RAG1 Core and V(D)J Recombination Signal Sequences Were Derived from Transib Transposons

The approximately 600-amino acid core region of RAG1 required for its catalytic activity is significantly similar to the transposase encoded by DNA transposons of the Transib superfamily.

“Here we show that the approximately 600-amino acid “core” region of RAG1 required for its catalytic activity is significantly similar to the transposase encoded by DNA transposons that belong to the Transib superfamily.”

Divergent proteins encoded by the sea urchin and lancelet genomes are 25%-30% identical to the RAG1 N-terminal domain and the RAG1 core.

“We also studied several divergent proteins encoded by the sea urchin and lancelet genomes that are 25%−30% identical to the RAG1 N-terminal domain and the RAG1 core.”

The critical DDE catalytic triad of RAG1 is shared with the Transib transposase as part of conserved motifs.

“Furthermore, the critical DDE catalytic triad of RAG1 is shared with the Transib transposase as part of conserved motifs.”

No significant similarities between the nine other known superfamilies of DNA transposons and RAG1 protein were found.

“However, based on PSI-BLAST searches, no significant similarities between these nine TPases and RAG1 protein were found (data not shown).”

The approximately 60-aa C-terminal portion of the Transib2_AG TPase is 35%-38% identical to the C-terminal portion of the RAG1 core.

“we found that the approximately 60-aa C-terminal portion of the Transib2_AG TPase was 35%−38% identical to the C-terminal portion of the RAG1 core (Figure S1).”

With 15 Transib TPases, Ei-values of matches between RAG1s and the new PSSM were approximately 10−6-10−3.

“Again, Ei-values of matches between RAG1s and the new PSSM derived from multiple alignment of 15 Transib TPases (the two GenBank proteins plus all our TPases) were much smaller (approximately 10−6–10−3; Table 1) than those obtained based on the PSSM constructed from the nine TPases at the preceding step (approximately 10−3–10−2).”

Recently sequenced genomes of honeybee, roundworms, fish, frog, mammals, sea squirts, plants, and fungi (except soybean rust) do not contain any detectable Transib transposons.

“At the same time, recently sequenced genomes of honeybee, roundworms, fish, frog, mammals, sea squirts, plants, and fungi (except soybean rust) do not contain any detectable Transib transposons (see Figure 1).”

Contig 6797 harbours two additional defective pseudogenes coding for different RAG1 core-like proteins.

“Interestingly, the 6,690-bp contig 6797 harbours two additional defective pseudogenes coding for different RAG1 core–like proteins (Figure 5).”

Nearly the entire sequences of several Transib TPases, excluding their 100-140-aa N-terminal domains, converged with an approximately 600-aa portion of RAG1 defined by positions approximately 360-1010.

“Nearly the entire sequences of several Transib TPases, excluding their 100–140-aa N-terminal domains, converged with an approximately 600-aa portion of RAG1 defined by positions approximately 360–1010 (Figure S3).”

RAG1 proteins appeared more similar to the PSSM created by PSI-BLAST based on multiple alignment of the Transib2_AG TPase and two Transib TPase-like proteins than to the solo Transib2_AG TPase in BLASTP.

“The RAG1 proteins appeared to be more similar to the position-specific scoring matrix (PSSM) created by PSI-BLAST based on multiple alignment of the Transib2_AG TPase and two Transib TPase-like proteins, than to the solo Transib2_AG TPase in the BLASTP search.”

The statistical significance of similarity between the Transib2_AG TPase and RAG1s was Ei = 0.025.

“The statistical significance of similarity between the TPase and RAG1s was measured by Ei = 0.025, where Ei is the E-value threshold for the first inclusion of RAG1 sequences into the PSI-BLAST iterations [18] (Materials and Methods).”

• cites [18] Altschul, 1997, Nucleic Acids Res sourceNot checked yet

  Not checked in the free run.

The C-terminal domain of RAG1 (residues 764-1011) is the portion of RAG1 most conserved between RAG1 and Transib TPases.

“The C-terminal domain of RAG1 (residues 764–1011) is the portion of RAG1 that is most conserved between RAG1 and Transib TPases.”

Some Transib families show high target site specificity, e.g., Transib-N1_AG and Transib-N2_AG integrate preferentially at cCASTGg and cCAWTGc, respectively.

“Moreover, some families show high target site specificity, e.g., Transib-N1_AG and Transib-N2_AG integrate preferentially at cCASTGg and cCAWTGc, respectively (TSDs are capitalized).”

The distances between the second D and E residues are much longer in Transib transposons (206-214 aa) and RAG1 (253 aa) than in DDE TPases from other studied superfamilies.

“This includes the distances between the second D and E residues, which are much longer in Transib transposons (206–214 aa) and RAG1 (253 aa) than in DDE TPases from other studied superfamilies (e.g., approximately 35-aa in Mariner/Tc1 [23], 2-aa in P [23], approximately 35-aa in Harbinger [24], with hAT as an exception (325-aa, [25]).”

• cites [23]Not checked yet

  Not checked in the free run.

• cites [24] Kapitonov, 2004, DNA Cell Biol sourceNot checked yet

  Not checked in the free run.

• cites [25] Zhou, 2004, Nature sourceNot checked yet

  Not checked in the free run.

Diverse Transib TPases are present in silkworm, red flour beetle, dog hookworm, freshwater flatworm, soybean rust, and hydra.

“we also found diverse Transib TPases in silkworm, red flour beetle, dog hookworm, freshwater flatworm, soybean rust, and hydra (Table S1).”

Genomes of plants and vertebrates seem to be free of any recognizable Transib transposons.

“Genomes of plants and vertebrates seem to be free of any recognizable Transib transposons (Figure 1).”

An intermediate RAG1-like sequence in Hydra magnipapillata, RAG1L_HM, is significantly similar to both RAG1 and Transib TPase.

“During the final revision of this manuscript, we identified an intermediate RAG1-like sequence in Hydra magnipapillata, called RAG1L_HM, which is significantly similar to both RAG1 and Transib TPase, as shown later.”

The corresponding positions are perfectly conserved in all Transib transposons, excluding the 85% conserved position 34.

“The corresponding positions are perfectly conserved in all Transib transposons (Figure 4A and 4B; excluding the 85% conserved position 34 in the Transib consensus that corresponds to position 5 in the RSS nonamer).”

Each catalytic residue of the DDE triad is part of a motif conserved in Transib TPases and RAG1 (motifs 4, 6, and 10).

“Moreover, each catalytic residue is a part of a motif that is conserved in the Transib TPases and RAG1 (motifs 4, 6, and 10 in Figures 3 and Figure S3).”

In PSI-BLAST, two unclassified proteins and RAG1s constituted the only group of GenBank proteins similar to the Transib2_AG TPase.

“we found that two unclassified proteins (GenBank gi 30923617 and 30923765; annotated as hypothetical proteins) and RAG1s constituted the only group of any GenBank proteins similar to the Transib2_AG TPase (Table 1).”

The 38-bp consensus TIR of Transib transposons consists of a conserved 5′-CACAATG heptamer separated by a variable 23-bp spacer from an AAAAAAATC-3′ nonamer.

“The 38-bp consensus TIR of Transib transposons consists of a conserved 5′- CACAATG heptamer separated by a variable 23-bp spacer from an AAAAAAATC-3′ nonamer.”

Several other motifs conserved in Transib TPases and RAG1 include amino acid residues shown experimentally to be important for specific functions in V(D)J recombination.

“Several other motifs conserved in Transib TPases and RAG1 include aa residues that have been shown experimentally to be important for specific functions in V(D)J recombination (Figure S3).”

All three core residues from the catalytic DDE triad in RAG1 proteins (residues 603, 711, and 965) necessary for V(D)J recombination are conserved in Transib TPases.

“All three core residues from the catalytic DDE triad in the RAG1 proteins (residues 603, 711, and 965) that are necessary for V(D)J recombination [21,22] are conserved in the Transib TPases (Figures 3 and Figure S3).”

• cites [21] Kim, 1999, Genes Dev sourceNot checked yet

  Not checked in the free run.

• cites [22] Landree, 1999, Genes Dev sourceNot checked yet

  Not checked in the free run.

The probability of the observed match between the RSS and Transib termini to occur by chance is less than 10−3.

“The probability of the observed match between the RSS and Transib termini to occur by chance is less than 10−3 (see Materials and Methods).”

Using RAG1 proteins as queries against sea urchin contigs, eight proteins were identified that are approximately 30% identical to portions of the RAG1 core and approximately 50% identical to each other.

“we identified eight proteins approximately 30% identical to portions of the RAG1 core and approximately 50% identical to each other (see Figures 2, 5, and S5).”

The similarity between Transib2_AG TPase and the RAG1 core in BLASTP was only marginally significant (E = 0.07).

“However, this similarity was only marginally significant (E = 0.07 where the E-value is an expected number of sequences matching by chance; Table 1).”

Combining seven known Transib TPases with GenBank noticeably dropped Ei-values for matches of RAG1s to a new PSSM based on alignment of nine Transib TPases.

“As a result, Ei-values for matches of RAG1s to a new PSSM based on alignment of nine Transib TPases (the two GenBank TPase-like proteins plus seven added TPases) noticeably dropped in comparison with the Ei-values obtained for the PSSM constructed in the previous step based on alignment of the three TPases (Table 1).”

With all 18 TPases, Ei values of matches between RAG1s and the Transib PSSM dropped to 10−9-10−4.

“When all 18 TPases were combined with the original GenBank set, the Ei values of matches between RAG1s and the Transib PSSM dropped significantly further (10−9–10−4; Table 1).”

A 597-aa protein sequence encoded by a single open reading frame (contig 29068) is 28% identical to nearly the entire RAG1 core (positions 461-1002 in human RAG1).

“We also identified a 597-aa protein sequence encoded by a single open reading frame (contig 29068, positions 1157–2944), which is 28% identical to nearly the entire RAG1 core (positions 461–1002 in the human RAG1, Figure S5).”

The starlet sea anemone and hydra genomes encode several families of the N-terminal RAG1 domain that appear to be separate from the RAG1 core-like proteins.

“Based on the same approach, our study found that the starlet sea anemone and hydra genomes also encode several families of the N-terminal RAG1 domain that appear to be separate from the RAG1 core–like proteins (data not shown).”

The hydra RAG1L_NM protein is significantly similar to the Transib TPase (26% identity; E-value approximately 10−14).

“Remarkably, the hydra RAG1L_NM protein turned out to be significantly similar to the Transib TPase (26% identity; E-value is approximately equal to 10−14 in a BLASTX search against GenBank proteins combined with the Transib TPase sequences).”

A third sea urchin protein on contig 768 contains unique motifs in its N-terminal regions that best match homologous regions of RAG1.

“We also found a third protein on contig 768 that contains unique motifs in its N-terminal regions that best match the homologous regions of RAG1.”

There is no significant sequence identity between hAT TPases and RAG1, even if one employs a PSI-BLAST search with most relaxed parameters.

“First, as we mentioned previously, there is no significant sequence identity between hAT TPases and RAG1, even if one employs a PSI-BLAST search with most relaxed parameters (i.e., E < 10, no filters, no composition-based statistics).”

None of the sea urchin and lancelet proteins align to the approximately 100-aa N-terminus of RAG1.

“None of the sea urchin and lancelet proteins align to the approximately 100-aa N-terminus of RAG1, which may indicate that this portion is missing from the genome or highly diverged and difficult to detect.”

The new conserved motifs are H-x3-L-x3-C-R-x-C-G and D-x3-I-h-P-x2-F-C-x2-C.

“The new conserved motifs are H-x3-L-x3-C-R-x-C-G and D-x3-I-h-P-x2-F-C-x2-C, and their function remains to be determined.”

BLASTP searches of the sea urchin sequences against all GenBank proteins, excluding RAG1, detected only the ring finger domain of the sea urchin sequences.

“BLASTP searches of the sea urchin sequences against all GenBank proteins, excluding RAG1, detected only the ring finger domain of the sea urchin sequences.”

Even without the ring finger motif, the remaining portions of the sea urchin sequences are significantly similar to the corresponding portions of RAG1.

“Even without the finger, the remaining portions of the sea urchin sequences were significantly similar to the corresponding portions of RAG1.”

The only exception was the sea anemone RAG1 core-like sequence, whose approximately 90-aa N-terminus is the ring finger.

“The only exception was the already mentioned sea anemone RAG1 core–like sequence. The approximately 90-aa N-terminus of the latter sequence is the ring finger (E < 10−7, multiple BLASTP matches against known ring fingers in GenBank).”

Unlike highly conserved RAG1 proteins, the RAG1 core-like proteins are remarkably diverse.

“However, unlike highly conserved RAG1 proteins, the RAG1 core–like proteins are remarkably diverse (see Figure 2).”

N-terminal RAG1 and RAG1-like proteins share an additional conserved motif W-x-p-h-x(3-6)-C-x2-C that resides between conserved motif 2 and the ring finger.

“It also likely that the N-terminal RAG1 and RAG1-like proteins share an additional conserved motif W-x-p-h-x(3–6)-C-x2-C that resides between conserved motif 2 and the ring finger (Figure 6).”

The RAG1 core-like protein in WGS contig 12509 seems to be encoded by the last exon starting at position 1650 of a hypothetical RAG1-like gene.

“The RAG1 core–like protein in whole genome shotgun (WGS) contig 12509 (Figure 5) also seems to be encoded by the last exon starting at position 1650 of a hypothetical RAG1-like gene.”

We could not find any RAG2-like sequences in the recently sequenced sea urchin, lancelet, hydra, and sea anemone genomes, which encode RAG1-like sequences.

“Also, we could not find any RAG2-like sequences in the recently sequenced sea urchin, lancelet, hydra, and sea anemone genomes, which encode RAG1-like sequences.”

The lancelet RAG1 core-like elements show no hallmarks of transposons.

“Again, as in the case of the sea urchin sequences, the lancelet RAG1 core–like elements show no hallmarks of transposons (data not shown).”

The acceptor splice site positions in the sea urchin RAG1 core-like proteins closely correspond to those in RAG1 from teleosts, in which RAG1 is split by an intron at position homologous to Gly460 in human RAG1.

“Remarkably, the acceptor splice site positions in the sea urchin RAG1 core–like proteins closely correspond to those in RAG1 from teleosts (i.e., most of the living ray-finned or bony fish), in which RAG1 is split by an intron at position homologous to Gly460 in human RAG1 [34].”

• cites [34] Willett, 1997, Immunogenetics sourceNot checked yet

  Not checked in the free run.

The RAG1 core shows 79% identity between sharks and mammals.

“as indicated by the observed conservation of the RAG1 core (79% identity between sharks and mammals).”

Detailed analysis of regions flanking the sea urchin RAG1-like DNA coding sequences revealed a variety of different transposable elements inserted in the proximity of the coding sequences.

“Detailed analysis of regions flanking the sea urchin RAG1-like DNA coding sequences revealed a variety of different transposable elements inserted in the proximity of the coding sequences (see Figure 5).”

Separate analysis of assembled sea urchin sequences yielded seven sequences encoding three diverse proteins significantly similar to the 380-aa N-terminal domain of RAG1 (BLASTX, E < 10−4), excluding the 100-aa N-terminus.

“A separate analysis of the assembled sea urchin sequences yielded seven sequences encoding three diverse proteins that were significantly similar to the 380-aa N-terminal domain of RAG1 (BLASTX, E < 10−4), excluding the 100-aa N-terminus (Figure 6).”

Only two cysteines of the zinc finger B (ZFB) C2H2 motif in RAG1 (residues 695-761) involved in its binding to RAG2 are perfectly conserved in the Transib TPases (motif 7).

“Interestingly, only two cysteines of the zinc finger B (ZFB) C2H2 motif in RAG1 (residues 695–761) involved in its binding to RAG2 [30,31] are perfectly conserved in the Transib TPases (motif 7; see Figures 3 and Figure S3).”

• cites [30] Rodgers, 1996, J Mol Biol sourceNot checked yet

  Not checked in the free run.

• cites [31] Aidinis, 2000, J Immunol sourceNot checked yet

  Not checked in the free run.

Hydra magnipapillata WGS trace sequences code for protein sequences 26%-30% identical to the RAG1 core.

“We also found several approximately 1000-bp WGS trace sequences of Hydra magnipapillata (Trace Archive IDs 688654311, 647073738, 666995387, 687186526, 688683890, and 688948453), coding for protein sequences 26%−30% identical to the RAG1 core (positions 753−995, E-value is approximately equal to 10−7 in a BLASTX search against GenBank).”

The ring finger motif is present in several sea urchin proteins unrelated to RAG1 but significantly similar to diverse proteins associated with immune and developmental systems as well as regulation of transcription.

“The ring finger motif itself is also present in several sea urchin proteins unrelated to RAG1 but significantly similar to diverse proteins associated with immune and developmental systems as well as regulation of transcription.”

We could not identify the 5-bp TSDs and TIRs characteristic of the Transib superfamily in the sea urchin RAG1-like sequences.

“We also could not identify the 5-bp TSDs and TIRs characteristic of the Transib superfamily.”

In starlet sea anemone, RAG1L_NV is approximately 30% identical to the human RAG1 core (positions 284-802).

“In starlet sea anemone, several approximately 1000-bp WGS trace sequences were found (e.g., GenBank Trace Archive IDs 668021618, 558173651, 568641192, and 599572062), which encode protein, called RAG1L_NV, that is approximately 30% identical to the human RAG1 core (positions 284–802, TBLASTN, 10−26 < E < 10−7).”

None of the Transib transposons identified so far encode any proteins other than the Transib/RAG TPase.

“However, none of the Transib transposons identified so far encode any proteins other than the Transib/RAG TPase.”

Based on the orientations and relative positions of these transposons, none of them appears to be associated with the RAG1-like sequences.

“Nevertheless, based on the orientations and relative positions of these transposons, none of them appears to be associated with the RAG1-like sequences (see Figure 5).”

The RAG1 core-like ORF in contig 29068 forms a terminal exon of an incomplete hypothetical gene composed of two exons.

“Using FGENESH [33], we detected that the RAG1 core–like open reading frame (ORF) in the contig 29068 forms a terminal exon (positions 1154–2947) of an incomplete hypothetical gene composed of two exons (internal and terminal; see Figure S6).”

• cites [33] Solovyev, 2002, Current Topics in Computational Biology: MIT PressNot checked yet

  Not checked in the free run.

In the sea urchin, lancelet, hydra, and starlet sea anemone genomes, the RAG1 core-like sequences and the N-terminal domain-like sequences do not appear to be linked to each other or to any other proteins.

“In the sea urchin, lancelet, hydra, and starlet sea anemone genomes, the RAG1 core–like sequences and the N-terminal domain–like sequences do not appear to be linked to each other or to any other proteins.”

The sea urchin, lancelet, hydra, and starlet sea anemone genomes harbor several highly divergent N-terminal-like domains, separate from the RAG1 core-like sequences and known transposable elements.

“Furthermore, given that the sea urchin, lancelet, hydra, and starlet sea anemone genomes harbor several highly divergent N-terminal–like domains, separate from the RAG1 core–like sequences and known transposable elements, it is very likely that the N-terminal–like domains of RAG1 also form a multigene family that can be traced back to a common ancestor of Deuterostomes (see Figure 1).”

The lancelet genome encodes five different proteins similar to the N-terminal domain of RAG1 (BLASTP E values 10−14-10−7).

“By analyzing the lancelet WGS traces, we also found that the lancelet genome encodes five different proteins similar to the N-terminal domain of RAG1 (BLASTP E values in searches against all GenBank proteins were in a range of 10−14–10−7).”

The 3′ terminal portion of the internal exon encodes a protein sequence marginally similar to an approximately 50-aa fragment of the RAG1 core (positions 394-454 in human RAG1).

“The 3′ terminal portion of the internal exon encodes a protein sequence that appears to be marginally similar to an approximately 50-aa fragment of the RAG1 core (positions 394–454 in human RAG1; Figure S5).”

The 102-aa N-terminal subdomain, corresponding to the first exon in genes split by two introns, appears to be missing in the sea urchin, lancelet, hydra, and starlet sea anemone N-terminal-like proteins.

“As indicated above, this subdomain, corresponding to the first exon in the genes split by two introns, appears to be missing in the sea urchin, lancelet, hydra, and starlet sea anemone N-terminal–like proteins.”

TIRs of hAT transposons are different from RSS both in terms of DNA sequence similarities and their conservation patterns.

“Third, unlike in the case of Transib transposons, TIRs of hAT transposons are different from RSS both in terms of DNA sequence similarities and their conservation patterns (Figure S7).”

The probability P1 that the 5′ terminus of a transposon from a particular Transib family would match by chance an RSS at its most conserved positions is 0.001.

“The value of P1 is 0.001, indicating a significant similarity between Transib TIRs and RSS.”

The lancelet genome encodes protein sequences approximately 35% identical to the RAG1 core (RAG1L_BF; BLASTP E-value = 10−34).

“we found that the lancelet genome encodes protein sequences approximately 35% identical to the RAG1 core (Figure S5; RAG1L_BF; BLASTP E-value is equal to 10−34).”

Unassembled WGS trace sequences encode two other sea urchin proteins, P4_SP and P5_SP, similar to the N-terminal RAG1 domain.

“Furthermore, we found that unassembled WGS trace sequences encode two other proteins, P4_SP and P5_SP, similar to the N-terminal RAG1 domain (Figure 6).”

The core-like sequences encoded by the sea urchin WGS sequence contigs 29068 and 12509 correspond to either the second or third RAG1 exon in teleosts.

“The core-like sequences encoded by the sea urchin WGS sequence contigs 29068 and 12509 correspond to either the second or third RAG1 exon in teleosts (depending on the number of introns), which is remarkably consistent with the fusion model.”

The only notable exception is the anemone RAG1 core-like protein sequence, which is capped by the 90-aa ring finger motif.

“The only notable exception is the anemone RAG1 core–like protein sequence, which is capped by the 90-aa ring finger motif.”

Recombination signal sequences (RSSs) were derived from terminal inverted repeats of an ancient Transib transposon.

“We demonstrate that recombination signal sequences (RSSs) were derived from terminal inverted repeats of an ancient Transib transposon.”

The structure and conservation patterns of the 38-bp termini of Transib transposons from 21 different families closely resemble those of RSSs, suggesting that RSSs were derived from termini of ancient Transib transposons.

“The structure and conservation patterns of the 38-bp termini of Transib transposons from 21 different families closely resemble those of RSSs, suggesting that the latter were derived from termini of ancient Transib transposons (Figures 4 and S4).”

Based on the conserved motifs, the function of these motifs in Transib TPases is expected to be similar to that in RAG1.

“Based on this information, the function of these motifs in Transib TPases is expected to be similar to that in RAG1.”

The structure of Transib TIRs corresponds closely to the structure of RSSs, which are composed of conserved heptamers 5′-CACAGTG separated by variable 22-bp spacers from ACAAAAACC-like nonamers.

“This corresponds closely to the structure of RSSs, which are composed of the conserved heptamers 5′- CACAGTG separated by a variable 22-bp spacers from ACAAAAACC-like nonamers [1,5–7].”

• cites [1] Tonegawa, 1983, Nature sourceNot checked yet

  Not checked in the free run.

• cites [5] Akira, 1987, Science sourceNot checked yet

  Not checked in the free run.

• cites [6] Ramsden, 1994, Nucleic Acids Res sourceNot checked yet

  Not checked in the free run.

• cites [7]Not checked yet

  Not checked in the free run.

Transib transposons belong to the only superfamily of eukaryotic DNA transposons with 5-bp TSDs generated upon insertions into the genome.

“Given the predominance of 5-bp TSDs, it is striking that Transib transposons belong to the only superfamily of eukaryotic DNA transposons with 5-bp TSDs generated upon insertions into the genome [17,24].”

• cites [17] Kapitonov, 2003, Proc Natl Acad Sci U S A sourceNot checked yet

  Not checked in the free run.

• cites [24] Kapitonov, 2004, DNA Cell Biol sourceNot checked yet

  Not checked in the free run.

Given that the only significant similarity of the RAG1 core was to the Transib TPase, the RAG1 core was re-confirmed as belonging to the Transib superfamily.

“Therefore, given that the only significant similarity of the RAG1 core was to the Transib TPase, the RAG1 core was re-confirmed as belonging to the Transib superfamily.”

RAG1-mediated transposition in vitro, both intermolecular and intramolecular, is most frequently accompanied by 5-bp TSDs.

“It has been known that RAG1-mediated transposition in vitro, both intermolecular and intramolecular, is most frequently accompanied by 5-bp TSDs [12,13].”

• cites [12] Agrawal, 1998, Nature sourceNot checked yet

  Not checked in the free run.

• cites [13] Hiom, 1998, Cell sourceNot checked yet

  Not checked in the free run.

The C-terminal domain of RAG1 is predicted to function analogously in Transib transposons.

“This domain is predicted to function analogously in Transib transposons.”

The 12/23 rule can be derived directly from the sequence structure of known Transib transposons.

“Therefore, even the 12/23 rule [1] can be derived directly from the sequence structure of known Transib transposons.”

• cites [1] Tonegawa, 1983, Nature sourceNot checked yet

  Not checked in the free run.

The results provide the first direct evidence linking RAG1 and RSSs to a specific superfamily of DNA transposons.

“Our results provide the first direct evidence linking RAG1 and RSSs to a specific superfamily of DNA transposons and indicate that the V(D)J machinery evolved from transposons.”

The V(D)J machinery evolved from transposons.

“Our results provide the first direct evidence linking RAG1 and RSSs to a specific superfamily of DNA transposons and indicate that the V(D)J machinery evolved from transposons.”

RAG1/2-mediated transpositions also show significant target specificity, presumably reflecting the original specificity of the Transib TPase.

“RAG1/2-mediated transpositions also show significant target specificity, presumably reflecting the original specificity of the Transib TPase [12].”

• cites [12] Agrawal, 1998, Nature sourceNot checked yet

  Not checked in the free run.

The observed sequence similarity between the RAG1 and Transib TPase protein can help to identify amino acid residues in the TPase that are crucial for transposition of Transib transposons.

“The observed sequence similarity between the RAG1 and Transib TPase protein can help to identify aa residues in the TPase that are crucial for transposition of Transib transposons.”

The remaining portion of the ZFB motif was probably lost in TPases of insect Transib transposons, which do not encode RAG2-like proteins.

“The remaining portion of the ZFB motif was probably lost in TPases of insect Transib transposons, which do not encode RAG2-like proteins.”

Most likely, Transib transpositions are also characterized by hairpin intermediates formed by the ends of the donor DNA double-strand breaks, as observed during V(D)J recombination and hAT transposition.

“Most likely, the Transib transpositions are also characterized by hairpin intermediates formed by the ends of the donor DNA double-strand breaks, as observed during V(D)J recombination and hAT transposition.”

Taken together with the fact that only the RAG1 core is significantly similar to Transib TPase, the data suggest that the vertebrate RAG1 represents a fusion of once separate proteins.

“Taken together with the fact that only the RAG1 core is significantly similar to Transib TPase, the data suggest that the vertebrate RAG1 represents a fusion of once separate proteins.”

• AssumptionRests on an assumption: The N-terminal domain of RAG1 is functionally distinct from the core domain.

The hydra RAG1 core-like protein provides the first direct link between the RAG1 core and Transib TPase.

“Therefore, the hydra RAG1 core–like protein provides the first direct link between the RAG1 core and Transib TPase.”

After its recruitment, the RAG1 core most likely went through a period of intensive transformations due to diversifying/positive selection, which further decreased its similarity to Transib TPase.

“Furthermore, after its recruitment, the RAG1 core most likely went through a period of intensive transformations due to diversifying/positive selection, which further decreased its similarity to Transib TPase.”

All four arguments strongly favor evolution of the V(D)J machinery from a Transib transposon.

“All four arguments strongly favor evolution of V(D)J machinery from a Transib transposon.”

• AssumptionRests on an assumption: RAG1/2-mediated transpositions are characterized by 5-bp (sometimes 4-bp) TSDs, whereas all known hAT transposons are characterized by 8-bp TSDs.

The RAG1 core is split into separate exons at the amino acid position corresponding to position 460 in the human RAG1 gene.

“As a result, the RAG1 core is split into separate exons at the aa position that corresponds to position 460 in the human RAG1gene [29,34,38].”

• cites [29] Spanopoulou, 1996, Cell sourceNot checked yet

  Not checked in the free run.

• cites [34] Willett, 1997, Immunogenetics sourceNot checked yet

  Not checked in the free run.

• cites [38] Venkatesh, 2001, Proc Natl Acad Sci U S A sourceNot checked yet

  Not checked in the free run.

Significant convergence of RAG1 and Transib TPases was observed to be independent of the particular type of substitution matrix (both BLOSUM62 and PAM70).

“Significant convergence of RAG1 and Transib TPases was observed to be independent of the particular type of substitution matrix (the same result was observed for both BLOSUM62 and PAM70 matrixes).”

The origin of the V(D)J recombination system in jawless vertebrates appears to be a culmination of earlier evolutionary processes rather than an isolated event associated with insertion of a single transposon.

“In any case, the origin of the V(D)J recombination system in jawless vertebrates appears to be a culmination of earlier evolutionary processes rather than an isolated event associated with insertion of a single transposon.”

The significant similarity between Transib TPases and RAG1 core, the common structure of Transib TIRs and RSSs, and the similar size of TSDs characterizing Transib transpositions and RAG1/RAG2-catalyzed transpositions directly support the hypothesis of a transposon-related origin of the V(D)J machinery.

“The significant similarity between the Transib TPases and RAG1 core, the common structure of the Transib TIRs and RSSs, as well as the similar size of TSDs characterizing transpositions of Transib transposons and transpositions catalyzed by RAG1 and RAG2, directly support the 25-year-old hypothesis of a transposon-related origin of the V(D)J machinery.”

The correspondence of sea urchin core-like sequences to teleost RAG1 exons is remarkably consistent with the fusion model.

“The core-like sequences encoded by the sea urchin WGS sequence contigs 29068 and 12509 correspond to either the second or third RAG1 exon in teleosts (depending on the number of introns), which is remarkably consistent with the fusion model.”

The presence of serine in this motif is expected to be crucial to Transib transpositions.

“Therefore, the presence of serine in this motif is expected to be crucial to Transib transpositions.”

The apparent lack of RAG2-like proteins in the sequenced portion of the sea urchin, lancelet, hydra, and sea anemone genomes, as well as in Transib transposons, suggests that RAG2 was introduced in a separate event in jawless vertebrates.

“Nevertheless, the apparent lack of RAG2-like proteins in the sequenced portion of the sea urchin, lancelet, hydra, and sea anemone genomes, as well as in Transib transposons suggests that RAG2 was introduced in a separate event in jawless vertebrates.”

• AssumptionRests on an assumption: The absence of detectable RAG2-like sequences in the sequenced portions of the sea urchin, lancelet, hydra, and sea anemone genomes indicates that RAG2 was not present in these genomes.

In teleosts, the RAG1 gene is divided into exons by either one or two introns.

“This is consistent with the observation that in teleosts, (bony fish) the RAG1 gene is divided into exons by either one or two introns.”

• cites [29] Spanopoulou, 1996, Cell sourceNot checked yet

  Not checked in the free run.

• cites [34] Willett, 1997, Immunogenetics sourceNot checked yet

  Not checked in the free run.

• cites [38] Venkatesh, 2001, Proc Natl Acad Sci U S A sourceNot checked yet

  Not checked in the free run.

It is very likely that the N-terminal-like domains of RAG1 form a multigene family that can be traced back to a common ancestor of Deuterostomes.

“Furthermore, given that the sea urchin, lancelet, hydra, and starlet sea anemone genomes harbor several highly divergent N-terminal–like domains, separate from the RAG1 core–like sequences and known transposable elements, it is very likely that the N-terminal–like domains of RAG1 also form a multigene family that can be traced back to a common ancestor of Deuterostomes (see Figure 1).”