Search

Mixed-linkage (1→3),(1→4)-β-d-glucan (MLG) is a biologically and technologically important hemicellulose, known to occur in three widely separated lineages: the Poales (including grasses and cereals), Equisetum (fern-allies), and some lichens e.g. Iceland moss (Cetraria islandica). Lichenase (E.C. 3.2.1.73) is widely assumed to hydrolyse all (1→4) bonds that immediately follow (1→3) bonds in MLG, generating predominantly the tetrasaccharide β-d-Glcp-(1→4)-β-d-Glcp-(1→4)-β-d-Glcp-(1→3)-d-Glc (G4G4G3G; MLG4), the corresponding trisaccharide (G4G3G; MLG3), and sometimes also laminaribiose (G3G; MLG2). The ratio of the oligosaccharides produced characterises each polysaccharide. We report here that digestion of MLG from barley (Hordeum vulgare), Equisetum arvense and C. islandica by Bacillus subtilis lichenase also yields the unexpectedly stable hexasaccharide, β-d-Glcp-(1→3)-β-d-Glcp-(1→4)-β-d-Glcp-(1→4)-β-d-Glcp-(1→4)-β-d-Glcp-(1→3)-d-Glc (G3G4G4G4G3G, i.e. MLG2-MLG4), identified by thin-layer chromatography, gel-permeation chromatography, HPLC (HPAEC), β-glucosidase digestion, (1)H/(13)C-NMR spectroscopy and mass spectrometry. On HPLC, G3G4G4G4G3G is the major constituent of a peak previously ascribed solely to the nonasaccharide G4G4G4G4G4G4G4G3G. Because it was widely presumed that lichenase would cleave G3G4G4G4G3G to MLG2+MLG4, our data both redefine the substrate specificity of Bacillus lichenase and show previous attempts to characterise MLGs by HPLC of lichenase-digests to be flawed. MLG2 subunits are particularly underestimated; often reported as negligible, they are here shown to be an appreciable constituent of MLGs from all three lineages. We also show that there is no appreciable yield of water-soluble lichenase products with DP>9; potential identities of products previously labelled DP>9 are suggested. Finally, this discovery also provides a opportunity to investigate the spatial distribution of subunits along the MLG chain. We show that MLG2 subunits in barley and Cetraria MLG are not randomly distributed, but predominantly found at the non-reducing end of MLG4 subunits.

Acid hydrolysis of cell wall-rich material from thalli of the hornwort Anthoceros caucasicus yielded substantial amounts of an unusual disaccharide ( 1 ). Hydrolysis of 1 yielded only GlcA, Gal and unhydrolysed 1 . Compound 1 was identified as α- d -Glc p A-(1→3)- l -Gal by 1 H and 13 C NMR spectroscopic analysis and by the susceptibility of its monosaccharide units to phosphorylation by enantiomer-specific kinases. Compound 1 was not detected in acid hydrolysates of other land plants including mosses, leafy and thalloid liverworts, lycopodiophytes and euphyllophytes; it was also absent from charophytes. The Anthoceros polysaccharide that yields 1 was partially extractable in cold aqueous buffer (pH 4.7) and Na 2 CO 3 , but not in EDTA or NaOH, suggesting that it was not a typical pectin or hemicellulose. The yield of 1 from various polysaccharide fractions correlated with the yields of Xyl, suggesting a previously unreported polymer containing d -GlcA, l -Gal and Xyl. The existence of a unique polysaccharide in an evolutionarily isolated plant ( Anthoceros ) supports the view that major steps in plant phylogeny were accompanied by significant changes in cell wall composition.

A high-throughput method is described by which Arabidopsis thaliana stems can be screened for variation in cell wall composition after hydrolysis with Driselase or trifluoroacetic acid (TFA). Driselase, a mixture of fungal enzymes, hydrolyses cellulose (to glucose) and all the major matrix polysaccharides (to monosaccharides and/or characteristic disaccharides); TFA hydrolyses the matrix polysaccharides, but not cellulose, to monosaccharides. Two different wild-type ecotypes, Columbia and Wassilewskija, showed only minor differences in wall carbohydrate composition. A small number of T-DNA-tagged populations that were screened contained individuals in which the proportion of cellulose, xyloglucan or xylan differed quantitatively from the wild-type. Differences from the wild-type were also observed in the susceptibility of the hemicelluloses to hydrolysis by Driselase, probably reflecting differences in wall architecture.

Driselase-digestion of cell walls from suspension-cultures of spinach (Spinacia oleracea L.), followed by anion-exchange chromatography, gel-permeation chromatography, preparative paper chromatography and preparative paper electrophoresis, yielded ten uronic acid-containing products in addition to free galacturonic acid (GalA). These included 4-O-methylglucuronic acid, alpha-L-rhamnopyranosyl-(1-->4)-D-glucuronic acid and several oligosaccharides containing GalA residues. The structures were unambiguously determined by a combination of 1- and 2-dimensional NMR spectroscopic techniques. Five of the six homogalacturonan-derived oligosaccharides purified contained 3-O-acetyl-GalA residues; however, methyl-esterified GalA residues occurred adjacent to both 2-O-acetyl-GalA and 3-O-acetyl-GalA residues. An acetylated, rhamnogalacturonan-I-derived oligosaccharide that was purified also contained 3-O-acetyl-GalA residues. Taken together with published data, our findings indicate considerable diversity in the patterns of pectin esterification. The implications for the action of pectin esterases are discussed.

Xyloglucan plays an important structural role in primary cell walls, possibly tethering adjacent microfibrils and restraining cell expansion. There is therefore considerable interest in understanding the role of xyloglucan endotransglucosylase/hydrolases (XTHs), which are encoded in Arabidopsis by a 33-member gene family. We compared the key catalytic properties of two very different Arabidopsis XTHs (heterologously produced in Pichia), both of which are aluminium-repressed. Reductively tritiated oligosaccharides of xyloglucan were used as model acceptor substrates. Untransformed Pichia produced no xyloglucan-acting enzymes; therefore purification of the XTHs was unnecessary. XTH15, a classical group-I/II XTH, had high XET and undetectable XEH activity in vitro; its XET Km values were 31 μM XXXGol (acceptor substrate) and 2.9 mg/ml xyloglucan (donor substrate). In contrast, XTH31, a group-III-A XTH, showed predominant XEH activity and only slight XET activity in vitro; its XET Km was 86 μM XXXGol (acceptor), indicating a low affinity of this predominantly hydrolytic protein for a transglycosylation acceptor substrate. The Km of XTH31’s XEH activity was 1.6 mg/ml xyloglucan. For both proteins, the preferred XET acceptor substrate, among five cellotetraitol-based oligosaccharides tested, was XXXGol. XTH31’s XET activity was strongly compromised when the second Xyl residue was galactosylated. XTH15’s XET activity, in contrast, tolerated substitution at the second Xyl residue. The two enzymes also showed different pH preferences, XTH31 exhibiting an unusually low pH optimum and XTH15 an unusually broad optimum. XTH31’s hydrolase activity increased almost linearly with decreasing pH in the apoplastic range, 6.2–4.5, consistent with a possible role in ‘acid growth’. In conclusion, these two Al3+-repressed XTHs differ, in several important enzymic features, from other members of the Arabidopsis XTH family.

Four isoenzymes of xyloglucan endotransglycosylase (XET; EC 2.4.1.207) were isolated from sprouting mung bean seedlings (M35, M45, M55a, M55b) and two from cauliflower florets (C30, C45). Purification in each case was by ammonium sulphate precipitation, reversible formation of a covalent xyloglucan–enzyme complex, and cation-exchange chromatography. The isoenzymes differed in pH optimum (range 5.0–6.5), K m for the nonasaccharide XLLGol (Gal 2 .Xyl 3 .Glc 3 .glucitol) as acceptor substrate, ability to utilise diverse oligosaccharides as acceptor substrate, and ability to bind to carboxymethyl-cellulose (and thus possibly to other polyanions such as pectin in the cell wall). None of the isoenzymes was particularly cold-tolerant, unlike one XET (TCH4) of Arabidopsis . The two cauliflower isoenzymes had higher K m values for XLLGol (70–130 μM) than the four mung bean isoenzymes (16–35 μM). We suggest that this difference is related to the major roles of the XETs in these two tissues: integration of new xyloglucan into the walls of the densely cytoplasmic cauliflower florets, and re-structuring of existing wall material in the rapidly vacuolating bean shoots.

Four peroxidases and four laccases were compared as to reaction rates catalysed with six phenolic substrates of relevance to the plant cell wall. When each phenolic substrate was tested at 670 μM and pH 6.0, in the presence of 670 μM H 2 O 2 or ∼270 μM O 2 as the electron acceptor, all the peroxidases and laccases had similar substrate preferences: reaction rates were in the order sinapyl>coniferyl> p -coumaryl alcohols, and feruloyl> p -coumaroyl esters. Specific activities were in the order basic peroxidase>acidic peroxidase≫laccase. The data are consistent with the view that peroxidases rather than laccases play a major role in phenolic cross-linking in the cell wall.

Cell-suspension cultures of rose (Rosa sp.) accumulated approximately micromolar concentrations of a novel, extracellular, non-reducing trisaccharide, characterised here as α- d -mannopyranosyl-(1→4)-α- d -glucuronopyranosyl-(1→2)-myo-inositol (MGI). Acid hydrolysis of [glucuronic acid-14C]MGI, isolated after incubation of the cells with d -[6-14C]glucuronic acid, readily yielded [14C]glucuronic acid-inositol, which was relatively resistant to further hydrolysis, as revealed by chromatography and electrophoresis. The complete structure and stereochemistry of MGI was elucidated using one and two dimensional NMR spectroscopic techniques. The structure of MGI suggests a biosynthetic origin from a membrane-bound glycophosphosphingolipid (GPS). We propose that GPSs represent a hitherto unrecognised source of oligosaccharin signalling molecules in plants.

Three representatives of a novel class of amide (isopeptide) glycoconjugates have been synthesised: Nα - d -galacturonoyl- l -lysine and Ne - d -galacturonoyl- l -lysine and Ne - d -polygalacturonoyl- l -lysine. Galacturonoyl–lysine amide bonds were labile in 2 M trifluoroacetic acid at 120° and in alkali, but relatively stable in cold acid. The amide bonds were resistant to digestion by Driselase, Pronase and trypsin. The polysaccharide backbone of Ne - d -polygalacturonoyl- l -lysine was hydrolysed by Driselase to yield two major ninhydrin-positive compounds which were shown by 1 H and 13 C NMR spectroscopy to be tri- and tetra- α -(1→4)- d -galacturonoyl- l -lysines. To investigate the possible natural occurrence of N -galacturonoyl isopeptide bonds, we fed cell-suspension cultures of spinach and tomato with d -[6- 14 C]glucuronic acid, which radio-labels pectic polysaccharides. The radioactive cell walls were digested with, sequentially, Driselase, mild acid, and proteinases. On electrophoresis at pH 2.0, several of the radioactive digestion-products were cathodic. Some of the cathodic products yielded [ 14 C]galacturonic acid upon complete acid hydrolysis. The existence of these products is compatible with the presence of novel N -galacturonoyl isopeptide bonds, which could serve as cross-links in plant cell walls.

Partial acid hydrolysis of cell wall material from Festuca arundinacea cell cultures yielded a novel O-feruloylated trisaccharide (3). Treatment of 3 with Driselase, which contains β- but not α- d -xylosidase, released xylose plus the known compound, β- d -xylopyranosyl -(1 → 2)-(5-O- feruloyl )- l -arabinose . Since 3 contained one NaIO4-resistant xylose residue, it was concluded to be β- d -xylopyranosyl-(1 → 3)-β- d -xylopyranosyl -(1 → 2)-(5-O- feruloyl )- l -arabinose . Partial acid hydrolysis of Festuca cell walls also yielded several higher-Mr feruloylated oligosaccharides, including a feruloylated pentasaccharide, 4 (sugar composition: Ara + Xyl2 + two non-pentose residues) and a feruloylated heptasaccharide, 5 (Ara + Xyl3 + three non-pentose residues). Compounds 4 and 5 were endogenously O-acetylated but 3 was not. Similar or identical compounds were found in hydrolysates of 20 additional species of the Gramineae. These products represent a series of complex side-chains which, in vivo, are attached via Araf residues to the parent xylan. Their possible biological roles are discussed.

Graminaceous cell walls contain arabinoxylans in which some of the Araf residues are 5-O-feruloylated. In the present study, mild acid hydrolysis of (pentosyl-3H)-labelled Festuca arundinacea cell walls yielded the well-characterized 5-O- Fer- l -Ara (1) and at least seven new (E)-feruloylated oligosaccharides, 2–8. Compound 2 was a feruloylated disaccharide, Fer -( d -Xyl p→ l -Ara ) ; electrophoresis in molybdate buffer, Smith degradation and methylation analysis showed a (1 → 2)-linkage. 1H and 13C NMR data and susceptibility to commercial β- d -xylosidase indicated a β-linkage. Graded alkaline hydrolysis of 2 showed no evidence for more than one ester group. Hydrolysis with commercial β-xylosidase yielded 5-O- feruloyl- l -arabinose . Compound 2 was thus 2-O-β- d -xylopyranosyl -(5-O- feruloyl )- l -arabinose . We propose that 2 was furanose-linked to a xylan backbone in the intact polymer. Compounds 3–8 also appeared to be O-feruloyl oligosaccharides with a former Araf residue at the reducing terminus. Mild acid hydrolysis of cell walls from twenty other species of grass and one palm also yielded compounds 1–3, and most or all of 4–8. Therefore, these complex feruloylated side-chains of arabinoxylans are widespread or universal in the Gramineae. Their possible biological roles are discussed.

Membrane preparations from cell-suspension cultures of maize produced 14C-polymeric material from [methoxy-14C]feruloyl-CoA under conditions reported to be suitable for the assay of feruloyl-CoA: polysaccharide feruloyltransferase activity. The major 14C-polymer formed did not contain O-[ 14 C ] feruloyl-arabinofuranosyl groups. Production of 14C-polymer was prevented by catalase, indicating a requirement for endogenous H2O2, and was Co2+-dependent but only in the presence of dithiothreitol (DTT). The membranes exhibited peroxidase activity with guaiacol and H2O2; this reaction was blocked by DTT. It is concluded that peroxidases and H2O2, endogenous to the isolated membranes, caused oxidative polymerization of [14C]feruloyl groups of feruloyl-CoA; DTT prevented polymerization by acting as a competitive substrate for peroxidases, and Co2+ restored polymerization by forming an insoluble complex with the thiol.

Angiosperms possess a retaining trans-α-xylosidase activity that catalyses the inter-molecular transfer of xylose residues between xyloglucan structures. To identify the linkage of the newly transferred α-xylose residue, we used [Xyl-(3)H]XXXG (xyloglucan heptasaccharide) as donor substrate and reductively-aminated xyloglucan oligosaccharides (XGO-NH(2)) as acceptor. Asparagus officinalis enzyme extracts generated cationic radioactive products ([(3)H]Xyl·XGO-NH(2)) that were Driselase-digestible to a neutral trisaccharide containing an α-[(3)H]xylose residue. After borohydride reduction, the trimer exhibited high molybdate-affinity, indicating xylobiosyl-(1→6)-glucitol rather than a di-xylosylated glucitol. Thus the trans-α-xylosidase had grafted an additional α-[(3)H]xylose residue onto the xylose of an isoprimeverose unit. The trisaccharide was rapidly acetolysed to an α-[(3)H]xylobiose, confirming the presence of an acetolysis-labile (1→6)-bond. The α-[(3)H]xylobiitol formed by reduction of this α-[(3)H]xylobiose had low molybdate-affinity, indicating a (1→2) or (1→4) linkage. In NaOH, the α-[(3)H]xylobiose underwent alkaline peeling at the moderate rate characteristic of a (1→4)-disaccharide. Finally, we synthesised eight non-radioactive xylobioses [α and β; (1↔1), (1→2), (1→3) and (1→4)] and found that the [(3)H]xylobiose co-chromatographed only with (1→4)-α-xylobiose. We conclude that Asparagus trans-α-xylosidase activity generates a novel xyloglucan building block, α-d-Xylp-(1→4)-α-d-Xylp-(1→6)-d-Glc (abbreviation: 'V'). Modifying xyloglucan structures in this way may alter oligosaccharin activities, or change their suitability as acceptor substrates for xyloglucan endotransglucosylase (XET) activity.

Incubation of non-radioactive pectic polysaccharides with α-(1→4)-Oligo- d -[6- 14 C]galacturonides (degree of polymerization two-12) in the presence of various enzyme preparations at pH 4–8 did not yield any detectable 14 C- transglycosylation products. Polysaccharides tested were polygalacturonate, citrus pectin, rhamnogalacturonan-II, soluble polysaccharides from the culture filtrates of rose ( Rosa sp.) cell suspensions, partially degraded pectic polysaccharides extracted by autoclaving the walls of cultured rose cells and citrus pectin after treatment with pectin methylesterase. Potential enzyme sources tested were culture filtrates and ionically bound wall proteins from rose and sycamore ( Acer pseudoplatanus ) cell suspension cultures, and extracts of several ripening fruits. In addition, in experiments designed to detect transglycosylation catalysed by an inextractable or labile apoplastic enzyme, 3 H from exogenous [reducing terminus-1- 3 H]pentagalacturonide was not incorporated into either wall-bound or soluble extracellular polymers by living cell suspension cultures. Thus, no evidence could be obtained for transglycosylation with oligogalacturonides as either donor or acceptor substrates under experimental regimes similar to those previously used to demonstrate transglycosylation with xyloglucan oligosaccharides.

Two new reducing glycoconjugates [ N - d -galacturonoyl-putrescinamide (GalA–Put) and N , N ′-di- d -galacturonoyl-putrescinamide (GalA–Put–GalA)] and homogalacturonan–putrescine (GalA n –Put) conjugates were synthesised as model compounds representing possible amide (isopeptide) linkage points between a polyamine and either one or two pectic galacturonate residues. The amide bond(s) were stable to cold acid and alkali (2 M TFA and 0.1 M NaOH at 25 °C) but rapidly hydrolysed by these agents at 100 °C. The amide bond(s) were resistant to Driselase and to all proteinases tested, although Driselase digested GalA n –Put, releasing fragments such as GalA 3 –Put–GalA 3 . To trace the possible formation of GalA–polyamine amide bonds in vivo, we fed Arabidopsis and rose cell-cultures and chickpea internodes with [ 14 C]Put. About 20% of the 14 C taken up was released as 14 CO 2 , indicating some catabolism. An additional ∼73% of the 14 C taken up (in Arabidopsis ), or ∼21% (in rose), became ethanol-insoluble, superficially suggestive of polysaccharide–Put covalent bonding. However, much of the ethanol-inextractable 14 C was subsequently extractable by acidified phenol or by cold 1 M TFA. The small proportion of radioactive material that stayed insoluble in both phenol and TFA was hydrolysable by Driselase or hot 6 M HCl, yielding 14 C-oligopeptides and/or amino acids (including Asp, Glu, Gly, Ala and Val); no free 14 C-polyamines were released by hot HCl. We conclude that if pectin–polyamine amide bonds are present, they are a very minor component of the cell walls of cultured rose and Arabidopsis cells and chickpea internodes.

Driselase-digestion of cell walls from suspension-cultures of spinach (Spinacia oleracea L.), followed by anion-exchange chromatography, gel-permeation chromatography, preparative paper chromatography and preparative paper electrophoresis, yielded ten uronic acid-containing products in addition to free galacturonic acid (GalA). These included 4-O-methylglucuronic acid, alpha-L-rhamnopyranosyl-(1--4)-D-glucuronic acid and several oligosaccharides containing GalA residues. The structures were unambiguously determined by a combination of 1- and 2-dimensional NMR spectroscopic techniques. Five of the six homogalacturonan-derived oligosaccharides purified contained 3-O-acetyl-GalA residues; however, methyl-esterified GalA residues occurred adjacent to both 2-O-acetyl-GalA and 3-O-acetyl-GalA residues. An acetylated, rhamnogalacturonan-I-derived oligosaccharide that was purified also contained 3-O-acetyl-GalA residues. Taken together with published data, our findings indicate considerable diversity in the patterns of pectin esterification. The implications for the action of pectin esterases are discussed.

Acid hydrolysis of cell wall-rich material from young leaves of the lycophyte Selaginella apoda (L.) Spring yielded substantial amounts of 3-O-methyl-D-galactose (1) in addition to the usual major monosaccharides (glucose, galactose, arabinose, xylose and galacturonic acid). The yield of 1 approximately equalled that of galacturonic acid. Compound 1 was identified as 3-O-methylgalactose by its 1H and 13C NMR spectra, and shown to be the D-enantiomer by its susceptibility to D-galactose oxidase. Compound 1 was detected in acid hydrolysates of the alcohol-insoluble residues from young leaves of all lycophytes tested, both homosporous (Lycopodium, Huperzia and Diphasiastrum) and heterosporous (Selaginella). It was not detectable in the charophyte green algae Coleochaete scutata, Chara coralina or Klebsormidium flaccidum, any bryophytes [a hornwort (Anthoceros), four liverworts and three mosses], or any euphyllophytes [a psilopsid (Psilotum), a horsetail (Equisetum), eusporangiate and leptosporangiate ferns, the gymnosperm Gnetum, and diverse angiosperms]. A high content of 1 is thus an autapomorphy of the lycophytes.

Alcohol-insoluble residues (AIRs) from Festuca and Zea cell cultures contained 7.4 and 35 nmol esterified ferulate mg −1, respectively. Driselase solubilised 79% of the feruloylated material from both AIRs. Of the feruloyl esters solubilised from Festuca and Zea AIRs, 72 and 56% respectively were small enough to be mobile on paper chromatography. The major feruloylated product of Zea AIR was the known 5-O-feruloyl-α-l-Araf-(1 → 3)-β-d-Xylp-(1 → 4)- d-Xyl (Fer-Ara-Xyl-Xyl). In contrast, the smallest major feruloylated product of Festuca AIR was a feruloyl pentasaccharide (3) containing 3 Xyl, 1 Ara and 1 non-pentose residue (NPR). The Ara and two of the three Xyl groups of 3 were resistant to NaIO4. Mild acid hydrolysis of 3 gave xylobiose, a feruloyl trisaccharide and β-d-Xylp-(1 → 2)-(5-O- feruloyl)-l-Ara. Compound 3 was therefore NPR-(1 → 3)-β-d-Xylp-(1 → 2)-(5-O-feruloyl)-β-l--Ara f-(1 → 3)-β-d-Xylp-(1 → 4)-d-Xyl. We conclude that the complex feruloyl oligosaccharide side-chains of Festuca arabinoxylan do not protect the polysaccharide against hydrolysis by the fungal glycanases present in Driselase. © 1997 Elsevier Science Ltd. All rights reserved

[ 14 C]Cinnamate was taken up very rapidly by cultured spinach cells and completely incorporated into low-MW conjugates within 20 min. The 14 C-labelled products were similar whether the [ 14 C]cinnamate was supplied continuously over a period of hours via a peristaltic pump or instantaneously. Radioactivity was slowly recruited from the low-MW pool into aromatic components of the cell-wall fraction. Saponification of the radioactive wall fraction yielded, in addition to radioactive ferulate and p -coumarate, large amounts of ethyl acetate-soluble radioactive material with the properties of oxidatively coupled phenols. The coupled material was associated with the most highly ‘Driselase’-resistant fractions of the cell wall. In contrast, ‘Driselase’ released most of the wall's ferulate and p -coumarate on disaccharide fragments. It is suggested that the oxidatively coupled phenols are formed from simpler phenols by peroxidase and that they cross-link the polysaccharides to which they are attached, making these polysaccharides relatively ‘Driselase’-resistant.

Gibberellic acid (GA) promoted cell expansion in a suspension culture of spinach. Preceding this response, weakly acidic pectinic acid was released into the medium. GA did not alter the concentration of sorbitol (0.7 M) needed to prevent growth, and sorbitol at this concentration did not block the action of GA upon the release of pectinic acid. GA suppressed the release into the medium of a peroxidase of MW ca 40 000 and favoured the intracellular accumulation of a polypeptide of similar MW. All the responses to GA were antagonized by 2,4-dichlorophenoxyacetic acid but not by kinetin. A hypothesis is presented in which the promotion of growth by GA is mediated by peroxidase, its phenolic substrates and the cell-wall pectic polysaccharides.