Ascorbate degradation: pathways, products, and possibilities.

A role for l-ascorbate as the precursor of several plant compounds adds to its already broad metabolic utility. There are many examples of plant species in which oxalate and l-threonate are formed from l-ascorbate breakdown, and a number of roles have been proposed for this: structural, physiological, and biochemical. On the other hand, the synthesis of l-tartrate from l-ascorbate remains limited to a very few species, amongst which we must be grateful to count the domesticated grapevine Vitis vinifera and its relatives on which wine production is based. Pathways for the degradation of ascorbate were first proposed ~50 years ago and have formed the basis of more recent biochemical and molecular analyses. The present review seeks to summarize some of these findings and to propose opportunities for future research.

Fruit softening: evidence for rhamnogalacturonan lyase action in vivo in ripe fruit cell walls.

BACKGROUND AND AIMS: The softening of ripening fruit involves partial depolymerization of cell-wall pectin by three types of reaction: enzymic hydrolysis, enzymic elimination (lyase-catalysed) and non-enzymic oxidative scission. Two known lyase activities are pectate lyase and rhamnogalacturonan lyase (RGL), potentially causing mid-chain cleavage of homogalacturonan and rhamnogalacturonan-I (RG-I) domains of pectin respectively. However, the important biological question of whether RGL exhibits action in vivo had not been tested. METHODS: We developed a method for specifically and sensitively detecting in-vivo RGL products, based on Driselase digestion of cell walls and detection of a characteristic unsaturated 'fingerprint' product (tetrasaccharide) of RGL action. KEY RESULTS: In model experiments, potato RG-I that had been partially cleaved in vitro by commercial RGL was digested by Driselase, releasing an unsaturated tetrasaccharide ('ΔUA-Rha-GalA-Rha'), taken as diagnostic of RGL action. This highly acidic fingerprint compound was separated from monosaccharides (galacturonate, galactose, rhamnose, etc.) by electrophoresis at pH 2, then separated from ΔUA-GalA (the fingerprint of pectate lyase action) by thin-layer chromatography. The 'ΔUA-Rha-GalA-Rha' was confirmed as 4-deoxy-ß-l-threo-hex-4-enopyranuronosyl-(1→2)-l-rhamnosyl-(1→4)-d-galacturonosyl-(1→2)-l-rhamnose by mass spectrometry and acid hydrolysis. Driselase digestion of cell walls from diverse ripe fruits [date, sea buckthorn, cranberry, yew (arils), mango, plum, blackberry, apple, pear and strawberry] yielded the same fingerprint compound, demonstrating that RGL had been acting in vivo in these fruits prior to harvest. The 'fingerprint' : (galacturonate + rhamnose) ratio in digests from ripe dates was approximately 1 : 72 (mol/mol), indicating that ~1.4 % of the backbone Rha→GalA bonds in endogenous RG-I had been cleaved by in-vivo RGL action. CONCLUSIONS: The results provide the first demonstration that RGL, previously known from studies of fruit gene expression, proteomic studies and in-vitro enzyme activity, exhibits enzyme action in the walls of soft fruits and may thus be proposed to contribute to fruit softening.

Allelochemical root-growth inhibitors in low-molecular-weight cress-seed exudate.

BACKGROUND AND AIMS: Cress seeds release allelochemicals that over-stimulate the elongation of hypocotyls of neighbouring (potentially competing) seedlings and inhibit their root growth. The hypocotyl promoter is potassium, but the root inhibitor was unidentified; its nature is investigated here. METHODS: Low-molecular-weight cress-seed exudate (LCSE) from imbibed Lepidium sativum seeds was fractionated by phase partitioning, paper chromatography, high-voltage electrophoresis and gel-permeation chromatography (on Bio-Gel P-2). Fractions, compared with pure potassium salts, were bioassayed for effects on Amaranthus caudatus seedling growth in the dark for 4 days. KEY RESULTS: The LCSE robustly promoted amaranth hypocotyl elongation and inhibited root growth. The hypocotyl inhibitor was non-volatile, hot acid stable, hydrophilic and resistant to incineration, as expected for K+. The root inhibitor(s) had similar properties but were organic (activity lost on incineration). The root inhibitor(s) remained in the aqueous phase (at pH 2.0, 6.5 and 9.0) when partitioned against butan-1-ol or toluene, and were thus hydrophilic. Activity was diminished after electrophoresis, but the remaining root inhibitors were neutral. They became undetectable after paper chromatography; therefore, they probably comprised multiple compounds, which separated from each other, in part, during fractionation. On gel-permeation chromatography, the root inhibitor co-eluted with hexoses. CONCLUSIONS: Cress-seed allelochemicals inhibiting root growth are different from the agent (K+) that over-stimulates hypocotyl elongation and the former probably comprise a mixture of small, non-volatile, hydrophilic, organic substances. Abundant components identified chromatographically and by electrophoresis in cress-seed exudate fitting this description include glucose, fructose, sucrose and galacturonic acid. However, none of these sugars co-chromatographed and co-electrophoresed with the root-inhibitory principle of LCSE, and none of them (in pure form at naturally occurring concentrations) inhibited root growth. We conclude that the root-inhibiting allelochemicals of cress-seed exudate remain unidentified.

Arabinogalactan-Proteins as Boron-Acting Enzymes, Cross-Linking the Rhamnogalacturonan-II Domains of Pectin.

Most pectic rhamnogalacturonan-II (RG-II) domains in plant cell walls are borate-bridged dimers. However, the sub-cellular locations, pH dependence, reversibility and biocatalyst involvement in borate bridging remain uncertain. Experiments discussed here explored these questions, utilising suspension-cultured plant cells. In-vivo pulse radiolabelling showed that most RG-II domains dimerise extremely quickly (<4 min after biosynthesis, thus while still intraprotoplasmic). This tallies with the finding that boron withdrawal causes cell wall weakening within 10-20 min, and supports a previously proposed biological role for boron/RG-II complexes specifically at the wall/membrane interface. We also discuss RG-II monomer ↔ dimer interconversion as monitored in vitro using gel electrophoresis and a novel thin-layer chromatography method to resolve monomers and dimers. Physiologically relevant acidity did not monomerise dimers, thus boron bridge breaking cannot be a wall-loosening mechanism in 'acid growth'; nevertheless, recently discovered RG-II trimers and tetramers are unstable and may thus underpin reversible wall loosening. Dimerising monomers in vitro by B(OH)3 required the simultaneous presence of RG-II-binding 'chaperones': co-ordinately binding metals and/or ionically binding cationic peptides. Natural chaperones of the latter type include highly basic arabinogalactan protein fragments, e.g., KHKRKHKHKRHHH, which catalyse a reaction [2 RG-II + B(OH)3 → RG-II-B-RG-II], suggesting that plants can 'enzymically' metabolise boron.

Same but different - pseudo-pectin in the charophytic alga Chlorokybus atmophyticus.

All land-plant cell walls possess hemicelluloses, cellulose and anionic pectin. The walls of their cousins, the charophytic algae, exhibit some similarities to land plants' but also major differences. Charophyte 'pectins' are extractable by conventional land-plant methods, although they differ significantly in composition. Here, we explore 'pectins' of an early-diverging charophyte, Chlorokybus atmophyticus, characterising the anionic polysaccharides that may be comparable to 'pectins' in other streptophytes. Chlorokybus 'pectin' was anionic and upon acid hydrolysis gave GlcA, GalA and sulphate, plus neutral sugars (Ara≈Glc>Gal>Xyl); Rha was undetectable. Most Gal was the l-enantiomer. A relatively acid-resistant disaccharide was characterised as ß-d-GlcA-(1→4)-l-Gal. Two Chlorokybus 'pectin' fractions, separable by anion-exchange chromatography, had similar sugar compositions but different sulphate-ester contents. No sugars were released from Chlorokybus 'pectin' by several endo-hydrolases [(1,5)-α-l-arabinanase, (1,4)-ß-d-galactanase, (1,4)-ß-d-xylanase, endo-polygalacturonase] and exo-hydrolases [α- and ß-d-galactosidases, α-(1,6)-d-xylosidase]. 'Driselase', which hydrolyses most land-plant cell wall polysaccharides to mono- and disaccharides, released no sugars except traces of starch-derived Glc. Thus, the Ara, Gal, Xyl and GalA of Chlorokybus 'pectin' were not non-reducing termini with configurations familiar from land-plant polysaccharides (α-l-Araf, α- and ß-d-Galp, α- and ß-d-Xylp and α-d-GalpA), nor mid-chain residues of α-(1→5)-l-arabinan, ß-(1→4)-d-galactan, ß-(1→4)-d-xylan or α-(1→4)-d-galacturonan. In conclusion, Chlorokybus possesses anionic 'pectic' polysaccharides, possibly fulfilling pectic roles but differing fundamentally from land-plant pectin. Thus, the evolution of land-plant pectin since the last common ancestor of Chlorokybus and land plants is a long and meandering path involving loss of sulphate, most l-Gal and most d-GlcA; re-configuration of Ara, Xyl and GalA; and gain of Rha.

Biochemical characterisation of cellulose and cell-wall-matrix polysaccharides in variously oxidised sugar-beet pulp preparations differing in viscosity.

Sugar-beet pulp (SBP) is an abundant, cellulose-rich, non-food by-product of agriculture. Oxidised SBP (oP) has valuable viscosity attributes, and different oxidation protocols yield higher- or lower-viscosity oP. We investigated how SBP polysaccharides change during oxidation, since these changes must define oP quality. Oxidation solubilised much pectin and hemicellulose; however, most cellulose stayed insoluble. Fresh SBP contains negligible 'hemicellulose a' (=alkali-extractable polysaccharides that precipitate upon acidification), but oxidation created abundant glucose-rich 'hemicellulose a' from SBP cellulose. We propose that the cellulose acquired COOH groups, conferring alkali-extractability and admitting more water, thereby augmenting viscosity. The pectin and hemicellulose molecules that were retained during oxidation had been partially depolymerised, and their median Mr correlated negatively with oP viscosity. We developed a novel procedure to explore cellulose's permeability by measuring the ingress of tritium from [3H]water into microfibrils and its retention during desiccation. In high-crystallinity Avicel, 75 % of the cellulose's OH groups were inaccessible to [3H]water, whereas filter-paper cellulose acquired the theoretical maximum 3H, indicating an open structure. Retention of 3H by oP preparations correlated positively with viscosity, indicating that increased cellulose accessibility generates a viscous oP. In conclusion, depolymerisation and solubilisation of matrix polysaccharides, accompanied by increasing water-accessibility of cellulose, enhanced SBP's viscosity.

Making and breaking of boron bridges in the pectic domain rhamnogalacturonan-II at apoplastic pH in vivo and in vitro.

Cross-linking of the cell-wall pectin domain rhamnogalacturonan-II (RG-II) via boron bridges between apiose residues is essential for normal plant growth and development, but little is known about its mechanism or reversibility. We characterized the making and breaking of boron bridges in vivo and in vitro at 'apoplastic' pH. RG-II (13-26 µm) was incubated in living Rosa cell cultures and cell-free media with and without 1.2 mm H3 BO3 and cationic chaperones (Ca2+ , Pb2+ , polyhistidine, or arabinogalactan-protein oligopeptides). The cross-linking status of RG-II was monitored electrophoretically. Dimeric RG-II was stable at pH 2.0-7.0 in vivo and in vitro. In-vitro dimerization required a 'catalytic' cation at all pHs tested (1.75-7.0); thus, merely neutralizing the negative charge of RG-II (at pH 1.75) does not enable boron bridging. Pb2+ (20-2500 µm) was highly effective at pH 1.75-4.0, but not 4.75-7.0. Cationic peptides were effective at approximately 1-30 µm; higher concentrations caused less dimerization, probably because two RG-IIs then rarely bonded to the same peptide molecule. Peptides were ineffective at pH 1.75, their pH optimum being 2.5-4.75. d-Apiose (>40 mm) blocked RG-II dimerization in vitro, but did not cleave existing boron bridges. Rosa cells did not take up d-[U-14 C]apiose; therefore, exogenous apiose would block only apoplastic RG-II dimerization in vivo. In conclusion, apoplastic pH neither broke boron bridges nor prevented their formation. Thus boron-starved cells cannot salvage boron from RG-II, and 'acid growth' is not achieved by pH-dependent monomerization of RG-II. Divalent metals and cationic peptides catalyse RG-II dimerization via co-ordinate and ionic bonding respectively (possible and impossible, respectively, at pH 1.75). Exogenous apiose may be useful to distinguish intra- and extra-protoplasmic dimerization.

An Arabidopsis thaliana arabinogalactan-protein (AGP31) and several cationic AGP fragments catalyse the boron bridging of rhamnogalacturonan-II.

Rhamnogalacturonan-II (RG-II) is a complex pectic domain in plant primary cell walls. In vivo, most RG-II domains are covalently dimerised via borate diester bridges, essential for correct cell-wall assembly, but the dimerisation of pure RG-II monomers by boric acid in vitro is extremely slow. Cationic 'chaperones' can promote dimerisation, probably by overcoming the mutual repulsion between neighbouring anionic RG-II molecules. Highly effective artificial chaperones include Pb2+ and polyhistidine, but the proposed natural chaperones remained elusive. We have now tested cationic peptide fragments of several Arabidopsis thaliana arabinogalactan-proteins (AGPs) as candidates. Fragments of AGP17, 18, 19 and 31 were effective, typically at ∼25 µg/ml (9-19 µM), promoting the boron bridging of 16-20 µM monomeric RG-II at pH 4.8 in vitro. Native AGP31 glycoprotein was also effective, and hexahistidine was moderately so. All chaperones tested interacted reversibly with RG-II and were not consumed during the reaction; thus they acted catalytically, and may constitute the first reported boron-acting enzyme activity, an RG-II borate diesterase. Many of the peptide chaperones became less effective catalysts at higher concentration, which we interpret as due to the formation of RG-II-peptide complexes with a net positive charge, as mutually repulsive as negatively charged pure RG-II molecules. The four unique AGPs studied here may serve an enzymic role in the living plant cell, acting on RG-II within Golgi cisternae and/or in the apoplast after secretion. In this way, RG-II and specific AGPs may contribute to cell-wall assembly and hence plant cell expansion and development.

Boron bridging of rhamnogalacturonan-II in Rosa and arabidopsis cell cultures occurs mainly in the endo-membrane system and continues at a reduced rate after secretion.

BACKGROUND AND AIMS: Rhamnogalacturonan-II (RG-II) is a domain of primary cell-wall pectin. Pairs of RG-II domains are covalently cross-linked via borate diester bridges, necessary for normal cell growth. Interpreting the precise mechanism and roles of boron bridging is difficult because there are conflicting hypotheses as to whether bridging occurs mainly within the Golgi system, concurrently with secretion or within the cell wall. We therefore explored the kinetics of RG-II bridging. METHODS: Cell-suspension cultures of Rosa and arabidopsis were pulse-radiolabelled with [14C]glucose, then the boron bridging status of newly synthesized [14C]RG-II domains was tracked by polyacrylamide gel electrophoresis of endo-polygalacturonase digests. KEY RESULTS: Optimal culture ages for 14C-labelling were ~5 and ~1 d in Rosa and arabidopsis respectively. De-novo [14C]polysaccharide production occurred for the first ~90 min; thereafter the radiolabelled molecules were tracked as they 'aged' in the wall. Monomeric and (boron-bridged) dimeric [14C]RG-II domains appeared simultaneously, both being detectable within 4 min of [14C]glucose feeding, i.e. well before the secretion of newly synthesized [14C]polysaccharides into the apoplast at ~15-20 min. The [14C]dimer : [14C]monomer ratio of RG-II remained approximately constant from 4 to 120 min, indicating that boron bridging was occurring within the Golgi system during polysaccharide biosynthesis. However, [14C]dimers increased slightly over the following 15 h, indicating that limited boron bridging was continuing after secretion. CONCLUSIONS: The results show where in the cell (and thus when in the 'career' of an RG-II domain) boron bridging occurs, helping to define the possible biological roles of RG-II dimerization and the probable localization of boron-donating glycoproteins or glycolipids.

Hemicellulose-remodelling transglycanase activities from charophytes: towards the evolution of the land-plant cell wall.

Transglycanases remodel cell-wall polymers, having a critical impact on many physiological processes. Unlike xyloglucan endotransglucosylase (XET) activity, widely studied in land plants, very little is known about charophyte wall-modifying enzymes - information that would promote our understanding of the 'primordial' wall, revealing how the wall matrix is remodelled in the closest living algal relatives of land plants, and what changed during terrestrialisation. We conducted various in-vitro assays for wall-remodelling transglycosylases, monitoring either (a) polysaccharide-to-[3 H]oligosaccharide transglycosylation or (b) non-radioactive oligosaccharide-to-oligosaccharide transglycosylation. We screened a wide collection of enzyme extracts from charophytes (and early-diverging land plants for comparison) and discovered several homo- and hetero-transglycanase activities. In contrast to most land plants, charophytes possess high trans-ß-1,4-mannanase activity, suggesting that land plants' algal ancestors prioritised mannan remodelling. Trans-ß-1,4-xylanase activity was also found, most abundantly in Chara, Nitella and Klebsormidium. Exo-acting transglycosidase activities (trans-ß-1,4-xylosidase and trans-ß-1,4-mannosidase) were also detected. In addition, charophytes exhibited homo- and hetero-trans-ß-glucanase activities (XET, mixed-linkage glucan [MLG]:xyloglucan endotransglucosylase and cellulose:xyloglucan endotransglucosylase) despite the paucity or lack of land-plant-like xyloglucan and MLG as potential donor substrates in their cell walls. However, trans-α-xylosidase activity (which remodels xyloglucan in angiosperms) was absent in charophytes and early-diverging land plants. Transglycanase action was also found in situ, acting on endogenous algal polysaccharides as donor substrates and fluorescent xyloglucan oligosaccharides as acceptor substrates. We conclude that trans-ß-mannanase and trans-ß-xylanase activities are present and thus may play key roles in charophyte walls (most of which possess little or no xyloglucan and MLG, but often contain abundant ß-mannans and ß-xylans), comparable to the roles of XET in xyloglucan-rich land plants.

Ancient origin of fucosylated xyloglucan in charophycean green algae.

The charophycean green algae (CGA or basal streptophytes) are of particular evolutionary significance because their ancestors gave rise to land plants. One outstanding feature of these algae is that their cell walls exhibit remarkable similarities to those of land plants. Xyloglucan (XyG) is a major structural component of the cell walls of most land plants and was originally thought to be absent in CGA. This study presents evidence that XyG evolved in the CGA. This is based on a) the identification of orthologs of the genetic machinery to produce XyG, b) the identification of XyG in a range of CGA and, c) the structural elucidation of XyG, including uronic acid-containing XyG, in selected CGA. Most notably, XyG fucosylation, a feature considered as a late evolutionary elaboration of the basic XyG structure and orthologs to the corresponding biosynthetic enzymes are shown to be present in Mesotaenium caldariorum.

Fruit softening: evidence for pectate lyase action in vivo in date (Phoenix dactylifera) and rosaceous fruit cell walls.

BACKGROUND AND AIMS: The programmed softening occurring during fruit development requires scission of cell wall polysaccharides, especially pectin. Proposed mechanisms include the action of wall enzymes or hydroxyl radicals. Enzyme activities found in fruit extracts include pectate lyase (PL) and endo-polygalacturonase (EPG), which, in vitro, cleave de-esterified homogalacturonan in mid-chain by ß-elimination and hydrolysis, respectively. However, the important biological question of whether PL exhibits action in vivo had not been tested. METHODS: We developed a method for specifically and sensitively detecting in-vivo PL products, based on Driselase digestion of cell wall polysaccharides and detection of the characteristic unsaturated product of PL action. KEY RESULTS: In model in-vitro experiments, pectic homogalacturonan that had been partially cleaved by commercial PL was digested to completion with Driselase, releasing an unsaturated disaccharide ('ΔUA-GalA'), taken as diagnostic of PL action. ΔUA-GalA was separated from saturated oligogalacturonides (EPG products) by electrophoresis, then subjected to thin-layer chromatography (TLC), resolving ΔUA-GalA from higher homologues. The ΔUA-GalA was confirmed as 4-deoxy-ß-l-threo-hex-4-enopyranuronosyl-(1→4)-d-galacturonic acid by NMR spectroscopy. Driselase digestion of cell walls from ripe fruits of date (Phoenix dactylifera), pear (Pyrus communis), rowan (Sorbus aucuparia) and apple (Malus pumila) yielded ΔUA-GalA, demonstrating that PL had been acting in vivo in these fruits prior to harvest. Date-derived ΔUA-GalA was verified by negative-mode mass spectrometry, including collision-induced dissociation (CID) fragmentation. The ΔUA-GalA:GalA ratio from ripe dates was roughly 1:20 (mol mol-1), indicating that approx. 5 % of the bonds in endogenous homogalacturonan had been cleaved by in-vivo PL action. CONCLUSIONS: The results provide the first demonstration that PL, previously known from studies of fruit gene expression, proteomic studies and in-vitro enzyme activity, exhibits enzyme action in the walls of soft fruits and may thus be proposed to contribute to fruit softening.

Cutin:xyloglucan transacylase (CXT) activity covalently links cutin to a plant cell-wall polysaccharide.

The shoot epidermal cell wall in land-plants is associated with a polyester, cutin, which controls water loss and possibly organ expansion. Covalent bonds between cutin and its neighbouring cell-wall polysaccharides have long been proposed. However, the lack of biochemical evidence makes cutin-polysaccharide linkages largely conjectural. Here we optimised a portfolio of radiochemical assays to look for cutin-polysaccharide ester bonds in the epidermis of pea epicotyls, ice-plant leaves and tomato fruits, based on the hypothesis that a transacylase remodels cutin in a similar fashion to cutin synthase and cutin:cutin transacylase activities. Through in-situ enzyme assays and chemical degradations coupled with chromatographic analysis of the 3H-labelled products, we observed that among several wall-related oligosaccharides tested, only a xyloglucan oligosaccharide ([3H]XXXGol) could acquire ester-bonds from endogenous cutin, suggesting a cutin:xyloglucan transacylase (CXT). CXT activity was heat-labile, time-dependent, and maximal at near-neutral pH values. In-situ CXT activity peaked in nearly fully expanded tomato fruits and ice-plant leaves. CXT activity positively correlated with organ growth rate, suggesting that it contributes to epidermal integrity during rapid expansion. This study uncovers hitherto unappreciated re-structuring processes in the plant epidermis and provides a step towards the identification of CXT and its engineering for biotechnological applications.

Cutin:cutin-acid endo-transacylase (CCT), a cuticle-remodelling enzyme activity in the plant epidermis.

Cutin is a polyester matrix mainly composed of hydroxy-fatty acids that occurs in the cuticles of shoots and root-caps. The cuticle, of which cutin is a major component, protects the plant from biotic and abiotic stresses, and cutin has been postulated to constrain organ expansion. We propose that, to allow cutin restructuring, ester bonds in this net-like polymer can be transiently cleaved and then re-formed (transacylation). Here, using pea epicotyl epidermis as the main model, we first detected a cutin:cutin-fatty acid endo-transacylase (CCT) activity. In-situ assays used endogenous cutin as the donor substrate for endogenous enzymes; the exogenous acceptor substrate was a radiolabelled monomeric cutin-acid, 16-hydroxy-[3H]hexadecanoic acid (HHA). High-molecular-weight cutin became ester-bonded to intact [3H]HHA molecules, which thereby became unextractable except by ester-hydrolysing alkalis. In-situ CCT activity correlated with growth rate in Hylotelephium leaves and tomato fruits, suggesting a role in loosening the outer epidermal wall during organ growth. The only well-defined cutin transacylase in the apoplast, CUS1 (a tomato cutin synthase), when produced in transgenic tobacco, lacked CCT activity. This finding provides a reference for future CCT protein identification, which can adopt our sensitive enzyme assay to screen other CUS1-related enzymes.

Defining natural factors that stimulate and inhibit cellulose:xyloglucan hetero-transglucosylation.

Certain transglucanases can covalently graft cellulose and mixed-linkage ß-glucan (MLG) as donor substrates onto xyloglucan as acceptor substrate and thus exhibit cellulose:xyloglucan endotransglucosylase (CXE) and MLG:xyloglucan endotransglucosylase (MXE) activities in vivo and in vitro. However, missing information on factors that stimulate or inhibit these hetero-transglucosylation reactions limits our insight into their biological functions. To explore factors that influence hetero-transglucosylation, we studied Equisetum fluviatile hetero-trans-ß-glucanase (EfHTG), which exhibits both CXE and MXE activity, exceeding its xyloglucan:xyloglucan homo-transglucosylation (XET) activity. Enzyme assays employed radiolabelled and fluorescently labelled oligomeric acceptor substrates, and were conducted in vitro and in cell walls (in situ). With whole denatured Equisetum cell walls as donor substrate, exogenous EfHTG (extracted from Equisetum or produced in Pichia) exhibited all three activities (CXE, MXE, XET) in competition with each other. Acting on pure cellulose as donor substrate, the CXE action of Pichia-produced EfHTG was up to approximately 300% increased by addition of methanol-boiled Equisetum extracts; there was no similar effect when the same enzyme acted on soluble donors (MLG or xyloglucan). The methanol-stable factor is proposed to be expansin-like, a suggestion supported by observations of pH dependence. Screening numerous low-molecular-weight compounds for hetero-transglucanase inhibition showed that cellobiose was highly effective, inhibiting the abundant endogenous CXE and MXE (but not XET) action in Equisetum internodes. Furthermore, cellobiose retarded Equisetum stem elongation, potentially owing to its effect on hetero-transglucosylation reactions. This work provides insight and tools to further study the role of cellulose hetero-transglucosylation in planta by identifying factors that govern this reaction.

Enzymically attaching oligosaccharide-linked 'cargoes' to cellulose and other commercial polysaccharides via stable covalent bonds.

The Equisetum enzyme hetero-trans-ß-glucanase (HTG) covalently grafts native plant cellulose (donor-substrate) to xyloglucan (acceptor-substrate), potentially offering a novel 'green' method of cellulose functionalisation. However, the range of cellulosic and non-cellulosic donor substrates that can be utilised by HTG is unknown, limiting our insight into its biotechnological potential. Here we show that HTG binds all celluloses tested (papers, tissues, hydrogels, bacterial cellulose) to radioactively- or fluorescently-labelled xyloglucan-heptasaccharide (XXXGol; acceptor-substrate). Glycol-chitin, glycol-chitosan and chitosan also acted as donor substrates but less effectively than cellulose. Cellulose-XXXGol conjugates were formed throughout the volume of a block of hydrogel, demonstrating penetration. Plant-derived celluloses (cellulose Iß) became more effective donor-substrates after 'mercerisation' in ≥3 M NaOH; the opposite was true for bacterial cellulose Iα. Cellulose-XXXGol bonds resisted boiling 6 M NaOH, demonstrating strong glycosidic bonding. In conclusion, HTG stably grafts native and processed celluloses to xyloglucan-oligosaccharides, which may carry valuable 'cargoes', exemplified by sulphorhodamine. We thus demonstrate HTG's biotechnological potential to modify various cellulose-based substrates such as textiles, pulps, papers, packaging, sanitary products and hydrogels.

High-Voltage Paper Electrophoresis (HVPE).

HVPE is an excellent and often overlooked method for obtaining objective and meaningful information about cell-wall "building blocks" and their metabolic precursors. It provides not only a means of analysis of known compounds but also an insight into the charge and/or mass of any unfamiliar compounds that may be encountered. It can be used preparatively or analytically. It can achieve either "class separations" (e.g., delivering all hexose monophosphates into a single pool) or the resolution of different compounds within a given class (e.g., ADP-Glc from UDP-Glc; or GlcA from GalA).All information from HVPE about charge and mass can be obtained on minute traces of analytes, especially those that have been radiolabeled, for example by in-vivo feeding of a 3H- or 14C-labeled precursor. HVPE does not usually damage the substance under investigation (unless staining is used), so samples of interest can be eluted intact from the paper ready for further analysis. Although HVPE is a technique that has been available for several decades, recently it has tended to be sidelined, possible because the apparatus is not widely available. Interested scientists are invited to contact the author about the possibility of accessing the Edinburgh apparatus.

Activity and Action of Cell-Wall Transglycanases.

Transglycanases (endotransglycosylases) are enzymes that "cut and paste" polysaccharide chains. Several transglycanase activities have been discovered which can cut (i.e., use as donor substrate) each of the major hemicelluloses [xyloglucan, mannans, xylans, and mixed-linkage ß-glucan (MLG)], and, as a recent addition, cellulose. These enzymes may play interesting roles in adjusting the wall's physical properties, influencing cell expansion, stem strengthening, and fruit softening.Activities discussed include the homotransglycanases XET (xyloglucan endotransglucosylase, i.e., xyloglucan-xyloglucan endotransglycosylase), trans-ß-mannanase (mannan -mannan endotransglycosylase), and trans-ß-xylanase (xylan -xylan endotransglucosylase), plus the heterotransglycanases MXE (MLG -xyloglucan endotransglucosylase) and CXE (cellulose -xyloglucan endotransglucosylase).Transglycanases acting on polysaccharide donor substrates can utilize small, labeled oligosaccharides as acceptor substrates, generating easily recognizable polymeric labeled products. We present methods for extracting transglycanases from plant tissues and assaying them in vitro, either quantitatively in solution assays or by high-throughput dot-blot screens. Both radioactively and fluorescently labeled substrates are mentioned. A general procedure (glass-fiber blotting) is illustrated by which proposed novel transglycanase activities can be tested for.In addition, we describe strategies for detecting transglycanase action in vivo. These methods enable the quantification of, separately, XET and MXE action in Equisetum stems. Related methods enable the tissue distribution of transglycanase action to be visualized cytologically.

Three highly acidic Equisetum XTHs differ from hetero-trans-ß-glucanase in donor substrate specificity and are predominantly xyloglucan homo-transglucosylases.

Transglycanases are enzymes that remodel the primary cell wall in plants, potentially loosening and/or strengthening it. Xyloglucan endotransglucosylase (XET; EC 2.4.1.207), ubiquitous in land plants, is a homo-transglucanase activity (donor, xyloglucan; acceptor, xyloglucan) exhibited by XTH (xyloglucan endotransglucosylase/hydrolase) proteins. By contrast, hetero-trans-ß-glucanase (HTG) is the only known enzyme that is preferentially a hetero-transglucanase. Its two main hetero-transglucanase activities are MLG : xyloglucan endotransglucosylase (MXE) and cellulose : xyloglucan endotransglucosylase (CXE). HTG is highly acidic and found only in the evolutionarily isolated genus of fern-allies, Equisetum. We now report genes for three new highly acidic HTG-related XTHs in E. fluviatile (EfXTH-A, EfXTH-H and EfXTH-I). We expressed them heterologously in Pichia and tested the encoded proteins' enzymic activities to determine whether their acidity and/or their Equisetum-specific sequences might confer high hetero-transglucanase activity. Untransformed Pichia was found to secrete MLG-degrading enzyme(s), which had to be removed for reliable MXE assays. All three acidic EfXTHs exhibited very predominantly XET activity, although low but measurable hetero-transglucanase activities (MXE and CXE) were also detected in EfXTH-H and EfXTH-I. We conclude that the extremely high hetero-transglucanase activities of Equisetum HTG are not emulated by similarly acidic Equisetum XTHs that share up to 55.5% sequence identity with HTG.