Portal de Periódicos da CAPES

Overlapping Destinations for Two Dual Targeted Glycyl-tRNA Synthetases in Arabidopsis thaliana and Phaseolus vulgaris

2001; Elsevier BV; Volume: 276; Issue: 18 Linguagem: Inglês

10.1074/jbc.m011525200

1083-351X

Anne‐Marie Duchêne, Nemo Peeters, André Dietrich, Anne Cosset, Ian Small, Henri Wintz,

Mitochondrial Function and Pathology

In plant mitochondria, some of the tRNAs are encoded by the mitochondrial genome and resemble their prokaryotic counterparts, whereas the remaining tRNAs are encoded by the nuclear genome and imported from the cytosol. Generally, mitochondrial isoacceptor tRNAs all have the same genetic origin. One known exception to this rule is the group of tRNAGly isoacceptors in dicotyledonous plants. A mitochondrion-encoded tRNAGly and at least one nucleus-encoded tRNAGly coexist in the mitochondria of these plants, and both are required to allow translation of all four GGN glycine codons. We have taken advantage of this atypical situation to address the problem of tRNA/aminoacyl-tRNA synthetase coevolution in plants. In this work, we show that two different nucleus-encoded glycyl-tRNA synthetases (GlyRSs) are imported into Arabidopsis thalianaand Phaseolus vulgaris mitochondria. The first one, GlyRS-1, is similar to human or yeast glycyl-tRNA synthetase, whereas the second, GlyRS-2, is similar to Escherichia coliglycyl-tRNA synthetase. Both enzymes are dual targeted, GlyRS-1 to mitochondria and to the cytosol and GlyRS-2 to mitochondria and chloroplasts. Unexpectedly, GlyRS-1 seems to be active in the cytosol but inactive in mitochondrial fractions, whereas GlyRS-2 is likely to glycylate both the organelle-encoded tRNAGly and the imported tRNAGly present in mitochondria. In plant mitochondria, some of the tRNAs are encoded by the mitochondrial genome and resemble their prokaryotic counterparts, whereas the remaining tRNAs are encoded by the nuclear genome and imported from the cytosol. Generally, mitochondrial isoacceptor tRNAs all have the same genetic origin. One known exception to this rule is the group of tRNAGly isoacceptors in dicotyledonous plants. A mitochondrion-encoded tRNAGly and at least one nucleus-encoded tRNAGly coexist in the mitochondria of these plants, and both are required to allow translation of all four GGN glycine codons. We have taken advantage of this atypical situation to address the problem of tRNA/aminoacyl-tRNA synthetase coevolution in plants. In this work, we show that two different nucleus-encoded glycyl-tRNA synthetases (GlyRSs) are imported into Arabidopsis thalianaand Phaseolus vulgaris mitochondria. The first one, GlyRS-1, is similar to human or yeast glycyl-tRNA synthetase, whereas the second, GlyRS-2, is similar to Escherichia coliglycyl-tRNA synthetase. Both enzymes are dual targeted, GlyRS-1 to mitochondria and to the cytosol and GlyRS-2 to mitochondria and chloroplasts. Unexpectedly, GlyRS-1 seems to be active in the cytosol but inactive in mitochondrial fractions, whereas GlyRS-2 is likely to glycylate both the organelle-encoded tRNAGly and the imported tRNAGly present in mitochondria. aminoacyl-tRNA synthetase glycyl-tRNA synthetase polymerase chain reaction green fluorescent protein Aminoacyl-tRNA synthetases (aaRSs)1 play a crucial role in protein synthesis by catalyzing the addition of amino acids to their cognate tRNAs. In plants, protein synthesis occurs in three cellular compartments: the cytosol, the mitochondria, and the chloroplasts. All tRNAs and aaRSs necessary for mRNA translation have to be present in these three compartments. In photosynthetic plants, cytosolic tRNAs are all nucleus-encoded, and chloroplastic tRNAs are all chloroplast-encoded (1Maréchal-Drouard L. Weil J.H. Dietrich A. Annu. Rev. Plant Physiol. Plant Mol. Biol. 1993; 44: 13-32Crossref Scopus (98) Google Scholar). By contrast, plant mitochondrial tRNAs have several genetic origins. Some are nucleus-encoded and imported from the cytosol. The others are mitochondrion-encoded and show 70–75% identity with their prokaryotic counterparts and less than 60% identity with their cytosolic counterparts. These mitochondrion-encoded tRNAs are of two types: the "native" tRNAs derived from authentic mitochondrial tRNA genes and the "chloroplast-like" tRNAs that are 97–100% identical to their chloroplast counterparts. The genes corresponding to the latter originated from the chloroplast and were inserted into the mitochondrial genome during evolution. The number of tRNAs in each category (imported, native, or chloroplast-like) and their identity can vary from one plant species to another (2Dietrich A. Small I. Cosset A. Weil J.H. Maréchal-Drouard L. Biochimie ( Paris ). 1996; 78: 518-529Crossref PubMed Scopus (59) Google Scholar,3Kumar R. Maréchal-Drouard L. Akama K. Small I. Mol. Gen. Genet. 1996; 252: 404-411Crossref PubMed Scopus (66) Google Scholar). Higher plant aaRSs are all encoded by the nuclear genome and post-translationally addressed to the different subcellular compartments. The fidelity of translation relies in part on the specificity of the aminoacylation reaction catalyzed by the aaRSs. Thus, strong coevolution is expected between the aaRSs and their cognate tRNAs (4Ribas de Pouplana L. Frugier M. Quinn C. Schimmel P. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 166-170Crossref PubMed Scopus (46) Google Scholar, 5Doolittle R. Handy J. Curr. Opin. Genet. Dev. 1998; 8: 630-636Crossref PubMed Scopus (66) Google Scholar). A few plant genes coding for mitochondrial aaRSs have been characterized, and it appears that the aaRSs used in mitochondrial translation have, in general, the same genetic origin as their substrate tRNAs. This is the case for tRNAsAla, tRNAsThr, and tRNAsVal and their cognate aaRSs in Arabidopsis thaliana. These tRNAs are most likely to be imported from the cytosol into mitochondria in A. thaliana, because the corresponding genes are absent from the mitochondrial genome (6Unseld M. Marienfeld J.R. Brandt P. Brennicke A. Nat. Genet. 1997; 15: 57-61Crossref PubMed Scopus (725) Google Scholar), and they were shown to be imported into mitochondria in other plants such as potato (3Kumar R. Maréchal-Drouard L. Akama K. Small I. Mol. Gen. Genet. 1996; 252: 404-411Crossref PubMed Scopus (66) Google Scholar). Similarly, cytosolic alanyl-tRNA synthetase (7Mireau H. Lancelin D. Small I. Plant Cell. 1996; 8: 1027-1039Crossref PubMed Scopus (68) Google Scholar), threonyl-tRNA synthetase, and valyl-tRNA synthetase (8Souciet G. Menand B. Ovesna J. Cosset A. Dietrich A. Wintz H. Eur. J. Biochem. 1999; 266: 1-8Crossref PubMed Scopus (55) Google Scholar) are also imported into mitochondria. In all three cases, the mitochondrial form and the cytosolic form of the enzyme are encoded by the same gene. Similarly, dual targeting to mitochondria and chloroplasts was observed for methionyl- (9Menand B. Maréchal-Drouard L. Sakamoto W. Dietrich A. Wintz H. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 11014-11019Crossref PubMed Scopus (80) Google Scholar), histidyl- (10Akashi K. Grandjean O. Small I. FEBS Lett. 1998; 431: 39-44Crossref PubMed Scopus (69) Google Scholar), cysteinyl-, and asparaginyl-tRNA synthetase (11Peeters N.M. Chapron A. Giritch A. Grandjean O. Lancelin D. Lhomme T. Vivrel A. Small I. J. Mol. Evol. 2000; 50: 413-423Crossref PubMed Scopus (72) Google Scholar), whereas the corresponding mitochondrial tRNAs were shown to be encoded by native (initiator tRNAMet, tRNACys) or chloroplast-like (elongator tRNAMet, tRNAHis, tRNAAsn) genes present in the mitochondrial genome (6Unseld M. Marienfeld J.R. Brandt P. Brennicke A. Nat. Genet. 1997; 15: 57-61Crossref PubMed Scopus (725) Google Scholar). An organelle-encoded native tRNAGly(GCC) is present in mitochondria of the dicotyledonous plants A. thaliana (6Unseld M. Marienfeld J.R. Brandt P. Brennicke A. Nat. Genet. 1997; 15: 57-61Crossref PubMed Scopus (725) Google Scholar,12Duchêne A.M. Brubacher S. Cosset A. Maréchal-Drouard L. Dietrich A. Moller I.M. Gardeström P. Glaser E. Plant Mitochondria: From Gene to Function. Backuys Publishers, Leiden, The Netherlands1998: 121-125Google Scholar), potato (Solanum tuberosum) (13Schock I. Maréchal-Drouard L. Marchfelder A. Mol. Gen. Genet. 1998; 257: 554-560Crossref PubMed Scopus (11) Google Scholar), and common bean (Phaseolus vulgaris) (14Brubacher-Kauffmann S. Maréchal-Drouard L. Cosset A. Dietrich A. Duchêne A.M. Nucleic Acids Res. 1999; 27: 2037-2042Crossref PubMed Scopus (25) Google Scholar). Because this tRNA cannot read all four GGN glycine codons, at least one other tRNAGlyis required in mitochondria for translation to occur, and a tRNAGly(UCC) was shown to be imported from the cytosol into S. tuberosum and P. vulgarismitochondria (14Brubacher-Kauffmann S. Maréchal-Drouard L. Cosset A. Dietrich A. Duchêne A.M. Nucleic Acids Res. 1999; 27: 2037-2042Crossref PubMed Scopus (25) Google Scholar). Coexistence of imported and organelle-encoded isoacceptor tRNAs in mitochondria has been reported only in a very few cases, i.e. tRNAsIle in higher plants (2Dietrich A. Small I. Cosset A. Weil J.H. Maréchal-Drouard L. Biochimie ( Paris ). 1996; 78: 518-529Crossref PubMed Scopus (59) Google Scholar) and tRNAsIle, tRNAsThr, and tRNAsVal inMarchantia polymorpha (15Akashi K. Tazkenaka M. Yamaoka S. Suyama Y. Fukuzama H. Ohyama K. Nucleic Acids Res. 1998; 26: 2168-2172Crossref PubMed Scopus (23) Google Scholar). The presence of isoacceptor tRNAs with different genetic origins in plant mitochondria raises the problem of the coevolution between tRNAs and aaRSs. In this work, we show that, along with the coexistence in the organelles of a cytosolic and a native mitochondrial tRNAGly, two different glycyl-tRNA synthetases are imported into mitochondria in dicotyledonous plants. Both are dual targeted proteins, one to the cytosol and mitochondria (called glycyl-tRNA synthetase-1 (GlyRS-1)) and the other to chloroplasts and mitochondria (called GlyRS-2). Unexpectedly, GlyRS-1 was shown to be active in the bean cytosol but not in mitochondria, whereas GlyRS-2 was able to aminoacylate either nuclearly or mitochondrially encoded tRNAsGly. A. thaliana total DNA was extracted from whole plants according to Dellaporta et al. (16Dellaporta S.L. Wood J. Hicks J.B. Plant Mol. Biol. Rep. 1983; 1: 19-21Crossref Scopus (6378) Google Scholar), digested with BglII, self-ligated, and used as a template for PCR amplification (17Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989Google Scholar) with divergent oligonucleotides AM10 (5′-GCGGGATCCGGATCGGAGCGATGGAGATTGGG-3′; theBamHI site for cloning is underlined) and AM12 (5′-CGCTCTAGAGGCTTCCCGTGCAGCTAAACCAG-3′; theXbaI site for cloning is underlined) (see Fig.1 A). Total RNA was extracted from 3-week-oldA. thaliana plants (18Goodall G.J. Wiebauer K. Filipowicz W. Methods Enzymol. 1990; 181: 148-161Crossref PubMed Scopus (149) Google Scholar), and poly(A+) RNA was prepared using a PolyATtract mRNA isolation system IV kit (Promega, Madison, WI). Primer extensions were performed with oligonucleotide AM16 (5′-GGATCGGAGCGATGGAGATTGGG-3′) (see Fig.1 A) and with 1 μg of poly(A+) RNA or 12 μg of total RNA (17Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989Google Scholar). Sequencing reactions were primed with oligonucleotide AM16 (17Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989Google Scholar). Predictions of intracellular targeting of proteins were made using ChloroP, MITOPROT, Predotar, PSORT, or TargetP. Isolation of organelles from A. thaliana plants gave a very poor yield, and extensive cytosolic contamination was observed. Organelles were therefore mainly isolated from bean, and to a lesser extent from wheat and potato. Mitochondria were extracted from 3-week-old A. thaliana plants, from 5-day-old etiolated bean hypocotyls or wheat plantlets, or from potato tubers in extraction buffers containing 0.6m mannitol, 0.3 m sucrose, or 0.3 mmannitol, respectively, as an osmoticum (19Maréchal-Drouard L. Guillemaut P. Cosset A. Arbogast M. Weber F. Weil J.H. Dietrich A. Nucleic Acids Res. 1990; 18: 3689-3696Crossref PubMed Scopus (101) Google Scholar) and were purified on continuous polyvinylpyrrolidone/Percoll gradients (20Neuburger M. Journet E.P. Bligny R. Carde J.P. Douce R. Arch. Biochem. Biophys. 1982; 217: 312-323Crossref PubMed Scopus (249) Google Scholar). ForA. thaliana mitochondria, a second, discontinuous polyvinylpyrrolidone/Percoll gradient was necessary to reduce cytosolic contamination (14Brubacher-Kauffmann S. Maréchal-Drouard L. Cosset A. Dietrich A. Duchêne A.M. Nucleic Acids Res. 1999; 27: 2037-2042Crossref PubMed Scopus (25) Google Scholar). Chloroplasts were extracted from leaves of 6-day-old bean plants (21Robinson C. Barnet L.K. Shaw C.H. Plant Molecular Biology: A Practical Approach. IRL Press at Oxford University Press, Oxford1988: 67-78Google Scholar). For protease treatment, mitochondria were incubated in the presence of 100 μg/ml proteinase K for 5 min at room temperature and 10 min on ice. Upon addition of 2 mm phenylmethylsulfonyl fluoride, organelles were recovered by centrifugation through a cushion of 27% (w/v) sucrose, 20 mm Hepes-KOH, pH 7.5, 1 mmphenylmethylsulfonyl fluoride. For subfractionation, mitochondria were resuspended in 5 mm potassium phosphate buffer, pH 7.5, and incubated on ice for 20 min to disrupt the outer membrane. Gentle homogenizing with a plunger was applied several times during incubation to help release of the outer membrane. The suspension was subsequently loaded on a 15/32/45/52% sucrose step gradient in a 10 mmpotassium phosphate buffer, pH 7.5, containing 2 mm EDTA and 0.2% (w/v) bovine serum albumin and centrifuged for 20 min at 125,000 × g. The outer membrane fraction and the mitoplasts were recovered at the 15/32% interface and the 45/52% interface, respectively. Both were washed in 10 mmpotassium phosphate, pH 7.5, 0.3 m sucrose, 1 mm EDTA, 0.1% (w/v) bovine serum albumin, 5 mmglycine and pelleted for 10 min at 175,000 × g. Mitoplasts were resuspended in the same buffer and disrupted by three freeze/thaw cycles and two sonication cycles of 10 s. The suspension was successively centrifuged for 5 min at 2500 ×g, to eliminate the remaining undisrupted material, and for 30 min at 175,000 × g, to separate the inner membrane fraction from the matrix. Denatured protein extracts for SDS-polyacrylamide gel electrophoresis were prepared in 10 mm Tris-HCl, pH 7.5, 5 mm EDTA, 0.3% (w/v) SDS, 5% (v/v) β-mercaptoethanol. Enzymatic extracts for aminoacylation and chromatography were prepared according to Maréchal-Drouard et al. (22Maréchal-Drouard L. Small I. Weil J.H. Dietrich A. Methods Enzymol. 1995; 260: 310-327Crossref PubMed Scopus (45) Google Scholar). Aminoacylations (19Maréchal-Drouard L. Guillemaut P. Cosset A. Arbogast M. Weber F. Weil J.H. Dietrich A. Nucleic Acids Res. 1990; 18: 3689-3696Crossref PubMed Scopus (101) Google Scholar) were performed in the presence of 10−4m[3H]glycine and 3 μg/μl yeast or Escherichia coli total tRNA, 0.15 μg/μl bean total leaf, mitochondrial, or chloroplastic tRNA (22Maréchal-Drouard L. Small I. Weil J.H. Dietrich A. Methods Enzymol. 1995; 260: 310-327Crossref PubMed Scopus (45) Google Scholar), or 0.04 μg/μl in vitrotranscribed tRNAGly. The sequence corresponding to the N terminus of GlyRS-1 (nucleotides 616–810 in Fig. 1 A) was amplified by PCR using the A. thaliana cDNA as a template and oligonucleotides SIGM.5 (5′-GATCTCTAGAAAAATGCGCATCTTCTCTACA-3′; theXbaI site is underlined) and SIGM.3 (5′-CATGCTCGAGAGTTACCCTGAGCTTCGAC-3′; the XhoI site is underlined). To fuse this sequence with the GFP gene, the PCR products were cloned into the SpeI/SalI sites of pOL-GFP-S65C (11Peeters N.M. Chapron A. Giritch A. Grandjean O. Lancelin D. Lhomme T. Vivrel A. Small I. J. Mol. Evol. 2000; 50: 413-423Crossref PubMed Scopus (72) Google Scholar), generating pSYGM in which the first AUG of the GlyRS-1 sequence was in a favorable context for initiation of translation. The sequence corresponding to the N terminus of GlyRS-2 (GenBankTM accession number AJ003069, nucleotides 32–275) was amplified by PCR using A. thaliana total DNA as a template and oligonucleotides SIGO.5 (5′-GATCTCTAGAAAAATGGCCATCCTCCATT-3′; the XbaI site is underlined) and SIGO.3 (5′-CATGCTCGAGCCTGGAGGCGTTGAA-3′; the XhoI site is underlined). The PCR products were cloned into theSpeI/SalI sites of pOL-GFP-S65C, generating pSYGO. The complete expression cassette containing the cauliflower mosaic virus 35S promoter, the GFP fusion, and the cauliflower mosaic virus 35S terminator was cut out of pSYGM and pSYGO withHindIII and cloned into the binary vector pGPTV-kan (23Becker D. Kemper E. Schell J. Masterson R. Plant Mol. Biol. 1992; 20: 1195-1197Crossref PubMed Scopus (540) Google Scholar). The resulting plasmids, pNP7 and pNP8, respectively, were used to transform the Agrobacterium tumefaciens LBA4404 strain (Life Technologies, Inc.). Transformation was performed by infiltration of Nicotiana benthamiana leaves (24Rubino L. Weber-Lotfi F. Dietrich A. Stussi-Garaud C. Russo M. J. Gen. Virol. 2001; 82: 29-34Crossref PubMed Scopus (49) Google Scholar) with a suspension of A. tumefaciens carrying pNP7 or pNP8. After 25–30 h, protoplats were prepared (25Nagy J.I. Maliga P. Z. Pflanzenphysiol. 1976; 78: 453-455Crossref Google Scholar) from the infiltrated leaves. Protoplasts were stained with a mitochondria-specific dye (MitoTrackerTM, CMTMRos, Molecular Probes, Eugene, OR) and analyzed using an epifluorescence microscope (9Menand B. Maréchal-Drouard L. Sakamoto W. Dietrich A. Wintz H. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 11014-11019Crossref PubMed Scopus (80) Google Scholar). The sequence corresponding to the N terminus of GlyRS-1 (nucleotides 616–742 in Fig. 1 A) was first amplified by PCR using the A. thaliana cDNA clone as a template and oligonucleotides AM26 (5′-CGCGCCATGGGCATCTTCTCTACATTCGTCTTTCATCGC-3′; theNcoI site is underlined) and AM36 (5′-CGCGGATCCTCGGCGTCAATCGGAATCTGGATC-3′; theBamHI site is underlined). This introduced a point mutation at position 735, replacing the second AUG codon with an AUU isoleucine codon. The PCR products were cloned into theNcoI/BamHI sites of pCK-GFP3 (9Menand B. Maréchal-Drouard L. Sakamoto W. Dietrich A. Wintz H. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 11014-11019Crossref PubMed Scopus (80) Google Scholar), yielding a GFP fusion. The obtained plasmid was named pAM160. TheEcoRI/HindIII fragments from pAM160 and pSYGO (see above), which contained the tobacco etch virus translation leader, the GlyRS-1 or GlyRS-2 N terminus, the GFP gene, and the cauliflower mosaic virus 35S terminator, were cloned into pBlueScript-KS (Stratagene, La Jolla, CA). These constructs were used as templates forin vitro transcription/translation carried out with a TNTTM coupled reticulocyte lysate system (Promega) in the presence of [35S]methionine. Import of35S-labeled fusion proteins into purified potato mitochondria was performed according to Wischmann and Schuster (26Wischmann C. Schuster W. FEBS Lett. 1995; 374: 152-156Crossref PubMed Scopus (76) Google Scholar) and analyzed by SDS-polyacrylamide gel electrophoresis. The sequence of the cytosolic form (690 amino acids) of GlyRS-1 was amplified by PCR using the A. thaliana cDNA as a template and oligonucleotides AM19 (5′-CGCGCCATGGACGCCACCGAGCAGTCTCTC-3′; theNcoI site is underlined) and AM20 (5′-GGCGGATCCGTCTGCAGCAGCAGAAGAATG-3′; the BamHI site is underlined) and cloned into pQE60 (Qiagen, Hilden, Germany). The resulting plasmid, pAM147, was used to transform theE. coli TG2 strain (17Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989Google Scholar). Overexpression in E. coli was induced with 2 mmisopropyl-1-thio-β-d-galactopyranoside. Protein extraction under denaturing conditions was performed according to the manufacturer's instructions. Extracts were fractionated by SDS-polyacrylamide gel electrophoresis. The polypeptide corresponding to GlyRS-1 was electroeluted from the gels (27Nguyen N.Y. Chrambach A. Hames B.D. Rickwood D. Gel Electrophoresis of Proteins. IRL Press at Oxford University Press, Oxford1981: 150-151Google Scholar) and injected into rabbits to raise antibodies. Alternately, native enzymatic extracts (22Maréchal-Drouard L. Small I. Weil J.H. Dietrich A. Methods Enzymol. 1995; 260: 310-327Crossref PubMed Scopus (45) Google Scholar) were prepared from the E. coli strain overexpressing GlyRS-1 and used for liquid chromatography fractionation. Proteins were separated by SDS-polyacrylamide gel electrophoresis, electrotransferred onto ImmobilonTM-P membranes (Millipore, Bedford, MA), and submitted to immunological detection following classical protocols (17Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989Google Scholar). Antisera were used at a 1/5000 dilution. Antibodies against mitochondrial superoxide dismutase were a gift from Dirk Inzé(Gent, Belgium), and antibodies against α-tubulin were from Amersham Pharmacia Biotech. Antisera against bean chloroplastic leucyl-tRNA synthetase were obtained previously (28Souciet G. Dietrich A. Colas B. Razafimahatratra P. Weil J.H. J. Biol. Chem. 1982; 257: 9598-9604Abstract Full Text PDF PubMed Google Scholar). Binding of the primary antibodies was revealed by chemiluminescence using peroxidase-conjugated secondary antibodies and ECL reagents (Amersham Pharmacia Biotech). Enzymatic extracts were fractionated by medium pressure chromatography on a 1-ml POROS 20 PE hydrophobic column (PerSeptive Biosystems, Framingham, MA) driven by a BioLogic integrated system (Bio-Rad). The samples (1 mg of proteins) adjusted to 1.5 m ammonium sulfate were loaded at 1 ml/min on the column equilibrated with a 20 mm Tris-HCl buffer, pH 7.5, containing 1.5 m ammonium sulfate, 1 mmMgCl2, 5% (v/v) 1,2-propanediol, 0.1 mm EDTA, 5 mm β-mercaptoethanol, 0.5 mmphenylmethylsulfonyl fluoride, 0.5 mmdiisopropylfluorophosphate. After washing with the same buffer (5 ml), elution was performed at 1 ml/min with a linear ammonium sulfate gradient (10 ml, 1.5 to 0 m). Fractions of 0.33 ml were collected, and aliquots were submitted to aminoacylation assays and Western blot analyses. Enzymatic extracts were also fractionated on a 1.3-ml UNOTM-Q1 anion exchange column (Bio-Rad). The samples were loaded at 0.5 ml/min on the column equilibrated with a 20 mm Tris-HCl buffer, pH 7.5, containing 10 mmNaCl, 1 mm MgCl2, 5% (v/v) 1,2-propanediol, 0.1 mm EDTA, 5 mm β-mercaptoethanol, 0.5 mm phenylmethylsulfonyl fluoride, 0.5 mmdiisopropylfluorophosphate. Elution was performed at 0.7 ml/min with a 15-ml linear gradient from 10 to 500 mm NaCl, followed by a 5-ml linear gradient from 500 to 1000 mm NaCl. Fractions of 0.5 ml were collected. Constructs encoding tRNAsGly were amplified by PCR with the relevant primers (see below) so that the tRNA gene sequence was directly fused to the T7 RNA polymerase promoter at the 5′ terminus and to aBstNI site at the 3′ terminus. PCR products were cloned into the EcoRI or EcoRI/BamHI sites of pUC19. After BstNI digestion, in vitrotranscription of these constructs with T7 RNA polymerase yielded mature sized, unmodified tRNA transcripts with a 3′ CCA end (29Dietrich A. Maréchal-Drouard L. Carneiro V. Cosset A. Small I. Plant J. 1996; 10: 913-918Crossref PubMed Scopus (59) Google Scholar). The following oligonucleotides were used as primers (EcoRI andBamHI sites are underlined; the T7 RNA polymerase promoter or BstNI site is in italics): for mitochondrial native tRNAGly(GCC), 5′ AGCAAGAATTC GAATTGTAATACGACTCACTATAGCGGAAATAGCTTAATGGTAG 3′ and 5′ GTACA GAATTC CCTGGAGCGGAAGGAGGGACTTGAAC 3′; for cytosolic tRNAGly(GCC), 5′ AGCAAGAATTC GAATTGTAATACGACTCACTATAGCACCAGTGGTCTAGTGGTAG 3′ and 5′ GTACAGAATTC CCTGGTGCACCAGCCGGGAATCGAAC 3′; for cytosolic tRNAGly(UCC), 5′ GCAAGAATTC GAATTGTAATACGACTCACTATAGCGTCTGTAGTCCAACGGTTAG 3′ and 5′ CGCGGATCC TGGTGCGTCTGCCGGGAGTCGAAC 3′. By analysis of cDNAs and inverse PCR products, we characterized a eukaryotic-type GlyRS gene in A. thaliana(GenBankTM accession number AJ002062) (30Duchêne A.M. Dietrich A. Plant Physiol. 1997; 115: 1730Google Scholar), and we termed this gene GlyRS-1. For this, an A. thaliana cDNA library (Strasbourg, France) was screened using as a probe an A. thaliana expressed sequence tag clone (GenBankTM accession number 117F7T7) (31Newman T. de Bruijn F. Keegstra K. Kende H. McIntosh L. Ohlrogge J. Raikhel N. Somerville S. Thomashow M. Retzel E. Somerville C. Plant Physiol. 1994; 106: 1241-1255Crossref PubMed Scopus (561) Google Scholar) showing similarity with known GlyRS genes. Several incomplete cDNAs, lacking the 5′ end, were recovered from this screening. The 5′ region of the gene (Fig.1 A) was cloned from A. thaliana genomic DNA by using inverse PCR. Southern blot analysis (17Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989Google Scholar) of total A. thaliana DNA digested with PstI,BglII, or EcoRI suggested that theGlyRS-1 gene is a single copy gene (data not shown), which was confirmed with the availability of the complete A. thaliana genomic sequence ((32); see the ArabidopsisGenome Initiative web site). A second gene showing similarities with GlyRS-1 and other eukaryotic GlyRS sequences was previously identified in the A. thaliana genome (GenBankTM accession number AC002534). However, this second gene was considered to be a pseudogene, because the sequence is partial, it contains a number of frameshifts, and no corresponding expressed sequence tags could be detected. The full GlyRS-1 coding sequence is 729 amino acids long. It is similar to other known eukaryotic GlyRS sequences (more than 45% amino acid identity and 60% similarity with Homo sapiens, Bombyx mori, and Caenorhabditis elegans GlyRS sequences) and presents no obvious similarity with prokaryotic GlyRS sequences such as the E. coli GlyRS, suggesting a cytosolic localization. A multiple alignment of GlyRS sequences is available on our web site devoted to A. thaliana tRNAs and aminoacyl-tRNA synthetases (web site address available from the corresponding author). The high level of similarity between GlyRS-1 and eukaryotic GlyRSs suggests that the native enzyme probably has an α2 quaternary structure, like other eukaryotic GlyRSs. The presence of a mitochondrial targeting peptide at the N-terminal end of the coding sequence was predicted by computer analysis. MITOPROT, Predotar, PSORT, and TargetP gave a score of 0.88, 0.90, 0.76, and 0.91, respectively, for a mitochondrial localization. Alignment of the GlyRS-1 amino acid sequence with that of the other GlyRS sequences starts around the end of the potential targeting sequence. It should also be noticed that a dual localization in the cytosol and in mitochondria was proposed for two homologs of GlyRS-1, in humans (33Mudge S.J. Williams J.H. Eyre H.J. Suthereland G.R. Cowan P.J. Power D.A. Gene. 1998; 209: 45-50Crossref PubMed Scopus (44) Google Scholar, 34Shiba K. Schimmel P. Motegi H. Noda T. J. Biol. Chem. 1994; 269: 30049-30055Abstract Full Text PDF PubMed Google Scholar) and in Saccharomyces cerevisiae (35Turner R.J. Lovato M. Schimmel P. J. Biol. Chem. 2000; 275: 27681-27688Abstract Full Text Full Text PDF PubMed Scopus (50) Google Scholar). Expression of the GlyRS-1 gene was investigated by primer extension. The results shown in Fig. 1 B indicate that there are two major transcripts with different 5′ ends that map 22 nucleotides upstream and 15 nucleotides downstream, respectively, of the first putative AUG initiation codon. The sequence shows a second in-frame AUG, 40 codons downstream of the first initiator AUG. This second AUG maps near the position corresponding to the C-terminal end of the predicted mitochondrial targeting peptide. The 5′ end of the shortest mRNA is located upstream of the second in-frame AUG codon, which could thus be used to initiate translation on this transcript, leading to the synthesis of a short, cytosolic form of GlyRS-1 (690 amino acids, 76 kDa). Translation of the long transcript could generate a long protein (729 amino acids, 80 kDa) with a putative mitochondrial targeting peptide. An expressed GlyRS gene of prokaryotic-type (GenBankTMaccession number AJ003069) was characterized by Uwer et al.(36Uwer U. Willmitzer L. Altmann T. Plant Cell. 1998; 10: 1277-1294Crossref PubMed Scopus (92) Google Scholar) upon analysis of an A. thaliana embryo development mutant due to a unique insertion of a DsA transposable element. The corresponding protein, which is 1068 amino acids long, has no significant similarity with GlyRS-1 but presents regions similar to the α and β subunits of E. coli GlyRS (59% identity with the α subunit and 36% identity with the β subunit), so that a dimeric structure of GlyRS-2 would reflect the tetrameric α2β2 structure of E. coli GlyRS. An alignment of the GlyRS-2 sequence with other GlyRS sequences is also available on our web site. Prediction of the subcellular localization run with ChloroP, Predotar, and TargetP gave a score of 0.57, 0.98, and 0.82, respectively, for a chloroplastic localization of GlyRS-2, and this enzyme was indeed shown to be imported into chloroplasts (36Uwer U. Willmitzer L. Altmann T. Plant Cell. 1998; 10: 1277-1294Crossref PubMed Scopus (92) Google Scholar). However, the corresponding GlyRS mutant has an embryo lethal phenotype. Embryo growth was stopped between the globular and heart stages of embryonic development, and germination of mutant seeds was never observed. Such a phenotype was not expected for a plastid enzyme, and the function of this GlyRS during plastidic translation does not provide a direct hint to its role during embryogenesis (36Uwer U. Willmitzer L. Altmann T. Plant Cell. 1998; 10