N-terminal domain of the dual-targeted pea glutathione reductase signal peptide controls organellar targeting efficiency

Import of nuclear-encoded proteins into mitochondria and chloroplasts is generally organelle specific and its specificity depends on the N-terminal signal peptide. Yet, a group of proteins known as dual-targeted proteins have a targeting peptide capable of leading the mature protein to both organelles. We have investigated the domain structure of the dual-targeted pea glutathione reductase (GR) signal peptide by using N-terminal truncations. A mutant of the GR precursor (pGR) starting with the second methionine residue of the targeting peptide, pGRdelta2-4, directed import into both organelles, negating the possibility that dual import was controlled by the nature of the N terminus. The deletion of the 30 N-terminal residues (pGRdelta2-30) inhibited import efficiency into chloroplasts substantially and almost completely into mitochondria, whereas the removal of only 16 N-terminal amino acid residues (pGRdelta2-16) resulted in the strongly stimulated mitochondrial import without significantly affecting chloroplast import. Furthermore, N-terminal truncations of the signal peptide (pGRdelta2-16 and pGRdelta2-30) greatly stimulated the mitochondrial processing activity measured with the isolated processing peptidase. These results suggest a domain structure for the dual-targeting peptide of pGR and the existence of domains controlling organellar import efficiency therein.     

Processing of the dual targeted precursor protein of glutathione reductase in mitochondria and chloroplasts

Pea glutathione reductase (GR) is dually targeted to mitochondria and chloroplasts by means of an N-terminal signal peptide of 60 amino acid residues. After import, the signal peptide is cleaved off by the mitochondrial processing peptidase (MPP) in mitochondria and by the stromal processing peptidase (SPP) in chloroplasts. Here, we have investigated determinants for processing of the dual targeting signal peptide of GR by MPP and SPP to examine if there is separate or universal information recognised by both processing peptidases. Removal of 30 N-terminal amino acid residues of the signal peptide (GRDelta1-30) greatly stimulated processing activity by both MPP and SPP, whereas constructs with a deletion of an additional ten amino acid residues (GRDelta1-40) and deletion of 22 amino acid residues in the middle of the GR signal sequence (GRDelta30-52) could be cleaved by SPP but not by MPP. Numerous single mutations of amino acid residues in proximity of the cleavage site did not affect processing by SPP, whereas mutations within two amino acid residues on either side of the processing site had inhibitory effect on processing by MPP with a nearly complete inhibition for mutations at position -1. Mutation of positively charged residues in the C-terminal half of the GR targeting peptide inhibited processing by MPP but not by SPP. An inhibitory effect on SPP was detected only when double and triple mutations were introduced upstream of the cleavage site. These results indicate that: (i) recognition of processing site on a dual targeted GR precursor differs between MPP and SPP; (ii) the GR targeting signal has similar determinants for processing by MPP as signals targeting only to mitochondria; and (iii) processing by SPP shows a low level of sensitivity to single mutations on targeting peptide and likely involves recognition of the physiochemical properties of the sequence in the vicinity of cleavage rather than a requirement for specific amino acid residues.     

How can organellar protein N-terminal sequences be dual targeting signals? In silico analysis and mutagenesis approach

Organellar nuclear-encoded proteins can be mitochondrial, chloroplastic or localized in both mitochondria and chloroplasts. Most of the determinants for organellar targeting are localized in the N-terminal part of the proteins, which were therefore analyzed in Arabidopsis thaliana. The mitochondrial, chloroplastic and dual N-terminal sequences have an overall similar composition. However, Arg is rare in the first 20 residues of chloroplastic and dual sequences, and Ala is more frequent at position 2 of these two types of sequence as compared to mitochondrial sequences. According to these observations, mutations were performed in three dual targeted proteins and analyzed by in vitro import into isolated mitochondria and chloroplasts. First, experiments performed with wild-type proteins suggest that the binding of precursor proteins to mitochondria is highly efficient, whereas the import and processing steps are more efficient in chloroplasts. Moreover, different processing sites are recognized by the mitochondrial and chloroplastic processing peptidases. Second, the mutagenesis approach shows the positive role of Arg residues for enhancing mitochondrial import or processing, as expected by the in silico analysis. By contrast, mutations at position 2 have dramatic and unpredicted effects, either enhancing or completely abolishing import. This suggests that the nature of the second amino acid residue of the N-terminal sequence is essential for the import of dual targeted sequences.     

Dual targeting of organellar seryl-tRNA synthetase to maize mitochondria and chloroplasts

Aminoacyl-tRNA synthetases (AARSs) play a critical role in translation and are thus required in three plant protein-synthesizing compartments: cytosol, mitochondria and plastids. A systematic study had previously shown extensive sharing of organellar AARSs from Arabidopsis thaliana, mostly between mitochondria and chloroplasts. However, distribution of AARSs from monocot species, such as maize, has never been experimentally investigated. Here we demonstrate dual targeting of maize seryl-tRNA synthetase, SerZMo, into both mitochondria and chloroplasts using combination of complementary methods, including in vitro import assay, transient expression analysis of green fluorescent protein (GFP) fusions and immunodetection. We also show that SerZMo dual localization is established by the virtue of an ambiguous targeting peptide. Full-length SerZMo protein fused to GFP is targeted to chloroplast stromules, indicating that SerZMo protein performs its function in plastid stroma. The deletion mutant lacking N-terminal region of the ambiguous SerZMo targeting peptide was neither targeted into mitochondria nor chloroplasts, indicating the importance of this region in both mitochondrial and chloroplastic import.     

A novel in vitro system for simultaneous import of precursor proteins into mitochondria and chloroplasts

Most chloroplast and mitochondrial precursor proteins are targeted specifically to either chloroplasts or mitochondria. However, there is a group of proteins that are dual targeted to both organelles. We have developed a novel in vitro system for simultaneous import of precursor proteins into mitochondria and chloroplasts (dual import system). The mitochondrial precursor of alternative oxidase, AOX was specifically targeted only to mitochondria. The chloroplastic precursor of small subunit of pea ribulose bisphosphate carboxylase/oxygenase, Rubisco, was mistargeted to pea mitochondria in a single import system, but was imported only into chloroplasts in the dual import system. The dual targeted glutathione reductase GR precursor was targeted to both mitochondria and chloroplasts in both systems. The GR pre-sequence could support import of the mature Rubisco protein into mitochondria and chloroplasts in the single import system but only into chloroplasts in the dual import system. Although the GR pre-sequence could support import of the mature portion of the mitochondrial FAd subunit of the ATP synthase into mitochondria and chloroplasts, mature AOX protein was only imported into mitochondria under the control of the GR pre-sequence in both systems. These results show that the novel dual import system is superior to the single import system as it abolishes mistargeting of chloroplast precursors into pea mitochondria observed in a single organelle import system. The results clearly show that although the GR pre-sequence has dual targeting ability, this ability is dependent on the nature of the mature protein.     

Translation rate underpins specific targeting of N-terminal transmembrane proteins to mitochondria

Protein biogenesis is a complex process, and complexity is greatly increased in eukaryotic cells through specific targeting of proteins to different organelles. To direct targeting, organellar proteins carry an organelle-specific targeting signal for recognition by organelle-specific import machinery. However, the situation is confusing for transmembrane domain (TMD)-containing signal-anchored (SA) proteins of various organelles because TMDs function as an endoplasmic reticulum (ER) targeting signal. Although ER targeting of SA proteins is well understood, how they are targeted to mitochondria and chloroplasts remains elusive. Here, we investigated how the targeting specificity of SA proteins is determined for specific targeting to mitochondria and chloroplasts. Mitochondrial targeting requires multiple motifs around and within TMDs: a basic residue and an arginine-rich region flanking the N- and C-termini of TMDs, respectively, and an aromatic residue in the C-terminal side of the TMD that specify mitochondrial targeting in an additive manner. These motifs play a role in slowing down the elongation speed during translation, thereby ensuring mitochondrial targeting in a co-translational manner. By contrast, the absence of any of these motifs individually or together causes at varying degrees chloroplast targeting that occurs in a post-translational manner.     

The N‐Terminal Sequences of Four Major Hydrogenosomal Proteins Are Not Essential for Import into Hydrogenosomes of Trichomonas vaginalis

The human pathogen Trichomonas vaginalis harbors hydrogenosomes, organelles of mitochondrial origin that generate ATP through hydrogen‐producing fermentations. They contain neither genome nor translation machinery, but approximately 500 proteins that are imported from the cytosol. In contrast to well‐studied organelles like Saccharomyces mitochondria, very little is known about how proteins are transported across the two membranes enclosing the hydrogenosomal matrix. Recent studies indicate that—in addition to N‐terminal transit peptides—internal targeting signals might be more common in hydrogenosomes than in mitochondria. To further characterize the extent to which N‐terminal and internal motifs mediate hydrogenosomal protein targeting, we transfected Trichomonas with 24 hemagglutinin (HA) tag fusion constructs, encompassing 13 different hydrogenosomal and cytosolic proteins of the parasite. Hydrogenosomal targeting of these proteins was analyzed by subcellular fractionation and independently by immunofluorescent localization. The investigated proteins include some of the most abundant hydrogenosomal proteins, such as pyruvate ferredoxin oxidoreductase (PFO), which possesses an amino‐terminal targeting signal that is processed on import into hydrogenosomes, but is shown here not to be required for import into hydrogenosomes. Our results demonstrate that the deletion of N‐terminal signals of hydrogenosomal precursors generally has little, if any, influence upon import into hydrogenosomes. Although the necessary and sufficient signals for hydrogenosomal import recognition appear complex, targeting to the organelle is still highly specific, as demonstrated by the finding that six HA‐tagged glycolytic enzymes, highly expressed under the same promoter as other constructs studied here, localized exclusively to the cytosol and did not associate with hydrogenosomes.     

Targeting capacity and conservation of PreP homologues localization in mitochondria of different species

Mitochondrial presequences and other unstructured peptides are degraded inside mitochondria by presequence proteases (PrePs) identified in Arabidopsis thaliana (AtPreP), humans (hPreP), and yeast (Cym1/Mop112). The presequences of A. thaliana and human PreP are predicted to consist of 85 and 29 amino acids, respectively, whereas the Saccharomyces cerevisiae Cym1/Mop112 presequence contains only 7 residues. These differences may explain the reported targeting of homologous proteins to different mitochondrial subcompartments. Here we have investigated the targeting capacity of the PreP homologues' presequences. We have produced fusion constructs containing N-terminal portions of AtPreP(1-125), hPreP(1-69), and Cym1(1-40) coupled to green fluorescent protein (GFP) and studied their import into isolated plant, mammalian, and yeast mitochondria, followed by mitochondrial subfractionation. Whereas the AtPreP presequence has the capacity to target GFP into the mitochondrial matrix of all three species, the hPreP presequence only targets GFP to the matrix of mammalian and yeast mitochondria. The Cym1/Mop112 presequence has an overall much weaker targeting capacity and only ensures mitochondrial sorting in its host species yeast. Revisiting the submitochondrial localization of Cym1 revealed that endogenous Cym1/Mop112 is localized to the matrix space, as has been previously reported for the plant and human homologues. Moreover, complementation studies in yeast show that native AtPreP restores the growth phenotype of yeast cells lacking Cym1, demonstrating functional conservation.     

Direct Targeting of Proteins from the Cytosol to Organelles: The ER versus Endosymbiotic Organelles

In eukaryotic cells consisting of many different types of organelles, targeting of organellar proteins is one of the most fundamental cellular processes. Proteins belonging to the endoplasmic reticulum (ER), chloroplasts and mitochondria are targeted individually from the cytosol to their cognate organelles. As the targeting to these organelles occurs in the cytosol during or after translation, the most crucial aspect is how specific targeting to these three organelles can be achieved without interfering with other targeting pathways. For these organelles, multiple mechanisms are used for targeting proteins, but the exact mechanism used depends on the type of protein and organelle, the location of targeting signals in the protein and the location of the protein in the organelle. In this review, we discuss the various mechanisms involved in protein targeting to the ER, chloroplasts and mitochondria, and how the targeting specificity is determined for these organelles in plant cells .     

The aconitase C-terminal domain is an independent dual targeting element

The tricarboxylic acid cycle enzyme aconitase in yeast is a single translation product, which is dual targeted and distributed between the mitochondria and the cytosol by a unique mechanism involving reverse translocation. There is limited understanding regarding the precise mechanism of reverse translocation across the mitochondrial membranes. Here, we examined the contribution of the mature part of aconitase to its dual targeting. We created a set of aconitase mutants harboring two kinds of alterations: (1) point mutations or very small deletions in conserved sites and (2) systematic large deletions. These mutants were screened for their localization by a α-complementation assay, which revealed that the aconitase fourth domain that is at the C-terminus (amino acids 517-778) is required for aconitase distribution. Moreover, fusion of this C-terminal domain to mitochondria-targeted passenger proteins such as dihydrofolate reductase and orotidine-5′-phosphate decarboxylase, conferred dual localization on them. These results indicate that the aconitase C-terminal domain is both necessary and sufficient for dual targeting, thereby functioning as an “independent signal”. In addition, the same C-terminal domain was shown to be necessary for aconitase efficient posttranslational import into mitochondria.     

Both metaxin and Tom20 together with two mitochondria-specific motifs support mitochondrial targeting of dual-targeting AtSufE1

Plant cells have two endosymbiotic organelles, chloroplasts, and mitochondria. These organelles perform specific functions that depend on organelle-specific proteins. The majority of chloroplast and mitochondrial proteins are specifically imported by the transit peptide and presequence, respectively. However, a significant number of proteins are also dually targeted to these two organelles. Currently, it is not fully understood how proteins are dually targeted to both chloroplasts and mitochondria. In this study, the mechanism underlying mitochondrial targeting of dual targeting AtSufE1 in Arabidopsis was elucidated. The N-terminal fragment containing 80 residues of AtSufE1 (AtSufE1N80) was sufficient to confer dual targeting of reporter protein, AtSufE1N80:GFP, in protoplasts. Two sequence motifs, two arginine residues at 15^th and 21^st positions, and amino acid (aa) sequence motif AKTLLLRPLK from the 31^st to 40^th aa position, were responsible for targeting to mitochondria a portion of reporter proteins amid the chloroplast targeting. The sequence motif PSEVPFRRT from the 41^st to 50^th aa position constitutes a common motif for targeting to both chloroplasts and mitochondria. For mitochondrial import of AtSufE1:N80, Metaxin played a critical role. In addition, BiFC and protein pull-down experiments showed that AtSufE1N80 specifically interacts with import receptors, Metaxin and Tom20. The interaction of AtSufE1N80 with Metaxin was required for the interaction with Tom20. Based on these results, we propose that mitochondrial targeting of dual-targeting AtSufE1 is mediated by both mitochondria-specific and common sequence motifs in the signal sequence through the interaction with import receptors, Metaxin and Tom20, in a successive manner.     

Trypanosome Alternative Oxidase Possesses both an N-Terminal and Internal Mitochondrial Targeting Signal

Recognition of mitochondrial targeting signals (MTS) by receptor translocases of outer and inner membranes of mitochondria is one of the prerequisites for import of nucleus-encoded proteins into this organelle. The MTS for a majority of trypanosomatid mitochondrial proteins have not been well defined. Here we analyzed the targeting signal for trypanosome alternative oxidase (TAO), which functions as the sole terminal oxidase in the infective form of Trypanosoma brucei. Deleting the first 10 of 24 amino acids predicted to be the classical N-terminal MTS of TAO did not affect its import into mitochondria in vitro. Furthermore, ectopically expressed TAO was targeted to mitochondria in both forms of the parasite even after deletion of first 40 amino acid residues. However, deletion of more than 20 amino acid residues from the N terminus reduced the efficiency of import. These data suggest that besides an N-terminal MTS, TAO possesses an internal mitochondrial targeting signal. In addition, both the N-terminal MTS and the mature TAO protein were able to target a cytosolic protein, dihydrofolate reductase (DHFR), to a T. brucei mitochondrion. Further analysis identified a cryptic internal MTS of TAO, located within amino acid residues 115 to 146, which was fully capable of targeting DHFR to mitochondria. The internal signal was more efficient than the N-terminal MTS for import of this heterologous protein. Together, these results show that TAO possesses a cleavable N-terminal MTS as well as an internal MTS and that these signals act together for efficient import of TAO into mitochondria.     

The role of the N-terminal domain of chloroplast targeting peptides in organellar protein import and miss-sorting

We have analysed 385 mitochondrial and 567 chloroplastic signal sequences of proteins found in the organellar proteomes of Arabidopsis thaliana. Despite overall similarities, the first 16 residues of transit peptides differ remarkably. To test the hypothesis that the N-terminally truncated transit peptides would redirect chloroplastic precursor proteins to mitochondria, we studied import of the N-terminal deletion mutants of ELIP, PetC and Lhcb2.1. The results show that the deletion mutants were neither imported into chloroplasts nor miss-targeted to mitochondria in vitro and in vivo, showing that the entire transit peptide is necessary for correct targeting as well as miss-sorting.     

Subcellular localization of rice histone deacetylases in organelles

Histone deacetylases (HDACs) are known to function in the nucleus. Here, we report on the organellar localization of three rice HDACs, OsSIR2b, OsHDAC6, and OsHDAC10. The 35S:OsSIR2b-GFP and 35S:OsHDAC10-GFP constructs were introduced into tobacco BY2 cells. Co-localization analysis of the green fluorescent protein and MitoTracker fluorescent signals in the transformed BY2 cells indicated that OsSIR2b and OsHDAC10 are localized in the mitochondria. Transgenic Arabidopsis lines harboring 35S:OsHDAC6-GFP and 35S:OsHDAC10-GFP constructs were similarly analyzed, revealing that OsHDAC6-GFP is localized exclusively in chloroplasts, whereas OsHDAC10-GFP is localized in both mitochondria and chloroplasts. The presence of OsHDAC6-GFP and OsHDAC10-GFP in chloroplasts was verified by immunodetection.     

Role of the Negative Charges in the Cytosolic Domain of TOM22 in the Import of Precursor Proteins into Mitochondria

TOM22 is an essential mitochondrial outer membrane protein required for the import of precursor proteins into the organelles. The amino-terminal 84 amino acids of TOM22 extend into the cytosol and include 19 negatively and 6 positively charged residues. This region of the protein is thought to interact with positively charged presequences on mitochondrial preproteins, presumably via electrostatic interactions. We constructed a series of mutant derivatives of TOM22 in which 2 to 15 of the negatively charged residues in the cytosolic domain were changed to their corresponding amido forms. The mutant constructs were transformed into a sheltered Neurospora crassa heterokaryon bearing a tom22::hygromycin R disruption in one nucleus. All constructs restored viability to the disruption-carrying nucleus and gave rise to homokaryotic strains containing mutant tom22 alleles. Isolated mitochondria from three representative mutant strains, including the mutant carrying 15 neutralized residues (strain 861), imported precursor proteins at efficiencies comparable to those for wild-type organelles. Precursor binding studies with mitochondrial outer membrane vesicles from several of the mutant strains, including strain 861, revealed only slight differences from binding to wild-type vesicles. Deletion mutants lacking portions of the negatively charged region of TOM22 can also restore viability to the disruption-containing nucleus, but mutants lacking the entire region cannot. Taken together, these data suggest that an abundance of negative charges in the cytosolic domain of TOM22 is not essential for the binding or import of mitochondrial precursor proteins; however, other features in the domain are required.     

The <Emphasis Type="Italic">Arabidopsis</Emphasis> calmodulin-like proteins AtCML30 and AtCML3 are targeted to mitochondria and peroxisomes,respectively

Calmodulin (CaM) is a ubiquitous sensor/transducer of calcium signals in eukaryotic organisms. While CaM mediated calcium regulation of cytosolic processes is well established, there is growing evidence for the inclusion of organelles such as chloroplasts, mitochondria and peroxisomes into the calcium/calmodulin regulation network. A number of CaM-binding proteins have been identified in these organelles and processes such as protein import into chloroplasts and mitochondria have been shown to be governed by CaM regulation. What have been missing to date are the mediators of this regulation since no CaM or calmodulin-like protein (CML) has been identified in any of these organelles. Here we show that two Arabidopsis CMLs, AtCML3 and AtCML30, are localized in peroxisomes and mitochondria, respectively. AtCML3 is targeted via an unusual C-terminal PTS1-like tripeptide while AtCML30 utilizes an N-terminal, non-cleavable transit peptide. Both proteins possess the typical structure of CaMs, with two pairs of EF-hand motifs separated by a short linker domain. They furthermore display common characteristics, such as calcium-dependent alteration of gel mobility and calcium-dependent exposure of a hydrophobic surface. This indicates that they can function in a similar manner as canonical CaMs. The presence of close homologues to AtCML3 and AtCML30 in other plants further indicates that organellar targeting of these CMLs is not a specific feature of Arabidopsis. The identification of peroxisomal and mitochondrial CMLs is an important step in the understanding how these organelles are integrated into the cellular calcium/calmodulin signaling pathways.     

Studies on the topology of the protein import channel in relation to the plant mitochondrial processing peptidase integrated into the cytochrome bc1 complex

The mitochondrial processing peptidase (MPP) specifically cleaves N-terminal targeting signals from hundreds of nuclear-encoded, matrix-targeted precursor proteins. In contrast to yeast and mammals, the plant MPP is an integral component of the respiratory cytochrome bc1 complex. The topology of the protein import channel in relation to MPP/bc1 in plants was studied using chimeric precursors containing truncated cytochrome b2 (cyt b2) proteins of 55-167 residues in length, fused to dihydrofolate reductase (DHFR). The DHFR domain could be tightly folded by methotrexate (MTX), generating translocation intermediates trapped in the import channel with only the cyt b2 pre-sequence/mature domain protruding into the matrix. Spinach and soybean mitochondria imported and processed unfolded precursors. MTX-folded intermediates were not processed in spinach but the longest (1-167) MTX-folded cyt b2-DHFR construct was processed in soybean, while yeast mitochondria successfully processed even shorter MTX-folded constructs. The MTX-folded precursors were cleaved with high efficiency by purified spinach MPP/bc1 complex. We interpret these results as indicating that the protein import channel is located distantly from the MPP/bc1 complex in plants, and that there is no link between protein translocation and protein processing.     

Chloroplast Import Signals: The Length Requirement for Translocation In Vitro and In Vivo

Protein translocation of cytosolically synthesized proteins requires signals for both targeting of precursor proteins to the surface of the respective compartment and their transfer across its membrane. In contrast to signals for peroxisomal and endoplasmic reticulum translocation, the signals for mitochondrial and chloroplast transport are less well defined with respect to length and amino acid requirements. To study the properties of signals required for translocation into chloroplasts in vitro and in vivo, we used fusion proteins composed of transit peptides and the Ig-like module of the muscle protein titin as passenger. We observed that about 60 amino acids—longer than the transit peptide length of many experimentally confirmed chloroplast proteins—are required for efficient translocation. However, within native chloroplast precursor proteins with transit peptides shorter than 60 amino acids, extension appears to be present as they are efficiently imported into organelles. In addition, the interaction of an unfolded polypeptide stretch of 60 or more amino acids with receptors at the chloroplast surface results in the unidirectionality of protein translocation into chloroplasts even in the presence of a competing C-terminal peroxisomal targeting signal. These findings prove the existing ideas that initial targeting is defined by the N-terminal signal and that the C-terminal signal is sensed only subsequently.     

Dual targeting and function of a protease in mitochondria and chloroplasts

Here we show, using the green fluorescent protein (GFP) fusion system, that an Arabidopsis thaliana zinc-metalloprotease (AtZn-MP) is targeted to both mitochondria and chloroplasts. A deletion mutant lacking the amino-terminal 28 residues, with translation initiation at the second methionine residue, was imported into chloroplasts only. However, a mutated form of the full-length targeting peptide, in which the second methionine residue is changed to leucine, was imported to both organelles. No GFP fluorescence was detected when a frame-shift mutation was introduced between the first and second ATG codons of the Zn-MP–GFP construct, suggesting no alternative translational initiation. Our results show that the dual targeting of the Zn-MP is due to an ambiguous targeting peptide. Furthermore, we show that the recombinant AtZn-MP degrades mitochondrial and chloroplastic targeting peptides, indicating its function as a signal peptide degrading protease in both mitochondria and chloroplasts.     

Sorting of precursor proteins between isolated spinach leaf mitochondria and chloroplasts

The precursors of the F₁-ATPase -subunits fromNicotiana plumbaginifolia andNeurospora crassa were imported into isolated spinach (Spinacia oleracea L.) leaf mitochondria. Both F₁ precursors were imported and processed to mature size products. No import of the mitochondrial precursor proteins into isolated intact spinach chloroplasts was seen. Moreover, the precursor of the 33 kDa protein of photosynthetic water-splitting enzyme was not imported into the leaf mitochondria. This study provides the first experimental report ofin vitro import of precursor proteins into plant mitochondria isolated from photosynthetic tissue and enables studies of protein sorting between mitochondria and chloroplasts in a system which is homologous with respect to organelles. The results suggest a high organellar specificity in the plant cell for the cytoplasmically synthesized precursor proteins.