Reprints 51-55

Date:: 1963-1964

Reference:: PP/CRI/M/1/2

Credit: Reprints 51-55. Source: Wellcome Collection.

8/192

$affects a single amino acid. For example, this is so far true of all the abnormal hemoglobins studied in detail. However, the most useful results have been obtained by treating the RNA of Tobacco Mosaic Virus (TM\ ) with nitrous acid, and then determining the alteration in se quence (if any) produced in the TMV protein (see Section IV,A for references). Again the typical change affects a single amino acid. No case has yet been reported of a change of two adjacent amino acids. This evidence makes overlapping codes rather unlikely, though it is possible, with ingenuity, to imagine a partly overlapping code that will fit the facts. Such a code has in fact been suggested by Wall (1962) and is discussed in Section VII,C,5. If the code is nonoverlapping there must be some way of deciding how the bases are grouped into codons, since the RNA structure itself has no obvious commas that could do this. Of course a particular base might function as a comma, though this does not seem very likely. At one time it was thought that perhaps the code might be of the comma- less type, and could be read in one way only. It can be shown that with four bases this can be done without imposing restrictions on the amino acid sequence if the number of amino acids does not exceed twenty (Crick et al., 1957). At the moment a comma-less code seems very improbable for a number of reasons. For example, with such a code one cannot easily explain the large variation in base composition of the DNA of different microorganisms. Polyuridylic acid should not act as a messenger, and other results of the cell-free system cannot be explained (see Section VII,C,1). B. Is the Message Read from One End? The best evidence that this is so comes from recent genetic studies on the A and B cistrons of the rll locus of bacteriophage T4 (Crick et al., 1961). There is no direct evidence that either of these two genes codes for a polypeptide chain, but there is indirect evidence [the ambi valent mutants of Benzer and Champe (1961)] that suggests they do. The argument turns on the nature of the mutations produced by acridines such as proflavine. Acridines act in this way only on the phage- infected cell, not on free phage, and the action does not need light. It is suspected that the mistake occurs during either replication or recom bination, from the acridine slipping in between adjacent base-pairs and forcing them apart (Lerman, 1961). It is postulated that the mutants produced come not from the alteration of a base (as in the case of nitrous acid) but from the deletion or addition of one or more of the bases. Mutants of this type, which includes many of the spontaneous mutants$

Content advisory

Reprints 51-55

Licence and re-use