Studies of Lignin Biosynthesis Using Isotopic Carbon. WIT

Studies of Lignin Biosynthesis Using Isotopic Carbon. WIT. Hydrogenoly sis Products of Lignin. Isolation of Radioactive. BY STEWART A. BROWN AND A. C...
0 downloads 0 Views 774KB Size
May 20, 1939

ISOLATION OF RADIO.\CTIVE HYDROGENOLYSIS PRUDVCTS OF LIGNIK

b.43 Mild alkaline treatment renders all three substances serologically inactive and susceptible to attack by phosphatase. The action of alkali on polyribophosphate was shown to result in dialyzable, non-reducing, chromatographically homogeneous fragments. Alkali-treated S V I likewise consisted of similar fragments which no longer separated upon the addition of alcohol, failed to precipitate antiserum and showed a decreased viscosity. The titration curve and pK’ of the fragments resembled those of a phosphate monoesterd0and existence of this linkage was confirmed by the action of phosphatase. Periodate oxidation of S VI suggested that the glucose and rhamnose were joined in 1,3-like linkages. The liberation of formaldehyde indicated that a -CHOH-CHOH group was present. Since formic acid was not produced, S VI has very few end groups, while the simultaneous disappearance of the galactose indicated t h a t this sugar is bound in 1,4- or 1,2- linkage. After alkaline hydrolysis of S VI, an additional mole of N a I 0 4 was consumed and one mole of formic acid was released, with no change in the (13) S. Zamenhof, G. Leidy, P. L. Fitzgerald, H E. Alexander a n d E. Chargaff, J . Bioi. C k P m , 203, 695 (1453).

[COSTRIBUTION FROM THE

PRAIRIE

2410

amount of formaldehyde. This would be expected if the 2- or 4 - 0 s of galactose were set free by alkaline hydrolysis, as, for example, in cleavage of the phosphate diester, S VI. The production of formaldehyde might come from oxidation of a 1,3-linked ribitol. The following formula is tentatively proposed for the structure of S VI

r

1

- -4

or 2-O-~-galactop?-ranosyl-(l + 3) -0-D-glucopy-

ranosyl-(l

+

3) -O-L-rhnmnopyranos~l( 1 -+ 1 or ?,)-ribitol-

1

OH i 3 or I-0-P-0- 1-

The ribitol phosphate residue might have either the D- or L-configuration. Structural studies are continuing and these indicate the correctness of the order given to the glucose and rhamnose residues. XEW BRUNSWICK, N. J.

REGIONAL LABORATORY, NATIONAL RESEARCH COUSCIL O F CANADA]

Studies of Lignin Biosynthesis Using Isotopic Carbon. WIT. Isolation of Radioactive Hydrogenolysis Products of Lignin BY STEWART A. BROWNAND A . C. NEISH RECEIVEDDECEMBER 8, 1958 After the feeding of several C14-labeled compounds t o wheat and maple, degradation of lignin by hydrogenolysis led to the isolation of dihydroconiferyl and dihydrosinapyl alcohols. The radioactivity of these products was taken t o indicate the degree of incorporation of the administered compounds into lignin. The pattern revealed was very similar to that previously found when the activity of the oxidative degradation products, vanillin and springaldehyde, was used, reinforcing the validity of conclusions from the oxidative degradation. However, labeled sinapic acid, which is converted to lignin yielding syringaldehyde, did not form lignin degradable to dihydrosinapyl alcohol. The active syringaldehyde was shown not to arise from sinapic acid bound t o lignin with alkali-labile linkages. Further evidence was obtained that vanillin can be incorporated into lignin, probably by condensation with a two-carbon metabolite. Acetic acid was converted only slightly to the lignin giving rise to hydrogenoly4s products; hence the Birch-Donovan acetate pathn ay does not contribute significantly to lignification.

Introduction Previous papers of this series? have reported comparisons of the efficiencies with which several plant species can convert numerous C14-labeled compounds to lignin. Although the criterion of the conversion efficiency should, in theory, be the specific activity of an intact lignin, i t is very doubtful whether such a compound has ever been isolated, and the ill-defined nature and variable composition of lignins isolated by standard methods have made this approach, in our opinion, impractical. As an alternative we have turned to lignin degradation products of low molecular weight which have known structures and either are crystalline or form (1) Presented in p a r t a t t h e E i g h t h Annual Research Conference on P l a n t Physiology, Hamilton, Ontario; November, 1957. Issued a s N . R . C . No. 5147. For P a r t VI1 of this series see reference 2e. 12) (a) S. A. Brown a n d A. C. Neish, Nature, 176, 688 (1955); (n) Can. J. Biochem. Physicl., 83, 948 (1955); ( c ) 34, 769 (1956); (d) D. Wright, S. A. Brown and A. C . Neish, ioid., 86, 1037 (1958); (e) S . A. Brown, D. Wright and A. C . Neish, i b i d . , 37, 2 5 (1959).

crystalline derivatives. The use of these degradation products enables a direct comparison of specific radioactivities on a molar basis between the administered compound and the lignin (as represented by its derivative), This alternative approach, too, suffers from certain disadvantages. There is no known reaction by which lignin can be degraded in anything approaching quantitative yield to products of known structure. The best reaction in this respect is nitrobenzene oxidation in alkaline medium, which yields phenolic aldehydes : vanillin, syringaldehyde and p-hydroxybenzaldehyde. It is this reaction which we have used in our studies to date. If one accepts the theory that the unit of the lignin polymer contains nine skeletal carbon atoms, the isolation of substituted benzaldehydes as lignin degradation products has the additional drawback t h a t only seven-ninths of the unit is available for study, the other two carbon atoms being lost in the

rxictation. In addition to providing an incoinplete picture, the recovery of these phenolic aldehydes makes necessary the labelling of phenylpropanoid precursors in either the ring or the adjacent carbon of the side chain, normally a more difficult task than the introduction of carbon-14 into either of the other two side chain carbons. In earlier work we have made the assumption, which appears to be generally accepted, t h a t vanillin and syringaldehyde have a unique origin in the guaiacpl and sy-ringyl lignin residues, respectively. If this assumption is valid, the measurement of the radioactivity of these aldehydes provides a relatively rapid and accurate way to determine the incorporation of carbon-14 into seven of the nine skeletal carbons of the lignin units. In the case of p-hydroxybcnznldehy~le such an assumption was niade only with distinct reservations, because i t had been cstablishcd that, under the conditions of the oxidation, tyrosine can give rise to this p r o d ~ i c t . ~Recent studies2e have confirmed that bound tyrosine can account for much, if not all, of the p-hydroxybenzaldehyde from oxidation of the cell wall residue obtained by the standard extraction procedure. it has been pointed out4,jthat in biosynthetic studies the use of phenolic aldehydes to represent lignin is, in itself, insufficient. Clearly, it is desirable to obtain Confirmation of the results obtained by this method, and this is possible by the isolation of other types of degradation product. Freudenberg and his co-workers have obtained such confirmation by isolating phenylpropanoid ethanolysis products of lignin after the administration of CI4labeled phrnylalaninesa and ferulic acid.6b In addition, phenylalanine was shown7 to be used by spruce for synthesis of coniferin, which is a lignin precursor.5,6a,8On the other hand, Freudenbergfibhas reported that ~ a n i l 1 i n - Cglucoside ~~ fed to spruce became incorporated into an insoluble fraction of the wood which did not, however, yield radioactive cthanolysis products. This contrasts with the finding, in this Laboratory, that free vanillin can be incorporated to an appreciable extent into the phenolic aldehydc-yielding lignin structures of

wheat.'^ Since more complete confirmation of the earlier findings was wanted, yet another degradative reaction has been employed for comparative purposes. The present paper reports comparisons of the specific activity found for vanillin and syringaldehyde with those of the analogously substituted products of lignin hydrogenolysis : dihydroconiferyl alcohol (DHCA) and dihydrosinapyl alcohol (DHSA), isolated after the administration of a series of CI4labeled compounds to maple and wheat. In several cases the specific activities of the phenolic aldehydes and isolated lignins have also been compared, ( 3 ) J. E. Stone, II J lll!indrIl :111cI R.G. T a n n e r , C n n .I. Chenz , 29, 734 (10513. f 4 ) K. F r ~ i i d e n b e r g :Inif . F r i ~Chr,m., 4 9 , 1384 (1997). f.5) K , Kratzl a n d C, R i l l r k , T a p p i . 4 0 , 269 (1957) fli) K P r e i i d e n h c r y , A1ie~718. C/rr,?ii , 6 8 , 84 (1936): ( b ) 68, 608 ( 1 "3(i).

(7) K . I'reudcnl,urg a n d I:. S i v d ~ r c i , r n . C ' / i t , i ~ i E r r , 9 1 , 691 (1955). (8) K. K r a t z l a u d G . IIofbunrr, -2lonafs. Chcm.. 8 9 , 90 (1958).

Experimental Cultivation of the Plants.-Wheat plants ( Triticit in w1gare Vill. var. Thatcher) were grown in gravel culture as described previously.2b Maple twigs ( A c c r neguizdo L. var. interius Sarg.) of the current season's growth were collected from t h e wild state. C14-Labeled Compounds.-~-Phenylalanine-R-C~4 (randomly labeled) and the sodium salt of acetic acid-2-C1Cr:ltnpton and LIayn;irtl9 to which tlie residue from bot11 species was subjected. Hydrogenolysis of the Cell Wall Residue.-The techniques used in the present work for the 115-drogenolysisof lignin arid isolation of the products are those described by nremer, Cooke and Hibbert,'O adapted t o a seirii-micro scale. The principal modification has heen the introduction o f chromatographic procedures not available t o t h e earlier workers. In a typical run the purified cell ~ r d residue, l weighing .i to 7 g., was transferred t o ;I high pressure hydrogeixitlon b o m b with about twelve times its weight of dioxane-water (1 1) containing 0.l''A hydrogen chloride. T h e use of tliis solvent gave yields cotnpar:ible to those obtained by us using ethanu-water, :inil possible cotnplications arising through the forniation of cthuxy groups" were avoided. *Ifter the adtlition of :tbout 10 g . o f settled Kaney nickel suspension,12the bomb was filled with hydrogen t o an 111itiul pressure of 100 atmospheres :ind 1ie:ttctl for 1 lir. a t 160-165'. Isolation of Hydrogenolysis Products.-Tlic reaction mixture, after cooling, mas filtered and the residue wlsllctl with 5076 dioxane. The mixture of filtrate and washings was concentrated in a rotary film evaporator until quite turbid and theii transferred t o a continuous liquid-liquid extractor, where a chloroform extraction was carried out for ca. 7 lir. Extraction in a separatory funnel invariabl!. led t o the formation of extremely intractable emulsions. Tlie chloroform extract was washed with a few milliliters of sodium bicarbonate solution and then extracted with 0 . 2 .L' sodium hydroxide. The extract, containing the plictiolic fraction, was acidified with hydrochloric acid and backextracted into chloroform. Evaporation of the chloroforili gave a light brown oil. A 7.5-g. lot of acid-washed diatomnceous earth (Celite 5 3 5 ) was mixed with 6.2 ml. of water, t h e mixture was slurried with n-butyl ether (alkali-washed and distilled from sodium) and packed with a plunger in a column 2 m i . in diameter. Above this w;rs packed 1.5 g. of anhydrous Celite 535 slurried in n-butyl ether. As much as possible of the chloroform-soluble oil W:IS taken up in a total of 1.2 1111. of warm water, and the solution was pipetted 011 t o the top of the column. Development was carried out with watersaturated n-butyl ether, and thirty 5-nil. fractions were

+

(9) E . W. C r a m p t o n a n d I.. A. M a y n a r d , J . Sritrilion, 16, 353 (1938). (10) C . P. Brewer, I,. M. Cooke and H. Hibbert, THISJ O U R N A L , T O , 57 (1918). (11) J . hT. Pepper and €1. Hibbert, i h i d . , T O , 67 (1948) ( 1 2 ) K . I l o z i n g o . "Organic Synthe.7ei." Coll \'