The degeneration of the proto-tRNAs' recognition loop might be through the extension of an inserted self-splicing intron. (a) The self-splicing intron might have been inserted as an in cis "aminoacyl-tRNA synthetase" ribozyme for proto-tRNAs. By self-splicing and self-inserting, the intron could mediate the interchange of the self-aminoacylation state and the charged mature state of a proto-tRNA. The drawing style is the same as that in Figure 2d, except that the thin solid line, denoting the intron, is added. The aa-binding site of a proto-tRNA (circled in the top-right subfigure) could transform into a large recognition loop to recognize a template RNA. (b) With the extension of the self-splicing intron, the residues no longer necessary in the template-recognition would be removed accompanying the intron's splicing. The proto-tRNA would then recognize a template RNA by a small anticodon loop (circled in the top-right subfigure). The inserting point of the intron was adopted in reference to experimental evidence [77–79]. Note: the intron was not drawn in the same scale as the proto-tRNA, and was also not drawn according to the real structure of a self-splicing intron, which might be much more complex.