The transposable element landscape of the model legume Lotus japonicus

Abstract

The largest component of plant and animal genomes characterized to date is transposable elements (TEs). The availability of a significant amount of Lotus japonicus genome sequence has permitted for the first time a comprehensive study of the TE landscape in a legume species. Here we report the results of a combined computer-assisted and experimental analysis of the TEs in the 32.4 Mb of finished TAC clones. While computer-assisted analysis facilitated a determination of TE abundance and diversity, the availability of complete TAC sequences permitted identification of full-length TEs, which facilitated the design of tools for genomewide experimental analysis. In addition to containing all TE types found in previously characterized plant genomes, the TE component of L. japonicus contained several surprises. First, it is the second species (after Oryza sativa) found to be rich in Pack-MULEs, with >1000 elements that have captured and amplified gene fragments. In addition, we have identified what appears to be a legume-specific MULE family that was previously identified only in fungal species. Finally, the L. japonicus genome contains many hundreds, perhaps thousands of Sireviruses: Ty1/copia-like elements with an extra ORF. Significantly, several of the L. japonicus Sireviruses have recently amplified and may still be actively transposing.

Figure 1.—

Phylogeny and transposon display of copia-like elements. (A) The phylogenetic tree was generated using the RT domain (orange bar) of elements from L. japonicus (yellow) and a representative of each lineage from A. thaliana (green) and B. oleracea (red) and rooted with the corresponding RT from the yeast Ty1 element. This tree, and all trees in subsequent figures, was generated using the neighbor-joining method and bootstrap values were calculated from 250 replicates. The 13 bracketed copia-like lineages are those identified in previous studies, and the three sublineages of copia_Endovir-like are labeled (2, 1A, 1B) and discussed in the text. The blue bar indicates the region in the LTR used to generate primers for transposon display analysis from the element noted with a blue circle. (B) Transposon display analysis of one of the copia_Endovir-like sublineages. Sublineage-specific primers were designed and PCR analysis was performed with these primers and with a Bfa1+T primer and resolved on a 6% polyacrylamide gel. Lanes 1–6, genomic DNAs from individual (siblings) plants from Miyakojima (M); lanes 7–12, individual plants from Gifu (G).

Figure 2.—

Structural organization of a representative copia_Endovir-like element from the three sublineages. (A) Lj_copia_Endovir-like_2, (B) Lj_copia_Endovir-like_1A, and (C) Lj_copia_Endovir-like_1B. For each sublineage, shaded boxes represent the different ORFs and solid arrowheads represent the LTRs. The length of the average ORF (in aa) and LTR (in base pairs) is shown. For B and C, the gag and pol ORFs overlap and a frameshift is presumed to occur to generate both proteins.

Figure 3.—

Phylogeny and transposon display of MULE elements. (A) The phylogenetic tree was generated using the catalytic domain (orange bar) of the transposase of elements from L. japonicus (yellow) and a representative of each lineage from A. thaliana (green) and B. oleracea (red) and rooted with the transposase from a fungus. The three bracketed lineages include two previously identified (MURA and Jittery) and the new Hop-like lineage. The solid blue circle indicates the elements and the blue bar indicates the region used to generate primers for transposon display (TD) analysis. (B) Transposon display analysis of L. japonicus Hop-like elements. Sublineage-specific primers were designed and PCR analysis was performed with these primers and with a Bfa1+T primer and resolved on a 6% polyacrylamide gel. Lanes 1–12 are the same as Figure 1B.

Figure 4.—

Legume- and fungal-specific MULE lineage. Phylogenetic tree of legume and fungal MULEs. The phylogenetic tree was generated using the catalytic domain of Hop-like MULEs from L. japonicus and from the indicated legume and fungal species.

Figure 5.—

Structure and genomic origin of gene fragments in a chimeric Pack-MULE. Pack-MULE–TIRs are shown as black arrowheads and black horizontal arrows indicate target-site duplications. Homologous regions are connected with solid lines and the GenBank accession number of the TAC sequences where the Pack-MULE or the genomic copy was found is indicated. The chromosomal location of TAC AP007527 is unknown. For the Pack-MULE and the receptor-like protein kinase gene, exons are depicted as colored boxes and introns as the lines connecting exons. The light blue box represents part of an exon where the sequence is of unknown origin. The identity of the two other genomic copies (accession nos. AP007878 and AP007527) is not known and are depicted as narrow boxes.

Figure 6.—

Phylogeny and transposon display of PIF/Pong-like elements. (A) The phylogenetic tree was generated as described in previous figures. Lineages identified in previous studies are indicated next to the brackets. The solid blue circle indicates the elements used to generate sublineage-specific primers for transposon display analysis. (B) Transposon display using primers derived from two distantly related Pong lineages. See Figure 1B for details and DNA analyzed.

Figure 7.—

Structure of Pong1A and 3A showing regions used for transposon display primer design. The ORF1 and the transposase (TPase) regions are indicated for each Pong. Solid arrows represent terminal inverted repeats. Shaded areas share sequence homology between the two Pongs. Diagonal lines indicate the corresponding regions and sequences between the Pongs that were used for primer design.