Curr Genet (2002) 41: 357–365 DOI 10.1007/s00294-002-0315-x
R ES E AR C H A RT I C L E
Charles Hedgcoth Æ Ahmed M. El-Shehawi Æ Ping Wei Melissa Clarkson Æ Dimitri Tamalis
A chimeric open reading frame associated with cytoplasmic male sterility in alloplasmic wheat with Triticum timopheevi mitochondria is present in several Triticum and Aegilops species, barley, and rye Received: 29 July 2001 / Revised: 15 May 2002 / Accepted: 22 May 2002 / Published online: 12 July 2002 Springer-Verlag 2002
Abstract Mitochondrial DNA from Triticum timopheevi has a chimeric gene, orf256, upstream of coxI. This gene is cotranscribed with coxI in cytoplasmic male sterile plants and produces a 7-kDa protein which is not produced in fertile or fertility-restored plants. T. aestivum, the nuclear donor in sterile plants, does not have orf256. Analysis by polymerase chain reaction of DNA from barley, rye, Aegilops bicornis, Ae. searsii, Ae. sharonensis, Ae. speltoides, Ae. tauschii, T. monococcum, and T. turgidum was done with oligonucleotide primers designed to detect orf256 or coxI sequences. Except for T. turgidum, these plants have various elements of the orf256 sequence over a 1-kb length of DNA immediately upstream of coxI in exactly the same arrangement as is found in the coxI region of T. timopheevi. Only T. timopheevi and Ae. speltoides have orf256 transcripts, and only cytoplasmic male-sterile plants involving these two species as maternal donors produce a protein from orf256. Part of an orf256-like sequence is present in T. turgidum but is at least slightly different in arrangement relative to coxI, as compared with the sequence in T. timopheevi. Neither maize nor sorghum have the orf256 sequence. Keywords Wheat Æ Mitochondrial DNA Æ coxI Æ orf256
Introduction Wheat cytoplasmic male sterility (cms) results from interspecific crosses, specifically the introduction of cytoplasm from Triticum timopheevi into T. aestivum (Wilson and Ross 1962). The cms trait probably results from
Communicated by D.R. Wolstenholme C. Hedgcoth (&) Æ A.M. El-Shehawi Æ P. Wei Æ M. Clarkson D. Tamalis Department of Biochemistry, Willard Hall, Kansas State University, Manhattan, Kansas 66506-3702, USA E-mail:
[email protected]
incompatibility between T. timopheevi mitochondria and the T. aestivum nucleus, as both parents are fertile (Leaver et al. 1988). Plants are restored to fertility by introducing nuclear restoration genes derived from T. timopheevi (Mann et al. 1984). Studies of the coxI gene regions in mitochondrial DNAs (mtDNA) from T. aestivum, T. timopheevi, cms lines (T. aestivum nucleus, T. timopheevi mitochondria), and fertility-restored lines (Rathburn and Hedgcoth 1991; Rathburn et al. 1993) revealed the presence of a DNA sequence in T. timopheevi mtDNA between the presumptive transcriptional promoter for coxI and its translational start site. This sequence is not present in T. aestivum mtDNA. The sequence consists of an open reading frame, orf256, which is transcriptionally functional in T. timopheevi mitochondria (Song and Hedgcoth 1994a). orf256 is expressed as a 7-kDa protein only in cms lines of wheat (Song and Hedgcoth 1994b). The –1 to –228 5¢ flanking sequence of orf256 is identical to the analogous region from coxI of T. aestivum. The first 11 amino acids encoded in orf256 are the same as the first 11 amino acids of COXI. The orf256 gene region appears to have been formed by a combination of 261 bp of the coxI gene region (228 bp of 5¢ flanking region, 33 bp of encoded N-terminus) and a segment of DNA from another source. A part of the 5¢ flanking region of the rice coxI gene (Suzuki et al. 1991) has 67% homology to orf256 but does not have an orf256-like open reading frame nor a chimeric segment from coxI. A closer relative to T. timopheevi, Aegilops speltoides, also gives male-sterile progeny and a 7-kDa protein (Song and Hedgcoth 1994b) when serving as a cytoplasmic donor in crosses with T. aestivum. In contrast to T. aestivum mtDNA, which does not hybridize with orf256 probe (Rathburn et al. 1993), Ae. speltoides appears to have a sequence equivalent to orf256 and rice has either vestiges of an orf256 nucleotide sequence or a nucleotide sequence which, later in evolutionary time, became orf256 in other species. T. aestivum is hexaploid with a genome constitution of AABBDD and was formed about 8,000 years ago
358
from a hybridization between T. turgidum (AABB) and Ae. tauschii (DD) (Friebe and Gill 1996; Kimber and Sears 1987). The A genome originated with T. urartu (AA), which is closely related to T. monococcum (AA). Although the exact origin of the B genome has not been determined, either because the donor is extinct or has not been found, the S genome of Ae. speltoides (SS) is the closest identified genome. Similarly, the A genome of T. timopheevi (AAGG) is thought to derive from T. urartu with the G genome originating from an ancestor like Ae. speltoides. Ae. speltoides, Ae. bicornis, Ae. searsii, and Ae. sharonensis appear to have diverged from a common ancestor at about the same time less than 9 million years ago (mya) (Sasanuma et al. 1996). The grass family (Poaceae) diverged about 50–80 mya into the subfamilies Pooideae (tribe Triticeae containing wheat, barley, rye, Aegilops sp.), Panicoideae (tribe Maydeae containing maize), and Bambusoides (tribe Oryzeae containing rice) (Clark et al. 1995; Doebley et al. 1990; Wolfe et al. 1989). Maize and sorghum diverged about 16.5 mya (Gaut and Doebley 1997); wheat and barley diverged about 10–15 mya (Wolfe et al. 1989), with wheat and rye diverging about 7 mya (P. Gornicki, personal communication). The cytoplasms of T. aestivum, T. timopheevi, and T. turgidum originate from an ancestor like Ae. speltoides (Tsunewaki 1996). Because of: (1) the close evolutionary history of T. aestivum and T. timopheevi, (2) the absence of orf256 in the mitochondrial DNA of T. aestivum, its presence in T. timopheevi, and the presence of a related sequence in rice, (3) the interesting transcriptional and translational characteristics of orf256 depending on the source of the nucleus and the relationship to cytoplasmic male sterility, and (4) the lack of a known function for orf256, we were interested in determining whether close relatives of these two plants and more distant cereals might have sequences related to orf256. We used the polymerase chain reaction (PCR) technique to probe for the presence and organization of orf256-like sequences in different grains, wheat progenitors, and wheat relatives. We found that barley, rye, and several species of Aegilops and Triticum have orf256-like sequences, possibly including the translational start codon, in the 5¢ flanking region of coxI. Only T. timopheevi and Ae. speltoides appear to express orf256 as an RNA product and neither of these as fertile plants expresses a corresponding protein product. Maize and sorghum, like T. aestivum, do not have orf256-like sequences responding to primers used in the PCR analysis nor to an orf256 hybridization probe.
auchira, speltoides ligustica, tauschii, Secale cereale (rye), Oryza sativa (rice)], David Johnston (Cargill Hybrid Seeds; cms wheat), S. Muthukrishnan [Department of Biochemistry; Hordeum vulgare (barley)], Charles Niblett (Department of Plant Pathology, University of Florida; T. aestivum), and Scott Hulbert [Department of Plant Pathology; Zea mays (maize)]. Sorghum bicolor was obtained directly from the field. Seeds were surface-sterilized (Speakman and Krueger 1983) and germinated in the dark at room temperature. Isolation of total DNA Etiolated shoots were harvested 7–10 days after germination, immediately frozen in liquid N2, and ground into powder using mortars and pestles previously treated with 0.1 M HCl and baked at 160 C for at least 6 h. Total plant DNA was isolated by a modification of the procedure of Jhingan (1992) by scaling down the amount of tissue used and extending the extraction time. About 50 mg of powdered tissue was extracted with 0.5 mL of 625 mM potassium ethyl xanthogenate (Jhingan 1992) for 15 min at 65 C. After centrifugation for 1 min at 12,000 g, DNA was recovered from the supernatant solution by precipitation with ethanol. PCR amplification of mtDNA sequences Primers for amplifying orf256-like sequences and coxI were: 5:–94 (5¢-GGTTCTTCTCTTCCAGCG-3¢), 5:1 (5¢-ATGACAAATATGGTTCGATGGC-3¢), 5:35 (5¢-GCAGGTTTACTGCTTTC-3¢), 5:253 (5¢-CTGAGCCTTTACGAGCAGG-3¢), 5:964 (5¢-AGCTTGCAGGAGTGATGGGCACATGCTTC-3¢), C:221 (5¢-GCTTGGGGATCCTGAATC-3¢), C:477 (5¢-GGAACGAAGCGCTTCATCGA-3¢), C:482 (5¢-AGATAGGAACGAAGCGC-3¢), C:972 (5¢-CACTCCTGCAATGGCACC-3¢), and C:1469 (5¢-GCTGTCACTAGAACGGACC-3¢). Primers designated ‘‘5:’’ represent sense strand sequences and those designated ‘‘C:’’ represent sequences on the complementary strand. Primer numbering is based on ‘‘1’’ as the first nucleotide of orf256 coding sequence. PCR (Mullis and Faloona 1987) conditions were: 2.5 units of Taq DNA polymerase (Fisher Scientific), 10 mM Tris-HCl, pH 8.3, 50 mM KCl, 2 mM MgCl2, 200 lM dNTPs, 600 nM of each forward and reverse primer, and 100 ng of total DNA in a final volume of 50 lL. The PCR reactions were done in a thermal cycler (Biometra Personal Cycler): denaturation at 94 C for 1 min, 35 cycles of denaturation at 94 C for 20 s, annealing at 55 C for 20 s, and extension at 72 C for 40 s, and a final extension at 72 C for 7 min. If necessary, a second round of amplification was done under the same conditions using 1 lL of the previous PCR product as template. PCR products were separated by electrophoresis in 1.5% agarose gels, along with a 100-bp ladder of marker DNA fragments (Promega), and stained with ethidium bromide. The coxI gene regions for T. aestivum and T. timopheevi and the locations of primer sequences are shown in Fig. 1a. Southern blot analysis of wheat relatives, barley, and rye Total DNA was digested with HindIII for gel electrophoresis and transferred to positively charged Nylon membranes. The PCRamplified orf256 coding region, using primers 5:35/C:477, and the coxI coding region, using primers 5:1,955/C:2,450, were used to prepare probes with [a-32P]dATP and the Prime-A-Gene kit (Promega).
Materials and methods Isolation of mitochondrial DNA and Southern blot analysis Sources of plant tissues Seeds were provided by Bikram Gill through the Wheat Genetics Resource Center [Department of Plant Pathology; Triticum species monococcum boeticum, monococcum monococcum, timopheevi, turgidum, Aegilops species bicornis, searsii, sharonensis, speltoides
Mitochondrial DNA was isolated from shoots (harvested 7– 10 days after germination) of T. aestivum, T. timopheevi, S. bicolor, and Z. mays; 10 lg was digested with HindIII for gel electrophoresis and transferred to a positively charged Nylon membrane (Rathburn et al. 1993). A DNA probe of the coding region of
359 Fig. 1a, b. The coxI regions of mitochondrial DNA from Triticum aestivum, T. timopheevi and rice. a The coxI regions of mitochondrial DNA from T. aestivum and T. timopheevi. The 5¢ (5:) and 3¢ (C:) primers used for PCR have bases numbered according to the sequence from T. timopheevi and are shown below each gene region with the size indicated for DNA fragments produced with the primer pairs. The stippled regions at the N-terminus of each gene represent the repeat sequence encoding the N-terminus of COXI. The base pair position numbers shown in parentheses for T. aestivum indicate the corresponding base pair positions in the T. timopheevi diagram. b Comparison of the upstream regions of coxI in mitochondrial DNA from rice, T. timopheevi, and T. aestivum. The encoded N-terminus repeat of coxI is indicated with an asterisk
orf256 was prepared by PCR using primers 5:35/C:477 and the cloned orf256/coxI region from T. timopheevi mitochondrial DNA (Rathburn and Hedgcoth 1991), which includes the conserved region found in O. sativa upstream of coxI (see Fig. 1b). The probe was labeled with digoxigenin-11-dUTP (Rathburn et al. 1993) for chemiluminescent detection following hybridization of a Southern blot.
Isolation of mitochondrial protein and Western blot analysis Protein was isolated from mitochondria of shoots (harvested 7– 10 days after germination) of a cytoplasmic male-sterile line of wheat, T. aestivum, and Ae. speltoides, and 20 lg from each source was subjected to polyacrylamide gel electrophoresis, transferred to a polyvinylidene difluoride membrane (Millipore), and probed with anti-orf256 antibody (Song and Hedgcoth 1994b).
Isolation of total RNA from shoots Total RNA was prepared from shoots harvested 7–10 days after germination, using TriReagent (Molecular Research Center; Chomczynski 1993). Contaminating DNA was removed by treatment with DNase I, using the MessageClean kit (GenHunter), and control PCR reactions were done to ensure that genomic DNA was removed by the DNase treatment. PCR amplification of orf256 and coxI mRNA For reverse transcription and PCR amplification, 5 lg samples of total RNA for each plant type were used. Part of orf256 or coxI mRNA was reverse transcribed into single strand cDNA, using either primer C:482 (orf256) or primer C:1,469 (coxI) and SuperScript II, following the supplied protocol (Life Technologies). Second-strand cDNA synthesis and amplification was done by PCR with primer pairs 5:253/C:482 (orf256) or 5:964/C:1,469 (coxI), using the PCR conditions outlined above.
Results Existence and organization of orf256-like mtDNA in barley and rye A BLAST (Altschul et al. 1990) search of sequence databases demonstrated an orf256-like sequence in rice with 67% identity to orf256 in a location upstream from coxI. The organization of the coxI gene regions for rice (Suzuki et al. 1991), T. timopheevi (Rathburn and Hedgcoth 1991), and T. aestivum (Bonen et al. 1987) is shown in Fig. 1b. Exploring whether orf256-like sequences are present in mtDNA of some other grains, primers 5:253 and C:482 were designed, based on conserved sequences in rice and T. timopheevi mtDNA. The
360
orf256 sequence for 5:253 corresponds to a location in the rice sequence at base pair 1,409 in Fig. 1b; the sequence for C:482 corresponds to rice base pair 1,638. The block of 230 base pairs defined by 5:253/C:482 is 69 base pairs farther upstream from coxI in rice, as compared with T. timopheevi, but the selected primer pair still defines a region of 230 base pairs in rice mtDNA with a 76% identity to the corresponding region in T. timopheevi mtDNA. With this primer pair, a 230-bp PCR band was obtained for both rice and T. timopheevi, as expected, and there was an absence of a band for T. aestivum consistent with the absence of orf256 or a related sequence in this species (Fig. 2a) (Rathburn and Hedgcoth 1991; Rathburn et al. 1993). Barley and rye also have orf256-like sequences, indicated by the presence of a band of about 230 bp, whereas there was no PCR product for maize and sorghum mtDNA (Fig. 2a). Because the orf256-like sequence of rice and orf256 of T. timopheevi are both upstream to coxI, additional PCR assays were done to determine whether the orf-like sequences of barley and rye are similarly located. With primer C:972 located 87 bp inside of coxI, the primer
Fig. 2a–c. PCR demonstration of the presence of an orf256-like sequence upstream from coxI in barley and rye. a Primer pair 5:253/C:482 for internal orf256-like sequence (both primer sites are located within T. timopheevi orf256 coding sequence). The bands are about the same size as that from T. timopheevi, 230 bp. b Primer pair 5:253/C:972 links an orf256-like sequence to coxI (one primer is within the orf256 coding sequence, the other is within the coxI coding sequence downstream of orf256). The T. timopheevi band is 720 bp and the barley band is about the same size, whereas the band for rye is slightly smaller; the rice band is about 790 bp. c Primer pair 5:1/C:482 demonstrates a coxI-like N-terminus upstream of primer C:482 (5:1 is the first position of both orf256 and coxI translation start sites, C:482 is within the orf256 coding sequence). The T. timopheevi band is 510 bp. M DNA size markers with a bright band at 500 bp, Os rice, Tt T. timopheevi, Ta T. aestivum, Hv barley, Sc rye, Sb sorghum, Zm maize
pair 5:253/C:972 should determine whether the orf-like sequence is 5¢ to coxI. The size of the expected PCR DNA fragment for T. timopheevi is 720 bp and that for rice is 788 bp; these fragment sizes were obtained (Fig. 2b). Barley gave a band size equivalent to that of T. timopheevi and the fragment for rye was slightly shorter, about 700 bp. T. aestivum, maize, and sorghum did not give PCR products (Fig. 2b). The orf256-like sequence in barley and rye is located upstream to coxI in about the same location as in T. timopheevi and rice mtDNAs. The first 11 amino acids encoded in orf256 are the same as the first 11 amino acids encoded in coxI, with 32 nucleotide identities of the first 33 nucleotides. Using the 22-nucleotide 5:1 primer, which starts at the first nucleotide of the start codon of orf256, with primer C:482 for PCR should determine whether the orf256-like sequences of barley and rye encode coxI-like N-terminal sequences. Rice did not give a PCR product with the primers; the rice orf256-like sequence does not contain a repeated coxI sequence (Fig. 2c). The orf256-like sequences of barley and rye provided DNA fragments of the same size as T. timopheevi, 482 bp (Fig. 2c). Thus, barley and rye have an orf256-like sequence with the repeat of the coxI N-terminus. Again, there were no products for T. aestivum, maize, or sorghum. All sources, including maize and sorghum, gave an 87-bp fragment representing the N-terminus of coxI with 5:1/ C:972 primers (Table 1). Because of the strong similarity between the T. timopheevi orf256 sequence and the orf256-like results for barley and rye, primers 5:–94/C:221 were used to determine whether barley and rye have an orf256-like 5¢ flanking DNA like that of T. timopheevi. The 315-bp DNA fragments in Fig. 3a confirm that the 5¢ flanking sequences of the orf256-like DNA in barley and rye are like that of orf256 in T. timopheevi. Although the 5¢ flanking sequence of orf256 from –1 to –228 is identical to the 5¢ flanking sequence of T. aestivum coxI from – 1 to –228 (Rathburn and Hedgcoth 1991; see Fig. 1a), there is no PCR product for T. aestivum, because there is no orf256 sequence to match the C:221 primer. Rice did not give a product because, in rice, the sequence upstream of the region with similarity to the orf256 coding region is unlike the T. timopheevi upstream sequence. Because there is equivalency of PCR results for barley and rye with those of T. timopheevi, the primer combination 5:–94/C:972 should produce for barley and rye mtDNA a DNA fragment of about 1,066 bp from the 5¢ flank of orf256 through the entire orf256 sequence and into coxI. Barley and rye mtDNA gave PCR fragments of about the same size as the fragment from T. timopheevi mtDNA (Fig. 3b). Lacking an orf256 sequence, T. aestivum mtDNA gave a 181-bp fragment (Fig. 3b). PCR analysis of wheat relatives for orf256 Assay for the presence of an orf256-like sequence in several Aegilops and Triticum species was done by PCR
361 Table 1. Summary of PCR results for wheat relatives, barley, and rye. Sizes shown for each primer pair are in base pairs DNA source
5:253/C:482
5:253/C:972
5:1/C:482
5:1/C:972
5:–94/C:221
5:–94/C:972
Aegilops bicornis Ae. searsii Ae. sharonensis Ae. speltoides Ae. tauschii Triticum monococcum T. turgidum Hordeum vulgare Secale cereale T. timopheevi T. aestivum
230 230 230 230 230 230 230 230 230 230 –
720 720 720 720 720 700 – 720 700 720 –
482 482 482 482 482 482 – 482 482 482 –
87 87 87 87 87 87 87 87 87 87 87
315 315 315 315 315 315 315 315 315 315 –
1,066 1,066 1,066 1,066 1,066 1,066 181 1,066 1,066 1,066 181
using primers 5:253/C:482. As for T. timopheevi, a 230bp fragment was found for all plants tested (Table 1). The primers 5:253/C:972 were used to determine whether the orf256-like sequence was upstream to coxI. Except for T. monococcum and T. turgidum, the fragments obtained were the same size as the fragment for T. timopheevi, about 720 bp, placing the orf256-like sequence upstream to coxI (Table 1). T. monococcum gave a fragment of about 700 bp (Table 1), whereas no fragment was obtained for T. turgidum. The probable presence of a coxI-like N-terminus as part of the orf256like sequence was shown with primers 5:1/C:482, yielding 482-bp fragments for wheat relatives, except for T. turgidum (Table 1). This primer combination (and the combination 5:1/C:972) also gave a 200-bp fragment (data not shown). Primer 5:1, when used as the only primer with DNAs closely related to wheat DNA, gave a 200-bp fragment, except for T. aestivum, T. timopheevi, and Ae. tauschii. Thus, most of the wheat relatives appear to have a nucleotide sequence involving inverted repeats of the coxI amino terminus about 200 bp apart at an unknown location in addition to the normal coxI N-terminus and the coxI N-terminus existing as part of orf256. For clarity, the 200-bp fragments are not listed in Table 1. T. monococcum boeticum and T. monococcum monococcum gave identical results; thus, results are shown and discussed as T. monococcum. Results for Ae. speltoides auchira and ligustica were also identical and are presented as Ae. speltoides. For the wheat relatives, PCR gave 87-bp fragments with primers 5:1/C:972 (Table 1), which are the same results obtained for barley, rye, T. timopheevi, and T. aestivum. This finding reflects a normal N-terminus region of coxI. Primer 5:1 has two binding sites in the orf256/coxI region: at the translation initiation sites of both orf256 and coxI. Therefore, priming the DNA from any of the plant sources that have orf256 upstream from coxI with the primers 5:1/C:972 should give both 972-bp and 87bp products from orf256 and coxI, respectively. However, even T. timopheevi mtDNA gives only the 87-bp product under the usual PCR conditions. With the excess of primers used in the PCR analysis, the downstream primer dominates the PCR reaction such that the
Fig. 3a–c. PCR analyses demonstrating the presence of an orf256like sequence and a link to coxI. a Presence of orf256-like upstream noncoding sequence in a similar position in T. timopheevi, barley, and rye. The primer pair 5:–94/C:221 produces 315-bp fragments for T. timopheevi, barley, and rye (5:–94 is an upstream noncoding sequence, C:221 is within the orf256 coding sequence). b Linking orf256 5¢ noncoding sequence, orf256 coding sequence and coxI for barley and rye, using a primer upstream of the orf256-like sequence (5:–94) and a primer within coxI (C:972). The fragments produced for barley and rye with primers 5:–94/C:972 are about the same size as the fragment for T. timopheevi, 1,066 bp, spanning from an upstream noncoding sequence through the gene and into coxI. c Linking orf256 5¢ noncoding sequence, orf256 coding sequence, and coxI, using a primer upstream of the orf256-like sequence (5:– 94) and a primer within coxI (C:972) for Aegilops and Triticum species. The larger fragments range from –94 in the noncoding region of orf256 through the gene and into coxI at position 972 and are about the same size as that obtained for T. timopheevi, 1,066 bp. The small fragment for T. turgidum is about 200 bp. Ata Ae. tauschii, Ash Ae. sharonensis, Ase Ae. searsii, Abi Ae. bicornis, Tm T. monococcum, Ttu T. turgidum
362
87-bp fragment is amplified well in excess of the 972-bp fragment, because synthesis from the primer site of the 87-bp fragment interferes with extension of synthesis from the upstream primer through the primer site in the 87-bp sequence. Results for the wheat relatives with primers 5:–94/ C:221 show that they have an orf256-like sequence with a 5¢ flanking sequence like that of T. timopheevi. DNA fragments of 315 bp were obtained for all plants, including T. turgidum (Table 1). Since the arrangement of the orf256-like sequence and the coxI gene appear to be nearly identical for the plants assayed, except for T. turgidum, it should be possible to obtain a single DNA fragment of about 1,066 bp, starting at –94 and extending through the orf256-like region and into coxI with the primers 5:–94/ C:972. DNAs from Ae. tauschii, Ae. sharonensis, Ae. searsii, Ae. bicornis, Ae. speltoides, and T. monococcum gave the expected fragment, slightly larger than 1 kb (Fig. 3c). The T. monococcum fragment is slightly smaller than the other fragments, which appear equivalent to the T. timopheevi 1,066-bp fragment. DNA from T. turgidum gave a fragment of about 200 bp (Fig. 3c), which is probably the same as the 181-bp fragment obtained with these primers and DNA from T. aestivum. The orf256 and coxI sequences of T. turgidum appear to be arranged differently from and falling between the wheat relatives and T. aestivum. T. turgidum has two – 94 sequences rather than one: one in the normal T. aestivum position upstream from coxI and one as in T. timopheevi upstream from the orf256-like sequence. At least part of the orf256 sequence is present in T. turgidum, as indicated by PCR results with primer pairs 5:–94/C:221 and 5:253/C:482. The failure of primers 5:1/C:482 to produce a fragment may be the result of the absence of a sequence like the 5:1 primer or only a change in the 3¢ nucleotide precluding the base pairing necessary for extension in the PCR reaction. Southern hybridizations were done with DNA digested with HindIII, using probes from either orf256 or coxI coding regions. The DNA from barley, rye, and all of the wheat relatives, with the exception of T. turgidum, have a HindIII fragment of about 3.1 kb that responds to the orf256 probe (Fig. 4a). Only a single fragment, slightly larger than 1 kb, responding to the orf256 probe is found for T. turgidum. Thus, the orf256 region of T. turgidum mtDNA is missing part of the gene region or has mutations which interfere with some of the PCR primers and has undergone a rearrangement in the mitochondrial genome or at least the addition of a new HindIII site. The DNA for all of the plants examined has only a single fragment responding to the coxI probe (Fig. 4b) with the gene region of T. turgidum probably similar to that of T. aestivum, whereas the coxI gene region of the other plants is like that of T. timopheevi with orf256 just upstream. Mismatching of primers and DNA sequence could explain the failure to detect any orf256-like sequence in maize and sorghum, although this would indicate the
Fig. 4a, b. Southern blot analysis indicating orf256-like sequence and coxI on the same size restriction fragment of DNA. a Probe specific for orf256 coding region. b Probe specific for coxI coding region. In T. aestivum mtDNA, coxI is located on a 3.3-kb HindIII fragment and, in T. timopheevi, the linked orf256 and coxI are located on a 3.1-kb HindIII fragment
possibility of considerable divergence of orf256-like sequence. This point was the basis for an experiment in which mtDNA from maize and sorghum was probed with orf256 DNA. Lack of hybridizable orf256-like sequence in maize and sorghum HindIII-digested mitochondrial DNA from T. aestivum, T. timopheevi, maize, and sorghum, following separation of fragments by agarose gel electrophoresis and transfer to Nylon membrane, was probed with orf256 DNA. A band was detected only for T. timopheevi mitochondrial DNA (Fig. 5) and it was of the expected size for a HindIII fragment containing orf256 (Rathburn et al. 1993). Thus, maize and sorghum, like T. aestivum, lack an orf256-like sequence. Expression of orf256 as RNA in various cereals Total RNA from each plant type was converted by reverse transcription into first-strand cDNA using primer C:482. Amplification of the cDNA to reveal mRNAs containing the orf256 coding sequence was done by PCR, using primers 5:253/C:482. Amplified DNA for orf256 was obtained only for T. timopheevi (Fig. 6a, b) and Ae. speltoides (Fig. 6b). The corresponding fragment which is amplified from rice RNA (Fig. 6a) is part of the 5¢ upstream region of rice coxI. It is not a complete coding sequence for orf256, but the primers are perfect matches and there is 76% nucleotide identity for the 230-bp primed region. In contrast, in a similar experiment with RNA from all plants, RNA for the coding region of coxI was detected by reverse transcription and PCR amplification (Fig. 6c, d).
363
7 kDa derived from orf256 and detectable by antibody (Song and Hedgcoth 1994b), but neither parent has this mitochondrial protein. Crosses between T. aestivum and Ae. speltoides with Ae. speltoides as maternal donor also produce male-sterile plants; these also have the 7-kDa protein (Song and Hedgcoth 1994b). Because Ae. speltoides has RNA containing the orf256 sequence, as does T. timopheevi, mitochondrial protein from Ae. speltoides was assayed for the presence of the 7-kDa protein by antibody in a Western blot. Although a T. aestivum/T. timopheevi cms hybrid had the 7-kDa protein, it was not present in mitochondrial protein from fertile Ae. speltoides nor, as expected, was it found in mitochondrial protein from T. aestivum (Fig. 7).
Discussion
Fig. 5. Southern blot analysis for the presence of orf256 sequence in mitochondrial DNA of T. aestivum, T. timopheevi, Sorghum bicolor, and Zea mays. A membrane blot of HindIII-digested mitochondrial DNA probed with DNA for the 35–477-bp region of orf256
The data demonstrate that barley, rye, and the Aegilops and Triticum species studied, with the exception of T. turgidum and T. aestivum, have a sequence upstream of the mitochondrial coxI gene with features resembling orf256 of T. timopheevi mtDNA: a similar 5¢ flanking region, a probable N-terminus like that of coxI, and correctly positioned sequences within a region corresponding to the coding region of orf256. Although the rice sequence (Suzuki et al. 1991) which led to this study has 67% nucleotide identity to the T. timopheevi sequence upstream of coxI (but lacks an orf256 and a repeat of the amino terminus in coxI), the number of PCR primers used and the relative positions of fragments produced indicate a greater conservation of sequence in barley, rye, and the Aegilops and Triticum species. In contrast, there is no evidence of the orf256-like sequence in maize and sorghum; nucleotide sequences in the
Fig. 6a–d. Presence or absence of orf256-like mRNA and coxI mRNA demonstrated by PCR amplification of first-strand cDNA products from reverse transcription of total RNA. a, b orf256-like mRNA. Reverse transcription first strand primer was C:482. PCR primers were 5:253/C:482 (orf256 coding sequence primers). NC A tube without RNA (negative control) carried through reverse transcription and PCR. c, d coxI mRNA. Reverse transcription first-strand primer was C:1,469. PCR primers were 5:964/C:1,469 (coxI coding sequence primers; coxI begins at nucleotide 886 when numbering begins with 1 at the orf256 translation start site)
Absence of a protein product for orf256 in Ae. speltoides The cms progeny of crosses between T. aestivum and T. timopheevi have a mitochondrial protein of about
Fig. 7. Western blot of mitochondrial protein from a cytoplasmic male sterile wheat line (S), T. aestivum (Ta), and Ae. speltoides (Asp) probed with anti-orf256 antibody. A positive response shows a band of about 7 kDa (Song and Hedgcoth 1994b)
364
5¢ noncoding region of coxI from maize and sorghum (Bailey-Serres et al. 1986; Isaac et al. 1985) do not have homology with the 3¢ terminus or noncoding region of orf256. The presence in the Triticum and Aegilops species of orf256 as an open reading frame located upstream of coxI, as in T. timopheevi, is confirmed by sequencing data of PCR fragments generated from 5:–94/C:972 (Clarkson, unpublished data). Only T. timopheevi and Ae. speltoides show expression of orf256 as RNA, and a protein product of orf256 is evident only in cms hybrids between T. timopheevi or Ae. speltoides and T. aestivum. These observations strengthen the association between orf256 and the cms state. T. turgidum has a part of the orf256 DNA sequence as shown by primer combinations 5:253/C:482 and 5:–94/ C:221 and hybridization with a probe from the orf256 coding region, but the sequences involved with the primers are not arranged like the analogous sequences from T. timopheevi immediately upstream of the coxI gene, because the primers 5:–94/C:972 give the 181-bp product as found for T. aestivum, in which there is no evidence of orf256 in the mtDNA. In contrast to the other wheat relatives, T. turgidum mtDNA appears to have two –94 sequences responding to the 5:–94 primer. Thus, the gene arrangement or sequence pattern differs from the other wheat relatives with the orf256-like sequence, because they do not have a –94 sequence matching the primer immediately upstream of coxI. Their –94 sequence is separated from coxI by orf256, resulting in a PCR product of about 1,066 bp with the 5:–94/C:972 primers, rather than the 181-bp fragment found for T. aestivum and T. turgidum. Furthermore, the coxI region of T. turgidum mtDNA appears to be on a HindIII fragment like that in T. aestivum and probably is like that of T. aestivum. Therefore, in T. turgidum, the orf256–coxI region of wheat relatives has been rearranged in some manner, such that the coxI region is like that of T. aestivum and the orf256-like sequence is on a separate HindIII fragment not linked to coxI as part of a dicistronic element. The mtDNA sequence of coxI from Ae. columnaris (Ikeda and Tsunewaki 1996) begins at a position corresponding to nucleotide 656 of the coding sequence of orf256. There is 100% homology with the same region of T. timopheevi extending through the termination codon and into the intergenic region. There is a 35-bp deletion in the intergenic region of the mtDNA of Ae. columnaris followed by a region of 96% homology extending to the position of the coxI start codon. This supports the results and conclusions of the work reported here that the orf256 gene region is highly conserved in Aegilops and Triticum species. The absence of an orf256-like sequence in maize and sorghum and its presence in barley, rye, and Triticum and Aegilops species is in accordance with the evolutionary relatedness of the various plants. It is interesting that barley, rye, and the evolutionarily close wheat relatives represented by Aegilops and Triticum species have the orf256 sequence, whereas the tetraploid T. turgidum
does not appear to have a complete copy and what is present is not in the same physical location as in other Triticum species. Significantly, the hexaploid T. aestivum, derived from T. turgidum, does not have an orf256 sequence. Thus, although Ae. speltoides is strongly supported as the donor of the T. turgidum and T. aestivum mitochondrial genomes (Terachi et al. 1990), the mtDNAs of T. turgidum and T. aestivum have diverged considerably from Ae. speltoides in the region immediately upstream of coxI. Acknowledgements This project was supported by the Kansas Agricultural Experiment Station (Publication 97-368-J) and by USDA Grant 94-37301-0379.
References Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ (1990) Basic local alignment search tool. J Mol Biol 215:403–410 Bailey-Serres J, Hanson DK, Fox TD, Leaver CJ (1986) Mitochondrial genome rearrangement leads to extension and relocation of the cytochrome c oxidase subunit 1 gene in Sorghum. Cell 47:567–576 Bonen L, Boer PH, McIntosh JE, Gray MW (1987) Nucleotide sequence of the wheat mitochondrial gene for subunit I of cytochrome oxidase. Nucleic Acids Res 15:6734 Chomczynski P (1993) A reagent for the single-step simultaneous isolation of RNA, DNA and proteins from cell and tissue samples. BioTechniques 15:532–535 Clark LG, Zhang W, Wendel JF (1995) A phylogeny of the grass family (Poaceae) based on ndhF sequence data. Syst Bot 20:436–460 Doebley J, Durbin M, Golenberg EM, Clegg MT, Ma DP (1990) Evolutionary analysis of the large subunit of carboxylase (rbcL) nucleotide sequence among the grasses (Gramineae). Evolution 44:1097–1108 Friebe B, Gill BS (1996) Chromosome banding and genome analysis in diploid and cultivated polyploid wheats. In: Jauhar PP (ed) Methods of genome analysis in plants. CRC Press, Boca Raton, pp 39–60 Gaut BS, Doebley JF (1997) DNA sequence evidence for the segmental allotetraploid origin of maize. Proc Natl Acad Sci USA 94:6809–6814 Ikeda TM, Tsunewaki K (1996) Deficiency of coxI gene expression in wheat plants with Aegilops columnaris cytoplasm. Curr Genet 30:309–514 Isaac PG, Jones VP, Leaver CJ (1985) The maize cytochrome c oxidase subunit I gene: sequence, expression and rearrangement in cytoplasmic male sterile plants. EMBO J 4:1617–1623 Jhingan AK (1992) A novel technology for DNA isolation. Methods Mol Cell Biol 3:15–22 Kimber G, Sears ER (1987) Evolution in the genus Triticum and the origin of cultivated wheat. In: Heyne EG (ed) Wheat and wheat improvement. American Society of Agronomy, Madison, pp 154–164 Leaver CJ, Isaac PG, Small ID, Bailey-Serres J, Liddell AD, Hawkesford MJ (1988) Mitochondrial genome diversity and cytoplasmic male sterility in higher plants. Philos Trans R Soc Lond B Biol Sci 319:165–176 Mann SS, Lucken KA, Bravo JM (1984) Genetic analyses of malesterility restoration in wheat. I. Chromosomal location of Rf genes. Crop Sci 24:17–20 Mullis KB, Faloona FA (1987) Specific synthesis of DNA in vitro via a polymerase-catalyzed chain reaction. Methods Enzymol 155:335–350 Rathburn HB, Hedgcoth C (1991) A chimeric open reading frame in the 5¢ flanking region of coxI mitochondrial DNA from cytoplasmic male-sterile wheat. Plant Mol Biol 16:909–912
365 Rathburn HB, Song J, Hedgcoth C (1993) Cytoplasmic male sterility and fertility restoration in wheat are not associated with rearrangements of mitochondrial DNA in the gene regions for cob, coxII, or coxI. Plant Mol Biol 21:195–201 Sasanuma T, Miyashita NT, Tsunewaki K (1996) Wheat phylogeny determined by RFLP analysis of nuclear DNA. 3. Intra- and interspecific variations of five Aegilops Sitopsis species. Theor Appl Genet 92:928–934 Song J, Hedgcoth C (1994a) Influence of nuclear background on transcription of a chimeric gene (orf256) and coxI in fertile and cytoplasmic male sterile wheats. Genome 37:203–209 Song J, Hedgcoth C (1994b) A chimeric gene (orf256) is expressed as protein only in cytoplasmic male-sterile lines of wheat. Plant Mol Biol 26:535–539 Speakman JB, Krueger W (1983) A comparison of methods to surface sterilize wheat seeds. Trans Br Mycol Soc 80:374–376 Suzuki T, Kazama S, Hirai A, Akihama T, Kadowaki K (1991) The rice mitochondrial nad3 gene has an extended reading
frame at its 5¢ end: nucleotide sequence analysis of rice trnS, nad3, and rps12 genes. Curr Genet 20:331–337 Terachi T, Ogihara Y, Tsunewaki K (1990) The molecular basis of genetic diversity among cytoplasms of Triticum and Aegilops. 7. Restriction endonuclease analysis of mitochondrial DNAs from polyploid wheats and their ancestral species. Theor Appl Genet 80:366–373 Tsunewaki K (1996) Plasmon analysis as the counterpart of genome analysis. In: Jauhar PP (ed) Methods of genome analysis in plants. CRC Press, Boca Raton, pp 271–299 Wilson JA, Ross WM (1962) Male sterility interaction of the Triticum aestivum nucleus and Triticum timopheevi cytoplasm. Wheat Inf Serv 14:29–30 Wolfe KH, Gouy M, Yang Y-W, Sharp PM, Li W-H (1989) Date of the monocot–dicot divergence estimated from chloroplast DNA sequence data. Proc Natl Acad Sci USA 86:6201–6205
