Chemical modification of viral RNA. VI. The action of N-methyl-N'-nitro-N-nitrosoguanidine.

N-Methyl-N'-nitro-N-nitrosoguanidine (MNNG, "nitrosoguanidine") has been extensively used as an effective mutagen for DNA-containing organisms and virusesl-4 but appears not to have been used with RNA viruses. Nothing definitive seems to have been reported about its mode of action. This paper deals with the effects of MNNG on the infectivity, mutation frequency, and base composition of tobacco mosaic virus (TMV) and TMV-RNA, as well as of model compounds. Methods and Materials.-MNNG was obtained from the Aldrich Chemical Company. All solutions were prepared fresh before use. At times, reaction mixtures were maintained at saturation (about 0.04 M at 25°) by repeated addition of solid MNNG. The preparation of TMV-RNA and the conditions of assay, usually after reconstitution, for both infectivity and mutation frequency were the same as used in previous studies in this series.Iae For reaction in predominantly nonaqueous media, the RNA was precipitated in 70 per cent ethanol and the precipitate taken up in formamide or 67 per cent dimethylformamide or dimethylsulfoxide (since it was not soluble in these pure solvents). TMV-RNA and other natural ribonucleic acids were freed of the reagent and its decomposition products by repeated alcohol precipitation, polynucleotides by passage through G-25 Sephadex, and nucleotides or dinucleoside monophosphates by removal of the reagent and its decomposition products by means of paper chromatography in butanol-water (86:14), which leaves the nucleotides at the origin. TMV was separated from the reaction mixtures by ultracentrifugation. Results.-Reaction of TMV-RNA: Aqueous solutions of MNNG rapidly decreased the infectivity of TMV-RNA over the entire pH range tested. The rate of inactivation of TM\V-RNA was not always strictly proportional to the amount of reagent used. It decreased with time, more rapidly at pH 4.5 and pH 7 than at pH 9. The observed and somewhat variable decreases in inactivation are attributed to the instability of the reagent in dilute buffered aqueous solution, as borne out by experiments in which the reagent caused less inactivation if it was added to the RNA after having been held under the usual experimental conditions for one hour. The rate of inactivation of TMV-RNA in the dispersing solvents-formamide, 67 per cent dimethylsulfoxide, and particularly 67 per cent dimethylformamidewas much slower than in pure aqueous media, and yet more so in the presence of phosphate buffer or other salts (Table 1). The amount of full-length RNA ('-30S) was not noticeably affected during 15 minutes of reaction in any solvents used. When the proportion of mutants among the infective particles was investigated, it was found somewhat higher for MNNG-treated RNA than for the mock treated controls, but it never approached the level obtained with nitrous acid, hydroxylamine, methoxyamine, or dyes plus light.0. e The mutation frequency of samples treated with MNNG in water ranged about a twofold and in formamide-MNNG-