Radiation of the Tnt1 retrotransposon superfamily in three Solanaceae genera

Abstract

Background: Tnt1 was the first active plant retrotransposon identified in tobacco after nitrate reductase gene disruption. The Tnt1 superfamily comprises elements from Nicotiana (Tnt1 and Tto1) and Lycopersicon (Retrolyc1 and Tlc1) species. The study presented here was conducted to characterise Tnt1-related sequences in 20 wild species of Solanum and five cultivars of Solanum tuberosum.

Results: Tnt1-related sequences were amplified from total genomic DNA using a PCR-based approach. Purified fragments were cloned and sequenced, and clustering analysis revealed three groups that differ in their U3 region. Using a network approach with a total of 453 non-redundant sequences isolated from Solanum (197), Nicotiana (140) and Lycopersicon (116) species, it is demonstrated that the Tnt1 superfamily can be treated as a population to resolve previous phylogenetic multifurcations. The resulting RNAseH network revealed that sequences group according to the Solanaceae genus, supporting a strong association with the host genome, whereas tracing the U3 region sequence association characterises the modular evolutionary pattern within the Tnt1 superfamily. Within each genus, and irrespective of species, nearly 20% of Tnt1 sequences analysed are identical, indicative of being part of an active copy. The network approach enabled the identification of putative "master" sequences and provided evidence that within a genus these master sequences are associated with distinct U3 regions.

Conclusion: The results presented here support the hypothesis that the Tnt1 superfamily was present early in the evolution of Solanaceae. The evidence also suggests that the RNAseH region of Tnt1 became fixed at the host genus level whereas, within each genus, propagation was ensured by the diversification of the U3 region. Different selection pressures seemed to have acted on the U3 and RNAseH modules of ancestral Tnt1 elements, probably due to the distinct functions of these regions in the retrotransposon life cycle, resulting in both co evolution and adaptation of the element population with its host.

Figure 1

Schematic representation of the amplified retrotransposon fragment. LTR: long terminal repeat. RNAseH: ribonuclease H; Linker: noncoding region with PPT (polypurine track); U3: unique 3' RNA region; R: repeat RNA; and U5: unique 5'RNA region. Thick arrows indicate the position of the primers used in the amplifications.

Figure 2

Sequence alignment of the partial amplified fragments. Representatives from Tnt1A, B and C; Retrolyc1A and B; and six sequences from Retrosol were included (names in bold). Underlined regions denote: RNAseH (ribonuclease), Linker, U3 (unique 3'RNA region), TATA, R (repeat RNA) and U5 (unique 5' RNA).

Figure 3

Phylogenetic analysis of Tnt1 Solanaceae superfamily. Phylogenetic analysis was performed with 197 sequences amplified from different species of Solanum (species are indicated by a code cited in Table 1 and different clones are indicated by numbers), and representative sequences from Tnt1 A, B and C, and Retrolyc1 A and B. The aligned nucleotide sequences span the last fragment of the RNAseH domain, the linker, U3, R and part of the U5 region.

Figure 4

Sequence divergence distribution along the amplified fragments. The values on the x-axis correspond to the nucleotide position of the fragment amplified. The values on the y-axis are the nucleotide diversity (π) measure, the percentage of divergent nucleotides relative to the number of informative bases, calculated using a sliding window of 10 bp and a step of 1 bp, by the DnaSp program [31]. The position of the TATA box and U3 are indicated. Data are shown for all sequences amplified and for pair-wise comparisons within groups without gaps.

Figure 5

Median-joining networks of the RNAseH coding region. The network was constructed using 443 Tnt1 superfamily sequences. Circles denote sequence types: yellow circles denote sequences types from Retrosol, blue circles denote sequences from Tnt1, and red circles denote sequences from Retrolyc1. (A) The size of the circle is proportional to the number of sequences, with one sequence type indicated in the bottom-left corner. Lines denote substitutions; with one-step distance indicated in the bottom-left corner. Reconstructed nodes are identified as black circles; they are absent in the dataset because they were not sampled or lost. (B) Zoom of the median-joining networks of the RNAseH coding region. The size of the circles is proportional to the number of sequences, with one sequence type indicated in the bottom-left corner. Lines denote substitutions and numbers on the lines indicate the substitution position on the sequence. One-step distance is indicated in the bottom-left corner. The labels in the nodes illustrate the nomenclature used to refer to the original master sequence of each genus, called α type: α1 is the master copy from Retrosol, α2 and α3 are from Tnt1 and α4, α5 and α6 are from Retrolyc1.