I. J. GOLDSTEIN, J. K. HAMILTON AND F. SMITH
6252 [ COSTRIBUTION FROM
Vol. 81
DEPARTMENT OF AGRICULTURAL BIOCHEMISTRY, UNIVERSITY
THE
OF
LXINNESOTA]
Reduction of the Products of Periodate Oxidation of Carbohydrates. X. Studies on Amylopectin Polyalcohol
Methylation
BY I. J. GOLDSTEIN, J. K. HAMILTON AND F. SMITH~ RECEIVED APRIL23, 1959 Amylopectin has been transformed into the corresponding polyalcohol I V (R = H ) by periodate oxidation followcd by reduction. Methylation of IV ( R = H ) has been effected and the product IV ( R = CH3) subjected to niethanolysis; the following cleavage products have been identified: 1,4-di-O-methylerythritol (VI), D-( 1-0-methylerythritol) (1.11) and methoxyacetaldehyde (VIII). The molecular ratio of LrI and VI1 has been determined and shown t o provide a new approach to the average size of the repeating unit of amylopectin.
In a previous communication2 i t was shown The amylopectin polyalcohol I V (li = H) that amylopectin could be oxidized with periodate was methylated with methyl sulfate and 405% and the resulting polyaldehyde reduced to the potassium hydroxide in the usual way. the corresponding polyalcohol. Since model experi- first methylation the product which was readily ments have s h ~ w nthat ~ , ~the alcohols I, I1 and I11 extracted by chloroform was treated with acetic prepared in the same manner, respectively, from anhydride and pyridine to eliminate any borate methyl P-D-xylopyranoside, methyl a-D-gluco- complexes6 and to facilitate complete methylation pyranoside and methyl p-cellobioside and methyl of the amylopectin polyalcohol. After fivc addiP-lactoside could be transformed into the corresponding fully methylated derivatives, i t became tional treatments with methyl sulfate and alkali feasible to study polysaccharide polyalcohols by the partially methylated polyalcohol was subjected the methylation technique. This paper is con- to the action of sodium and methyl iodide in liquid cerned with methylation studies on amylopectin ammonia and then two methylations with silver polyalcohol (IV, R = H). The methylation oxide and methyl iodide. CH,OH
I
OH
OH
4
J
p
y
OH OH
yy, -
0
H
111
YH2 CH2OPL OR OR
CHAOR
,
1
I CH, CH,O '3 b R OR -J ?i,
I \' CH20CH3
CH20CH3 H-A-OH
I CHOH I
CH*OCH:,
v
I H-C-OH 1
CHgOCH3 vI
CH20CH3 H-&-OH d - O H
1
CHpOH VI1
CHO
I
CHZOCHa VI11
studies on other polysaccharide polyalcohols will form the subject of later communications. Amylopectin was subjected to prolonged periodate oxidation and the polyaldehyde so formed was reduced to the polyalcohol I V (R = H) with sodium b~rohydride.~ (1) Paper KO. 4095, Scientific Journal Series, Minnesota Agricultural Experiment Station, St. Paul, hfinn. Presented a t t h e 131st A.C.S. Meeting, Miami, Fla., April, 1957. ( 2 ) J. K . Hamilton and F. Smith, THIS J O U R N A L , 78, 5907, 5910 (1956). ( 3 ) J . K . Hamilton, G. \V. Huffman and F. Smith, i b i d . , 81, 2173 (1959). (4) J . K . Hamilton, G. \\'. Huffman and F. Smith, i b i d . , 81,2176 (1959). ( 5 ) hI. Ahdel-Akher, J. K. Hamilton and F. Smith, i b i d . , 73, 4691 (1951).
Hydrolysis of the fully methylated amylopectiii polyalcohol (IV, R = CH3) with boiling niethanolic hydrogen chloride afforded a mixture of 1,4-di-0methylerythritol (VI), D-( 1-0-methylerythritol) (VII) and methoxyacetaldehyde (VIII) . The 1,4-di-O-methylerythritol (VI) which arises from the 1,4-linked-~-glucose residues was identified as the characteristic 2,3-di-p-t0luenesulfonate.~ The D-( 1-0-methylerythritol) (VII), derived from those glucose residues a t which branching occurs, was identified readily as the crystalline 2,3,4-tri-Pnitrobenzoate. The methoxyacetaldehyde VI11 which arises from C1and C2of all the building units in IV, R = CH3,readily formed the characteristic crystalline p-nitrophenylhydrazone. Since the l,-l-di-0-niethylerythritol (13) :irises from the non-terminal glucose units m d the D(1-O-methylerythritol) (VII) from the glucose residues a t which branching occurs, then the molecular ratio of VI to VI1 should correspond to the average value of the repeating unit (the r'itio of non-terminal to terminal non-reducing residues)
Dec. 5 , 1959
METHYLATION O F AXYLOPECTIN POLYALCOHOL
6253
in amylopectin. The molar ratio of VI to VI1 tions with vigorous stirring, half of the reagents being added a t room temperature and the rest a t 55'; acetone was added was calculated from the periodate consumed by VI when necessary to keep the partially methylated polyalcohol and VI1 in the hydrolyzate of IV (R = CH3) in solution. The product was extracted from the cooled and the formaldehyde which is formed from VI1 reaction mixture with chloroform. The combined chloroa t the same time. Since the molar proportion of form extracts were washed with water to remove salts and to dryness giving a sirupy product. formaldehyde produced from VI1 is equal t o half evaporated Acetylation of the Partially Methylated Amylopectin the molar periodate consumed by VII, the differ- Polyalcohol.-In order to cleave any borate complexes6 and ence between the total molar periodate consumed to facilitate completion of the methylation, the sirupy prodby VI and VII, and the periodate consumed by VI1 uct obtained in the previous experiment was dissolved in (50 ml.) and treated with acetic anhydride (35ml.). corresponds to the number of moles of VI present pyridine After heating to 40" the reaction mixture was kept overnight in the mixture. The average of four determinations and poured with stirring into ice-water (300 ml.). The stiff of the ratio of VI to VI1 was about 17, a result in sirupy product which settled immediately was separated by good agreement with those obtained by direct decantation from the upper, aqueous layer and repeatedly methylation studies6 and by periodate oxidation washed by trituration with water. The aqueous layer was extracted five times with chloroform (100 ml.) and the comand determination of the formic acid p r o d ~ c e d . ~ , ~ bined chloroform extracts were washed with sodium bicarThe amount of 1,3-di-O-methylglyceritol (V) bonate solution and with water. The chloroform extract formed from IV (R = CH3) would also provide a was combined with the sirupy product and the whole value for the average repeating unit in amylopectin. evaporated in vacuo to give a sirup. Complete Methylation of Amylopectin Polyalcohol.-The This cleavage fragment (V) is, however, relatively sirupy product obtained from the previous acetylation ex~ o l a t i l e it , ~ is not readily detected on paper chro- periment was methylated in acetone solution with methyl matograms and its determination therefore is much sulfate (200 ml.) and 45% aqueous potassium hydroxide more difficult to accomplish than that of either V I (750 ml.) as already described. Five methylations were applied in this manner, the product being isolated after each or VII. methylation by extraction with chloroform. The product These findings therefore demonstrate that meth- after the fifth methylation was dissolved in chloroform and ylation studies on polyalcohols can be used to the solution washed with water and evaporated to give the throw light on the structure of polysaccharides. methylated amylopectin polyalcohol as a yellow sirup which dried by heating in vacuo a t 60" for 3 hr. A search is now being made to ascertain whether wasThe dried residue was dissolved in liquid ammonia (750 those glucose units of amylopectin that are stable ml.) and treated with sodium (3.3 g.) and methyl iodide (9 to periodate can be detected as a minor component ml.) as described p r e v i o ~ s l y . ~After ~ ~ evaporation of the in the hydrolyzate of the methylated amylopectin ammonia in a stream of dry air the methylated amylopectin polyalcohol was dissolved in chloroform and the solution was polyalcohol. washed with water until the washings were neutral. The
Experimental
Oxidation of Amylopectin with Sodium Periodate.Waxy corn starch amylopectin (40.0 9.) was defatted by two three-hour extractions with boiling methanol (200 ml.). The amylopectin was filtered and dried first in air and then in vacuo (moisture content 7.6%). The amylopectin was suspended in water (450 ml.) and heated with stirring to 75" to effect gelatinization. After cooling to room temperature the solution was treated with a cooled ( 5 " ) aqueous solution (500 ml.) containing sodium periodate (85.6 g.) and the volume quickly adjusted to 2 liters with cold (5") water. The reaction mixture was kept in the dark at 5'. After 10 weeks the periodate consumpton had reached 1.08 moles per glucose residue; the formic acid p r o d u c t i o ~ i , ~0.052 ~' mole per glucose residue, corresponded to an average repeating unit of 19. The amylopectin polyaldehyde which had separated from solution was recovered (centrifuge) and washed with ice-cold water until the washings no longer gave a positive test for iodate (tested with acidified potassium iodide). The product was washed (centrifuge) successively with ethanol, diethyl ether, petroleum ether and dried in air (yield 40.0 g . ) , The polyaldehyde was a fine, white powder which closely resembled the original amylopectin. Reduction of Amylopectin Polyaldehyde with Sodium Bor~hydride.~-Toa suspension of the polyalcohol (20.0 g.) in water (200 ml.) was added rapidly with stirring a solution of sodium borohydride (10 g.) in water (100 ml.). The reaction mixture foamed considerably and a gel-like semisolid formed momentarily. After standing 20 min. the solution was opalesceiit and after standing overnight a clear pale yellow solution resulted which did not reduce Fehling solution. Methylation of Amylopectin Polyalcohol.-The solution obtained in the previous experiment was treated with methyl sulfate (300 ml.) and 40% aqueous potassium hydroxide (900 ml.), the reagents being added during 2 hr. in 12 equal por(6) E. L. Hirst, It.M . T. Plant and M. D. Wilkinson, J. Chem. Soc., 2376 (1932); W. 2. Hassid and R. M. McCready, THISJOURNAL, 65, 1157 (1943). (7) F. Brown, S. Dunstan, T.G. Halsall, E. L. Hirst a n d J. K. N. Jones, N a t u r e , 156, 785 (1945); A. L. Potter and W. 2. Hassid, T H I S J O U R N A L , 7 0 , 3488 (1948).
chloroform solution was dried (MgSO,), evaporated to dryness in vacuo and the product retreated with sodium and methyl iodide in liquid ammonia as before. The methylated amylopectin polyalcohol was subjected t o four successive treatments with Purdie reagents, methyl iodide (100 ml.) and silver oxide (10 g.), the latter being added in five portions in the usual way. The methyl iodide was distilled and the product isolated by extraction with diethyl ether (150 d . ) . Dry petroleum ether (b.p. 30-60') was added until the solution became turbid after which centrifugation removed all inorganic impurities and a clear yellow solution resulted. Evaporation of solvent and heating a t 60" a t 1 mm. gave the methylated am>-lopcctinpolyalcohol IV ( R = CHI) as a viscous yellow sirup (yield 20.0 g.). A solution of the product in methanol ( c 4 ) was optically inactive. Anal. Calcd. for CgH1EOj: OCH3, 45.1. Found: OCHs, 44.1, 44.2. Methanolysis of the Methylated Amylopectin Polyalcohol.-A solution of the methylated polyalcohol IV ( R = CHI, 1.016 g.) in 3 7 , methanolic hydrogen chloride (40 inl.) was boiled for 18 hr. The reaction mixture was neutralized (Ag2COa),filtered, and concentrated (atm. press.) to give a residue R ( 2 ml.) and a distillate D. Treatment of this distillate, D, as before3s4 with an acidified solution of p-nitrophenylhydrazine readily gave methoxyacetaldehyde p-nitrophenylhydrazone, m.p. and mixed m.p. 116". When the distillate D was treated with 2,4-dinitrophenylhydrazine a crystalline product was formed which was separated mechanically into two crystalline forms, orange needles, m.p. 116O, and yellow plates, m.p. 124'. Anal. Calcd. for C9H1006N4: C, 42.5; H , 3.94; hT, 22.03; OCH3, 12.2. Found (for the orange needles): C,42.99; H,4.06; N,21.6; OCHa, 11.8. The residue R (2 ml.) was extracted with acetone, filtered and evaporated (bath temp. 30-35") a t the water-pump. Paper chromatographic analysis, using butanone-water azeotrope as the irrigating solvent and Tollens spray reagent, 0.61) revealed the presence of 1,4-di-O-methylerythritol (Rf and D-( 1-0-methylerythritol) (Rr 0.29). Separation and Identification of 1,4-Di-O-rnethylerythrito1 and D-( 1-0-Methylerythritol).-The sirupy product from
6254
I. J. GOLDSTEIN, J. K. HAMILTON AXD F. SMITH
Vol. 81
the previous experiment was dissolved in the minimum The solution was cooled and the volume adjusted to 50 ml. amount of butanone-water azeotrope and transferred t o a with water. This solution containing the mixture of 1,4-dicellulose-hydrocellulose column.* Separation of the com- 0-rnethylerythritol and D-( 1-0-methylerythritol) is referred ponents was effected with butanone-water azeotrope using t o as solution A. a n automatic fraction collector, fractions of 10 ml. being Total Periodate Consumption.lo-To a n aliquot ( 5 ml.) of collected every 30 min. Examination of the contents of the solution A was added an aqueous solution (10 mi.) containtubes in the usual way using the Tollens reagent spray showed ing periodic acid (0.20 g.) and the volume was adjusted to that the 1,4-di-O-methylerythritolwas in tubes 18-27 and 25 ml. with water. The reaction mixture together with a the D-( 1-0-methylerythritol) in tubes 58-70. "blank" was kept at 20" in the dark. The periodate conIdentification of 1,4-Di-O-methylerythritol .-Evaporasumption was determined by adding a I-ml. aliquot t o a tion (in vacuo at 35-40') of the eluate in tubes 18-27 gave a saturated aqueous solution of sodium bicarbonate ( 10 ml.) colorless sirup (450 mg.). Some of the product distilled containing 0.1 N arsenite (1 ml.). After standing 15 min. a with the solvent and i t was detected in the distillate by paper crystal of potassium iodide was added and the excess of the chromatography. Treatment of the sirupy product with arsenite back-titrated with 0.01 N iodine. The results p-toluenesulfonyl chloride in pyridine in the usual manner4 showed that the periodate consumed by the 1-ml. aliquot afforded crystalline 1,4-di-0-rnethyl-2,3-di-O-tosylerythri- corresponded t o 3.7 ml. of 0.01 N iodine. Hence the periotol, m.p. and mixed m.p. 139' (after recrystallization from date consumed by 5 ml. of the solution A = (3.7 X 0.01)/2 ethanol). X 25/1 mmoles. Had all the hydrolyzate been treated, the Anal. Calcd. for CmH2608S2: C, 52.4; H , 5.78. Found: periodate consumption would have been = (3.7 X 0.01)/2 X 25/l X 50/5 = 4.62 mmoles. This corresponds to the total C , 52.5; H, 5.72. periodate consumed by the mono- and di-0-methylerythriIdentification of D-( 1-0-Methylerythritol).-Concentratol. tion of the coiltents of tubes 58-70 gave D.( 1-0-methylalierythritol) as a colorless sirup (15 mg.) which showed [ a ] * 3 ~ Determination of the Mono-0-methylerythrito1.-An -5" in ethanol ( c 0.8). Treatment of the D-(l-O-methyl- quot (1 ml.) of solution A oxidized by periodate according to the procedure of Lambert and Neish11 was found, by refererythritol) with p-nitrobenzoyl chloride in dry pyridine in the ence to a standard curve for 1-0-methylerythritol to conusual manner gave the 2,3,4-tri-p-nitrobenzoateas fine, tain 159 mg. of formaldehyde. This corresponds to a total white needles, m.p. and mixed m.p. 162.5-164", [ ( Y ] ~ * D of 42.8 mg. or 0.315 mmoles of 1-0-methylerythritol insolu80' in chloroform (c 1) after recrystallization from acetonetioil .-i. In determining the standard curve relating absorbethanol. An authentic specimen of D-( 1-0-methylerythriance and concentration of 1-0-methylerythritol by the chrotoll tri-p-nitrobenzoateg showed m.p. 162.5-164", [a]23D motropic acid method an amount of 1,4-di-O-methylerythritol 85" in chloroform ( c 1). Determination of the Ratio of D-( 1-0-Methylerythritol) to mas added before periodate cleavage was applied t o generate 1,4-Di-O-methylerythriitol in the Hydrolyzate of Methylated formaldehyde. This ensured that the same amount of methoxyacetaldehyde was present in the standard solution D-( 1-0-methylerythritol) Amylopectin Polyalcohol.-The as in the reaction mixture from the methylated aniylopectin was determined by periodate oxidation and determination of polyalcohol being analyzed. the formaldehyde produced from Cd. Since the molar proCalculation of the Average Repeating Unit of Amylopecportion of periodate consumed in this reaction is twice the is equal tin.-The total amount of 1,4-di-O-methylerythritol molar proportion of formaldehyde generated, then subtracto the total molar consumption of periodate (4.62 mmoles) tion of this calculated molar amount of periodate consumed less the malar consumption (0.63 mmole) of periodate by the by the D-( I-0-methylerythritol) from the total molar amount 1-0-methyIerythrito1(0.315mmole) (since 1mole of the latter of periodate consumed by the mixture of D-( 1-0-methylerJ thritol) and 1,4-di-O-methylerythritol gives the molar reacts with 2 moles of periodate) = 4.62 - 0.63 = 3.99 mmoles or 3.99 X 155 = 618 mg. of 1,4-di-O-methylerythritvl amount of periodate taken up b y the 1,4-di-O-methyl(mol. wt. 1%). erythritol. It was thus possible to arrive at a n approximate Since demethylation results in the formation of 1.3 ing. value for the molar ratio of 1,4-di-O-methylerythritol to of mnno-0-methylerythritol from 100 mg. of 1,4-di-0D-( 1-0-methylerythritol) which corresponds t o the average methylerythritol, then 618 mg. of 1,4-di-O-methylerythritol value of the repeating unit in the parent amylopectin. A correction was applied for a small amount of demethyl- will give rise to 618 X (1.3/100) = 8 mg. of monu-0-methylation of 1,4-di-0-methylerythritol that occurred during the erytliritol. Hence the D-( 1-0-methylerythritol) derived dimethanolpsis of the methylated amylopectin polyalcohol. rectly from the methylated polyalcohol = 42.8 - 8 = 34.8 By means of the chromotropic acid methodll i t was shown mg. or 0.256 mmole. Therefore the inolar ratio of 1,4-di-0that 100 mg. of 1,4-di-O-methylerythritolgave rise t o 1.3 methyl to D-( 1-0-methylerythritol) = 3.99/0.256 = 15.6/1. I n additional experiments the molar ratio was 16.3, 18.2, mg. of mono-0-methylerythritol. T h a t this reaction took 17.5; average, 17. place was confirmed in a separate experiment by paper chromatography. Acknowledgment.-The authors thank the Corn The experimental procedure adopted after several trial Industrial Research Foundation, the Office of experiments is presented: a solution of the methylated amylopectin polyalcohol (1.062 9.) in methanol (30 ml.) Ordnance Research, U. S . Army, and the National containing hydrogen chloride (0.6 g.) was refluxed for 3 hr. Science Foundation for support.
+ +
( 8 ) J. D. Geerdes, B. A. Lewis, R. Montgomery a n d I1. Smith, A i f a l . Cirnm., 26, 264 (1954). (9) 1. J. Goldstein, B. A. Lewisand F. Smith, d b s t r a c t s 132nd -1.C.S. Xeeting. X e w York, N. Y.,Sept., 1957, p . 10-D.
ST.PAPL, MI^. (10) P. Fleury a n d J . Lange, J. p h a r m . chi>!:., [SI 17, 107 (1!)33) M ,Lambert a n d 4. C. Neish. Cali. J . Chein.. 288. 83 (195[!).
(11)
