1072
1701. 78
COMMUNICATIONS TO THE EDITOR
Acknowledgment.-We would like to express our deep appreciation t o Professor E. M. Purcell for a valuable discussion on several of the above points. This work was made possible by generous grants from E. I. du Pont de Nemours and Co., Inc., and from the Esso Education Foundation.
0.04 AI a slight broadening, of the order of one oersted, was observed. If the broadening arises from the electron exchange reaction, the rate constant for the bimolecular exchange is 4 X lo8 liter mole-lsec.-l. The absence of broadening a t lower concentrations establishes this value as an upper limit, ie., k 6 4 x 108 liter rnole-'sec.-'. MALLINCKRODT CHEMICAL LABORATORY HARVARD UNIVERSITY AKSEL A. BOTHSER-BY Previous work with radioactive tracers3v4 has CAMBRIDGE, MASSACHUSETTS RICHARD E. GI ICK indicated that k> 4 X l o 4liter mole-lsec.-'. RECEIVED DECEMBER 25, 1955
PARAMAGNETIC RESONANCE OF OCTACYANOTCNGSTATEiV\L
Sir : The paramagnetic resonance of crystalline potassium octacyanotungstate and its aqueous solutions has been observed. A polycrystalline sample of composition K3\V(CN)~.0.55H20yields a single almost symmetrical resonance a t room temperature. The g value a t the center of the peak is 1.98. The breadth of the peak between points of extreme slope is 30 oersteds. An aqueous solution, approximately 0.01 M in U7(CN)s---, yields a symmetrical resonance of three lines. The central intense line occurs a t g = 1.972 with, breadth between points of extreme slope of 9.3 oersteds. The two satellites, each of intensity 7 + 3y0 of that of the central peak are separated by 52 oersteds. The central peak in all probability arises from ions containing Wla4which has zero nuclear spin and is 86y0 abundant; the satellites are the hyperfine components associated with \V183which has spin 1/2 and is 147, abundant. The spectrum of the aqueous solution is unusual in its sharpness. Owing to rapid relaxation processes, most paramagnetic compounds of heavy elements have lines so broad a t room temperature, that the resonances are not easily observed. The hyperfine coupling constant is unusually large, corresponding to a magnetic field of 6 X lo5 oersteds a t the tungsten nucleus. The results suggest the possible usefulness of TV(CN)S--- in experiments involving alignment of the nuclei of some of the radioactive isotopes of tungsten. The conventional description of W(CN)a--ascribes the paramagnetism to an unpaired electron occupying a d orbital. The isotropic hyperfine interaction here described requires admixture of a configuration containing unpaired electrons in s orbitals.* Further experiments with dilute solid solutions of TI'(CN)8--- in single crystals of a diamagnetic substance are required for more complete determination of the nature of the electronic wave function. An attempt was made to measure the rate of electron exchange between W(CN)s--- and UT(CN)s--- by the observation of the spectrum of the former in the presence of the latter. At concentrations of 1V(CN)8--- lower than 0.02 M no broadening of the hyperfine components of \TVla3(CK)s--- was observed. With \V(CP\T)8--- a t ( I ) T h i s work has been supported in p a r t b y t h e United S t a t e s Air Force through t h e Office of Scientific Research of t h e Air Research a n d Development Command and b y t h e United S t a t e s Atomic Energy Commission under Contract A T ( l l - 1 ) - 3 4 with t h e University of California. (2) A. Abragam, J. Horowitz a n d M. 13. L. Pryce, P r o c . R o y . Soc. (London), AWO, 169 (1955).
(3) E . L. Goodenow a n d C. S. Garner, THIS JOURNAL. 77, 5 2 7 2 (1955). (4) H . Baadsgaard and W. D . Treadwell, Helu. Ckim. Acta, 38, 1669 (1955).
WASHINGTON UNIVERSITY ST. LOVIS,MISSOURI UXIVERSITY OF CALIFORSIA
S. I. XEISSMAN
AT LOS A S G E L E S CLIFFORD S. GARNER Los ANGELES,CALIFORNIA RECEIVED FEBRUARY 6, 19SG
THE PARTIAL STRUCTURE OF NOVOBIOCIN (STREPTONIVICIN).'
Sir:
11.
The isolation and properties of novobiocin have been d e ~ c r i b e d . ~ -The ~ present studies, added to those previously reported,5 indicate that it is a CS sugar attached glycosidically to the 7-position of 2 - butenyl) - benzamidol3- [4-hydroxy-3-(3-rnethyl-
4,7-dihydroxy-8-methylcoumarin. Novobiocin (I) (C31H36N2011), is cleaved by the action of hot acetic anhydride to yield 4-acetoxy-3(3-methyl-2-butenyl)-benzoic acid (11) and a neutral compound, C23H26N2010 (111).5 Hydrolysis of I by 4 N hydrochloric acid in 607, ethanol gave an optically inactive acid, C22H21h'06 (IV), which upon cleavage with hot acetic anhydride yielded 2,Z-dimethyl-6-chromancarboxylic acid (VI) and an optically inactive neutral compound, C14HllN05(VII).5 In addition to IV, the acid hydrolysis affords an (1) T h e Upjohn C o m p a n y Registered T r a d e Xfark for novobiocin i s Albamycin. Our previous Communication5 on t h e s t r u c t u r e of this antibiotic is listed under our former generic n a m e , streptonivicin, now abandoned. T h e isolation of t h e same material b y t h e Xletck group has been described in THISJ O U R N A L77, , 6401 (195.5). T h e comparisons establishing identity are described b y H e n r y Welch a n d W. W. n'right i n Antibiotics oizd Chemothevnpy, 6 , 6 i O (1935). (2) (a) Streptonivicin, A S e w Antibiotic. I . Discovery and Biologic Studies, C. G. Smith, A . Dietz, W. T. Sokolski and G. M. Savage, Anlibinlics and Chemolhernpy, i n press, February, 1956. (b) 11. Isolation and Characterization, H . Hoeksema, hl. E. Bergy, W. G. Jackson, J. W . Shell, J . nT.Hinman, A. E. Fonken, G . A. Boyack, E. L. Caron, J. H . F o r d , W.13. Devries, a n d G. C r u m , ibid. (c) 111. I n V i f v o and I n V i v o Evaluation, J . R . Wilkins, C. Lewis a n d A. R. Barbiers, ibid. ( d ) 11'. .A Biological Assay for Body Tissues a n d Fluids, R . SI. Taylor, W. T . Sokolski, G. R I . Savage and %I. J . Vander Brook, ibid. (e) V. Absorption, Distribution a n d Excretion, R. L f . Taylor, W. L. 3Ziller a n d M .J . Vander Brook, ibid. ( f ) V I . Toxicology, E. John Larson, N. E. Connor, 0. F. SwoaP, R. A . Runnells, \ I . C. Prestrud, T . E . Eble, D'. A. Freyburger. W. Veldkamp and R. I f , T a y l o r , ibid., hfarch, 19.56. (8) Streptonivicin (Albamycin) A New 4ntibiotic; Preliminary Report, F. R . Heilman, D. R . Sichols, W. E. n'ellman, a n d J. E . Geraci, P r o c . St&'.Ileetings M a y o Clinic, 30, 540 (1955). (4) Streptonivicin, Laboratory and Climcal Studies in t h e Pediatric Age Group, Feng-Kai 1,in and L. L. Coriell, T h i r d Annual Symposium on Antibiotics, November 2-4, 1955; "Antibiotics Annual, 19551056," Welch a n d I I a r t i - I b a n e z , Xfedical Encyclopedia, I n c . , iXew York, N . Y . , i n press. (5) H e r m a n Hoeksema, James L. Johnson, a n d Jack W. H i n m a n , THISJ O U R N A L , 7 7 , 6710 (1955). New analytical d a t a for I and 111 have been obtained as follows: Calcd. for CiiHJeNiOii (I): C, 60.77; H , 5.92; N, 4.58. F o u n d : C , 60.62; H , 5.91; N,4.54. Calcd. for CzaH~aNrOio(111): C , 56.33; H , 5.35; N, 5.71. Found: C , 56.53; H . 5.33; N, 5.68.
COMMUNICATIONS TO THE EDITOR
March 5, 1956
1073
prepared from 3-amino-4-hydroxycoumarin has characteristic ultraviolet fine structure identical with that of 111, VI1 and VIII. The spectra of 111, VII, VI11 and IX differ only slightly from those of the analogous models, and in a manner attributable to the effect of a phenolic group on the aromatic ring. Compound IX gives positive Benedict and ninhydrin reactions and nitrous acid converts CH3 IV it to a diazo compound analogous to that reported I by Link.6 The location of the 7-hydroxyl and 8-methyl groups was established by degradation of I X to known compounds. Fusion of IX with potassium hydroxide yields an ether-extractable dihydroxytoluene, m.p. 117-1 19' (calcd. for C~Hs02: C, 67.73; H, 6.49. Found: C, 67.89; H, 6.21). The latter is not oxidized a t a rotating platinum or a dropping mercury electrode in direct contrast to hydroquinone, p-toluyuinol, and 3-methylgentisic acid. Identification of the compound as 2methylres~rcinol~was confirmed by direct comCH3 VI1 , parison with an authentic sample. Treatment of I X with 1 N sodium hydroxide for several days a t room temperature or with hot Benedict reagent gives an ether-extractable acid purifiable by chromatography over Florisil (ether) and sublimation. This dihydroxytoluic acid (calcd. for CsH~04: HO C, 57.14; H, 4.8; C-CH3 8.99. Found: C, 57.3; H , 5.2; C-CH3, 9.18) melts with sublimation a t CHI IX X optically active neutral substance (V), m.p. about 200°, and was identified as 2,4-dihydroxy-m173-175' (calcd. for C11H2lNOs: C, 50.18; H, toluic acid by comparison with an authentic sample 8.04; N, 5.32; 1OCH3, 11.79; 1 OEt, 17.11. Found prepared by the procedure of Shah and Laiwalla.8 The optically active, neutral substance, V, apC, 50.68; H, 8.16;N, 5.25; OCH3, 11.71; OEt, pears to be the ethyl glycoside of a methoxy sugar, 17.01). The reaction with acetic anhydride thus apparently involves an amide linkage and the acid and X appears to be the corresponding methyl glycoside (calcd. for C10H19NO6: C, 48.19; H, hydrolysis cleaves a glycosidic linkage. Deacetylation of I11 with sodium methoxide in 7.68; N, 5.62; 2 0-CH3, 24.9; 1 C-CH3, 6.03. boiling methanol yields a neutral substance (VIII), Found: C, 48.11; H , 7.71; N, 5.61; 0-CH,, with ultraviolet absorption es- 21.8; C-CH3, 1.9). Infrared bands a t 1702 and m.p. 276-250', sentially the same as I11 (calcd. for Cz1H24Nz09: 1625 cm.-' are indicative of a urethan grouping. C, 56.34; H , 5.40; N, 6.25. Found: C, 56.07; Boiling rnethanolic hydrogen chloride causes H, 5.54; N, 6.12). Hot methanolic hydrogen elimination of the neutral nitrogen of X as chloride cleaves VI11 t o amphoteric IX and neu- ammonium chloride with the formation of another tral X, separable by fractional crystallization from neutral substance (XI) (calcd. for C10H1606: C, acid methanol and ether. Similar treatment of 51.72; H , 6.95; 2 O-CH3,26.73. Found: C, 51.97; VI1 affords I X . Under appropriate hydrolytic H, 6.91; O-CH3, 27.59) with infrared adsorpconditions, VI1 is converted to 7-acetyl-IX or to tion and chemical properties indicative of a cyclic N-acetvl-IX, each subsequently convertible to carbonate ester. Reaction with barium hydroxide IX. Recrystallization of IX from water or in aqueous solution a t room temperature causes aqueous ethanol gives the free base (calcd. for precipitation of the theoretical amount of barium CIOH~XO~: C, 57.99; H, 4.38; N , 6.77; C-CH3, carbonate with formation of X I I . The nitrogen 7.3. Found C, 57.69; H, 4.32; N, 6.25; C-CH3, of the glycoside moiety appears, therefore, in the 6.25). Compound IX and its free base decompose form of a carbamate. The single acetylatable gradually above 200'. The acidic benzoyl deriva- hydroxyl (indicated by the transformation I11 +. tive is obtained under Schotten-Baumann condi- VIII), the methoxyl group and a bridge oxygen tions, m.p. 309-310' (calcd. for C17HlnNOs:C, 65.59; account for all of the remaining functionality of X. H, 4.21; N, 4.50. Found: C, 65.63; H, 4.19; N, The low C-CHs analyses are consistent with the 4.42). When this is treated with hot acetic anhy- demonstration of twin infrared bands of gemdimethyl groups a t 1382 and 1366 ern.-' in X, X I dride and pyridine, VI1 is obtained in good yield. The reactions VI1 + N-acetyl-IX + I X , and X I I . The facile conversion of X to the carIX + VII, and IX + N-benzoyl-IX + VI1 proceed bonate ester (XI) and the ready reduction of one identically with those carried out on model com- mole of periodate by XI1 limit the acetylatable pounds prepared from 3-amino-4-hydroxycouma- hydroxyl and carbamate groups to positions on adrin generously supplied by Prof. K. P. Link, or pre- jacent carbon atoms. pared according to his procedures.6 The oxazole ( 7 ) E . T. Jones and A . Robertson, J . Chem. Sric., ID00 (1932).
+
( 6 ) C. F. Heubner a n d R . P L i n k , THISJ O U R N A L , 67, 99 (1945).
(8) R . C. S h a h a n d M . C. Laiwalla, ibid., 1828 (1938).
1074
Vol. 7s
COMMUNICATIONS TO THE EDITOR
These data permit the following partial structure for novobiocin :
(? H , S H C 0
present as indicators of the inversion.
/CHJ
UDPG
UDPGal -Y+ U-DPG UDPGA f Z U P S I I
+ 2 D P S - -+
(3)
+ 2H-
(4)
\A/CHzCH-C 'CH3 LVhen
mrified galacto-waldenase was incubated with UDPGil in the absence of the indicator system, the inversion reaction did not proceed. It could be initiated, however, by adding catalytic amounts of DPN. This is illustrated We acknowledge with gratitude the contribu- in Table I. That the cofactor is DPN and not some tions of Drs. M. Calvin, H. E. Carter, D. J. Cram, i:npurity in the preparation is borne out by experiJ. L. Johnson and E. C. Olson, Mrs. ,4. E. Fonken ments using purified Seurospora, D P N ~ s ~DPK .~ and Mr. W. -1.Struck to this work. preincubated with DPNase until it was no longer active when assayed with alcohol dehydrogenase was RESEARCH LABORATORIE s J A C K \V. HISMAS THEUPJOHNCOMPASV HERMAN HOEKSEMA not active in the galacto-waldenase system. : I F ; a ~ . ~ ~ r a z MICHIGAN oo, E. LOUISCAROT control with DPN similarly preincubated with \V,G. JACKSOS heat inactivated DPNase was active. DPNH RECEIVED ~;E:~RI.ARY 9, 1933 was inactive in the system unless it was oxidized by preincubation with acetaldehyde and alcohol deDIPHOSPHOPYRIDINE NUCLEOTIDE, A COFACTOR hydrogenase. DPX could not be replaced by T P K . FOR GALACTO-WALDENASE' Sly:
The interconversion of Gal-1-P and G-1-P in galactose adapted and in mammalian tissue4 has been found to occur according to the equations G'il-1-P
+ UDPG I-UUPG\(,r,v \ I
i i \ l
41mI
Illitittiti.,
1
