"General nature of the genetic code"

Date:: 1964

Reference:: PP/CRI/I/1/23

Catalogue details

Licence: In copyright

Credit: "General nature of the genetic code". Source: Wellcome Collection.

5/16

46 GENERAL NATURE OF THE GENETIC CODE one codon. This now seems very unlikely, since almost all mutants which have been studied show a change of only a single amino acid. If the code were an overlapping one we should certainly expect to find mutants which were changed in two or three (adjacent) places in the polypeptide chain. The evidence comes from the work of Dr. Yanofsky already mentioned and from studies on abnormal human haemoglobins and the protein of To bacco Mosaic Virus (Wittmann and Wittmann-Liebold, 1963; Tsugita, 1962.) It is not yet certain that the bases making up a codon are adjacent on the polypeptide chain, although we have no reason to think they are not (Dr. Stent, in discussion, pointed out that the recent biochemical work of Leder and Nirenberg (1964) makes it very likely that codons consist of three adjacent bases). If the code is of the non-overlapping type there must be something which decides how the message is split up correctly into groups of three bases. Otherwise the message might be read out of phase and thus be totally incorrect. There are, of course, two ways of reading a triplet code out of phase, one corresponding to beginning one base too early and the other to beginning one too late. From evidence (Crick et al. 1961) obtained by study ing two adjacent genes joined together by a genetic deletion we believe this is done in the most obvious way. That is, by beginning at some fixed point in the sequence and reading along from there in steps of three bases at a time. This genetic work also suggested very strongly that codons consist of three bases, rather than four or five or some other number. The evidence was obtained using acridine mutants of the rll locus of phage T4. These mutants behave as though they were due to the addition or subtraction of a base (or bases) in the genetic message, rather than to one base being changed into another, though this interpretation will be more certain when it becomes possible to study the amino acid changes produced by such mutations. It was found that when three mutants of like type (each the addition of a single base, say) were combined by genetic methods into one gene, the gene still functioned, whereas the same mutants, either singly or in pairs, made the gene completely inactive. Such results are most easily understood if every codon consists of three bases. The recent biochemical work of Leder and Nirenberg (1964) appears to confirm this. Our recent more detailed knowledge of the genetic code has come from experiments using cell-free systems which synthesize proteins in the test-tube. The advantages of such systems are not only that the separate steps of protein synthesis can be more easily studied in this way, but also that one can add molecules to them which cannot penetrate easily, if at all, into intact cells. The system most used is obtained from E. coli, though other sources, includ F. H. C. CRICK 47 ing mammalian ones are also employed. In outline it consists of ribosomes, S-RNA molecules, activating enzymes and various soluble factors, which are probably protein, in addition to amino acids, ATP (as an energy source) and GTP. The recent review by Watson (1964) should be consulted for details. In the cell the genetic information in DNA is copied onto single- stranded RNA by the enzyme RNA polymerase, and the messenger RNA (m-RNA) thus produced is then read by the ribosomes. In the cell-free system various artificial m-RNA molecules can be added and their effect on amino acid incorporation studied. The break-through came when Nirenberg and Matthaei (1961) showed that polyuridylic acid, used as m-RNA, pro moted the incorporation of phenylalanine into polypeptide chains. Thus one codon for phenylalanine is presumably UUU. Subsequent work mainly by Nirenberg and his colleagues and by Ochoa and his group have shown that polyadenylic acid codes lysine, and that polycytidylic acid probably codes proline. In addition by using mixed polynucleotides, with random or semi- random sequences, it has been possible to allocate a fair number of the 64 possible triplets to the various amino acids, though triplets high in guanine have proved difficult to study, since polynucleotides tend to have too much secondary structure if their guanine content is high. Of course by these methods alone it has not been possible to obtain the order of the three bases in a particular codon, but only the composition of the codon. Two reviews (Nirenberg et al. 1963; Speyer et al. 1963) provide a good summary of the results obtained by these methods. Among other things the evidence strongly suggested that there is, in general, more than one codon for each amino acid; that is, that the code is degenerate. It is suspected that there is one type of S-RNA for each codon, but this has not yet been proved rigorously. It is not yet known whether the genetic code is identical in all living things, but preliminary studies (see the article by Weinstein, 1963) show that it is certainly rather similar in a number of rather widely separate organisms. In order to obtain the order of bases within any codon it is necessary to use polynucleotides having their bases in a known order. Khorana's group is attempting to do this by synthesizing polynucleotides with repeating sequences. Doty's group is also studying polynucleotides with known or partly known sequences. Meanwhile Nirenberg has hit upon an ingenious method for obtaining codons using trinucleotides. Only preliminary results have so far been published (Leder and Nirenberg, 1964) but it is known that Nirenberg and his colleagues have already determined many codons in this way and the results will probably be in print before this article appears.