Ultraviolet radiation promotes the formation of a cyclobutane ring between adjacent pyrimidine residues on the same DNA strand to form a pyrimidine dimer. Such dimers may be restored to their monomeric forms through the action of a light-absorbing enzyme named DNA photolyase. The redox-active cofactor involved in the light-induced electron transfer reactions of DNA repair and enzyme photoactivation is a noncovalently bound FAD. In this paper, the FAD cofactor of Escherichia coli DNA photolyase was characterized as the neutral flavin semiquinone by EPR spectroscopy at 9.68 and 94.5 GHz. From the high-frequency/high-field EPR spectrum, the principal values of the axially symmetric g-matrix of FADH(*) were extracted. Both EPR spectra show an emerging hyperfine splitting of 0.85 mT that could be assigned to the isotropic hyperfine coupling constant (hfc) of the proton at N(5). To obtain more information about the electron spin density distribution ENDOR and TRIPLE resonance spectroscopies were applied. All major proton hfc's could be measured and unambiguously assigned to molecular positions at the isoalloxazin moiety of FAD. The isotropic hfc's of the protons at C(8alpha) and C(6) are among the smallest values reported for protein-bound neutral flavin semiquinones so far, suggesting a highly restricted delocalization of the unpaired electron spin on the isoalloxazin moiety. Two further hfc's have been detected and assigned to the inequivalent protons at C(1'). Some conclusions about the geometrical arrangement of the ribityl side chain with respect to the isoalloxazin ring could be drawn: Assuming tetrahedral angles at C(1') the dihedral angle between the C(1')-C(2') bond and the 2p(z)() orbital at N(10) has been estimated to be 170.4 degrees +/- 1 degrees.

En outre, si u11 kchantillon d'uraniuin doit Stre analysk pour d6terminer les traces de fer, de mangani.se e t de cobalt, et s'il est dissous en HN03, le nitrate d'uranyle cristallisk peut &tre facilement dissous dans l'kther-HNOa, 2%. Colllpte tenu des rksultats de cette ktude, on pourra &re sClrs que les traces des trois lnktaux passeront compl&tement en solution dans l'kther, pour une successive skparation chromatographique.

The hyperfine electron-spin resonance spectra of a number of methyl-substituted 9,10-anthrasen~iquinones have been determined. The spectra have been analyzed by computer techniques, and splitting constants have been assigned for all protons. T h e results indicate that, while the parent quinones are essentially free of steric hindrance, the methyl groups in the cr position of the reduced quinones are not free to rotate. T h e methyl groups in t h e (3 position of the semiquinones show nor~nal hyperconjugative behavior. These results are consistent with the different redox potentials of these compounds revealed in a previous study.In a polarographic study of substituted 9, Cra~vford et al. (1) showed that substitution of a illethyl group for a P proton produced little change in the polarographic reduction potential. However, substitution of a methyl group for an a: proton caused a considerable increase in the polarographic reduction potential or decrease in the oxidation potential. This was shown to be due to steric factors by judicious choice of suitably substituted anthraquinones. In particular, the steric factor was assuilled t o operate by causing a resistance to planarity in the reduced quinone compared with the quinone itself, which must be some\vhat nonplanar (2). Gill and Stonehill (3) suggested that steric hindrance pushed both the =O and large a! substituents out of the ring plane.In this work a n attenlpt is made to investigate the hyperconjugation" (4) and restricted rotation by means of the electron-spin resonance (e.s.r.) hyperfine splitting constants of methylated anthrasemiquinones (5). The results are compared with preliminary observations on the nuclear magnetic resonance (n.n~.r.) and infrared spectra of the corresponding quinones. E X P E R I M E N T A LThe quiriones used in this investigation, except those described below, were provided through the kindness of Dr. R. J. Crawford of the University of Alberta. They were prepared by known procedures and purified by careful recrystallization from isopropyl alcohol. T h e observed melting points were anthraquinone. 286-288"; 1-inethylanthraquinone, 172-173'; 2-methylanthraquinone, 175-177"; 1,2-dimethylanthraquinone, 159-161"; 1,3-dimethylanthraquinone, 163-165"; 2,6-diinethylanthraquinone, 245-247"; 2,7-dimethylanthraquinone, 171-173"; and 1,4-dimethylanthraquinone, 151-142". 2-CD3-Anthraquino~zeToluene-CD3, RiIerclc, was allowed to react with phthalic anhydride in the presence of aluminium chloride. The resulting 4-CDj-benzoylbenzoic acid was cyclized by treating it with fuming sulfuric acid. T h e resulting 2-CD3-anthraquinone had a melting point of 172-174 "C, and no stretching vibration in the infrared caused by aliphatic CH bonds. 1-CD3-A ~tthraqz~inoneToluene-CDa, AiIerck, was brominated in the presence of nitric acid (6). The product contained both oand p-bromotoluene-CD3 which were separated by preparative gas chron~atography using a Wilken's Aerograph model 600 gas chromatograph with benzoquinoline on firebrick packing. o-Bromotoluen...