3228
COMMUNICATIONS TO THE EDITOR
chondria were used immediately after isolation. All data are based on from two to five independent determinations. The data reveal that the uncoupling efficiency of a given concentration of D N P depends upon M, the substrate metabolized. At 5 X D K P uncoupled approximately 50% of the phosphorylation associated with a-ketoglutarate oxidation. This same concentration of D N P had relatively little effect on the phosphorylation coupled with the oxidation of a-glycerophosphate, succinate and glutamate. At a higher level of DNP, 1 X M, phosphorylation associated with aketoglutarate oxidation was almost completely eliminated. I n contrast, phosphorylation coupled with glutamate oxidation was only slightly reduced. 113th the respiratory substrates, succinate and aglycerophosphate, moderate decreases in the phosphate esterified were observed. Neither concentration of D N P caused marked alterations in the respiratory rate. Other experiments with a furM, comther five-fold increase in DNP, 5 x pletely uncoupled phosphorylation with all substrates. These data indicate that phosphorylations coupled to electron transfer are not equally sensitive to DNP. This implies that during oxidative phosphorylation steady state concentrations of the components of the system are established and that this balance regulates the response to a given concentration of DNP. In our experiments the different substrates may have modified this equilibrium by influencing the steady state levels of respiratory enzymesg or by having different phosphorylation sites.’O Other evidence demonstrates that the oxidation of these substrates in the fly is mediated by the same respiratory chain components as found in mammalian preparations.”,l2 In both organisms the initial electron acceptor for a-ketoglutarate and glutamate is pyridine nucleotide. This coenzyme is not a participant in the mitochondrial oxidation of a-glyceropho~phate.’~Thus it is unlikely that the apparent contrast in our results with some previously reported data with mammalian tissues14 can be attributed to gross differences in the electron transport system. This comparison therefore suggests that the differing effects of a given concentration of inhibitor may be a reflection of differing equilibrium states. (9) B. Chance a n d G. Williams, J . Bid. Chem., 2 1 7 , 409 (1955). (IO) B. Chance a n d G. Williams, i b i d . , 217, 439 (1955). (11) B. Sacktor, J . Gen. Physiol , 35, 397 (1952). (12) B. Sacktor, Arch. B i o c h e m B i o > h y s , 46, 3.19 (1953). (13) B . Chance and B. Sacktor, unpublished results. (14) F. Hunter, in “Phosphorus Metabolism,” (W. D. McElroy a n d B. Glass, editors), T h e Johns Hopkins Press, Baltimore, Md., 1, 297 (1951).
Vol. 78
contains a 5- or 6-hydroxybenzimidazole moiety‘, and appears to differ from vitamin BIZ only by having this moiety in place of 3,C,-diniethylbenzimidaiole. The following evidence shows that the nucleoside of Factor 111 is a >-hydroxybenzimidazole glycoside, probably an a-ribofuranoside.4 Factor I11 was kindly supplied by I’rofcssor Dr. K. Bernhauer. The nucleoside (I)’ obtained by acid hydrolysiq of Factor I11 was not isolated in crystalline form but was purified by paper chromatography using 1.5) system an n-butanol-acetic acid-water (4: (X,0.46). The absorption spectruni in methanol showed, , ,A a t 249 and 296 mp, and the [ m ] ” D was about - 11 ’. Since 5-hydroxybenzimidazole could result from degradation of either a 5- or 6-h~dro~ybexizimidazole glycoside the position of the hydrouyl group in I was shown as follows R
1 .,
I1
CHa
The Factor I11 nucleoside I was methylated with diazomethane and the product was converted to the methiodide by refluxing in methanol with methyl iodide. The methiodide was cleaved to an o-phenylenediamine by treatment with hot methanolic sodium hydroxide, conditions known to cleave N,N’-dialkylbenzimidaiolium salts to N,N’-dialky-1-o-phenylene diamine^.^ In the case Of I , these conditions would also be expected to cause removal of the sugar moiety, giving a monomethyldiamine in which the unsubstituted amino group is derived from the original glycosidic nitrogen. For characterization, the diamine from I was converted into 1-methyl-6-methoxybenzimidazole(11), m.p. 67-68’, by reaction with formic acid. I1 was synthesized by first reacting 3-bromo-lnitroanisole and methylamine to give N-methyl-6nitro-m-anisidine. This was hydrogenated to the
DIRECTORATE OF MEDICAL RESEARCH (1) F.hl. Robinson, I . X I . Miller, J. F. hfcPherson and K . Folkers, CHEMICAL ~ ‘ A R F A R ELABORATORIES BERTRAM SACKTORTHISJ O U X K A L , 7 7 , 5192 (1955). A R M Y CHEMICAL CENTER,MARYLASD DOXALDCOCHRAN (2) W. Friedrich and K. Bernhauser, Angew. Cheni , 67, GI9 (l95,7> (3) W. Friedrich and B. Bernhauser, Z . Soturfot’schung, 9b, RECEIVED MAY4, 1956
VITAMIN BIZ. XXVII. STRUCTURE OF THE FACTOR I11 NUCLEOSIDE AND SYNTHESIS OF HYDROXYAND METHOXYBENZIMIDAZOLE RIBOSIDES
Sir: It has been reported previously that Factor I11
(1954). (4) B y private communication, we have been informed b y Professor Dr. K. Bernhauer t h a t he and Dr. Friedrich have methylated Fact o r I11 directly, and then by degradation have obtained G-hydroxy-1-methylbenzimidazole. Thus, they have established t h a t t h e nucleoside of Factor 111 is a 5hydroxybenzimidazole glycoside (W. Friedrich and K. Bernhauer, A x g w . Chem., in press). ( 5 ) K. Hoffman, “Imidazole and Derivatives,” P a r t I, Interscience Publishers, Inc., New York, N. IT., 1953, p. 2S0.
July 5 , 1956
3229
COMMUNICATIONS TO THE EDITOR
corresponding diamine which reacted with formic since to our knowledge the chromophore repreacid to give I1 (calcd. for CsHloNz0: C, 66.65; sented by the aliphatic a,P-unsaturated azoxy group H, 6.22; N, 17.28. Found: C, 66.89; H, 5.99; has not previously been reported.2 The previous characterization indicated I as an N, 17.40). A mixed melting point of the isolated and synthetic compounds was not depressed. optically active, neutral oil, C13H2~N203,with one Therefore, I was a 5-hydroxybenzimidazole glyco- alkoxyl and two terminal methyl groups. The side. hydrogen content allowed two double bonds and The available knowledge made i t seem likely the ultraviolet spectrum (A, 237.5, e = 11,000) that the Factor I11 nucleoside, like a-ribazole,G showed them in conjugation. The infrared spectrum of I indicated an -OH is an a-D-ribofuranoside. For comparison, two methoxybenzimidazole ribofuranosides were syn- and/or possibly an -NH. The infrared spectrum thesized. Reaction of the chloromercuri deriva- of the monoacetate derivative, b.p. 54-90" (0.5 p) ; tive of 5methoxybenzimidazole (which probably A,, 237.3, E = 11,000; [ C Y ] ~ ' D+25.3', 3% in was a mixture of 1-chloromercuri-5- and -6-meth- ethanol (calcd. for C15H2sNz0.r: C, 59.97; H, oxybenzimidazoles) with triacetyl-a-D-ribofurano- 9.39. Found: C, 60.28; H, 9.70), which had no syl chloride followed by deacetylation gave two absorption in the 3 p region, indicated an acetate products which were purified by countercurrent ester and thus confirmed the presence of the only distribution in a butanol-water system and sepa- acetylatable hydrogen as an -OH. rated as their picrates (111) m.p. 144-146' and I and the acetate of I absorbed two moles of (IV) 78-82' (calcd. for C19H19N5012: C, 44.80; hydrogen (Pt, alcohol) to give 11,3 b.p. 70-80' H, 3.76; N, 13.75. Found: (111) C, 44.74; (4 p ) ; A,, 229, E = 5300; [a]% +3.7', 5% in H, 3.40; N, 13.88; (IV) C, 44.36; H, 3.94; N, chloroform (calcd. for C13H&202: C, 63.89; H, 13.58). Regeneration of the bases by use of IR- 11.55; N, 11.46. Found: C, 63.31; H, 11.44; 45 resin gave from I1 a crystalline base (V), N, 10.80), and the acetate of 11, b.p. 63-67' m.p. 162-164', [(YIz2D -33', and from 111 a non- (0.5 M); A,, 230, E = 4700 (calcd. for CljHaoNzOa: crystalline base (VI), [a]"D -36'. The nearly C, 62.46; H, 10.48; N, 9.71. Found: C, 62.40; equal [ a ] D values indicate the same glycosidic H, 10.60; N, 9.53). configuration and therefore a difference in the posiThe facile reductive loss of a single oxygen atom tion of the methoxy group. The positions of these indicated an N-oxide, which, with the alcohol and groups were determined by the method described alkoxyl groups, accounted for the three oxygen above. From V was obtained 1-methyl-6-methoxy- atoms. I1 and the acetate of I1 were easily hybenzimidazole and from VI, 1-methyl-5-methoxy- drolyzed to n-octanol, providing the skeleton of the benzimidazole,' m.p. 110-113'. Therefore V was Cg moiety. a 5-methoxybenzimidazole riboside and VI was 0 CHzOCH3 the 6-methoxy isomer. T I Demethylation of V with 48% hydrobromic acid TZ-C~H~~CH=CH-N=K-CH i gave a non-crystalline product (VII) which was CH-OH purified by paper chromatography (Rf 0.47), I I [ a I z 5 D about -59'. The ultraviolet absorption CH3 spectra of VI1 and I were nearly identical. If the CH20CHa assumption that I is a D-ribofuranoside is correct, I n-CsHiaCHzCH=N-NH-CH it is the a-form and the synthetic compound VI1 i is the 0-form. CH-OH The similarities in the behavior of the natural I1 I hydroxy- and methoxybenzimidazole glycosides CH3 0 CH20CH3 with that of the corresponding synthetic ribosides II I lend support to the postulate that I is an CY-DCHaC--I\"-CH 0 ribofuranoside. Studies are now in progress to obi II tain further evidence concerning structure. CH-0-C-CH, (6) N. G. Brink and K . Folkers, THISJOURNAL, 74, 2856 (1952).
(7) A. M. Simonov and P. A. Uglov, J . Gen. Chem., 21, 884 (1951).
I11
1
CH3
The fact that in the acetic acid I absorbed four CONTRIBUTION FROM THE CLIFFORD H. SHUNK MERCK,SHARP& DOHME FRANKLIN hl. ROBINSON moles of hydrogen (Pt) or I1 absorbed two moles of RESEARCH LABORATORIES JAMESF. MCPHERSON hydrogen with the formation of two amines showed DIVISIONOF MERCK& Co., INC. MARJORIE M. GASSER RAHWAY, h-EW JERSEY KARLFOLKERSthat the two fragments were joined in the original via the two nitrogen atoms and clearly indicated RECEIVED JUNE 5 , 1956 THE STRUCTURE OF ELAIOMYCIN, A TUBERCULOSTATIC ANTIBIOTIC
Siy: Structure I is proposed for the antibiotic Elaiomycin.' This antibiotic is chemically unique, (1) T. H. Haskell, Q. Bartz, et al. [Antibiotics and Chemotherapy, 4, 141, 338 (1954) I described the isolation, purification, spectra, chemical characterization, and biologic studies of Elaiomycin [4-methoxy-3-(1octenyl-NON-azoxy)-2-butanol].
an azoxy group. The amines were identified as n-octylamine and a C5 amine, isolated as the completely acetylated acetate-amide 111, b.p. 85-90'
(2) Recently the plant poison, Macrozamin, has been shown by Lythgoe [J. Chem. Soc., 2311 (1951)l to contain an asymmetrical aliphatic azoxy group. Hydroscopin A [J. Antibiotics ( J a p a n ) , 7 , 329 (1954)l may also be similar although no structural work has been reported. (3) I1 was formulated as an aliphatic hydrazone since the ultraviolet and infrared spectra differed from those of an azo compound but resembled those of a model compound C~HW-C(CH~)=N-NH-CHI (X-X 229, e = 5300).
