Molecular Characterization of Type II Transposable Elements in Cowpea [ Vigna unguiculata ( L . ) Walp ]

Previous genetic studies in cowpea [Vigna unguiculata (L.) Walp] have shown that an active bipartite transposable element (TE) is responsible for a range of mutant phenotypes of its leaf, stem and flower. Since type II TEs have not been characterized at the molecular level in cowpea, this study was initiated to survey the presence of type II TEs in the cowpea genome. Type II TEs: Enhancer/Suppressor-mutator (En/Spm) and Miniature Inverted-repeat Transposable Elements (MITEs) were isolated and characterized. The sequence identity between the EnSpm TE clones was 46% at the nucleotide level (NL) and 30% at the amino acid level (AL) while that of MITEs was 71% at NL and 63% at AL. These cowpea En/Spm TEs were 80% homologous with En/Spm elements of other crops at NL and 46% at AL. The MITEs were 96% similar at NL and 18% homologous at AL. DNA gel blot analysis confirmed the presence of the En/Spm TEs in cowpea. RT-PCR (reverse transcriptase polymerase chain reaction) analysis showed that the VuEnSpm-3 and the MITE clone, VuPIF-1 were actively transcribed in wild type and mutant cowpea tissues. Overall, our data show that multiple, divergent lineages of En/Spm and MITEs are present in the cowpea genome, some of which are actively transcribed. Our findings also offer new molecular resource to further investigate the genetic determinants underlying previously described mutant cowpea phenotypes.

In plants, TEs contribute significantly to the size, structure and plasticity of the genome [7].For example, in maize and other plant species, up to 80% of the genome consists of transposable elements [8].TEs also play an active role in genome evolution [3] [9] [10] by helping their hosts adapt to new conditions by conferring useful traits [11].The identification and characterization of transposons in a given host can greatly assist genetic studies of that organism [12].Transposable elements have been used for improvement in many crop plants such as sorghum, tomatoes, rice and maize [1] [4]; in maize, TEs were used to develop Striga tolerant lines [13], and they have been used as markers to assess genetic segregation in sorghum [14], for phylogenetic studies [15], gene tagging and reverse genetics in plants [16]- [18].
In eukaryotes, TEs are classified into two classes based on their mode of transposition: class I elements (retrotransposons), which move via an RNA intermediate, produced via reverse-transcription, before being inserted into the genome in a "copy and paste" manner.By contrast, class II elements move by a "cut and paste" mechanism without RNA intermediate.Type II TEs consist of an autonomous element-a transposase-and non-autonomous elements with terminal inverted repeats (TIR) of 10 -200 bp [3].They are usually characterized by target site duplications.Class II elements are classified into several super families [1], including the Enhancer/ Suppressor-mutator (En/Spm) [19] [20]), the Activator/Dissociation (Ac/Ds), Mutator-Like Elements (MULEs), Mariner-Like Elements (MLEs) and Miniature Inverted-repeat Transposable Elements (MITEs) [21].The En/ Spm share a common sequence-5'-CACTA-3' at their TIR and the transposase is highly conserved among plants.They have been found and characterized in Gramineae [22], Leguminosae, Solanaceae, Chenopodiaceae, Alliaceae species [2] and Euphorbiaceae [23].MITEs are characterized by their small size, usually less than 500 bp, lack coding capacity and have short TIRs.They can be further classified as Tc1/Mariner-like or PIF/Harbinger-like based on their association with established super families [24].
Cowpea [Vigna unguiculata (L.) Walp] is a drought-tolerant, fast-growing, and highly nutritious legume of particular importance in the semi-arid regions of tropical countries in Africa, Asia and southern America.Previous genetic studies in cowpea have shown that an active bipartite transposable element system is responsible for a range of mutations affecting leaf, stem and flower morphology and pigmentation [25]- [29].As characteristic of most TE-induced mutations, the mutants were not stable; an example is that of the flower mutation, both mutant and wild type flowers were found on the same plants [30].In this study, we present, for the first time, the isolation and molecular characterization of type II TEs, including members of the En/Spm and MITE superfamilies based on the analysis of these unstable mutants.Findings from this study will contribute to and possibly enhance the application of TEs in future breeding efforts of cowpea.

Plant Material and Nucleic Acid Extraction
Cowpea accessions and parental lines used in this study were mostly obtained from the Department of Crop Protection and Environmental Biology, University of Ibadan, Ibadan, Nigeria as listed on Table 1.Reduced Pet-  [31].RNA was extracted from leaves, stem and flowers as described by [32] and with the RNeasy kit following the manufacturer's instructions (Qiagen, Valencia, CA, USA).

Pedigree of Mutants
Mutants were obtained from crosses made by earlier workers on cowpea TEs [26]- [29], their pedigree is shown in Figure 1.

PCR Amplification of Fragments
Total DNA (250 ng) was amplified by polymerase chain reaction (PCR) in a PTC 100 Thermal Cycler (MJ Research Inc.) using the conditions and primers described by [2] [23] for the amplification of En/Spm-like transposases, and those used by [33] for the amplification of PIF/Harbinger-like MITEs.Others included primers for Mariner-like elements [24], Ac/Ds elements [34], Zaba elements [35] and Mutator-like elements [36].Specific primers were used for the reverse transcriptase polymerase chain reaction (RT-PCR) were designed using Primer 3 program [37].In all cases, PCR amplification was performed in a 50 μl reaction using 1 unit of Taq DNA polymerase (Bioline, USA).

Cloning Procedures and DNA Sequencing
PCR amplicons obtained from RFM were gel purified using the Qiaex gel extraction kit (Qiagen, USA).PCR amplicons were ligated into a pCR8-GWTOPO (Invitrogen, CA, USA) or pDrive (Invitrogen, CA, USA) and transformed into Escherichia coli DH5α competent cells according to standard procedures [31].Following blue white selection protocol, recombinant clones were grown in liquid LB medium and plasmid DNA was isolated using ultra-pure plasmid kit (Baygene, USA).Inserts were sequenced using universal primers by the Iowa state University sequencing facility (Iowa, USA).All clones were sequenced in both orientations.The DNA sequences were manually edited and any sequence ambiguities were resolved by re-sequencing.For each PCR reaction, three to five independent plasmid clones were sequenced to enable detection of any size or sequence heterogeneity present in the clones.

Database Searches and Sequence Comparison
The sequences obtained from this study were compared with other sequences in the database by using BLASTN searches against the GenBank non-redundant database of the National Center for Biotechnology Information (NCBI) using default parameters [38].Cowpea TEs sequences were aligned using the programs QAlign [39], MEGA [40], and CLUSTAL W. Phylogenies of the clones were obtained using the relaxed dissimilarity algorithm of the Neighbour Joining program.Trees were constructed using Tree View [41].

Southern Blot Analysis
Twenty microgram cowpea DNA was digested overnight using EcoR1 and Hind III at 37˚C and separated on a 1% TAE gel following digestion.DNA was transferred to Hybond N+ nylon membrane (Amersham, USA) using standard procedures [31] as follows: the membrane was prehybridized at 42˚C in DIG easy HyB buffer (Roche, USA) in a hybridization oven.Probes were prepared using the DIG labeling kit (Roche, USA), and hybridization was carried out as recommended by the supplier (Roche, USA).After hybridization, the membrane was exposed to an X-ray film and developed using standard techniques.

RT-PCR Analysis
Primers were designed from the transposase gene of VuEnSpm-1, VuEnSpm-3 and VuPIF-1.The lists of the primers used are shown in Table 2. RT-PCR was carried out using RNA extracted from mutant and wild type plant tissues, including leaves, stems and flowers.The RNA was reverse transcribed into cDNA using the Superscript III reverse transcriptase of Invitrogen (CA, USA) according to the manufacturer's specifications.RT-PCR products were analyzed on a 2% agarose TAE gel.

Identification of Type II Transposable Elements in Cowpea
PCR amplicons were obtained from the all the type II TEs investigated namely: En/Spm, MITEs, Ac/Ds, MLEs and mutator elements.However, subsequent BLAST searches showed that sequences with significant homology to previously characterized TE were observed only for En/Spm and MITEs.In case of Ac/Ds, MLE and the mutator elements, GenBank searches did not retrieve sequences with significant sequence homology to previously identified Ac/Ds and mutator elements.Four En/Spm clones of about 650 bp each were sequenced and analyzed.These sequences were submitted to the NCBI GenBank with the following accession numbers: FJ526201, FJ526202, FJ526203, and FJ526204.A total of five MITEs clones of approximately 500 bp each were sequenced with only two showing significant homology to previously identified MITEs, and in particular to the PIF/Harbinger subfamily.These two sequences were also submitted to GenBank with accession numbers GQ422756 and GQ422757.The En/Spm-like elements described here are the first class II TE reported in Vigna unguiculata using molecular techniques.En/Spm-like elements were present in all mutant and wild type cowpea genotypes used in this study as assessed by PCR analysis (Figure 2

Comparisons within the Cowpea En/Spm and MITE Clones
For each type of TE, the different clones were aligned to assess the level of diversity present within their class.Sequence analysis of the three clones designated VuEnSpm1, VuEnSpm2 and VuEnSpm3, showed only 46% similarity at the nucleotide level and 30% at the amino acid level.Thus, the VuEn/Spm elements showed considerable sequence diversity.VuEnSpm1 and VuEnSpm2 are highly homologous with 72% nucleotide identity, while VuEnSpm3 showed only 41% and 52% nucleotide sequence identity to VuEnSpm1 and VuEnSpm2, respectively.All three clones contained an ORF of 192 amino acids encoding part of the En/Spm transposase as expected, which is an indication of possible active transposons.A similar alignment was made for the two clones of VuPIF, designated VuPIF1 and VuPIF2.These sequences were 71% similar at the nucleotide level and 63% at the amino acid level.

Comparison between TEs from Cowpea and Other Plants
A multiple sequence alignment of cowpea TEs and other plant species followed by phylogenetic analysis showed the genetic relationship between VuEnSpm1, VuEnSpm2, VuEnSpm3 and En/Spm-like transposons (Figure 3) as follows: VuEnSpm1 and VuEnSpm2 were more closely related to En/Spm of Pisum sativum while VuEnSpm3 was more closely related to En/Spm of Cicer arietinum and Manihot esculenta.However, En/Spmlike elements from Beta and Allium spps were grouped with other leguminous plants such as Lens culinaris, Cajanus cajan and Lens esculentum.Overall, top BLAST hits were obtained for leguminous plants (Figure 3) and they showed a higher level of relatedness to one another.Similarly, when a BLAST search was performed using the VuPIF sequences, top hits included Sorghum halenpense and Dendrocalamus minor (Table 1).In addition, more hits with sequences from mostly monocot plants such as Pennisetum glaucum, Zea mays and Saccharum hybrid cultivar clones, with a nucleotide sequence identity ranging from 90% to 99% and amino acid identity of 67% to 72% were obtained from the BLAST search.When the conserved regions of the top hits were aligned with the VuPIF clones, the phylogenetic analysis showed two groups with the two VuPIF clones together and the second group consisted of 2 sub-groups of monocots and the only dicototyledonous crop (Figure 4).The overall mean distance between the amino acid of PIF TEs from other crops was 82%, hence a similarity of 18% from the analysis with MEGA version 4.

Reverse Transcriptase PCR (RT-PCR) Analysis
To assess whether the TE isolated from cowpea are actively transcribed, RT-PCR reactions were conducted using cDNA from both the wild (non-mutant) and mutant types of different plant tissues.RT-PCR amplicons were obtained for VuEnSpm3 and VuPIF1 while VuEnSpm1 did not show any transcript in all the mutants and wild type tested.Tissue-specific amplifications showed that VuPIF-1 transcripts are present in both mutant and wild type tissues with expression of two copies (faint) of MITEs transcripts in the leaves of the unifoliate leaf mutant (LM1) compared to wild type leaf tissues (Ife brown) (Figure 5).In addition, only VuPIF-1 transcripts were present in the stem tissues of the branching habit mutant LM2 (Figure 5) while it was absent in the stem of wild type (Tvu 1) (not shown).A similar result was obtained in the flowers of the Rose-like Flower Mutant (RFM) where VuEnSpm-3 transcripts are more pronounced in the mutants than in the wild type (not shown).Figure 5 shows typical amplification obtained from the different plant parts.

Discussion
This study aims to authenticate the presence of type II TEs in cowpea, Vigna unguiculata and to provide an insight into the type of TEs potentially implicated in mutations reported in previous studies through classical breeding.To our knowledge, this is the first report of molecular characterization of type II TEs in cowpea.The protein BLAST result of the TEs showed that there are partial open reading frames (ORFs) that encode products involved in transposition.The VuEnSpm encodes a transposase while the VuPIF, was found to be a non-functional transposase protein.These findings will constitute a useful source of information to the crop's genome annotation [42].
Also, it has shown the degree of genetic similarity of clones of these types of TEs in cowpea .Similar levels of divergence were obtained in the En/Spm of chickpea [2].The multiple alignment and phylogenetic analysis with similar types of TEs in other crops shows the relationship between these organisms relative to cowpea TEs.This finding is comparable to the findings of [2], who reported that En/Spm-like transposon sequences from legume species cluster together.Overall, our data suggests that lineages of En/Spm and MITEs are present in cowpea.En/Spm-like transposable element was present in low copy in all the cowpea varieties investigated.The southern blot analysis revealed about 5 -6 bands, which may be an indication that En/Spm-like transposons are present in the cowpea genome in low copies.This is similar to the report of [2] of medium and low copies of this type of TEs in the plants they studied.A low copy number has been reported for many other En/Spm-like elements   in plants, including other legumes [2], Poaceae [22] as well as others plant species [43]- [46].
A substantial proportion of plant genomes are made up of transposable elements and they contribute both to their structure and evolution [24].Therefore, the amplification obtained for the En/Spm in all the lines analyzed is an indication that the high mutability observed in cowpea mutants are likely due to the activities of transposable elements; with the En/Spm and MITEs also possibly responsible for these mutations.Therefore, the results obtained in this study corroborates the presence of these types of class II TEs in cowpea and therefore supports the fact that the mutations observed through classical breeding might have been as a result of the activities of transposable elements.[25]- [28] described the previous work done in obtaining these mutants, which revealed that there was somatic instability and possible insertion and activity of TEs in the tissues in which the mutation took place.

Vitis vinifera protein
The RT-PCR analysis provides additional evidence that these TEs are transcribed differently in the different plant parts/tissues investigated both in the mutants and in the parental lines/wild type.Firstly, the fact that RT-PCR amplicons were obtained for VuEnSpm3 and VuPIF1 while VuEnSpm1 did not show any transcript in all the mutants and wild type tested, suggests that VuEnSpm-3 is actively transcribed in the tissues analyzed as opposed to VuEnSpm-1.This corroborates the activity of TEs as a result of somatic instability by earlier workers.

Conclusion
In conclusion, this study presents evidence to confirm the presence of EnSpm TEs and MITEs in the cowpea genome.Our results will contribute significantly to the improvement of this crop by providing important genomic resource that was previously unavailable and open up avenues for cowpea TEs to be used in molecular marker development [14] or gene tagging [18].Availability of complete cowpea genome sequence will further expand the discovery and application of TEs in cowpea genomic studies.
(a)).Likewise, MITE-like sequences were present in all cowpea genotypes based on PCR analysis (data not shown).The southern blot analysis confirmed the presence of EnSpm (Figure2(b)).Five to six distinct bands were obtained in the cowpea lines under Figure2.

Figure 3 .
Figure 3. Dendrogram showing the relationship between VuEn/Spm clones and those from other plants from BLAST search.

Figure 4 .
Figure 4. Dendrogram showing the relationship between VuPIF clones and those from other plants from BLAST search.

Table 1 .
The list of the mutants and wild types used for this study.Ife brown and Ife BPC.The lines Tvu 1 and Tvu 1509 were obtained from the International Institute of Tropical Agriculture, Ibadan, Nigeria and Tvu 940151 was obtained from the University of California Davis, Davis, CA, USA.Total DNA was extracted from young leaves as described by

Table 2 .
Characteristics of cowpea type II TE characterized in this study.