Dec. 5 , 1958
GLYCOSIDES OF D-GALACTOSAMINE
mixture of methylated neutral sugars was applied to strips of filter paper, pretreated with water, and was resolved with solvent system F. Appropriate sections were located with guide strips, eluted with water and diluted to a suitable volume. The concentration of the solutions was determined by the o-aminodiphenyl method,'O the relation between concentration and absorbance for each sugar having been determined previously.'e The average value of three determinations was used. Hypoiodite Oxidation of the Hemicellulose .-The oxidation was carried out with a 0.1 N iodine solution buffered to a PH of 10.6.'O The amount of iodine consumed after 2.5 hr. in the dark corresponded to one reducing end-group per 62 anhydroxylose units. Determination of the Molecular Weight of the Methylated Hemicellulose .-Osmotic pressure measurements were carried out with the osmometer of Zimm and Myerson6z as improved by Stabin and I m m e r g ~ t . ~ eGel cellophane membrane@* which had never been allowed t o dry, were used, the solvent was chloroform-ethanol (9: 1 v./v.) and the temperature was 30 f 0.01'. The osmotic pressure was determined a t six different concentrations by the static method and the values of the reduced osmotic pressure, h/w, as plotted against w , were extrapolated to zero concentration (Table 111). The number-average molecular weight,
6325
a,, was calculated from the relationship an= 25,700/ @ / w ) o and the correspondin& degree of polymerization,
P,, from the equation p, = M, X n / M R , where n was the number of xylose residues present per acid side group and i V 1 ~was the molecular weight of the repeating unit of the fully methylated polysaccharide (2457). Determination of the Intrinsic Viscosity of the Hemicellulose .-Reduced viscosities (7gp/C) of the potassium salt of the hemicellulose were determined a t seven different concentrations with a Craig-Henderson" viscometer in M cupriethylenediamine. Extrapolation to zero concentration according to Hugginss4 gave an intrinsic viscosity, [7], of 0.812 dl./g. and a value of k' of 0.428. Kinetic energy corrections were negligible. According to the relationship 7, = 212[7], developed earlier for a similar polysaccharideag this would correspond to a number-average degree of polymerization of 172. Estimation of the Polymolecularity of the Hemicellulose. -Hemicellulose (4 9.) was dissolved in 5% aqueous potassium hydroxide (200 ml.) and the solution was diluted to 700 ml. with water and ethanol, until the polysaccharide barely remained in solution. Fractionation was carried out a t 25" by gradual addition of ethanol until the solution became cloudy, after which stirring was continued for 10 min. Fourteen fractions were collected on the centrifuge and purified in the same way as the original hemicellulose. TABLE 111 Final drying was from petroleum ether (b.p. 30-60") zn OSMOMETRY DATAOBTAINEDFOR T H E METHYLATED HEMI- vacuo. The fractions were weighed and their intrinsic viscosity was determined. The frequency distribution CELLULOSE had one maximum located a t a P , value of 170 and exhibited Wa hb h/a a slight negative skewness. 4.815 9.371 1.946 Acknowledgment.-The authors wish to express 4.415 7.897 1.789 their gratitude to Dr. C. P. J. Glaudemans and Dr. 3.814 6.718 1.761 A. Taurins for valuable help and advice. They 3.511 6.061 1.726 are also thankful to Miss A. M. Time11 for as2.760 4.835 1.752 sistance in collecting milkweed pods. F. W. B. 1.904 3.168 1.664 wishes to express his gratitude t o the American ... 1.5 1 0 Viscose Corporation for two fellowships and to the * Osmotic height in Pulp and Paper Research Institute of Canada for a Concentration in g./kg. solution. cm. solvent. two summer stipends.
68, 911 (1946) (62) B. H. Zimm and I Myerson, THISJOUXNAL, (63) Kindly supplied by Dr. R. H . Marchessault, American Viscose Corporation, Marcus Hook, Pa.
[COSTRIBUTION FROM
THE
(64) M. L Huggins, THISJOURNAL, 64, 2716 (1942).
MONTREAL, QUEBEC
DEPARTMENT O F MEDICINE, HARVARD MEDICALSCHOOL, MASSACHUSETTS GENERAL HOSPITAL]
AND THE MEDICAL SERVICES OF THE
Acetohalogeno Derivatives and Glycosides of D-Galactosamine BY
ZOFIA TAR.4SIEJSKA AND
ROGER\Ir.
JEANLOZ~
RECEIVED JUNE 24, 1958 2-Acetamido-3,4,6-tr~-O-acetyl-2-deoxy-~~-~-galactop~~ranosyl chloride and bromide have been synthesized from D-galaCtosamine CY- and P-pentaacetates. Under various conditions, they afforded methyl and ethyl Z-acetamid0-3,4,6-tri-Oacetyl-2-deoxp-~-~-galactopyranosides, which were further hydrolyzed into methyl and ethyl 2-acetamido-2-deoxy-P-~galactopyranosides. The 1-chloro derivative was shown to transpose into 1,3,4,6-tetra-O-acety1-2-amino-2-deoxy-~~-~galactopyranose hydrochloride. Opening of the 1,6-anhydro ring of 2-acetamido-3,4-d~-~-acetyl-1,6-anhydro-2-deoxy-~-~galactopyranose with concomitant acetylation was successfully achieved.
Investigations on the metabolism of aminosugars have demonstrated the need for the availability of phosphate esters of D-galactosamine possessing known chemical s t r ~ c t u r e s . ~The pre(1) Aminosugars. XIX. This is publication No. 236 of T h e Robert W. Lovett Memorial Foundation for the Study of Crippling Disease, Department of Medicine, Harvard Medical School, Massachusetts General Hospital, Fruit Street, Boston 14. This investigation has been supported by research grants from the National Science Foundation (Grant 9-2312) and the National Institute of Arthritis and Metabolic Diseases, National Institutes of Health, Public Health Service (Grant A-148-C4). (2) Special Investigator of the Arthritis and Rheumatism Foundation. (3) C. E. Cardini and L. F. Leloir, Arch. Biochem. el Bioghys., 46,
55 (1953).
requisite intermediates in the synthesis of the 1phosphate esters of D-galactosamine are the acetohalogeno derivatives. Whereas synthesis of this type of derivative in the N-acetyl-D-glucosamine series has been the subject of recent numerous publication^,^-^^ the analogous derivatives of (4) R . Kuhn and W. Kirschenlohr, Chem. Ber., 66, 1331 (1953); 87, 384 (1954). ( 5 ) R. R. Baker, J. P. Joseph, R. E. Schaub and J. H. Williams, J . Org. Chem., 19, 1786 (1954). (6) F. Micheel, F. P. van de Kamp and H. Wulff, C h e m . Ber., 88, 2011 (1955:. (7) L. F. Leloir and C. E. Cardini, Biockim. et Bioghys. A c f a , 20, 33 (1956). (8) G. Fodor and L. 6 t v o s . Chern. Ber., 89, 701 (1956); A n n . , 604, 29 (19.57).
AT-acetyl-D-galactosaniine have not yet been re- 111 with absolute ethanol afforded ethyl 2-acetamido-3,4,G-tri-O-acety~-%-deoxy-~-~-galactopyra~ioported and are the subject of the present paper. The lack of success of Moggridge and Neuber- side (VI), subsequently hydrolyzed to ethyl 2-acetager14 in their attempt t o isolate the 2-acetamido- mido-2-deoxy-B-D-galactoside (1'11). The same 3,4,&tri-O-acetyl- 2 -deoxy- D -glucopyranosyl bro- glycoside VI1 had been obtained previously during mide has been shown t o be due to the instability the recrystallization of 2-acetamido-2-deoxy-~-g~1acof t h e 1-bromo derivative which isomerizes during tose16 from absolute ethanol and i t was identified the process of recrystallization into 1,3,4,G-tetra- by transformation into VI. The preparation of the a-aiionier of the ethyl 0-acetyl-2-amino-2 -deoxy- a - D-glucopyranose hyIn order t o obtain the $-glyco- glycoside was carried out b y treating ~-galactostlrobromide.6~11~12 sides, t h e l-bromo derivative was not isolated b u t amine pentaacetate with ethanol in the presence of was treated directly with an alcohol in the pres- hydrochloric acid according t o the classical proence of silver oxide, l 2 silver carbonatell or mercuric cedure.15 Since derivatives of ?-acetanlido- 1 ,Ganh ydro-2cyanide4,l3 as a condensing agent. Use of t h e more stable crystalline 1-chloro derivative5 leads deoxy-3-D-galactopyranose are valuable interalso t o the successful synthesis of ,!3-glycosides.10~11~13 mediates in the synthesis of substituted derivHowever, the chloro compound is also somewhat atives of D-galactosamine, the 1,&anhydro ring unstable, resulting in its partial transformation was opened by acetolysis with the purpose of transforming the resulting acetate into the reactive into the hydrochloride during crystallization. The present studies were conducted with D- 1-halogenogalactosarriine. Despite the previous galactosamine, and led t o results identical with those report of an unsuccessful attempt to acetolyze observed in the D-glucosamine series. Action of 2-amino-1 ,G-anhydro-2-deoxy- 3 - D - galactopyranose hydrobromic acid iri glacial acetic acid on either hydrochloride, 17 its .\T--acetyl-3,$-di-O-acetylderivaof the ariomers of D-galactosamine pentaacetate tive 1711 showed L: change of rotation from -37 t o 102'. when treated with acetic anhydride, acetic producetl 2-acetainirlo-3,1,B-tri-O-acetyl-2-deoxy-Dgalactopyranosyl bromide (I). Because of t h e acid and a catalytic amount of sulfuric acid.I8 known instability of the analogous glucosamine de- Isolation of the resulting product gave the crystalrivative, no attempt was macle to isolate I. I t s line cy- and P-anomers of ~-acetamido-l,X,~,Gl5 in yields solution was treated with methanol in the presence tetra-O-acetyl-2-deoxy-D-galactopyranose of iriercuric cyanide or silver oxide t o give methyl of 61 and 5yo,respectively. Reaction of the sirupy ,G-tri-O-acetyl-2-dec,xy-P-D-galacto- mixture of the pentaacetate with hydrobromic ?-acetamido-:3.~t pyranoside (XI). Thth catalysts gave similar acid in glacial acetic acid, followed by direct treatyields, but removal of the silver salts was easier. ment with methanol and silver oxide, gave the Thc total yield of TI from n-galactosamine penta- previously described methyl 2-acetnmido-3,4,& acetate TWS TCiC.;. Sicce 8-glycosides of D - ~ ~ U C tri-O-acety~-2-deoxy-~-~-ga~actopyranoside O(11). amine also have been obtained starting from the E~perimental'~ chloro dcrix-atis-e, which is more convenient t o 2-Acetamido-l,3,4,6-tetra-0-acety~-2-deoxy-~,,~-~-ga~actomanipulate than the bromo deriva,tive, 2-acet- pyranose.-The mctliotl described b y Staccy'j was modiaiiiido-3,~4,Ci - tri - 0acetyl - 2 -deoxy- LY -n- galactopy- f i e ~ I . ~ oA solution of 2.0 g. of D-galacto unine hydrochloride ranosyl chloride (111) was synthesized from either in a mixture of 15 ml. of anhydrous p ridine and 10 nil. of acetic anhydride was left overnight at room tenipernture. cy- or 3-anomers of D-galactosamine pentaacetate. The mixture rras diluted with 300 nil. of cliloroform and 'l'wo iiiethods were tested, hydrogen chloride in washed at 0" successively with 15 ml. of water, two times with ether5 arid titnnium tetrachloride"; the latter was 10 ml. of saturated sodium bicarbonate solution and then with fount1 to he simp!er. Ti'hen the chloride I11 was 10-ml. portions of a IO':> so1utio:i of cupric sulfate until disappearance of the dcep blue pyridine-copper complex, treatetl with methanol, in the presence of silver and finally with water. After drj-ing over sodium sulfate, oxide, the $glycosick I1 was isolated in a yield of the chloroform \!-as removctl iit Y ~ C Z ~and O t h e sirupy residue
+
was crystallized bl- addition of n snx~.llamount of absolute
Like the glucosamine derivative, 111 was foulid ethanol. n'heii the &momer is present, i t crystallizes However, in some preparations, only tlic a-form was to be unstable on recrystallization and to transform first. isolated. The total yield was improvcd to 62y6 b y clirointo 1,3,4,G-tetra-O-acetyl-2-amino-2 -deoxy- a - ~ - matography on silicic of the niotlier liquors, the CYThe struc- anomer being eluted first with pure ethyl acetate. galactopyranose hydrochloride (IT'). Methyl 2-Acetamido-3,4,6-tri-O-acetyl-2-deoxy-~-r~-gature of I\- was ascertained by LIT-acetylationt o give lactopyranoside (11) via 1.-A solution of 500 rng. o f 2the known 2-acetamido-1 ,3,4,6-tetra-O-acety1-2acetamido-l,3,4,6-tetra -0 -acetyl - 2 -deoxy-a - n -galactopytleoxy-a-D-galactopyranose.15 ranose in 4 ml. of glacial acetic acid saturated at 0' with Alkaline hydrolysis of I1 gave the crystalline hydrobromic acid was left overnight at room temperature methyl 2-acetamido-2-deoxy-p-D-galactopyrdnosidein the dark, according t o Inouyc, et ~ 1 . ' ~After dilution with 40 ml. of chloroform, the solution \vas washed rapidly (V). Similarly reaction of the 1-chloro derivative with small portions of ice-cold water, then two times with ail (!)) Ti. X I d e y , G. F. Xlaley a n d H. h. Lardy, THISJOURNAL,'18, .5:30:3 ( ! < i 5 G ) . (10) C!,. J nlore!, f i v p e v i o r t i n , 12, 4 I 9 (1
