1404
[COSTRIi?l:TIOS FROM
TIIE
DEPARTYEST
OF
AGRICULTURAL BIOCIIEMISTRY, UNIVERSITY OF hfISNESOTA]
Acetolysis of the Glucomannan of Iles Mannan' B Y F.SMITH .4ND H. c. SRIVASTAVA RECEIVED NOVEMBER 16, 1955 The glucomannan of Iles mannan, the polysaccharide extracted from the tubers of Amorphophallus plants gives on acetolysis followed by deacetylation a mixture of oligosaccharides in addition t o D-glUCoSe and D-mannose. Three of the oligosaccharides have been obtained in a crystaIline form and have been shown t o be 4-O-~-~-glucopyranos~.l-cu-~-ln3nnopyranose, 4-0-~-~-glucopyranosyl-(3-~-glucopyranose (cellobiose) a n d 4-O-~-~-mannopyranosyl-u-~-glueopyranose. The structure of the last-named hitherto unknown disaccharide has been proved by methylation studies. The structural pattern of the glucomannan polysaccharide is discussed.
Iles mannan obtained from the tubers of the Amorphophallus oncophyllus and A . variubilis has been shown2to be a mixture of two polysaccharides, a glucomannan and an amylose-like polyglucosan, in the ratio of approximately 6: 1. Methylation studies2 showed that the glucomannan component was composed of n-glucopyranose and n-mannopyranose units linked together by 1,4'-p-linkages, the ratio of mannose to glucose being approximately 2 : 1. Information concerning the sequence of the sugar residues in the glucomannan may be obtained by stepwise degradation and characterization of oligosaccharides of varying degree of polyrnerizatiori. This paper is concerned with the partial degradation of the glucomannan component of Iles mannan by acetolysis and the isolation and characterization (1) Paper No. 3459, Scientific Journal Series, Minnesota Agricultural Experiment Station. This paper will form part of a thesis t o be submitted b y H. C. Srivastava t o the University of Minnesota in partial fulfillment of the requirements for the degree of Ph.D. (2) P. A . Rebers and F. Smith, THISJ O U R N A L , 76, 6097 (1954).
of three oligosaccharides, namely, 4-o-B-~-glucopyranosyl-p-D-glucopyranose(cellobiose), 4-0-p-nglucopyranosyl-a-n-mannopyranose and 4-0-p-nmannopyranosyl-a-D-glucopyranose. The glucomannan was separated from Iles rnannan meal by precipitation as a copper complex and subsequent decomposition of the latter with dilute acid as previously described. Preliminary experiments were carried out in order to establish the conditions under which acetolysis produced the largest amounts of oligosaccharides. The maximum yield of the oligosaccharide acetates was obtained when the acetolysis3 was conducted initially a t 0" and after 24 hours completed by heating. The nature and relative amounts of the products formed by acetolysis was determined by deacetylation of the mixture of acetates followed by qualitative paper chromatography. Acetolysis was found to give several oligosaccharides as well as n-glucose and D-mannose. The mixture of mono- and oligosac(3) Cf..K . S. Barclay, E J. Bourne, M . Stacey and M . Webb,
1.Chcm. S o c ,
1501 (1954).
charides was resolved into its componentsbychroma- syl-@-D-glucopyranose (cellobiose). Its identity tography first on a charcoal-celite column4 and was confirmed by converting it into the correthen on a cellulose column.6 I n certain cases sheet sponding known 0-octaacetate. The third oligosaccharide C having m.p. 203paper chromatography proved to be effective for 204' and [ C X ] ' ~ D+30" + +19" in water gave upon final purification of the oligosaccharides. T h a t the oligosaccharides were not reversion hydrolysis glucose and mannose in equal amounts. products was shown by the fact t h a t when a mixture I t s molecular weight determined by oxidation with of glucose and mannose (in the ratio of 1:2) was alkaline hypoioditeg and by bromine water corretreated with acetic anhydride and sulfuric acid sponded t o that of a disaccharide. Hydrolysis of under conditions similar t o those in the acetolysis the bromine oxidation product yielded mannose of the glucomannan and the reaction product de- and gluconic acid. Moreover, @-glucosidase had acetylated, chromatographic analysis on paper no hydrolytic action on this oligosaccharide. I t showed no evidence for the formation of oligosac- was apparent therefore that C was a mannosylglucose and since the original glucomannan was known charides. Three of the oligosaccharides produced by acetol- to be composed of glucose and mannose units ysis and subsequent deacetylation have been crys- joined by 1,4'-glycosidic bonds, the oligosacchatallized and identified as 4-O-@-~-glucopyranosyl-a- ride C must also contain the 1,4'-linkage. Proof of this followed from the observation that niethylaD-mannopyranose (A), 4-O-~-~-glucopyranosyl-@D-glucopyranose (cellobiose, p) and 4-O-@-~-manno- tion of C to give D tollowed by hydrolysis of the pyranosyl-a-D-glucopyranose (C). latter gave rise to 2,3,4,G-tetra-0-1ncthyl-~-lnanThe oligosaccharide A which had m.p. 134- nose, identified as the anilide,1° and 2,3,G-tri-O139' and showed [ a ] ' l ~ +4.7" (final) in water gave methyl-D-glucose identified as the 1,4-bis-p-nitroequal parts of glucose and mannose upon acid hy- benzoate.2 From the fact that the oligosacchadrolysis as ascertained by paper chromatographic ride has a low initial rotation (+30", water) i t is beanalysis. Bromine oxidation followed by hydrolysis lieved that the biose linkage is of the P-type and afforded glucose and mannonic acid. This showed since it displays a downward mutarotation, the t h a t the oligosaccharide was a glucosyl-mannose. configuration a t C1 is probably a. The disaccharide Since the glucomannan had already been shown to is therefore designated 4-O-p-D-mannopyranosyl-apossess only 1,4'-linkages, it appeared likely t h a t n-gluco pyranose. this disaccharide was a 4-glucosyl-mannose. This Since it has already been established that the proved t o be correct for the disaccharide was found glucomannan has a linear structure* and is comt o be identical with the 4-O-P-~-glucopyranosyl-a- posed of two parts of mannose and one part of gluD-mannopyranose previously synthesized.6,' More- cose, and if it may be assumed t h a t the structural over, it gave rise t o octa-0-acetyl-4-O-@-~-glucopy-pattern is repeated throughout the polymer moleranosyl-D-mannose8 upon acetylation with sodium cule, then the isolation of cellobiose only, and none acetate and acetic anhydride. of its related homologs such as cellotriose or celloThe oligosaccharide B which had m.p. 227O and tetraose requires that the structural pattern of the showed [ff]""D + 2 8 O + +34" in water and gave repeating unit in the polysaccharide be tentatively only glucose upon treatment with @-glucosidase formulated as proved to be identical with 4-O-@-~-glucopyrano-46-D-G Pl-4fl-0-Gp 1-(4fl-~-hfan p I),(4) R. I. Whistler and D. F. Durso, TEISJOURNAL, 1 2 , 677 (1950). ( 5 ) L. Hough. J. K. N. Jones and W. H. Wadman. J. Chcm. Soc., 2511 (1949). (6) M. Bergmann and H. Schotte, Ber., 54, 1570 (1921). (7) W. N. Haworth, E. L. Hirst, H. R L. Streight, H. A. Thomas and J. I. Webb, J. Chem. SOL, 2636 (1930)' ( 8 ) D. H. Brauns, TEISJOURNAL, 48, 2776 (1926).
By joining two of these together the formation of the two other disaccharides, 4-O-@-~-glucopyranosyl-D-mannose and 4-O-@-~-mannopyranosyl-~-glu(9) J. L. Baker and H. F. E. Hulton, Biochem J . , 1 4 , 754 (1920). (10) J . C. Irvine and D. McNicoll, J . Chcm. Soc., 9 1 , 1449 (1910).
F. SMITH AND H. C. SRIVASTAVA
1406
cose [luring degratlatioii o f the glucoinannan becomes apparent. Since 110 niannobiose or rnnnnotriose was isolated, it woultl appear that the linkage between the mannose units is cleaved more readily than the others during the process of acetolysis. Further support for the above forniulation for the glucomannan is now being sought in the identification of the other oligosaccharides produced by acetolysis.
tion of the ncctatc ntisture i t i ~ i i c t l t y l;ilcoholic solution with a catalytic amount of sodium1* gave a mixture of reducing sugars and oligi isaccliaricles which were examined by paper cliromatogr.ipliy usiiig solvent 1 and spray reagent a. T h e results a r c given in Taljlc I . Experiment (b).--In a sccotid esperitncnt the gluconianIian ( 0 . 5 g . ) NYI, shaken with acetic anhydride (10 ml.) and sulfuric acid (0.3 nil. j for 38 hours. The yield of acetolysis product which separated on pouring the reaction mixture iato watcr \vas n.4 g . The components produced upon deacety1;:ticin are givi.11 in l a h l e I .
Experimental All evaporations were done under reduced prcssure a t 35-45". T h e nicltiiig points are uncorrected. Tlic folloiving solvents w r e used for partition clirriinatiigrapliy: (1) !ayer),tl pyridine-ethyl acctate-wLiter ( 1 : 2.5:3.5-upper ( 2 ) 1-propanol: water azcotrnpe, ( 3 ) 2-hutaiioiie-~raterazcotrope,'2 ( 4 ) l-hut;iliol--ethanoi-\v~tter (5: 1 :4--upper layer), (5) berizene-etlianol-~v~iter(200: 47: 15-uppcr h y e r J L 3antl (6) 1-butanol-acetic acid-ivater ( 4 : 1 :5-uppcr layer). The following spray reagents wcre used for the detection of sugars and their derivatives: ( a ) acetonic silver nitrate and :ilcoltolic caustic ( h ) p-ariisidine trichloroaeetate,'j (c 1 periodate-l~ciizitline'6 a i i d ( d ) liydrosylamine-ferric cl~loride.~~ Isolation of the Glucomannan from Iles Mannan.-To Iles mannan flour (70 9 . ) was added sodium xylene sulfoiiate (40 nil., 507c solution j and the inanitan allowed t o swell.2 Sodium hydroxide (2f)il mi., 3070 by weight) was then added and the mixture allowed to stand for 0.5 hour when the material swelled still further. The mixture was heated at GO" and water (1800 mi.) in portions of 100 ml. was added with stirring to effect solutioit. The solution was centrifuged to remove the undissolved residue and t o the supernatant liquid Fchling solution B (400 mi.) and Fehling solution A (300 ml.) were added with stirring. T h e copper complex, which precipitated as a jelly, was filtered, mashed with dilute Fettling srilutiun and decomposed by suspending it in water a t 0" and adding .V hydrochloric acid dropwise. Tlie portion ( A ) of the complex ivliich did not dissolve was removed (centrifuge). The supernatant was poured into 05% ethaiiol when there separated a white jelly-like precipitate with a greenish-blue tinge. T h e color due t o copper \I as rcmoved by trituration with acetoneacetic acid. Tlie portion A of the copper complex, which had not dissolved in 11' hydriichloric acid, was dissolved in ammonium hydroxide (sp. gr., 0 . 9 ) and the deep-blue solution was dialyzed against distilled water until no more color could bc removed. The glucomannan was precipitated by pouring the dialyzed solution into 95% ethanol. T h e light blucgreen color of the pol iccharidc \vas removed by trituratiiiii with acetone-acetic cid as described previously. The two fractions of the glucomannan polysaccharide thus obtained from the decoin~)~isition of the copper complex were combined and reprccipitatcd b!. dissolving in sodium hydroside (3070) antl atltliiig I'cliliiig solutirm, the copper coinples of the pol>.saccli:iridc being ticciirnposcd with 2 12' h y drochloric acid a\ i l l the previou\ n u n n e r . The polysaccharide was prcci1)itated b y pciuriiig tlie solution itito a~cohol and the coppcr i o i i , renioved, with aeetone-acctic acid. The glucornannart t h u s obtaiiicd ~ v , i za white arni~rphous granular powder (16.3 g . ) which 1i:id [ a ] * ' D -32" ( c 1) in 30y0 sodium hydroside. Acetolysis of the Glucomannan, Experiment (a).-The glucomannan (0.5 g.) w;is sliakeii wit11 a mixture of acetic anhydride ( 2 . 5 ml.), acetic acid j 2.8 nil.) m d concentrated sulfuric acid (0.14 ml.) for 6.5 d a y s a t room temperature. U'hen the reaction mixture \va\ poured tvitli stirriitg into water no precipitate was furmed. Estr:wtioii uf the mixturi with chloroform gave a sirupy acctate (0.4 g.). Deacetyls(11) E. F . hIcI'.irren, E.Brand a n d 11. R . R u t k u n s k i , A l l u ; . : - ' / ! < , I I I . , 23, 11-iti i1051). r I ? ! I.. I ~ o K ~ I.s , S . Cucndet. I . Ehrcntli.il, R . K o c h a n d 1.' hinith, .\-otjtrc, 166, 520 (1050). ( 1 3 1 G . A Adams, Can. J. C k e n . . 33, 50 (19Z.5). (14) W. E . Trevelyan, D. P. Procter and J. S. Harrison, N a t u r e , 166,
444 (1950). (15) L. Hough, J. K . N . J o n e s a n d W. H. Wadman,
J . Chem. Soc., 1702 (1950). (IO) J . A . Cifonell! and F. Smith, A n a l . Chem.. 46, 1132 (1954). (17) M . Abdel-Akher and F. Smith, THIS J O U R N A L , 73, 58.59 (1951).
Vol. 78
TABLE I CHKO\lATOCR:\PIiIC
ASALYSIS OF TFIE
COMPOSESTS PRO-
n~ ;\CETOLTSIS OF GLTCOMASNAS Ii,,,,,,,,,,, ( R,) \-aitica (soI\-eiit 1, spray rcagctit a ) DYCEU
Experiment ( a i Compi~nent
Experiment (b) Component
I\'rn
R," 0.10 .32 .45
1 2 3 4
0 1;
5 (glucose) 6 (mannose) i (maltose staiillartl)
5 (glucose) . 83 6 (mannose) 1.OO 0.55 7 (maltose s t : t n t l d ) 0.55
;t
. R i 11 .-I9 c . 70 , Sf; 1.0
I t is believed t h a t eoriipoiieiits 2 and 3 of cxperirnent ( a ) are similar if not identicdl with components b and c of experiment ( b ) . Component a of cspcriment ( b ) having R , 0.1, however, is believed to be a higher ,oligosaccharide and is not idcntieal with component 1 of experiment (a). Experiment (c).-The glucomannan (4.G g.) in small portions was added with stirring to a mixture of acetic anhydride (17.5 ml.) and sulfuric acid (1.8 mi.) cooled t o 0". After all the polysaccharide was added, the mixture was stirred (at 0") for another 2 hours, and thcn a t room temperature for 12 hours. More acetic ar,hydridc (4.6 ml.) was added and tlic mixture stirred for a further period of 33 hours after which time the reaction was completed by Itcatittg a t 80-90" for 15 minutes (rapid darkening of the solution took place during heating). The reaxtion mixture was poured with stirring into cold water when a grey-white precipitate \vas obtained. I t was filtered, washed with water and dried, (7.8 g.). T h e mixture of acetates gave upon dc:tcetyluttoiiig four oligosaccharides in addition t o glucose and m;utitose. X i i attempt to resolve the mixture of the acetates of the mono- arid oligosaccharides using a calcium carbonate column according t o the procedure of Bredereck, et a!.,1P failed. Examination of Possible Reversion during Acetolysis.T o a mixture of glucose (0.5 g . ) and mannose (1.0 8.) was added a cold ( 0 ' ) mixture of acetic anhydride (6.6 ml.) and sulfuric acid (0.67 ml.) and the reaction mixturc stored in the cold ( 5 ' ) with occasional shaking for 24 hours. The reaction mixturc x a s further allowed t o stand a t room temperature (20-23') for another 24 hours, and then heated at 80-90" for 15 minutes. The reaction mixture was poured into water with stirring and the acetates, after extraction with chloroform, isolated in the usual manner. Upon deacetylation of the sirupy mixture of acetates and chromatographic analysis, only glucose and mannose were found t o be present. Separation of the Mono- and Oligosaccharides on a Charcoal Column.-A portion (1.0 9.) of the mixture of sugars rbtained i n ( b ) was fractionated on a charcoal-celite column4 i n tlic usual way. The column was eluted successively with water ( 2 1.>,*0 570 aqueous ethanol (1.5 I,), 15% aqueous eth:iriol ( 1 1.) and 50% aquccius ethanol. The r ) r d u c t eluted wit11 50% rthanol was a ticgraded %i ri~plcn,Bcr , 6 9 , 12;s f 192Ij!. (19) I1 Hrrdereck, H . D u r r and I(. R u c l . B e r . , 87, 5Pii (19.541. (20) I n other experiments, it was found better to pack the charcoal Celite mixture a s a slurry in ?.5yoaqueous ethanol. and to begin the separation of the mixture of sugars with the same solvent rather than with water. This procedure results in a more effective separation of the monosaccharides from the oligosaccharides than i s the case with water (Allene Jeanes, C . A. Wilham. R . W. Jones, H M . Tsurhiya and C . E Rist, THISJ O U R N A L , 75, ,5911 (1953))
(18) G
(211 W. L. Porter and S . H,>ban,A n d C h e i n . , 2 6 , 184G (1954).
1408
Vol. 78
F. SMITH AND H. C. SRIVASTAVA
matographed on a sheet of paper using solvent 1. T h e zones containing glucose and mannose were cut out by the help of marginal guide strips and the sugars, after extraction from the paper with water, determined colorimetrically by the method of Dubois, et al.,*' using phenol and sulfuric acid. T h e oligosaccharide was found to be composed of equal parts of glucose and mannose. When a solution of the oligosaccharide was incubated with &glucosidase in the manner described above, no hydrolysis occurred. Methylation of 4-0-P-D-MannOpyranOSyl-D-glUCOpyranOSe. -To a solution of C (137 mg.) in water ( 5 ml.) cooled in ice, potassium hydroxide (5 ml., 50%) and methyl sulfate ( 1 ml.) were added dropwise with vigorous stirring. In this manner 6 ml. of methyl sulfate and 30 mi. of potassium hydroxide solution were added over a period of 1.5 hours. During the addition of the reageuts the temperature was maintained below 10'. T h e solution was stirred for another 6 hours and then more methyl sulfate (6 nil.) and potassium hydroxide (30 ml., 50%) were added a t room temperature ( 2 5 " ) . The reaction mixture which a t this stage was nonreducing, was then heated t o 50-60' and the methylation continued by the dropwise addition of methyl sulfate ( 5 ml.) and potassium hydroxide (25 ml., 40%) in the previous manner. T h e reaction mixture was heated a t 90-100" for 1 hour and the product extracted from the cooled reaction mixture with chloroform. T h e extract on drying and evaporation gave a sirup (24 mg.). Since the yield of the methylated material was poor, the aqueous solution left after chloroform extraction was neutralized (sulfuric acid) and then diluted with an equal amount of methanol to precipitate the sodium sulfate. The precipitate was filtered and t h e filtrate evaporated. The material thus obtained was combined with t h e chloroform soluble product and, after adding acetone (20 ml.), methylated with methyl sulfate (10 ml.) a n d potassium hydroxide (50 nil., 40%).at 50-60' in the manner described above. The reaction mixture was heated a t 90-100' for 1.5 hours, and, after cooling, extracted with chloroform. Evaporation of the dried (MgSO,) chloroform extract afforded a sirup (75.5 mg.). T h e aqueous solution left after chloroform extraction was worked up in the previous manner and the material thus obtained was methylated separately with methyl sulfate and alkali. I n this way 19 mg. more of chloroform soluble methylated material was obtained. It was combined with the product obtained previously, dissolved in ether, filtered and the filtrate evaporated t o give a sirup (94 mg.). The latter was methylated twice with methyl iodide and silver oxide and the sirupy methylated product (82 mg.) isolated in the usual manner.
mg.) was refluxed with N sulfuric acid ( 5 i d . ) for 18 hours, and the resulting solution, after neutralization (BaC03) and filtration evaporated t o a sirup. T h e sirup was chromatographed on paper using solvents 3 and 5 and the methylated sugars detected by spray reagent b. The presence of 2,3,0tri-0-methylglucose (component I ) and 2,3,4,6-tetra-0methylmannose (component 11) was recognized by comparison with authentic samples of thcsc two sugars (see Table 111). TAULE I11 Solvent 3
Methylated s~igars
K c U
Solvent 5
KMb
0.68 0.30 Component I .os 0.97 Component I1 2,3,4,6-Tetra-O-~iictliyl-~-rlialltluie .os 1 .00 2 , 3 , 6 - T r i - O - m e t h y l - ~ - ~ ~ i a 1 i ~ ~ o ~ c .(io 0.39 2,3,6-Tri-0-111cthyl-~-glucusc . (i8 .R1 2 , ~ , 4 , ~ - T C t r a - ~ - l l l C t l l y l - l ~ - g l U ~ ~ JI ~. ~o .!I1 I?, = I< ( 2 , 3 , 4 , 0 - t c t r a - 0 - r l l c t l l q . l - l l - ~ l ~ l ~b~H,, ~ ~ ~= ~ ) I< . (2,3,4,~-tctra-~-rlict~iyl-~-1n~1li1i~~e).
The niixturc of nicthylated sugars, obt:tiiietl oil liytlrolyais of the methylated oligosaccharide, was resolved into pure components I and I1 by chromatography on sheets of pnpcr using solvent 3, when 33 mg. of 2,3,fi-tri-0-methyl-~-glucose ( I ) and 39 mg. of 2,3,4,6-tetra-0-n1ethyl-~-m~tl~1iose (II), both in the form of sirup, were obtained. Identification of 2,3,6-Tri-0-methyl-~-g~ucose.-~o11iponent I ( 2 2 mg.) which had [ c Y ] ~ ~+43" D in methanol (c 0.7) was converted into d bis-p-nitrobenzoate by treatment of it in pyridine solution with p-nitrobenzoyl chloride.2 After recrystallization from methanol the derivative (10 mg.) had m.p. and mixed m.p. with an authentic sample of 2,3,6-tri0-methyl-D-glucose 1,4-bis-p-nitrobenzoate 192' and showed [ a I z 8-34' ~ in chloroform ( c 2).2 Identification of 2,3,4,6-Tetra-O-methyl-~-mannose.Component I1 (20 mg.), which had [ C I ] ~ ~+D2 4 O in methanol (c 0.7), was converted into the corresponding anilide by boiling for 7 hours with ethanol (3 ml.) containing aniline (100 mg., freshly distilled). After keeping the reaction mixture overnight, the solvent and excess of aniline were removed in vacuo when the anilide crystallized. Recrystallization from ethanol-petroleum ether gave needles of 2,3,4,6tetra-0-methyl-D-mannose anilide,lo m.p. and mixed m.p. 144', [ ( u ] ~ ~-7" D (equil. value) in methanol (c 0.8). b'hen mixed with 2,3,4,6-tetra-0-methyl-~-glucose anilide the m.p. was 118-127'.
authors wish to thank Hydrolysis of Octa-O-methyl-4-O-P-D-mannopyranosyl-D- Acknowledgment.-The Isbell for a sample of 4-o-p-D-ghCOmethylated oligosaccharide D (75 Dr. H.
glucopyranose.-The
(24) M. Dubois, K.A. Gilles, J. K . Hamilton, P. A. Rebers and F. Smith, Anal. Chcm., 18, 350 (1956).
s.
pyranosyl-D-mannose. ST.PAUL,MINNESOTA
