Plant Science 158 (2000) 97–105

Identification and characterization of thetrnS/pseudo-tRNA/nad3/rps12 gene cluster from Coix lacryma-jobi

L: organization, transcription and RNA editing

Sandra Martha G. Diasa, Susely F. Siqueiraa, Bernard Lejeuneb, Anete P. de Souzaa,c,*a Centro de Biologia Molecular e Engenharia Genetica (CBMEG), Uni6ersidade Estadual de Campinas (UNICAMP),

Cidade Uni6ersitaria ‘Zeferino Vaz’, 13083-970 C.P. 6010 Campinas SP, Brazilb Institut de Biotechnologie des Plantes, Uni6ersite de Paris Sud, Bataille 630, 91405 Orsay Cedex, France

c Departamento de Genetica e E6olucao, Instituto de Biologia (IB), Uni6ersidade Estadual de Campinas (UNICAMP),Cidade Uni6ersitaria ‘Zeferino Vaz’, 13083-970 C.P. 6109 Campinas SP, Brazil

Received 13 April 2000; received in revised form 31 May 2000; accepted 2 June 2000

Abstract

During a study of mitochondrial sequence conservation between the liverwort Marchantia polymorpha and several Angiospermspecies, as revealed by heterologous hybridization experiments, the trnS/pseudo-tRNA/nad3/rps12 gene cluster in Coix lacryma-jobi L., an Asian grass species from the Andropogoneae, was identified using the mitochondrial probe orf 167 from M.polymorpha. The Coix gene cluster was cloned and sequenced, and its expression analyzed. The gene sequence and gene locusorganization were found to be similar to the corresponding cluster in wheat and maize. Northern hybridization and reversetranscription-polymerase chain reaction analyses indicated that nad3 and rps12 genes were co-transcribed as a 1.25 kb RNAmolecule. The transcript displayed 20 and six RNA edition sites, in the nad3 and rps12 genes, respectively, that changed the codonidentities to amino acids, which are better conserved in different organisms. Twenty-three cDNA clones were analysed for theedition process and revealed different partial editing patterns without apparent sequential processing. © 2000 Elsevier ScienceIreland Ltd. All rights reserved.

Keywords: Mitochondrial DNA; NADH dehydrogenase subunit 3; Ribosomal S12; RNA editing; Coix lacryma-jobi L.

www.elsevier.com/locate/plantsci

1. Introduction

The mitochondrial genomes of higher plantsare much larger than those of non-plant organ-isms, from which they differ in general structure,variable gene arrangement, encoding capacity,and gene expression [1,2]. Almost all of thegenes for protein complexes in the respiratorychain, encoded by animal mitochondrial

genomes, have also been identified in higherplant mitochondria. The mitochondrial genomeof Arabidopsis thaliana contains 57 genes with atleast 42 putative open reading frames [3]. RNAediting occurs widely in higher plant mitochon-dria and involves C to U and, less frequently, Uto C alterations [4–6].

General understanding of the information con-tent of plant mitochondrial genomes has beengreatly advanced by the sequencing of the entiremitochondrial genome of the liverwort Marchan-tia polymorpha [7] and of the higher plant A.thaliana [3]. In the M. polymorpha mitochondrialgenome, 28 open reading frames (orf ) were pre-dicted as being possible genes. Five of these (orf228, 509, 169, 322 and 277) are homologous to

* Corresponding author. Present address: Centro de BiologiaMolecular e Engenharia Genetica (CBMEG), Universidade Estadualde Campinas (UNICAMP), Cidade Universitaria ‘Zeferino Vaz’,13083-970 C.P. 6010 Campinas SP, Brazil. Tel.: +55-19-7881132;fax: +55-19-7881089.

E-mail address: anete@unicamp.br (A.P. de Souza).

0168-9452/00/$ - see front matter © 2000 Elsevier Science Ireland Ltd. All rights reserved.

PII: S 0 1 68 -9452 (00 )00308 -3

S.M.G. Dias et al. / Plant Science 158 (2000) 97–10598

genes required for cytochrome c biogenesis inRhodobacter capsulatus, a photosynthetic bac-terium. Homologous genes to those already men-tioned have also been found in higher plant species[8–16].

Genes are widely dispersed in the mitochondrialgenome of higher plants and their dispositionvaries considerably among plant species. This vari-ability is the result of frequent rearrangementswithin mitochondrial genomes [2]. Co-transcrip-tion of adjacent genes has been found to occur inthe mtDNA of several plant species. Polycistronictranscripts such as the rrn18/rrn5 gene in wheat[17], and rps3/rpl16 [18] and atpA/atp9 genes inmaize [19] may contain sequence coding for re-lated products. However, co-transcribed genessuch as atp9/rps13 in tobacco [20], nad3/rps12 inwheat and maize [21], and orf 25/cox3 in rice [22]are involved in different metabolic pathways. Theco-transcription of genes encoding proteins, actingin different metabolic pathways, indicates thatpost-transcriptional and/or translational regula-tion is important for the control of gene-productabundance [16].

Ribosomal protein genes are generally scatteredthroughout the mitochondrial genome of An-giosperms and, in some cases, may be linked tonon-ribosomal protein genes [8,9,16,21]. Some ofthese associations have been conserved over largeevolutionary distances. For example, the nad3gene (encoding mitochondrial NADH-ubiquinone-oxidoreductase subunit 3) and the rps12 gene(small subunit ribosomal protein 12) are closelylinked and co-transcribed in the mtDNA of An-giosperm families as distant as the monocotgrasses (Sorghum [23], rice [24], maize and wheat[21]) and the dicot Brassicaceae (Arabidopsis [25],radish [26], rapeseed and other Brassica species[16]) and also in the mtDNA from Pinus syl6estrisand other Gymnosperms [27]. In maize and wheat,the trnS and a pseudo-tRNA gene are locatedupstream to nad3 [21].

In this report, the analysis of a 1.4 kb BglIIfragment from the mtDNA of Coix lacryma-jobiL. (a south-east Asia grass species from the An-dropogoneae tribe) identified by heterologous hy-bridization with DNA sequence of themitochondrial orf 167 of M. polymorpha, was de-scribed. This orf I67 was found to hybridize withC. lacryma-jobi nad3 gene, located on a genecluster that includes trnS, pseudo-tRNA and rps12

genes. The organization, sequence analysis, tran-scription and editing pattern of this gene clusterare described.

2. Materials and methods

2.1. Isolation and analysis of mtDNA and mtRNA

Mitochondrial DNA (mtDNA) was isolatedfrom etiolated seedlings of C. lacryma-jobi L. cvAdlay, maize (Zea mays, cv AGF352), pea (Pisumsati6um, cv Mikado), soybean (Glycine max, cvIAC-5) and alfalfa (Medicago sati6um, from alocal market), potato tubers (Solanum tuberosum,cv Binje) and cauliflower inflorescence (Brassicaoleracea, from a local market). Purified mitochon-dria were prepared as previously described [28]:mtDNA was obtained after mitochondrial lysisand CsCl-ethidium bromide centrifugation. Mito-chondrial RNA was prepared as described previ-ously [29], treated with DNAse I, to remove theremaining DNA, then phenol extracted and pre-cipitated. Standard procedures [30] were used forrestriction enzyme digestions and agarose gelelectrophoresis.

2.2. Southern and Northern hybridizations

Restriction-digested mtDNAs were transferredto Hybond-N filters (Amersham, UK) by standardprocedures [30]. DNA fragments used as probeswere purified from gel slices by electroelution andlabeled by random hexamer priming. Het-erologous hybridizations, using M. polymorphamitochondrial orf 167 as a probe against BamHI-digested mtDNA from different Angiosperm spe-cies, were performed under low stringencyconditions: the filters were pre-hybridized (45°C, 4h) and hybridized (45°C, 18 h) in 5×SSC, 10×Denhardt’s solution, 0.1% (w/v) sodium dodecylsulfate (SDS), 20 mM Tris–HCl (pH 7.5) and 0.1mg/ml denatured salmon-sperm DNA, plus aprobe which had been 32P-labeled by a multiprimereaction (Prime-a-Gene kit; Promega). Filters werewashed twice for 15 min in 1×SSC, 0.1% SDSsolution at 50°C and then exposed for autoradiog-raphy. Homologous hybridizations and washeswere performed under stringent conditions: pre-hybridization (42°C, 4 h) and hybridization (42°C,overnight) were carried out in a solution contain-

S.M.G. Dias et al. / Plant Science 158 (2000) 97–105 99

ing 5×SSC, 10×Denhardt’s, 1.0% (w/s) SDS, 20mM Tris–HCl (pH 7.5), 0.1 mg/ml denaturedsalmon-sperm and 50% formamide. The filter wasthen washed twice for 10 min in 2×SSC, 1% SDSat room temperature, and twice for 10 min with0.1×SSC, 0.1% SDS at 42°C, and then exposedfor autoradiography. Isolated mtRNA (5.0 mg/lane) was denatured at 55°C for 15 min in asolution containing 18% formaldehyde, 50% for-mamide and 0.5×MOPS buffer. The sampleswere fractionated on 1.2% agarose gel containing1×MOPS buffer and 0.66 M formaldehyde andblotted onto Hybond N. New membrane stripswere used for each hybridization experiment. Pre-hybridization and hybridization were performedunder stringent conditions. Probe labeling, hy-bridization and washes were as already described.Transcript sizes were determined using RNA sizestandards from Gibco-BRL (USA).

2.3. cDNA synthesis and polymerase chainreaction amplification

Reverse transcription was carried out on 1.0 mgDNAseI-treated mtRNA using 10 pmol from anappropriate primer. After heating the mixture for10 min at 65°C, the buffer from an ‘Expanded™Reverse Transcriptase’ kit (Boehringer, Germany),10 mM dithiothreitol, 1 mM of each dNTP, 20 URNAsine (Gibco-BRL) and 50 U Expand ReverseTranscriptase was added. The reaction was incu-bated for 1 h at 42°C and stopped by heating for2 min at 95°C. The resulting cDNAs were am-plified by polymerase chain reaction (PCR) using10 pmol of the appropriate primers in a buffersupplied by the manufacturer, 0.15 mM of eachdNTP, 2.5 mM MgCl2 and 2.5 U Taq DNAPolymerase (Gibco BRL) in a final reaction vol-ume of 100 ml. The annealing temperature used forthe primer combinations fnad3–drps12 anddnad3–drps12 was 61°C. After 3 min at 94°C, 25cycles of amplification were carried out (94°C for1.5 min, 61°C for 2 min, 72°C for 2 min), followedby a final extension of 10 min at 72°C. Theamplified products were purified by agarose gelelectrophoresis before cloning. Control for cDNAsynthesis was performed replacing reverse tran-scriptase by water in the reaction mixtures andverifying through PCR that there was no amplifi-cation product from that template.

2.4. cDNA cloning and sequencing

Standard procedures were used in the prepara-tion, isolation and analysis of recombinant clonesof Escherichia coli [30]. Overlapping DNA restric-tion fragments from the region of interest werecloned into pBluescript vectors (Stratagene, USA)and the cDNAs generated by reverse transcriptase-polymerase chain reaction (RT-PCR) were clonedinto the pGEM-T vector System I (Promega,USA). Nucleotide sequences were determined us-ing the dye terminator-cycle method on an ABIPRISM 310 sequencer (Applied Biosystems,USA). Both strands of the cDNA and genomicclones were sequenced. The sequence data wasanalyzed using the Lasergene System (DNA Star,USA).

2.5. PCR amplification of orf 167 from M.polymorpha mtDNA

M. polymorpha total DNA (10–20 ng) (kindlyprovided by M.-C. Boisselier-Dubayle, MuseumNational d’Histoire Naturelle, Laboratoire deCryptogamie, Paris, France) were mixed with 2.5mM MgCl2 on a buffer supplied by the manufac-turer, containing 0.1 mM of each dNTP and 2.5 UTaq DNA Polymerase (Gibco BRL) in a finalreaction volume of 100 ml. After 3 min at 94°C, 30cycles of amplification were carried out (92°C for1 min, 55°C for 1.5 min, 72°C for 2 min), followedby a final extension of 10 min at 72°C. Theproducts of amplification were cloned into thepGEM-T vector and sequenced to confirm theiridentity before using them as probes in het-erologous hybridizations.

2.6. Oligonucleotides

For PCR amplification of the orf 167 from totalDNA of the M. polymorpha, the primers usedwere: marpo 167a, 5%-AGT TGG AGG AGATAG GAT TTC GTG T-3%; and marpo167b, 5%-GTT ACT TCT TTT GCG GCT GTT TTC T-3%.For cDNA synthesis and RT-PCR amplification,the primers were: fnad3, 5%-GCG AGA GAACGA AGT GGG-3%; dnad3, 5%-GCT TTG GTGATG TCG GAA T-3%; and drps12, 5%-GAG GCATCT TCC ATT CAT TTA G-3%.

S.M.G. Dias et al. / Plant Science 158 (2000) 97–105100

3. Results and discussion

3.1. Isolation of the trnS/pseudo-tRNA/nad3/rps12locus from Coix mt DNA

Several reports have shown that M. polymorphamitochondrial orfs, with initially unknown func-

tions, have been conserved in higher plants. Theidentity of these orfs was discovered by compari-sons with sequence data banks [8–16]. To identifyM. polymorpha mitochondrial orfs conserved inhigher plants, the extent of mitochondrial se-quence conservation between M. polymorpha andAngiosperms were assessed. Several mitochondrialorfs from the liverwort have been used as probesagainst mtDNA from maize, Coix, cauliflower,potato, pea and soybean in heterologous hy-bridization experiments. Appropriate oligonucle-otides were designed to amplify some chosenmitochondrial orfs from M. polymorpha totalDNA. The experimental hybridization conditionswere chosen to identify mtDNA sequences show-ing as little as 50% sequence similarity with theprobe together with keeping a good signal/noiseratio. One of the probes, namely orf 167, hy-bridized with only one or two restriction frag-ments from each of the species listed giving astronger signal with maize and Coix mitochondrialDNAs (Fig. 1). Therefore, orf 167 from M. poly-morpha could be considered a potentially con-served orf in Angiosperms. The hybridizingsequence in Coix mt was located on a BglII 1.4 kbfragment (data not shown) that was cloned into apBluescript and characterized by restriction map-ping (Fig. 2). The orf 167 similarity region waslocalized on the restriction map of this BglII frag-ment by hybridization (results not shown).

3.2. Sequence analysis of thetrnS/pseudo-tRNA/nad3/rps12 locus fn CoixmtDNA

The 1.4 kb BglII fragment was subcloned intopBluescript to obtain appropriate overlappingfragments for DNA sequencing on both strands.The complete nucleotide sequence of the fragment(Fig. 2) was compared with DNA sequence data-bases. The comparisons identified significant simi-larities between the 1.4 kb BglII sequence and acluster of mitochondrial genes of wheat and maize,encoding the following genes: tRNASer (trnS),pseudo-tRNA, subunit 3 of NADH dehydroge-nase (nad3) and ribosomal protein subunit S12(rps12) [21]. Since the sequence of this rps12 genewas interrupted by one of the terminal BglII sitesof the fragment, primer drps12 (rps12 maize-wheatsequence based) in combination with fnad3 primerwas used to amplify the missing portion of the

Fig. 1. Identification of mtDNA restriction fragments con-taining sequences homologous to M. polymorpha mitochon-drial orf 167. Mitochondrial DNA from alfalfa (1), potato (2),Coix (4), cauliflower (5), pea (6), maize (7), soybean (8) andCoix ctDNA (3) was digested with BamHI and fractionatedon 1% agarose gel. The DNA was then transferred to a nylonfilter and hybridized with a 32P-labeled fragment containingorf 167. M, Molecular marker l DNA digested with HindIIIand fX 174 DNA digested with HaeIII.

Fig. 2. Physical map of the restriction sites of the Coix 1.4 kbBglII/BglII fragment. Genes are represented by dark boxes.The four probes used in Northern and Southern blotting areindicated and correspond to nad3 (probe 1), rps12 (probe 2),tRNASer (probe 3) and pseudo-tRNA (probe 4). The arrowsindicate gene orientation. The restriction sites are: A, A6aI;Bg, BglII; E, EcoRI; Ps, PstI; Pv, P6uII; S, SstI; S, SpeI; X,XhoI. The PstI and SstI sites are from the BlueScript plasmidvector, from which 1.4 BglII/BglII was cloned.

S.M.G. Dias et al. / Plant Science 158 (2000) 97–105 101

Fig. 3. Nucleotide sequence of the tRNASer/pseudo-tRNA/nad3/rps12 locus in Coix mtDNA (GenBank accession numberAF239263). The amino acid sequence deduced from the nucleotide sequence is shown in single letter code above the nucleotidesequence. RNA editing sites are indicated by a lowercase ‘c’. The predicted amino acid before and after editing are indicated.Horizontal arrows, synthetic oligonucleotides prepared for cDNA synthesis and PCR amplification; dashed line, pseudo-tRNAinsertion; stars, stop codons. Pseudo-tRNA and tRNASer are underlined. The sequence used for drps12 synthesis was based ona consensus sequence between maize and wheat [21].

rps12 gene from Coix mtDNA. The amplificationproduct was cloned and sequenced, and its se-quence added to that of the 1.4 kb BglII fragment(Fig. 3).

The stringency conditions for heterologous hy-bridizations using orf 167 were chosen for detectinglow similarity with mtDNA sequences. Sequencecomparisons and alignment between the sequencesof orf 167 and the BglII fragment revealed thatorf 167 displayed 45% similarity with the Coix nad3gene; this similarity was evenly distributed over thegene sequence and responsible for the hybridizationsignal as checked by hybridization controls with the

nad3 sequence from wheat (data not shown). Thisvalue is to be compared with the more significant82% similarity found when comparing this sameCoix nad3 gene with the actual homologous nad3from M. polymorpha. Although orf 167 has beendescribed as a hypothetical 18.6 KD protein in thenad3–nad7 intergenic region, with no relationshipto the NAD3 protein, the origin of the similaritywith nad3 reported here is intriguing. It may befortuitous or it may reveal an ancient sequenceduplication in the mitochondrial genome of M.polymorpha having occurred in the vicinity of theactual nad3 sequence.

S.M.G. Dias et al. / Plant Science 158 (2000) 97–105102

The nucleotide sequences of the Coix nad3/rps12 gene show 100–99.4 and 100–99.7% simi-larity with the same gene sequences from maizeand wheat, respectively. In Coix, the intergenicspacer sequence between nad3 and rps12 is 44 bplong, very similar to the 47 bp intergenic spacerfound in maize and wheat [21].

The Coix tRNASer sequence has 100% identitywith the same gene from maize, wheat and rice[21,24,31]. There is a pseudo-tRNA sequence lo-cated 259 bp downstream of the trnS gene, ho-mologous to the sequence found in maize, wheatand rice mitochondrial genomes [21,24,31]. TheCoix pseudo-tRNA gene presents a 50 bp inser-tional sequence in the same position as in othermonocots examined so far (Fig. 3), which is 99%similar to a 48 bp insertion found in the maizepseudo-tRNA. As pointed out previously [24], therice pseudo-tRNA gene seems to have evolvedfrom tRNAPhe because of its high sequence ho-mology with tRNAPhe [32]. Sequence comparisonsexcluding insertional sequences have shown thatCoix and maize pseudo-tRNA genes are 93% sim-ilar with the potato tRNAPhe. Thus, in all mono-cots examined so far [21,24,31], this pseudo-tRNAgene seems to have been inactivated in the courseof evolution. This inactivation may have origi-nated from an intramolecular recombination fol-lowed by a sequence insertion in a commonancestor sequence, probably the mitochondrial

tRNAPhe gene.Southern blot hybridizations using the nad3 and

pseudo-tRNA genes as probes against Coix,maize, cauliflower, potato, soybean, pea and al-falfa mitochondrial genomes digested by BamHI,showed that the nad3 gene is present as a singlecopy in mtDNA from Coix and other species(data not shown). However, of the presented spe-cies, only the maize and Coix mtDNAs gavestrong hybridization signals with pseudo-tRNAgene (data not shown). These results suggest thatthis pseudo-tRNA gene originated from a recom-bination event that took place in a monocot an-cestor, after the dicot/monocot divergence.

3.3. Transcription of thetrnS/pseudo-tRNA/nad3/rps12 locus

Northern blot hybridizations were carried outwith total mitochondrial RNA from Coix usingprobes corresponding to each of the genes in thecluster: a 0.5 kb P6uII/SpeI fragment (probe 1)for nad3, a 0.3 kb SstI/XhoI fragment (probe 2)for rps 12, and a 0.6 P6uII/PstI fragment (probe3) for trnS (Fig. 2). Only a single abundant tran-script of approximately 1250 nts was detectedwith probes 1 and 2, indicating that the Coix nad3and rps12 genes were co-transcribed, as confirmedby RT-PCR amplification (see later). This tran-script pattern was similar to that of nad3/rps12transcription in rice, where a single 1.2 kb tran-script was identified [24]. However, maize andwheat have shown a more complex transcriptionpattern, with two and at least four transcripts,respectively [21], and the mRNA is 0.9 kb inlength. The mRNA transcript size difference be-tween Coix and maize was surprising, consideringthe high sequence identity (95–98%) shared be-tween the 1455 pb Coix trnS/pseudo-tRNA/nad3/rps12 locus and the same region in maize andwheat. Considering that the two mapped 5% endsof the wheat transcript [21] are within a sequenceregion showing almost 100% identity with theCoix sequence, it can be supposed that the differ-ence observed in transcript size would be ex-plained by the 3% end transcript change. However,ends of the nad3/rps12 Coix transcript need to bemapped to confirm this hypothesis.

Northern hybridizations were used to testwhether the trnS gene and pseudo-tRNA weretranscribed in Coix. In this way, probes 3 and 4(Fig. 2) were hybridized with Coix total mtRNA.

Fig. 4. RNA hybridizations of the nad3, rps12 and tRNASer

transcripts in Coix. Five micrograms of total mtRNA weredenatured and applied to each lane of a 1.2% agarose/formal-dehyde gel, transferred to a nylon membrane and hybridized32P-labeled fragments of nad3 (1), rps12 (2) and tRNASer (3).An RNA ladder (9.5–0.24 kb, BRL) was used to estimatemolecular size.

S.M.G. Dias et al. / Plant Science 158 (2000) 97–105 103

Table 1Degree of editing of the nad3 and rps12 genes in the 23 cDNA samples analyzed

nad3Editing rps12

CDNANumber of edited sitesCDNA Number of edited sitesclonesclones

% NN %

26 20Completely edited 126 53 6Incompletely but extensively edited 10 43 18–19 4 17 5Barely edited 2 9 1 3 13 2

22 0 45 17Completely unedited 023Total 23

Only probe 3 hybridized to a transcript (Fig. 4),indicating that trnS is transcribed in Coix whereasthe pseudo-tRNA is not.

As already mentioned, nad3-rps12 co-transcrip-tion was confirmed by RT-PCR amplification.Three primers (fnad3, dnad3 and drps12) weredesigned from the coding regions of the nad3 andrps12 genes (Fig. 3). An approximately 800 bpamplification product was obtained for the nad3/rps12 cotranscript.

3.4. RNA editing pattern of the nad3/rps12transcripts in Coix

To characterize the extent and sites of RNAediting in nad3/rps12 transcripts, cDNA coveringthe coding region was obtained by RT-PCR usingthe primers fnad3 or dnad3 and drps12 (Fig. 3).The comparison of the genomic and 23 cDNArevealed 20 C-to-U RNA editing events in nad3and six in the rps12 coding region (Fig. 3). Twenty-six C-to-U editing sites resulted in 16 and six codonmodifications in the nad3 and rps12 genes, respec-tively, corresponding to 13.6% of the NAD3 and4.8% of the RPS12 amino-acid sequence. Most ofthe editing sites in the Coix nad3–rps12 genes havealready been found in wheat [21], Oenothera [33],Pinus [27], Magnolia, onion, sunflower [34], Petu-nia [35], Sorghum [23] and Brassica [16]. A com-parison of the nad3 editing positions in Coix andwheat [36] showed four differences: codons 13, 42and 46 were edited only in wheat, and codon 62(editing site six in Fig. 3) only in Coix, whereas therps12 editing sites in Coix and wheat were identi-cal. Contrary to the case with wheat, these fourediting events in Coix nad3 were identical to those

in Oenothera [33] and only one of them was differ-ent in Pinus [27]. These sites probably existed priorto the evolutionary separation of monocots anddicots, and the species-specific differences origi-nated from genomic point mutations in some lin-eages during evolution.

Sequence analysis of 23 cDNA clones revealedunedited and partially edited clones with no evi-dent polarity for the editing process. The extent ofediting differed between nad3 and rps12 tran-scripts: more than 50% of them presented com-pletely edited rps12, whereas only 26% of themwere fully edited in the nad3 region (Table 1). Thisfinding suggests that the rps12 editing sites in Coixare edited first or faster than the nad3 editing sites.This pattern differs from that in Magnolia whereall 23 editing sites in the PCR-derived nad3 cDNAwere found to be fully edited, while all of theediting sites in rps12 were only partially altered[34]. Mixtures of partially or differentially editedcDNA clone populations derived from the nad3–rps12 loci have also been found in the mitochon-dria of other plants such as wheat [36] and pine[27].

In conclusion, a high conservation of the trnS/pseudo-tRNA/nad3/rps12 sequence, gene organi-zation and editing pattern of the nad3/rps12 genesamong Coix and monocots were found, especiallybetween maize and Coix, indicating that evolutionhas produced a high degree of conservation in thislocus in these species. This fact is surprising consid-ering; first, Coix is a distant relative of maize (anAsian species of the Andropogoneae tribe); second,the high recombinationary ability of the mitochon-drial genome plant; and, finally, the general non-conservation of different gene clusters, even in veryclose evolutionarily species.

S.M.G. Dias et al. / Plant Science 158 (2000) 97–105104

Acknowledgements

The authors wish to thank Dr. M.-C. Boisselier-Dubayle (Museum National d’Histoire Naturelle,Laboratoire de Cryptogamie, Paris, France) forthe gift of Marchantia polymorpha total DNA.This work was financed by grants to A.P.S. fromFundacao de Amparo a Pesquisa de Sao Paulo(FAPESP; 96/03520-8). A.P.S. was also the recipi-ent of a research fellowship from Conselho Na-cional de Desenvolvimento Cientıfico eTecnologico (CNPq). S.M.G.D. and S.F.S. weresupported by graduate fellowships from FAPESP.

References

[1] M.W. Gray, Origin and evolution of DNA, Annu. Rev.Cell Biol. 5 (1989) 25–50.

[2] W. Schuster, A. Brennicke, The plant mitochondrialgenome: physical structure, information content, RNAediting, and gene migration to the nucleus, Annu. Rev.Plant Physiol. 45 (1994) 61–78.

[3] M. Unseld, J.R. Marienfeld, P. Brandt, A. Brennicke,The mitochondrial genome of Arabidopsis thaliana con-tains 57 genes in 366 924 nucleotides, Nat. Genet. 15(1997) 57–61.

[4] P.S. Covello, M.W. Gray, RNA editing in plant mito-chondria, Nature 341 (1989) 662–666.

[5] J.M. Gualberto, G. Bonnard, L. Lamattina, J.H. Weill,J.M. Grienenberger, RNA editing in wheat mitochon-dria results in the conservation of protein sequences,Nature 341 (1989) 660–662.

[6] R. Heisel, B. Wissinger, W. Schuster, A. Brennicke,RNA editing in plant mitochondria, Science 246 (1989)1632–1634.

[7] K. Oda, K. Yamato, E. Ohta, Y. Nakamura, M. Take-mura, N. Nozato, K. Akashi, T. Kanegae, Y. Ogura, T.Kohchi, K. Ohyama, Gene organization deduced fromthe complete sequence of liverwort Marchantia polymor-pha mitochondrial DNA: a primitive form of plantmitochondrial genome, J. Mol. Biol. 223 (1992) 1–7.

[8] D.H. Gonzalez, G. Bonnard, J.M. Grienenberger, Agene involved in the biogenesis of cytochromes is co-transcribed with a ribosomal protein gene in wheatmitochondria, Curr. Genet. 24 (1993) 248–255.

[9] W. Schuster, B. Combettes, K. Flieger, A. Brennicke, Aplant mitochondrial gene encodes a protein involved incytochrome c biogenesis, Mol. Gen. Genet. 239 (1993)49–57.

[10] W. Schuster, The highly edited orf 206 in Oenotheramitochondria may encode a component of a heme trans-porter involved in cytochrome c biogenesis, Plant Mol.Biol. 25 (1994) 33–42.

[11] G. Bonnard, J.M. Grienenberger, A gene proposed toencode a transmembrane domain of an ABC transporteris expressed in wheat mitochondria, Mol. Gen. Genet.246 (1995) 91–99.

[12] W. Jekabson, W. Schuster, Orf 250 encodes a secondsubunit of an ABC-type transporter in Oenothera mito-chondria, Mol. Gen. Genet. 246 (1995) 166–173.

[13] M. Nakazono, Y. Ito, N. Tsutsumi, A. Hirai, The genefor a subunit of an ABC-type heme transporter is tran-scribed together with the gene for subunit 6 of NADHdehydrogenase in rice mitochondria, Curr. Genet. 29(1996) 412–416.

[14] H. Handa, G. Bonnard, J.M. Grienenberger, The rape-seed mitochondrial gene encoding a homologue of thebacterial protein Ccl1 is divided into two independentlytranscribed reading frames, Mol. Gen. Genet. 252 (1996)292–302.

[15] R. Menassa, N. El-Rouby, G.G. Brown, An open read-ing frame for a protein involved in cytochrome c biogen-esis split into two parts in Brassica mitochondria, Curr.Genet. 31 (1997) 70–79.

[16] K. Itani, H. Handa, Rapeseed mitochondrial ccb206, agene involved in cytochrome c biogenesis, is co-tran-scribed with the nad3 and rps12 genes: organization,transcription, and RNA editing of the nad3/rps12/ccb206 locus, Curr. Genet. 34 (1998) 318–325.

[17] D. Falconet, B. Lejeune, F. Quetier, M.W. Gray, Evi-dence for homologous recombination between repeatedsequences containing 18S and 5S ribosomal RNA genesin wheat mitochondrial DNA, EMBO J. 3 (1984) 297–302.

[18] M.D. Hunt, K.J. Newton, The NC3 mutation-genetic-evidence for the expression of ribosomal-protein genes inZea mays mitochondria, EMBO J. 10 (1991) 1045–1052.

[19] P.S. Covello, M.W. Gray, Sequence-analysis of wheatmitochondrial transcripts capped in 6itro : definitive iden-tification of transcription initiation sites, Curr. Genet. 20(1991) 245–251.

[20] M.M. Bland, C.S. Leavings III, D.F. Matzinger, Thetobacco mitochondrial ATPase subunit 9 gene is closelylinked to an open reading frame for a ribosomal protein,Mol. Gen. Gen. 204 (1986) 4–8.

[21] J.M. Gualberto, H. Wintz, J.H. Weil, J.M. Grienen-berger, The genes coding for subunit 3 of NADH dehy-drogenase and for ribosomal protein S12 are present inthe wheat and maize mitochondria genomes and areco-transcribed, Mol. Gen. Genet. 215 (1988) 118–127.

[22] A.W. Liu, K.K. Narayanan, C.P. Andre, E.K. Kaleikau,V. Walbot, Co-transcription of orf 25 and coxIII in ricemitochondria, Curr. Genet. 21 (1992) 507–513.

[23] W. Howad, F. Kempken, Sequence analysis and tran-script processing of the mitochondrial nad3–rps12 genesfrom Sorghum bicolor, Plant Sci. 129 (1997) 65–68.

[24] T. Suzuki, S. Kazama, A. Hirai, T. Akihama, K. Kad-owaki, The rice mitochondrial nad3 gene has an ex-tended reading frame at its 5% end: nucleotide sequenceanalysis of rice trnS, nad3, and rps12 genes, Curr. Genet.20 (1991) 331–337.

[25] P. Brandt, S. Sunkel, M. Unseld, A. Brennicke, V.Knoop, The nad4L gene is encoded between exon c ofnad5 and orf 25 in the Arabdopsis mitochondrialgenome, Mol. Gen. Genet. 236 (1992) 33–38.

[26] C.A. Makaroff, J.D. Palmer, Mitochondrial DNA rear-rangements and transcriptional alterations in the malesterile cytoplasm of Ogura radish, Mol. Cell Biol. 8(1988) 1474–1480.

S.M.G. Dias et al. / Plant Science 158 (2000) 97–105 105

[27] B. Karpinska, S. Karpinski, J.E. Hallgren, The genesencoding subunit 3 of NADH deydrogenase and riboso-mal protein S12 are co-transcribed and edited in Pinussyl6estris (L.) mitochondria, Curr. Genet. 28 (1995)423–428.

[28] F. Vedel, F. Quetier, Physico-chemical characterizationof mitochondrial DNA from potato tubers, Biochim.Biophys. Acta 340 (1974) 374–387.

[29] D.B. Stern, K.J. Newton, Isolation of plant mitochon-drial RNA, Methods Enzymol. 118 (1986) 488–496.

[30] J. Sambrook, E.F. Fritsch, T. Maniatis, Molecularcloning: a laboratory manual, 2nd ed. (1989), ColdSpring Harbor Laboratory, Cold Spring Harbor, NY.

[31] H. Wintz, J.M. Grienenberger, J.H. Weil, D.M. Lons-dale, Location and nucleotide sequence of two tRNAgenes and a tRNA pseudogene in maize mitochondrialgenome: evidence for the transcription of a chloroplastgene in mitochondria, Curr. Genet. 13 (1988) 247–254.

[32] L. Marechal, P. Guillemault, J.M. Grienenberger, G.Jeannin, J.H. Weil, Structure of bean mitochondrial

tRNAPhe and localisation of the tRNAPhe on the mito-chondrial genomes of maize and wheat, FEBS Lett. 184(1985) 289–293.

[33] W. Schuster, B. Wissinger, M. Unseld, A. Brennicke,Transcripts of the NADH-dehydrogenase subunit 3 geneare differentially edited in Oenothera mitochondria,EMBO J. 9 (1990) 263–269.

[34] G. Perrotta, T.M. Regina, L.R. Ceci, C. Quagliariello,Conservation of the organization of the mitochondrialnad3 and rps12 genes in evolutionary distant an-giosperms, Mol. Gen. Genet. 251 (1996) 326–337.

[35] B. Lu, M.R. Hanson, A single nuclear gene specifies theabundance and extent of RNA editing of a plant mito-chondrial transcript, Nucleic Acids Res. 20 (1992) 5699–5703.

[36] J.M. Gualberto, G. Bonnard, L. Lamattina, J.M.Grienenberger, Expression of the wheat mitochondrialnad3–rps12 transcription unit: correlation between edit-ing and mRNA maturation, Plant Cell 3 (1991) 1109–1120.

.