CH=C(CHdCH2and ring - C H 3 ; 8.46 (32) m, -CH,CH=C(CH,)CHZ-. The methoxy groups in coenzyme Qloreadily undergo I1 Ia, R = R' = C6H5C0 alcoholysis in basic The two methoxy groups b, R = C6HsC0, R'= CeHa in coenzyme Q l o appear to have comparable reactivity in C, R = R'=CaHs such displacements since the monoethoxy homolog was Irradiation of a benzene solution of IIb4 resulted in shown to be a mixture of the two possible isomers.5 partial conversion to the previously unreported cis It appeared that ammonolysis of coenzyme Qlo to give isomer Ib, m.p. 69.5-71' (cyclohexane); ' ": : :A synthetic rhodoquinone was possible. This was ac240 mp ( E 14,100); v~::Oimul' (cm.-l) 764 s, 748 w, complished by treating coenzyme Qlo with ammonium 721 s, 700 s, 694 s, 681 s, and 648 m. A solution of hydroxide in a solvent of diethyl ether-ethanol (1 : 1). I l b and Ib (0.026 and 0.025 M , respectively, in benzene) The reaction gave a mixture of several unidentified reached an apparent photostationary state (1Ib : Ib = products in addition to a purple quinone, m.p. 39-45', 1.2, determined by gas chromatography o n column 1 which gave the same Rf values as natural rhodoquinone at 210" after 6 hr.6 All components were found on alumina (0.37) and silica gel G (0.33) thin layer to be stable under the conditions of analysis. plates in 40% ether in n-hexane and in reverse-phase Photointerconversion of IC and Ilc could also be paper chromatography on silicon-impregnated paper effected by direct irradiation, but the conditions had to (0.75) in a solvent of water-1-propanol (1:4), The be altered and side reactions became significant. paper chromatographic comparison of synthetic and Irradiation of 0.1 M benzene or cyclohexane solutions natural rhodoquinone, which was carried out under of cis- or trans- 1,2-diphenylcy~lopropane~ in quartz conditions used t o distinguish side-chain length^,^ in vessels with 2537 A. light8 at 40' induced photoisomericonjunction with the nuclear magnetic resonance speczation. The ratio IIc :IC was approximately 0.65, trum of natural rhodoquinone (showing ten vinyl prodetermined after brief irradiation of mixtures approachtons), suggests that the natural product contains a ing this composition. The products, after prolonged side chain of ten isoprene units. Comparison of the melting points of the synthetic and natural rhodoquinone (39-45 vs. 69-70') suggests that the synthetic material is a mixture of the two isomeric aminoquinones and that rhodoquinone is formed enzymatically within the bacteria rather than artifactually by the ammonium ion in the growth medium. The infrared, ultraviolet, irradiatjon (8.5 hr.) of IJc in benzene, were IC (22 %),6b and nuclear magnetic resonance spectra of synthetic IIc (2073, 111 ( 7 7 3 , IV ( 1 6 z ) , V ( 7 7 3 , and VI (trace). and natural rhodoquinone are indistinguishable. All were isolated by gas chromatography (column These data, in conjunction with the observation by 3,5a 180') and identified by their infrared spectra and Rudneys that coenzyme Qlo is a biosynthetic interretention volumes. A trace of 1,3-diphenylpropane mediate to rhodoquinone in R. rubrum, show that the could have escaped detection since its retention time and structure of natural rhodoquinone is either 2-amino-3infrared spectrum resembled those of V. methoxy- or 2-methoxy-3-amino-5-methyl-6-[3'-meth- The isomerization and the formation of compounds yl-2 '-butenylen a k i s-(3 '-met hyl-2'-bu tenylene)]-l,4111-VI may be rationalized on the assumption that bond benzoquinone. a in IC or IIc cleaves t o form a species resembling a diradical, represented here by VII.g Scisson of bond (5) B. 0. Linn, N . R . Trenner, B. H . Arison, R . G. Weston, C. H. b to give the much less stable VI11 apparently does not Shunk, and I
