X. TMV STUDIES IN GENETIC CODING 499 



tion. Further work from other hiboratorics confirmed these observations, 

 and it now appears that with average hit numbers greater than 1 or 2, 

 almost all deamination survivors are mutants, though only a fraction 

 of these give the differential host test (i.e., local lesions in X. sijlvestris 

 or Java tobacco). It has been established beyond doubt that deamination 

 is per se truly mutagenic, and that selection does not play a significant 

 role in this phenomenon (Mundry, 1959). 



Having discussed the reactions which were found to effect mutations 

 in RNA, a consideration of the presumed nature of the mutagenic event, 

 in terms of RNA structure, appears advisable. The reaction most fre- 

 quently employed is that with nitrous acid. In the case of DNA, evidence 

 was adduced that the two reactions involving the 6-amino position 

 (adenine -» hypoxanthine, and cytosine—> uracil) can be mutagenic, but 

 that deamination of the guanine is only lethal (Vielmetter and Schuster, 

 1960). This is interpreted in terms of the current concepts of the replica- 

 tion mechanism based on the Watson-Crick scheme of base pairing. In 

 RNA, and only in RNA, is nitrous acid an astonishingly effective muta- 

 gen, leading to mutation of almost all survivors. This has been attributed 

 to the obligatory mutagenic effect of the deamination of cytosine since 

 this leads to another genetically fully competent symbol in the RNA 

 language but not in the DNA language, i.e., uracil (Fraenkel-Conrat, 

 1961; Tsugita, 1961). In contrast, the deamination of adenine leads to a 

 base equally unnatural to RNA and to DNA. Thus, nitrous acid acting 

 on RNA is the only direct mutagenic reaction known. Its effect contrasts 

 with the various indirect mutagenic reactions, such as the deamination 

 of the purines, or of cytosine in DNA, the methylation of the purines, 

 the action of i¥-bromosuccinimide or hydroxylamine on the pyrimidines, 

 the biological incorporation of base analogs into DNA, and others. All of 

 the indirect mutagens are believed to increase the probability of a mis- 

 reading of the coded information, but only the deamination of cytosine 

 is believed to directly alter that information. 



B. PROTEIN COMPOSITION OF CHEMICALLY PRODUCED MUTANTS 



When mutants became available through chemical modification of the 

 RNA, research programs were initiated, both at our laboratory and at 

 the Max-Planck-Institut fiir Biologie, to determine the composition 

 and the structure of the proteins of these mutant strains. Since the 

 amino acid sequence of the wild type protein was known, and analytical 

 methods for the detection and localization of single amino acid exchanges 

 were available, there was reason for hope that this approach would 

 reveal significant information concerning the relationship between the 

 structure of the template RNA and tlie protein product. 



