Chloroplast-like transfer RNA genes expressed in wheat mitochondria

Volume 17 Number 14 1989

Volume 17 Number 14 1989

Nucleic Nucleic Acids Acids Research Research

Chloroplast-like transfer RNA genes expressed in wheat mitochondria

Paul B.M.Joyce and Michael W.Gray*

Department of Biochemistry, Dalhousie University, Halifax, Nova Scotia B3H 4H7, Canada

Received May 15, 1989; Accepted June 20, 1989

EMBL accession nos X15118, X151 19

ABSTRACT In the course of a systematic survey of wheat mitochondrial tRNA genes, we have sequenced chloroplast-like serine (trnS-GGA), phenylalanine (trnF-GAA) and cysteine (trnC-GCA) tRNA genes and their flanking regions. These genes are remnants of 'promiscuous' chloroplast DNA that has been incorporated into wheat mtDNA in the course of its evolution. Each gene differs by one or a few nucleotides from the authentic chloroplast homolog previously characterized in wheat or other plants, and each could potentially encode a functional tRNA whose secondary structure shows no deviations from the generalized model. To determine whether these chloroplast-like tRNA genes are actually expressed, wheat mitochondrial tRNAs were resolved by a series of polyacrylamide gel electrophoreses, after being specifically end-labeled in vitro by 3'-CCA addition mediated by wheat tRNA nucleotidyltransferase. Subsequent direct RNA sequence analysis identified prominent tRNA species corresponding to the mitochondrial and not the chloroplast trnS, trnF and tmnC genes. This analysis also revealed chloroplast-like elongator methionine, asparagine and tryptophan tRNAs. Our results suggest that at least some chloroplast-like tRNA genes in wheat mtDNA are transcribed, with transcripts undergoing processing, post-transcriptional modification and 3'-CCA addition, to produce mature tRNAs that may participate in mitochondrial protein synthesis.

INTRODUCTION Considering its large size (200-2400 kbp) [1], plant mitochondrial (mt) DNA appears to contain a surprisingly limited number of tRNA genes [2]. These fall into two discrete classes. One group encodes tRNA species that display 65-80% sequence similarity with the homologous eubacterial and chloroplast tRNAs [2-19]. The genes for these 'native' tRNAs are assumed to have originated from the eubacteria-like endosymbiont(s) that provided the mitochondrial genome of plant and other eukaryotic cells [20]. The other group of plant mitochondrial tRNA genes is specifically related ( >90% sequence identity) to tRNA genes encoded by chloroplast (ct) DNA [17,21 -27]. These 'ct-like' tRNA genes represent promiscuous ctDNA [28] that has been incorporated into the plant mitochondrial genome [29] in the course of evolution. Such sequences are widely distributed, seemingly in random fashion, in plant mtDNAs [30]. Although such ct-like tRNA genes may sustain inactivating mutations following their appearance in mtDNA [21], most have remained identical or almost identical to their chloroplast counterparts, and as such are potentially functional. That some actually are functional is supported by the isolation from plant mitochondria of mature, ct-like tRNAs that are clearly distinguishable from their counterparts in the chloroplast of the same organism [31]. The intriguing possibility that chloroplast tRNA genes have been recruited to function in plant mitochondria raises questions about the distribution of such genes in different plant © IRL Press

5461

Nucleic Acids Research mtDNAs, as well as the overall composition of the tRNA population in plant mitochondria. Here we describe three potentially functional, ct-like tRNA genes encountered during a systematic survey of tRNA genes in wheat mtDNA. We also present evidence for the existence in wheat mitochondria of abundant, mature tRNA species corresponding to each of these genes, as well as ct-like elongator methionine, asparagine and tryptophan tRNAs. Our results indicate that ct-like tRNAs, transcribed from promiscuous ctDNA, constitute a prominent fraction of the wheat mitochondrial tRNA population. These observations strengthen the view that tRNAs of distinctly different evolutionary origin function together in translation in plant mitochondria.

MATERIALS AND METHODS Isolation of DNA from wheat mitochondria and wheat chloroplasts Wheat mtDNA was prepared as described previously [32]. Wheat ctDNA was isolated according to the protocol of Brookjans et al. [33], except that (i) the procedure was scaled down for 50 g of wheat leaves, and three additional low speed centrifugations were carried out to exclude contaminating mitochondria and nuclei from the chloroplast preparation, and (ii) chloroplast RNA was eliminated by digestion with RNase TI and RNase A, rather than by passage through a Sephacryl S-1000 column. Cloning and DNA sequencing Recombinant plasmids pH-S3/Fl and pH-C 1 were selected from a pUC9/HindIII clone bank of wheat mtDNA on the basis of their hybridization to a 32P-labeled probe of wheat mitochondrial 4S RNA, as described previously [2]. These clones contained Hindm inserts of 1.6 kbp (H-S3/Fl) and 4.8 kbp (H-Cl). H-S3/Fl was transferred directly to M13mpl9 [34] for production of deletions by the CycloneTM system [35] and subsequent sequencing by the dideoxy chain termination method [36] using the SequenaseTM system [37]. A 0.8 kbp Sau3AI fragment of H-Cl (containing the only tRNA gene on H-C 1, as deduced by hybridization with total mtRNA) was transferred into M13mpl9 in both orientations for sequencing from each end. An AccI subclone of H-Cl that overlapped the Sau3AI subclone was also used to provide sequence information. Each nucleotide position was determined at least twice from the same or opposite strand; both strands were sequenced over the regions containing the three tRNA

genes

discussed below.

Probe preparation and hybridizations For hybridization experiments, labeled DNA probes were synthesized by allowing the sequencing reactions [36] to proceed in the absence of dideoxynucleotides. Probes were produced from appropriate templates (deletions of H-S3/F1 and the Sau3AI subclone of H-C1) in which tRNA coding sequences were positioned immediately upstream of the site of initiation of probe synthesis. Total mitochondrial tRNAs were 3'-end-labeled with [5'-32P]pCp [38]. Total chloroplast nucleic acid (recovered prior to the stage of RNase treatment) was 3'-end-labeled in the same way and subjected to polyacrylamide gel electrophoresis. The region co-migrating with the 4S fraction of wheat mitochondrial RNA (run on the same gel) was excised, and the RNA was extracted [39] and used as a chloroplast 4S RNA probe. Southern hybridization experiments were performed using BIOTRANS nylon membranes according to the protocols provided by the supplier (ICN). All probes were heated in a boiling waterbath for 10 min and cooled on ice immediately before hybridization. 5462

Nucleic Acids Research AAGCTTTTCCCTTTCTATATATTCTTGAGAATAAGTAAATACCATCCTTTTTCCTATCAGCATCTCTCTA TCTGACAGTAAATATTGCTTTCAATCGATCAGGCACTAGGTTATTAAAGTATGGATATGCCATATAATAA GGGAAAAGGCCTCCAGTTTTTTTTAGCGCAAAAGAAAAGCCGATTCCAAATTTTCTATTTCCCGGCAAGT CTATGAGAGTCCTATTACGGATCCTCTGATCTTTAGCTTGAATGGTCGTGTTTTTATATAATTAGGTCTT TCTTCTGAGC.TCTTCTATTACGGTACGGAAAAATCAAACCTATCACTCTCAAAGAATGGGTAAAAAGAAT TCAAGCTAGGGGAGCCTACACTCTCTAACTACCTATACTACGCCCCCATTGATTGGAAACCCTCGGGGAA AATGATCCCATAAACAAAGGAATTATACAGTACGAAATAGCAAAAAAACAGACTAATTCGATTCTAAAAA ATCGATAATAGATATTTTATTCTGCTTAAATTGTCCCGACAAGGAGAGATATTCAATATTTGAATCAGAT

70 140 210 280 350

TGGATTTCGTAGTATACCAATGTGAGCCAAACTATTACTATTTCATCTAACTTGAATAACCAAGGACTGA ATTTGACTATGGATTCTGATAACTCGAGAAGTTTTTATTTGGTTATGATCCAAAAAGAAAAAAGAAAGGT TTAACAAGCCATGAGAAAATAGAGTAAAGTAACCATACGTTCTGTTTGCATAACGTGTATACGCCACGCC ATACAATCGAAATAGAAACATGGGACGATTCCAGAATTTTGAATAGATCCGTGGGGATGAGAGAAGTTGT

420 490 560 630 700 770 840

TGTTGAGAAATATGAAATTGGAAAGGAAGTCAAACTCCCTATGGAACCTGTGATCGGTTGTACCTGTACT

910

-l0 -35 TTAGTGATACGAAAACTCGCTATTCACTCAGTTTCTGGTCAATAATAAGATTATGT AGAGAGATGGCCG

980

-~~~~~~~~~~~ ~Ser AGTGGTTCAAGGCGTAGCATTGGAACTGCTATGTAGGCTTTTGTTTACCGAGGGTTCGAATCCCTCTCTT

1050

TTCTGTTAATTCACCAACGTTACCGACCACAATGAAATTTCTAGTAAGAG4FAAATCCGTCGACT

1120

TTCTAAATCGTGAGGGCGTCAAGTCCCTCTATCCCC4ItTAAAAAGCCCATTTTACTTCCTAACTATTGTACC

1190

TCCGAACTGACTTCTTTGGATCTTATCCCATTCATACAAATGAACATATTATAGTATAGGCAAGTAATCC CTATTATTAAGTAAGTCATTCACAATCCATATCATTATCCTGATATTTACTAAGTCCAATTTGAGAATAC

1260 1330 1400

TCC

. _ _----

%IpLeu AACCTCTATTTTTCATTAATGGTTCAAACAAGATTCACTATCTTTCTTATTCTACTCTTTCACAAACGGA __

__

-35

-10

TTTTTTCTTTTTGAGTCCCTTTAATTGACATAGATACAAGTACTCTACTAGGATGATGCACAATAAAT

TCAGGATAGCTCAGTTGGTAGAG!at C

1470

GGATTGAAAATCCTCGTGTCACCAGTTCAAATCTGGTTCCTGG 1540

AGAACGAACAATGAATGCCTTTTCGGGAAAGAAAAGGGCCACATATTTTTTTTAGGGAAAAGCTT

1608

Figure 1. Nucleotide sequence of the 1608 bp HindlII/HindIll fragment H-S3/F1, cloned from wheat mtDNA. Coding regions for the ct-like trnS-GGA (Ser) and trnF-GAA (Phe) genes are delineated by the solid lines, while the 3'-portion of the ct-like trnL-UAA gene (&Leu) is enclosed by the dashed line. The latter is likely a pseudogene, as suggested by the deletion of one of the two T residues within the highly conserved GTTC sequence normally characteristic of a functional tRNA sequence (underlined GTC). The dotted underline encompasses a junction that appears to be the result of a recombination event that has eliminated a 2.6 kbp portion of the region of wheat ctDNA from which H-S3/F1 was derived (see text and Fig. 4 and 5). The overlined '-10' and '-35' positions are equivalent to the putative promoter sequences upstream of the tmS-GGA and trnF-GAA genes in maize ctDNA

[46].

Isolation and sequence analysis of individual tRNA species Total wheat mitochondrial tRNA was selectively 3'-end-labeled by incubation with [a-32P]ATP in the presence of a tRNA nucleotidyltransferase preparation (isolated from wheat mitochondria and kindly supplied by P.J. Hanic-Joyce). Individual tRNAs were subsequently isolated by two dimensional polyacrylamide gel electrophoresis [40]. Labeled tRNAs were applied to a 10% gel (33x40x0.04 cm) containing 4 M urea and electrophoresed at 500 V and 4°C until the xylene cyanol marker had migrated 38 cm. A strip of gel encompassing the partially fractionated tRNA sample was then set into place on top of a 20% polyacrylamide gel (33 x40 x0.08 cm) containing 4 M urea, and the sample was electrophoresed at 500 V and 4°C for 160 h. Resolved tRNAs were located by autoradiography, extracted from the gel [39], and applied to a 15% polyacrylamide gel (33 x40 x0.04 cm) containing 7 M urea; electrophoresis was carried out at 1900 V and room temperature until the xylene cyanol had reached the bottom of the gel. By this procedure, some (apparently homogeneous) tRNAs from the second dimension gel were

5463

Nucleic Acids Research C ys t tt GATCGGA T *GCGGCATGGCCAAGCGGTAAGGCAGGGGACTGCAAATCCTTTATCCCCAGTTCAAATC

70

ACCAAAAAAATACTTAGGAT¶TCTTATTGTACTGATTGAACTTGACAAATTCTTGCCC

140

TGCATAAGCAAAGGAAAAGAiCTAGGCCTGTCGATACTGGATTTTACCATAGTTCTAGTTCTAAAAAAGA

210 280 350 420 490 560 630 700 770 804

GGTGTCGCC

CTGTCAATTTCTTCTTAAAAATAGGGCTCTGGCAGGGTTCTGGGTAGTGGCCCAAACCGATACCAATAGG GATGAGGGATATATACTTATGAATTTCATAATTGATTCTATTCCGCAATTCATAACTACGAAAGTAAGAG GTCGGAAATCTTTTATCCAAAGGTCCCTAAAGGACATTCCTTTTTTGCTCCTAGGATTGAAGAAGAGATT ATACGAAAGACTATCAAGACTTCCTAATTATCTGACACTACCTACACTAAAGTAAGGTCTACTGACTCTG TCTGGAATTAGATTGGATAGGCTGATGGGAAATTTAGGAGTTGTGGAAAGAAGAGACTTTCTTTCCATAC TTACGAGAGACTCCGAGTACTCAAACATTCAATCAGGATGGTCGGTTAGCTCTTATCTTTTAGAATAACT TAGTAAAAAAAAAGTAAAAGTAATAAAGAAAATGAATCTCCTAACTCGGTAGCCCATGAGGTTCCAATGG TTAGTTTTGGATTTTTTCAGGCAAGAAAGCAAAGAAGAAGGGAGGATTCTCTATTACCAACTTCCCCCAT TGTCCTGTCCTCATTGGTCATTCATCTTTCGATC

Figure 2. Nucleotide sequence of an 804 bp Sau3Al/Sau3Al fragment subcloned from the 4.8 kbp H-Cl HindIII restriction fragment of wheat mtDNA. The trnC-GCA gene (Cys) is enclosed by the solid line. Nucleotide differences within the corresponding region of wheat ctDNA [44] are shown as lower case letters above and below the mtDNA sequence. The vertical arrow marks the end of the published ctDNA sequence [44]. G G- C G- C A-T a

-T

A £-770

C

0

G-

8

G

C

0

IA

-C

A-

JIPAGCC G

TCAAGGcG

11111

GAGGG C-70 C 0TT

oil I

G_A

iTTT TT:~ GTAI@GT

-

-

c

so$I

T

TGGA

G

Ser

G- C C-G

10\

cGAACCG III G 20

TA

AA

~60

TGGGTCT IiIII

GTAAGGCA

C-40

-70

0

A-

TT-sO TT

Phe

T -

C

CGTG

G

T

A

ACCAGC

T-A

20

TGGTCT A

kT

1T1TcccT

A

A

CCCAGTTC TTATso * C- 40

30-e

-T

G

cA

Cys Figure 3. Potential secondary structures of the ct-like Cys (GCA), Ser (GGA) and Phe (GAA) tRNA sequences identified in wheat mtDNA. Differences between the mitochondrial and homologous maize (Ser, Phe) or wheat (Cys) chloroplast genes are boxed, with arrows pointing to the chloroplast residue.

5464

S~oL .-'~~ .. ..

.... ..

'

Nucleic Acids Research

wheat mtDNA

...

._..tobacco ctDNA

200 bp | _ l

g

lnuc8S

RT

~~~~~~~~~Oenothera mtDNA

1

S

rps4

Figure 4. Diagram showing portions of wheat mtDNA (H-S3/F1, Fig. 1) and Oenothera mtDNA [24] that are homologous to the region of tobacco ctDNA between positions 46,300 and 50,320 [42]. Sections of high sequence similarity (>70% identical) are delineated by the stippled blocks between wheat or Oenothera mtDNA and tobacco ctDNA. Coding regions are designated by black boxes: S, tRNAser; rps4, gene for ribosomal protein S4; T, tRNAThr; ORF70A, unidentified open reading frame; L, tRNAIeU (cross-hatching denotes an intron interrupting this gene); F, tRNAPhe. 'RT' and 'nuc 18S' designate unique regions in Oenothera mtDNA that encompass, respectively, a reverse transcriptase-like protein and a region homologous to the nuclear-encoded 1 8S rRNA gene [24]. Vertical arrows denote the positions of two directly repeated CAACAA motifs that are implicated in the recombinational formation of H-S3/Fl from the corresponding region of wheat ctDNA (see text and Fig. 5).

resolved into more than one species. Purified tRNAs were recovered as above [39], and after two ethanol precipitations were subjected to sequence analysis by both chemical and enzymatic degradation procedures, as previously described [41]. In the enzymatic procedure, anomalies in the migration of the smallest products [42] made it impossible to obtain the sequence of the first 6-9 residues from the 3'-end. In contrast, in the chemical sequencing procedure the sequence was readily determined starting at the penultimate position of the -CCA terminus. Usually, at least 40-50 nucleotides from the 3'-end could be determined by these procedures. RESULTS Sequence and organization of tRNA genes in H-S3/FJ and H-Cl Two tRNA genes were found in H-S3IF1, the complete sequence of which is shown in Fig. 1. These genes encode serine (trnS-GGA) and phenylalanine (trnF-GAA) isoacceptors. A cysteine tRNA gene (trnC-GCA) was found in a Sau3AI subfragment of H-Cl (Fig. 2). The potential secondary structures of these tRNA sequences (Fig. 3) are strictly conventional and contain the expected conserved primary and secondary structural motifs of a typical tRNA, as has proven to be the case for other plant mitochondrial tRNA sequences characterized to date (e.g. [2]). The three tRNA sequences identified in H-S3/F1 and H-Cl are much more similar to the corresponding chioroplast tRNA sequences than they are to the homologous eubacterial tRNA sequences (Table 1). Thus, they belong to the ct-like group of plant mitochondrial 5465

Nucleic Acids Research tobacco ctDNA 47,214 wheat mtDNA tobacco ctDNA

1,068 49,787

GGGCTCAAATAGCA ... ... *****CG****CG*1*C****kl*******C*A***...

47,247

CAACAArGAAATTTATCGTA...

49,820

...ACCAATTTTACTAACAACAA

1,101

...CATGTCAATACCGG

Figure 5. Putative recombination junction located between the tRNASer coding sequence and the tRNALeu pseudogene in the H-S3/F1 fragment of wheat mtDNA (dotted underline in Fig. 1). The box encloses two directly repeated CAACAA motifs present at different locations within the 46,300-50,320 region of tobacco ctDNA (vertical arrows in Fig. 4). The alignment demonstrates that positions 1068- 1101 of H-S3/FI show a pronounced similarity (* = identical nucleotides) with the tobacco sequence immediately upstream of the first CAACAA motif and immediately downstream of the second motif.

tRNA genes. In fact, the greater part of H-S3/F1 is homologous to a region between positions 46,300 and 50,320 in tobacco ctDNA (Fig. 4). In the latter, the trnS and trnF genes are much farther apart than they are in H-S3/F1, being separated by rps4, tmnT, and ORF70A genes and a split tRNALeU (trnL) gene [43]. In H-S3/F1, only the 3'-half of the trnL gene is present between the trnS and tmnF genes (Fig. 1 and 4). Examination of the tobacco ctDNA sequence [43] shows that the region that is effectively deleted in H-S3/F1 is flanked by short repeated sequences (CAACAA; positions 47,228-47,233 and 49,801-49,806 (Fig. 5)). Assuming that the corresponding region of wheat ctDNA Table 1. Percentage similarity between wheat mitochondrial Ser (GGA), Phe (GAA) and Cys (GCA) tRNA sequences and their homologs in other systems. Ser

Phe

Cys

90.3 97.7 98.9 98.9

94.5 95.9 95.9 98.6

94.4 91.5 90.9

Chloroplast Marchantia polymorpha (liverwort) spinach tobacco maize wheat

98.6

Eubacterial Bacillus subtilis Escherichia coli

68.2 64.0a

71.2 79.5

88.7 64.8

Mitochondrial Drosophila yakuba human Aspergillus nidulans yeast

30.1b 40.3b 50.9b 58.6b

38.8 41.7 57.5 59.3

38.8 43.8 40.3c 47.2

54.8 63.0

62.9

65.3

55.4

Nuclear yeast wheat

Archaebacterial Halobacterium volcanii

61.9

All sequences are listed in [49] except those from Marchantia chloroplast, which may be found in [53]. a S252 in [49] b UGA anticodon c C410 in [49]

5466

Nucleic Acids Research

ct

mt

mt

kbp 23 9.4

23

9.4

6.7

H-C

-*

ct kbp

6.7

43

463

2.3 20

2.3 2.0

H-S/FI -l

(A)

*

(B)

Figure 6. Autoradiograms showing the hybridization of 32P-labeled probes specific for tRNACYS (A) and tRNAPhe (B) to HindIII restriction digests of wheat mtDNA (mt) and wheat ctDNA (ct). Numbers refer to the sizes (kbp) of marker HindlIl fragments of X DNA. The result shown in (B) was duplicated in a separate experiment with a tRNASer-specific probe.

contains similar direct repeats, recombination between them (a process that is common in plant mtDNA [44]) could have eliminated a 2.6 kbp region of promiscuous ctDNA after its incorporation into the wheat mitochondrial genome, thus accounting for the structure of H-S3/F1. The same trnS-GGA gene, differing from the wheat mitochondrial gene described here by a single nucleotide (T rather than C at position 21), has been found in primrose (Oenothera berteriana) mtDNA, also within a piece of promiscuous ctDNA [24]. In this case, however, the homologies to tobacco ctDNA are different than in the case of H-S3/F1, and include about half of the rps4 gene (Fig. 4). From a comparison of sequences common to wheat mtDNA, Oenothera mtDNA, and tobacco ctDNA in the vicinity of the trnS gene, it is not immediately apparent whether the plant mitochondrial trnS-GGA region was acquired from ctDNA before the divergence of monocotyledons and dicotyledons, subsequently undergoing different rearrangmenents in the two lineages, or whether this tRNA gene was contributed by the chloroplast genome of higher plants in two separate events. Determination of the relevant wheat and Oenothera ctDNA sequences might help to decide between these two scenarios. The sequence of the wheat chloroplast trnC-GCA gene and flanking regions has been published [45] and part is shown in Fig. 2, together with the sequence of the wheat mitochondrial trnC-GCA gene and flanks. Only limited sequence is available for comparison outside of the gene itself; nevertheless, it appears that this region of wheat mtDNA has also been derived from ctDNA. A trnC-GCA gene identical to that in wheat mtDNA has recently been found in maize mtDNA [17]. Homology between wheat and maize mtDNA continues in the downstream 104 nucleotides (which are 94.2% similar), after which evident sequence identity abruptly disappears. Although an origin of H-S3/F1 and H-Cl from ctDNA is clearly indicated, sequence comparisons with ctDNA argue against the possibility that these restriction fragments arise from contaminating ctDNA in the wheat mtDNA preparation used for cloning. Further evidence against this possibility comes from hybridization experiments using the cloned tRNA genes as probes. In each case, the probe hybridized to the expected HindI fragment of wheat mtDNA, but to a different-sized HindIll fragment of wheat ctDNA (Fig. 6). 5467

Nucleic Acids Research L:NA4 M

42--

A

c tt

rt t

r^. ut

ct

r

aP

--

_ 4m~~

s. ~~~~~~~~~~~~~~~01

(A)

(B)

(C)

(D)

Figure 7. Autoradiograms showing the hybridization of [a-32P]ATP-labeled (via -CCA addition) wheat 4S RNA with wheat organellar DNAs: mt, mitochondrial; ct, chloroplast. Triangles in (B) and (D) point to the positions of H-S3/F1 (open) and H-Cl (closed) (1.6 and 4.8 kbp, respectively). Solid circles in (D) denote hybridization signals that probably derive from trace contamination of the mtDNA preparation with ctDNA. Arrow in (C) indicates the region of the ctDNA restriction profile that shows the most intense hybridization with chloroplast 4S RNA. Numbers refer to the sizes (kbp) of marker HindlIl fragments of X DNA.

These results also suggest that the three tRNA genes described here are single copy genes in wheat mtDNA. Chloroplast-like tRNAs in the wheat mitochondrial tRNA population The tRNA-containing clones described here were selected on the basis of their hybridization to a highly purified 4S RNA fraction prepared from isolated wheat mitochondria. This suggests that the three genes are expressed. However, the wheat mitochondrial trnF-GAA and trnS-GGA genes differ by only one and two nucleotides, respectively, from the corresponding maize chloroplast tRNA genes [46] (Fig. 3), while the wheat mitochondrial trnC-GCA gene is identical to its wheat chloroplast counterpart [45] except at a single nucleotide position (Fig. 3). Thus, it is possible that these genes are not functional, but that they were detected by the corresponding chloroplast tRNAs present as significant contaminants of the wheat mitochondrial tRNA probe. Both the organelle purification procedure itself and the use of embryos rather than seedling tissue should ensure a low level of contamination of the mitochondrial fraction by chloroplasts or chloroplast-derived nucleic acids. Supporting this expectation is the fact that chloroplast rRNA (23S, 16S or 5S) was not detected in previous studies of wheat mitochondrial RNA prepared by this procedure [47,48]. To further assess the question of cross-contamination, we carried out experiments in which either wheat chloroplast or wheat mitochondrial 4S RNA was used as a hybridization probe against HindIll digests of either wheat ctDNA or mtDNA. As shown in Fig. 7, wheat mitochondrial 4S RNA hybridized surprisingly poorly to 5468

Nucleic Acids Research 10Di

120:. 36'251

41 48

28 39.

80

/

@7

42O

18k*31A6

2

2

35

Figure 8. Autoradiogram showing the two dimensional electrophoretic separation of wheat mitochondrial tRNAs specifically labeled with [ax-32P]ATP in the presence of tRNA nucleotidyltransferase isolated from wheat mitochondria. Conditions of electrophoresis (partially denaturing in each dimension) are specified in the text. Material in the numbered regions was recovered for further electrophoretic resolution under fully denaturing conditions (see text and Figure 9).

ctDNA (A). This same result was obtained in two separate experiments with different preparations of wheat mitochondrial 4S RNA, one 3'-end-labeled with [5'-32P]pCp in the presence of RNA ligase and the other labeled with [ct-32P]ATP through nucleotidyltransferase-mediated -CCA addition. The homologous hybridization patterns (mt 4S RNA vs. mtDNA (B); ct 4S RNA vs. ctDNA (C)) are distinct from one another, with major bands in the mtDNA blot (B) clearly different from major bands in the ctDNA blot (C). [Note that those ctDNA restriction fragments that do hybridize with mitochondrial 4S RNA (A) are not the ones that hybridize most strongly with chloroplast 4S RNA (C).] Chioroplast 4S RNA detected both H-S3/Fl and H-Cl in mtDNA (arrowheads in (D)), but hybridization was less intense than when mitochondrial 4S RNA was used as probe (compare (B) and (D)). On the other hand, the chioroplast probe did reveal what appears to be a slight contamination of mtDNA with ctDNA, as evidenced in the mtDNA track by faint bands (solid circles, (D)) that correspond to prominent bands in the ctDNA lane (compare (C) and (D)). This ctDNA contamination was not apparent when mitochondrial 4S RNA was used as probe (compare (B) and (C)); in particular, no detectable hybridization of mitochondrial 4S RNA is evident in the vicinity of those ctDNA restriction fragments that hybridize most intensely with chioroplast RNA (arrow). Taken together, these results indicate that contaminating chioroplast tRNA does not contribute in a significant way to the hybridization of wheat mitochondrial 4S RNA with 5469

Nucleic Acids Research C0N

~ ~

_

~

~

~

~

~

OCO40NC.4~ ~

r

N;

-

Mm~~~~~~~~~ 0

.cQ

LO

'c

Ne

-.

Figure 9. Autoradiogram showing the further resolution obtained when tRNAs recovered after two dimensional electrophoresis (see Fig. 8) were subjected to a third electrophoretic run in a fully denaturing polyacrylamide gel. Separated components are designated a, b, c, etc., in order of increasing mobility.

restriction fragments of wheat mtDNA. Isolation and sequence analysis of chloroplast-like tRNA species present in wheat mitochondria In order to obtain definitive evidence that the ct-like tRNA genes described here are expressed in wheat mitochondria, we isolated and sequenced mature tRNA species that correspond to each of these genes. Initially, [5'-32P]pCp-labeled tRNAs were isolated by hybridization to individual HindIll restriction fragments of wheat mtDNA. In this way, one (H-Cl) or two (H-S3/F) RNAs of the expected size were selected. However, each of these was heterogeneous in length, a reflection of the metabolic lability of the 3'-CCA terminus. Subsequently, the tRNA population was specifically labeled with [a-32P]ATP via nucleotidyltransferase-mediated 3 '-CCA addition (which eliminated 3 '-end heterogeneity). Labeled tRNAs were then separated by a two dimensional polyacrylamide gel electrophoretic procedure [40]. The resulting autoradiographic pattern (Fig. 8) proved to be highly reproducible from run to run. After recovery from individual spots or clusters (as indicated in Fig. 8), tRNAs were subjected to an additional electrophoretic separation (Fig. 9). Some tRNAs were homogeneous at this stage, whereas others were further resolved into two or more bands (designated a, b, c, etc.). Some clusters from the partially denaturing two-dimensional gel (e.g., # 15 and #22, Fig. 8) migrated as single species in the fully denaturing 'third dimension' gel (Fig. 9), suggesting that some of the complexity of the initial pattern (Fig. 8) may have been due to the separation of metastable conformers of individual tRNA species. Table 2 lists those tRNAs identified to date among the 65 separated species that we have subjected to RNA sequence analysis. In several cases, different versions of the same tRNA were resolved by gel electrophoresis; these did not differ perceptibly over the region sequenced, so they may have been resolved on the basis of conformation or possibly aminoacylation status (because a relatively crude tRNA nucleotidyltransferase was used in the present study, it is possible that this preparation also contained aminoacyl-tRNA synthetases capable of charging some -CCA-end-labeled tRNAs). Alternatively, the multiple species may differ in primary sequence and/or post-transcriptional modification within the unsequenced 5'-terminal region (e.g. [6]). With the exception of a tRNAMSP, which was not detected in the present study, we found species corresponding to all of the native wheat mitochondrial tRNA genes we have previously characterized [2,3,15,16,19]. In addition, we identified abundant species corresponding to the ct-like serine, cysteine and phenylalanine tRNA genes described here. 5470

Nucleic Acids Research Table 2. Wheat mitochondrial tRNAs separated by polyacrylamide gel electrophoresis (see Fig. 8 and 9) and identified by RNA sequence analysis. Amino Acid

Anticodon

Number

E

C

Gene

[ref.] Chloroplast-like (mtDNA-encoded)

Asn

Cys

GUU GCA

eMet

CAU

Phe Ser Trp

GAA GGA CCA

Gln

UUG

49b 12b 49c 7 46b 9 1 46a

+ + + + + + + +

30 31 43 17 26 49a 16 32 19 34 45 13 2 3 22 47b

+ + + + + + + +

+ + +

+

lOa lOb 6 15

+

+

+ +

+ +

+

+

b

+

[27]C

+ +

b b

[23]

Native (mtDNA-encoded) uu ~UUC GluGlu UUU Lys eMet

CAU

fMet

CAU

Pro Ser

UGG GCU

Ser Tyr

UGA GUA

Gly

GCC

[15] +

+ +

+

+ + + +

[25]a

+ +

[ 14jd [19]

[4]e [3]

+

+

+

[2,16]

[2] [2]

[2]

Cytoplasnic (nDNA-encoded) Leu

Val

CAA GAC

Numbering is that used in Fig. 8 and 9. E = sequence obtained by enzymatic procedure; C = sequence obtained by chemical procedure. a In lupine mtDNA b This report c In soybean and Arabidopsis mtDNA d In soybean mtDNA e In maize mtDNA

Sequence analysis also revealed a ct-like tRNArrP (corresponding to the wheat mitochondrial gene reported by Marechal et al. [23]) as well as two other ct-like tRNAs homologous to sequenced genes from other plant mtDNAs: asparagine (lupine mtDNA [25]) and elongator methionine (soybean and Arabidopsis mtDNA [27]). A soybean probe containing the latter gene (kindly provided by H. Wintz) gave a strong hybridization signal with wheat mtDNA (unpublished results), supporting the expectation that a ct-like elongator tRNAMet gene is encoded by the wheat mitochondrial genome. The corresponding 5471

Nucleic Acids Research Table 3. Positions which distinguish between mtDNA-encoded ct-like tRNA genes and their ctDNA-encoded counterparts DNA

Cys (GCA) Phe (GAA Ser (GGA) Trp (CCA)

eMet (CAU)

66 32 16 50 28 30 44 9 16 28 41 43

wheat mt DNA RNA

wheat ct maize ct DNA DNA

T T T G T T A

C

U U

C C A

G U A G C U/Ca

C G G A T T

U/Ca

C

A

G

All chloroplast sequences are listed in [49] with the exception of that for wheat chloroplast tRNATrP [23]. RNA sequence analysis was ambiguous at these positions

a

elongator tRNAMet has also been isolated from bean mitochondria [11]. Table 3 summarizes those positions that differentiate five wheat mitochondrial, ct-like tRNAs from their authentic chloroplast homologs (from wheat or maize). Included in the table are data from RNA sequence analysis, with relevant portions of RNA sequencing gels for the serine (A), phenylalanine (B) and cysteine (C) acceptors presented in Fig. 10. It is clear from these results that the ct-like tRNAs in wheat mitochondria correspond to the mitochondrial and not the chloroplast genes. The case is particularly strong for the cysteine and tryptophan tRNAs, where the wheat mitochondrial and chloroplast genes are both available for comparison. The trnF coding sequence differs from that of the maize chloroplast gene at one position (residue 32), where there is a T in the wheat mitochondrial sequence but a C in the maize chloroplast one (Fig. 3). This difference is reflected in the structure of the wheat mitochondrial tRNAPhe characterized here (Fig. 10). Although the sequence of the wheat chloroplast trnF gene is not known, it is highly unlikely that there is a T at position 32, because this residue is a C in all other chloroplast tmF-GAA sequences that have been determined to date [49]. It is worth emphasising that among the collection of wheat mitochondrial tRNAs (65 in all) that we have so far characterized by direct sequence analysis, there are no examples of authentic chloroplast species that could represent either contaminants or imported species.

DISCUSSION We have reported here the characterization of three ct-like tRNA genes encoded by wheat mtDNA, and have presented evidence that these genes are transcribed. We have found and partially sequenced abundant wheat mitochondrial tRNAs corresponding to each gene, and differing from the homologous gene present in wheat or maize ctDNA. In addition, in the course of this study we have identified three other ct-like tRNAs (tryptophan, elongator methionine and asparagine) in the wheat mitochondrial tRNA population. In these cases, also, direct RNA sequence analysis supports the idea that the tRNAs are transcribed from mitochondrial (but ct-like) genes. From the structure of the trnC-GCA, trnS-GGA, tmF5472

Nucleic Acids Research GAUC

277

GA UC

U',U

3_2

-

*

A,-* A4

38

-

G

A-*

(A)

(B) C--

(C

Figure 10. Autoradiograms of portions of sequencing gels of wheat mitochondrial Ser (A), Phe (B) and Cys (C) tRNAs. Arrows highlight residues (circled) that distinguish these mitochondrial tRNAs from their chloroplast counterparts. Numbering corresponds to that in Fig. 3. Enzymatic sequencing was carried out in (A) and chemical sequencing in (B) and (C) (see [41]). The faint U at position 32 in (B) was also seen in enzymatic sequencing gels; the resiudue at this position may therefore be a modified U (e.g., X).

GAA (this report) and trnW-CCA [23] coding and flanking sequences, it is clear that these been incorporated into the wheat

genes are parts of promiscuous ctDNA that has mitochondrial genome in the course of evolution.

The implication of these and previous [ 1 1,17,31 ] results is that chloroplast tRNA genes have not only been transferred to plant mtDNA, but have also been recruited to function there. While it formally remains to be demonstrated that the mitochondrial, ct-like tRNAs that we and others [11,31] have characterized are functional in protein synthesis in plant mitochondria, the indirect evidence currently available strongly suggests this is the case. These tRNA species are mature, in that they contain modified nucleoside constituents ([11,31] and our unpublished results) and (as shown here) are substrates for 3'-terminal Table 4. Distribution of potentially functional chloroplast-like tRNA genes in the mtDNA of various plants. Asn

Arabidopsis lupine

Cys

His

Phe

Ser

Trp

D

D D R

D,R

D,R

D D

maize Oenothera bean

D

D

R D

soybean wheat

eMet

Da,R

D,R

R,H

D,R

D = from DNA sequences analysis; R = from RNA sequence analysis; H = from hybridization with a heterologous probe (soybean). a Preliminary sequence analysis has now identified this gene (P.B.M. Joyce, unpublished results)

5473

Nucleic Acids Research -CCA addition by a tRNA nucleotidyltransferase isolated from wheat mitochondria. They are present at levels that approximate those of tRNA species originating from native mitochondrial tRNA genes. With the exception of a separate, non-ct-like elongator tRNAMet, we have so far encountered no other tRNAs specific for asparagine (GUU), cysteine (GCA), phenylalanine (GAA), serine (GGA) and tryptophan (CCA) in the wheat mitochondrial tRNA population. We do not know yet whether the six ct-like tRNA sequences discussed here constitute the complete set of such (potentially functional) genes encoded by wheat mtDNA. With the exception of a ct-like tRNAHiS gene in maize mtDNA [22] (whose possible presence in wheat mtDNA we have not yet examined), no other ct-like tRNA genes have been reported in other plants. On the other hand, each of these six tRNAs/tRNA genes has been found in at least one other plant species, and the elongator methione and tryptophan tRNAs have each been found in a total of four plants, both monocotyledons and dicotyledons (Table 4). A systematic survey and detailed analysis of ct-like tRNA genes in plant mtDNA would be very useful in assessing how universal this phenomenon is, whether the same tRNA gene has been recruited independently in different plant lineages, and when in evolutionary time these transfers have occurred. Such studies would presumably indicate whether some ct-like tRNA genes were transferred (and acquired function) before the divergence of mono- and dicotyledons. The existence of two distinct classes of tRNA genes in plant mtDNA raises further questions about their expression and their joint participation in mitochondrial protein synthesis. Although the upstream flanking regions of at least some of the ct-like tRNA genes have retained the typical ' -10' and '-35' motifs (see Fig. 1) found upstream of many authentic chloroplast tRNA genes [46], it is unlikely that these putative regulatory sequences are used in the mitochondrion. Sequence comparisons have so far failed to reveal any eubacteria-like regulatory signals flanking native wheat mitochondrial genes [2], while transcript mapping experiments have not identified obvious candidates for precursors of any of the ct-like tRNA genes found in wheat mtDNA (T.Y.K. Heinonen and P.S. Covello, unpublished observations). For the most part, these and the native tRNA genes are solitary in the wheat mitochondrial genome, with no evident physical or co-transcriptional linkage to each other or other genes. Possibly they are processed from long precursors transcribed from remotely located promoters. In this regard, we note that artificial transcripts containing the ct-like tRNA sequences are correctly processed by a wheat mitochondrial extract that also mediates the accurate and efficient processing of transcripts of native tRNA genes (P.J. Hanic-Joyce and M.W. Gray, in preparation). We have previously suggested that use by the mitochondrial translation system of tRNAs of distinctly different genetic origins may place constraints on the divergence of mitochondrially encoded tRNA sequences [50]. Such constraints might account in part for the decidedly normal structures of plant mitochondrial tRNAs, compared with their counterparts in many other mitochondrial systems. The translation system in plant mitochondria may be further constrained by its use of yet a third, genetically distinct class of tRNAs, encoded in the nucleus and imported from the cytoplasm. Evidence in support of the import of cytoplasmic-type tRNAsLeu into bean mitochondria has been presented [51,52], and by RNA sequence analysis in the present study we have identified this same cytoplasmic-like tRNAL"U, as well as ones for glycine and valine, in wheat mitochondria. While rigorous control experiments need to be carried out in our case to exclude cytoplasmic contamination, it is noteworthy that tRNAs for these particular amino acids are not 5474

Nucleic Acids Research represented among the set of six ct-like and ten native tRNA genes so far identified in wheat mtDNA. Our working hypothesis, based on available evidence, is that the wheat mitochondrial tRNA population is comprised of three distinct classes, two of which are encoded by the mitochondrial genome and the third by the nuclear genome. How this situation arose is a most intriguing evolutionary question whose answer will require much further comparative analysis. In the meantime, we are continuing to work toward a complete characterization (by both DNA and RNA sequence analysis) of the wheat mitochondrial tRNA population.

ACKNOWLEDGMENTS We thank P.J. Hanic-Joyce for a generous gift of tRNA nucleotidyltransferase from wheat mitochondria and for advice on conditions for its use; P.H. Boer for the wheat mitochondrial HindIH clones analyzed in this study; and H. Wintz for providing a clone containing the soybean mitochondrial elongator tRNAMet gene [27]. We also thank Lisa Laskey for her expert assistance in preparing this manuscript for publication. Salary support in the form of a Medical Research Council of Canada Studentship and a Walter C. Sumner Memorial Fellowship to P.B.M.J. is gratefully acknowledged, as is the grant of operating funds that supported this research (MT-4124 from the Medical Research Council of Canada to M.W.G.). M.W.G. is a Fellow in the Evolutionary Biology Program of the Canadian Institute for Advanced Research. *To whom correspondence should be addressed

REFERENCES 1. 2. 3. 4. 5. 6. 7.

8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25.

Sederoff,R.R. (1984) Adv. Genet. 22, 1-108. Joyce,P.B.M., Spencer,D.F., Bonen,L. and Gray,M.W. (1988) Plant Mol. Biol. 10, 251-262. Gray,M.W. and Spencer,D.F. (1983) FEBS Lett. 161, 323-327. Parks,T.D., Dougherty,W.G., Levings,C.S.lII and Timothy,D.H. (1984) Plant Physiol. 76, 1079-1082. Gottschalk,M. and Brennicke,A. (1985) Curr. Genet. 9, 165-168. Marechal,L., Guillemaut,P. and Weil,J.-H. (1985) Plant Mol. Biol. 5, 347-351. Marechal,L., Guillemaut,P., Grienenberger,J.-M., Jeannin,G. and Weil,J.-H. (1985) FEBS Lett. 184, 289-293. Parks,T.D., Dougherty,W.G., Levings,C.S.III and Timothy,D.H. (1985) Curr. Genet. 9, 517-519. Bartnik,E. and Borsuk,P. (1986) Nucleic Acids Res. 14, 2407. Borsuk,P., Sirko,A. and Bartnik,E. (1986) Nucleic Acids Res. 14, 7508. Marechal,L., Guillemaut,P., Grienenberger,J.-M., Jeannin,G. and Weil,J.-H. (1986) Plant Mol. Biol. 7, 245-253. Grabau,E.A. (1987) Curr. Genet. 11, 287-293. Runeberg-Roos,P., Grienenberger,J.M., Guillemaut,P., Marechal,L., Gruber,V. and Weil,J.H. (1987) Plant Mol. Biol. 9, 237-246. Wintz,H., Chen,H.-C. and Pillay,D.T.N. (1987) Nucleic Acids Res. 15, 10588. Joyce,P.B.M. and Gray,M.W. (1988) Nucleic Acids Res. 16, 1210. Joyce,P.B.M., Spencer,D.F. and Gray,M.W. (1988) Plant Mol. Biol. 11, 833-843. Wintz,H., Grienenberger,J.-M., Weil,J.-H. and Lonsdale,D.M. (1988) Curr. Genet. 13, 247-254. Izuchi,S.-i. and Sugita,M. (1989) Nucleic Acids Res. 17, 1248. Joyce,P.B.M. and Gray,M.W. (1989) Biochim. Biophys. Acta, in press. Gray,M.W. and Doolittle,W.F. (1982) Microbiol. Rev. 46, 1-42. Dron,M., Hartmann,C., Rode,A. and Sevignac,M. (1985) Nucleic Acids Res. 13, 8603-8610. Iams,K.P., Heckman,J.E. and Sinclair,J.H. (1985) Plant Mol. Biol. 4, 225-232. Marechal,L., Runeberg-Roos,P., Grienenberger,J.M., Colin,J., Weil,J.H., Lejeune,B., Quetier,F. and Lonsdale,D.M. (1987) Curr. Genet. 12, 91-98. Schuster,W. and Brennicke,A. (1987) EMBO J. 6, 2857-2863. Karpidska,B. and Augustyniak,H. (1988) Nucleic Acids Res. 16, 6239.

5475

Nucleic Acids Research 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43.

44.

45. 46. 47. 48. 49.

50. 51. 52. 53.

Schuster,W., Grienenberger,J.M., Weil,J.-H. and Brennicke,A. Nucleic Acids Res. 16, 7737. Wintz,H., Chen,H.-C. and Pillay,D.T.N. (1988) Curr. Genet. 13, 255-260. Ellis,J. (1982) Nature 299, 678-679. Stern,D.B. and Lonsdale,D.M. (1982) Nature 299, 698-702. Stern,D.B. and Palmer,J.D. (1984) Proc. Natl. Acad. Sci. U.S.A. 81, 1946- 1950. Marechal,L., Guillemaut,P., Grienenberger,J.-M., Jeannin,G. and Weil,J.-H. (1985) Nucleic Acids Res. 13, 4411 -4416. Bonen,L. and Gray,M.W. (1980) Nucleic Acids Res. 8, 319-335. Brookjans,G., Stummann,B.M. and Henningsen,K.W. (1984) Anal. Biochem. 141, 244-247. Yanisch-Perron.C., Vieira,J. and Messing,J. (1985) Gene 33, 103-119. Dale,R.M.K., McClure,B.A. and Houchins,J.P. (1985) Plasmid 13, 31-40. Sanger,F., Nicklen,S. and Coulson, A.R. (1977) Proc. Natl. Acad. Sci. U.S.A. 74, 5463-5467. Tabor,S. and Richardson,C.C. (1987) Proc. Natl. Acad. Sci. U.S.A. 84, 4767-4771. Peattie,D.A. (1979) Proc. Natl. Acad. Sci. U.S.A. 76, 1760-1764. Schnare,M.N., Heinonen,T.Y.K., Young,P.G. and Gray,M.W. (1985) Curr. Genet. 9, 389-393. Fradin,A., Gruhl,H. and Feldmann,H. (1975) FEBS Lett. 50, 185-189. Schnare,M.N., Spencer,D.F. and Gray, M.W. (1983) Can. J. Biochem. Cell. Biol. 61, 38-45. Winter,G. and Brownlee,G.G. (1978) Nucleic Acids Res. 5, 3129-3139. Shinozaki,K., Ohme,M., Tanaka,M., Wakasugi,T., Hayshida,N., Matsubayasha,T., Zaita,N., Chunwongse,J., Obokata,J., Yamaguchi-Shinozaki,K., Ohto,C., Torazawa,K., Meng,B.Y., Sugita,M., Deno,H., Kamogashira,T., Yamada,K., Kusuda,J., Takaiwa,F., Kata,A., Tohdoh,N., Shimada,H. and Sugiura,M. (1986) Plant Mol. Biol. Rep. 4, 111-147. Lonsdale,D.M., Brears,T., Hodge,T.P., Melville,S.E. and Rottmann,W.H. (1988) Phil. Trans. R. Soc. Lond. B, 319, 149-163. Quigley,F., Grienenberger,J.M. and Weil,J.H. (1985) Plant Mol. Biol. 4, 305-310. Steinmetz,A.A., Krebbers,E.T., Schwarz,Z., Gubbins,E.J. and Bogorad,L. (1983) J. Biol. Chem. 258, 5503 -5511. Spencer,D.F., Bonen,L. and Gray,M.W. (1981) Biochemistry 20, 4022-4029. Schnare,M.N. and Gray,M.W. (1982) Nucleic Acids Res. 10, 3921-3932. Sprinzl,M., Hartmann,T., Meissner,F., Moll,J. and Vorderwulbecke,T. (1987) Nucleic Acids Res., Sequences Supplement 15, r53-rl88. Boer,P.H. and Gray, M.W. (1988) Curr. Genet. 14, 583-590. Green,G.A., Marechal,L., Weil, J.-H. and Guillemaut,P. (1987) Plant Mol. Biol. 10, 13-19. Marechal-Drouard,L., Weil,J.-H. and Guillemaut,P. (1988) Nucleic Acids Res. 16, 4777-4788. Ohyama,K., Fukuzawa,H., Kohchi,T., Shirai,H., Sano,T., Sano,S., Umesono,K., Shiki,Y., Takeuchi,M., Chang,Z., Aota,S.-i., Inokuchi,H. and Ozeki,H. (1986) Plant Mol. Biol. Rep. 4, 149-175.

This article, submitted on disc, has been automatically converted into this typeset format by the publisher. 5476