A genetic code with only 32 codons in LUCA (569.1)

We have found that the over 600 actinobacteria whose genomes have been reported lack ribosomal protein S21. Actinobacteria include several species in which the GC content of their DNA is greater than 74%. In several of these species the GC content in the wobble base position is greater than 97%. In these species codons ending in A or T are confined to hypothetical proteins having no homologues in the gene bank. Nine of the codons ending in A or T do not appear in any ribosomal proteins or tRNA sythetases of these species. No tRNAs cognate to these unused codons are found in these species. The reliably identified proteins have a marked amino acid bias favoring the four residues encoded by the eight GC‐only codons (Gly, Ala, Arg, and Pro). The vast majority of the putative proteins in these six species have full length multiple open reading frames making hypothetical protein gene identification particularly ambiguous. The absence of S21, the sever codon bias, and the absence of tRNAs corresponding the unused codons, suggest that these species are closely related to the last surviving universal common ancestor of all species. The data also suggest that a genetic code of 32 codons ending in G and C may have prefigured the modern code.