The Genetic Code for Proteins

Proteins consist of polypeptide chains which are formed by
joining together amino acids head to tail. A typical chain may be
some hundreds of residues long. Only 20 different amino acids are
found in proteins no matter what the biological origin of the protein.

There are two sorts of nucleic acid, DNA and RNA. Each has
a regular backbone consisting of an alternating sequence of phos-
phate and sugar groups. The sugar is ribose in RNA and the closely
related deoxyribose in DNA, To each sugar is attached a flat
heterocyclic molecule called a base. There are four sorts of bases
in both DNA and RNA. In DNA they are adenine, guanine, cytosine
and thymine. In RNA thymine (which is 5-mthyluracil) is replaced
by uracil.

The genes cosist of part of a nucleic acid molecule. It is
typically some hundreds of bases long. There is evidence that the
amino acid sequence of any protein is determined by its own
particular gene. Thus in some way a sequence of four different
things (the bases in nucleic acid) determines the sequence of 20
different things (the amino acids in protein). ‘The problem of
how this is done is known as the coding problem.

There is in all cells an intricate biochemical mechanism for

protein synthesis, parts of which are fairly well understood. In
- 2

brief, the synthesis takes place on particles about 200 A in
diameter called ribosomes. The genetic message is conveyed from
the DNA of the gene to the ribosome by means of a special copy
of the gene called messenger RNA. It is possible to break open cells
and isolate the components necessary for protein synthesis and to
combine them in such a way that a limited synthesis of protein
takes place in the testtube. It is also possible to add arti-
ficial messenger RNA and to observe which amino acids are
incorporated into protein as a result.

It has been shown by genetic experiments that in all
probability three bases code one amino acid and that the message
4s read into the correct groups of three by starting at some fixed
point, probably an end, and reading along three at a time. The
work was done using the B gene of the ryylocus of bacteriophage 1T4
which attacks E. Coli. The mutants studied were the type produced
by acridines which are believed to have one or more bases added or
subtracted from the normal gene. Such mutants are quite distinct
from those produced by chemicals which change one base into
another. It was found that the acridine mutants could all be given
either the sign + or the sign -, such that when joined into pairs
by genetic recombination the pairs (++) or (~~) were inactive,
whereas many of the pairs (=~) gave an active or partly active gene.
The striking result was that combinations of the type (+++) or

(---) were also active. This suggested that three bases (or, less
~3-

likely, a multiple of three bases) stood for one amino acid.

No$ all of the combinations (#-) were active, especially
if they were rather far apart on the genetic map. A plausible
explanation of this phenomenon made it likely that only a small
number of the 64 possible triplets did stand for an amino acid,
and that therefore the code was degenerate; that is, each amino
acid would be represented by more than one triplet.

Recently biochemical work, espeétally by Nirenberg and
Ochoa and their respective colleagues, has uncovered some of the
details of the genetic code. Nirenberg and Matthaei found that
if poly U (the RNA in which every base is uracil) is used as a
messenger in the cell-free system it produced polyphenylalanine;
that is a "protein" in which every amino acid residue is phenyl-
alanine. This suggests that the triplet UUU stands for phenyl-
alanine. More recent work using random co-polymers of two or more
bases has suggested triplets for many of the other amino acids.
There is good evidence that at least two triplets stand for the
amino acid leucine and it is probable that each amino acid is
represented by several triplets. However, the triplets coding one
amino acid all appear to be rather similar. It is not yet known
whether the 64 triplets can be put into 20 groups in some logical
fashion or whether the groupings found are mainly due to historical

accident.