6462
COMMUNICATIONS TO THE EDITOR
Vol. so
mine water oxidation and hydrolysis, gave D- B and C are identical, their isolation and nonisoserine' (positive rotation,' isolated and identified identity allow the assignment of chemicalI6 and from papergram) from C-4, C-5 and C-6 of neos- antibacterial16.l9 differences of the two antibiotics amine C. These data assign neosamine C the same to the neosamines. stereochemistry as that of D-glucosamine a t C-2 This investigation was supported in part by a and C-5. research grant, KO.E-1278, from the National It has been shown for a number of hexoses that Institute of Allergy and Infectious Diseases, Public the replacement of a hydroxyl group by an amino Health Service. We also wish to express our group does not alter appreciably the magnitude of thanks to the Upjohn Company for the generous their specific rotation^.^,^ The rotation of neos- gift of neomycin samples. amine C dihydrochloride, [ c Y ] ~ ~+07", D thus sug(19) 0.K.Sebek, J . Bncfcriol., 1 6 , 199 (1958). (20) Robert F. Carr Fellow, 1957-1958. gests that it probably has the stereochemistry of D-glucose (D-glucosamine hydrochloride, [cy ] 2 0 ~ DEPARTMENT OF CHEMISTRY +73°)10 or, somewhat less likely, that of D-galaCtOSe AND CHEMICAL ENGINEERINGKENNETHL. RINEIIART,JR. OF ILLINOIS PETERW. K. Wooao (D-galactosamine hydrochloride, [q(]20D +96O),lo UNIVERSITY ILLINOIS ALEXANDER D. ARGOUDELIS but not that of D-gulose (D-gUlOSar" hydrochlo- URBANA, RECEIVED SEPTEMBER 9, 1958 ride, [ a I 2 l D - 19°)9 or D-allose ( [CY]~'D +14O),I1 the two other aldohexoses allowed by the stereochemical assignments a t C-2 and C-5. INTRAMOLECULAR ELECTRON EXCHANGE I N ANIONS OF The high positive rotation of neobiosamine C PARACYCLOPHANES (dihydrochloride: [MID +33,700)' s ~ g g e s t s ' ~ ~ ~ ~ an cy-D-glycosidic link between neosamine C Sir: As Cram and his co-workers have demonstrated, (dihydrochloride: [AI ID +16,800) and D-ribose ([hI]D -3,450),3an assignment strengthened by an the paracyclophanes are well suited for studies of infrared band a t 544 cm.-I in the spectrum of N,N'- inter-ring interactions. I have examined the dibenzoylneobiosaminol C , attributed14 to equa- anions of the [1.8], [ 2 . 2 ] , [3.4], [4.4] and [6.6] torial (p) anonieric C--H deformation. The most paracyclophanes for inter-ring electron exchange probable stereochemistry of neobiosamine C is by the method of electron magnetic resonance. The hyperfine structure in the magnetic resonance then represented by I.3a Xcic! hydrolysis of methyl neobiosaminide B1 absorption of a paracyclophane anion appears to gave hygroscopic neosamine B dihydrochloride, result from coupling with the nuclei in one ring [aIz7D + 1 7 O (c 0.92, water)15 [,4nd. Found: c, only when the exchange is slow, and with the nuclei 25.41; H, 6.61; N, 10.761, which gave positive in both rings when the exchange is rapid. "Slow" reactions with ninhydrin, aniline acid phthalate and "rapid" are taken relative to the frequency of the hyperfine interaction. In the intermediate and Fehling solution. The isomeric neosamines B and C represent to range the nature of the spectrum depends on the our knowledge the first two dianiinohexoses iso- kinetics of the exchange. The paracyclophane anions were prepared by lated, though mono-aminohexoses have been the subject of extensive recent investigation^^.^^^^^ reduction of solutions in 1,2-dimethoxyethane with and the related antibiotic kanamycin has been alkali metals. The [t.S] and [ 2 . 2 ] compounds are shown to contain both B-amino-6-deoxy-~-g~ucosereduced in good yield a t room temperature by and 3-amino-3-deoxy-~-glucose. l8 Since other por- potassium. The [3.4] and [4.4] and [6.6] comtions (neamine, D-ribose) of the isomeric neomycins pounds on the other hand are appreciably (but incompletely) reduced only a t temperatures below (7) K.Freudenberg, B c Y . ,47, 2027 (1914). -50'. At room temperature only a blue diamag( 8 ) P. W. Kent and hl. W , Whitehouse, "Biochemistry of the netic solution of rubidium in the solvent is found. Aminosugars." Butterworths Publications, Ltd., London, 1955,p. 202. As the solution is cooled the blue color is replaced (9) E. E. van Tamelen, J. R. Dyer, H. E. Carter, J. V. Pierce and E. E. Daniels, TEISJOURNAL, 78,4817 (1956). by the green color of the anions. The reaction is (IO) Kef. (81, p. 171. reversible, the blue color returning when the solution (11) "The hlerck Index of Chemicals and Drugs," 6th ed., Merck is warmed. The cycle may be repeated many times. and Co., Inc., Rahway, AT.J., 1952,p. 34. The magnetic resonance spectra of the anions 31, 66 (1909); 58, 1566 (1916); (12) C. S. Hudson, THE JOURNAL, 46,483 (1924). indicate that the unpaired electron is found almost (13) A. Neubergcr and R . V. Pitt-Rivers, Biochcm. J . , S3, 1580 exclusively a t the unsubstituted ring positions in (1939); 6.also ref. (S), p. 59. accord with earlier observations of the anions of (14) S. A. Barker. E. J. Bourne, M. Stacey and D. H. Whiffen, toluene and p - ~ y l e n e . ~[l.S]- has a spectrum of J . Chrm. Soc., 171 (1954); S. A. Barker, E. J. Bourne and D. H. Whiffen,M r f h o d i ojBiochcm. Anal., S , 213 (1956). (15) A similar compound, [u]n +18", has been obtained[ M.-M. Janot, H. PCnau, D. Van Stolk, G. Hagemann and L. Penasse, Bull. soc. chim. Froncc, 1458 (1954)1 as a degradation product of the antibiotic frsmycetin (Soframycin) by a procedure exactly paralleling that employed in the present studies. The nearly identical properties of tbir and other degradation products1***"" of framycetin (hydrochloride: [uln +579 suggest its identity with neomycin B (hydrochloride: [m]D +63O," +5401). (16) J. H. Ford, ?*I. E. Bergy, A. A. Brook.. E. R. Garrett, J. Alberti, J. R. Dyer and H.E. Carter, TEISJOURNAL, 77,5311 (1955). (17) R.Kuhn. r l ol.. Anrtw. Chrm., 69, 23 (1957). (18) (a) M.J . Cron. 0. B. Pardig, D. L. Johnson, D. F. Whitehead, I. R. Hoopcr and R . U. Lemieux, THIS JOURNAL. 80, 2342 (1958); (b) ibid., 80,4741 (1958).
(1) This work has been supported by the U. S. Air Force through the Air Force Office of Scientific Research of the Air Research and Development Command under Contract AF 49-638-464. Reproduction in whole or in part is permitted for any purpose of the U. S. Government. (2) D. J. Cram and H. Steinberg. THISJOURNAL, 73, 5691 (1951); D. J. Cram and N. L. Allinger, ibid., 7 6 , 726, 2362 (1954); J. Abell and D. J. Cram, i b i d . , 7 6 , 4406 (1954); D. J. Cram N. L. Allinger and H.Steinberg, ibid., 76, 6132 (1954); D. J. Cram and J. Abcll, ibid.. 77, 1179 (1955); D. J. Cram and R. W. Hierstod. {bid., 71, 1186 (1955); D. J. Cram and N. L. Allinger, i b i d , 7 7 , 6289 (1953); D. J Cram snd R. A. Reeves, ibrd., 80, 3094 (1958); K. C. Dewhirst and D. J. Cram, ibid.. 80, 3113 (1958). (3) T. R. Tuttle, Jr., and S. I. Weiwman, ibad., 80, 5342 (1958).
COMMUNICATIONS TO THE EDITOR
Dec. 5, 1958
6463
nine evenly spaced lines with approximately binomial intensity distribution. The interval between lines is 2.7 oersteds. [2.2]- has a poorly resolved spectrum of a t least nine components. [4.4]and [6.6]- have indistinguishable spectra of five evenly spaced lines with binomial intensity distribution. The lines are completely resolved and are separated by 5.5 oersteds. [3.4]- has a spectrum resembling those of [4.4]- and [6.6]- but with wider lines and broad tails extending beyond the 22 oersted spread of the former. The spectra may be interpreted as follows: In [4.4]- and [6.6]- the electron exchange proceeds less frequently than 3 X IO6 see.-' (the line breadth in frequency), while in [1.8]- and [2.2]- it proceeds more rapidly than 1.5 X lo7set.-' I n [3.4] the exchange rate is somewhat higher than 3 x lo6 sec.' but less than 1.5 X lo7set.-'. Further work on other paracyclophanes and on the temperature dependence of the rates is in progress. I am deeply indebted t o Professor Cram for all the paracyclophanes which have been used in these experiments.
periodate with formation of no formaldehyde. The periodate product was oxidized with bromine water, then hydrolyzed vigorously, t o give glycine [papergram-violet ninhydrin spot, Rr 0.445 ( P Y W ) , ~0.375 ( P h e r ~ W ) ~authentic ; sampleviolet, Rf 0.446 (PyW), 0.363 (PhenW)] and serine [purple spot, Rf 0.531 (PyW), 0.305 (PhenW); authentic sample-purple, Rf 0.528 (PyW), 0.286 (PhenW)]. Methyl N,N'-dibenzoylneobiosaminide C(I1) consumes two moles of periodate, thus must have one -CHOHCHOHgrouping in each monosaccharide (ribose and neosamine) moiety. This periodate oxidation product was treated with bromine water, then hydrolyzed vigorously to give isoserine [papergram-yellow ninhydrin spot, Rf 0.43 (PyW), 0.27 (PhenW); authentic sampleyellow, Rr 0.42 (PyW), 0.26 (PhenW)], as well as glycine [violet spot, Rr 0.39 (PyW), 0.34 (PhenW) ; authentic sample-violet, Rr 0.39 (PyW) 0.33 (PhenW)]. The formation of isoserine establishes an aldohexose (rather than a ketohexose) structure for neosamine C, while demonstration of a vicinal glycol grouping in this moiety of I1 shows neosamine C to contain a pyranose (rather than DEPARTMENT OF CHEMISTRY furanose) ring in I. The structure of neosamine C S. I. WEISSMAN in neomycin C is, therefore, that shown in I. WASHINGTON UNIVERSITY SAINT LOUIS, MISSOURI Methyl neobiosaminide C1 (111) was hydrolyzed RECEIVED OCTOBER 4, 1958 under mild conditions to I, which was reduced with sodium borohydride, then N-benzoylated t o give N, N '-dibenzoylneobiosamin ol C, m .p. 2 17-227 ' CHEMISTRY OF THE NEOMYCINS. 111. THE STRUCTURE OF dec. [Found: C, 56.12; H, 6.31; N, 5.101. VigorNEOBIOSAMINE C ous hydrolysis of this derivative gave ribitol Sir : [papergram-colorless with aniline acid ~ h t h a l a t e , ~ It was reported recently1.2 that neobiosamine C, Rf 0.616 (BAW 221)4; authentic sample--colorless, approximately one-half of neomycin C, is a di- R, 0.612 (BAW 221)], thus confirming the earlier saccharide composed of a diaminohexose (for which assignment' of neobiosamine C as a neosaminidowe now propose the name neosamine C, cf. accom- ribose, rather than a ribosidoneosamine, which panying Communication) and D-ribose. We pre- would have given ribose [authentic sample-red sent here evidence showing neobiosamine C to have spot, Rf0.584 (BAW 221)]. N,N'-Dibenzoylneothe structure I. biosaminol C consumed three moles of sodium periodate, with formation of one mole of acid. 6C:H2NHR' 'CH'OH A glycosidic bond to the ribose C-5 position would I require a four-mole periodate uptake, with formaH-'C -KHCOChH,5 tion of two moles of acid. The ribose C-3 position I HO H0eC-H is eliminated per se by the uptake of one mole of Neosamine C NHR,\o I ' I periodate in the ribose moiety of methyl N,N'H-4c - OH dibenzoylneobiosaminide C (cf. above) , (impossible I with a C-3 link since two adjacent hydroxyl groups H-'C -OH are prohibited for either pyranose or furanose I forms). The ribose C-4 position is also extremely 'CHsNHCOCoH5 unlikely, since oxidation of neobiosamine with IV bromine water gave a product whose infrared D -Ribose spectrum contains a carbonyl band a t 1765 cm.-l, I, R R' = If (Neobiosamine C) indicative of a y-lactone,e hence of a free C-4 11, R CHI, R = coc&b hydroxyl. The sole remaining ribose position is 111, R = CHs, R' = H C-2 and the structure of neobiosamine C is I. Cleavage of the glycosidic linkage in methyl This investigation was supported in part by a N,N'-dibenzoylneobiosaminide C' (11) by mild research grant, No. E-1278, from the National acidic hydrolysis, followed by sodium borohydride Institute of Allergy and Infectious Diseases, Public reduction, gave N,N'-dibenzoylneosaminol C (IV), Health Service. We also wish to express our m.p. 186-1S8.5'. [Found: C, 61.80; H, 6.23; thanks t o the Upjohn Company for the generous N, 7.281, which reduced two moles of sodium (4) P y W is 97.6 parts pyridine $52.5 parts water. PhenW is 0
(1) K. L. Rinehart, Jr., P. W. K . Woo. A. D. Argoudelis and A. M. Gicsbrecht, THIS JOURNAL, 79, 4667 (1957). (2) K . L. Rinehart, Jr., P. W. K.Woo and A. D. Argoudelis, ibid., 79, 4568 (1957). (3) K. L. Rinehart, Jr., P. W. K.Woo and A. D. Argoudelis, ibid., 80, 6461 (1958).
phenol saturated with an aqueous solution containing 3.7% sodium dihydrogen phosphate and 6.3% sodium citrate. BAW 221 is t r r t buty1alcohol:acetic acid:water:2:2:1. (5) S. M. Partridge, Nalutc, 164, 443 (1949). (6) S. A. Barker, E. J. Bourne, R . M. Pinkard and D. H. Whiffen, Chcm. and I n d . (London), 658 (1958).
