Protein Transport into Mitochondria Is Conserved between Plant and Yeast Species*

Protein targeting into plant mitochondria was investigated by in vitro translocation experiments. The precursor of the mitochondrial F1-ATPase beta subunit from Nicotiana plumbaginifolia was synthesized in vitro, translocated to, processed, and assembled in purified Vicia faba mitochondria. Transport (but not binding) required a membrane potential and external nucleotides and was conserved among plant species. beta subunit precursors from the yeasts Saccharomyces cerevisiae and Schizosaccharomyces pombe were imported and correctly processed in plant mitochondria. This translocation used protease-sensitive components of the outer membrane. Conversely, the N. plumbaginifolia beta subunit precursor was efficiently translocated and cleaved in yeast mitochondria. However, a precursor for a chloroplast protein was not targeted to plant or yeast mitochondria. We conclude that the machinery for protein import into mitochondria is specific and conserved in plant and yeast organisms. These results are discussed in the context of a poly- or monophyletic origin of mitochondria.

We conclude that the machinery for protein import into mitochondria is specific and conserved in plant and yeast organisms. These results are discussed in the context of a poly-or monophyletic origin of mitochondria.
Plant mitochondria and chloroplasts contain their own genome which has a limited coding capacity. Most of mitochondrial and chloroplast proteins are encoded in the nucleus, synthesized as precursors on cytoplasmic ribosomes, and subsequently transported into organelles. Precursors generally contain an amino-terminal extension which directs the protein to the target membrane and is removed after import into the organelle.
The chloroplast protein import has been better studied than the targeting process into plant mitochondria (for review, see Keegstra et al., 1989;Weisbeek at al., 1989). However, the mechanism of mitochondrial import has been well characterized in fungal (Saccharomyces cerevisiae and Neurospora crassa) and mammalian cells (for review, see Rosenberg et al., 1987;Attardi and Schatz, 1988;Hart1 et al., 1989). The majority of mitochondrial targeting presequences share a positively charged amphiphilic structure (von Heijne, 1986) involved in binding to proteinaceous receptors on the mitochondrial surface (Pfaller et al., 1988). Translocation into mitochondria requires a membrane potential across the inner membrane (Schleyer et al., 1982;Gasser et al., 1982) and in addition ATP, possibly for the maintenance of precursors in an unfolded form (Pfanner et al., 1987;Chen and Douglas, 1987;Verner and Schatz, 1987). Once introduced in the mitochondrial matrix, presequences are removed by a processing machinery whose components have been identified (Hawlitschek et al., 1988;Witte et al., 1988;Yang et al., 1988). Processed proteins are finally directed to their submitochondrial compartment.
The presence of chloroplasts in plant cells may require a more stringent sorting machinery. This raises the question whether proteins are imported into plant mitochondria as in fungal and mammalian mitochondria. We have previously shown that the p subunit precursor of mitochondrial F1-ATPase from Nicotiana plumbaginifolia is synthesized as a larger precursor which is processed during or after mitochondrial uptake (Boutry et al., 1987a). Similar results were obtained with other precursors introduced in maize ( White and Scandalios, 1987), broad bean (Whelan et al., 1988), or pea (Unger et al., 1989) mitochondria. The precursor of the F1-ATPase p subunit from N. plumbaginifolia contains a presequence which is rich in basic and hydroxylated amino acids but poor in acidic residues (Boutry and Chua, 1985). It could possibly form a positively charged amphiphilic (Y helix (von Heijne, 1986). Its involvement in mitochondrial targeting was demonstrated by showing that 89 NHz-terminal amino acid residues of the precursor were capable of specifically targeting the passenger protein, bacterial chloramphenicol acetyltransferase, into mitochondria in transgenic plants (Boutry et al., 198713). The presequence length of the N. plumbaginifolia /3 was estimated to be 55 residues long from comparison with the sequence of the mature @ subunit of sweet potato (Kobayashi et al., 1986).
In the present work, we report in vitro experiments demonstrating that the 0 subunit precursor is correctly translocated in plant mitochondria, cleaved, and assembled in the F1-ATPase complex. This import process requires both energized membranes and external nucleoside triphosphates. We compared in vitro targeting of the 8 subunit from plant and yeast species into either plant or yeast mitochondria and found that the process was conserved between those organisms.

Mitochondrial
Protein Transport 16857 DNA Construction-The vector used for in vitro transcription was pTZl8 (United States Biochemical Corporation) which positions the Ti RNA polymerase promoter adjacent to the multilinker region. The EcoRI cDNA fragment containing the F,-ATPase fl subunit gene of N. plumbaginifolia (Boutry and Chua, 1985) was inserted into the EcoRI site of pTZ18U (pTZ-atpP-1). A SalI-Hind111 fragment containing the S. cereuisiae ATPL gene (Takeda et al., 1985) was inserted into the corresponding sites of the polylinker region of pTZ18U (pTZ-ATPZ-C). An AccI fragment containing the Schizosaccharomyces pombe atp2 gene' was filled in with the Klenow DNA polymerase and inserted into the SmaI site of pTZ18R. This fragment contains the whole coding sequence together with 96 nucleotides of the 5'-noncoding-transcribed sequence. Hind111 restriction analysis confirmed the fragment orientation (pTZ-ATPB-P). A XmnI fragment containing a N. plumbaginifolia gene for the chlorophyll a/b-binding protein* was inserted into the SmaI site of pTZ18R. This fragment contains the whole coding sequence with 114 nucleotides of the 5'-noncodingtranscribed sequence. Sac1 restriction analysis confirmed the fragment orientation (pTZ-ab).

In Vitro Transcription
Translation-The plasmids pTZ-ATP2, pTZ-ATPP-C, and pTZ-ab were linearized with HindIII. The plasmid pTZ-ATPZ-P was linearized with XbaI. The transcription reaction included in a 50-rl reaction: 40 mM Tris-HCl (pH 8.0), 15 mM MgClz, 10 mM dithiothreitol, 0.5 mM each of ATP, CTP, UTP, 0.5 mM GpppG, 50 units of RNasin, 300 rg/ml bovine serum albumin, 5 pg of linearized DNA, and 12 units of T? RNA polymerase. Following an incubation of 2 min at 37 "C to allow RNA capping, GTP was added to 0.1 mM and the mixture was incubated 30 min at 37 "C. The RNA was then purified through a l-ml Sephadex G-50 column equilibrated with 10 mM Tris-HCl (pH 7.5), 1 mM EDTA. Five ~1 of RNA were translated for 1 h at 30 "C in 44 ~1 of a nuclease-treated reticulocyte lysate supplemented with 100 &i of [35S]methionine.
In the case of chloroplast import, 5 ~1 of RNA were translated for 1 h at 30 "C in 35 ~1 of a wheat germ extract supplemented with 100 &i of [35S]methionine.
In Vitro Mitochondrial Import-Vicia faba mitochondria were isolated from dark-grown hypocotyls as previously outlined (Boutry et al., 1987b). Mitochondria were resuspended in suspension medium (0.4 M mannitol, 10 mM KH,POI, pH 7.5 (KOH)). Yeast mitochondria were isolated as described (Genga et al., 1986). In oitro import reactions were performed in 160 ~1 of a medium containing 250 mM mannitol, 20 mM Hepes3-KOH (pH 7.5), 50 mM KCl, 2 mM MgC12,l mM ATP, 1 mM KHzPOI, 1 mM dithiothreitol, 20 pM ADP, 10 mM malate, 5 pl of translation mix and 20-40 pg of mitochondria. The mixture was incubated for 30 min at 28 'C. Modifications of the reaction are indicated in the legend to figures. Proteinase K digestion (2.4 pg/ml), where indicated, was performed for 15 min at 0 "C. Mitochondria, after addition of 1 mM PMSF, were reisolated by pelleting through a l-ml mannitol cushion (0.6 M mannitol, 10 mM KH,PO, (pH 7.5) (KOH), 100 mM KCl, 1 mM PMSF) and washed in 0.5 ml of suspension medium supplemented with 1 mM PMSF. Import products were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis through 9-13% gradient gels as described (Laemmli, 1970). Proteins were then transferred to nitrocellulose membranes (Towbin et al., 1979) which were subsequently exposed to x-ray films. The volume of reticulocyte lysate loaded as a control on the gels was B-fold lower than the volume added in the import reaction.
In Vitro Chloroplast Import-Chloroplast isolation from pea leaf and import were as described by Cline et al. (1985).

F,-ATPase
Purification-Mitochondrial membrane fractions were obtained from purified mitochondria which were diluted (5 mg/ml) in hyposmotic buffer (4 mM Tris, 2 mM ATP, 1 mM EDTA (pH 8.0) (HCl)) and centrifuged at 100,000 X g for 1 h in a Kontron TST-60.4 rotor. The pellet was resuspended in the same buffer at a concentration of 10 mg/ml at room temperature, mixed in an Eppendorf tube with an equal volume of chloroform neutralized by 100 mM Tris (pH 8.0) (HCl). The two phases were mixed by vortexing for 20 s. The aqueous phase obtained after 5 min of centrifugation was centrifuged a second time to remove any remaining chloroform and followed by a 30-min centrifugation in the A-100 rotor of the Beckman Airfuge at 100,000 x g. The chloroform-solubilized fraction was layered onto a 3.8-ml linear 5-30% (w/v) sucrose gradient in 4 mM Tris, 1 mM ' P. EDTA, 2 mM ATP (pH 8.0) (HCl) and centrifuged for 4 h in a Kontron TST-60.4 rotor at 50,000 rpm (260,000 x g) at 20 "C. Fractions (20 in total) were collected from the bottom, and 75-~1 aliquots were tested for ATPase activity (Pullman et al., 1960). The peak of ATPase activity normally sedimented in fractions 8-10 from the bottom.

RESULTS
The p Subunit Precursor Is Imported and Processed in Purified Mitochondria Isolated from Both Dicotyledon and Monocotyledon Species-A complete atp2-1 cDNA clone encoding the mitochondrial F,-ATPase /3 subunit of N. plumbaginifolia (Boutry and Chua, 1985) was placed under the control of a T7 polymerase promoter in the expression vector pTZ18. In uitro transcribed RNA was subsequently translated in a rabbit reticulocyte lysate into a /3 polypeptide precursor with the expected size of M, = 59,000 (Fig. LA, lane I). The labeled precursor was incubated with purified V. faba mitochondria in the presence of a respiratory substrate (malate) and ATP and found to be partly cleaved to a mature form (lane 2) which had the same electrophoretic mobility as the mature /3 subunit of a partially purified V. faba F1-ATPase FIG. 1. The F,-ATPase /3 subunit precursor is imported and processed into mitochondria (Mito) from both monocotyledon and dicotyledon species. Panel A, reticulocyte lysate containing radiolabeled F1-ATPase p precursor (lane I) was incubated with V. faba mitochondria as described under "Materials and Methods" (lanes 2-5). After reaction, the mixture was treated with proteinase K (lane 3, 0.6 pg/ml; lane 4, 1.2 rg/ml; lane 5, 3.1 fig/ml) for 15 min at 0 "C. Mitochondria were then washed and reisolated. The products were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis on a 9-13% polyacrylamide gel and transferred to a nitrocellulose membrane. The autoradiograph of the resulting membrane is shown (lanes 1-5). Precursor (p) and mature (m) forms are indicated. Lane 6, Amido-black staining of the partially purified V. faba F,-ATPase analyzed in the same gel. Panel B, radiolabeled F1-ATPase /3 precursor (lane 1) was incubated with Allium porrum mitochondria (lanes 2 and 3) and treated with 2.4 pg/ml of proteinase K (Rot. K) (lane 3).
(lane 6). Addition of increasing amounts of proteinase K after was probably associated with the F1-ATPase not totally assembled. Thus, once imported into the mitochondrial matrix, the p subunit precursor from N. plumbuginifoliu was cleaved and assembled into a functional F1-ATPase complex associated with the inner membrane. import did not degrade the processed form (lanes 3-5) which was thus located inside the organelle. On the contrary, the precursor as well as two unidentified translated products of lower apparent size were not incorporated into the organelle since they were degraded by the protease treatment (lanes 3-5). Interestingly, the ,L? precursor synthesized in a wheat germ extract was binding to mitochondria but was poorly translocated into the organelle (results not shown). Our import experiments were conducted with mitochondria isolated from V. fuba hypocotyls because this material facilitated the isolation of intact and functional mitochondria as indicated by their high respiratory controls (not shown). The import reaction was also observed with Nicotiunu tubucum mitochondria isolated from green leaves although with a much lower efficiency (Boutry et al., 1987a). This probably reflects the reduced respiratory activities of mitochondria obtained from green material. Finally, the @ precursor, obtained from a dicotyledon species, was also translocated and correctly processed in mitochondria isolated from Allium porrum, a monocotyledon species (Fig. 1B). This result indicates that plant import and processing machinery are well conserved in the plant kingdom.
The Imported and Processed /3 Subunit Is Assembled into F,-ATPuse Complex--We tested whether the mature fl subunit was assembled into an F1-ATPase complex. Mitochondria were lysed by osmotic shock after in uitro import of the labeled precursor. The F1-ATPase was solubilized by a chloroform treatment of the membrane fraction and purified by centrifugation on a sucrose gradient (Boutry et al., 1983). Electrophoretic analysis of the different fractions revealed the presence of an F1-ATPase in the fractions B-10 ( Fig. 2A) where the ATPase activity sedimented (not shown). The autoradiograph (Fig. 2B) shows that the mature /3 subunit peaked in the same fractions. An additional peak of mature p subunit was observed in the top fractions of the gradient and Reticulocyte lysate containing radiolabeled precursor protein was incubated with V. fuba mitochondria as described under "Materials and Methods." Mitochondria from 20 import reactions (-I1 mg) were used for FL-ATPase purification ("Materials and Methods"). The fractions collected from sucrose gradient were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Coomassie Blue staining of a gel obtained in a control experiment is shown in panel A. The fractions resulting from the import experiment are shown in panel B (autoradiograph). The ATPase activity sedimented in the fractions labeled with a dot. Stars indicate the F1-ATPase subunits: 01, 8, y, 6 and 6'.

Import of /3 Subunit Precursor into Plant Mitochondriu Requires Both Trunsmembrune
Potential and External Nucleoside Triphosphutes-The mitochondrial import process was found to be sensitive to reagents abolishing the membrane potential across the mitochondrial inner membrane. Indeed, the addition of 2.5 ELM antimycin A, an inhibitor of the respiratory chain, dramatically reduced the amount of imported protein (Fig. 3, lane 2). Addition of 1 pM of an oxidative phosphorylation uncoupler (carbonylcyanide m-chlorophenylhydrazone) or a K' ionophore (valinomycin) completely blocked the translocation of the /3 subunit precursor to mitochondria (lanes 3 and 4). Translocation of Mn-superoxide dismutase into maize mitochondria was also found to be sensitive to valinomycin (White and Scandalios, 1987). However, in the latter case, valinomycin also abolished binding of the precursor to mitochondria. This data contrasts with our results and those reported for other organisms.
In order to study nucleotide requirements, the translation mixture was passed through a Sephadex G-50 column to remove small molecules. In the absence of nucleotides, the import and cleavage of the /3 precursor was severely reduced (Fig. 4A, lane 1) or almost completely abolished (Fig. 4B, lane I). When increasing levels of ATP were added, import process resumed to a maximal efficiency at about 1 mM ATP (lanes 2-7). Requirement was not specific for ATP since the /3 precursor was translocated and cleaved to the mature form as well in the presence of 1 mM of GTP, CTP, or UTP (Fig. 4B). Thus, by the requirement of a membrane potential and external nucleoside triphosphates, protein translocation in plant mitochondria is similar to mitochondrial import in N. crussu (Schleyer et al., 1982;Pfanner and Neupert, 1986) and S. cereuisiue (Gasser et al, 1982;Chen and Douglas, 1987;Eilers et al., 1987).

Both Plant and Yeast p Precursors Are Trunslocuted into Mitochondriu
of Both Organisms-In order to evaluate whether the mitochondrial import machinery was conserved among various organisms, we compared the in vitro mitochondrial targeting of the precursor for the F,-ATPase /3 subunit from N. plumbuginifolia and two yeast species, S. cerevisiue (Takeda et al., 1985) and S. pombe.' Although the amino acid sequences of the three mature Reticulocyte lysate containing radiolabeled precursor protein was passed through a Sephadex G-50 column equilibrated with import buffer ("Materials and Methods") lacking ATP and ADP. The protein fraction was then added in import reactions containing V. faba mitochondria and nucleotides. Panel A, import reactions in the presence of increasing amounts of ATP. Lanes: I = 0 mM; 2 = 0.05 mM; 3 = 0.1 mM; 4 = 0.25 mM; 5 = 0.5 mM; 6 = 1 mM; 7 = 2 mM. Panel B, import reaction in the presence of different nucleotides. Lanes: 1 = without nucleotide; 2 = 1 mM ATP; 3 = 1 mM GTP, 4 = 1 mM CTP; 5 = 1 mM UTP. The products were analyzed as in Fig. 1 (Boutry and Chua, 1985), S. pombe' and S. cereuisiae (Vassarotti et al., 1987) are depicted. The beginning of the conserved sequences of the mature part is underlined. Arrows indicate the cleavage sites, determined by sequencing of the mature protein (S. pombe and S. cereukiae) or by comparison (iV. plumbaginifolia) with the sequence of the mature p subunit from sweet potato (Kobayashi et al., 1986). Charged residues are indicated (+ or -).
proteins share over 72% homology,' the amino-terminal extensions of the precursors are quite different in length and primary structure (Fig. 5). They exhibit, however, common features of mitochondrial presequences: they are rich in positively charged and hydroxylated residues, lack acidic amino acids, and are capable of forming an amphiphilic helix (von Heijne, 1986).
The three p subunit precursors were synthesized in vitro and incubated with V. faba mitochondria. Each precursor was partly processed to a mature form (Fig. 6A, lanes 2, 5, and 8) which was inaccessible to externally added protease (lanes 3, 6, and 9). The radiolabeled mature yeast p subunits possessed the same electrophoretic mobilities as the B subunits of the purified F1-ATPase from both yeast strains (lanes 6' and 9') indicating that in both cases, the cognate cleavage site was recognized by the plant matrix protease. However, it appeared that the S. cereuisae precursor import was less efficient (lanes 8 and 9). Reciprocally, plant and yeast precursors were incubated with isolated S. cereuisiae mitochondria. The precursors were all imported and cleaved to the correct mature form protected from external protease (Fig. 6B).
An early step in the mitochondrial import process is the interaction between precursors and protease sensitive components, later identified as receptors and which appear to be responsible for the specificity of the imported precursor (Gasser et al., 1982;Zwizinski et al., 1983;Pfaller et al., 1988). In our study, trypsin pretreatment of V. faba mitochondria reduced import of plant and yeast fl subunit precursors (Fig. 7) without modifying the integrity of mitochondria as monitored by respiratory controls (results not shown). Thus, the translocation of the plant and yeast /3 subunits in plant mitochondria requires a protease-sensitive component of the outer membrane. In both cases, a low residual translocation was observed. This was also the case for the N. crassa proteins translocation  and interpreted as a bypass: once receptors have been destroyed by protease treatment, import can still occur but with low efficiency . Bypassing showed little specificity since import of a chloroplast precursor into mitochondria (Hurt et al., 1986) uses this pathway . However, other chloroplast precursors were not found translocated in yeast mitochondria (Smeekens et al., 1987). In order to investigate this phenomenon in plant mitochondria, we performed import reactions with the light harvesting chlorophyll a/b-binding (Cab) precursor of N. plumbaginifolia.
This precursor contains a transit peptide which directs the protein into the thylakoid membranes (Schmidt et al., 1981). Indeed, the Cab precursor synthesized in a wheat germ extract was imported and processed in chloroplast isolated from pea (Fig. 8A). Different mature forms were observed as already reported for other species (discussed in Clark et al., 1990). However, when the Cab precursor was synthesized in a reticulocyte lysate followed by incubation with plant ( Fig. 8B) or yeast (Fig. 8C) mitochondria, no processed form was observed, and the precursor was completely degraded by external protease indicating that, in our experimental conditions, this chloroplast precursor was not imported into mitochondria. A similar result was obtained when the precursor was synthesized in a wheat germ extract and incubated with plant mitochondria (results now shown).

DISCUSSION
The B subunit precursor of mitochondrial F1-ATPase from N. plumbaginifoliu was imported, processed in purified V. faba mitochondria, and assembled into an F,-ATPase complex ( Figs. 1 and 2).
In addition to a membrane potential (Fig. 3), the transport of the F,-ATPase /3 subunit precursor into plant mitochondria requires ATP in the external medium (Fig. 4A). This is also the case for other organisms (for review, see Hart1 et al., 1989). Other nucleoside triphosphates could also sustain this process (Fig. 4B). This observation confirms those of Chen and Douglas (1987) who explained the lack of nucleotide specificity by the presence of a nucleoside diphosphokinase associated with the mitochondrial outer membrane and intermembrane space.
Several reports suggest that a function of ATP (or analogues) in mitochondrial protein import is to confer import competence to precursor proteins (Pfanner et al., 1987; Verner  (lanes 2, 3, 5, 6, 8, and 9). After reactlon, the mixtures were treated with 2.4 &ml protemase K (hot K) for 15 min at 0 "C (lanes 3, 6, and 9). The products were analyzed as in Fig. 1 and Chen and Douglas, 1987). This process possibly involves the 70-kDa heat shock protein family which could maintain mitochondrial proteins in an unfolded structure prior to their translocation across the membrane (Deshaies et al., 1988). In our conditions, a putative "unfoldase" could originate directly from the mitochondrial preparation or from the reticulocyte lysate. In this respect, it should be recalled that a fi precursor synthesized in a wheat germ extract is not translocated in mitochondria. However, it cannot be excluded that nucleoside triphosphates are required for further import steps. Indeed, it was recently shown that NTP is required in the matrix for protein translocation (Hwang and Schatz, 1989;Hart1 and Neupert, 1990). Moreover, in chloroplast, ATP is required internally for precursor binding and translocation (for review, see Keegstra et al., 1989 plumbaginifolia, S. cerevisiae, and S. pombe are quite different in length and primary structure (Fig. 5), they share the capacity to form an amphiphilic a-helix where positively charged and hydrophobic residues are on opposite sides. This common feature could be important for interaction with the import machinery since the three precursors are individually recognized and translocated by the plant and yeast machineries (Fig. 6). Import of both the S. pombe and N. plumbaginifolia /3 precursors in plant mitochondria requires a trypsin-sensitive protein probably located in the outer membrane (Fig. 7). This protein may be a receptor as identified in yeast (Vestweber et al., 1989) and in Neurospora (SolIner et al., 1989) or another enzyme involved in the early steps of translocation. Finally, the enzymatic machinery responsible for processing the precursor seems conserved since the yeast and plant fl precursors were correctly cleaved in heterologous systems. These observations are confirmed by in uivo experiments. A plant Mn-superoxide dismutase was efficiently imported and correctly processed by yeast mitochondria (Bowler et al., 1989) and inversely, the presequence of the mitochondrial tryptophanyl-tRNA-synthase from yeast imported the bacterial fi-glucuronidase into mitochondria of transgenic plants (Schmitz and Lonsdale, 1989).
These results have an important implication concerning the origin of mitochondria. The endosymbiotic origin of mitochondria (and chloroplasts) is now widely favored. However, the possibility of a multiple origin for mitochondria has been raised (Raven, 1970;Stewart and Mattox, 1984). For instance, the major differences in mitochondrial structure and expression between plant and other organisms together with rRNA sequence comparison support the hypothesis that mitochondria are at least partly of polyphyletic origin (i.e. that the mitochondrial genomes from plants and other organisms partly originated from distinct endosymbiotic events) (Gray et al., 1989). The uniqueness of the mitochondrial protein translocating system, which could have developed only after endosymbiosis of a bacteria-like progenitor of mitochondria, strongly supports a unique origin of mitochondria before fungi and plants diverged. However, this scenario does not exclude the possibility of an additional and more recent symbiotic event in the plant phylum leading to a transfer of genetic information (e.g. of rRNA) from the new endosymbiont to the established mitochondria (Gray et al., 1989). Alternatively, fusion of mitochondrial membranes with the plasma membranes of the new endosymbiote would have readily provided the latter with the receptors necessary for protein import from the cytosol and thus allowed a rapid evolution of the new nucleosymbiote interactions which eventually led to the disappearance of the original mitochondrion.
Although chloroplast and mitochondrial protein targeting systems share common properties (for review see Keegstra et al., 1989;Weisbeek et al., 1989), both processes are expected to be specific. However, a chloroplast presequence from Chlamydomonos reinhardtii was demonstrated to direct passenger proteins into yeast mitochondria (Hurt et al., 1986) but this import process occurred with low efficiency and seemed to bypass protease sensitive receptors . Interestingly, structure analysis of this and other presequences from C. reinhardtii indicated that they are more similar to mitochondrial than to chloroplast targeting presequences from higher plants (Franzen et al., 1990). In other experiments, several chloroplast presequences from higher plants did not import different proteins into yeast mitochondria (Smeekens et al., 1987). Moreover, no misrouting was observed in transgenic plant cells in which a mitochondrial or a chloroplast presequence from IV. plumbuginifolia specifically addressed the chloramphenicol acetyltransferase into mitochondria or chloroplasts, respectively (Boutry et al., 1987b).
Our in vitro experiments showed that the chloroplast precursor to the chlorophyll a/b-binding protein was not imported into plant or yeast mitochondria (Fig. 8). Thus, it appears that chloroplast and mitochondrial targeting presequences must contain distinct information required for specific addressing. Indeed, analysis of primary and predicted structures of chloroplast and mitochondrial targeting presquences indicated distinct properties for both types (von Heijne et al., 1989