METHOSULFATE, STREPTONIGRIN, AND RUBIFLAVIN IN BACTERIAL SUSPENSIONS*

The existence of an EPR signal in a biological sample provides direct evidence for the presence of free radicals in the sample. If hyperfine structure is discernible in the signal, the free radicals may be identifiable by comparison with known spectra. In a signal which is due to a known radical, the quality of the hyperfine structure can indicate the extent of binding of the radical to macromolecules.2, 3 We report here that an EPR signal develops when any one of the inhibitorsphenazine methosulfate, streptonigrin, or rubiflavin-is added to a suspension of bacteria. The existence and structure of these signals may prove important to understanding the mode of action of the inhibitors. Materials and Methods.-Bacteria: Aerated cultures of Escherichia coli strain B, growing exponentially in a tris-buffered minimal medium4 or in trypticase soy broth (Difco), were harvested, washed twice with 0.15 M NaCl, and resuspended in minimal medium (without glucose) at a concentration of 1010 cells/ml. Bacillus megaterium was harvested while in exponential growth in trypticase soy broth, washed as in the case of E. coli, and resuspended at a concentration of 108 cells/ml. Bacterial suspensions were kept at 0?C until used and gave satisfactory results for at least 24 hr. Streptonigrin (NSC 45383) was a gift from Dr. Tom J. McBride of the John L. Smith Memorial for Cancer Research. Rubiflavin (batch AA-3250-115-1) was a gift from Dr. Adorjan Aszalos of the Squibb Institute for Medical Research. Phenazine methosulfate was purchased from Nutritional Biochemicals Corporation. These inhibitors were used without further purification. EPR observations were made with a Varian V-4502 EPR spectrometer employing 100-kc magnetic field modulation. The sample was placed in the Varian aqueous sample cell. Magnetic field calibration was carried out with an aqueous solution of peroxylamine disulfonate in saturated sodium carbonate. Estimates of the number of electron spins/ml were made by comparison with a standard sample (pitch inKC1) supplied by Varian. Such estimates give only order of magnitude because of the difference in geometry between the standard and the sample cell. Results.-EPR spectra obtained from free radicals of PMS, streptonigrin, and rubiflavin are presented in Figures 1, 2, and 3. The free radicals were prepared in each case both by chemical and biological reduction of the corresponding unreduced form. Biological reduction was effected in suspensions of log phase cells of either B. megaterium or E. coli, with the exception that with rubiflavin only E. coli gave a signal. The quinone and a source of electrons, such as ethanol, were added to a small volume of the bacterial suspension, and a portion was charged in the aqueous sample cell. As a result of metabolic oxidation of the electron source, the contents of the cell became anaerobic quite rapidly, and thereafter the quinone acted as the principal electron acceptor. Ethanol was found to provide electrons at a convenient rate for observation of the free radicals. Other possible electron sources were screened, using streptonigrin and E. coli. Glucose, acetaldehyde, n-butyl, n-propyl, and isopropyl alcohols gave no signal. Methanol gave a very weak signal; acetone was somewhat better. It is possible that the ethanol performs some other function than simply to act as a source of electrons, since glucose 887