Carbohydrate Research, 87 (1950) 249-256

Elsevier Scientific Publishing Company, Amsterdam - Printed in The Netherlands

STRUCTURE OF THE GLUCOMANNAN OF Aloe barbademfs MILLER*

GAURHARI MANDAL**
AND AhlALENDU DAs+

ISOLATED

FROM

THE

LEAVES

Dcpartmerrt of Chemistry, Jada~~purUniversii_v, Calculta-700032

(Imiia)

(Received June 23rd, 1980; accepted for publication, July 17th, 1980)

ABSTRACT

The polysaccharide mixture obtained pulp was fractionated by stepwise treatment solution. This yielded a pure glucomannan in the molar ratio of 1 :22. Methylation 2,3,4,6-tetra-, 2,3,6-tri-, and

by hot-water extraction of A. harhadmsis with calcium chloride solution and Fehling fraction containing glucose and mannose analysis of the glucomannan furnished and 2,3,6-tri-0-methylglucose

2,3-di-0-methylmannose,

in the molar ratios of 1.3: lS.3: 1.2: 1.0. The glucomannan reduced one molar equivalent of periodate per hexosyl residue, and none of the monosaccharides survived the oxidation_ Smith degradation furnished mainly erythritol and a trace of glycerol. From these results, a structure has been assigned to the repeating unit of the glucomannan. The number-average, molecular weight (Rn) of the permcthylated glucomannan was found to be 1.5 x 1O5.
INTRODUCTION

Previously,

we

have

reported

that

the

polysaccharide

composition

of

the

leaves of A. barbadensis have helped in lowering

changes with the season of the year. This the uranic acid content to - lo%, by CaCI,

change seems to treatment of the

crude, polysaccharide mixture. The resulting polysaccharide mixture enriched in hexose (especially mannose) could be readily separated into (i) a pure glucomannan fraction (as its copper complex), and (ii) a mixture of a galactan, an arabinan, and pectic acid. We now report the structure of the glucomannan.
RESULTS AND DISCUSSION

A. barbademis

The crude

leaves used in this investigation were collected in October, 1977. polysaccharide (A) was isolated from the pulp by extraction with hot water,

*Characterization of the Polysaccharides of A/or Dar-bademis Miller, Part II. For Part 1, see rcf_ I_ **Present address: Department of Chemistry, Mahishadal Raj College, P-0. Mahishadal, Dist. Midnapur, West Bengal, India. tTo whom enquiries should be addressed. 02.25, @ 1980 0008-621 s/so/ OOOO4OOO/S Elsevier Scientific Publishing Company

230

as already described*, and found to contain GalA, 70_5j& and total hcxose, 9.95x,:,. Fractionation of A by treatment with CaClt solution furnished polysaccharidc A, containing GalA, 35.2;/,, and a repetition of the purification process furnished polysaccharide A, containing sugars of A, were Man, traces of Glc and Rha. a low value to such GalA, 9.96 ?A, and total Gal, and Ara in the molar It is significant that the uranic the isolation hexosc, 65.7 ;:,_ The major, neutral ratios of 4.7s :2_67 : 1.0, along with acid content
could

not bc lessened C,, see Part I).

during

of the u-galactan

(Fraction

Also, Man, which was present as a trace in polysaccharide C,, was now the major component of A,. Polysaccharide A2 showed no ester bands (ri:., at 1730 and 1260 cm-) in its i-t-. spectrum. Hence, considerin, n the method of isolation, it would be highly lyzed_ improbable As mannans that are all of the O-acetyl amenable to groups originally with present had been hydro-

complexing

fractionation

was achieved

by treating

the aqueous

copper solution, solution of AZ with

further

solution at 4: this resulted in the isolation of polysaccharidc fractions precipitated, copper complex) and A, (from the supernatant liquor). observation was that, although A, itself was quite soluble in water, its sub-fraction A, (containing the glucomannan) was very insoluble, and sub-fraction A; (containing the galactan, arabinan: and galacturonan components) was fairly soluble. Because of its insolubility in water or alkali, further purification of A, was possible only by dispersion in water, and centrifugation. For component identification, A, was dissolved in 7OTL H,SOa, the solution diluted. and the polysaccharide hydrolyzed_ The component monosaccharides were found to be GIc and Man in the molar ratio of I :21.9. The hexose content of A, (as cstimatcd, after hydrolysis, by g.l.c., using nzw-inositol as the internal standard) was found to bc 90(x,, and this probably reflects the very hygroscopic nature of the material. The compositions of fractions and A, are given in Table I_ A, A,, AZ, A,, Methylated A, was obtained by one Hakomoria and two Put-die treatments: the product contained OMe, 43.9%, and had ~~~~~~~~ - 10.2. Attempted fractionution of the permethylated product by adding petroleum cthcr to its chloroform solution furnished essentially one fraction, which was indistin~uishablc from the parent compound in respect of methoxyl content and value of specific rotation: this indicated that the permethylated product was homogeneous. Mcthylation analysis

Fehling A, (from the An interesting

of the product furnished (a) X,3,4,6-tetra-, (h) 2,3,6-tri-, and (c) 2,3-di-O-mcthylmannose, and (d) 2,3,6-tri-O-methylglucosc, in the molar ratios of I .3 : IS.3 : I.2 : 1.O (see Table II). The molar ratio (tl): (U i- h + c) = l.O:ZO.S, bein, u also of the same order as that tif Glc and Man in A,. corroborated the homogeneity of the glucomannan. As g.1.c. was not conclusive enough to identify the tetru-0-methyl sugar, the latter was identified by demethylatin, 0 it, and characterizing the resulting mannose in the usual way. From these results, it is possible to deduce some conclusions rcgardiny the structure of the repeating unit of the glucomannan. Characterization of the tri-O-

252
TABLE II
OBTAIXED FROhl THE GLUCOMAKSKAX

G.

MANDAL,

A.

DAS

IDEX~IFICATION THE METHYL~TED OF SUGARS

OF

A. barbadensis

Components (as alditol acetates of)

RRP Cal. I
1.00 (1-W

Mole ratio Cal. II

2,3,4,6-Tetra-0-methylmannose 2,3,6-Tri-0-methylmannose
2_3.6-Tri-0-methylglucose

1.00

1.3:1 18.3:1
1.0:

2.24

(0.99)
2.06
(2.03) 2.35 (2.32) 3.79 (3.69)

(2.20) 2.52 (2.50)

2,3-Di-0-methylmannosc

5.00 (4.83)

1.2: 1

aRelative retention-times with respect to I,Edi-O-acetyl-2,3,4,6-tetra-0-methyl-D-glucitol col. = column. Literature valuess are given in parentheses.

as unity;

methyl sugars indicated that the hexosyl residues (glucose and mannose) are linked through O-1 and O-4 in the main chain, and, possibly, also in the side chains in the repeating unit. The occurrence of 2,3-di-O-methylmannose indicated that this polymer is branched, and that the branching took place only from the mannosyl residues, probably through O-6. The isolation of only tetra-O-methylmannose indicates that the nonreducing ends of the repeating unit are composed of mannosyl residues_ In view of these results and the molar ratios of the methylated sugars, structure 1 may be suggested as being the average repeating-unit of the glucomannan.

T .
1 Man 4
-

_ t.

1 Man
I

where (s + _V-i- z) = 20 mannosyl residues. In structure 1, the length of the side chain (z) and the location of the glucosyl residue (in the main or side chain) remain uncertain_ The negative specific rotation - 10.2) of the permethylated glucomannan indicates a preponderance of /I-linkages. On periodate oxidation, A, reduced one molar equivalent of the oxidant per hexosyl residue within 48 h (constant thereafter). During oxidation, the solution

253

0123456

10

11

Concentration

C (mg/mL)

Fig. 1. Plot of x/C againsr C for permethylated thp_ value of (zr/C)c_o \\:a found to be 1.17.) became probably acidic, but the results of the

A:s (glucomannan

fraction).

(From

the intcrccpt,

formic

acid

estimation

were

not

reproducible,

due to the initial, heterogeneous state of the reaction mixture. Smith degradation of the oxidized product furnished erythritol as the major component, and a trace of glycerol, and none of the monosaccharides survived_ These findings are also consistent with the structure proposed. The number-average, molecular weight (mn) of pcrmcthylatcd A, was dcrermined by osmometry as described earlier, using the same solvent system (Cl-ICI,-

to EtOH). From the plot of z/C against C (see Fig. I), the intercept corresponding was found to be 1.17, and, from this, the m,, value was calculated to be (&x-o 1.5 x 1os. It is evident, therefore, that this glucomannan is significantly diNerent from
those previously lower molecular
ESPERlhlENTAL

described 7-8 in being neither linear nor acetylated, weight.

and it has

much

~I~~te~inls or~clrwthotls. - Paper partition-chromatography (p.c.)was conducted on Whatman Nos. 1 and 3 MI papers, using the following solvent systems (v/v): (A) 8 :2 : 1 ethyl acetate-pyridine-water, (B) 4: I : 5 I-butanol-ethanol-water (uppc~ layer), and (C) 40 : I I : 19 I-butanol-ethanol-water. The staining reagents used wcrc (a) alkaline silver nitrate, (h) benzidinc periodate, and (c) aniline hydmgrnoxalate. Unless otherwise stated, diminished pressure. Analytical earlier. columns (100-200 partially
mesh)

all evaporations were conducted at 35-40 under methods and instruments were the same as reported

Specific rotations were recorded at equilibrium, at ~20~. For g.l.c., glass (I.53 m x 6 mm) containing (I) 3 y<, of ECNSS-M on Gas Chrom Q mesh) at 190 (for alditoi acetates of neutral sugars) and at 170 (for methylated alditol acetates), and (II) 5%, of OV-225 on SIL. RUB. (SO-100 at 170 (for partially methylated alditol acetates) were used.

hydrolyzed7, and the tetra-O-methyl sugar was isolated in the pure state by preparative p.c. (solvent C) on 3 MM paper. This compound was demethylntedS, and the resulting product was found to contain mannose (p-c.: solvent rl, stain (I), along with some partially methylated sugars. Periodate osiciation. Polysaccharide A, (4 mg, in dupiicate) was dispersed in water, and subjected to periodate oxidation at 10. The initial turbidity disappeared within 18 h, when the rate of uptake could be determined spectrophotometrically. The uptake became constant in 4-S h. with reduction of 0.93 mol of the

oxidant per hexosyl residue. In another experiment, A, (16 mg) in IH20 (S mL), in duplicate, was oxidized with periodate under the same conditions_ Aliquots were titrated for formic acid with O-01&1 NaOH at 20, 30, 40, and 45 h in the usunl way. but the results were not satisfactorily reproducible. Remaining portions of the duplicate reaction-mixture were pooled. and decomposed with ethylene giycol (I mL). This mixture was dialyzed, and the solution concentrated by Iyophilization. Tha resultins polysaccharide was reduced wit!., KBH& (100 mg) for 24 h at 4, and the reduction product (-9 mg) isolated in the usual way 7. Hydrolysis OF the reduced polysaccharide revealed the presence of erythritol (major. RG_,, I .S5: solvent .<I_ stain h) and a trace of glycerol (RcaI 2.3 I : solvent A. stain b). No hexosc survived the oxidation_ Detcwnifwtiolr of ~~iokcwrlarwci.yht. The number-average, ~~~olcculniweight (R,,) of permethylated A, was determined by OsInometry (see Part I). Mc:tsurcments \kere made by using duplicate solutions, of conccntraiions (mg/mL) 10.49, 5.215. and 2.6225, in 24: 1 (v/v) CHCI,-EtOH at 30.
ACKNOWLEDGSIENTS

The authors thank the U.G.C. (New Delhi) and Mahishadal Raj College I-~I providing an F-1-P. Fellowship and study leave to G. M. They also thank Dr. N. N. Das of IJIRA, Calcutta, for helpful suggestions. REFERENCES

3 J. K. N.
4 S.-I.

JONES AXD Ii. J. STO~DLI-:Y, ~W~lro~b C~ri50l1~lr.

Clam.,

5 (1963)

36-X.

HAROMORI, J. Biodxwt.
AXD .I.

(Tok_w).
f. Clrrm. HAMILTOS.
ASD

55 f 1964) 205-705.
SW_, R.

5 T. PURDIE

C.
J.

IKVIM. K.

53 (1903) 1021; S5 (190-I) 1049-1070.

A~I)

6 M. ALIDEL-AKHER, (1951) 4970-4971. 7 B. S. PAIILSEN, E. S D. C. Gow~n, B.

9
A. R. N.

MOXXOX~LR~,

F.

S\II-I~L J. Am.

C~IPI~~. SM..,

71

F.~GEttbiElbl.

E.
AND

@WXIWI:.
Y. V.

CCWIJO/I_W/I~_ 60 ( 19%) 34%351_ Rec.,

Cur-hol~_wfr.

NEELISIDDAIAH.

ANJAXEYALLJ.

Rcs..

72 (!979)901-205.
444-145.

GORROD AND J. K. N. Jorts.

J. C/rem.
C/lrcw..

Sm..

( 19.51) 7522.
Nmrw, 1042-1045. 166 (1950)

Iti W. E. TXEVELYAN, D. P. PS~O~IER. I I J. A. CIFO~<ELLI AND F. .%lITH, A//u/.

12
R H. HORRGCU WD G.

AX\DJ. S. HARRISOS, Lorrc~r, (1949)

26 (1954)1132-l134. 34 ( :96S) 471-479.

B. M.~NNIN~_

13 W. F.

LEHN~~ARDT AND R. J. C.IXS%LER, Cl~rot~:r~rogr.. J.

256
14

G. NANDAL,

A. DAS

L. HOUGI-I, J. K. N. JOGS, AND W. H. \~ADXlAN, J. Chenz. Sot., (1950) 1702-1706. 15 E. A. KAEAT AND M. M. MAYER, Esperitnental Intmt~ltocltenlistr~l, 2nd edn., C. C. Thomas, Springfield, Ill., 1967, pp. 542-549. 16 T. G. HALSALL, E. L. Hmsr, AXD J. K. N. JONES, J. Cilenr. Sue., (1947) 1427-1432. 17 N. N. DAS ASD A_ DM, Curboltydr. Res., 56 (1977) 337-349. IS P.-E_ JASSSOS, L. KESNE, H. LIEDGREN, B. LINDBERG, AND J. LGNNGREN, Cknt. Conumm., Univ. Stocklmlm, S (1976) 22-23.