2712
H. E. ST.IVELY, 0 . WINTERSTEINER, J . FRIED, H . L. WHITEAND 11. MOORE
[CON~RIBUTION FROM
THE
Vol. 69
SQUIBB INSTITICTE FOR MEDICAL RESEARCH, OF OKCAUIC CIIEMISTRY~ DIVISION
Streptomycin. VI. Some Derivatives and Reactions of Dihydrostreptobiosamine BY H . E. ST\VEI,Y, 0.!,YINTI:RSTEINER, J. FRIED,H. L. WHITEAND M. MOORE
Dihydrostreptomycin' ' is cleaved by inethanolic hydrogen chloride into streptidine dihydrochloride and methyl dihydrostreptobiosaminide hydrochloride. The latter compound is an amorphous product which yields on acetylation two crystalline pentaacetates differing by anomerism a t the hrmiacetalic carbon atom 1 of the streptose moiety ( a-form,2b3,4,5 /3-form5). Another glycosidic derivative of dihydrostreptobiosamine is the amorphous ethyl thiodihydrostreptobiosaminide hydrochloride,6 which has been characterized by the preparation of crystzlline N-acetyl6 and pentaacety17 derivatives. The structure of dihydrostreptobiosamine (11) is implied in that of streptobiosamine (I) recently established by Kuehl, Flynn, Brink and Folkers.*
appear that the a- and /3-hexaacetates are anomers differing from each other by stereoisomerism a t C1 of the dihydrostreptose moiety, and that they correspond sterically to the a- and @-formsof methyl pentaacetyldihydrostreptobiosaininide,j the specific rotations of which are -117' and -&lo, respectively. Since streptose has been shown to be an L-sugar: this assignment of prefixes is in accord with the established conventions for designating anomers on the basis of their specific rotations. The nature of the highly levorotatory y-hexaacetate remains to be elucidated. There can be little doubt that it represents a definite, homogeneous entity, as the properties, especially the rotation values, of preparations from different runs showed excelrent agreement, and remained - -CHOH r-CHOH CHunchanged on chromatographing. The I I I I CH-0--K 1 CH--O-R analytical data for nitrogen, total and 0CHaNH.CH 0 i,OH 0 1 OH l acetyl indicate that it is a hexaacetate deI R = O rived from dihydrostreptobiosamine, but Y'CHO I the carbon values were consistently 0.5--CH HOCH -- CH $CH*OH I 0.8% too high for a compound isomeric I I I- --__ CHI CHB CH with the a- and p-hexaacetates. Furthermore, unlike the latter, it gives evidence I I1 CHzOH I of beinz unsaturated (tetranitromethane. That the primary alcoholic group in I1 is derived permanganate7 and decomposes with pigment for: from the free aldehyde group in I follows from the mation on short heating a t 100' in the dry state conversion of a-methyl streptobiosatninide di-' or in aqueous solution. This instability is not due methyl acetal hydrochloride to methyl pentaace- to the presence of a free aldehyde group, since tyldihydrostreptobiosaminide by selective hy- the Schiff test, as well as the more reliable spectrodrolysis of the dimethyl acetal group, catalytic photometric test with thiosemicarbazide,'O by reduction of the free aldehyde group thus liber- which the preformed aldehyde group in streptoated, and acetylation.* .3 mycin and streptobiosamine can be readily demonIn the present communication we describe a strated, was completely negative. The compound few additional derivatives of dihydrostreptobios- is being further investigated. amine which were obtained in the course of early By hydrolysis with boiling water the a- and ,i3structure studies on this entity. isomers of hexaacetyldihydrostreptobiosamine are Acetylation of dihydrostreptobiosamine hydro- both transformed, by loss of the acetyl group a t chloride with acetic anhydride and pyridine carbon atom 1 of the dihydrostreptose portion, yielded as the main product a hexaacetate m. p. into an amorphous product which appears to be l U O , [ a ] D -10So, designated a , and smaller essentially an equilibrium mixture of anomeric amounts of two other hexaacetates designated /3 pentacetates. Nevertheless the proportion of an(m. p. 1W0, [ a ] -36.2O) ~ and y (m. p. 13(io, omers in the crystalline products secured from [.ID -175'). From the optical data it would these mixtures varied in dependence on the isomer serving as the starting product. The experiment (1) R . L. Peck, C. E. Hoffhine and K. Frilkers, THIS JOURNAL, 68, 1390 (1946). starting from the a-hexaacetate yielded a penta(2) Q. R. Bartz, J. Controulis, H. M. Crooks, Jr.: and M. C. acetate m. p. 140-142' which, to judge from its Rebstock, i b i d . , 68, 2163 (1946). rotation (- 98O) and the fact that it could be re(3) J. Fried and 0. Wintersteiner, i b i d . , 69, 70 (1947). acetylated in good yield to the a-hexaacetate, is ( 4 , I . R . Hooper, L. H. Klemm, W.J. Polglase and M. I.. Wolfrom, ibid., 68, 2120 (1946). probably substantially pure a-pentaacetyldihy( 5 ) PIT. G. Brink, F. A . Kuehl, Jr., E. H . Flvnn and K. Folkers, drostreptobiosamine. On the other hand, the ibid , 68, 2657 (1946). crystalline pentaacetate fractions derived from the ( 6 ) F. A. Kuehl, Jr.. E. H. Flynn, N. G. Brink and K. Fdkers, a-hexaacetate were anomeric mixtures having spei h i d , 68, 2006 (3940). L
I
(7) R U. Lemieux. W. J . Polglase, C. W. I)c Walt and M . I-. Wolfrom, i W . , 68, 2747 (1946). ( 8 ) F. A. Xuehl, Jr., E. H. Flynn, N. G. k i n k and K. Folkers. i b i d . . 68, 26% (1946).
( 9 ) J. Fried, D. E. Walz and 0. Wintersteiner, ibid., 68, 2746 (19413). (IO) R. Donovick. G. Rake and J. Fried, J . B i d . Chem., 164 173 (1046).
Nov., 1917
DERIVATIVES AND REACTIONS OF DIHYDROSTREPTOBIOSAMINE
2743
cific ntations close to that attained at equilibrium tetrahydroanhydrostreptobiosamine would be repin boiling water (about -80'). A product with resented by the morpholine structure 111. similar properties was also obtained by hydrolytic CHa removal, with mercuric chloride, of the ethylthio I group of ethyl thiodihydrostreptobiosaminide pentaacetate? The latter reaction demonstrates that the acetyl group eliminated from the anomeric hexaacetates by hydrolysis with water is indeed the one situated at carbon atom 1 of the dihydrostreptose moiety. The ethyl thiodihydrostreptobiosaminide pentaacetate used in the above reaction was prepared by treating dihydrostreptobiosamine hydrochlo0 CH? 1 ride with concentrated hydrochloric acid and I AcOCH ethyl mercaptan and acetylating the resulting I crude product, whereas Lemieux, et aLI7obtained 1 CH it by acetylation of the amorphous but pu,.ified I ethyl thiodihydrostreptobiosaminide hydrochlo111 CH~OAC ride6 resulting from the mercaptolysis of dihydroPotentiometric titration of the pentaacetyl streptomycin with ethyl mercaptan and anhydrous hydrogen chloride. While in the latter case the derivative revealed that the acid-binding group formation of an ethyl thioglycoside might be ex- has a much lower basicity (PKa' 4.6) than would pected, it is sornewhat surprising that the same be expected from the tertiary amino group in 111. product was obtained from free dihydrostrepto- However, the influence of environmental factors, biosamine, beca.use aldo-pentoses and -hexoses that is of steric hindrance or of inductive effects generally react with ethyl mercaptan in the pres- exerted by the neighboring acetoxy groups, canence of aqueous hydrochloric acid to yield the di- not be predicted, so that this fact per se does not ethyl thioacetak derived from their aldehydo necessarily argue against the above structure. forms. Evidently the furanose ring in the strep- More surprising was the observation that drastic tose moiety is more stable under these conditi0n.s acid hydrolysis, mder conditions which are known than is the case in other furanoses, for instance in to cleave the glycosidic linkage between the two 1,2-isopropylidene-S-desoxy-L-arabinose,which moieties of dihydrostreptobiosamine, failed to liberate the reducing group of the N-methyl-L-glureadily forms the diethyl thioacetal. IYhile pentaacetyldihydrostreptobiosamine in cosamine portion. That no extensive hydrolysis aqueous solution fails to take up hydrogen in the of any kind had occurred was evidenced by the prescnce of platinum oxide, dihydrostreptobios- recovery of some of the pentaacetyl derivative on amine hydrochloride is reduced under these condi- acetylation of the acid-treated product. It would tions with the consumption of one mole of hydro- then have to be assumed that the acetalic linkages gen. Acetylation of the reduction product af- incorporated in the morpholine and pyranose rings of I11 are unusually resistant to acid. However, forded a crystalline compound m. p. 154-155' which, however, was not the expected heptaacetyl- this is not without precedent in anhydro structetrahydrostreptobiosamine, but a base hydrochlo- tures of this type. A similar stability toward ride of the composition Cz3HzoOsN(CH3CO)s. acid hydrolysis is exhibited by difructopyranose HCl.C2HiOH. The mole of ethanol, which cannot anhydride12 and difructofuranose anhydridei3 in bc removed by drying at 80°, is derived from the which the reducing groups of the two fructose resisol\.ent used in the crystallization and is probably dues are parts of a dioxane ring as well as of the iiot an integral part of the molecule. Otherwise pyranose or furanose rings. The hydrochloride, m. p. l55', can be obtained the change produced by the reduction correspmds to the addition of 2 hydrogen atoms and in good yield from streptobiosamine hydrochlothe loss of one molecule of water. Accordingly the ride directly in the same manner as from dihydrostreptobiosamine. The requisite two moles of term tetrnhydronnhydrostreptobiosamine is pro- hydrogen are consumed without a discernible posed for the constituting free base. The un- break in the hydrogenation curve after the absorpacetylated product, in contradistinction to dihy- tion of the first mole. This result tends to confirm drostreptobiosamine, does not reduce Fehling the supposition that the aldehydic group of the solution. This fact, together with the retention dihydrostreptose moiety is involved in the addiof basic properties and the lack of an N-acetyl tion of the second mole, because when this group group after acetylation, suggests that the hydro- is blocked by glycoside formation, as in inethyl genation effects a reductive condensation of the streptobiosaminide or streptomycin, the retluctioii altlchydc group df the dihydrostreptose moiety stops at the stage of thc dihydro derivati\-cs. with the methylimino group, so that pentaacetyl(13) I I . 11. Schlubach and C . Behre, A7rn., 508, l ( i (193.4).
j
( 1 1 ) P. A. Levene and J. Compton,
(1936).
J. Bioi. Chcm., 116, 189
(13) W. N. Haworth and H. R. L. Streight, Helu. Chim. Acla, 16, 693 (1932).
2744
H. E. STAVELY, 0. WINTERSTEINER, J. FRIED, H. L. WHITEAND M. MOORE
further requirement for the occurrence of the second reduction step is the presence of the imino hydrogen atom, as is evident from the unreactivity of pentaacetyldihydrostreptobiosarnine, in which the nitrogen atom instead of the aldehydic group is blocked. In experiments designed to convert a-methyl pentaacetyldihydrostreptobiosdrninide to u-hexacetyldihydrostreptobiosdmine by the acetolysis procedure of Hann and Hudson,l* the methyl glycoside was treated with a mixture of acetic anhydride and acetic acid containing catalytic amounts of sulfuric acid. The ensuing reaction was accompanied by a fall of the levorotation to low levels. The mixture was worked up in the usual manner (bicarbonate-chloroform) . The crystalline material extractable with chloroform consisted for the most part of N-methyl-L-glucos:miine Npentaacetate. From the amount of material recovered in this fraction i t appeared that the diwccharide had been nearly quantitatively cleaved by acetolysis into the component sugars. Since the dihydrostreptose portion apparently had remained in the aqueous phase, the procedure of working up the reaction mixture was modified in various ways so as to render feasible the isolation of this moiety. These attempts were unsuccessful, largely on account of the difficulty encountered in separating the water-soluble products derived from the starting product from the sulfoacetic acid, SOsHCHzCOOH, which is formed almost qu;iiititatively from the sulfuric acid present in thr. xcetolysis mixture by reaction with acetic anh\-tlritlc,l5 and which inconveniently forms a water-soluble barium salt. % hen ' acid conditions prevailed during the decomposition of the acetic anhydride in the acetolysis mixture, the yield of chloroform-extractable material was generally much lower. Reacetylation of the water-soluble products in two such experiments yielded additional chloroform-soluble material from which, by chrc)iiiatographing, the hitherto undescribed Nriiethyl-L-glucosamine P-pentaacctate was isolated. I t appears that the acids present in the aqueous phase effected the hydrolytic removal of at least the 1-acetyl group in this moiet? as well as the subsequent inversion to the @-form. I t should be mentioned that the /3-pentaacetate i \ alsi) iorrned, besides the preponderant u-isoI I I ~ T , ' in ~ the acetylation of the crude N-methyl[.-glucosaniinc resulting from the reduction of synthetic S-methyl-L-glucosaminic acid lact.one.i6,Ifia Finally we describe in the experimental part a crystalline, high-melting product of unknown f I I J R \I. I i d n n a n d C . S. Iiudson, T H r s J O U R S A I . , 56,. 2465 \ 1!l34). (1.3) I'rru:hiniout, Cumpl. r e n d . , 21, 1054 (1881); R e c lruu. c h i r n . ,
7, 2 5 ( I d b h i , T . P. SIurray a n d IV. 0. Keuyori, 1 2 3 0 (1346,.
Tiiii J O C R S A L ,
61,
Vol. 69
structure which was obtained in early attempts to prepare an acetone derivative of methyl dihydrostxeptobiosaminide. The method used was that of Dangschat and Fischerl' by which these authors succeeded in preparing monoisopropylidene-mesoinositol. The oily product obtained by this procedure from methyl dihydrostrep tobiosaminide hydrochloride was acetylated and chromatographed. The isolated crystalline substance (rn. p. 289' dec.) lacks the methoxyl group of the starting product, but is undoubtedly a derivative of dihydrostreptobiosamine and not a fragment. The analytical data are not quite conclusive, but fit best for a dimer C42H64021N2 containing two isopropylidene groups and five acetyl groups, one of which is bound to nitrogen. Experimental Acetylation of Dihydrostreptobiosamine Hydrochloride. -The disaccharide was prepared by hydrolysis of dihydrostreptomycin trihydrochloride with 1 Ssulfuric acid a t 45' as previously d e ~ c r i b e d ,except ~ that in later runs the hydrolysis period was shortened t o fifteen hours. Dihydrostreptobiosamine hydrochloride ( 4 9.) w t h shaken with acetic anhydride (8 cc.) and pyridine (2 cc.) until i t was dissolved (three hours) and the solution was then kept at 45" overnight. The solvents were removed in vacuo, and the residue was extracted three times with benzene. The benzene solution was concentrated to about 20 cc. and passed through a column of aluminum oxide 1 X 20 cm. The column was washed with benzene and then with benzene-ether 7:3. On trituration of the residues with ether crystallization occurred in the later benzene atid most of the benzene-ether eluates. The crystalline material was purified by treating it in acetone solution with norit, and then by recrystallization from acetone-ether or ethyl acetate-hexane. The a-hexaacetate thus obtained forms rosets of prisms m. p. 143-144°,18 [ a ] 2 a D -108O ( c , 1.0 in chloroform). Yield was 0.8-1.0 g. Anal. Calcd. for Cl3Hi8O9N(CH3C0)6: C, 50.76; H, 6.30; N , 2.37; GCH3C0, 43.7; 0-Acetyl, 36.4. Found: C, 50.80; H , 6.58; N, 2.44; CH,CO, 42.30; 0-Acetyl, 36.0. The 6- and y-hexaacetates were obtained only after milder conditions for the acetylation had been adopted. Though a part of the acetylated material then crystallized directly, these crystals proved t o be a mixture from which the preponderant a-isomer could not be obtained in pure form by fractional crystallization. Chromatographing therefore not be dispensed with for securing the pure isomers. To dihydrostreptobiosamine hydrochloride (4.8 g.) pyridine (:C3 CC.) and acetic anhydride (19 CC.) were added. The mixture was placed into the refrigerator and shaken ocr~tsioti:illy. After about one hour all the material except it small amount of n d:irk brown sediment wiis diizolved. After forty-six hours of stnridiiig a t 4" the solvents were removed in z'ucuo. The residue was takrri u p in chloroforiri, and the filtered solution after chilling !\as washed successively with ice-cold water, ciilutv hytlrocshloric acitl, sodium bicarbonate solution atid water. The rehidue obt;rincd on evaporation of the dried chloroform 5olutiori (7.1 g.) was dissolved i n ethyl acetate ( 3 . 5 cc.). The crystalline product formed on addition of ether ( , 3 J c c . ) and seeding w i t h thc a-hexaacetate was collected :mtl w.ihc(1 with ethvr (2.17 g., m . p . 13i-11flo). After s:veraI rrc.ryit;tlli~atir,i~~ from ethyl acet:ite-c~licr the I)ro(luit , t i ) jiidge from apptxr;iiicc ant1 phyyic;il i)rop.rti
