Ty1-copia group retrotransposons and the evolution of retroelements in several angiosperm plants: evidence of horizontal transmission.

The phylogenetic relationships among thirty-seven new Ty1-copia group retrotransposons in seven angiosperm plants were examined by reverse transcriptase and ribonuclease H sequence analysis. Distribution pattern of the retrotransposons of closely related plant species generally reflects a close phylogenetic relationship. In contrast, we found that several retrotransposon sequences from the same genome exhibited a high degree of divergence and had a relatively high degree of identity versus retrotransposon sequences from widely divergent species, including an ancestral phytopathogen fungus. This finding supports the hypothesis that the horizontal transmission from phytopatogen organism to the host flowering plants could have played a role in the evolutionary dynamics of Ty1-copia group retrotransposons.

Background

The retrotransposons constitute the most common class of repetitive DNA in a broad range of taxa, interspersed within the genome and ubiquitous in eukaryotes. It is currently believed that they have played a role in shaping genome architectures contributing significantly to the remarkable variations in genome size and evolution by changing the structures and expression patterns of genes [1]. The advent of PCR methods for amplifying conserved domains of retrotranspons has led to the rapid increase of finding on sequence evolution and phylogenetic relationships of the Ty1-copia in plant genomes [2, 3]. In this scenario, the prevailing view is that the Ty1-copia retrotransposons existed in the early phase of eukaryote life and diverged into heterogeneous subgroups before modern plant species aroused [4]. Previous studies have shown large heterogeneity within LTR retroelements as a result of mutations accumulated during evolution. These mutations originated different lineages of retrotransposons, characterized by differences in activity and chromosomes distribution [5]. Retrotransposon are usually located in intergenics regions and tend to accumulate in centromeres, telomeres and heterochromatic regions. In many cases, the same retrotransposon is found in different cultivars of the same species and in several related species, but not in far distant taxa [2, 3].

Phylogenetic analysis of reverse transcriptase (RT) sequences from the Ty1-copia retrotransposons showed that the sequence heterogeneity is generally proportional to the evolutionary distance of their host plant species, implying that the vertical transmission has been a major factor in the evolution of Ty1- copia retrotransposons within the plants [6]. On the other hand, based on the high sequence conservation across large evolutionary distances and a lower sequence similarity between elements from the same genome and related species, it has been postulated that the horizontal transmission (HT) of Ty1-copia retrotransposons can occur in plants [3]. Few cases of HT involving transposable elements have been reported, and most involved Class II transposable elements [7]. In this study we have isolated and characterized new Ty1-copia retrotransposons in seven dicotyledon plants and carried out a phylogenetic analysis. Distinct evolutionary lineages of Ty1-like were identified, and according to several authors significant variation among Ty clade elements were observed. The similarity between Solanum melongena and Phytophthora infestans retroelements shown by molecular analysis and phylogenetic relationships was intriguing since these two species are unrelated. This similarity suggested that HT of the LTR retrotransposon named Phythophthora (GenBank accession no. CAI72292.1) has occurred within the Solanum melongena genome.

Methodology

Plant material and genomic DNA isolation:

Taxonomic classification of the studied angiosperm species is listed in Table 1 (see supplemental material) Plants were obtained from a nursery in South-Italy (Caserta). Total DNA from each plant species was extracted from the fresh leaf tissues according to Fantaccione and coworkers [8].

PCR strategy, internal controls, cloning and sequencing:

Total DNA (2 µg) was completely digested with EcoRI (Roche) and digested samples (200 ng) were used for PCR amplifications with three combinations of degenerate oligonucleotide primers (RT1/RNase H1; RT2/RNase H1; RT3/RNase H1). The sense RT1 (5'- GATGTDAARACKRCNTTYYTD-3'), RT2 (5'- ATGGAGCARCCDGAMGGHTTY-3'), RT3 (5'- TATGTDGATGAYATGYTDATT-3') and the antisense RNase H1 (5'-CCTCACATCWATRTGYTTBGW-3') degenerate oligonucleotide primers, which respectively correspond to the conserved RT (DVKTA(T)FL, MEQPE(D)GF, YVDDMLI) and ribonuclease H (RNase H) (T(S)KHIDVR) peptide motifs of the Ty1-copia group retrotransposons were used. For each species, PCR amplifications (0.8-1 kb) were performed in 50 µL volume of 2 mM MgCl2, 200 µΜ of each dNTP, 200 µΜ of each primer and 2 U of FastStart Taq DNA Polymerase (Roche). After an initial denaturation step for 4 min at 94°C, the PCR reactions were subjected to 30 cycles of amplification consisting of 1 min denaturation at 94°C, 1 min annealing at 44°C and 1 min extension at 72°C with a 10 min final extension at 72°C. PCR products were purified from the agarose gel, cloned in pDrive vector (Qiagen) and sequenced using ABI 377 automated DNA sequencer (Applied Biosystems). Internal PCR controls: Pinci1- 26 forward (5'-GTATGTGGGGCTCCAAGCC-3') and Pinci1-26 reverse (5'-GCGGCTGCACGGTCATTG-3') were designed on Pinci1-26 noncoding infection-specific mRNA sequence of P. infestans (GenBank accession no. EF091740). The control PCR was carried out as described above. All clones were named as follows: The first letter meant cloning product, the second and third letter indicated the genus and species (pC.a.= C. Annuum, pC.m.= C. Melo, pC.s.= C. Scolymus, pH.m.= H. Macrophylla, pP.a.= P. Avium, pS.m.= S. Melongena, pV.v.= V. Vinifera) the fourth letter identified the numeric sequence of the screened clones.

Databases, sequence analysis and phylogenetic trees:

Nucleotide sequences were compared to the GenBank-NCBI database using the BLAST network service (http://www.ncbi.nlm.nih.gov/BLAST/) and analyzed using CENSOR software (http://www.girinst.org/censor). Multiple sequence alignments were performed utilizing the Clustal (1.8) program [8] from the DDBJ Homology Search system (http://www.ddbj.nig.ac.jp) with blosum matrix. The phylogenetic trees were constructed by Neighbor-Joining method [9] and plotted by DrawTree using the PHYLIP software package [10]. Nonparametric bootstrap resampling (1000 replicates) was used to estimate the clade robustness. Cvalues of genome plants were estimated using the Angiosperm DNA C-values database available at http://www.rbgkew.org.uk./cval/database1.thml site.

Analysis of synonymous and nonsynonymous substitution per site and codon bias:

pC.a.D and pS.m.D retrotransposons were compared to the P. infestans RT-ORF sequence (GenBank accession no. CAI72292.1) to examine the differences in synonymous and nonsynonymous changes using the SNAP (Synonymous/Nonsynonymous Analysis Program) tool in the HIV Sequence Database (http://www.hiv.lanl.gov). This program uses the algorithm devised by Nei and Gojobori [11]. Codon bias as determined by the Nc value was computed using the Codon W program (http://bioweb.pasteur.fr/seqanal/interfaces/codonw.html).

Discussion

Identification of Ty1-copia retrotranspons in various plant species:

In order to study the evolution of Ty1-copia, we performed a PCR-based assay using genomic DNAs from seven plant species. By using the degenerate oligonucleotide primers described in methodology, single products of the expected size (approx. 1000 bp) were obtained and 52 clones were sequenced Table 1 (see supplemental material). Based on a search against the GenBank database and CENSOR software, 37 sequences showed clear similarity (40%-60%) to RT-RnaseH domains Table 1 (see supplemental material). The majority of detected Ty1-copia elements revealed defective translated products for the presence of stop codons and short indels that resulted in frameshift and frame stop mutations, resulting transcriptionally or transpositionally inactive. The variations within the inactive elements can contribute to the evolution of the plant genome, leading to gene duplication events and providing new properties in the retrotransposition mechanisms [12].

Phylogenetic analysis of Ty1-copia-like retrotransposons in plants:

Phylogenetic relationships were determined by comparing the thirty-seven clones among them or with retrotransposons from other plant species (Figure 1, Figure 2, Figure 3, Figure 4 & Table 1, 2 see supplemental material) Phylogenetic analysis in (Figure 3) showed that the retrotransposons were clustered in three clades (A, B and C). Clade A was further divided in four subclades (AI, AII, AIII and AIV), as judged by the bootstrap values. The subclades AI and AIII contained members from monocotyledon and dicotyledon plants. These results could be explained by the universal idea that the different retrotransposons already existed before the divergence between monocotyledon and dicotyledon plants and have been vertically transferred [3]. Vertical relationships were found also among several copia-like retrotransposons from related plant species belonging to the Solanaceae family (L. esculentum, N. tabacum, L. chilense, C. annuum, S. melongena and I. batatas), as shown in subclades B and C (Figure 3). However, vertical and horizontal transmissions are not mutually exclusive [9]. In fact, pS.m.D and pC.a.D retrotransposons were clustered in the subclass AI together with same dicotyledon plant retrotransposons (B. rapa and A. thaliana). In addition, pC.s.D retrotransposon and a monocotyledon retrotransposon from O. sativa genome clustered together in the subclade AIII (79% of positivity score). Similar phylogenetic relationships were observed about plant retrotransposons belonging to the Rosids subclass, too (Figure 4).

Figure 1

Phylogenetic evidence of horizontal transmission from P. infestans to their host flowering plants. Bootstrap values of 50% and higher are shown (1,000 trials).

Figure 2

Phylogenetic trees of the Ty1-copia group retrotransposons from seven plant species based on RT and RNase H domains. Numbers on the branches are the bootstrap percentages for 1,000 replicates.

Figure 3

A phylogenetic tree based on the RT and RNase H domain analysis from Ty1-copia group retrotransposons in Asterids species and in different plant using the Neighbor- Joining method. Numerals adjacent to branches indicate are the bootstrap values for 1,000 replicates. Ty1-copia retrotransposons previously identified are shown as accession number and source species.

Figure 4

A phylogenetic tree based on the RT and RNase H domain analysis from Ty1-copia group retrotransposons in Rosids species and in different plant using the Neighbor-Joining method. Bootstrap values of 50% and higher are shown (1,000 trials). Ty1-copia retrotransposons previously identified are shown as accession number and source species.

Interestingly, when we submitted the pC.a.D and pS.m.D retrotransposons from C. annuum and S. melongena genomes (Asterids subclass and Solanaceae family) to a Blastx search (http://www.ncbi.nlm.nih.gov/BLAST/), the P. infestans fungal hit (GenBank accession no.CAI72292.1) matched with 59% and 63% positivity scores, respectively. This plant pathogen parasitizes a large number of host species, including S. melongena and C. annuum [13, 14]. Although fungi and angiosperm plants last shared a common evolutionary ancestor about 2.5 billion years, pC.a.D and pS.m.D elements are included in the same group with P. infestans retrotransposon (Figure 1). Codon bias is a possible selective constraint on synonymous nucleotides [15]. A bias in codon usage occurs when synonymous codons are not all used at the same frequency in coding DNA and such bias in codon usage could result from mutational pressure or from selective pressure [16]. A common measure of codon bias is the effective number of codons (Nc), which can vary from 21 (when only one codon is effectively used for each amino acid) and 61 (when codons are used randomly) [17]. The P. infestans, S. melongena and C. annuum retrotransposon sequences have an Nc value of 52.81, 51.84 and 47.60, respectively, consistent with only relatively moderate codon bias. Since selection operates much more efficiently on nonsynonymous base substitution [17], a common measure of selection for function is the ratio of synonymous to nonsynonymous substitution per site (dS/dN). When RT sequences from pC.a.D and pS.m.D retrotransposons were separately compared to P. infestans putative polyprotein using the SNAP program, the corrected frequencies of nonsynonymous substitutions for this region were 0.3408 and 0.3481, and the corrected frequencies of synonymous substitutions ranges were 0.8415 and 0.9013. Importantly, the ratio dS/dN for these two analysis pairs was only 2.4695 and 2.5895, consistent with a weak selection for function at the amino acid level. To eliminate any possible DNA contamination from host-oomycete interaction, we instituted internal PCR controls using primers designed on Pinci1-26 pathogen sequence from P. Infestans [13] and we found no PCR amplification products that are no contamination.

Conclusion

Our analysis showed that the phylogenetic relationships between Ty1-copia retrotransposons in several plant genomes are complex and consistent with the universal idea that today the different retrotransposons pervade the entire plant kingdom as large and highly heterogeneous populations. The high degree of similarity between retrotransposons from divergent species and, in addition, the sequence divergence in the same species and/or related species suggests that the horizontal transmission events occurred in Ty1-copia evolution. In this scenario, we have attempted to understand whether the relatively high degree of the sequence conservation between the retrotransposons from pathogen oomycete P. infestans and two Asterids species may be a new example of horizontal gene transfer.

Taken together, our data support the horizontal transmission hypothesis since it seems unlikely that this high degree of the sequence conservation over such a length of time is the product of standard vertical transmission. Our example is a minor contribution and further studies are needed to exhaustively identify retrotransposons from plant species to gain further insight into the evolutionary history of the relationship between retrotransposons and their genome hosts.