Figure 3.

A hypothesis for RAG evolution; an alternative to the transposon/split receptor gene model. The evolution of RAG genes begins as an ancestral Transib transposon with symmetric TIRs. Insertion of an N-RAG-TP transposon and recombination with the ancestral Transib gave rise to a new transposon with all the core features of an extant RAG1 gene, including identical TIR sequences and the N terminal region. An active RAG2 precursor GE was inserted close to or inside the RAG1L gene that became an ancestral hypothetical RAG1/2 transposon that was mainly further transmitted vertically. In amphioxus, the protoRAG transposon lost its PHD domain but maintained its transposition activity based on 27/31 asymmetric TIR spacers. In sea urchins, the RAG1/2L genes were domesticated and may have gained function, which is currently unknown. According to our guardian of the genome hypothesis (lower left box), RAG1/2 was also domesticated early in the jawed vertebrate lineage to protect vertebrate genomes from insertion and excision of harmful TEs (labeled as 1 in the box). Later, a Transib transposon with 12/23 asymmetric TIRs entered and survived inside a jawed vertebrate antigen receptor exon, splitting it into V and J segments (2 in the box). As an ex-transposon endonuclease, the RAG1/2 machinery was co-opted and trained to excise the asymmetric TIR-flanked Transib transposon threat in the V(D)J receptor loci. Fortuitously, the excision process in the V(D)J loci provided an immunological advantage and therefore the RAG1/2 complex gained its V(D)J recombinase function in the germline (3 in the box). Eventually, under selective pressure, the V(D)J machinery was selected evolutionarily to work primarily in the lymphoid cell lineages (4 in the box).