280 Raphael Falk
nucleotide-triplet being acodon), and Brenner [1957], upon analyzing dimmers of
amino-acids from known peptides, concluded that there were more adjacent amino-
acid combinations than and overlapping code will allow. First attempts to decipher
the genetic code compared changes of the “wild type” amino-acid at a specific site
in the polypeptide of the enzyme tryptophan-synthetase ofE. coliby two different
“mutant” amino-acids, with mutations at different though recombinationally very
close sites in the cistron, presumably belonging to the same codon [Hennig and
Yanofsky, 1962]. However, the code was deciphered by molecular, rather than by
genetic methods. Polyribonucleotide strands of known structure were “fed” to cell
extracts and the synthesized polypeptides were extracted and identified [Lengyel
et al., 1961; Nirenberg and Matthaei, 1961]. Still, the crucial insight that the
genetic code is read in-frame from a given starting point was obtained by genetic
analysis of acridine induced mutations in therIIcistrons of the T4 bacteriophage
[Cricket al., 1961]. It was reasoned that acridine molecules intercalate in the DNA
molecule to cause a deletion or an insertion of one nucleotide in the replicating
DNA molecule. First order intra-cistronic acriflavin-induced repressor mutants
(designated –) of a givenrIImutant (designated +) were found to be by them-
selvesrIImutants. They were followed by second order intra-cistronic repressor
mutants of the repressors (designated, again, +) and so on, to higher order sup-
pressors. All suppressor mutationsper se, i.e., without their respective suppressed
mutation, acted as genuinerII mutants. Suppressor mutants of different orders
were recombined incis. Whereascis-recombinants of a + and — mutation gave
as a rule the suppressed, nearly wild-type phenotype,cis-recombinants of pairs of
+s (or –s) mutants gave the mutant phenotype. However,cis-recombining three
+s (or –s), even of different orders, gave again a nearly wild-type phenotype. This
supported the notion that acroflavine mutants were either deletions or additions
of a singe nucleotide (or a small constant multiple of nucleotides) and that the
genetic code is a triplet (or a small multiple of a triplet) code, in which reading
starts at a given site and continues, comma-less, one triplet (codon) after another,
from the initiation codon to the termination codon (three termination codons have
been identified among the 64 combinations of four nucleotide triplets, 61 coding
for 20 amino acids, one to six codons per amino-acid, the code being redundant).
This ingenious analysis as well as the chemical analysis of amino-acid attributions
to the codons were elegantly verified for sequences of amino acids of peptide frag-
ments of the gene coding for lysozyme ofE. coli, in which an acriflavin mutant
was induced, and then, its inducedcis-suppressor was selected. Comparing the
amino-acid sequences in the wild-type and between-mutants fragments verified the
model of the open reading frame of the code as well as the code assignments of
the chemical analysis [Streisingeret al., 1967]. (For a comprehensive review see
[Kay, 2000].)