Carbohydrate Polymers 44 (2001) 319–324 www. elsevier. com/locate/carbpol Hydrolysis of diethyl diferulates by a tannase from Aspergillus oryzae ? M. -T. Garc? a-Conesa a,*, P. Ostergaard b, S. Kauppinen b, G. Williamson a a Phytochemicals Team, Division of Diet, Health & Consumer Sciences, Institute of Food Research, Norwich Research Park, Colney, Norwich NR4 7UA, UK. b ? Screening Biotechnology, Enzyme Research Novo Nordisk A/S, Novo Alle, bldg. 1BM1. 05 DK-2880 Bagsvaerd, Denmark. Abstract Diferulic acid forms cross-links in naturally occurring plant cell wall polymers such as arabinoxylans and pectins.
We have used model ethyl esteri? ed substrates to ? nd enzymes able to break these cross-links. A tannase from Aspergillus oryzae exhibited esterase activity on several synthetic ethyl esteri? ed diferulates. The ef? ciency of this esterase activity on most diferulates is low compared to that of a cinnamoyl esterase, FAEA, from Aspergillus niger. Of the diferulate substrates assayed, tannase was most ef? cient at hydrolysing the ? rst ester bond of the 5–5- type of dimer. Importantly and unlike the cinnamoyl esterase, tannase from A. ryzae is able to hydrolyse both ester bonds from the 8–5-benzofuran dimer, thus forming the corresponding free acid product. These results suggest that tannases may contribute to plant cell wall degradation by cleaving some of the cross-links existing between cell wall polymers. 2001 Elsevier Science Ltd. All rights reserved. Keywords: Aspergillus oryzae; cell wall polymers; Diethyl diferulates 1. Introduction Micro-organisms need to produce a combination of enzymes, primarily carbohydrases and ‘esterases’ (able to remove side chain substituents) that act synergistically, in order to increase digestibility of the plant cell wall.
Crosslinking of the cell wall polymers by ferulic acid dehydrodimers: 8–5-, 8–O–4-, 5–5- and 8–8- diFAs (Ralph, Quideau, Grabbber & Hat? eld, 1994) (Fig. 1), is a major obstacle which limits the accessibility of main chain-degrading enzymes to the structural polysaccharides and reduces cell wall digestibility (Grabber, Hat? eld & Ralph, 1998a,b). The breakage of one or both ester bonds from these dehydrodimer cross-links between plant cell wall polymers is essential for degradation of plant cell wall.
An esterase from Aspergillus niger, FAEA, is able to cleave both ester bonds from the 5–5-diferulate and the 8–O–4-diferulate and can release the corresponding free acids from xylanase solubilised plant material (Bartolome ? et al. , 1997; Kroon, Garc? a-Conesa, Fillingham, Hazlewood & Williamson, 1999). However, this esterase was able to cleave only one ester bond from the 8–5-benzofuran type of dimer and, although this may be suf? cient to facilitate the subsequent action of main polymers-degrading enzymes, it * Corresponding author. Tel. : 44-1603-507723; fax: 44-1603255267. ? E-mail address: maria. [email protected] ac. uk (M. -T. Garc? a-Conesa). 0144-8617/01/$ – see front matter PII: S0144-861 7(00)00248-4 ? was not possible to release the free acid (Garc? a-Conesa et al. , 1999). To date, there are no reports on the release of this dimer from plant material. Microbial esterases are a broad group of enzymes able to hydrolyse the ester bond of a variety of naturally occurring esters of ? avonoids and hydroxycinnamates present in plants. De? nition and characterisation of these enzymes are based mainly on their speci? city towards a range of synthetic and puri? ed substrates.
The availability of new substrates provides researchers with a tool to re-examine the speci? city of esterases for hydrolysing the variety of components present in plant cell walls. For example, synthetic dehydrodiferulates (Ralph et al. , 1994; Ralph, ? Garc? a-Conesa & Williamson, 1998) can be used to search for other esterases able to cleave plant cell wall cross-links, perhaps even more speci? c than FAEA from A. niger, and in particular, to search for esterases able to cleave the second ester bond from the 8–5-benzofuran diferulate, contributing to the digestion of the plant cell wall.
Tannin acyl hydrolases (EC 3. 1. 1. 20), commonly referred to as tannases, are inducible enzymes produced by fungi, mainly by Aspergillus and Penicillium species (Lekha & Lonsanne, 1997) but they have also been described in yeast (Aoki, Shinke & Nishira, 1976), bacteria (Deschamps, Otuk & Lebeault, 1983; Nelson, Pell, Scho? eld & Zinder, 1995) and plants (Niehaus & Gross, 1997). Tannases produced by microorganisms probably serve as a mode of invasion into the host plant by hydrolysing tannins that are present in many herbaceous and woody plants (Lekha & 2001 Elsevier Science Ltd. All rights reserved. 320 M. -T. Garc? a-Conesa et al. / Carbohydrate Polymers 44 (2001) 319–324 Ara Xyl Xyl Gal Xyl Xyl Xyl Xyl Ara O Xyl Xyl GlcA O Xyl Ara Gal O Xyl Xyl Xyl Xyl Ara O Xyl Xyl Xyl Xyl Xyl Xyl Ara O O 5-5 Xyl Xyl Xyl Xyl OCH3 Xyl Ara O O O OCH3 OH H3CO OH Xyl Xyl Xyl Ara O OCH3 O O OCH3 8-5-Benzofuran O Ara Xyl Ara Xyl Xyl O Ara Xyl Xyl Gal Xyl Ara OH Xyl Xyl OCH3 8-O-4 OH Fig. 1. Scheme showing the main diferulate cross-links described in plant cell walls (8–O–4-, 5–5- and 8–5-benzofuran diferulates). Lonsanne, 1997). Tannases have been mostly characterised by their activity on these complex polyphenolics.
Tannases are esterases able to hydrolyse the ‘ester’ bond (galloyl ester of an alcohol moiety) and the ‘depside’ bond (galloyl ester of gallic acid) in substrates such as tannic acid, methylgallate and m-digallic acid (Lekha & Lonsanne, 1997). Two separate isoenzymes, tannase I and tannase II, with esterase and depsidase activity, respectively, have been described in Aspergillus oryzae (Beverini & Metche, 1990) but more recently, some pure tannases exhibiting both activities have been obtained (Barthomeuf, Regerat & Pourrat, 1994; Niehaus & Gross, 1997).
Tannase has also been reported to act on substrates such as chlorogenic acid, ( )-epicatechin gallate and ( )-epigallocatechin-3-gallate (Lekha & Lonsanne, 1997). The activity of tannase is markedly inhibited in the presence of diisopropyl ? uorophosphate suggesting that this enzyme is a typical serine esterase (Barthomeuf et al. , 1994). It is possible that the esterase and/or depsidase activity of tannases is also able to cleave other phenolics present in plant cell walls such as the dehydrodimer cross-links. In this report, the activity of a tannase from A. oryzae on a range of synthetic ethyl esteri? ed diferulates is investigated.
The results are compared to those obtained with another esterase, FAEA from A. niger, assayed under the same conditions. (FAEA) was puri? ed from culture supernatants of A. niger grown on oat spelt xylan according to a published procedure (Faulds & Williamson, 1994). 2. 2. Substrates Diethyl 8–5-benzofuran diferulate was synthesised using peroxidase-H2O2 and puri? ed by ? ash chromatography (Ralph et al. , 1998). Ethyl 8–5-benzofuran diferulate was obtained from controlled enzymatic hydrolysis of diethyl 8– 5-benzofuran diferulate incubated with FAEA, and puri? cation of the monoester product by reverse-phase preparative ? hromatography (Garc? a-Conesa, Plumb, Kroon, Wallace & Williamson, 1997). Diethyl 5–5-diferulate was prepared from acetylated divanillin (Richitzenhain, 1949). Diethyl 8–O–4-diferulate was synthesised according to the method of Ralph et al. (1994) and puri? ed by reverse-phase prepara? tive chromatography (Garc? a-Conesa et al. , 1997). Epicatechin-3-gallate was a kind gift from Dr Alan Davies (Unilever, Colworth House, Sharnbrook, UK). Methyl ferulate was purchased from Apin Chemicals Ltd (Oxon, UK) and gallic acid (3,4,5-trihydroxybenzoic acid monohydrate) was obtained from Sigma (UK).
All other chemicals were of AnalaR or HPLC-grade purity. 2. 3. Protein determination Protein concentration of an aqueous solution of pure tannase from A. oryzae was estimated using a commercial assay kit (Coomassie Plus, Pierce, Rockford, IL, USA) based on the method of Bradford (1976). Bovine serum albumin (0–25 mg/ml in water) was used as standard protein. Protein concentration of a solution of pure FAEA was calculated by determining its absorbance at 280 nm 2. Experimental 2. 1. Enzymes Pure tannase from A. oryzae was kindly provided by Novo Nordisk (Denmark).
The cinnamoyl esterase ? M. -T. Garc? a-Conesa et al. / Carbohydrate Polymers 44 (2001) 319–324 321 OH OH HO O HO O OH OH OH OH OH Tannase (+)-Epicatechin OH O OH O OH pH 6. 0, 37C OH HO OH O OH Gallic acid (+)-Epicatechin 3-gallate Fig. 2. Hydrolysis of ( )-epicatechin 3-gallate by tannase to form ( )-epicatechin and gallic acid. (e : 43,660 M 1 cm 1, estimated on the basis of amino acid composition). 2. 4. Esterase activity Esterase activity was assayed as described by Faulds and Williamson (1991). Methyl ferulate (MFA; 1. 0 mM ? nal concentration) was prepared in Mops buffer (pH 6. ) and incubated with an appropriate amount of enzyme at 37 C. The reaction was terminated by the addition of acetic acid (pH 2. 0). Release of free ferulic acid was measured by reverse-phase HPLC with detection at 325 and 280 nm (Waldron, Parr, Ng & Ralph, 1996), and quanti? ed by reference to a ferulic acid calibration curve. All enzyme assays were performed in duplicate and with appropriate blanks to allow for correction for any background reactions. One unit (U) of activity was de? ned as the amount of enzyme releasing 1 mmol of ferulic acid per min at pH 6. 0 and 37 C. . 5. Tannase activity Tannase activity was determined using ( )-epicatechin 3-gallate (ECG) as substrate, which is hydrolysed by the enzyme to form gallic acid and ( )-epicatechin (Fig. 2). ( )-epicatechin 3-gallate (1. 8 mM ? nal concentration) was prepared in Mops buffer (pH 6. 0) and incubated with an appropriate amount of enzyme at 37 C. The hydrolysis process was terminated by boiling the reaction mixture for 10 min. Samples were ? ltered (0. 2 mm) and injected (100 ml) onto a Prodigy ODS(3) reverse-phase column ? (25 cm ? 4:6 mm i. d. , 100 A, Phenomenex, UK).
The mobile phase included solvent A, consisting of tetrahydrofuran (THF) in water (2:98), and solvent B, acetonitrile. The ? ow rate was 1. 0 ml/min. Separation was effected with gradient elution, starting at 0 min with 95% solvent A (5% solvent B) up to 7 min, decreasing to 0% solvent A (100% solvent B) at 10. 0 min, held isocratically at 0% solvent A, 100% solvent B for a further 7 min, and followed by reconditioning the column. Gallic acid was determined by HPLC/Diode Array with detection at 280 nm, where the absorbance of gallic acid shows its maximum, and was identi? d by spectroscopic analysis with diode array detection from 220 to 400 nm. Quanti? cation was by integration of peak areas at 280 nm, with reference to a calibration curve. The response factor calculated for gallic acid at 280 nm was 26; 200 ^ 1000 area units/ng; the standard curve was linear over the range: 50 to 3000 ng/100 ml. All enzyme assays were performed in duplicate and with appropriate blanks to allow for correction for any background reactions. One unit (U) of activity was de? ned as the amount of enzyme releasing 1 mmol of gallic acid per min at pH 6. and 37 C. 2. 6. Activity on synthetic diferulates The activity on esteri? ed diferulates was assayed using a range of synthetic ethyl diferulates. Reaction mixtures contained the esteri? ed substrate (? nal concentration: 0. 2 mM, diethyl 5–5-diferulate; 0. 3 mM, diethyl 8–O–4diferulate; 0. 05 mM, diethyl 8–5-benzofuran diferulate; 0. 2 mM, ethyl 8–5-benzofuran diferulate), in 20% (v/v) DMSO in buffer (100 mM Mops, ? nal pH 6. 1). The activity was initiated by addition of the enzyme and incubations were performed at 30 C.
Activity on the 5–5- and 8–5diesters was measured after 2 h of incubation whereas the activity on the 8–5- monoester and the 8–O–4- diester was determined after 16 h of incubation. The reaction was terminated by the addition of acetic acid (? nal pH 2. 0) and samples were ? ltered (0. 2 mm) prior to analysis by HPLC. The release of product was monitored by HPLC/Diode Array using the method described by Waldron et al. (1996) with detection at 280 and 325 nm. The response factors for the reaction products, monoesters and free acids, were calculated carefully from appropriate solutions of puri? ed 322 M. -T. Garc? a-Conesa et al. / Carbohydrate Polymers 44 (2001) 319–324 Table 1 Esterase and tannase activities for tannase from A. oryzae and FAEA from A. niger (activity values are expressed in nmol min 1 mg 1) Tannase (A. oryzae) Protein (mg/ml) Activity on MFA Activity on ( )-epicatechin 3-gallate a ity to hydrolyse the hydroxycinnamate ester linkage in MFA. 3. 2. Activity on synthetic esteri? ed diferulates HPLC analysis of reaction mixtures obtained after incubation of tannase with any of the four synthetic substrates tested indicated that, under the conditions of our assay, tannase from A. ryzae was active on the three ethyl-diesteri? ed substrates and on the ethyl-esteri? ed substrate, forming monoesters and free acid products. Tannase was able to cleave both ester bonds from the 5–5-diethyl diester forming two products that were identi? ed as 5–5-monoester and 5–5-diferulic acid. The 8–O–4- diester was a very poor substrate for tannase and only minor quantities of the two possible monoesters products were detected. No free 8–O– 4-diferulic acid was formed after 16 h of incubation. Tannase was also able to hydrolyse the 8–5-benzofuran diester forming predominantly a monoester product.
However, two other small peaks were detected and identi? ed on the basis of their retention time and spectra as a second monoester benzofuran product and as 8–5-benzofuran diferulic acid (Fig. 3). The formation of free 8–5benzofuran diferulic acid was con? rmed when the enzyme was incubated directly with the 8–5-benzofuran monoester (Fig. 4). These results indicated that tannase from A. oryzae was able to cleave both ester bonds from the benzofuran type of dimer releasing the free acid. The ability of tannase to hydrolyse the diferulate compounds was compared to that of FAEA from A. iger, under the same assay conditions. The values obtained for both enzymes are presented in Table 2 and are expressed in 8-5-Benzofuran diester 8-5-Benzofuran monoester Blank + Tannase 20 Minutes FAEA (A. niger) 1. 33 29400 Nd a 1. 98 2. 93 26300 Nd: not detected. synthetic compounds. All assays were performed in duplicate. Blanks containing the substrate plus the enzyme in 20% DMSO in buffer were incubated with acetic acid to correct for background peaks. 3. Results and discussion 3. 1. Esterase and tannase activities The activities on MFA and on ECG for the pure enzymes, tannase from A. ryzae and FAEA from A. niger, are shown in Table 1. Tannase exhibited a high activity on ECG (26. 3 U/mg) and some esterase activity on the hydroxycinnamate substrate ( 10 4-fold less than FAEA). However, no gallic acid was formed when incubating the ( )-epicatechin 3-gallate with FAEA, even after incubation for 16 h with a large amount of pure enzyme (80 mg of pure FAEA). These results indicated that FAEA was not able to cleave the ester bond of the ?avonoid molecule whereas tannase showed some abil- Relative abs (325 nm) 0 10 8-5-Benzofuran diFA 8-5-Benzofuran monoester 0 Fig. 3. HPLC elution pro? le at 325 nm of diethyl 8–5-benzofuran (diester substrate) and the products of its hydrolysis (monoester and free acid) by a tannase from A. oryzae. ? M. -T. Garc? a-Conesa et al. / Carbohydrate Polymers 44 (2001) 319–324 323 Relative abs (325 nm) 8-5-Benzofuran diFA 8-5-Benzofuran monoester Blank + Tannase 0 10 Minutes 20 30 Fig. 4. HPLC elution pro? le at 325 nm of ethyl 8–5-benzofuran (monoester substrate) and the product of its hydrolysis (free acid) by a tannase from A. oryzae. total nmol of product formed per mg of protein.
As the incubation periods were very long, 2 h for the diethyl 5– 5- and the diethyl 8–5-benzofuran diferulates and 16 h for the ethyl 8–5-benzofuran and the diethyl 8–O–4-diferulates, these values are not initial rates. The results show that tannase from A. oryzae, is able to hydrolyse the ester bonds from the various synthetic diferulate substrates tested although the ef? ciency of this enzyme acting on the diferulates is much lower than that of FAEA from A. niger. This could be related to a poorer recognition of the hydroxycinnamate moiety by tannase. FAEA is an esterase much more highly speci? for hydrolysis of some of the dehydrodiferulates cross-links. However, whilst FAEA is not active on the second ester bond from the 8– 5-benzofuran type of dimer and did not form 8–5-benzoTable 2 Hydrolysis of synthetic ethyl esteri? ed diferulates by a tannase from A. oryzae and by FAEA from A. niger. Values are expressed in total nmol of product per mg of protein. Incubation of diethyl 5–5- and diethyl 8–5benzofuran diferulates was for 2 h; incubation of ethyl 8–5-benzofuran and diethyl 8–O–4-diferulates was for 16 h Substrate Diethyl 5–5-diFA Tannase (A. oryzae) FAEA (A. iger) Nd b (monoester) 2760 (free acid) 1122 (monoester) Nf c (free acid) 5430 (monoester) 2820 (monoester) 870 (free acid) 166. 0 (monoester) a 12. 0 (free acid) Diethyl 8–5-benzofuran diFA 34. 0 (monoester) Ethyl 8–5-benzofuran diFA 32. 0 (free acid) Diethyl 8–O–4-diFA 15. 0 (monoester) 11. 0 (monoester) a furan diferulic acid (even after an extended period of incu? bation with large amounts of enzyme) (Garc? a-Conesa et al. , 1999), tannase was able to hydrolyse this second ester bond and formed some free acid (32 nmol of 8–5-benzofuran diferulic acid per mg of protein after 16 h of incubation).
Tannases show high speci? city for the phenolic site of the substrate. Substrates derived from benzoic acid carrying two ortho-hydroxyls seem to be better substrates than cinnamic acid derivatives (Niehaus & Gross, 1997; Scalbert, 1991). Esters with cinnamoyl residues, e. g. methyl cinnamates and sinapoyl glucose, were not accepted as substrates by a tannase puri? ed from oak leaves and chlorogenic acid was hydrolysed at a very low rate (Niehaus & Gross, 1997). In the present report, a tannase with some activity on hydroxycinnamates derivative substrates is presented.
Of the diferulate substrates assayed, tannase showed the best ef? ciency hydrolysing the ? rst ester bond of the 5–5- type of dimer, followed by hydrolysis of one ester bond from the 8–5benzofuran dimer, thus forming the corresponding monoester products. The 8–O–4- diester was a very poor substrate. The preferences shown by tannase are similar to those exhibited by FAEA but importantly and unlike the cinnamoyl esterase, tannase was able to recognise the second ester bond from the 8–5-benzofuran type of diferulate dimer. 4. Conclusions Tannase from A. oryzae exhibits some esterase activity on several synthetic ethyl esteri? d diferulates. The ef? ciency of this activity is lower than that of other esterases, i. e. FAEA from A. niger. However, tannase from A. oryzae is able to hydrolyse the second ester bond from the 8–5-benzofuran dimer forming the corresponding free acid (Fig. 5). The Product. Nd: not detected (under the described conditions, the monoester was all converted to free acid). c Nf: not formed (even after extended period of incubation with large amounts of enzyme). b 324 ? M. -T. Garc? a-Conesa et al. / Carbohydrate Polymers 44 (2001) 319–324 O O O O O O O O OMe H O H Tannase O O OMe H O H Tannase O
O OMe H O H OMe OH OH OMe OH OMe 8-5- Benzofuran diester 8-5- Benzofuran monoester 8-5- Benzofuran diFA Fig. 5. Scheme showing the activity of tannase from A. oryzae on both ester bonds from the 8–5-benzofuran coupled diferulate. results suggest that tannases may contribute to plant cell wall degradation by cleaving some of the cross-links existing between polymers. Acknowledgements The research was funded by the Biotechnology and Biological Sciences Research Council, UK, and the European Union (FAIR-CT95-0653 and FAIR-CT96-1099). References Aoki, K. , Shinke, R. , & Nishira, H. (1976). Puri? ation and some properties of yeast tannase. Agricultural and Biological Chemistry, 40, 79–85. Barthomeuf, C. , Regerat, F. , & Pourrat, H. (1994). Production, puri? cation and characterisation of a tannase from Aspergillus niger LCF 8. Journal of Fermentation and Bioengineering, 77, 320–323. Bartolome, B. , Faulds, C. B. , Kroon, P. A. , Waldron, K. W. , Gilbert, H. J. , Hazlewood, G. , & Williamson, G. (1997). An Aspergillus niger esterase (FAE-III) and a recombinant Pseudomonas ? uorescens subsp. cellulose esterase (XYLD) release a 5–5-ferulic dehydrodimer (‘diferulic acid’) from barley and wheat cell walls.
Applied Environmental Microbiology, 63, 208–212. Beverini, M. , & Metche, M. (1990). Identi? cation, puri? cation and physicochemical properties of tannase of Aspergillus oryzae. Sciences des Aliments, 10, 807–816. Bradford, M. (1976). A rapid and sensitive method for the quanti? cation of microgram quantities of protein utilising the principle of protein-dye binding. Analytical Biochemestry, 72, 248–254. Deschamps, A. M. , Otuk, G. , & Lebeault, J. M. (1983). Production of tannase and degradation of chestnut tannin by bacteria. Journal of Fermentation Technolology, 61, 55–59. Faulds, C.
B. , & Williamson, G. (1991). The puri? cation and characterisation of 4-hydroxy-3-methoxycinnamic (ferulic) acid esterase from Streptomyces olivochromogenes. Journal of General Microbiology, 137, 2339–2345. Faulds, C. B. , & Williamson, G. (1994). Puri? cation and characterisation of a ferulic acid esterase (FAE-III) from Aspergillus niger: speci? city for the phenolic moiety and binding to micro-crystalline cellulose. Microbiology, 144, 779–787. ? Garc? a-Conesa, M. T. , Plumb, G. W. , Kroon, P. A. , Wallace, G. , & Williamson, G. (1997). Antioxidant properties of ferulic acid dimers.
Redox Report, 3, 239. ? Garc? a-Conesa, M. T. , Kroon, P. A. , Ralph, J. , Mellon, F. A. , Colquhoun, I. J. , Saulnier, L. , Thibault, J-F. , & Williamson, G. (1999). An esterase from Aspergillus niger (FAEA) can break plant cell wall cross-links without release of free diferulic acids. European Journal of Biochemistry, 266, 644–652. Grabber, J. H. , Hat? eld, R. D. , & Ralph, J. (1998a). Diferulate cross-links impede the enzymatic degradation of non-ligni? ed maize walls. Journal of the Science of Food and Agriculture, 77, 193–200. Grabber, J. H. , Ralph, J. , & Hat? eld, R. D. 1998b). Diferulate cross-links limit the enzymatic degradation of synthetically ligni? ed primary walls of maize, . Journal of Agriculture and Food Chemistry, 46, 2609–2614. ? Kroon, P. A. , Garc? a-Conesa, M. T. , Fillingham, I. J. , Hazlewood, G. P. , & Williamson, G. (1999). Release of ferulic acid dehydrodimers from plant cell walls by feruloyl esterases. Journal of the Science of Food and Agriculture, 79, 428–434. Lekha, P. K. , & Lonsanne, B. K. (1997). Production and application of tannin acyl hydrolase: state of the art. Advances in Applied Microbiology, 44, 215–260. Nelson, E. Pell, A. N. , Scho? eld, P. , & Zinder, S. (1995). Isolation and characterisation of an anaerobic ruminal bacterium capable of degrading hydrolysable tannins. Applied Environmental Microbiology, 61, 3293–3298. Niehaus, J. U. , & Gross, G. G. (1997). A gallotannin degrading esterase from leaves of pedunculate oak. Phytochemistry, 45, 1555–1560. Ralph, J. , Quideau, S. , Grabber, J. H. , & Hat? eld, R. D. (1994). Identi? cation and synthesis of new ferulic acid dehydrodimers present in grass cell walls. Journal of the Chemical Society, Perkin Transactions 1, 3485–3498. ? Ralph, J. Garc? a-Conesa, M. T. , & Williamson, G. (1998). Simple preparation of 8–5-coupled diferulate. Journal of Agriculture and Food Chemistry, 46, 2531–2532. Richitzenhain, H. (1949). Enzymatische versuche zur entstehung des lignins. Chemische Berichte, 82, 447–453. Scalbert, A. (1991). Antimicrobial properties of tannins. Phytochemistry, 30, 3875–3883. Waldron, K. W. , Parr, A. J. , Ng, A. , & Ralph, J. (1996). Cell wall esteri? ed dimers: identi? cation and quanti? cation by reverse phase high performance liquid chromatography and diode array detection. Phytochemical Analysis, 7, 305–312.