The mitochondrial genomes of seed plants are exceptionally fluid in size, structure, and sequence content, with the accumulation and activity of repetitive sequences underlying much of this variation. We report the first fully sequenced mitochondrial genome of a legume, Vigna radiata (mung bean), and show that despite its unexceptional size (401,262 nt), the genome is unusually depauperate in repetitive DNA and "promiscuous" sequences from the chloroplast and nuclear genomes. Although Vigna lacks the large, recombinationally active repeats typical of most other seed plants, a PCR survey of its modest repertoire of short (38–297 nt) repeats nevertheless revealed evidence for recombination across all of them. A set of novel control assays showed, however, that these results could instead reflect, in part or entirely, artifacts of PCR-mediated recombination. Consequently, we recommend that other methods, especially high-depth genome sequencing, be used instead of PCR to infer patterns of plant mitochondrial recombination. The average-sized but repeat- and feature-poor mitochondrial genome of Vigna makes it ever more difficult to generalize about the factors shaping the size and sequence content of plant mitochondrial genomes.
Citation: Alverson AJ, Zhuo S, Rice DW, Sloan DB, Palmer JD (2011) The Mitochondrial Genome of the Legume Vigna radiata and the Analysis of Recombination across Short Mitochondrial Repeats. PLoS ONE6(1): e16404. https://doi.org/10.1371/journal.pone.0016404
Editor: Orian S. Shirihai, Boston University, United States of America
Received: September 23, 2010; Accepted: December 18, 2010; Published: January 20, 2011
Copyright: © 2011 Alverson et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: National Institutes of Health Ruth L. Kirschstein NRSA Postdoctoral Fellowship (1F32GM080079-01A1) to AJA, NIH research grant RO1-GM-70612 to JDP, and the METACyt Initiative of Indiana University, funded in part through a major grant from the Lilly Endowment, Inc. to JDP. This work was performed under the auspices of the U.S. Department of Energy's Office of Science, Biological and Environmental Research Program, and by the University of California, Lawrence Berkeley National Laboratory under contract No. DE-AC02-05CH11231, Lawrence Livermore National Laboratory under Contract No. DE-AC52-07NA27344, and Los Alamos National Laboratory under contract No. DE-AC02-06NA25396. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
The mitochondrial genomes of seed plants are exceptionally fluid in size, structure, and sequence complexity, making the adage "no two are alike" applicable in ways that are unparalleled by other organelle genomes. Much of this diversity reflects the accumulation and activity of repetitive sequences. Repeats of diverse size and number have been characterized from the roughly 20 seed plant mitochondrial genomes so far sequenced. At one extreme, the nearly 1 Mb Cucurbita mitochondrial genome contains tens of thousands of short (20–40 nt) dispersed repeats that comprise >30% of its genome , whereas other genomes contain small numbers of large (1–120 kb) and mostly species-specific segmental duplications . The size and number of repeats in a plant mitochondrial genome is important because they are also the sites of intramolecular recombination, so repeats ultimately underlie much of the known structural diversity in plant mitochondrial genomes as well. Recombination across inverted repeats inverts the intervening sequences, whereas recombination across directly oriented repeats separates the genome into pairs of subgenomic molecules , . These processes create a structurally dynamic assemblage of genomic molecules in vivo and have led to a virtual scrambling in the gene orders of closely related species  and even conspecific genetic lines , , . Recombination can also cause sequence duplications and deletions, resulting in rapid and sometimes substantial shifts in genome size. For example, although the mitochondrial genomes of five maize cytotypes have virtually identical sequence complexities, a set of large (0.5–120 kb), cytotype-specific duplications has led to >25% variation in genome size . Likewise, a male-sterile strain of Beta vulgaris contains an 87 kb duplication that is absent from its fertile counterpart , . Recombinationally derived deletions, some of which have important deleterious consequences , , are common as well.
Recombination frequency is proportional to the size of the repeat: large (>1 kb) repeats recombine at high frequency, intermediate-sized (100–1000 nt) repeats recombine sporadically, and short (<100 nt) repeats are thought to recombine rarely, if ever , , . Evidence for repeat-mediated recombination traditionally comes from physical mapping of overlapping clones , restriction fragment analysis , and Southern hybridization studies . More recently, whole-genome sequencing projects based on paired-end sequencing of clone libraries have used conflicting signals in genome assemblies to infer patterns of intramolecular recombination –. Finally, PCR across predicted recombination boundaries has also been used to detect recombinant genotypes . The ability of PCR to amplify low-concentration templates is thought to make it particularly well suited for detection of rare recombinants involving short repeats , , .
In addition to repeat content, seed plant mitochondrial genomes also show substantial variation in gene content, reflecting ongoing gene loss and functional gene transfer to the nucleus , . Most gene losses involve ribosomal protein genes and two respiratory genes, sdh3 and sdh4 , . A survey of some 300 diverse seed plants revealed only two losses of the remaining 24 genes. One of these genes, cox2, was found to be universally present across all 300 taxa, save one recent functional transfer to the nucleus in a group of papilionoid legumes –. We sequenced the mitochondrial genome of one of these legumes, Vigna radiata (mung bean), confirmed the absence of the cox2 gene, and discovered a genome in an ongoing state of reduction with respect to gene content. In addition, a comparative analysis of repeat content in the fully sequenced seed plant mitochondrial genomes shows that Vigna has a paucity of repeats of all size classes, including the large recombinationally active repeats present in most seed plants. Although PCR revealed evidence of recombinational activity for numerous short repeats, a novel set of control assays showed that methodological artifacts undermine any firm conclusions about the extent of in vivo recombination in the Vigna mitochondrial genome.
Results and Discussion
Genome Assembly and Sequence Content
The Vigna mitochondrial genome was sequenced to an average read-depth of roughly 8× following standard protocols for shotgun Sanger sequencing. This included ligation of random 3-kb DNA fragments into plasmid vectors followed by transformation of E. coli with the recombinant plasmids. The genome contains one region that is apparently recalcitrant to cloning. A sequence of approximately 100 nt in length, occupying positions 120136–120243 in the genome, was not covered by any of the roughly 2,300 clones generated for the project. PCR and sequencing of this region closed the assembly and revealed two copies of an 11-nt inverted repeat that might have inhibited cloning.
The Vigna mitochondrial genome assembled into a single, circular-mapping molecule of length 401,262 nt and 45.1% GC content, both of which are near the median values of fully sequenced seed plant mitochondrial genomes. The genome contains 31 protein, 3 rRNA, and 16 tRNA genes (Fig. 1). Two identical copies of the atp9 gene are present in the genome. Vigna has one of the most protein-gene-poor mitochondrial genomes so far sequenced in plants, with only two caryophyllids, Beta and Silene, having fewer intact genes , . Like other genome projects (see ref.  for discussion), the Vigna genome sequence confirms the high accuracy of the inferences of mitochondrial gene content made by Adams et al.  in their Southern blot assay of 280 diverse angiosperms. This first completely sequenced legume mitochondrial genome also confirms the absence of the cox2 gene. The cox2 gene loss, originally inferred by Southern blot hybridization , represents the best-studied case of recent functional transfer of an organellar gene to the nuclear genome, with the transfer restricted to a subset of papilionoid legumes –. Although most other respiratory genes have never been found to have been lost during angiosperm evolution, 17 genes (15 ribosomal protein and 2 respiratory) are known to have been lost frequently , , . Nine of these 17 genes are either absent from the Vigna mitochondrial genome (rpl2, rpl10, rps2, rps11, rps13, sdh3) or are present as pseudogenes in various stages of attrition (rps7, rps19, sdh4). The sdh4 gene is the most intact of these, with just a single 10-nt insertion located roughly 30 amino acids upstream of the conserved stop codon. Although the insertion drastically alters the downstream reading frame, it does not introduce a premature stop codon, raising the possibility that the sdh4 gene in Vigna is functional, having 1) co-opted a stop codon roughly 15 amino acids downstream of the conserved stop codon, and 2) tolerated substantial 3′ extension and drastic amino acid divergence in the last ∼20% of the conserved length of the gene. Functional studies of the mitochondrial sdh4 gene, or demonstration of functional transfer of sdh4 to the nuclear genome, will help resolve these possibilities. Roughly half of the ∼300 nt rps19 gene is present, albeit in two disparately spaced pieces, whereas just a single 58 nt fragment of the ∼450 nt rps7 gene remains in the genome. Although sensitive BLAST searches of the genome found a few short DNA fragments (27 and 33 nt in length) with ≥93% similarity to cox2, these could easily represent spurious matches. All pseudogene fragments have retained a relatively high (92–98%) sequence similarity to their intact homologs in Citrullus, suggesting that pseudogenes are "disappearing" via deletions and/or recurrent reshuffling rather than gradual sequence decay. This stands in sharp contrast to the retention of an essentially full-length rps14 pseudogene in grasses for some 80 million years .
Features on transcriptionally clockwise and counter-clockwise strands are drawn on the inside and outside of the circle, respectively.
The Vigna mitochondrial genome contains a conserved set of 17 cis-spliced and five trans-spliced group II introns (Fig. 1). Seed plant mitochondrial genomes typically require trans-splicing of the intron separating exons 3 and 4 of the nad5 gene to create a full-length nad5 transcript. In Vigna, exon 3 is identically oriented and less than 3 kb apart from exon 4 (Fig. 1), raising the possibility of a recent reversion to cis-splicing of this intron.
As in other seed plants, genes and introns comprise a relatively small fraction, just 16.4%, of the overall Vigna mitochondrial genome. BLAST searches revealed only trace amounts of chloroplast- and identifiably nuclear-derived DNA in the intergenic regions, with these two sequence types comprising just 0.5% and 1.6% of the total sequence, respectively (Table 1). These two "promiscuous" sources of DNA typically constitute a more substantial fraction of seed plant mitochondrial genomes , . Most nuclear fragments showed similarity to transposable elements, and one fragment matched a lectin protein kinase pseudogene previously found in the mitochondrial genomes of two cucurbits . A large fraction of the non-coding DNA (29.3%, excluding chloroplast- and nuclear-derived sequences) resembles plant mitochondrial DNA from previously sequenced plant mitochondrial genomes or from plant genome projects in the NCBI whole-genome shotgun database (e.g., Lotus, Medicago, and Ricinus), based on a BLAST expect cutoff of 1e-6 (Table 1). One of these regions shows sequence similarity to a group B DNA polymerase and a DNA-directed RNA polymerase, a syntenic arrangement similar to intra- and extrachromosomal plasmids found in other plant mitochondria , . The genome also contains two regions with similarity to mitovirus-like RNA polymerases from Ricinus and Vitis .
Although the current sample of fully sequenced seed plant mitochondrial genomes is still taxonomically sparse, some preliminary trends in repeat content are emerging. For example, compared to Cycas and most eudicots, the nine grass genomes have, on average, a greater proportion of their genomes occupied by large (>1 kb) repeats. Coverage by large repeats varies considerably within grasses and underlies substantial changes in sequence complexity between relatively recently diverged taxa (e.g., Oryza and Bambusa) as well as subspecies (Fig. 2) and genetic lines  of maize. By contrast, eudicot mitochondrial genomes show greater disparities in genome size, but with the exception of the male-sterile genetic line of Beta, lower overall coverage by large repeats (Fig. 2). This trend is particularly evident in rosids, in which coverage by large repeats does not exceed 6% for any one species, and in which the two largest sequenced mitochondrial genomes (Vitis and Cucurbita) contain no large repeats (Fig. 2). Despite these apparent trends, the current sample of genomes is still too sparse, or in some cases too biased (e.g., monocots are represented solely by grasses), to draw firm conclusions about the evolution of repeat content in plant mitochondrial genomes.
Genome coverage by repeats <1 kb in length is shown in blue, and coverage by repeats ≥1 kb in length is shown in red. Short repeats are sometimes contained, either partly or entirely, within large repeats; genome coverage by these sites is shown in green. Coverage by non-repetitive portions of the genome is shown in white, so the repetitive and non-repetitive fractions sum to the entire size of the genome. The number of repeats <1 kb and ≥1 kb is indicated directly above each bar. These numbers over-estimate the number of unique repeat coordinates in the genome (see Materials and Methods for details). The four Zea genomes are: 1, Zea mays subsp. mays; 2, Zea mays subsp. parviglumis; 3, Zea perennis; and 4, Zea luxurians.
With fewer repeats than all previously sequenced seed plant mitochondrial genomes, Vigna represents an extreme with respect to repeat content (Figs. 2 and 3). Repeats contribute very little to the overall size of the Vigna genome (just 2.7% coverage compared to 8–62% coverage in other genomes; Fig. 2). The Vigna mitochondrial genome is skewed towards fewer and shorter repeats when compared to comparably sized, repeat-poor genomes (e.g., Bambusa) or even the much smaller genomes of Silene and Brassica (Fig. 2). Most Vigna repeats are less than 100 nt in length, and most of these are less than 40 nt in length (Fig. 3). The largest repeat in the Vigna mitochondrial genome contains a duplicate copy of the atp9 gene, and at just 297 nt in length, is substantially shorter than the largest repeat in all other fully sequenced seed plant mitochondrial genomes. Vigna contains only one copy of the 314-nt recombining repeat that is well-characterized from the mitochondrial genomes of several Phaseolus species (a closely related legume) . Finally, Vigna is one of a small number of sequenced mitochondrial genomes (including Bambusa, Vitis, and Cucurbita) that lacks the large (>1 kb) recombining repeats that are otherwise characteristic of seed plant mitochondrial genomes (Fig. 3). Mapping studies have shown that Brassica hirta lacks large repeats as well . Thus, as they do with genome size , mutation rate , and RNA editing frequency , seed plant mitochondrial genomes also show substantial differences in repeat content and, presumably, recombinational activity.
We detected one chimeric sequencing read that conflicted with the main assembly in that it spanned a predicted recombination boundary involving a 175-nt direct repeat. The discovery of this short and apparently recombinationally active repeat, coupled with the absence of large repeats in the genome, prompted us to screen this and 35 additional short repeats (Dataset S1) for evidence of recombinational activity using the PCR strategy illustrated in Figure 4. Using purified mitochondrial DNA as the template, PCR detected recombinant products for every repeat in our survey, regardless of length (38–297 nt), sequence similarity (93–100%), and orientation (direct or inverted) (Fig. S1). Direct sequencing of PCR products invariably gave results consistent with the expectation for repeat-mediated recombination. The characteristics of six representative repeats from our survey are shown in Table 2, and the corresponding recombinant DNA sequences are available in Dataset S2.
Arrows show the orientations of one direct (red) and one inverted (blue) repeat. Arrowheads show the locations and orientations of PCR primers used to detect mitochondrial recombination, relative to the main genome assembly (A). Recombination across a direct repeat (red) divides the genome into two circular subgenomic molecules. The altered arrangement of primers dir-F and dir-R permits PCR-based detection of recombinant product A→D (B). Recombination across an inverted repeat (blue) inverts the intervening sequences, enabling PCR amplification of recombinant product E→G with primers inv-F and inv-R (C).
PCR-mediated recombination poses a potential problem when amplifying any kind of repetitive target region (e.g., multigene families and microsatellites) . Although PCR recombination has not, to the best of our knowledge, been reported for the kinds of assays of intramolecular recombination reported here, we wanted to determine whether in vitro recombination during PCR could create the patterns observed here and in other PCR-based studies on plant mitochondrial recombination , , . To do so, we identified four single-copy regions of varying length (55, 90, 148, and 639 nt) and high sequence similarity (94–100%) in the mitochondrial genomes of two different species, Vigna radiata and Cucurbita pepo, and treated these regions as surrogate repeats in a set of PCR-based recombination assays similar to those described above (Figs. 4 and S2). We used two different PCR templates for these assays. The first was a 1∶1 mixture of total DNA from each species, and the second was an artificial template with substantially higher concentrations of the target regions. We created the latter by separately amplifying the regions of interest from each species then combining the amplicons into a 1∶1 mixture. To test for recombination, we performed PCR using either the total DNA or an artificial amplicon mixture as template, together with primers designed to amplify a bi-species PCR recombination product (Figs. 4 and S2). Because each primer bound to the DNA of a different species, and because our template DNA contained no contiguous and naturally occurring recombinant molecules, PCR-mediated recombination is the only plausible means by which a positive PCR result could be obtained. Using the total DNA mixture as the template, we amplified the intended target region for just two of the eight potential recombination products. These two recombinant products (148 H←F and 639 A→D) were recovered in relatively low yields (Fig. S3). For each of the four bi-species amplicon mixtures, however, we obtained high yields of both possible recombinant products from both dilutions of the artificial template (Fig. S3). Direct sequencing of one high-yield amplicon for each of the eight recombinant products confirmed our prediction of a chimeric, half-Vigna/half-Cucurbita PCR product.
The higher incidence of PCR recombination in the amplicon templates is consistent with previous findings of increased rates of PCR recombination with increased concentration of template DNA , . This is also supported by PCR amplifications of the recombinant configurations from Vigna total DNA, as these reactions contained only about 1/70th the level of mitochondrial genomes (see Methods) as the purified mitochondrial DNA template used in the assays described at the beginning of this section. The total-DNA assays gave quite variable results compared to the assays that used purified mitochondrial DNA, yielding (depending on the repeat) either no detectable product, lower levels of product, of comparable levels of product (not shown). Because the total DNA derives from an unidentified and potentially different genetic line than the purified mitochondrial DNA, it is formally possible that mitochondrial repeat content differs somewhat among genetic lines.
These results, together with the bi-species control assays, suggest that many of the Vigna recombination products are either present in vivo in very low abundance  or are actually absent in vivo, with their recovery a consequence of PCR-mediated recombination. The bi-species control experiments show that very short regions of sequence identity are sufficient to mediate PCR recombination, the result of either template exchange by Taq polymerase  or premature extension termination within the repeat and subsequent illegitimate priming by incompletely extended products . Although it is now clear that PCR recombination can mimic patterns of naturally occurring intramolecular recombination in plant mitochondrial genomes, we cannot rule out that at least some, perhaps many, of the Vigna repeats actually do recombine in vivo, as has been reported for a number of similarly short repeats in the mitochondrial genomes of Arabidopsis ,  and Phaseolus . The recovery of a recombinant clone involving a short, 175-nt repeat indicates that at least one of the Vigna repeats probably does recombine (or has recombined) in vivo, but that the recombination products exist at a low enough level that most of them would not be recovered in our relatively low-depth (∼8×) genome assembly. Indeed, quantitative real-time PCR on two recombination products showed that recombinant configurations exist, whether through in vivo or in vitro recombination, at levels 40–100× less than the main assembly (not shown).
Although Southern blot hybridizations might provide corroborating qualitative and semi-quantitative evidence concerning the recombinational activity of the Vigna repeats, Southerns can be insufficiently sensitive for detection of very-low-level recombinant products associated with repeats as short as those in the Vigna genome , , , resulting in false-negative evidence concerning recombination. Taken together, the shortcomings of PCR and Southern hybridizations are probably best overcome with whole-genome, paired-end shotgun sequencing. Inexpensive, high-throughput sequencing technologies have the potential to produce deep enough coverage to quantify the relative in vivo proportions of dominant and low-level recombinant mitochondrial genome configurations throughout the genome. In the case of Vigna, accurate estimation of the relative levels of minor genome configurations will require sequencing the genome to a depth of perhaps 1000–10,000×. Strategies that merge traditional Southern hybridizations with paired-end shotgun data have also proven powerful for understanding the qualitative and quantitative aspects of plant mitochondrial DNA recombination . In the end, high-depth sequencing of the mitochondrial genome of Vigna, or any of the growing number of seed plants without large repeats, will ultimately show whether mitochondrial recombinational activity is as notoriously variable across seed plants as are mitochondrial genome size and sequence content , mutation rate , and RNA editing frequency .
Materials and Methods
Mitochondrial DNA Isolation, Genome Sequencing and Assembly
Mitochondria were isolated from etiolated seedlings of Vigna radiata cv. Berken using the DNAse I procedure , and mitochondrial DNA was purified from lysed mitochondria by CsCl centrifugation . A single 3-kb library was constructed, cloned, and Sanger sequenced by the U.S. DOE Joint Genome Institute (JGI) in Walnut Creek, California. Detailed protocols are available at http://www.jgi.doe.gov/sequencing/protocols/prots_production.html. The vast majority of sequence reads were assembled into a single, circular-mapping contig with Phrap (www.phrap.org). Consed was used to visualize and validate the final assembly, and to design PCR primers for filling gaps and augmenting regions of low sequence coverage . The annotated genome sequence is available from GenBank (accession HM367685).
Protein, rRNA, and tRNA genes were annotated as described in Alverson et al. . The mitochondrial genome was also compared to a database of all previously sequenced seed plant mitochondrial genomes with BLAST to identify putatively functional conserved syntenic regions . Briefly, these regions include genes, introns, and the conserved sequences immediately flanking them. The latter are delimited using both syntenic- and sequence-level conservation as determined by BLAST comparison of the Vigna genome to a database of all fully sequenced seed plant mitochondrial genomes. These regions are likely to contain promoters, untranslated regions, and trans-spliced introns. Chloroplast-derived sequences were identified by comparing the Vigna mitochondrial genome to a database of representative seed plant chloroplast genomes with BLASTN, and non-coding mitochondrial-like sequences were identified by searching the Vigna genome against a database of all fully sequenced seed plant mitochondrial genomes. All regions that did not match conserved syntenic regions and chloroplast-derived sequences were extracted and searched against the Repbase repetitive element database (ver. 13.05)  and the following databases maintained by the National Center for Biotechnology Information (NCBI): the non-redundant (nr) nucleotide and protein databases, the whole genome shotgun (wgs) database, and the est_others database. All NCBI-BLASTN (ver. 2.2.22+) searches used the following settings: word_size 9, gapopen 5, gapextend 2, reward 2, penalty –3, dust no.
Repeats and Recombination Analyses
Repeated sequences in Vigna and other seed plant mitochondrial genomes were identified as described previously . Briefly, the genome was searched against itself using WU-BLAST with the following settings: M = 1, N = 3, Q = 3, and R = 3, kap, span, B = 1×109, and W = 7. All BLAST hits with a BLAST e-value ≤1 were considered repeats. We predicted recombination boundaries for 36 repeats in the Vigna genome that varied in length, orientation, and sequence identity, and used Consed  to design PCR primers that would amplify one or both predicted recombination products. PCRs were carried out in 25 µL volumes: 18.25 µL water, 2.5 µL 10X buffer (New England Biolabs), 1 µL (400 µM) dNTPs, 0.25 µL Taq polymerase (New England Biolabs #M0267L), 1 µL (0.8 µM) per primer, and 1 µL (40 ng) of purified Vigna mitochondrial DNA (from cv. Berken) or 2 µL (30 ng) of total Vigna DNA (from material of unknown genetic ancestry purchased at local grocery store). Because mitochondrial DNA comprises only about 2% of Vigna total DNA , the effective concentration of mitochondrial template molecules in PCR carried out using purified mitochondrial DNA was about 70 times that using total DNA. PCR conditions were as follows: 94°C for 3 m, 35 cycles of (94°C for 30 s, 55°C for 30 s, 72°C for 60 s), and final extension at 72°C for 10 m. PCR products were purified using ExoSAP-IT (United States Biochemical, Cleveland, OH), and most were sequenced to verify that we had amplified the expected products. Dataset S1 lists the 36 repeats assayed for recombinational activity in the Vigna mitochondrial genome. Recombination primers for six representative repeats (Table 2) are listed in Table S1, and FASTA-formatted sequences of sequenced PCR products are available in Dataset S2.
It is possible that positive PCR results do not reflect the existence of naturally occurring recombinant molecules but instead result from PCR-mediated recombination, which is a concern when amplifying any kind of repetitive target region . To determine whether PCR recombination can give false-positive evidence of intramolecular recombination, we identified identical or near-identical regions shared between the Vigna and Cucurbita (GenBank GQ856148) mitochondrial genomes (Table S2). As described in Results and Discussion and illustrated in Figure S2, we treated these shared regions as surrogate repeats and performed the same kind of PCR-based assays used to detect recombination in the Vigna mitochondrial genome (Fig. 4). PCR conditions were the same as above. The artificial template described in the Results and Discussion was generated by separately amplifying the repeat-containing regions from Vigna and Cucurbita templates with PCR, gel-extracting the products with a QIAquick Gel Extraction Kit (Qiagen Inc.), then pooling equal volumes of the two PCR products into a single mixture (Fig. S2). Primer sequences for these experiments are listed in Table S3, and FASTA-formatted sequences for sequenced PCR products are available in Dataset S2.
We calculated genomic coverage by repeats and estimated the number of repeats for each of the seed plant mitochondrial genomes shown in Figure 2. Coverage is a non-redundant measure of the number of sites occupied by repeats, as determined by a WU-BLAST of each genome to itself (see above). Short repeats are sometimes contained, either partly or entirely, within larger repeats. When calculating coverage, sites in the genome that fall within two or more such overlapping repeats are counted only once. Repeat number estimates (Fig. 2) are based on the number of unique begin–end coordinates of BLAST hits in the genome. In some cases, this number will over-estimate the actual repeat number, especially for genomes that contain large numbers of imperfect, multi-copy repeat families. For example, Silene latifolia contains a family of six recombining direct repeats with a core length of 1362 nt, but with up- and downstream repeat extensions that differ among the six copies . The number of unique begin–end coordinates for a six-copy repeat can range from six (for a six-copy perfect repeat family) to 30 (for a six-copy family of imperfect, variably sized repeats). In this example, WU-BLAST identified 25 different begin–end coordinates for this repeat family (Fig. 2), arguably over-estimating the actual number of repeats by as much as a factor of four.
Short repeats in the Vigna mitochondrial genome that showed evidence for recombinational activity. Repeats vary in length (38–297 nt), sequence similarity (93–100%), and orientation (direct or inverted).
Outline of an assay to determine whether PCR recombination can mimic plant mitochondrial recombination. BLAST comparison of the Vigna and Cucurbita mitochondrial genomes identified surrogate repeats, i.e., regions of identical or near-identical sequence of lengths similar to the repeats in our recombination survey. In all cases, the sequence flanking each side of the "repeat" is unique both within and between the two genomes. Arrows show the orientation of the repeats, and arrowheads mark the location and orientation of PCR primers (A). Regions containing the surrogate repeats, shown by gray boxes, were amplified with primer combinations V3+V4 for Vigna and C3+C4 for Cucurbita, gel-extracted, and the two products were then combined into a 1:1 mixture (B). This mixture was used as the template for PCR wherein one primer matched a unique flanking region in Vigna and the other matched a unique flanking region in Cucurbita. In vitro PCR recombination is the only plausible means of obtaining a positive PCR result. Sequencing of this product should reveal a chimeric, half-Vigna/half-Cucurbita fragment (C).
Results of PCR recombination assays. Four identical or near-identical regions, each with unique flanking sequences, shared between the Vigna and Cucurbita mitochondrial genomes served as surrogate repeats for the PCR recombination assays illustrated in Figure S1. The four "repeats" were 55, 90, 148, and 639 nt in length. Lanes are marked as follows: V, PCR-amplified "repeat" region from Vigna; C, PCR-amplified "repeat" region from Cucurbita; V+C/10, a mixture of the Vigna and Cucurbita amplicons diluted ten-fold; T, amplification of recombination products from a mixture of Vigna and Cucurbita total DNAs; A, amplification of recombination products from undiluted mixture of the V and C PCR products; A/10, amplification of recombination products from a mixture of the V and C PCR products, diluted 10-fold. We assayed both possible recombination products, which are labeled according to Figure 4.
Primers for PCR assays of intramolecular recombination in the Vigna mitochondrial genome.
Regions of the Vigna and Cucurbita mitochondrial genomes used for PCR recombination experiments.
Primers used for PCR recombination experiments.
General Feature Format file with locations, orientations, and percent similarities for 36 repeats assayed for recombinational activity in the Vigna mitochondrial genome.
We thank three anonymous reviewers for comments on an earlier version of the manuscript.
Conceived and designed the experiments: AJA DWR DBS JDP. Performed the experiments: AJA SZ JDP. Analyzed the data: AJA SZ DWR DBS JDP. Contributed reagents/materials/analysis tools: AJA DWR JDP. Wrote the paper: AJA.
- 1. Alverson AJ, Wei XX, Rice DW, Stern DB, Barry K, et al. (2010) Insights into the evolution of mitochondrial genome size from complete sequences of Citrullus lanatus and Cucurbita pepo (Cucurbitaceae). Molecular Biology and Evolution 27: 1436–1448.AJ AlversonXX WeiDW RiceDB SternK. Barry2010Insights into the evolution of mitochondrial genome size from complete sequences of Citrullus lanatus and Cucurbita pepo (Cucurbitaceae).Molecular Biology and Evolution2714361448
- 2. Allen JO, Fauron CM, Minx P, Roark L, Oddiraju S, et al. (2007) Comparisons among two fertile and three male-sterile mitochondrial genomes of maize. Genetics 177: 1173–1192.JO AllenCM FauronP. MinxL. RoarkS. Oddiraju2007Comparisons among two fertile and three male-sterile mitochondrial genomes of maize.Genetics17711731192
- 3. Lonsdale DM, Hodge TP, Fauron CM (1984) The physical map and organisation of the mitochondrial genome from the fertile cytoplasm of maize. Nucleic Acids Res 12: 9249–9261.DM LonsdaleTP HodgeCM Fauron1984The physical map and organisation of the mitochondrial genome from the fertile cytoplasm of maize.Nucleic Acids Res1292499261
- 4. Palmer JD, Shields CR (1984) Tripartite structure of the Brassica campestris mitochondrial genome. Nature 307: 437–440.JD PalmerCR Shields1984Tripartite structure of the Brassica campestris mitochondrial genome.Nature307437440
- 5. Palmer JD, Herbon LA (1989) Plant mitochondrial DNA evolves rapidly in structure, but slowly in sequence. Journal of Molecular Evolution 28: 87–97.JD PalmerLA Herbon1989Plant mitochondrial DNA evolves rapidly in structure, but slowly in sequence.Journal of Molecular Evolution288797
- 6. Satoh M, Kubo T, Mikami T (2006) The Owen mitochondrial genome in sugar beet (Beta vulgaris L.): possible mechanisms of extensive rearrangements and the origin of the mitotype-unique regions. Theoretical and Applied Genetics 113: 477–484.M. SatohT. KuboT. Mikami2006The Owen mitochondrial genome in sugar beet (Beta vulgaris L.): possible mechanisms of extensive rearrangements and the origin of the mitotype-unique regions.Theoretical and Applied Genetics113477484
- 7. Arrieta-Montiel MP, Shedge V, Davila J, Christensen AC, Mackenzie SA (2009) Diversity of the Arabidopsis mitochondrial genome occurs via nuclear-controlled recombination activity. Genetics 183: 1261–1268.MP Arrieta-MontielV. ShedgeJ. DavilaAC ChristensenSA Mackenzie2009Diversity of the Arabidopsis mitochondrial genome occurs via nuclear-controlled recombination activity.Genetics18312611268
- 8. Kubo T, Nishizawa S, Sugawara A, Itchoda N, Estiati A, et al. (2000) The complete nucleotide sequence of the mitochondrial genome of sugar beet (Beta vulgaris L.) reveals a novel gene for tRNA(Cys)(GCA). Nucleic Acids Research 28: 2571–2576.T. KuboS. NishizawaA. SugawaraN. ItchodaA. Estiati2000The complete nucleotide sequence of the mitochondrial genome of sugar beet (Beta vulgaris L.) reveals a novel gene for tRNA(Cys)(GCA).Nucleic Acids Research2825712576
- 9. Satoh M, Kubo T, Nishizawa S, Estiati A, Itchoda N, et al. (2004) The cytoplasmic male-sterile type and normal type mitochondrial genomes of sugar beet share the same complement of genes of known function but differ in the content of expressed ORFs. Molecular Genetics and Genomics 272: 247–256.M. SatohT. KuboS. NishizawaA. EstiatiN. Itchoda2004The cytoplasmic male-sterile type and normal type mitochondrial genomes of sugar beet share the same complement of genes of known function but differ in the content of expressed ORFs.Molecular Genetics and Genomics272247256
- 10. Lilly JW, Bartoszewski G, Malepszy S, Havey MJ (2001) A major deletion in the cucumber mitochondrial genome sorts with the MSC phenotype. Current Genetics 40: 144–151.JW LillyG. BartoszewskiS. MalepszyMJ Havey2001A major deletion in the cucumber mitochondrial genome sorts with the MSC phenotype.Current Genetics40144151
- 11. Yamato KT, Newton KJ (1999) Heteroplasmy and homoplasmy for maize mitochondrial mutants: A rare homoplasmic nad4 deletion mutant plant. Journal of Heredity 90: 369–373.KT YamatoKJ Newton1999Heteroplasmy and homoplasmy for maize mitochondrial mutants: A rare homoplasmic nad4 deletion mutant plant.Journal of Heredity90369373
- 12. André C, Levy A, Walbot V (1992) Small repeated sequences and the structure of plant mitochondrial genomes. Trends in Genetics 8: 128–132.C. AndréA. LevyV. Walbot1992Small repeated sequences and the structure of plant mitochondrial genomes.Trends in Genetics8128132
- 13. Marechal A, Brisson N (2010) Recombination and the maintenance of plant organelle genome stability. New Phytologist 186: 299–317.A. MarechalN. Brisson2010Recombination and the maintenance of plant organelle genome stability.New Phytologist186299317
- 14. Lonsdale DM, Brears T, Hodge TP, Melville SE, Rottmann WH (1988) The plant mitochondrial genome: homologous recombination as a mechanism for generating heterogeneity. Philosophical Transactions of the Royal Society of London Series B-Biological Sciences 319: 149–163.DM LonsdaleT. BrearsTP HodgeSE MelvilleWH Rottmann1988The plant mitochondrial genome: homologous recombination as a mechanism for generating heterogeneity.Philosophical Transactions of the Royal Society of London Series B-Biological Sciences319149163
- 15. Falconet D, Lejeune B, Quetier F, Gray MW (1984) Evidence for homologous recombination between repeated sequences containing 18S and 5S ribosomal RNA genes in wheat mitochondrial DNA. Embo Journal 3: 297–302.D. FalconetB. LejeuneF. QuetierMW Gray1984Evidence for homologous recombination between repeated sequences containing 18S and 5S ribosomal RNA genes in wheat mitochondrial DNA.Embo Journal3297302
- 16. Ogihara Y, Yamazaki Y, Murai K, Kanno A, Terachi T, et al. (2005) Structural dynamics of cereal mitochondrial genomes as revealed by complete nucleotide sequencing of the wheat mitochondrial genome. Nucleic Acids Research 33: 6235–6250.Y. OgiharaY. YamazakiK. MuraiA. KannoT. Terachi2005Structural dynamics of cereal mitochondrial genomes as revealed by complete nucleotide sequencing of the wheat mitochondrial genome.Nucleic Acids Research3362356250
- 17. Sloan DB, Alverson AJ, Štorchová H, Palmer JD, Taylor DR (2010) Extensive loss of translational genes in the structurally dynamic mitochondrial genome of the angiosperm Silene latifolia. BMC Evolutionary Biology 10: 274.DB SloanAJ AlversonH. ŠtorchováJD PalmerDR Taylor2010Extensive loss of translational genes in the structurally dynamic mitochondrial genome of the angiosperm Silene latifolia.BMC Evolutionary Biology10274
- 18. Sugiyama Y, Watase Y, Nagase M, Makita N, Yagura S, et al. (2005) The complete nucleotide sequence and multipartite organization of the tobacco mitochondrial genome: comparative analysis of mitochondrial genomes in higher plants. Molecular Genetics and Genomics 272: 603–615.Y. SugiyamaY. WataseM. NagaseN. MakitaS. Yagura2005The complete nucleotide sequence and multipartite organization of the tobacco mitochondrial genome: comparative analysis of mitochondrial genomes in higher plants.Molecular Genetics and Genomics272603615
- 19. Shedge V, Arrieta-Montiel M, Christensen AC, Mackenzie SA (2007) Plant mitochondrial recombination surveillance requires unusual RecA and MutS homologs. Plant Cell 19: 1251–1264.V. ShedgeM. Arrieta-MontielAC ChristensenSA Mackenzie2007Plant mitochondrial recombination surveillance requires unusual RecA and MutS homologs.Plant Cell1912511264
- 20. Woloszynska M, Trojanowski D (2009) Counting mtDNA molecules in Phaseolus vulgaris: sublimons are constantly produced by recombination via short repeats and undergo rigorous selection during substoichiometric shifting. Plant Molecular Biology 70: 511–521.M. WoloszynskaD. Trojanowski2009Counting mtDNA molecules in Phaseolus vulgaris: sublimons are constantly produced by recombination via short repeats and undergo rigorous selection during substoichiometric shifting.Plant Molecular Biology70511521
- 21. Adams KL, Palmer JD (2003) Evolution of mitochondrial gene content: gene loss and transfer to the nucleus. Molecular Phylogenetics and Evolution 29: 380–395.KL AdamsJD Palmer2003Evolution of mitochondrial gene content: gene loss and transfer to the nucleus.Molecular Phylogenetics and Evolution29380395
- 22. Adams KL, Qiu Y-L, Stoutemyer M, Palmer JD (2002) Punctuated evolution of mitochondrial gene content: high and variable rates of mitochondrial gene loss and transfer to the nucleus during angiosperm evolution. Proceedings of the National Academy of Sciences of the United States of America 99: 9905–9912.KL AdamsY-L QiuM. StoutemyerJD Palmer2002Punctuated evolution of mitochondrial gene content: high and variable rates of mitochondrial gene loss and transfer to the nucleus during angiosperm evolution.Proceedings of the National Academy of Sciences of the United States of America9999059912
- 23. Adams KL, Rosenblueth M, Qiu Y-L, Palmer JD (2001) Multiple losses and transfers to the nucleus of two mitochondrial succinate dehydrogenase genes during angiosperm evolution. Genetics 158: 1289–1300.KL AdamsM. RosenbluethY-L QiuJD Palmer2001Multiple losses and transfers to the nucleus of two mitochondrial succinate dehydrogenase genes during angiosperm evolution.Genetics15812891300
- 24. Adams KL, Song K, Roessler PG, Nugent JM, Doyle JL, et al. (1999) Intracellular gene transfer in action: dual transcription and multiple silencings of nuclear and mitochondrial cox2 genes in legumes. Proc Natl Acad Sci U S A 96: 13863–13868.KL AdamsK. SongPG RoesslerJM NugentJL Doyle1999Intracellular gene transfer in action: dual transcription and multiple silencings of nuclear and mitochondrial cox2 genes in legumes.Proc Natl Acad Sci U S A961386313868
- 25. Covello PS, Gray MW (1992) Silent mitochondrial and active nuclear genes for subunit 2 of cytochrome c oxidase (cox2) in soybean: evidence for RNA-mediated gene transfer. Embo Journal 11: 3815–3820.PS CovelloMW Gray1992Silent mitochondrial and active nuclear genes for subunit 2 of cytochrome c oxidase (cox2) in soybean: evidence for RNA-mediated gene transfer.Embo Journal1138153820
- 26. Daley DO, Adams KL, Clifton R, Qualmann S, Millar AH, et al. (2002) Gene transfer from mitochondrion to nucleus: novel mechanisms for gene activation from cox2. Plant J 30: 11–21.DO DaleyKL AdamsR. CliftonS. QualmannAH Millar2002Gene transfer from mitochondrion to nucleus: novel mechanisms for gene activation from cox2.Plant J301121
- 27. Daley DO, Clifton R, Whelan J (2002) Intracellular gene transfer: reduced hydrophobicity facilitates gene transfer for subunit 2 of cytochrome c oxidase. Proc Natl Acad Sci U S A 99: 10510–10515.DO DaleyR. CliftonJ. Whelan2002Intracellular gene transfer: reduced hydrophobicity facilitates gene transfer for subunit 2 of cytochrome c oxidase.Proc Natl Acad Sci U S A991051010515
- 28. Nugent JM, Palmer JD (1991) RNA-mediated transfer of the gene coxII from the mitochondrion to the nucleus during flowering plant evolutuion. Cell 66: 473–481.JM NugentJD Palmer1991RNA-mediated transfer of the gene coxII from the mitochondrion to the nucleus during flowering plant evolutuion.Cell66473481
- 29. Qualmann SR, Daley DO, Whelan J, Pratje E (2003) Import pathway of nuclear-encoded cytochrome c oxidase subunit 2 using yeast as a model. Plant Biology 5: 481–490.SR QualmannDO DaleyJ. WhelanE. Pratje2003Import pathway of nuclear-encoded cytochrome c oxidase subunit 2 using yeast as a model.Plant Biology5481490
- 30. Ong HC, Palmer JD (2006) Pervasive survival of expressed mitochondrial rps14 pseudogenes in grasses and their relatives for 80 million years following three functional transfers to the nucleus. BMC Evol Biol 6: 55.HC OngJD Palmer2006Pervasive survival of expressed mitochondrial rps14 pseudogenes in grasses and their relatives for 80 million years following three functional transfers to the nucleus.BMC Evol Biol655
- 31. Kubo N, Arimura SI (2009) Discovery of the rpl10 gene in diverse plant mitochondrial genomes and its probable replacement by the nuclear gene for chloroplast RPL10 in two lineages of angiosperms. DNA Research 17: 1–9.N. KuboSI Arimura2009Discovery of the rpl10 gene in diverse plant mitochondrial genomes and its probable replacement by the nuclear gene for chloroplast RPL10 in two lineages of angiosperms.DNA Research1719
- 32. Mower JP, Bonen L (2009) Ribosomal protein L10 is encoded in the mitochondrial genome of many land plants and green algae. BMC Evol Biol 9: 265.JP MowerL. Bonen2009Ribosomal protein L10 is encoded in the mitochondrial genome of many land plants and green algae.BMC Evol Biol9265
- 33. Knoop V, Unseld M, Marienfeld J, Brandt P, Sunkel S, et al. (1996) copia-, gypsy- and LINE-like retrotransposon fragments in the mitochondrial genome of Arabidopsis thaliana. Genetics 142: 579–585.V. KnoopM. UnseldJ. MarienfeldP. BrandtS. Sunkel1996copia-, gypsy- and LINE-like retrotransposon fragments in the mitochondrial genome of Arabidopsis thaliana.Genetics142579585
- 34. Handa H, Itani K, Sato H (2002) Structural features and expression analysis of a linear mitochondrial plasmid in rapeseed (Brassica napus L.). Molecular Genetics and Genomics 267: 797–805.H. HandaK. ItaniH. Sato2002Structural features and expression analysis of a linear mitochondrial plasmid in rapeseed (Brassica napus L.).Molecular Genetics and Genomics267797805
- 35. McDermott P, Connolly V, Kavanagh TA (2008) The mitochondrial genome of a cytoplasmic male sterile line of perennial ryegrass (Lolium perenne L.) contains an integrated linear plasmid-like element. Theoretical and Applied Genetics 117: 459–470.P. McDermottV. ConnollyTA Kavanagh2008The mitochondrial genome of a cytoplasmic male sterile line of perennial ryegrass (Lolium perenne L.) contains an integrated linear plasmid-like element.Theoretical and Applied Genetics117459470
- 36. Goremykin VV, Salamini F, Velasco R, Viola R (2009) Mitochondrial DNA of Vitis vinifera and the issue of rampant horizontal gene transfer. Mol Biol Evol 26: 99–110.VV GoremykinF. SalaminiR. VelascoR. Viola2009Mitochondrial DNA of Vitis vinifera and the issue of rampant horizontal gene transfer.Mol Biol Evol2699110
- 37. Woloszynska M, Kieleczawa J, Ornatowska M, Wozniak M, Janska H (2001) The origin and maintenance of the small repeat in the bean mitochondrial genome. Molecular Genetics and Genomics 265: 865–872.M. WoloszynskaJ. KieleczawaM. OrnatowskaM. WozniakH. Janska2001The origin and maintenance of the small repeat in the bean mitochondrial genome.Molecular Genetics and Genomics265865872
- 38. Palmer JD, Herbon LA (1987) Unicircular structure of the Brassica hirta mitochondrial genome. Current Genetics 11: 565–570.JD PalmerLA Herbon1987Unicircular structure of the Brassica hirta mitochondrial genome.Current Genetics11565570
- 39. Mower JP, Touzet P, Gummow JS, Delph LF, Palmer JD (2007) Extensive variation in synonymous substitution rates in mitochondrial genes of seed plants. BMC Evol Biol 7: 135.JP MowerP. TouzetJS GummowLF DelphJD Palmer2007Extensive variation in synonymous substitution rates in mitochondrial genes of seed plants.BMC Evol Biol7135
- 40. Sloan DB, MacQueen AH, Alverson AJ, Palmer JD, Taylor DR (2010) Extensive loss of RNA editing sites in rapidly evolving Silene mitochondrial genomes: selection vs. retroprocessing as the driving force. Genetics 185: 1369–1380.DB SloanAH MacQueenAJ AlversonJD PalmerDR Taylor2010Extensive loss of RNA editing sites in rapidly evolving Silene mitochondrial genomes: selection vs. retroprocessing as the driving force.Genetics18513691380
- 41. Meyerhans A, Vartanian JP, Wainhobson S (1990) DNA recombination during PCR. Nucleic Acids Research 18: 1687–1691.A. MeyerhansJP VartanianS. Wainhobson1990DNA recombination during PCR.Nucleic Acids Research1816871691
- 42. Qiu X, Wu L, Huang H, McDonel PE, Palumbo AV, et al. (2001) Evaluation of PCR-Generated Chimeras, Mutations, and Heteroduplexes with 16S rRNA Gene-Based Cloning. Applied and Environmental Microbiology 67: 880–887.X. QiuL. WuH. HuangPE McDonelAV Palumbo2001Evaluation of PCR-Generated Chimeras, Mutations, and Heteroduplexes with 16S rRNA Gene-Based Cloning.Applied and Environmental Microbiology67880887
- 43. Lahr DJG, Katz LA (2009) Reducing the impact of PCR-mediated recombination in molecular evolution and environmental studies using a new-generation high-fidelity DNA polymerase. BioTechniques 47: 857–863.DJG LahrLA Katz2009Reducing the impact of PCR-mediated recombination in molecular evolution and environmental studies using a new-generation high-fidelity DNA polymerase.BioTechniques47857863
- 44. Odelberg SJ, Weiss RB, Hata A, White R (1995) Template-switching during DNA synthesis by Thermus aquaticus DNA polymerase I. Nucleic Acids Research 23: 2049–2057.SJ OdelbergRB WeissA. HataR. White1995Template-switching during DNA synthesis by Thermus aquaticus DNA polymerase I.Nucleic Acids Research2320492057
- 45. Kolodner R, Tewari KK (1972) Physicochemical characterization of mitochondrial DNA from pea leaves. Proceedings of the National Academy of Sciences of the United States of America 69: 1830–1834.R. KolodnerKK Tewari1972Physicochemical characterization of mitochondrial DNA from pea leaves.Proceedings of the National Academy of Sciences of the United States of America6918301834
- 46. Palmer JD (1982) Physical and gene mapping of chloroplast DNA from Atriplex triangularis and Cucumis sativa. Nucleic Acids Res 10: 1593–1605.JD Palmer1982Physical and gene mapping of chloroplast DNA from Atriplex triangularis and Cucumis sativa.Nucleic Acids Res1015931605
- 47. Gordon D, Abajian C, Green P (1998) Consed: a graphical tool for sequence finishing. Genome Res 8: 195–202.D. GordonC. AbajianP. Green1998Consed: a graphical tool for sequence finishing.Genome Res8195202
- 48. Jurka J (2000) Repbase update: a database and an electronic journal of repetitive elements. Trends Genet 16: 418–420.J. Jurka2000Repbase update: a database and an electronic journal of repetitive elements.Trends Genet16418420
- 49. Palmer JD, Thompson WF (1980) Studies on higher plant chloroplast and mitochondrial DNA. Carnegie Institution of Washington Year Book 79: 120–123.JD PalmerWF Thompson1980Studies on higher plant chloroplast and mitochondrial DNA.Carnegie Institution of Washington Year Book79120123