Sept. 5 , 1958
COMMUNICATIONS TO THE EDITOR
4745
proceeds with inversion of carbon-3 to form 1,2,5,6di - 0 - isopropylidene - 3 - deoxy - 3 - amino - D - allose (111). Compound I11 can be prepared in 83% yield by hydrogenolysis of the 1,2 ;5,6-di-O-isopropyhdene3-deoxy-3-hydrazino-~-hexose(IV) of Freudenberg and Braunsj in ethanol a t 80' using Raney DEPARTMENT OF INORGANIC & PHYSICAL CHEMISTRY A. I(.HOLLIDAY nickel catalyst and 40 p.s.i. of hydrogen. The adTHEUNIVERSITY OF LIVERPOOL vantage of this route is that, whereas the aminoLIVERPOOL, ENGLAND A. G. MASSEY lysis of I1 proceeds in 16Yc yield,3 the hydrazinolRECEIVED JULY 14, 1958 ysis proceeds in 60% yield.6 Acid hydrolysis of I11 produced an aminosugar of different paperT H E ACTION OF POTASSIUM AMIDE IN AMMONIA chromatographic properties than that obtained on O N O-CHLOROPHENYLACETONE: CORRECTION.' the hydrolysis of methyl 4,6-benzylidene-3-deoxySir: 3-amino-a-D-glucoside diacetate. The X-acetyl derivative of I11 (V), m.p. 127We recently reported a new principle of ring ~ (c, 2 in chloroform), [calcd. for closure, the essence of which is the intramolecular 128", [ a ] +71.3" addition of a nucleophilic center to a benzyne C11H2306N:C, 55.80; H I 7.69; N, 4.65. Found: C, 55.87; H, 7.73; N, 4.85.1 was hydrolyzed for structure.2 Amongst examples of the new principle, the con- two hours a t 100" in 0.2 N hydrochloric acid. Nversion of o-chlorophenylacetone t o indan-2-one) Acetylation was then accomplished by the addition through the action of potassium amide in liquid of acetic anhydride to the neutralized hydrolyzate. ammonia, was described. We have now dis- Evaporation of the solvent and extraction of the covered an error: the compound reported a s indan- residue with ethanol gave a 1,2-0-isopropylidene(VI), m.p. 1542-one (m.p, 60-61 ") is actually 2-methylindole 3-deoxy-3-acetamido-a-~-hexose (m.p. The compound obtained, in 25% 156". Calcd. for Cl1H19OeN: C, 50.56; H, 7.33; yield in recent experiments, strongly depresses N, 5.36. Found: C, 50.25; H, 7.44; N, 5.29. Periodate oxidation of VI, reduction of the prodthe mixed melting point with authentic indan-2one, but does not alter the mixed melting point uct with sodium borohydride and acetylation of with authentic 2-methylindole. Also, the com- the reduced product gave a crystalline substance pound gives a positive sodium fusion test for'nitro- which was hydrolyzed to aminosugar using N gen. All our efforts to isolate authentic indan-2- hydrochloric acid a t 100'. On evaporation, a one from the reaction in question have been fruit- crystalline product was isolated whose infrared spectrum (KBr disc) and X-ray powder diagram less. It is of theoretical interest that 2-methylindole were identical t o those obtained with an authentic is formed in this reaction; the matter will be dis- sample of 3-deoxy-3-amino-~-ribosehydrochloride (VII).6 These results establish the allo-configuracussed in a future publication. tion for 111-VI and provide a new synthesis of VI1 (1) Research supported in part by the Office of Ordnance Research, U. S. Army. which is a constituent of the antibiotic p u r ~ m y c i n . ~ (2) B. P. Hrutfiord and J. F. Bunnett, THISJOURNAL, 80, 2021 Acknowledgments.-The authors wish to (1958). thank Dr. M. Przybylska for the X-ray analyses. (3) H.D.Porter and C. M. Suter, ibid., 67, 2022 (1935).
continuous nature of the discharge, may account for the increased rate of production of diboron tetrachloride in a d.c. arc. Work is in progress to increase the yield still further, and an investigation of electrode materials other than mercury is planned.
(4) L. Marion and
C.W. Oldfield, Can. J. Research, BSB, 1 (1947).
( 5 ) K. Freudenberg and F. Brauns, Bbr., 6 6 , 3233 (1922).
VENABLECHEMICAL LABORATORY BJORN F. HRUTFIORD ( 6 ) B.R. Baker and R. E. Schaub, J. Org. Chem., 19,648(1954). (7) P . W. Fryth, C.W. Waller, B.L. Hutchmgs and J. H. Williams, UNIVERSITY OF NORTH CAROLINA CHAPELHILL,N. C. J. F. BUNNETT THISJOURNAL, 80,2736 (1958). RECEIVED AUGUST4, 1958 DEPARTMENT OF CHEMISTRY UNIVERSITY OF OTTAWA R . U. LEMIEUX OTTAWA,CANADA P A U L CHU 1,2 ;5,6-DI-O-ISOPROPYLIDENE3-DEOXY-3-AMINORECEIVED JULY 21, 1958 ~-D-ALLOSE
Sir: Agf (CYCLOOCTATETRAENE) Hydrolysis of the new antibiotic kanamycin has STRUCTURE OF THECOhlPLEX been found' to yield 3-deoxy-3-amino-~-glucose Sir: (I).2 Complexes between metal ions and organic n The crystalline amine which is formed3 on the aminolysis of 1,2;5,6-di-O-isopropylidene a-D-glUC0- bonding systems have been of interest for some and are germane to general discussions8 of furanose tosylate (11) has been c h a r a c t e r i ~ e d ~ time'sa *~ weak complexes and their r e a c t i v i t i e ~ . ~ ,Aside ~.~ as 1,2;5,6-di-O-isopropylidene-3-deoxy-3-amino-cfrom the Ag+ (Benzene) complex,6pB few such comD-glUCOSe. These reactions therefore appear to (1) S. Winstein and H. J. Lucas, THISJOURNAL, 60, 836 (1938); provide a convenient route for the preparation of I. This communication is to report that, a s sus- 79, 4339 (1957). (2) L.J. Andrewn, Chrm. Rcrr. 64, 713 (1954). pected by Cope and ShenJ4the aminolysis actually (3) R. S. Mulliken, J. Ckhsm. Phys., 19, 514 (1951); THISJOURNAL, (1) M. J. Cron, D. L. Evans, F. M. Palermiti, D. F. Whitehead, I. R . Hooper, P. Chu and R . U. Lemieux, THISJOURNAL, 80, 4741
(1968). (2) S. Peat and L.F. Wiggins, J . Chrm. SOC.,1810 (1938). (3) IC. Freudenberg, 0. Burkhart and E. Braun, Bar., 59, 714 (19261. (4)
A. C. Cope and T. Y.Shen, TEISJOURNAL, 78, 3177 (1956).
71, 600 (1950); 74,811 (1952). (4) K. E. Rundle and J. D. Corbett, ibid., 79, 757 (1957). (5) R. Rundle and J. Goring, ibid., 72, 6337 (1950); for a structural study of the styrentPdClt complex see J. Holden and N. Baenziger, i b i d . , 77, 4987 (1955). (6) H. G. Smith and R . E. Rundle, Abstracts, American Crystallographic Society Meeting, Milwaukee, Wisconsin, June 23-27 (1958).
COMMUNICATIONS TO THE EDITOR
4740
plexes have been subjected to complete structure determination, and hence we have carried out a study of Ag+(Cyclooctatetraene)NOa- by the Xray diffraction method. We anticipate a close relation between Ag+(Olefin) complexes1 and the ethylene-like Ag +(COT) complex.' The strongest bonding in the complex (Fig. 1) is between one Ag+ and one COT; these Ag7'(COT) units are then joined more weakly into infinite
A.
(11) R. E. Dicherson, P. J. Wheatley. 1'. A. Howell and W. S . I.ipscomb, J . C h e m . Phrs.. 27, 200 (1957). UNI\'ERSITY OF hf1NiXESOTA
I
so
5% and B = 91'7'. Valuesll of R = 0.113 and r = 0.056 for the 1136 observed rcflections indicate that refinement is nearly complete. We wish to thank Professor S. 11.' Fenton for suggesting this investigation, the Office of Kava1 Research for support, and the Minneapolis Honeywell Company for a fellowship to F.S.11.
Scrioor, OF CHEXISTKY \
Vol.
MINSRAPOI.IS 13, MINNESOTA
F. Sc0.r-r M A r I r E n s WILLIAM X. LIPSCOMII
RECEIVED JULY 21, 1058
A NEW ADENOSINE DINUCLEOTIDE ISOLATED FROM MUSCLE EXTRACTS' Sir:
1 H Fig. 1.-The
,
" t
'
'1
s
A dinucleotide of adenosine (XAD) which has no color a t PH 6 but which turns to green a t pH 9 has been found in muscle. The adenine base was determined by its ultraviolet spectrum (values E %80/~260:PH 2 , 0.368; PH 7 , 0.122: reference values2: PH 2, 0.375; pH 7, 0.125) and by chromatography (Table I). The inference that the other base is a pteridine is supported by its blue fluorescence, ultraviolet spectrum and the indication that it contains 4 nitrogen per mole (calculated from the Am value worked out from the determination of phosphate and ribose). The spectral characteristics, composition and chromatographic behavior of XXD and its products of acid hydrolysis (mononucleotide (XRP), nucleoside (XR) and free pteridine (X)) are giwn in Fig. I and Table I.
Ag+(COT) complex joined in infinite chains along the c axis; distances are in b.
chains along the c axis of the crystal. The closest Xg. . .O distance to a Nos- is 2.43 pi., a value between the ionic and covalent sums of radii. The closest h g . . .C distances of 2.5 h., distinctly greater than the sum (2.3 A.) of covalent radii, are sufficiently short that the bonding in COT may be perturbed. The COT molecule has the D z d tub form, as expected from earlier s t u d i e ~ , and * ~ ~ has average C=C = 1.37 and C-C = 1.46 distances, both *0.04 k . Corresponding distances are 1.31 and 1.46 ( = t O . O l A,) in a refinementlo of the COT crystal dsta, and 1.334 and 1.462 (optimistically =t;O.OOl A.) in a recent electron diffraction study. Thus we claim plausibility, but not significance, for the increased C=C distance. In the infinite Xg+(Benzene) complex6the average C-C distance is, however, 1.40 A.,essentially the same as in benzene itself, Ag. . .C distances6 of 2.50 and 2.63 A. in the .lg+(Benzene) complex are comparable with those in Fig. 1 but indicate more distortion in the &+(Benzene) complex. The unit cell is monoclinic, the space group is W1/a and there are four Ag+(COT)N03- in the cell. Parameters are a = 16.84, b = 8.94, c = (7) A . C. Cope a n d Ii. A . Hochstein, THISJOURNAL, 72, 2515 (1950). (8) H. S. Kaufman, H. M a r k and I. Fankuchen, Nature. 161, 165 (1948). (9) 0. Bastiansell, I,. Hedberg, a n d K. Hedberg, J. Chem. Phys., a7. 1311 (1957). (10) J Brrgnian, private communicatiorr.
Fig. 1.-Absorption spectra of S A D aiid X R P : 1,spectruin of XAD at pH 2 ; 2, spectrum of X R P at PH 2 ; 3, spectruni of X R P a t PH 12.
Adenine and the pteridine derivatives were isolated from the products of hydrolysis of XAD by electrophoresis (by this method adenine remains in the origin) and purified by chromatography with the solvents Is0 and Eth. The spectrum of XRP a t pH 2 shows a peak a t 256 mp (Am = 10.0 X lo3) and another a t 366 mp (Am = 15.5 X lo3). When (1) Supported by research grants from the A-ational Institute of Health, U . S. Public Health Service (Nos. A-1174 and H-1889). ( 2 ) W. E. Cohn in "htethods in Enzymology," S. P. Colowick and N. 0. Kaplan, Editors, A c ~ r ~ d e mI'ress, ~r Inc., New York, N. Y., 1957, Vol 111, p . 710.