Evolutionary scenario for the origin of Paramecium IESs. (a) Putative timing for transposon invasion in the ciliate phylum (ciliate tree adapted from [28]). Identification of closely related domesticated piggyBac transposases in Tetrahymena and Paramecium (grey boxes) led to the hypothesis that a piggyBac transposon invaded the germline genome of one of their common ancestors (grey arrow), prior to the divergence between these two ciliate lineages. Because of the absence of TA-IESs in Tetrahymena, only the germline genome of Paramecium is thought to have undergone subsequent invasion by Tc1/mariner transposons (red arrowhead). TA-IESs (red box flanked by two black squares) and related transposons were found in the more distant ciliate Euplotes, but the protein(s) required for their developmentally programmed excision have not been identified. (b) In the revisited version of the IBAF model in Paramecium, the ancestor of the PGM gene (in grey) was already present when the first Tc1/mariner transposon (yellow box) started to invade the MIC. During the blooming step, Pgm may have been recruited to rid the genome from deleterious transposon insertions within genes. Thanks to the preexistence of the Pgm domesticated transposase and to the NHEJ repair pathway, Tc1-related transposons could be excised precisely from the somatic genome of the next sexual generation (programmed elimination from the MAC is represented by black dotted arrows), between the two duplicated copies of their TA target site (black squares). This has allowed invading Tc1/mariners to spread throughout the germline genome as a consequence of transposition catalyzed by their own transposase (mobility inside the MIC is symbolized by red arrows). During evolution, most copies of Tc1/mariner transposons have lost their coding capacity and have shortened in size, while being kept under selection pressure for their Pgm-dependent precise excision from the MAC, to give the currently known IESs (red boxes).