Point Counter Point

Schultz and Yarus (1996) propose that codon reassignment takes place through ambiguously translating intermediates, meaning ‘‘the supposition that codons are read ambiguously by two tRNAs (a tRNA and an RF [release factor] in the case of terminations), specifying insertion of more than one amino acid (or an amino acid as well as stop).’’ In simple terms, they propose that a codon can have two meanings and that it switches from the first to the second meaning to bring about a change in the code. This is in contrast to the scheme for codon capture that we have proposed, in which a codon disappears from coding sequences and later reappears in a new role (Osawa and Jukes 1989; Osawa et al. 1992). The article by Schultz and Yarus (1996) is largely a recapitulation of their previous publication (Schultz and Yarus 1994), which we have discussed and rebutted (Osawa and Jukes 1995), and we apologize for reiteration of some of our 1995 discussion. Their proposal is difficult to reconcile with fidelity of the genetic code, which is essential to maintaining the constancy of composition of proteins, without which the continued viability of living organisms in natural populations would be lessened, even though such organisms may survive for several generations in laboratories. Schultz and Yarus also say that ‘‘total disappearance (of a codon) should be an extremely slow occurrence because mutation pressure and genetic drift in large populations are among the weaker evolutionary forces, producing only very slow changes in genomic composition. Furthermore, back mutation increases in rate as the goal is approached because of the accumulation of codons related to the disappearing codon by single mutation.’’ But these forces, in spite of ‘‘back mutation,’’ are responsible for many evolutionary changes. An increase in back mutation such as CGY to CGG is not supported by experimental evidence. Directional mutation pressure is exerted not only on codons but also on tRNAs, including anticodons. The amount of an anticodon varies in parallel with that of the corresponding codon. For example, the use of codon CGG diminishes under AT pressure together with a decrease in the amount of anticodon CCG, so conversion of CGY to CGG becomes less and less likely to occur and finally CGG disappears. Codon CGG has disappeared from Mycoplasma capricolum; indeed, this was noted by Schultz and Yarus (1994). But the anticodon for CGG has also disappeared (Oba et al. 1991) and translation of a synthetic mRNA containing CGG in frame, in a Mycoplasma cell-free system, stops at this point, so the synthesized peptide is not released from the ribosomes. Evidently, reading of this codon by another tRNA, as might be expected in the scheme of Schultz and Yarus, does not occur. Because of these mutual constraints on the codon and anticodon, the frequency of back mutations would decrease rather than increase. The final result would be disappearance of both codon and anticodon (Yamao et al. 1991; Osawa 1995, pp. 58–70). The effect of anticodon on codon usage caused by mutation pressure is sometimes very strong (Osawa 1995, pp. 53–56). Codons UUA and AGA were absent from about 3,500 codons examined in Micrococcus luteus, showing the effects of GC pressure. Also, on the spc operon of M. luteus, 12 codons (UUA, CUU, CUA, AUU, AUA, GUU, GUA, CCA, ACA, CAA, AAA, and AGA) were absent and their function was replaced by higher G + C counterparts (UUG, CUC, CUG, AUC, GUC, GUG, CCG, ACG, CAG, AAG, and AGG), which is a striking example of the effects of GC pressure. See also Kano et al. (1991). The reverse effect is shown by use of high A + U anticodons in Mycoplasma capricolum (Table 3 of Osawa et al. 1992); Ohama et al. (1990a). Production of an unassigned codon is not a rare event. Examples are in Table 5.3 of Osawa (1995). It may be instructive to examine the transition from using UGG as the sole tryptophan codon to the use of both UGG and UGA for this purpose in Mycoplasma capricolum. In this species, the tRNA genes for UCA J Mol Evol (1997) 45:1–8