Search

Additional file 9: File S2. Identified RAGL loci mapped onto nucleotide sequence.

Additional file 8: File S1. Detailed presentation of the sequence information about the new RAGL loci identified in this study: detection relevant data (e-values and query-target alignments from TBLASTN), TIR and TSD detection information, predicted protein sequences, and additional relevant observations regarding some of the sequences.

Additional file 2: Figure S2. (a) Alignments of four groups of protostome RAGL transposon TIRs. Sequences of 5′-TIR and 3′-TIR (reverse complement) pairs are aligned from four groups: C. virginica (I), P. imbricata (II and III) and nemertean N. geniculatus (IV). Despite both being mollusks, the C. virginica and P. imbricata TIR sequences are very dissimilar both in sequence and length. None of the 4 groups contain an RSS nonamer-like region, however, the C. virginica 5′ and 3′ TIRs exhibit a length difference of 11 bp, reminiscent of the 11 bp difference in lengths of the spacers in the 12RSS and 23RSS that underlies the 12/23 rule of V(D) J recombination [3, 6]. Similarity color code: dark grey, fully conserved; light grey, partially conserved; white, no similarity. (b) Potential RAG-derived non-autonomous transposable elements in protostomes and deuterostomes. Summary of the potential RAG-derived non-autonomous elements identified in WGS database in organisms where the RAGL transposon was detected. The configuration of these elements, as well as TSD sequences (if present) are indicated. The e-values shown derive from blastn searches of the WGS project of each organism using previously detected RAGL TIR-containing margins (200 bp) as queries.

Additional file 6: Table S1. Genomic and transcriptomic data availability for bilaterian invertebrates. List of the bilaterian species for which there are Transcriptome Shotgun Assembly (TSA) and/or Whole Genome Shotgun (WGS) projects in the NCBI database as of February 26, 2019. Gnathostomata species are not listed. The taxonomic identifier is given in the NCBI taxid column and corresponds to that used in NCBI databases. The number of projects available is indicated. In the Number of TSA projects column, transcriptomic projects that were not marked as “TSA project” are indicated in parentheses.

Additional file 7. Alignment S1 Multiple sequence alignment of (a) RAG1/RAG1L and (b) RAG2/RAG2L predicted proteins. Domains, sequence motifs, secondary structure assignment, protein-protein and protein-DNA contact interactions (within 5 Å) displayed above the alignment derive from the BbeRAGL cryo-EM structure (PDB: 6B40). Additionally, for RAG1/RAG1L (a), acidic catalytic residues, red; active site residue mouse H795, purple; zinc coordinating residues within ZDD (*) and ZnC2 and ZnH2 (#) are indicated above the sequences, while for RAG2/RAG2L (b) the beta sheet regions of each kelch-type blade and the GG motifs are shown above the alignment. Locations at which coding sequences span exon boundaries are underlined. Sequence descriptions including references to genomic, transcriptomic or protein databases are shown at the end of the alignment, along with a legend of the symbols used.

Additional file 6: Table S1. Genomic and transcriptomic data availability for bilaterian invertebrates. List of the bilaterian species for which there are Transcriptome Shotgun Assembly (TSA) and/or Whole Genome Shotgun (WGS) projects in the NCBI database as of February 26, 2019. Gnathostomata species are not listed. The taxonomic identifier is given in the NCBI taxid column and corresponds to that used in NCBI databases. The number of projects available is indicated. In the Number of TSA projects column, transcriptomic projects that were not marked as “TSA project” are indicated in parentheses.

Additional file 4: Figure S4. Complementary RAG1L phylogenetic analyses using PhyML implementation [49, 50]. Trees are displayed as in Fig. 2b except that branches with bootstrap numbers below 50% were not collapsed together. Different substitution models were used as follows: (a) LG + G + I + F model selected via AIC minimization (b) LG + G + I model selected via BIC minimization (c) WAG substitution model.

Additional file 3: Figure S3. Additional phylogenetic analyses. (a-d) Detailed phylogenetic trees of RAG1 and RAG1L protein sequences including several from: (a) Cephalochordata - indicated with blue shading., (b) Echinodermata - blue shading, (c) Mollusca - orange shading and (d) including one from cnidaria - green shading (e) RAG2/RAG2L phylogenetic trees. Trees were built using Maximum Likelihood and WAG substitution model as implemented in MEGA X [27] and are displayed as in Fig. 2b except that branches with bootstrap numbers below 50% were not collapsed together. Trees were built from a variable number of significant positions of their alignment: (a) 289, (b) 398, (c) 354, and (d) 469 respectively .

Additional file 4: Figure S4. Complementary RAG1L phylogenetic analyses using PhyML implementation [49, 50]. Trees are displayed as in Fig. 2b except that branches with bootstrap numbers below 50% were not collapsed together. Different substitution models were used as follows: (a) LG + G + I + F model selected via AIC minimization (b) LG + G + I model selected via BIC minimization (c) WAG substitution model.

Additional file 5: Figure S5. (a) Conservation of functional relevant amino acids in RAG1 and RAG1L. The alignments depict the extensive conservation of some of the most important and well characterized amino acids in RAG1/RAG1L proteins (numbers given for BbeRAG1L): from left to right, catalytic carboxylates (D701, E764, D818, and E1063), residues implicated in controlling coupled versus uncoupled cleavage (A1064 and V751), a residue that facilitates transposition (M949), a histidine component of the active site (H894), and zinc-coordinating residues (C830, C833, H1035 and H1040). Residue E649 in mouse RAG1 contributes to synchronous, or “coupled”, cleavage by RAG at two RSSs, in part through its ability to form a hydrogen bond with S963 [22, 31]. The BbeRAG1L/2 L complex (BbeRAGL) exhibits less propensity for coupled cleavage in part because E649 has been replaced with V751 [22]. Valine is highly conserved at this position in RAG1L proteins from protostomes, suggesting that DNA cleavage by these proteins, if it occurs, would more likely resemble the uncoupled cleavage activity of BbeRAGL. Mutation of the charged residue R848 in mouse RAG1 to the hydrophobic residue methionine, as is found in BbeRAG1L, strongly activates the transposition activity of RAG [22]. Virtually all invertebrate RAG1L proteins, including those from protostomes, have a hydrophobic amino acid, most often methionine, at this position. (b) RAG2/RAG2L PHD domain alignment. The pattern of conserved cysteine and histidine residues (marked with X) are different between invertebrate RAG2L (top) and jawed vertebrate RAG2 (bottom). Variability logo (top) shows relative entropy (bits) calculated on the invertebrate alignment group. Amino acid color code as in Fig. 5.

Additional file 5: Figure S5. (a) Conservation of functional relevant amino acids in RAG1 and RAG1L. The alignments depict the extensive conservation of some of the most important and well characterized amino acids in RAG1/RAG1L proteins (numbers given for BbeRAG1L): from left to right, catalytic carboxylates (D701, E764, D818, and E1063), residues implicated in controlling coupled versus uncoupled cleavage (A1064 and V751), a residue that facilitates transposition (M949), a histidine component of the active site (H894), and zinc-coordinating residues (C830, C833, H1035 and H1040). Residue E649 in mouse RAG1 contributes to synchronous, or “coupled”, cleavage by RAG at two RSSs, in part through its ability to form a hydrogen bond with S963 [22, 31]. The BbeRAG1L/2 L complex (BbeRAGL) exhibits less propensity for coupled cleavage in part because E649 has been replaced with V751 [22]. Valine is highly conserved at this position in RAG1L proteins from protostomes, suggesting that DNA cleavage by these proteins, if it occurs, would more likely resemble the uncoupled cleavage activity of BbeRAGL. Mutation of the charged residue R848 in mouse RAG1 to the hydrophobic residue methionine, as is found in BbeRAG1L, strongly activates the transposition activity of RAG [22]. Virtually all invertebrate RAG1L proteins, including those from protostomes, have a hydrophobic amino acid, most often methionine, at this position. (b) RAG2/RAG2L PHD domain alignment. The pattern of conserved cysteine and histidine residues (marked with X) are different between invertebrate RAG2L (top) and jawed vertebrate RAG2 (bottom). Variability logo (top) shows relative entropy (bits) calculated on the invertebrate alignment group. Amino acid color code as in Fig. 5.

Additional file 1: Figure S1. Genomic organization of RAGL and RAGL transposons identified in this study. (a) Genomic organization of RAGL copies identified in deuterostomes, mollusks, nemerteans and cnidarians. Only the most relevant RAG1L/RAG2L pairs are shown. The legend for panels (a) and (b) is provided at the bottom of panel (b). Loci that are likely to be pseudogenized are indicated with a white box. Supporting transcriptomic data are indicated along with corresponding TSA entry. Green and gray arrows indicate transcripts corresponding to coding and untranslated regions, respectively. Unmapped regions of transcripts are shown as unfilled rectangles outside of the gene track. (b) Genomic organization of incomplete and potentially pseudogenized RAG1L/RAG2L loci in cnidarians. Most of these regions either have stop codons or low sequence coverage and are therefore shown with vertical stripes. The P. damisconis locus is incomplete as it is located at the margin of the scaffold and might encode a complete protein. Assembly gaps near the detected loci are shown as black boxes.

Additional file 7. Alignment S1 Multiple sequence alignment of (a) RAG1/RAG1L and (b) RAG2/RAG2L predicted proteins. Domains, sequence motifs, secondary structure assignment, protein-protein and protein-DNA contact interactions (within 5 Å) displayed above the alignment derive from the BbeRAGL cryo-EM structure (PDB: 6B40). Additionally, for RAG1/RAG1L (a), acidic catalytic residues, red; active site residue mouse H795, purple; zinc coordinating residues within ZDD (*) and ZnC2 and ZnH2 (#) are indicated above the sequences, while for RAG2/RAG2L (b) the beta sheet regions of each kelch-type blade and the GG motifs are shown above the alignment. Locations at which coding sequences span exon boundaries are underlined. Sequence descriptions including references to genomic, transcriptomic or protein databases are shown at the end of the alignment, along with a legend of the symbols used.