Why are many mRNAs translated to the vicinity of mitochondria: A role in protein complex assembly?
Introduction
From 1972 to 1975 a series of experiments conducted by Butow and colleagues (Kellems and Butow, 1972, Kellems and Butow, 1974, Kellems et al., 1974, Kellems et al., 1975) lead to convincing evidence that a subclass of cytoplasmic ribosomes are bound to the surface of yeast mitochondria. A few years later, work in several laboratories showed that proteins destined to be imported into mitochondria are made as precursors which can be synthesized in vitro and taken up by isolated mitochondria in the absence of concomitant protein synthesis. While apparently contradictory, these two observations might suggest that some mRNAs or a certain fraction of a mRNA can be translated to the vicinity of mitochondria. However, after unambiguous confirmations of Butow's analyses, the widely accepted conclusion was that while mitochondria-bound polysomes may contribute to mitochondrial protein import, they do not appear to be obligatory for this process and it was clearly suggested that they only represent experimental artefacts (Suissa and Schatz, 1982). Recurrently, however, published experiments sustained diverse roles for post-translocation process in the mitochondrial import of cytoplasmic proteins. For instance, Fujiki and Verner (1993) presented evidence for tight coupling between protein synthesis and mitochondrial import and they concluded that, although precursors such as the beta-subunit of the F1F0-ATPase could be imported post-translationally, this was not its normal manner of import. Also the delivery of a large set of nascent polypeptides to the mitochondrial surface was shown to be an important problem for eukaryotic cells (Beddoe and Lithgow, 2002).
More recently, we demonstrated that overproduction of the karyopherin Pse1p/Kap121p improves the import of the hydrophobic protein Atm1p, an ABC transporter of the mitochondrial inner membrane and is correlated with an enrichment of the corresponding transcript in cytoplasmic ribosomes associated with mitochondria (Corral-Debrinski et al., 1999). This observation prompted us to reconsider the question of the localization of cytoplasmic ribosomes which translate mitochondrially localized proteins. We used genome-wide approaches to analyze the mRNA content of polysomes which copurify with mitochondrial fractions under the conditions previously described by Butow's group. Different types of microarray analyses reproducibly show that about 50% of the mRNAs coding for a mitochondrially localized product were primarily associated with polysomes that were bound to mitochondria. Also, it was consistently observed, by in vivo GFP-labeling experiments, that some mRNAs are actually localized to the vicinity of mitochondria, a process which was shown to be controlled, at least in part, by the 3′UTR region of the relevant mRNAs. MLR values, reflecting the observed Mitochondrial Localization of mRNAs, could be obtained for many yeast genes (Marc et al., 2002) and diverse searches for correlations between MLR and a variety of parameters could be undertaken. Surprisingly enough no correlation was found with amino-terminal recognition peptides or protein hydrophobicity. Instead a very strong positive correlation was found between the subcellular location of mitochondrial transcripts and the putative age of mitochondrial genes. Using a fairly stringent BLAST cut-off value, Karlberg et al. (2000) observed that only half of the considered 423 yeast mitochondrial proteins have homologues in prokaryotes, These genes that are believed to have derived from the endosymbionts (or the last common ancestor) have generally a high MLR value and are thus probably translated to the vicinity of mitochondrial outer membrane. On the basis of this correlation several hypotheses can be proposed; for instance, it is tempting to speculate that the polysomes that translate mitochondrial proteins of bacterial origin are faster and more prone to import abundant proteins, also one can imagine that the correct assembly of complex mitochondrial structures may require a locus-specific import process dependent of a mitochondria-proximal translation site. In this work we will mainly scrutinize the second possibility by considering several features of genes which code for elements of complex mitochondrial structures.
Section snippets
The 3′UTR plays a critical role in the mitochondria-bound translation process: the case of ATP2
Atp2p and Atp3p are important subunits of ATP synthase F1 domain (Duvezin-Caubet et al., 2003). Using the RNA-labeling system previously developed (Beach et al., 1999) we could localize the ATP2-3′UTR fragment fused to CP binding sites, in the living cells, to the vicinity of mitochondria (Margeot et al., 2002). When we replaced the ATP2-3′UTR with the ADH1-3′UTR, this led to a strain which grows poorly in glycerol ((Margeot et al., 2002), Fig. 1). To better understand the phenotype of the
Mitochondria-bound polysomes: a well-defined class of genes is involved
After the pioneering work of Butow and colleagues (Kellems et al., 1974), microarray experiments and in vivo localization analyses (Marc et al., 2002, Sylvestre et al., 2003) showed that nearly half of the nuclear genes coding for mitochondrially localized products are translated to the vicinity of mitochondria. As far as ATP2 gene is concerned, this process is highly dependent on the properties of the 3′UTR region which is required for a correct assembly of the corresponding complex (Fig. 1).
Acknowledgments
We wish to thank K. Wolf, L. del Giudice and R. Butow for their encouragement to publish this work. We thank G. Dujardin and F. Devaux for their critical reading of this manuscript. The precious advice of our colleagues of the Transcriptome platform (http://www.transcriptome.ens.fr/sgdb/) were greatly appreciated. This work was supported by grants from ARC 3310.
References (36)
- et al.
ATP13, a nuclear gene of Saccharomyces cerevisiae essential for the expression of subunit 9 of the mitochondrial ATPase
FEBS Lett.
(1991) - et al.
Cytochrome oxidase in health and disease
Gene
(2002) - et al.
COX23, a homologue of COX17, is required for cytochrome oxidase assembly
J. Biol. Chem.
(2004) - et al.
Delivery of nascent polypeptides to the mitochondrial surface
Biochim. Biophys. Acta
(2002) - et al.
The mature AEP2 gene product of Saccharomyces cerevisiae, required for the expression of subunit 9 of ATP synthase, is a 58 kDa mitochondrial protein
FEBS Lett.
(1995) - et al.
Coupling of cytosolic protein synthesis and mitochondrial protein import in yeast. Evidence for cotranslational import in vivo
J. Biol. Chem.
(1993) - et al.
ATP22, a nuclear gene required for expression of the F0 sector of mitochondrial ATPase in Saccharomyces cerevisiae
J. Biol. Chem.
(2003) - et al.
Cytoplasmic-type 80 S ribosomes associated with yeast mitochondria. I. Evidence for ribosome binding sites on yeast mitochondria
J. Biol. Chem.
(1972) - et al.
Cytoplasmic type 80 S ribosomes associated with yeast mitochondria. 3. Changes in the amount of bound ribosomes in response to changes in metabolic state
J. Biol. Chem.
(1974) - et al.
Cytoplasmic type 80 S ribosomes associated with yeast mitochondria: II. Evidence for the association of cytoplasmic ribosomes with the outer mitochondrial membrane in situ
J. Biol. Chem.
(1974)