54 Self and the Phenomenon of Life
b2726 Self and the Phenomenon of Life: A Biologist Examines Life from Molecules to Humanity “9x6”20 amino acids. The discrepancy between the number of tRNAs and the
number of AARSs has to be solved. Put it another way, how can 20 AARS
attach 20 different amino acids to 61 different kinds of tRNAs and come
up with a correct match (obviously not a one-to-one match)? For exam-
ple, the amino acid serine is coded by six different codons, some of which
bear no similarity at all. The single AARS for serine is able to bring the
amino acid to all six different tRNAs assigned to serine, despite varia-
tions in the anticodon in these tRNAs. The explanation is that only some
of AARSs recognize the anticodons on the tRNAs; the others recognize
structural features in the tRNA molecules other than the anticodons,
so that their matching task is unaffected by changes in the anticodons.
Therefore, for AARS, the recognition of the general molecular struc-
ture of a tRNA takes precedence over its anticodon. This fact is strong
evidence that protein-RNA interaction came before genetic coding in
the origin of life. (The remarkable matching ability of an AARS can be
compared to that of a clever matchmaker who recognizes a boy not only
by his legal name but also by all his five nicknames, so that no identity
error is committed when attempting to pair him with a certain girl.)
The last step in translation is the binding of the amino acid-charged
tRNA to mRNA through codon/anticodon matching. Once multiple
tRNAs are lined up along the messenger RNA on the ribosome, the sim-
ple task of joining the amino acids together is performed by the peptidyl
transferase in the ribosome.
3.17 Speculations on the Primitive Translation System
The modern translation machine is a highly sophisticated system. It
would be reasonable to assume that the primitive counterpart would be
much simpler and might even be drastically different. For example, we
can imagine that the 3-letter genetic code might have only two nucle-
obases to choose from (instead of the current four), generating only
eight possible “words,” permitting eight amino acids to be coded for.
The peptide length could be short and variable, and the coding could be