COMMUNICATIONS TO THE EUITUK
Xug. 5 , 1!15!1
4111
dependent of any added cofactor. N-Acetylhexosamine disappearance from the reaction mixture was measured by loss in reducing power2 or t y a modification of the Morgan-Elson reaction. Reaction mixtures contained these additions in a final volume of 1 ml.: 1 pmole of K-acetylhexosamine, 3 pnioles of phosphate buffer, $H 7 . 2 , and enzyme (30-200 p g . ) . Incubations were carried out by shaking a t room temperature and in an air atmosphere for 90 minutes. It was noted that cne-half mole of oxygen was consumed for each mole of N-acetylhexosamine and reducing sugar that disappeared. The disappearance of substrate was dependent on an aerobic mechanism as shown by a doubling of the rate when the reaction was incubated under oxygen rather than air. Furthermore, incubation under nitrogen completely inhibited N-acetylhexosamine disappearance. These data suggested that the reaction catalyzed by the enzyme involved a n oxidation of carbon 1 to yield the corresponding N-acetylhexosaminic acids. The products (I) of enzyme action on the Nacetylhexosamines were chromatographed on paper in three dirTerent solvent systems. Only the Rf values found in a butanol, acetic acid and water system (30: l5:25) are reported even though comparable results were obtained with all three systems. Rfvalues of 0.34 were observed for I when it was adjusted to pH S prior to spotting and visualized with C l - ~ t a r c h - K I . ~Furthermore, I or authentic N-acetylhexosaminic acids, when acidi'' A' fied to pH 1 before applying to the papers, gave H spots which reacted readily with the hydroxylamine-FeCla reagents,j indicating lactciie formation which is characteristic of the S-acetylhexosaminic acids6 Hydrolysis of I in 2 ,liHC1 for 2 hours a t 100" converted I t o a compound (11) t h a t reacted with ninhydrin and had an Rfvalue of 0.22. Synthetic glucosaminic and galactosaminic acids (Rr 0.22) and I1 behaved identically in three solvent systems. Chromatographically, it was not possible to separate N-acetylglucosaminic acid from Kacetylgalactosaminic acid, nor was i t possible to distinguish between glucosiminic acid and galactosFigures VIII, X-XI. aminic acid. These compounds were identified by converting them to their corresponding pen( 7 ) D H R Barton, T. Bruun and A S. Lindsey, J . Cheirz. Soc , 2210 (1932). toses.' The product of N-acetylglucosamine oxidaDEPAXTMENT OF CHEMISTRY MARVIN CARMACKtion gave rise to a compound that was identified INDIASA CNIVERSITY DAXAW. MAYO as arabinose ( R f 0.26, butanol :ethanol: water, BLOOMINGTOS, INDIANA JAMESP. FERRIS 4 : 1 : 1). The product of N-acetylgalactosamine RECEIVED M A Y 26, 1969 oxidation gave rise to a compound identified as lyxose (Rr 0.32, butanol :ethanol : water, 4: 1 : 1). ENZYMATIC OXIDATION OF When treated in the same manner as I, authentic N-ACETYLHEXOSAMINES TO samples of the corresponding N-acetylhexosaminic N-ACETYLHEXOSAMINIC ACIDS' acids and the hexosaminic acids behaved identically. Sir: These data show that the enzymatic reaction Crude extracts obtained from a strain of Proteus zwlgaris 31 XI contained an enzyme that catalyzed catalyzed by Proteirs uulgnris involves a direct the disappearance of N-acetylglucosamine or N- oxidation of the free N-acetylhexosamiiies t o the ( 2 ) A?. Ss., 62, silver nitrate and which undergoes rapid solvolysis in aqueous methanol. The hydroxyl thus exhibits a pattern of reactivity quite unlike t h a t found in the similar bicyclo [4,3,l]decane system of pcaryophyllene alcoh01.~ Furthermore the formation of dehydrodesoxydeltaline (V) under mild conditions6 cannot be satisfactorily interpreted with Structure X because the introduction of a double bond a t the bridgehead position, R2,would require an unacceptable violation of Bredt's rule. Models show, however, that an essentially coplanar bond can fit a t the CI3-C5 position of Structure V (no bridgehead). A perhydrophenanthrene skeleton was suggested but not favored by Cookson and Trevett4b as a possible precursor to demethylenedelpheline. We believe that the latter and demethylenedeltamine are best represented by X I (R1= -OH, R 2 = -H, R 3 = -OH, R 4 = -OH) and VI11 (R1= -OH, R 2 = ---OH, R 3 = -OH, R 4 = -OH), respectively, and that desoxylycoctonine is the 0methyl ether of X I (R1= -0CHJ.
11
Service.
373 (1964).
an isocyanide stretching f r e q u e i ~ c y . ~R strong broad band with maximum absorption a t 728 cm. -', considerably displaced from the Sic1 band at 800 cm.-' for Sic14 and a t 810 cm.-l for HSiCla, is undoubtedly the Sic1 band since it is the only other major band in the spectrum. A more detailed study of the spectrum for this (8) L . F. Leloir, C. E. Cardini and J. hl. Olsoarria, Arch. Biochem. compound is indicated before one can draw any Bioghys., 74, 84 (1968). well-founded conclusions concerning its structure. DEPARTMEXT O F MEDICAL MICROBIOLOGY, UNIVERSITY However, the features so far observed are compatOF SOUTHERN CALIFORSIA SCHOOL OF MEDICISE Los ANGELES,CALIFORNIA, AND L. I. HOCHSTPIN ible with either a very rapid cyanide-isocyanide 1. B. ~ V O L F E SCRIPPS CLINICAND RESEARCH FOITNDATION LA JOLLA, CALIFORNIA ~ iI.. N A K A D A equilibrium3 greatly favoring the cyanide form, or a cyanide model with asymmetry introduced by backRECEIVED J U N E 3 , 195% bonding involving the 3d orbitals of the silicon. This explanation could also account for the shift of the Sic1 band to longer wave lengths. PREPARATION AND SOME PROPERTIES OF TRICHLOROCYANOSILANE' An unsuccessful attempt to prepare SiClnCN Sir: has been reported4; Goubeau and Reyhing exam'l'reatnient of mercury(I1) cyanide with disilicon ined several metathetic reactions involvilig various hexachloride liquid or vapor a t approximately tetravalent silicon halides and differelit group I 100" results in a volatile, colorless liquid, melting cyanides. On the basis of the failure of this prrpoint -46.2 + 0.2", which can be separated from vious attempt and the conditions of the prcscnt unchanged disilicon hexachloride by distillation preparation, a mechanism involving addition of in z'acuo through traps maintained a t -63 and cyanyl radical to the silicon-silicon bond is sug-78'. The -63" trap retains unchanged di- gested. Further investigations of the chemical properties silicon hexachloride, identified by its - 1' melting point. The -78" trap retains the colorless liquid of the new compound and its derivatives are i n progress. which exhibits these vapor pressures :
corresponding N-acetylhexosaminic acids, thus differing from previously reported mechanisms for the metabolism of N-acetylhexosamines.8 The purified enzyme preparation did not catalyze the disappearance of glucosamine, galactosamine or glucose.
1, 'C.
P,, (obs.) P,,(calcd.)
-45.2
2.3 2.25
-30.7
6.2 6.24
-22.9
10.0 10.3
00.0
10
20
37.6 37.8
62.2 62.1
101.6 99.3
The calculated values are obtained from the equation log P,, = i . 7 5 1 - ( 1 6 8 i / T ) from which a AfTvap of '7,720 calories per niolc and an extrapolated boiling point of 73.2" can be calculated. Thus the Trouton constant for this liquid is 22.2. The formula SiClaCh' was established for this compound by analysis corresponding to the formula Sil.opC12.98(CN)o.s6 and by the vapor density measurement a t 2'7.8" corresponding to an apparent molecular weight of 158.8; calculated for SiC13CN, 160.4. The new compound is stable indefinitely a t -78" in vacuo and in the vapor phase a t room temperature. I n the liquid phase at room temperature the compound undergoes a slow decomposition, producing silicon tetrachloride and nnnvolatile brown solids. Trichlorocyanosilane undergoes rapid hj-drolysis. n'ith limited aniounts of water vapor hydrogen cyanide and hexachlorosiloxane result. The water solution from complete hydrolysis gives a strong Turnbull's Blue test for CN-. The infrared absorption spcctruiii of tlic vapor shows a strong sharp peak a t 2200 cni.-l characteristic of C N stretching2n3and a moderately strong sharp peak at 2080 cm.-', previously assigned as (1) T h e authors wish gratefully t o acknowledge t h e partial support of this work b y t h e Research Corporation under a Frederick Gardner Cottrell G r a n t . (2) H. R . Linton a n d E. R . Xixon, Speclrochim. Acln, 10, 299 (19.58). (3) T . A. Bither, W. H . K n o t h , R . V. Lindsay, Jr.. and U'. €1. Sharkey, THIS J O U R N A L , 80, 4151 (1968).
(4) J. Goubeau a n d J. Reyhing, 2. anorg. u . o i l g e m . C h e m . , 294, 96 (1958).
DEPARTMEST OF CHEMISTRY PURDUE USIVERSITY ALEXANDERKACZMARCZYK LAFAYETTE, INDIAXA GRAST UKRV RECEIVED JUNE 10, 1959
SIMPLE
SYNTHESES OF PYRIMIDINE-Z'-DEOXYRIBONUCLEOSIDES'
Sir : Recent studies with 5-fluoro-2'-clcoxyuridinc (P-FUDR) and j-fluoro-Z'-deoxycytidine (P-FCDK) have demonstrated their usefulness as anti-tumor agents in several experimental tumors2f3and in clinical trial^.^ P-FUDR was prepared5 by enzymic procedures while P-FCDR was synthesized6 from P-FUDR. I n view of the need for Fi-fluorinated-2'-deoxynucleosides, we report the total syntheses of pyrimidine-2'deoxyribonucleosides by the mercuri p r o c e d ~ r e . ~It~ was ~ found that crys(1) T h i s investigation was supported in p a r t ( t o t h e Sloan-Ketterinp Institute) b y f u n d s from t h e National Cancer Institute a n d t h e National I n s t i t u t e s of Health, Public Health Service (Grant No. CY3190). (2) C . Heidelberger, L. Griesbach, 0 C r u z , R . J . Srhnitzer a n d E
