498 A. TSUGITA AM) H. !• RAKXKKL-CONRAT 



to 50%. These evcnls aw accoiiiijaiiicd l)y more or lesjs extensive inaeti- 

 vation (no inactivation with fluorouraeil ), but no mutants ha\c hcen 

 obsen'cd as a consequence of these reactions. 



The chemical modification reactions are performed (Urcctly with the 

 virus, or jircferably with the isohited RNA. Formaldeiiyde adds to or 

 condenses with the amino groups of the RNA (Fraenkel-Conrat, 19o5bj, 

 and is a relatively inefficient inactivating agent, probably because of the 

 reversibility of the reaction (Staehelin, 1958). Glyoxal and similar com- 

 pounds have a particular affinity for guanine residues, and are better 

 inactivating agents (Staehelin, 1959a,b). No mutants have l)een obtained 

 as the result of aldehyde treatment. However, it must be stressed that 

 the only technique which renders the detection of relatively infrequent 

 mutants a not too laborious task, the differential host method, has not 

 been systematically nor frequently applied to RNA modified by the 

 agents discussed in this and the preceding paragraph. 



The next class of agents to be discussed are the alkylating agents. 

 These vary greatly in their affinity for RNA, with mustard gas being the 

 most reactive and iodoacetate the least reactive of the series that was 

 investigated (Fraenkel-Conrat, 1961). It appears probable that the 

 introduction of one or two alkyl groups per mole of RNA (6400 nucleo- 

 tides) causes inactivation, regardless of the reagent used. The most 

 readily alkylatable sites are the guanine and adenine residues. When 

 various alkylated RNA preparations were surveyed for the presence of 

 mutants the methylated dcrivates were found to contain significant 

 numbers of mutated molecules among the survivors, but all other bigger 

 substituents appeared to be not clearly mutagenic under the same test 

 conditions. 



One other agent, A'-bromosuccinimide, which seems to introduce 

 bromine predominantly into the pyrimidine rings, was reported in pre- 

 liminary manner to be mutagenic to a similar extent as mcthylation 

 (Fraenkel-Conrat, 1961). Hydroxylamine and similar agents which at- 

 tack the pyrimidine bases have also been reported to effect mutations 

 (Schuster, 1961). 



The reaction which has proven particularly fruitful in tlie field of 

 vims mutagenesis has been the deamination reaction by means of nitrous 

 acid. Schuster and Schramm (1958) and Vielmetter and Schuster (1960) 

 established that the purine bases were deaminated at similar rates and 

 slightly faster than cytosine in RNA, and that one hit per 6000 nucleo- 

 tides had a 50% chance of inactivating the molecule. Oierer and INfundiy 

 (1958) subsequently showed that many of the non-lethal hits were muta- 

 genic. The single-hit nature of the curve relating mutation fi-efjuency to 

 reaction time showed that a single deamination could result in a muta- 



