (Riezman et al., 1983) generation of mitoplasts (Glick et al., 1992), and immunoblotting (Haid und Suissa, 1983). Fluorograms were quantified using a computerized densitometer (Molecular Dynamics). Hydrophobic moments were calculated according to Eisenberg et al. (1984).

Cell 630

Acknowledgments The corresponding author for this paper is G. S. This study was supported by grants from the Swiss National Science Foundation (3-26169.69). from the United States Public Health Service (2 ROl GM37603), the Human Frontier Science Program Organization, and the Human Capital and Mobility Program of the European Community. We are indebted to Dr. P. Jend for help with peptide sequencing; Drs. 8. S. Glick, K. Hannavy, G. von Heijne, T. Lithgow, H. Riezman, and M. Spiess for critical comments on the manuscript; V. Zellweger for secretarial help; the members of our laboratory for helpful discussions; T. Junne and H. Briitsch fortechnical assistance; M. Jaggi, V. Grieder, and L. Miiller for the artwork; and Dr. R. Schekman (University of California, Berkeley) for the antiserum against the Kar2 protein. Received

June

15, 1994; revised

August

Garcia, P. D., Ou, J.-H., Rutter, W. J., and Walter, P. (1966). Targeting of the hepatitis B precore protein to the endoplasmic reticulum membrane: after signal peptide cleavage, translocation can be aborted and the product released into the cytoplasm. J. Cell Biol. 706, 1093-1104. Gietl, C. (1992). Malate dehydrogenase isoenzymes: cellular and role in the flow of metabolites between the cytoplasm organelles. Biochim. Biophys. Acta 7700, 217-234.

locations and cell

Gietz, R. D., and Sugino, A. (1966). New yeast-E. colishuttle constructed with in virro mutagenized yeast genes lacking pair restriction sites. Gene 74, 527-534.

vectors six-base

Gietz, R. D.. St. Jean, A., Woods, R. A., and Schiestl, R. H. (1992). Improved method for high efficiency transformation of intact yeast cells. Nucl. Acids Res. 20, 1425.

31, 1994.

References

Glick, B. S. (1991). Protein import Meth. Cell Biol. 34, 369-399.

Baker, K. P., Schaniel, A., Vestweber. D., and Schatz, G. (1990). A yeast mitochondrial outer membrane protein essential for protein import and Cell viability. Nature 348, 605-609. Beasley, E. M., Miller, S., and Schatz, G. (1993). The signal that sorts yeast cytochrome b2 to the mitochondrial intermembrane space contains three distinct functional regions. EMBO J. 12, 2303-2311. Behrens, M., Michaelis, G., and Pratje, E. (1991). Mitochondrial inner membrane protease I of Saccheromyces cerevisiae shows sequence homology to the Escherichia co/i leader peptidase. Mol. Gen. Genet. 228, 167-176. Bibus, C. R., Lemire, B. D., Suda, K., and Schatz, restoring import of a yeast mitochondrial protein presequence. J. Biol. Chem. 263, 13097-13102.

Fujiki, Y., Hubbard, A. L., Fowler, S., and Lazarow, P. B. (1962). Isolation of intracellular membranes by means of sodium carbonate treatment: application to endoplasmic reticulum. J. Cell Biol. 93, 97-102.

G. (1966). Mutations with a nonfunctional

yeast

mitochondria.

Glick, 8. S., Brandt, A., Cunningham, K., Miiller, S., Hallberg, R. L., and Schatz, G. (1992). Cytochrome c, and cytochrome b2 are sorted to the intermembrane space of yeast mitochondria by a stop-transfer mechanism. Cell 69, 609-622. Guiard, B. (1965). Structure, expression and regulation of a nuclear gene encoding a mitochondrial protein: the yeast L(+)-lactate cytochrome c oxidoreductase (cytochrome bZ), EMBO J. 4, 3265-3272. Haid, A., and Suissa, M. (1963). lmmunochemical identification membrane proteins after sodium dodecyl sulfate-polyacrylamide electrophoresis. Meth. Enzymol. 96, 192-205.

of gel

Ham, F. U., and Neupert, W. (1990). Protein sorting to mitochondria - evolutionary conservations of folding and assembly. Science 247, 930-936.

Boguta, M., Hunter, L. A., Shen, W.-C., Gillman, E. C., Martin, N. C., and Hopper, A. (1994). Subcellular localization of MOD5 proteins: map ping of sequences sufficient for targeting to mitochondria and demonstration that mitochondrial and nuclear isoforms comingle in the cytosol. Mol. Cell. Biol. 14, 2296-2306. Borgese, N., and Longhi, R. (1990). Both the outer mitochondrial membrane and the microsomal forms of cytochrome bs reductase contain covalently bound myristic acid -quantitative analysis on the polyvinylidene difluoride-immobilized proteins. Biochem. J. 266, 341-347. Borgese, N., and Pietrini, G. (1966). Distribution of the integral membrane protein NADH-cytochrome bs reductase in rat liver ceils, studied with a quantitative radioimmunoblotting assay. Biochem. J. 239,393403. Brandt, A. (1991). Pulse labeling of yeast cells as a tool to study mitochondrial protein import. Meth. Cell Biol. 34, 369-376. Broach, J. R., Strathern, J. N., and Hicks, J. B. (1979). Transformation in yeast: development of a hybrid cloning vector and isolation of the CAN7 gene. Gene 8, 121-133. Caplan, A. J., Tsai, J., Casey, P. J.. and Douglas, M. G. (1992). Farnesylation of YDJlp is required for function at elevated growth temperatures in Saccharomyces cerevisiae. J. Biol. Chem. 267,16690-16695. Chevalier, M.-R., Block, J. C., and Lacroute, F. (1960). Transcriptional and translation expression of a chimeric bacterial-yeast plasmid yeasts. Gene 77, 11-19.

into isolated

in

Csukai, M., Murray, M., and Orr, E. (1994). Isolation and complete sequence of CBR, a gene encoding a putative cytochrome b reductase in Saccharomyces cerevisiae. Eur. J. Biochem. 279, 441-446. Daum, G., Boehni, P. C., and Schatz, G. (1962). Import of proteins into mitochondria. Cytochrome b2 and cytochrome c peroxidase are located in the intermembrane space of yeast mitochondria. J. Biol. Chem. 257,13026-13033. De Santis, A., and Melandri, B. A. (1964). The oxidation of external NADH by an intermembrane electron transfer in mitochondria from the ubiquinone-deficient mutant E3-24 of Saccharomyces cerevisiae. Arch. Biochem. Biophys. 232, 354-365. Eisenberg, D., Schwarz, M., Komaromy, M., and Wall, Ft. (1964). Analysis of membrane and surface protein sequences with the hydrophobic moment plot. J. Mol. Biol. 778, 125-142.

Hase, T., Muller, U., Riezman, H., and Schatz, G. (1964). A 70-kd protein of the yeast mitochondrial outer membrane is targeted and anchored via its extreme amino terminus. EMBO J. 3, 3157-3164. Hase, T., Nakai, M., and Matsubara, H. (1966). The N-terminal 21 amino acids of a 70 kDa protein of the yeast mitochondrial outer membrane direct E. co/i 6galactosidase into the mitochondrial matrix space in yeast cells. FEBS Lett. 797, 199-203. Heim, R., Hug, M., Strehler, E. E., and Carafoli, E. (1992). Microdiversity of human plasma membrane calcium pump isoforms generated by alternative RNA splicing in the N-terminal coding region. Eur. J. Biochem. 205,333-340. Hewick, R. M., Hunkapillar, M. W., Hood, L. E., and Dreyer, W. J. (1961). A gas-liquid solid phase peptide and protein sequenator. J. Biol. Chem. 256, 7990-7997. Hildebrandt, A., and Estabrook, R. W. (1971). Evidence for the participation of cytochrome bs in hepatic microsomal mixed function oxidation reactions. Arch. Biochem. Biophys. 743, 66-79. Hines, V., and Schatz, G. (1993). Precursor binding to yeast mitochondria. A general role for the outer membrane protein Mas7Op. J. Biol. Chem. 268.449-464. Horst, M., Jend, P., Kronidou, N. G., Bolliger, L., Oppliger, W., Scherer, P., Manning-Krieg, U., Jascur, T., and Schatz, G. (1993). Protein import into yeast mitochondria: the inner membrane import site protein ISP45 is the MPl7 gene product. EMBO J. 72, 3035-3041. Hurt, E. C., Pesold-Hurt, B., and Schatz, G. (1964). Theamino-terminal region of an imported mitochondrial precursor polypeptide can direct cytoplasmic dihydrofolate reductase into the mitochondrial matrix. EMBO J. 3,3149-3156. Jensen, R. E., Schmidt, S., and Mark, R. J. (1992). Mutations in a 1Bamino acid hydrophobic region of the yeast cytochrome c, presequence prevent sorting to the intermembrane space. Mol. Cell. Biol. 12, 4677-4666. Kubota, S. (1977). tem of anaerobically of NADH-cytochrome

Studies on the microsomal electron-transport sysgrown yeast. IV. Purification and characterization bs reductase. J. Biochem. 87, 167-195.

Maniatis, T., Fritsch, E. F., and Sambrook, J. (1962). Molecular Cloning: A Laboratory Manual (Cold Spring Harbor, New York: Cold Spring Harbor Laboratory Press).

One Gene 839

Encodes

Two Mitochondrial

lsoproteins

Manning-Krieg. U. C., Scherer, P. E., and Schatz, G. (1991). Sequential action of mitochondrial chaperones in protein import into mitochondria. EMBO J. 10, 3273-3280.

Schneider, A., Behrens, M., Scherer, P., Pratje, E., Michaelis, G., and Schatz, G. (1991). Inner membrane protease I, an enzyme mediating intramitochondrial protein sorting in yeast. EMBO J. 10, 247-254.

Martin, J., Mahlke, K., and Pfanner, N. (1991). Role of an energized inner membrane in mitochondrial protein import. J. Biol. Chem. 266, 18051-18057.

Schwarz, E., Seytter, T., Guiard, B., and Neupert, W. (1993)Targeting of cytochrome& into the mitochondrial intermembrane space: specific recognition of the sorting signal. EMBO J. 12, 2295-2302.

McBride, H. M., Millar, D. G., Li, J.-M., and Shore, G. (1992). A signalanchor sequence selective for the mitochondrial outer membrane. J. Cell Biol. 179, 1451-1457.

Scott, E. M., and Griffith, I. V. (1959). The enzymic defect of hereditary methemoglobinemia: diaphorase. Biochim. Biophys. Acta 34, 584586.

Mall, T., Tebb, G., Surana, U., Robitsch, H., and Nasmyth, K. (1991). The role of phosphorylation and the CDC28 protein kinase in cell cycleregulated nuclear import of the S. cerevisiae transcription factor SW15 Cell 66, 743-758.

Simons, K., and Fuller, S. D. (1985). Annu. Rev. Cell Biol. 7, 243-248.

Nakai, M.. Harabayashi, M., Hase, T., and Matsubara, l-l. (1989). Protein sorting between outer and inner mitochondrial membranes: submitochondrial localization of cytochrome c, whose presequence is replaced by the amino-terminal region of a 70 kDa outer membrane protein. J. Biochem. 106, 181-187.

Spiess, M., Handschin, C., and activity of hydrophobic sequences J. Biol. Chem. 264, 19117-19124.

Southern, fragments

Nakai, M., Endo, T.. Hase, T., and Matsubara, H. (1993). Isolation and characterization of the yeast MSP7 gene which belongs to a novel family of putative ATPases. J. Biol. Chem. 268. 24262-24269. Nasmyth, K. A.. and Tatchell, K. (1980). The structure yeast mating type loci. Cell 79, 753-784. Nishino, H., and Ito, A. (1986). Subcellular chrome b-mediated NADH-semidehydroascorbate in rat liver. J. Biochem. 700, 1523-1531.

of transposable

Cell surface

polarity

in epithelia.

E. M. (1975). Detection of specific sequences among DNA separated by gel electrophoresis. J. Mol. Biol. 98,503-517. Baker, K. P. (1989). Stop transfer depends on the translation system.

Tomatsu, S., Kobayashi, Y., Fukumaki, Y.. Yubisui, T.. Orii, T.. and Sakaki, Y. (1989). The organization and the complete nucleotide sequence of the human NADH-cytochrome bs reductase gene. Gene 60, 353-381. Wessels. H. P., Hansen, G. H., Fuhrer, C., Look, A. T., Sjiistr6m, H., Noreen, O., and Spies% M. (1990). Aminopeptidase N isdirectlysorted to the apical domain in MDCK cells. J. Cell Biol. I7 7, 2923-2930.

distribution of OM cytoreductase activity

Yaffe, M. P., and Schatz, G. (1984). Two nuclear mutations that block mitochondrial protein import in yeast. Proc. Natl. Acad. Sci. USA 87, 4819-4823.

Oshino, N., Imai, Y., and Sato, R. (1971). A function of cytochrome bg in fatty acid desaturation by rat liver microsomes. J. Biochem. 69, 155-l 67.

Yanisch-Perron, C., Vieria, J., and Messing, J. (1985). Improved Ml3 phage cloning vectors and host strains: nucleotide sequences of the M13mp18 and pUCI9 vectors. Gene 33, 103-119.

Ozols, J., Korza, G., Heinemann, F. S., Hediger, M. A., and Strittmatter, P. (1985). Complete amino acid sequence of steer liver microsomal NADH-cytochrome bs reductase. J. Biol. Chem. 260, 11953-l 1981.

Yubisui, T., Miyata, T., Iwanaga, S., Tamura, M.. and Takeshita, M. (1966). Complete amino acid sequence of NADH-cytochrome bs reductase purified from human erythrocytes. J. Biochem. 99, 481-422.

Passon, P. G., and Hultquist, D. E. (1972). reductase from human erythrocytes. Biochim. 73.

Yubisui, T., Naitoh, Y., Zenno, S., Tamura, M., Takeshita, M.. and Sakaki, Y. (1987). Molecular cloning of cDNAs of human liver and placenta NADH-cytochrome bg reductase. Proc. Natl. Acad. Sci. USA 84, 3609-3613.

Soluble cytochrome bg Biophys. Acta 275,62-

Perlman, D., Raney, P., and Halvorson, H. 0. (1984). Cytoplasmic and secreted S. cerevisiae invertase mRNAs encoded by one gene can be differentially or coordinately regulated. Mol. Cell. Biol. 4,16821688. Pfanner, N., Tropschug, M., and Neupert, W. (1987). protein import: nucleoside triphosphates are involved import-competence to precursors, Cell 49, 815-823.

Mitochondrial in conferring

Pietrini. G., Aggujaro, D., Carrera, P., Malyszko, J., Vitale, A., and Borgese, N. (1992). A single mRNA, transcribed from an alternative, erythroid-specific, promoter, codes for two non-myristylated forms of NADH-cytochrome bs reductase. J. Cell Biol. 177, 975-986. Pratje, E., and Guiard, B. (1988). One nuclear gene controls the removal of transient pre-sequences from two yeast proteins: one encoded by the nuclear, the other by the mitochondrial genome. EMBO J. 5, 1312-1317. Prosser, spinach

I. M., and Lazarus, C. M. (1990). Nucleotide sequence nitrate reductase cDNA. Plant Mol. Biol. 75, 187-190.

of

Ramage, L., Junne, T., Hahne, K., Lithgow, T., and Schatz, G. (1993). Functional cooperation of mitochondrial protein import receptors in yeast. EMBO J. 12, 4115-4123. Reddy, V. V. R., Kupfer, D., and Caspi, E. (1977). Mechanism of C-5 double bond introduction in the biosynthesis of cholesterol by rat liver microsomes -evidence for the participation of microsomal cytochrome b5. J. Biol. Chem. 252, 2797-2801. Riezman, H., Hay, R., Gasser, S., Daum, G., Schneider, G., Witte, C., and Schatz, G. (1983). The outer membrane of yeast mitochondria: Isolation of outside-out sealed vesicles. EMBO J. 2, 1105-l 111. Raise, D., and Schatz, G. (1988). Chem. 263, 4509-4511.

Mitochondrial

presequences.

J. Biol.

Sanger, F., Niklen, S., and Coulson, A. R. (1977). DNA sequencing with chain terminating inhibitors. Proc. Natl. Acad. Sci. USA 74,54635467. Schatz, G. (1993). The protein tein Sci. 2, 141-146.

import

machinery

of mitochondria.

Pro-