|Home | About | Journals | Submit | Contact Us | Français|
It is known that the miniature inverted-repeat terminal element (MITE) preferentially inserts into low-copy-number sequences or genic regions. Characterization of the second largest subunit of low-copy nuclear RNA polymerase II (RPB2) has indicated that MITE and indels have shaped the homoeologous RPB2 loci in the St and H genome of Eymus species in Triticeae. The aims of this study was to determine if there is MITE in the RPB2 gene in Hordeum genomes, and to compare the gene evolution of RPB2 with other diploid Triticeae species. The sequences were used to reconstruct the phylogeny of the genus Hordeum.
RPB2 regions from all diploid species of Hordeum, one tetraploid species (H. brevisubulatum) and ten accessions of diploid Triticeae species were amplified and sequenced. Parsimony analysis of the DNA dataset was performed in order to reveal the phylogeny of Hordeum species.
MITE was detected in the Xu genome. A 27–36 bp indel sequence was found in the I and Xu genome, but deleted in the Xa and some H genome species. Interestingly, the indel length in H genomes corresponds well to their geographical distribution. Phylogenetic analysis of the RPB2 sequences positioned the H and Xa genome in one monophyletic group. The I and Xu genomes are distinctly separated from the H and Xa ones. The RPB2 data also separated all New World H genome species except H. patagonicum ssp. patagonicum from the Old World H genome species.
MITE and large indels have shaped the RPB2 loci between the Xu and H, I and Xa genomes. The phylogenetic analysis of the RPB2 sequences confirmed the monophyly of Hordeum. The maximum-parsimony analysis demonstrated the four genomes to be subdivided into two groups.
Nuclear RNA polymerases in eukaryotes belong to three different classes, i.e. RNA polymerase I, II and III. Each enzyme is composed of two large (>100 kDa) and several smaller subunits, each of which is typically encoded by a unique single-copy gene (Oxelman and Bremer, 2000). RPB2 encodes the second largest subunit of nuclear RNA polymerase II, and is responsible for the transcription of protein encoding genes (Kolodziej et al., 1990; Sawadogo and Sentenac, 1990). This gene is found in all eukaryotes (Oxelman and Bremer, 2000) and has highly conserved regions in them (Sweetser et al., 1987; Kolodziej et al., 1990; Denton et al., 1998). It has been demonstrated that RPB2 is encoded by a single gene in Arabidopsis (Larkin and Guilfoyle, 1993), tomato (Warrilow and Symons, 1996) Rhododendron and Hordeum vulgare (Denton et al., 1998). Oxelman et al. (2004) reported that some dicots had two copies, while the studied monocot Dioscoreales contained only one copy. Sequencing analysis of the RPB2 gene suggested the presence of only one copy in several diploid Triticeae species with the H genome (Hordeum), St genome (Pseudoroegneria), P genome (Agropyron) and W genome (Australopyrum) (Sun et al., 2008).
The Triticeae genus Hordeum comprises 31 species (including cultivated barley, H. vulgare ssp. vulgare; classification follows von Bothmer et al., 1995) and exists at the diploid, tetraploid and hexaploid levels with a basic chromosome number x = 7. Wild Hordeum species occur under a wide variety of climates in Eurasia and the New World (von Bothmer et al., 1995). Attempts to resolve the interspecific relationships in Hordeum have involved numerious kinds of data and analytical methods. Based on morphological data, von Bothmer and Jacobsen (1985) classified the Hordeum species into four sections; viz. sect. Hordeum, sect. Anisolepis, sect. Critesion and sect. Stenostachys. Cytogenetic data have shown that the diploid species can be divided into four monogenomic groups (see Wang et al., 1994, for genome designations): the I-genome group (H. vulgare and H. bulbosum), the Xa genome group (H. marinum; formerly X), the Xu genome group (H. murinum; formerly Y) and the H genome group (the remaining diploid species; von Bothmer et al., 1986a, 1987). The genomic groups do not correspond to the four sections. Isoenzyme analysis (Jørgensen, 1986) and C-banded karyotypes (Linde-Laursen et al., 1992) supported the four basic genomes. The presence of four basic genomes was also supported by restriction site variation in chloroplast DNA (Baum and Bailey, 1991; Doebley et al., 1992), restriction fragment length polymorphism with repetitive DNA (Svitashev et al., 1994), nuclear gene DMC1 (disrupted meiotic cDNA1) sequences (Petersen and Seberg, 2003), and rDNA ITS sequences (Blattner, 2004). The phylogeny based on nuclear data is incongruent with phylogenies based on data from plastid genomes (Petersen and Seberg, 2003).
In the Triticeae tribe, molecular evolution of the RPB2 gene has only been investigated in allotetraploid StH genomic species of Elymus (Sun et al., 2007). These results indicated that three transposable element indels have shaped the homoeologous RPB2 loci in the St and H genome of Elymus species, and that RPB2 is an excellent tool for investigating the phylogeny and evolutionary dynamics of speciation (Sun et al., 2007). Here we characterize the molecular evolution of the RPB2 gene in the four genomes from 26 taxa of Hordeum and compare their evolution with several other diploid Triticeae species, and find that transposable element have shaped the RPB2 loci in the H, I, Xa and Xu genomes. The nucleotide sequences were also used to reconstruct the phylogeny of the diploid species in the genus Hordeum.
Twenty-six taxa representing all diploid species of Hordeum, the tetraploid species H. brevisubulatum and ten accessions of diploid Triticeae species from Pseudoroegneria (St genome), Agropyron (P), Australopyrum (W), Lophopyrum (Ee), Thinopyrum (Eb) and Dasypyrum (V) were included in the study (Table 1). Classification of Hordeum species follows von Bothmer et al. (1995). The genome designations follow Wang et al. (1994), in which Xa corresponds to X, and Xu corresponds to Y in von Bothmer et al. (1986a). Bromus catharticus was used as an outgroup based on a previous phylogenetic analysis of Poaceae (Hsiao et al., 1995).
The RPB2 sequences were amplified by polymerase chain reaction (PCR) using the primers P6F and P6FR: protocols are given in Sun et al. (2007). PCR products were purified using the QIAquick™ PCR purification kit (QIAGEN Inc) according to the manufacturer's instruction, and then sequenced. In order to avoid the error induced by Taq DNA polymerase during PCR amplification, each sample was independently amplified three times and sequenced: Taq errors that cause substitutions are mostly random, and it is unlikely that any two sequences would share identical Taq errors to create a false synapomorphy.
The PCR products from the tetraploid species H. brevisubulatum were cloned into the TOPO-TA kit from Invitrogen (Carlsbad, CA) according to the manufacturer's protocol. Ten colonies were randomly selected for screening. Each was transferred to 10 µL of LB broth with 0·1 mg mL–1 antibiotics. These solutions were incubated at room temperature for 20 min before using 2 µL for PCR to check for the presence of an insert. This PCR reaction used the primers included in the TOPO kit according to the recommended protocol. For those solutions that were confirmed to contain the insert, the remaining 8 µL of solution was transferred to 5 mL LB broth (with antibiotics) and incubated at 37 °C overnight. Plasmid DNA was isolated using the Promega Wizard® Plus Minipreps DNA Purification System (Promega Corporation, Madison, WI) according to the manufacturer's instructions.
Both strands of the amplified product and plasmid DNA were sequenced using the P6F and P6FR primers. Direct dideoxytermination sequencing was performed using a fluorescent dye terminator Cycle Sequencing Kit (Applied Biosystems, Perkin-Elmer Warrington, UK).
Automated sequence outputs were inspected visually using chromatographs. Multiple sequence alignments were made using ClustalX with default parameters (Thompson et al., 1997). Phylogenetic analysis using the maximum-parsimony (MP) method was performed with the computer program PAUP* ver. 4 beta 10 Win (Swofford, 2003). All characters were specified as unweighted and unordered, and gaps were treated as missing data. Most-parsimonious trees were obtained by performing a heuristic search using the Tree Bisection-Reconnection (TBR) option with MulTrees on, and ten replications of random addition sequences with the stepwise addition option. Multiple parsimonious trees were combined to form a strict consensus tree. Overall character congruence was estimated by the consistency index (CI), and the retention index (RI). In order to infer the robustness of clades, bootstrap values with 1000 replications (Felsenstein, 1985) were calculated by performing a heuristic search using the TBR option with MulTrees off. Parsimony methods try to minimize the number of substitutions, irrespective of the branch lengths on the tree.
A total of 26 taxa of Hordeum were amplified with the P6F/F6FR primer combination. The size of amplified fragments ranged from ~850 bp to ~950 bp. Hordeum murinum ssp. glaucum showed the largest amplified fragment with size of ~950 bp (Fig. 1, lane 5). One tetraploid species, H. brevisubulatum, was also amplified: the patterns appeared to display two bands but they were very close together and it is difficult to resolve them in 1·5 % agarose gel (Fig. 1, lane 16). The sequence results confirmed that two different fragments were amplified from this tetraploid species.
Sequence comparisons from the Hordeum species found a miniature inverted-repeat transposable element (MITE) in the H. murinum ssp. glaucum sequences (H52 and H74), which contained a 15-bp inverted repeat (GTACTCCCTCCGGTT; Fig 2A, underlined with an arrow) that contains a TA and the 12-bp CTCCCTCCGGTT with 99 bp between them. The 99-bp sequence formed a more-or-less hairpin-like structure. The 5′ flanking sequences of the repeated sequence are duplicated TAA (Fig. 2A, box with dashed outline). Two copies of sequences were identified from the tetraploid H. brevisubulatum. One copy showed a 59-bp deletion (H10227L). A 14-bp deletion was observed in Xa genome species [H. marinum ssp. marinum (H121) and ssp. gussoneanum (H581); Fig. 2A, box with solid outline].
Comparison of the sequences of the Hordeum species with Bromus catharticus and several other diploid Triticeae species found that the Hordeum sequences in this region are shorter. Two-to-three relative large insertions were detected in other diploid Triticeae species in this region: the first insertion with 15–18 bp in the Ee, Eb, P, W, V and St genomes; the second insertion with 15 bp in the Ee, Eb, W, V and St genomes; and the third with 39 bp in the St genome of P. spicata (PI 506274, PI 610986) and P. stipifolia (PI 325181), and 96 bp in the V genome of D. villosum (PI 368886) (Fig. 2B).
Apart from the numerous minor length differences observed in Hordeum species, one additional major indel (insertion/deletion) of 27–36 bp was discovered (Fig 2C). A BLAST search using this fragment against the transposable elements (TEs) stored in the TREP (Triticeae Repeat) database did not return perfect matches, but showed that the insertion sequences also belong to transposable-like elements, with part of the sequence matching with CACTA elements and part matching with copia, or gypsy. A 27-bp deletion was discovered in the Xa genome species, while the indel pattern for H genome species is complex. Hordeum patagonicum ssp. setifolium (H1352), ssp. santacrucense (H1353), ssp. magellanicum (H1342) and H. pubiflorum (H1236) did not have a deletion in this region, while another two subspecies of H. patagonicum displayed the deletion, ssp. mustersii (H1358) with a 36-bp deletion and ssp. patagonicum with a 27-bp deletion. The other eleven H genome species showed a 36-bp deletion, and five H (six sequences) genomes have a 27-bp deletion. Interestingly, the H genome species with a 36-bp deletion all are New World Hordeum species, whilst those with a 27-bp deletion are Eurasian species except for H. patagonicum ssp. patagonicum (H6052). In addition, a 6-bp insertion (GCATAA) was detected in the H genome of H. patagonicum ssp. setifolium (H1352), ssp. santacrucense (H1353), ssp. magellanicum (H1342) and H. pubiflorum (H1236) (Fig. 2C).
The RPB2 sequences were used to reveal the phylogeny of Hordeum species and their relationship to other diploid species in the tribe Triticeae. Phylogenetic analysis was conducted using B. catharticus as the outgroup. Parsimony analysis produced 386 equally parsimonious trees with a consistency index (CI) of 0·845 and a retention index (RI) of 0·919. A strict-consensus tree was constructed from these trees, as shown in Fig. 3. The phylogenetic analysis well separated the Hordeum species from the several diploid species in Triticeae. The four genomes in Hordeum were divided into two clades: I + Xu (bootstrap value = 59 %) and H + Xa (bootstrap value = 100 %). The I-genome species, Hordeum vulgare and H. bulbosum, were positioned in one monophyletic group with 100 % bootstrap support. The I-genome group is sister to the Xu genome species H. murinum ssp. galucum with 59 % bootstrap support. The H and Xa genomes were positioned in one group, the clade being supported by a bootstrap value 100. Hordeum patagonicum ssp. setifolium, ssp. magellanicum, ssp. santacrucense and H. pubiflorum were grouped in a highly supported clade (bootstrap value = 100 %) within the H + Xa clade and were separated from the other H-genome species, whereas the other two H. patagonicum subspecies, mustersii and patagonicum, were positioned with other Horderum species: mustersii with New World species and patagonicum with the Asian species H. roshevitzii. Two sequences from the tetraploid H. brevisubulatum grouped together with a highly supported bootstrap value (95 %), with H. bogdanii as sister group with moderate support (bootstrap value 78 %). The two Xa genomic taxa H. marinum ssp. gussoneanum and marinum formed a highly supported group (bootstrap value 98 %) within the H + Xa clade.
Since the discovery of transposons by McClintock, a variety of them have been shown to be major components of eukaryotic genomes (for a review, see Kidwell, 2005). Transposable elements (TEs), comprising both Class I retrotransposons and Class II DNA transposons, constitute the major fractions of repetitive sequences in eukaryotes, and their activity can substantially alter genome structure, produce novel exon combinations, and affect gene expression (Kidwell and Lisch, 2001). Miniature inverted-repeat terminal elements (MITEs) are particular Class II TEs, and play an important role in genome evolution because they have very high copy numbers and display recurrent bursts of transposition (Bureau and Wessler, 1994). Several characterized MITEs are found to preferentially reside in low-copy, genic regions of plant genomes, underscoring their possible roles in the evolution of structure and function of plant genes (Bureau and Wessler, 1992; Wessler et al., 1995; Zhang et al., 2000). MITEs do not encode any TPase or TPase remnant: their classification has been based on shared terminal inverted repeat (TIR) and target site duplication (TSD) sequences. Two superfamilies, Tourist-like and Stowaway-like, have been found to be very common in plants (Feschotte et al., 2002). The putative target site sequences of Tourist and Stowaway are 5′-TAA-3′ and 5′-TA-3′, respectively. MITE Stowaway and Tourist elements are common in Triticeae species (Petersen and Seberg, 2000; Wicker et al., 2003) and in many cereal grasses (Bureau and Wessler, 1994). Our previous study found a 39-bp MITE Stowaway element insertion in the RPB2 gene for all tetraploid Elymus St genomes and diploid P. spicata and P. stipifolia St genomes (Sun et al., 2007). The 39-bp MITE Stowaway element was not detected in the Hordeum genomes I, H and Xa in this study. However, a MITE element was observed in this region (Fig. 2A) in the Xu-genome species, which is different from the previously reported 39-bp MITE Stowaway in Triticeae. The 5′-flanking sequences of a conserved 15-bp TIRs (GTACTCCCTCCGGTT) is TAA TAA, which is a preference target site duplication for the Tourist element. The conserved 15-bp TIR (GTACTCCCTCCGGTT) is different from the consensus sequence (5′-GGCCTTGTTCGGTT-3′) found in the Tourist family (Bureau and Wessler, 1992) and the TIR (GGCCAGTCACAATGG) of the Tourist-like MITE mPing (Jiang et al., 2003). The 15-bp conserved sequences contain a TA and CTCCCTCCGGTT, which is a typical structure of Stowaway elements (Bureau and Wessler, 1994). Thus, it is likely that the MITE element detected in the Hordeum Xu genome originated from the Stowaway nested within Tourist element. The first Stowaway element was found as an insertion in a sorghum Tourist element (Bureau and Wessler, 1994). Such MITE multimers have also been reported in rice (Tarchini et al., 2000). It has been proposed that MITEs could be preferential targets for other MITEs (Feschotte and Mouchès, 2000). The transposable element retrotransposons nested within TRIM (terminal-repeat retrotransposons in miniature) elements have been detected in several plant genomes (Witte et al., 2001). Our results also suggest that the amount of activity and impact of the MITE element in the RPB2 region on the Hordeum genome has been highly genome-specific and represents a phylogenetic signal.
An additional 27–36-bp indel was observed in the Hordeum genomes (Fig. 2C). A search against the TREP (Triticeae Repeat) database suggested that the 27–36-bp sequences may belong to transposable-like elements with part of the sequence matching with CACTA elements, and part of the sequence matching with copia and gypsy. The same size of indel in this region was reported for the RPB2 gene in several Triticeae genomes (Sun et al., 2007). The 27–36-bp sequence was found in the I and Xu genome, but deleted in the Xa genome and some H genomes. The Xa genome showed a 27-bp deletion, while a 36-bp deletion was detected in twelve H-genome species and a 27-bp deletion was detected in five H-genome species (six sequences). The H-genome species with a 36-bp deletion are all New World Hordeum species, whilst those with a 27-bp deletion are Eurasian species except for H. patagonicum ssp. patagonicum (H6052). The indel length of regions in the H genomes corresponds well to the species' geographical distribution, and this may demonstrate adaptive evolution in Hordeum species. Comparison of the sequences from Hordeum genomes indicated that MITE elements and indels have shaped the RPB2 loci in the H, I, Xa and Xu genomes.
The phylogenetic analysis of the RPB2 sequences confirmed the monophyly of Hordeum. The most-parsimonious (MP) analysis demonstrated the four genomes to be subdivided into two groups, thus confirming the results of Komatsuda et al. (1999). One is a clade containing H and Xa, the other contains the Xu and I genomes (Fig. 3). The H + Xa clade is highly supported in MP analysis (100 %); however, the proposed sister group relationship between the Xu species H. murinum ssp. glaucum and three I-genome taxa could only be weakly confirmed by our data in the MP analysis, without statistical support (59 % bootstrap). Hordeum vulgare, H. bulbosum and H. murinum s.l. have been placed in a same section, Hordeum, by von Bothmer et al. (1995). Cytological data suggested that H. vulgare and H. bulbosum contained the same genome, I, whereas H. murinum had a separate genome, Xu, only distantly related to the I genome (von Bothmer et al., 1995). Our RPB2 data clearly shows that H. vulgare and H. bulbosum constitute a monophyletic group that corresponds well with those obtained by isoenzyme analysis (Jørgensen, 1986), chloroplast DNA RFLP data (Doebley et al., 1992), nuclear EF-G sequences (Komatsuda et al., 1999) and rDNA ITS sequences (Blattner, 2004). The position of H. murinum ssp. glaucum as the sister taxon to the H. vulgare and H. bulbosum clade is weakly supported in our MP tree (Fig. 3). Conflicting results on the relationship between I genome and Xu genome species have been reported. Jørgensen (1986) considered the H. murinum (Xu) complex as a sister group of the H. vulgare and H. bulbosum complex based on isoenzymic data; Pelger and von Bothmer (1992) arrived at a similar conclusion based on hordein variation. The sister relationship between the Xu and I genomes was weakly supported (57 % bootstrap) by rDNA ITS sequences (Blattner, 2004), strongly supported (96 % bootstrap) by EF-G data (Komatsuda et al., 1999), but in contrast not supported by repetitive DNA sequences (Svitashev et al., 1994) and cpDNA analysis (Doebley et al., 1992). Other studies have encountered the same problem in placing H. murinum (Xu genome; De Bustos et al., 2002; Petersen and Seberg, 2003). An analysis of ribosomal DNA suggested that H. bulbosum is more closely related to the H. murinum complex than to H. vulgare (Molnar et al., 1989): our data did not support this conclusion. The transposable element is the major cause of genome differentiation between I and Xu in the RPB2 sequence, and has shaped the I and Xu genomes in Hordeum species.
The MP tree clearly shows that H. marinum ssp. marinum and ssp. gussoneanum constitute a highly supported clade (Fig. 3). The results agree well with data from two single-copy nuclear genes (EF-G and DMC1; Komatsuda et al., 1999, 2001; Petersen and Seberg, 2003), rDNA ITS sequences (Blattner, 2004), repeated DNA polymorphism analysis (Svitashev et al., 1994) and RAPD analysis (Marillia and Scoles, 1996). Our results further confirm that the two subspecies carry the same genome (von Bothmer et al., 1987). Hordeum marinum was incorporated in a clade with the H-genome group in the MP tree, which corresponds well with the consensus tree based on EF-G sequences when Komatsuda et al. (1999) included insertion–deletions in their phylogenetic analysis. Crosses and the meiotic behavior of chromosomes in hybrids revealed no chromosomal homology between H. marinum and other Hordeum species, and a unique genome was proposed for H. marinum (von Bothmer et al., 1986a, 1987). Isoenzyme analysis (Jørgensen, 1986), repeated DNA polymorphism (Svitashev et al., 1994) and C-banding patterns (Linde-Laursen et al., 1992) also separated H. marinum from the H-genome species and other groups. ITS sequence analysis separated H. marinum from the H-genome species, but strongly supported the sister group relationship between H. marinum and the H-genome species (Blattner, 2004). On the other hand, RAPD analysis showed that the cluster of H. marinum was included in the cluster of section Stenostachys, the largest group in the genus Hordeum (Marillia and Scoles, 1996). As also shown in a phylogenetic tree based on combined analysis of sequence data of the nuclear genes EF-G and DMC1, Hordeum marinum was incorporated in a clade with the H-genome group (Petersen and Seberg, 2003). All nuclear gene sequence data support the contention that Hordeum marinum is closely related to the H-genome group (Komatsuda et al., 1999, 2001; Petersen and Seberg, 2003; Blattner, 2004; and our RPB2 data).
The large group of H-genome species showed high variability in RPB2 sequences, but was highly homogenous in nuclear EF-G and DMC1 sequences (Komatsuda et al., 1999; Petersen and Seberg, 2003). The ITS data resulted in a well-supported split between the Old World and New World H genome species (Blattner, 2004). Differentiation of the New World H-genome Hordeum species from the Old World diploid species was indicated by meiotic analysis of interspecific hybrids within and between these regions (von Bothmer et al., 1995), and was confirmed by restriction fragment length polymorphism (Dubcovsky et al., 1997). C-banding (Linde-Laursen et al., 1992) and repeated DNA sequences (Svitashev et al., 1994) showed some differentiation among American and Eurasian H-genome species. Our RPB2 data also separated all New World H-genome species except H. patagonicum ssp. patagonicum from the Old World H-genome species, and hence support this funding of separation. Our data indicated that two diploid, central-Asiatic species, H. bogdanii and H. roshevitzii, are not as closely related to one another as previously suggested (Doebley et al., 1992). The third Asiatic species, H. brevisubulatum, is tetraploid and two distinguishable copies of sequences were isolated from this species. Phylogenetic analysis showed that the two copies are closely related to one another even though one copy showed a 59-bp deletion (H10227L). A relatively close affinity among H. brevisubulatum, H. bogdanii and H. roshevitzii has been suggested (Linde-Laursen et al., 1980; von Bothmer et al., 1986a; Jørgensen, 1986; Blattner, 2004). A relatively close relationship between H. brevisubulatum and H. bogdanii could be moderately confirmed by the RPB2 data (78 % bootstrap value; Fig. 3).
The RPB2 sequences provided a higher resolution within the H-genome group compared with all previous studies using single gene sequences. Even within H. patagonicum, separation could be found with ssp. setifolium, magellanicum and santacrucense forming a well-supported clade, whereas ssp. patagonicum was grouped with the Asian H. roshevitzii with high support, and ssp. mustersii was put into the New World clade (Fig. 3). As also shown with ITS rDNA data (Blattner, 2004), H. patagonicum ssp. patagonicum was separated from ssp. magelanicum, setifolium and santacrucense, whereas ITS sequences of ssp. mustersii were quite diverse, some in the ssp. patagonicum cluster, and some in the clusters of ssp. magelanicum, setifolium and santacrucense. The H. patagonicum group shows wide and complex patterns of morphological variation, and substantial chromosomal differentiation. Meiotic pairing suggested that the five taxa in this group shared the same ‘basic’ genome and are closely related (von Bothmer et al., 1986b). RPB2 (this study) and ITS sequence data (Blattner, 2004) strongly suggest that the H. patagonicum group is of paraphyletic origin and is highly differentiated. The treatment of these taxa as H. patagonicum subspecies is thus open to discussion and further investigations are required for a better understanding of the nature of this group.
On the basis of ITS rDNA sequences, Blattner (2004) suggested that H. pubiflorum is closely related to the H. patagonicum group, and might even have originated with a paraphyletic H. patagonicum. Our results confirm the close relationship between H. pubiflorum and the three H. patagonicum subspecies, setifolium, magellanicum and santacrucense.
Characterization of the RPB2 gene in Hordeum species found a miniature inverted-repeat terminal element (MITE) in the Xu genome. The 5′-flanking sequences of a conserved 15-bp TIR is duplicated TAA. Comparison of the conserved 15-bp TIR (GTACTCCCTCCGGTT) of the MITE in the Xu genome with that in the Tourist family (Bureau and Wessler, 1992) and that in the Tourist-like MITE mPing (Jiang et al., 2003) suggests that it is likely that the MITE element detected in the RPB2 region on the Hordeum Xu genome originated from the Stowaway nested within the Tourist element. An additional 27–36-bp indel was observed in Hordeum genomes. The indel length in H-genomes corresponds well with their geographical distribution. Our results suggest that the amount of activity and impact of the MITE on the Hordeum genome has been highly genome-specific in Hordeum species.
The maximum-parsimonious analysis demonstrated the four genomes in Hordeum to be subdivided into two groups, thus confirming the results of Komatsuda et al. (1999). Within the H-genome group the RPB2 sequences provided a higher resolution compared with all previous studies, thus suggesting that RPB2 may be useful for further phylogenetic analysis of the polyploid species in the genus Hordeum.
This research was supported with grants from NSERC (discovery grant 238425), Canadian Foundation for Innovation, a Senate Research Grant at Saint Mary's University, Japan Society for the Promotion of Science (JSPS) and Ministry of Agriculture, Forestry and Fishery of Japan (TRC1004). Thanks also go to two anonymous reviewers for valuable comments.