Biochimie (1996) 78, 518-529 Soci~t~franqaise de biochimie et biologie mol6culaire / Elsevier, Paris

Editing and import: Strategies for providing plant mitochondria with a complete set of functional transfer RNAs
A D i e t r i c h a, I S m a l l b, A C o s s e t a, J H Well a, L M a r 6 c h a l - D r o u a r d a
alnstitut de Biolbogie Mol6culaire des Plantes du CNRS, Universitd Louis-Pastelo; 12, rue du G~n~ral-Zimmo; F-67084 Strasbourg cedex; Station de Gdn~tique et d'Amdlioration des Plantes. INRA, Route de St-CyJ; 78026 Versailles cedex, France

(Received 24 April 1996; accepted 7 August 1996) Summary - - The recombinations and mutations that plant mitochondrial DNA has undergone during evolution have led to the inactivation or complete loss of a number of the 'native' transfer RNA genes deriving from the genome of the ancestral endosymbiont. Following sequence divergence in their genes, some native mitochondrial tRNAs are 'rescued' by editing, a post-transcriptional p.rocess which changes the RNA l"Ne primary sequence. According to in vitro studies with the native mitochondrial tRNA from potato and tRNA111S from larch, editing is required for efficient processing. Some of the native tRNA genes which have been inactivated or lost have been replaced by tRNA genes present in plastid DNA sequences acquired by the mitochondrial genome during evolution, which raises the problem of the transcriptional regulation of tRNA genes in plant mitochondria. Finally, tRNAs for which no gene is present in the mitochondrial genome are imported from the cytosol. This process is highly specific for certain tRNAs, and it has been suggested that the cognate aminoacyl-tRNAs]tnthetases may be responsible for this specificity. Indeed, a mutation which blocks recognition of the cytosolic Arabidopsis thaliana tRNA"*"*by the corresponding alanyltRNA synthetase also prevents mitochondrial import of this tRNA in transgenic plants. Conversely, no significant mitochondrial co-import of the normally cytosol-specific tRNAasp was detected in transgenic plants expressing the yeast cytosolic aspartyl-tRNA synthetase fused to a mitochondriai targeting sequence, suggesting that, although necessary, recognition by a cognate aminoacyl-tRNA synthetase might not be sufficient to allow tRNA import into plant mitochondria.

aminoacyl-tRNA synthetase / chloroplast / editing / evolution / import / mitochondria / plant / transfer RNA Introduction
Although recent data suggest that mitochondria have a monophyletic endosymbiotic origin Ill, the mitochondrial genome of higher plants is exceptionally large compared to that of other eukaryotes [2], It exhibits structural instability and consists of subgenomic molecules usually able to recombine via repeated sequences [3, 4]. The large size of plant mitochondrial genomes is apparently only partly due to a greater gene content and mainly due to interspersed non-coding sequences of varied and mostly unknown origin, as well as to insertion of large amounts of foreign DNA of plastid and nuclear origin [5, 61. Although most mitochondrial polypeptides are nuclearly encoded and impored, some ribosomal proteins and key polypeptides invol :ed in the enzymatic complexes of the respiratory chain are synthesized by the organelle translation machinery. Hence, plant mitochondria require a full set of functional transfer RNAs (tRNAs). Generally, mitochondrial genomes contain between 22 and 29 tRNA genes I7-10], sufficient to decode all codons given relaxed base-pairing rules (U:N wobble). In contrast, mitochondrial genomes of vascular higher plants contain only 10-! 2 'native' (ie deriving from the genome of the ancestral endosymbiotic bacterium) tRNA genes [11-13]. This is manifestly insufficient to provide for the needs of the mitochondrial translation s~,stem, and thus plant mitochondria rely on alternative sources of tRNAs. Some of the missing native genes have been replaced by functional tRNA genes carried by chloroplast DNA fragments inserted into the mitochondrial genome during the course of evolution (eg [ 14--16]). Taken together, the native tRNAs and these so-called 'chloroplast-like' tRNAs still do not account for the 20 amino acids found in proteins. Thus, a number of higher plant mitochondrial tRNAs are no longer encoded by the organellar genome, but are encoded in the nucleus and imported from the cytosol [17, 181. Finally, as a further peculiar feature, it recently appeared that, to be functional, some of the higher plant native mitochondrial tRNAs need to be modified by editing, a post-transcriptional process which introduces changes into the RNA sequence [19, 20]. These processes, which together provide plant mitochondria with a complete set of non-redundant functional tRNAs

Abbreviations: AlaRS, alanyl-tRNA synthetase; AspRS, aspartyltRNA synthetase; LysRS, lysyl-tRNA synthetase; RT-PCR, reverse transcription-polymerase chain reaction. Note: Larix x leptoeuropaea results from the crossing of Lari.: kaempferi and Larix decidua.

519 (the chloroplast-like and nucleus-encoded species do not have native counterparts recognizing the same codon), raise intriguing questions regarding evolution, and rely on so far poorly understood fundamental cellular mechanisms: specific RNA sequence conversion, transfer of genetic information between different compartments, regulation of mitochondrial gene expression, transport of nucleic acids through the double mitochondrial envelope. The present article reviews the progress made by our groups during the last few years in addressing these problems. It also includes our most recent experiments using transgenic plan~s to analyze mitochondriai tRNA import. Our studies on the three plant mitochondrial tRNA classes (native, chloroplast-like, nucleus-encoded) aim at a better understanding of both evolutionary events and functional processes and may potentially have practical applications. sion gene by RNase prolection to identify those plants '~ hM~ had integrated the AspRS fusion gene as well as the resistance marker.

Isolation d'mitochondria fi'mn transgenic planLs

Mflochondria from transgenic potato tubers and leaves were isolated on Percoll gradients [291 and submitted to an osmotic shock to disrupt the outer membrane. For that purpose, mitochondria were resuspended in 50 lal isotonic buffer and subsequently diluted in 650 ~1 sterile water. After 30 s, the osmotic pressure was readjusted using 650 tal of two times concentrated buffer. To remove residual protein contamination, mitoplasts were incubated for 30 min on ice in the presence of 0.25 mg/mL proteinase K and finally recovered by centrifugation (3 min at 8000 g) after addition of i mM phenylmethylsulfonylfluoride as a protease inhibitor. To remove residual RNA contamination, mitoplasts were incubated for 5 min at 23C in the presence of an RNase mixture (40 p.g/mL RNase A, 400 units/mL RNase TI, I unit/mL phosphodiesterase) and recovered by centrifugation.

Materials and methods

Isolation of ~ytoplasmic tRNAs and aminoacylatimt assays
For aminoacylation studies, total cytoplasmic tRNA was extracted from potato tubers by phenol extraction and recovered by ethanol precipitation 121]. Large RNAs were eliminated in the presence of l M NaC! and tRNAs were further purified by DEAE-cellulose chromatography 1221. The proportion of mitochondrial material in such preparations usually does not exceed 0.5%. Aminoacylation of total potato cytoplasmic tRNA was conducted at 37C in the presence of purified yeast aspartyl-tRNA synthetase (AspRS). The aminoacylation reaction mixture contained 50 mM Tris-HCl buffer (pH 7.5), 10 mM ATP, 15 mM MgCI_,,0.4 mM glutathione, 0.1 mg/mL bovine serum albumin and 25 ~tM L-I 3H lasparlic acid (Isotopchim, France) at 6.5 Ci/mmole. Pure yeast AspRS 1231 was a gili from Jean Gangloff and Gilbert Eriani. Enzyme activity control tests involved commercial yeast total IRNA.

Western blot analysis

For Western blot analysis, proteinase K-treated mitoplasts were lysed in a [0 mM Tris-HCl (pH 7.5), ! mM EDTA. 2% (v/v) Triton X-100 buffer. After addition of I volume of denaturing buffer 1301, samples were centrifuged for 15 min at 50 000 g and the supemarant was loaded on 12% (w/v) denaturing polyacrylamide gels [30]. Pure yeast AspRS was also loaded on the gels as a control, as well as molecular mass protein markers (Sigma). After migration, gels were electrotransferred onto nylon membranes (lmmobilon-P, Millipore) for 90 rain at 500 mA in 25 mM Tris, 192 mM glycine, 15% (v/v) methanol. Non-specific binding was blocked by incubation Ibr 2 h at 37C in 10 mM Tris-HCI (pH 7.5), 150 mM NaCI, 0.05% (v/v) Tween-20, 5% (w/v) dry milk and 5% (v/v) normal goat serum. Incubation with tile polyclonal rabbit antiserum against yeast AspRS 1311 (a gift from Jean Gangloff) was for 2 h at room temperature with a 1/10 000 (v/v) final dilution in the blocking medium. Binding of the primary antibody was revealed by chemical luminescence using a peroxidase-conjugated antiserum againsl rabbit lgGs and ECL reagents (Amersham).

DNA constrm'ts
The coding sequence tor the yeast cytosolic AspRS Ii'om the plasmid pTG908A70 1241 Ca gilt from Jean Gangloff~ was cloned in a translational fusion behind a sequence encoding the first 90 amino acids of the l~-subunit of the mitochondrial ATP synthase from Nicotiana plumbaginifidia 125 I. This region includes the mitochondrial targeting sequence. The expression of the fusion gene was controlled by the promoter from the class I patatin gene B33 1261 and the poly A signal from cauliflower mosaic virus, taken from the plasmid pABDI 1271, The vector used for these experiments was Bluescript KS+ (Stratagene).

Isolation of mitochondrial tRNAs

Total mitochondrial tRNA was extracted from RNase-digested mitoplasts by phenol extraction and recovered by ethanol precipitation as previously described [21J. For Northern blotting, mitochondrial tRNAs were fractionated by electrophoresis on 15% (w/v) polyacrylamide gels under denaturing conditions [321. Northern analysis was as described [33 I.

Results and discussion

Plant transformation
Sixty lag of the plasmid containing the AspRS fusion gene and 50 lag of pABD! (containing a neomycin phosphotransferase gene [27]) were used to transfoml 2 x 10" potato protoplasts (potato clone 2 i 7, a cross of Solanum tuberosum and Solamtm chacoense). Protoplast preparation, electroporation and regeneration of transgenic plants was as described by Masson et al 1281. Antibiotic resistant plants were screened for transcripts from the AspRS fuWe have addressed different questions raised by the complex genetic origin of plant mitochondrial tRNAs, including evolution of tRNA genes in the plant mitochondrial genolnes, editing anti tRNA function, regulation of the expression of both native and chloroplast-like tRNA genes in plant mitochondria, as well as tRNA targeting and import into plant mitochondria. As already obvious from previous

520 reports (eg [11,18]) and further shown by our recent work [34], important differences in the chloroplast-like and imported tRNA species appear between and within plant lineages. Thus, tRNA gene evolution justifies the choice of the piant/tRNA systems studied and is present as a backdrop to more functional aspects throughout the present work. The main sections are organized around the three classes of plant mitochondrial tRNAs, native, chloroplast-like and nucleusencoded, with emphasis on editing, control of gene expression and import mechanisms, respectively. Editing as a strategy to yield fimctional native mitochondrial tRNAs A first strategy developed by plant mitochondria to keep a complete tRNA set is to correct at the RNA level limited sequence divergence which has occurred in some of the native mitochondrial tRNA genes. This is achieved by editing, a widespread post-transcriptional process first identified in kinetoplastids from Trypanosoma brucei [35] and which changes the primary sequence of RNAs, as compared to that of the corresponding DNA templates (for reviews see eg [36--38]). It is mainly observed in mitochondria, to a lesser extent in chloroplasts and in a few cases in transcripts of nuclear genes. In higher plant mitochondria and in chloroplasts, RNA editing results generally in the conversion of cytidines to uridines. Although editing was first found almost exclusively in messenger RNAs, where it contributes to the conservation of functional proteins, editing events affecting ribosomal RNA, intron sequences and tRNAs have now been observed in various organism,,, (references in [201). As described below, it appears in particular that some higher plant native mitochondrial tRNAs would not be functional in the absence of editing. Potato and bean mitochondrial native tRNAm,,"and larch mimchondrial native tRNAHis By comparing the sequences of the bean (Phaseolus vul. garis) and potato (Solanum tuberosum) native mitochondrial tRNA Ph~ genes to those of the corresponding mature tRNAs, we demonstrated that in both dicotyledonous angiosperms (flowering plants) the C encoded at position 4 in the gene is converted into a U in the mature tRNA [391. This nucleotide change corrects the mismatched C4:A69 basepair which appears when folding the gene sequence into the cloverleaf structure. Aminoacylation of in vitro synthesized unedited and edited potato tRNAPhe transcripts in the presence of a plant mitochondrial enzymatic extract showed that editing has no significant influence on tRNA Paecharging with phenylalanine. However, RT-PCR amplifications of natural potato mitochondrial tRNAP~,e precursors and processing assays with in vitro synthesized precursors showed that Ca to U4 editing is a prerequisite for efficient processing of potato native mitochondrial tRNAPhe[20]. Indeed, no unedited tRNAPhe has been found in vivo at the detection level of purification and direct sequencing [39]. Based on sequence data, we speculated [20] that a U4:A6,~ Watson-Crick base-pairing is essential to maintain the cloverleaf folding of the tRNA part in the potato mitochondrial tRNA Phe precursor RNA and to avoid an alternate stem-loop folding involving the upstream region of the precursor which is unlikely to be recognized by RNAse E C4 to U4 editing also occurs in the mitochondrial tRNAP~eof oenothera (Oenothera berteriana), another dicotyledonous angiosperm, and is also required for processing [19, 40]. Similarly, folding the sequence of the native mitochondrial tRNA His gene expreszed in larch (Larix x leptoeuropaea, a gymnosperm; gymnosperms constitute a less evolved group of vascular plants than angiosperms and are mainly represented by conifers) into the cloverleaf structure yields three C:A mismatched base-pairs (C6:A67, CI2:A23 and A29:C41 using classical tRNA nucleefide numbering). Again sequencing of tRNA ni~ precursor molecules and in vitro processing assays demonstrated that all three C:A mispairings are edited into classical U:A base-pairs and that unedited tRNAHi~ precursors are not processed [41]. Native mitochondrial tRNA Hi~and tRNA Ph~are found in liverwort (Matz'hantia polymorpha, a bryophyte) [8]. In this non-vascular primitive plant, no editing has ever been detected and a U, instead of a C, is found at the gene level in all the positions mentioned above as editing sites in the larch tRNAHi.~ and potato tRNA Phe. Editing therefore appears to correct the sequence of some native plant mitochondrial tRNAs following sequence drift in their genes and thereby 'rescues' these tRNAs which would otherwise be non-functional. Potato mitochondrial nqtive tRNAn,' Originally, the 1erm 'RNA editing' referred to primary sequence changes in messenger RNAs leading to codon modification 1351. To define the limits of what should be called editing when tKNAs are concerned is much less obvious, as they undergo a ,~ariety of posHranscriptional modifications in addition to primary sequence changes. The C to U transitions in plant mitochondrial RNAs are likely to occur via a simple deamination reaction [42, 431, a process which is comparable to many other nucleotide modification reactions occurring in tRNAs. Also, many nucleotide modifications in tRNAs can be considered to be as site-specific as the primary sequence changes discussed above. It would therefore be conceivable to call 'tRNA editing' all nucleotide modifications with clear functional consequences. One post-transcriptional modification of the potato mitochondrial native tRNAne would obviously fall into such an extended view of editing. The mitochondrial gene corresponding to this tRNA contains a C residue at the first position of the anticodon and codes therefore for a methionine-specific (CAU) anticodon. The C is post-transcriptionally modified so that the mature tRNA contains a iysidine-like nucleotide (L*, a C residue modified with an unidentified derivative of the amino acid lysine) at that position and caL, be aminoacylated with isoleucine but not with

521 methionine [44]. As for E. coti tRNA ne (LAU) [45] the aminoacylation and codon specificity of the potato mitochondrial native tRNA ne (L*AU) is likely to rely on the ~editing' of C34 into L'34. This process might occur in all 'prokaryotic' systems, as a simitar tRNA ~le is not only present in bacteria and mitochondria of dicotyledonous angiosperms, but probably also in mitochondria of monocotyledonous angiosperms like wheat (Triticum aestivum) [151 and maize (Zea mays) [46], in mitochondria of the gymnosperm larch [34] and in chloroplasts [47]. Table I, The origin of tRNAs in different plant groups. N, native; C, chloroplast-like; I, imported; ?, unknown. * indicates IRNAs which are C-to-U edited. The data in this table are drawn moslly from [8, 85] (liverwort), [! 3, 151 (monocotyledonous plants), [11, 12, t8] (dicotyledonous plants) and [341 (larch, monocotyledonous plants, dicotyledonous plants).

IRNA
Ala Arg Asn Asp Cys Gin Glu Giy His
lie

Livepwort Larch Dicotyledonous Monocotyledonou plants s plants

N N
N N

I '~
9 N

I 1
C N

I !
C N

Replacement of native mitochondrial tRNA genes by chloroplast-derived tRNA genes

Expression of chloroplast-originating genes present on promiscuous plastid DNA fragments inserted into the mitochondrial genome in the course of evolution is a second strategy used by plant mitochondria to complement their tRNA set. The isolation from bean mitochondria of a m~.ture chloroplast-like tRNA'rrp clearly distinguishable frov~ its bean chloroplast counterpart [14] was the first direct evidence for the expression of a chloroplast-like tRNA gene in mitochondria. Chloroplast-like tRNA genes have been found in a number of different plants [48], showing tholethis phenomenon is general (table I). Although some of these genes may have inactivating mutations, most have remained identical or almost identical to their authentic chloroplast counterparts. One of the main problems raised by this situation is to understand how the transcription of these plastid genes is controlled in mitochondria.

N N N N N
N/I b

'~ N N 1 N*
N/I b

N* N N N C
N/I b

C N N I C (I~a
N/! ~

Leu Lys Met Phe Pro Ser Thr Trp Tyr Val

N N N N N N N/I f N N N

I I 9 I I l 1 C N I

I N N/Cc N* N N/Cd (l) e I C N I (N?) g

I N N/Cc C N N (C)d I C N I

Chloroplast-like tRNA eh,', tRNACy~'~ tRNA m~ and In contrast to dicotyledonour; plants like potato, bean or oenothera, the native tRNA eh~ gene is a pseudogene in mitochondria of the monocotyledonous plants wheat [49] and maize !161 (table I). When the sequences of these pseudogenes are lblded into the cloverleaf structure, a C4-Gt,9 base-pair is found instead of the C4-A69 mispairing deduced from the bean, potato and oenothera gene sequences. Processing of potato precursor tRNA Phe containing a C4-G69 base-pair, instead of the U4-A69 base-pair normally produced by editing, is only 40% efficient I20]. It can therefore be speculated that an A69 to G69 mutation appeared in the native tRNAPhe gene of monocotyledonous angiosperms after their divergence from dicotyledonous angiosperms, giving rise to a tRNA Phe transcript which could no longer be edited and was inefficiently processed. In a further evolutionary step, the transfer of the chloroplast tRNA ph~ gene into the mitochondrial genome of monocotyledonous plants [15] might have enabled efficient expression and processing of an alternative tRNAPh~, leading to further divergence of the native tRNA eh~ gene into a pseudogene. A similar sequence of events might have happened to the mitochondrial native tRNACy ~which is still expressed and, at least in oenothera, edited in dicotyledonous plants I ! 8, 40], but has been replaced by the plastid-derived tRNACy.~in the mitochondria

aWheat mitochondria import tRNAHis [34]. btRNAtle (L*AU) is a native tRNA in all plants [48], tRNA 1re (IAU) is imported [18, 34, 86]. CtRNAnVletis native and tRNA manetis chlorop)ast-like in monocotyledonous and dicotyledonous plants [48]. UMitochondria of wheat and dicotyledonous plants, except sunflower, express a chloroplast-liketRNAaer (GGA) 1481that is absent from maize [ 131. eSunflower mitochondria import tRNAs~er, unlike .mitochondriaof other dicotyledonous ~lants examined to date 1341.tAt least one nucleus-encoded tRNAT M is imported into liverwort mitochondria (Akashi K, ~,laa~camaK, personal communication),gPotatomitochondrial tRNAs ' hybridize to mitochondrial DNA [18].

of monocotyledonous plants [ 15, 16]. Moreover, a chloroplast-like tRNACys pseudogene seems to be present in the potato mitochondrial genome [18]. Finally, the native tRNA tti~ present in the bryophyte liverwort [8], and expressed and edited in the gymnosperm larch [41] is absent from angiosperms where it has been replaced in most cases by a chloroplast-like tRNA riis [ 18, 50, 51 ]. Again, liverwort appears to be the simplest system studied so far, as its mitochondria contain no plastid-derived tRNAs [8]. It must be stressed, however, that for other plant species the situation has evolved differently, as nuclearly-encoded tRNA TM is used in larch mitochondria [34], and tRNA His is imported from the cytosol into wheat mitochondria [41].

522

Transcriptional regulation of mitochondrial tRNA genes

The existence of both native and chloroplast-like tRNA genes in plant mitochondria makes their transcriptional regulation interesting. Moreover, whereas the coding sequences of plastid-derived tRNA genes have often remained almost unchanged since their integration into the mitochondrial DNA, various rearrangements and/or sequence alterations have frequently occurred in their flanking regions (fig 1). Not all of these plastid-derived tRNA genes are expressed. Although attempts have been made [52], no absolute rule can be proposed to predict from the general sequence features whether a chloroplast-like tRNA gene present in mitochondria will be expressed or not. We have isolated two identical potato mitochondrial chloroplast-like tRNAAs. genes which exhibit very different flanking regions in their near vicinity, highly similar to the corresponding authentic chloroplast regions in one case and unrelated to chloroplast sequences in the other case (Mar6chalDrouard L, Desprez T, Chavant L, Dietrich A, in preparation; accession numbers in the EMBL nucleotide sequence database X93575 and X93576). The presence of the tRNAAshgene copy still flanked by sequences of chloroplast type seems to be restricted to some Solanaceae. Similarly, although the chloroplast-like mitochondrial tRNAHi~ gene is identical or almost identical in potato (Mar~chal-Drouard L, Desprez I', Chavant L, Dietrich A, in preparation; accession number in the EMBL nucleotide sequence database X93577), oenothera [53], sunflower (Helianthus annuus, another dicotyledonous plant) [52] and maize [501, and the regions flanking this gene are highly similar between potato, oenothera and sunflower, the chloroplast sequences maintained in the vicinity of the maize mitochondrial tRNAHisgene are completely unrelated. As all these chloroplast-like tRNA a,,. and tRNAHi~genes are active, it seems that their expression in plant mitochondria is not dependent upon their immediate flanking sequences. In chloroplasts, expression of tRNA genes relies upon 'prokaryotic-like'-10 and -35 boxes, internal promoters, or co-transcription with other genes (for a review see eg [54]). Knowledge about promoter sequences of tRNA genes in plant mitochondria is quite sparse at the present stage. The transcription initiation site of the oenothera mitochondrial native tRNAOhe (GAA) gene, which is probably co-transcribed with tRNA ho (UGG), has been the best studied so far. This transcription initiation site, located in an upstream purine rich region containing the consensus motif CRTAaGaGA found in putative mitochondrial messenger RNA and ribosomal RNA promoters of dicotyledonous plants [55, 56], was shown to be active in a pea (Pisum sativum) mitochondrial in vitro transcription system [571. However, in potato mitochondria, the native tRNA s~ (GCU), tRNAPhe (GAA) and tRNA Pro (UGG) are co-transcribed and the above mentioned consensus motif is not present at the initiation site of this transcription unit [581. Thus, no clear picture emerges yet and it remains to be defined to what extent chloroplast-like tRNA genes are ex-

pressed in plant mitochondria thanks to appropriate promoter sequences or to insertion into already transcribed regions of the plant mitochondrial genomes. In this context, it is interesting to note that, in all angiosperm mitochondria studied so far, the chloroplast-like tRNATq' gene is always expressed; in contrast, a chloroplast-like tRNAPro gene, usually located about 150 nt upstream of this tRNArrp gene, has never been or is no longer expressed [59, 60].

Replacement of native mitochondrial tRNAs by import of nuclearly encoded species from the cytosol
The third strategy used by plant mitochondria to complement their tRNA set is to share some nucleus-encoded species with the cytosolic translation machinery. Beside the inability to find in plant mitochondrial genomes the minimal set of tRNA genes needed to support protein synthesis (even taking into account the chloroplast-like genes) (eg [11-13]), evidence for tRNA import from the cytosol came from the presence in plant mitochondria of 'cytosol-like' tRNAs hybridizing to the nuclear genome [15, 17, 18, 32, 61, 62]. It appeared from sequence analysis that some nuclearly encoded tRNAs partition between the cytosol and the mitochondria, as the only difference identified between the cytosolic and mitochondrial forms of a given tRNA was the post-transcriptional modification of Gi8 into Gm~s in tRNAs TM [17, 32, 61, 62]. Most of our efforts aim at understanding the selection and transport mechanisms involved in this still intriguing process.

Nuclearly encoded mitochondrial tRNAs

Direct proof for mitochondrial import of a nuclearly encoded tRNA was obtained in transgenic plants. The bean nuclear gene encoding tRNALeu (C*AA) [32] was introduced into potato. Specific probing showed that the corresponding heterologous tRNA was expressed and was indeed partitioning between the cytosolic and the mitochondrial fractions [331. By mutating the transgene, a four base-pair insertion could even be introduced into the anticodon of the tRNAL~"without compromising mitochondrial import [33]. Comparison of the mitochondrial tRNA populations of larch (gymnosperm), potato and sunflower (dicotyledonous angiosperms), as well as maize and wheat (monocotyledonous angiosperms) showed a number of differences within or between plant lineages in the identity of the imported species, sometimes between relatively closely related plants [34]. Some tRNAs, namely tRNAAJa,tRNAsLeu, tRNATM, turned out to be imported from the cytosol into the mitochondria in all plants tested (table I). Conversely, other tRNAs systematically appeared to be encoded by the mitochondrial DNA: tRNAAsp, tRNA oln, tRNA ~lu, tRNA ne (L*AU) and tRNATyr. At least some of these have peculiarities dating from their prokaryotic origins which might make them difficult to replace by typically eukaryotic nuclearly-encoded equivalents: tRNAGIn because glutaminyltRNACla derives from glutamyl-tRNAo'n (eg [63]) and

523

trnP

trnW
cp higher #ant mt m a i z e rat wheat mt sugar beet

not expressed

expressed

trnH
~.~.~.~.~.~.Z.~.~e.~
c p higher pMnt

.... i ~:, -

'

' lm~mmmllm~i.~iii.i.~ii...il

rat sunflower potato

C
~ .

trnN
" / ~ Z l ~ ~ , / ' 2 " 2 " 2 ' # ' # ' # ' 2 Z .... i

cp mgher pJa.t

~ ~ i i i

mm~~ll
~ ~

~
i

........................... W B ~ I I B I B I rnt ~unflower, : Oenothera, potato 1

~

trnY

i,.~

...... -

trnD

rnt wheat

rat potato2

Fig 1. Evolutionary aspects of chloroplast-like tRNA genes inserted into plant mitochondrial genomes. A. Organization of the chloroplast-like tRNAPro (trnP) and tRNATrp (trnW) gene clusters present in the mitochondrial genome of maize [87], wheat [60] and sugar beet (Beta vulgaris) [591, a.~ compared to the corresponding region of the chloroplast genomes. B. Organization of the region containing the chloroplast-like tRNAms gene (trnH) in the mitochondriai ge,aome c,f maize I50], sunflower [52] and potato (Mar6chal-Drouard L, Desprez T, Chavant L, Dietrich A, in preparation; accession number in the EMBL nucleotide sequence database X93577), as compared to the corresponding region of the chloroplast genomes. C. Organization of the region(s) containing the chloroplast-like tRNAAsh gene(s) (trnN) in the mitochondrial genome of wheat 188[, sunflower [89] and polato (M~r~chal-Drouard L. Desprez T, Chavant U Dietrich A, in preparation; accession numbers in the EMBL nucleotide sequence database X93575 and X93576), as compared to the corresponding region of the ~loroplast genomes, hi the case of potato mitochondria, two chloroplast-like trnN genes, numbered 1 and 2, have been identified. The tRNA lyr gene (traY), located downstream oftrnN in sunflower, oenothera and potato (trnN n) corresponds to a native tRNA species [90]. cp refers to the chloroplast genome, mt to the mitochondrial genome. Black boxes represent tRNA genes, hatched boxes correspond to chloroplast-like intergenic sequences and dotted boxes to native intergenic sequences. Arrows indicate the direction of transcription.

524 tRNAne (L*AU) because of its peculiar post-transcriptional ~nodification (see above). The observations made suggest that the ability to import each tRNA has been acquired independently at different times during the evolution of higher plants [34]. precursor to the mitochondrial LysRS, which is a distinct enzyme [72]. In Trypanosoma brucei, even heterologous tRNAs can be imported [73]. This suggests loose recognition rules of individual tRNAs by the import machinery in trypanosomatids, although binding to receptors in the mitochondrial membrane thanks to a defined sequence in the D loop has been proposed [69]. In yeast, selection for import is very strict, as only a single tRNA is imported [74]. In plants, mitochondrial tRNA import is also selective (eg [17, 18, 34]), implying specific recognition of the tRNAs to be imported by one or several factors of the transport pathway. To test whether aminoacyl-tRNA synthetases are involved in the recognition of tRNAs imported into plant mitochondria, we transformed plants with a heterologous gene encoding a tRNA which is no longer recognized by its cognate aminoacyl-tRNA synthetase and looked at the fate of the corresponding transcript.

General considerations on the plant mitochondrial tRNA import process We. have isolated a number of bean nuclear tRNA genes czdlng for species presumably imported into mitochondria (tRNATM and tRNA TM) or present only in the cytosol (tRNAr~o and tRNASer). Alignment of these tRNA gene sequences, of previously obtained tRNA sequences, as well as of sequences available in the literature, did not reveal any sequence or secondary structure motif which is conserved and characteristic for the nuclearly encoded tRNAs importe~ into plant mitochondria, either in the tRNAs themselves, or in the gene flanking sequences. Therefore, there is no indication that there exist particular characteristics in these tRNAs, or their precursors, which would be specific for mitochondrially imported tRNAs (Ramamonjisoa D, Kauffmann S, Choisne N, Green G, Small I, MarfchalDrouard L, Dietrich A, in preparation; accession numbers in the EMBL nucleotide sequence database X98179 to X98184). This lack of conserved features, together with the varied pattern of tRNAs imported in different plants (eg [341), suggests that each imported tRNA is recognized by a specific factor. It is difficult to believe that there is a specific import channel for each tRNA. A more plausible hypothesis is the involvementof specific protein carriers with mitochondriai targeting sequences that would direct the relevant cytosolic tRNAs to the protein import channel. The protein import apparatus has been implicated in tRNA import in yeast 1641, and DNA oligonucleotides cross-linked to peptide carriers can be imported into yeast and mammalian mitochondria [65, 66], The search for natural carriers has centered on proteins capable of tightly binding to and shielding tRNAs. Mitochondrial aminoacyl-tRNA synthetases were suggested as ideal candidates for such a role at a very early stage [67]. Recognition by a cognate aminoacyl-tRNA synthetase seems necessary but not sufficient for tRNA import into mitochondria in transgenic plants
The involvement of aminoacyl-tRNA synthetases in the tRNA import process has been tested, with contradictory results, for trypanosomatid and yeast mitochondria. Import of tRNAs into isolated Leishmania mitochondria does not require added protein factors [68, 69], and non-aminoacylatable tRNAs or precursor tRNAs can be mitochondrially imported in Leishmania and Trvpanosoma brucei [70, 711. In contrast, import of tRNAt-r., (CUU) into yeast mitochondria only occurs following aminoacylation by the cytosolic lysyl-tRNA synthetase (LysRS) and in the presence of the

Expression and mitochondrial import of tRNA ata variants in transgenic plants In all higher plants studied so far, the mitochondrial tRNAAla is a m;~Oearlyencoded species imported from the cytosol. In bacteria [75] and animals [76], the major identity motif in tRNA Ala recognized by alanyl-tRNA synthetase (AlaRS) is the G3:U70 wobble pair in the acceptor stem. Restoring normal Watson-Crick base-pairing by altering UT0to C70 abolishes binding of AIaRS to tRNAAlaand hence abolishes aminoacylation. We have shown that the same mutation has an identical effect on the cytosolic/mitochondrial tRNAAla from plants, both by in vitro aminoacylation assays, and by amber suppressor tests in vivo [77]. The wildtype (UT0) and the mutated (C70) Arabidopsis thaliana (a dicotyledonous angiosperm) tRNAala were expressed in transgenic tobacco (another dicotyledonous angiosperm) plants. Whereas both forms were present at a normal level in the cytosolic fraction, only the wild-type, aminoacylatable tRNAARawas recovered in the mitochondrial fraction [78]. A four-nucleotide insertion (TCGA) was also introduced into the anticodon loop of these U70/C70 tRNAA~a constructs, with no significant consequence on the aminoacylation properties of the corresponding tRNAs. Again, replacement of UT0with C70 led to the lack of the heterologous tRNAA~ain the mitochondrial fraction, whereas the insertion in the anticodon loop had no apparent effect on the partitioning between the two fractions [78]. Complementary in vitro experiments involving wild type or mutated tRNA Ala transcripts and tobacco mitochondrial extracts suggested that the absence of the G3:C70tRNAs in mitochondria of the transgenic plants is unlikely to be due to differential stability versus the G3:UT0tRNAs I781. It appears from these experiments that a single inactivating base change is sufficient to abolish both aminoacylation and mitochondrial import of tRNAAI',. This implies that interaction with AIaRS is a prerequisite for tRNA Ala import into plant mitochondria. The plant tRNAAlaimport mechanism is therefore likely to resemble that of tRNALy~ (CUU)

525 in yeast [72], apart from the fact that the plant cytosolic and mitochondrial AIaRS are encoded by the same gene [79]. Further attempts were carried out in which the G3:U?0 identity motif of tRNAA~a, an imported tRNA, was introduced into cytosolic IRNA Phe,a tRNA which is not imported into the mitochondria of the angiosperms studied so far. The phenylalanine to alanine aminoacylation identity switch worked reasonably well with the A thaliana cytosolic tRNA phe, as demonstrated by in vitro aminoacylation and in vivo transient expression tests [77]. Unfortunately, failure to detect expression of the corresponding gene constructs in transgenic tobacco plants has prevented us from establishing whether its new alanine identity enables this modified tRNA phe to be imported into mitochondria.
!

0,20

~
0
o ~

0,10

0,00 ~ ~

Expression of a chimeric mitochondrial precursor form of yeost cytosolic aspartyl-tRNA synthetase in transgenic plants As a general rule, mitochondrial aminoacyl-tRNA synthetases are all nuclearly encoded, but only the imported cytosolic tRNAs can be aminoacylated by the corresponding mitochondrial enzymes; non-imported cytosolic tRNAs are not recognized efficiently by the mitochondrial enzymes. We decided to try and set up a situation in which transgenic plants were provided with an additional aminoacyl-tRNA synthetase able both to be imported into mitochondria and to recognize a normally non-imported cytosolic tRNA species. The purpose was to determine whether this was sufficient to drive a new tRNA into mitochondria. We chose yeast aspartyl-tRNA synthetase (AspRS) and tRNAAsp for these experiments, because in mitochondria of both monocotyledonous and dicotyledonous plants, tRNAAsp is a native, mitochondrially encoded species [ 18, 80] and cytosolic potato tRNAAso is a good substrate for purified yeast cytosolic AspRS (fig 2). A translational fusion between the first 90 amino acids (including the mitochondrial targeting transit peptide) of the 13- subunit of the Nicotiana plumbaginifolia mitochondrial ATP synthase [25] and yeast cytosolic AspRS [24] was expressed in transgenic potato plants. Mitochondrial protein and tRNA fractions were prepared from both tubers and leaves of the transgenic lines. Protein fractions were tested for the presence of the yeast AspRS by Western blotting. A signal corresponding to a protein of the expected size (fig 3) was specifically obtained with mitochondrial extracts from plants carrying the yeast transgene, as compared to control plants. This band was taken to be proof of both expression and mitochondrial import of the yeast AspRS in transgenic potato. Transfer RNA fractions prepared from the same mitochondrial samples were then tested by Northern blo~ting for the presence of the cytosolic tRNAAsp using a labeled oligonucleotide probe derived from the sequence of the soybean nuclear gene encoding this tRNA [81 ]. Despite the reproducible presence of the yeast AspRS in the mitochondrial fractions, no significant co-import of the cytosolic tRNAAsp was ever detected.

10

20

30

Time (rain) Fig 2. Aminoacylation kinetics of total potato cytoplasmic tRNA (3 mg/mL) in the presence of a saturating amount of purified yeast AspRS (20 ~tg/mL).

mr (kDa)
97

67

AspRS 45
Fig 3. In vivo expression and mitochondrial import of yeast cytosolic AspRS in transgenic potato plants. Mitochondrial extracts from control (1) and transgenic (2) potato tubers were fractionated by polyacrylamide gel electrophoresis and analyzed by Western blotting with a polycional antiserum raised against yeast cytosolic AspRS. Samples corresponding to equivalent protein amounts (as determined according to [91]) were loaded. The migration of protein molecular mass markers is indicated. The position of purified yeast cytosolic AspRS is shown by an arrow. The apparent difference in size between purified yeast AspRS and the protein specifically detected by the antiserum against AspRS in mitochondrial extracts of ~ransgenic potato is likely to be due to extra amino acid residues located downstream of the processing site. as the miao chondrial presequence of the N plumbaginifolia FbATP synthase [~-subunit corresponds to 55 out of the 90 residues encoded by the upstream plant sequence in the chimeric transgene [92] and six amino acids are encoded by the linker sequence.

526 The outcome of these experiments implies that recognition by a mitochondrially imported aminoacyl-tRNA synthetase is not necessarily enough to promote efficient import of a tRNA. The yeast mitochondrial LysRS precursor alone is not sufficient to drive import of tRNAt-y~ into isolated yeast mitochondria [72]. However, it cannot be excluded at the present stage that the expression level of the yeast AspRS construct in our transgenic potato plants was too low to allow a detectable tRNAAsp import or that, in contrast to the normal form (',f yeast cytosolic AspRS, the chimeric precursor form of this enzyme introduced into the transgenic plants does not e~ficiently recognize the plant tRNAAsp.

] trnF ]
C4-.~ U4

polato

Native trnF
u4 Liverworl

expression

Larch

Conclusions and prospects

The different origins of tRNAr'he, tRNA Hi~and tRNACy~ in various plants illustrate well the strategies developed by plant mitochondria during evolution to maintain a complete tRNA set (fig 4). These tRNAs are native mitochondrial species needing no editing in the primitive, non-vascular plant liverwort, a bryophyte. In larch (gymnosperm) mitochondria, the native tRNA His gene is still expressed but sequence divergence makes it necessary to edit the tRNA. In both dicotyledonous and some monocotyledonous angiosperms, this native gene has been lost and tRNA HL, is obtained from a plastid-derived gene acquired by the mitochondrial genome. Finally, in the monocotyledonous angiosperm wheat, the mitochondrial genome does not code for a tRNAais and wheat mitochondria import this tRNA from the cytosoi. Similarly, following sequence divergence, the native tRNAPh is maintained in mitochondria of dicotyledonous plants thanks to editing, but has been replaced by a plastid-derived species in monocotyledonous plants and is imported from the cytosol in larch 1341. In the same way, depending o n the plant lineage, mitochondrial tRNACys corresponds to either an edited native, a chloroplast-like or perhaps a nuclearly encoded species (fig 4). Interestingly. although liverwort remains the root species in all cases, the trees based on the genetic origin and editing of mitochondrial tRNAPhe, tRNAHi~ and tRNACy~ do not correspond to the phyiogenetic tree of the different plant species considered. The phylogenetic tree would first separate gymnosperms (larch) from angiosperms and subsequently split angiosperms into dicotyledonous plants (bean, oenothera, potato, sunflower) and monocotyledonous plants (maize, wheat). It is evident from the available data that the mitochondrial tRNA species falling into each of the different categories (native, edited native, plastid-derived and imported from the cytosol) differ to some extent within and between plant lineages, However, there is a clear evolutionary tendency towards loss of native tRNA genes, which drop in number from 29 in liverwort [8] to 12 in petunia (Pettmta hybrida, another dicotyledonous plant) [111, 11 in sunC6, el2. C 4 ~ U6, ILl12,U411 Larch

Native trnH -J

U6,UI2,U41~ / iverworl

/ .,1,] Cp-like trnH [ 2 ,, sunflower,maize ~["-..~....,,,.,z,.~,.s" Potaio,

r-"

Loss o[ native gene expression

Wheal

I trnC ]
IC28~U2~

Native trnC
U28
Liverworl

Maize,wheat

Lossof hativegene
expression

~ .

3~"~NucleartrnC?"
!
Larch

Fig 4. Evolution of the gene~ (trnF, trnHand maC) coding for mitochondrialtRNAme, tRNArtls and tRNA~-ysas an illustrationof the strategiesdevelopedby plant mitochondriato keep a complete set of functionaltRNAs. Existenceof editing events,expressionof chloroplast-liketRNA genes and importof nucleus-encodedtRNA species are numbered !, 2 and 3 respectively.Cp-like stands for chloroplast-like. All relevantreferencesare cited in the text or in the legendsto table I and figure I,

flower [12] and 10 in maize Ii3] for instance. It is interesting to consider the order of events involved in the replacement of native tRNAs. As mitochondrial genomes are

527 present in multiple copies per cell, inactivation or loss of a tRNA gene from a proportion of the genomes is not immediately lethal to the cell, but this may subsequently create a selection pressure towards alternative modes of expression. Conversely, the mitochondria may first acquire the al~emafive tRNA (either imported or chloroplast-like) and only subsequently lose expression of the native tRNA. This latter sequence of events more closely resembles that inferred for the transfer of protein-coding genes to the nucleus [82]. The fact that replacement of native tRNAs by nuclearly encoded tRNAs differs in different plant lineages raises the problem of how the selectivity of the intport process c;n vary in this way. If aminoacyl-tRNA synthetases are the k,'y components for import specificity, then the mitochondrial precursors of these enzymes would be expected to differ between plants which import a given cytosofic tRNA into mitochondrla and those which do not. Extensive studies on plant aminoacyl-tRNA synthetases, still a poorly documented domain, will be of great help in answering these questions. Cloned aminoacyl-tRNA synthetase genes should also provide tools for producing mitochondrial precursor forms of these enzymes in large quantities. Such precursors might be required for the rcconstitution of tRNA import with isolated mitochondria. Deciphering the plant mitochondrial tRNA import mechanism may, in the long term, yield knowledge and molecular tools of practical benefit. The possibility of targeting RNAs transcribed from nuclear transgenes to mitochondria would be a way of transferring novel genetic information into these organelles. Genetic transformation of plant mitochondria has yet to be achieved. Defining the key components of mitochondrial tRNA import in plants and other organisms may also lay the foundations for the genetic treatment of patients suffering from cytopathies due to point mutations in mitochondrial tRNA genes, the rationale being nuclear expression, targeting and mitochondrial import of a functional tRNA. Although it is generally considered that nuclearly encoded tRNA species are not naturally imported into mammalian, and especially human, mitochondria, there is some evidence for the import of the RNA moiety of the mitochondrial RNA processing endonuclease (MRP RNase) (eg [83]) and of the HIV RNA [84], suggesting that the pathway potentially exists and might be exploited by supplementing human cells with the appropriate heterolognus components. In addition, recent experimental evidence implies import of a cytosolic tRNA into the mitochondria of marsupials (M0rl M, P~iabo S, personal communication).
T Desprez, AM Duch~ne-Louarn, G Green, S Kauflmann, D Ramamonjisoa, C Remacle, G Souciet, H Wmtz (Strasbourg) and K Akama, K Akashi, VTC Carneiro, N Choisne, R Kumar, D Lancelin, H Mireau (Versailles). Thanks are also due to P Keltz, P Klethi and R Wagner (IBMP, Strasbourg) for care of plants in the greenhouse. This work was funded by the tnstitut National'de la Recherche Agronomique (1NRA), the Cenlre National de la Recherche Scientifique (CNRS) and the Universit6 Louis Pasteur (ULP, Strasbourg).

References
I Burger G, Lang BE Reith M. Gray MW (1996) Genes encoding the same three subunits of respiratory complex II are present in the mitochondrial DNA of two phylogenetically distant eukaryotes. Proc Nail Acad Sci USA 93, 2328-2332 2 Qu6tier F (1995) Los g6nomes des organites: 30 ans de recherche. Bitiaur 146 (June), 12-16 3 Fauron C, Casper M. Gao Y, Moore B (1995) The maize mitochondrial genome: dynamic, yet functional. Trends Genet I 1,228-235 4 Bonen L, Brown GG (1993) Genetic plasticity and its consequences: perspectives on gone organization and expression in plant mitochondria. Can J Bot 71,645--660 5 Schuster W, Brennicke A (1987) Plastid, nuclear and reverse transcriptase sequences in the mitochondrial genome of Oenothera: is genetic information transferred between organdies via RNA? EMBO J 6. 2857-2863 6 Knoop V, Unseld M, Marienfeld J, Braladt P, Sunkel S, Ullrich H, Brennicke A (1996) Copia-, gypsy- and LINE-like retrotransposon fragments in the mitochondrial genome of Arabidopsis thaliana. Geneti~'s 142, 579-585 7 Anderson S, Bankier AT, Barrell BG, De Bruijn MHL, Coulson AR, Drouin J, Eperon IC, Nierlich DP. Roe BA. Sanger E Schreir PH, Smith AJH, Staden R. Young IG (1981) Sequence and organization of the human mitochondrial genome. Nature 290, 457--465 8 0 d a K, Yamato K, Ohta E, Nakamura Y, Takemura M, Nozato N, Akashi K, Ohyama K (i992) Transfer RNA genes in the mitochonthial genome from a liverwort, Marchamia polymorpha: The absence of chloroplast-like transfer RNAs. Nucleic Acids Res 20, 3773-3777 9 Tzagaloff A, Myers AM (1986) Genetics ~f mitochondrial biogenesis. Annu Roy Biochem 55,249-285 10 Leblanc C, Boyen C. Richard O, Bonnard G, Grienenberger JM, Kloareg B (1995) Complete sequence of the mitochondri:d DNA of the Rhodophyte Chondrus crispus (Gigartinales). Gone content and genome organization. J Mol Biol 250, 484-.495 I I Weber-Lotfi E Mar6chaI-Drouard L, Folkerts O, Hanson M, Grienenberger JM (1993) Localization of tRNA genes on the Petunia hvbrida 3704 mitochondrial genome. Plant Mol Biol 2 I. 403--407 12 Ceci L, Veronico P. Gallerani R (1996) Identification and mapping of tRNA genes on the Helianthus ammus mitochondrial genome. DNA Sequem'e, in press 13 SangarE A, Well JH, Grienenberger JM, Fauron C. Lonsdale D (1990) Localization and organization of IRNA genes on the mitochondrial genomes of fertile and male-sterile lines of maize. Mol Gen Genet 223, 224-232 14 Mar~chal L, Guillemaut P. Grienenberger JM, Jeannin G, Well JH (1985) Sequence and codon recog~itic,n of bean mitochondria and chloroplast tRNAs'rq': evidence for a high degree of homology. Nm'leic Acids Res 13,4411--4416 15 Joyt,e PBM, Gray MW (1989) Chh)roplast-like transfer RN?, genes expressed i.1 wheat mitochondria. Nt~,1oh A~ ids Rcs t7, 5461--5476 16 Wintz H, Grienenberger JM, Well 2H, Lonsdale DIvl (1988) Location and nucleotide sequence of two tRNA genes and a tRNA pseudo-gone in the maize mitochondrial genome: evidence for the transcription of a chloroplast gene in mitochondria. Curt Genet t3, 247-254

Acknowledgments
The authors are very grateful to J Gangloff ~md G Eriani lbr the gift of the yeast AspRS gene, pure yeast AspRS and an antiserum against yeast AspRS and to L WiUmitzer for the gift of the patatin promoter. We wish to thank our co-workers who are or have been involved in these studies on plant mitochondrial tRNAs: L Chavant,

528
17 Mar+chal-Drouard L, Weil JH, Guillemaut P (1988) Import of several tRNAs from the cytoplasm into the mitochondria in bean Phaseohts vulgaris. Nucleic Acids Res 16, 4777--4788 18 Mar+chaI-Drouard L, Guillemaut P. Cosset A, Arbogast M, Weber E Well JH, Dietrich A (1990) Transfer RNAs of potato (Solam+m tuberosum) mitochondria have different genetic origins. Nucleic Acids Res 18, 3689--3696 19 Marehfelder A, Brennicke A, Binder S (1996) RNA editing is required for efficient excision of tRNA (Phe) from precursors in plant mitochondria. JBiolChem 271, 1898-1903 20 Mar~chaI-Drouard L, Cosset A, Remacle C, Ramamonjisoa D. Dietrich A (1996) A single editing event is a prerequisite for efficient processing of potato mitochondrial phenylalanine tRNA. Mol Cell Bio/ 16, 3504-3511 21 Mar~chal-Orouard L, Small I, Weil J-H, Dietrich A (1995) Transfer RNA import into plant mitochondria. Methods Enzymol 260. 310-327 22 Guillemaut P. Weil JH (1990) Purification and identification of higher plant organellar transfer RNAs. In: Transfer RNAs and other soluble RNAs (Cherayil JD, ed) CRC Press. Boca Raton, 55-67 23 Eriani G, Pr6vost G, Kern D, Vincendon P, Dirheimer G. Gangloff J (1991) Cytoplasmic aspartyl-tRNA synthetase from Saccharomyces cerevisiae: Study of its functional organisation by deletion analysis. Eur J Biochem 200, 337-343 24 Sellami M, Fasio;o F, Dirheimer G, Ebel JP, Gangloff J (1986) Nucleotide sequence of the gene coding for yeast cytoplasmic aspartyl-tRNA synthetase CAPS); mapping of the 5' and 3' termini of AspRS mRNA. Nucleic Acids Res 14, 1657-1666 25 Boutry M, Nagy F. Poulsen C, Aoyagi K, Chua NH (1987) Targeting of bacterial chloramphenicol acetyltransferase to mitochondria in transgenic plants. Natttre 328, 340-342 26 Rocha-Sosa M, Sonnewald U, Frommer W. Stratmann M, Schell J, Willmitzer L (1989) Both developmental and metabolic signals activate the promoter of a class I patatin ~c.~. MBO ,I 8, 23-29 27 Paszkowski J, Pisan B, Schillito RD, Hohn T, Hohn B. Potrykus I (1984) Direct gene transfer to plants. EMBO J 3, 2717-2722 28 Masson J, Lancelin D, Bellini C, Lecerf M. Guerche P, Pelletier G (1989) Selwtion of somatic hybrids between diploid clones of potato (Solanum tuberosum L) transformed by direct gene transfer. TheorAppl Genet 78, 153-159 29 Neuburger M. Journct EP, Bligny R, Cardc JP. Douce R (1982) Purification of plant mitochondria by isopycnic centrifugation in density gradients of Percoll. AJ~'hBitwhem Biophys 217, 312-323 30 Lacntmli U K (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Natm'e 227, 680-685 31 P~vost G, Eriani G, Kern D, Dirheimer G, Gangloff J (1989) Study of the arrangement of the functional domains along the yeast cytoplasmic aspartyl-tRNA synthetase. Era' J Biochem 180, 351-358 32 Green G, Mar~chal L, Wci[ JH, Guillemaut P (1987) A Phaseolus vulgaris mitochondrial tRNA"cu is identical to its cytoplasmic counterpart: sequencing and in vivo transcription of the gene corresponding to the cytoplasmic tRNALcu. Plant Mol Biol I0, 13-19 33 Small !, Mar~chaI-Drouard L, Masson J, Pelletier G+ Cosset A, Well JH, Dietrich A (1992) In vivo intport of a ~:~rmal or mutagenized heterologous transfer RNA into the mitochondria of transgenic plants: Towards novel ways of influencing mitochondrial gene expression? EMBOJ I I, 1291-1296 34 Kumar R, Mar~chaI-Drouard L, Akama K, Small I (1996) Striking differences in mitochondrial tRNA import antong different plant species, Mol Gen Genet 251, in press 35 Benne R, Van den Burg J, Brakenhoff J, Sloof P, Van Boom JH, Tromp MC (1986). Major transcript of the frameshift coxll from trypanosome mitochondria contains four nucleotides that are not encoded in the DNA. Cell 46, 819-826 36 Hanson MR, Sutton CA, Lu B (1996) Plant organdie gene expression: Altered by RNA editing, Trends Plant &'i I, 57-.-64 37 Cattaneo R (1991) Different types of messenger RNA editing. Anuu Rer Genet 25, 71-88 38 Cattaneo R (1992) RNA editing - In chloroplast and brain. Trends Bh~chem Set 17, 4--5 39 Mar~chal-Drouard L, Ramamonjisoa D, Co,~set A, Well JH, Dietrich A (1993) Editing corrects mispairing in the acceptor stem of bean and potato mitochondrial phenylalanine tranffer RNAs. Nucleic Acids Res 2 I, 4909--4914 40 Binder S, Marchfelder A, Brennicke A (1994) RNA editing of IRNAPI;e and IRNAcy~ in mitochondria of Oenothera herteriana i~ initiated in presursor molecules. Mol Gen Ge,et 244. 67-74 41 Mar~chal-Drouard L, Kumar R, Remacle C, Small l (1996) RNA editing of larch mitochondrial tRNAH~s precursors is a prerequisite for processing, Nucleic Acids Res 24, in press 42 Yu W, Schuster W (1995) Evidence for a site-specific cytidine deamination reaction involved in C to U RNA editing of plant miiochondria. J Biol Chem 270, 18227-18233 43 Rajasekhar VK+ Mulligan RM (1994) RNA editing in plant mitochondria: alpha-phosphate is retained during C to U conversion in raRNAs. Plant Cell 5, 1843-1852 44 Weber E Dietrich A, Well Ill, Mar~chaI-Drouard L 11990) A potato mitochondrial isoleucine tRNA is coded for by a niilochondrial gene possessing a melhionine anticodon. Nm'leh' At'ids Res 18, 5027-5030 45 Muramatsu T, Nishikawa K, Nemoto E Kuchino Y, Nishimura A, Miyasawa I", Yokohama S (1988) Codon and aminoacid spccificities of a transfer RNA are both converted by a single post-transcriptional modification. Nature 336, 179-18 I 46 Sangar6 A, Lonsdale D, Weil JH, Grienenberger JM (1989) Sequence analysis of the tRNAT~+and tRNALYS, genes and evidence for the transcription of a chloroplast-like tRNA..... in maize miiu~.h,mJr~a. C,,, Genet 16, ! 95-20 I 47 Francis MA, Dudock BS (1982) Nucleotide sequence of a spinach chloroplast isoleucine tRNA. J Biol Chem 257, 11195-11197 48 Vemnico P,Gallerani R, Ceci LR (1996) Compilation and classification of higher plant mitochondrial tRNA genes. Nucleic Acids Res 24, 2199-2203 49 Gualberto JM, Wintz H, Weil JH, Grienenberger JM (1988) The genes coding for subunit 3 of NADH dehydrogenase and for ribosomal prorein S12 are present in the wheat and maize mitochondrial genomes and are co-transcribed. Mol Gen Genet 215, I 18-127 50 Irons KP, Heckman JE, Sinclair JH (1985) Sequence of histidyl tRNA present as a chloroplast insert in mtDNA of Ze, mays. Phmt Mo/Biol 4, 225-232 51 Ceci IR. Ambrosini M. Siculella L, Gallerani R (1989) Location of a single IRNA-His gene tm Ihc master chromosome of sunflower mitochondrial DNA. Plant Sci 61,219-275 52 Ambrosini M, Ceci LR, Fiorella S, Gallerani R (1992) Comparison of regions coding for I,'ansfer RNA (itis) genes of milochondri:d and chlornplasl DNA in sunllower: A proposal concerning the classification of cp-like transfer RNA genes. Plant Mot Biol 20, I--4 53 Binder S, Schuster W, Gri.enenberge.rJM, Well JH, BroLnickeA (199.9) (Jl~, tin,, Lys Phc ~t:l lya Genes fortRNA ,tRNA ,IRNA L tRNA ".tRNA" ' andtRNA are encoded in Oeuothera mitochondrial DNA. Cm'r Genet 17, 353358 54 Mar6chaI-Drouard L. Weil JH, Dietrich A (1993) Transfer RNAs and transfer RNA genes in plants. Atom Rer Plant Physhd 44, 13-32 55 Binder S, Brennicke A (1993) A tRNA gene transcription initiation site is similar to nnRNA and rRNA promoters in plant nnitochondria. Nucleic Acids Res 21, 5012-5019 56 Binder S, Brennicke A (1993) Transcription initiation sites in mitochondria of Oenothera herteriana. J Biol Chem 268, 7849-7855 57 Binder S, Hatzack E Brennicke A (1995) A novel pea mitochondrial in vitro transcription system recognizes homologous and heterologous mRNA and tRNA promoters. J Bhd Chem 270, 22182-22189 58 Remacle C, Mart~chaI-DrouardL (1996) Characterization of the potato mitochondrial transcription unit containing +nalive' n',S (GCU), troF (GAA) and tr,:e (UGG). Plant Mol Biol 30, 553-563 59 Kubo T, Y,mai Y, Kinoshita T. Mikami T (1995) The chloroplast trnPtruW-petG gene cluster in the mitochondrial genomes of Beta vulgaris, B trigyna and B webhiana: Evolutionary aspects. Curt Genet 27, 285289 60 Mar~chal L. Runeberg-Roos P, Grienenberger JM, Colin J, Weil JH, Lejeune B, Qu6tieLE Lonsdale DM (1987) Homology in the re,,ion containing a tRNA rp gene and a (complete or partial) tRNAP' gene

529
in wheat mitochondrial and chloroplast genomes. C,rr Gem't 12.9198 61 Mar6chaI-Drouard L, GLfillemaul P (19881 Nuc]eotidc sequence of bean mitochondrial |RNALCU4 and its cytoplasmic counterpart. Re examination of the modified nucieotide |yesent in position 12 in bean mitochondrial and cytoplasmic IRNAt~Ul sequences. Ntwleic Acids Re.~ 16, 11812 62 Mar~chaI-Drouard L, Neuburger M, Guillemaut P, Douee R, Weil JH, Dietrich A (19901 A nuclear-encoded potato (Solanum tuherosum) mitochondrial tRNALeuand its cytosol; ~ counterpart have identical nucleotide sequences. FEBS Lett 262, 170-172 63 Rogers KC. $611 D (19951 Divergence of glutamate and glutamine aminoacylation pathways: Providing tile evolutionary rationale for mischarging. J Mol Evo140, 476---48I 64 Tarassov I, Entelis N, Martin RP ( 19951 An intact protein translocating machinery is required for mitochondrial import of a yeast cytoplasmic tRNA. J Mol Biol 245, 315-323 65 Seihel P, Trappe J, Villani G, Klopstock T. Papa S, Reichmann H (1995) Transfection of mitochondria: Strategy towards a gene therapy of mitochondrial DNA diseases. Nucleic Acids Res 23, 10-17 66 Vestweber D, Schatz G (19891 DNA-protein conjugates can enter mitochondria via the protein import pathway. Nature 338, 170-172 67 Suyama Y ( 19861 Two dimensional polyacrylamidegel electrophoresis analysis of Tetrahymena mitochondrial tRNA. Curt Getter 10.41 I--420 68 Mahapatra S, Ghosh 1", Adhya S (19941 Import of small RNAs into Leishmanhe mitochondria in vitro. Nucleic AcMs Res 22, 3381-3386 09 Mahapatra S, Adhya S (19961 hnport of RNA into Leishm,nia mitochondria occurs through direct interaction with membrane-bound receptors. J Biol Chem, in press 70 Chen DT, Shi X, Suyama Y (19941 In vivo expression and mitochondrial import of normal and mutated IRNATM in Leishmania. Moi Biochem Parasitol 64, 121-133 71 Schneider A, Martin J, Agabian N (19941 A nuclear-encoded IRNA of To,pamJsoma brm'ei is imported into mitochondria. Mol Cell Biol 14, 2317-2322 72 Tarassov IA, Entelis N. Marlin R (19951 Mitochondrial import of a cytoplasnfic lysine-tRNA in yeast is mediated by cooperation uf cytoplasmic and mitochondrial lysyl-lRNA synthetascs. EMBO,I 14.3461347 I 73 Hauser R, Schneider A ( 19951 tRNAs arc imported into mitochondria of Trylntnosoma h~m'ei independently of their genomic context arid genetic origin. EMBO ,I 14, 4212--422(I 74 Marlin RP, Schneller JM, Stahl A, Dirheimer G (1979) Impor! of nuclear deoxyribonucleic acid coded lysine accept ing transle'r ribonucleic acid (anticodon CUU) into yeast mitochondria. Bio~'hemistt3' 18, 4600460,'~ 75 Hou YM, Schimmel P (19881 A simple structural feature is a major delemfinant of the identity of a transfel RNA. Nature 333. 140-145 76 Hou YM, Schimmel P (19891 Evidence that a major determinant for the identity of a transfer RNA is conserved in evolution. Biochemistry 28, 6800-6804 77 Cameiro VTC, Dietrich A, Mar6chaI-Drouard L, Cosset, A, Pelleticr G. Small I ( 19941 Characterization of some major identity elements in plant alanine and phenylalanine Iransfer RNAs. Phmt Mol Biol 26.
1843-1853

78 Dietrich A, Mardchal-Drouard L, Carneiro V. ( ~:~sselA. Small | (19%) A single base change prevents import of cymsollc tRNAAI~'into mito' chondria in transgenic plants. Plant ,J 10. in press 79 Mireau H. Lancelin D. Small ID (1996) The same: Arahidopsis gene encodes both cylosolic and mitochondrial alanyl-lPNA synthetases. Plant Cell 8, 1027-1039 80 Joyce PBM, S12encer D F, Bonen L, Gray.. MW (19881 Genes for tRNAp'sp.tRNAvr, tRNA ~yrand two tRNAs ''er in wheal mitochondrial DNA. Plant Mol Biol 10, 251-262 81 Waldron C. Wills N, Gesteland RF (19851 Plant tRNA genes: Putative soybean genes for tRNA-Asp and tRNA-Met. J Mol Appl Genet 3, 7-17 82 Brennicke A, Grohmann L Hiesel R, Knoop V, Schuster W ( 19931 The mitochondrial genome on its way to the nucleus: Different stages of gene transfer in higher plants. FEBS Lett 325, I,~0--145 83 Li K, Smagula CS, Parsons WJ, Richardson JA, Gonzalez M, Hagler HK, Williams RS (19941 Subcellular partitioning of MRP RNA assessed by ultrastructural and biochemical analysis. ,t Cell Biol 124, 871-882 84 Somasundaran M, Zapp ME Beattie LK, Pang L, Byron KS, Bassell GJ, Sullivan JL. Singer RH ( 19941 Localization of HIV RNA in mitochondria of infected cells: Potential role in cytopathogenicity. J Cell Biol 126, 1353-1360 85 Akashi K, Sakurai K. Hirayama J, Fukuzawa H, Ohyama K (19961 Occurrence of nuclear-encoded IRNAlie in the mitochondria of a liverwort, Morchantia polymoq~ha. Curr Genet, in press 86 Glover K, Gray MW (1994) The genetic origin of transfer RNAs in wheat mitochondria. Poster Abstract 135. 4th International Congress ql"Plant Molecular Biology, Amsterdam, June 19-24 87 Leon P, Walbot V, Bedinger P (19891 Molecular analysis of the linear 2.3 kb plasmid of maize mitochondria: Apparent capture of IRNA genes. Nucleic Acids Res 17, 4089-4099 88 Joyce PBM, Gray MW (19891 Aspartate and asparagine IRNA genes in wheat mitochondrial DNA: A cautionary note on the isolation of tRNA genes from plants. Nucleic Acids Res 17, 7865-7878 89 Ceci LR, Ambrosini M, Fiorella S, Gallerani R (19941 Detection of a conserved arrangement of three IRNA genes in the sunflower mitochondrial genome. Identification, mapping and expression of tr, CtrnN-trnY genes. Biochemisn3" attd Molecular BiohJgy International 32, 1161-1172 q0 Mar6chal L. Guillemaut R Well JII (19851 Sequence of two bean miIochondrial tRNAs%'r which differ in the level of post-transcriptional modification and have a prokaryotic-like large extra-loop. Plant Mol Biol 5, 347-35 I 91 Br:tdford MM (1976) A rapid and sensitive method for the quanlilation of microg,'am quantities of protein utilizblg the principle of protein-dye binding. Anal Biochem 72, 248-254 92 Chaumont F, de Castro Silva Filho. M, Thomas D, Lelerme S, Boutry M (1994) Truncated presequences of mitochondrial FI-ATPase []subunit from Nicotiana plumhagin(folia transport CAT and GUS proteins into mitochondria of transgenic tobacco. Plant Mol Biol 24. 631-641