Baixe Plant Biochemistry - sdarticle (7) e outras Manuais, Projetos, Pesquisas em PDF para Química, somente na Docsity! 5 Carbohydrate Metabolism; Structural Carbohydrates J.S. Grant Reid 5.1 Introduction 205 5.2 The plant cell wall or extracellular matrix 205 5.3 Structures and interactions of plant cell wall polysaccharides . 209 5.4 Supramolecular interactions of structural polysaccharides in cell walls 218 5.5 Biosynthesis of structural polysaccharides 223 5.6 Metabolic turnover of structural components 231 5.7 Conclusions 234 References 235 5.1 INTRODUCTION Apart from water, structural carbohydrates are the main chemical constituents of most plant tissues and most plant cells. This is because carbohy- drates form the bulk of the plant cell's supporting structure - the cell wall or extracellular matrix. Consequently plant structural carbohydrates together form the most abundant natural com- pounds available on earth. They are clearly our most important renewable natural resource, and will be used increasingly both as a source of energy and as raw materials for industrial processes. 5.2 THE PLANT CELL WALL OR EXTRACELLULAR MATRIX 5.2.1 Significance of the cell wall With very few exceptions, plant cells are enclosed within a wall. The wall has mechanical strength, and defines the shape and size of the cell. Under the light microscope, the walls separating the cells in a plant tissue are usually clearly visible. The walls of adjacent cells meet at a dividing-line known as the middle lamella^ which can be distinguished with the electron microscope (Fig. 5.1a,b). There is strong cell-cell adherence across the middle lamella, since the individual cells PLANT BIOCHEMISTRY ISBN 0-12-214674-3 within a plant tissue do not normally fall apart when the tissue is placed under mechanical stress. Indeed the mechanical strength of a plant tissue is a function of the properties of the walls of cells within the tissue. The walls of plant cells of different types differ greatly in appearance. Young cells, which still retain the capacity to divide and/or elongate, invariably have a very thin cell wall (0.1-1/im in cross-section). This is the primary cell wall (Fig. 5.1a,b). Primary cell walls are of funda- mental importance in the process of cell expan- sion. It is evident that primary cell walls must yield to allow a cell to grow. It is equally clear that cell growth is accompanied by an increase in the total area of the wall. Thus if, as is observed, wall thickness is maintained during cell expansion, growth must be accompanied by intensive bio- synthesis of cell wall constituents. It must be also accompanied by the rearrangement of those molecular interactions which confer rigidity on the cell wall. There is convincing evidence, derived mainly from measurements of cell mechanical properties and from measurements of cell turgor pressure, that cell expansion is in fact controlled by metabolic events within the primary cell wall (Taiz, 1984). Young plant tissues, the cells of which have only thin primary walls, are relatively soft. Their rigidity is maintained by cell turgor., that is Copyright © 1997 Academic Press Ltd All rights of reproduction in any form reserved 2 0 6 J.S. GRANT REID (̂ ) mFj^'.fm^ Figure 5.1 Electron micrographs of plant cell walls, (a) Micrograph of developing lupin cotyledons, showing primary cell walls (W). The middle lamella is the central, darker-staining line. Bar represents l//m. Photo reproduced from Parker (1984), with permission, (b) Micrograph showing contact between two fiber cells (F) and a parenchyma cell (VP) in linden wood. The layered secondary walls (Si,S2) of the fiber cells are visible. L.Pi, middle lamella plus primary cell wall. Photo reproduced from Vian et al. (1986), with permission. hydrostatic pressure exerted by the protoplast on the wall and resisted by the latter. If such tissues are subjected to water-stress to the extent that turgor pressure cannot be maintained, wilting of the tissue occurs. When plant cells lose the capacity to grow and divide, they may differentiate into cells of different types, some of which have cell walls which are very thick. Such thickenings take the form of deposits laid down inside the primary cell wall, that is between the primary cell wall and the plasma membrane of the cell. The thickenings, which are usually many times the thickness of the primary cell wall, are termed secondary, and the whole wall internal to the primary cell wall is termed the secondary cell wall (Fig. 5.1b). Cells with secondary wall-thickenings are very impor- tant in conferring rigidity on plant tissues, inde- pendently of cell turgor. Sclerenchyma cells (fiber cells or stone-cells) have this function, whereas tracheids and xylem elements both strengthen tissues and function in water-conduction. The walls of all of the above cell types are not only rigid, but extremely hard. Hardness is conferred by the deposition throughout the cell wall, after the completion of secondary thickening, of the phenolic polymer lignin. Lignin deposition (ligni- fication) is usually followed by cell death and the disappearance of the cytoplasmic contents. It is noteworthy that the secondary wall thickenings are not always uniform over the cell surface. In xylem elements, for example, the secondary thickenings take the form of elaborate patterns (spirals, reticular etc.), implying precise targeting of carbohydrate deposits to particular areas of the wall during its biosynthesis. It is significant also that the spiral and reticular thickenings of the water-conducting cells are exactly those best- suited to strengthening a tubular vessel subjected to radial stress (in this case lower hydrostatic pressure inside than outside). In contrast to primary cells walls, walls which have undergone lignification are relatively inert metabolically. Although 'cell wall' is the most commonly used expression for the carbohydrate-rich structure surrounding the plant cell, research workers have sought an alternative descriptor lacking the idea of inertia and immovability associated with the word 'wall'. This is because, in most cells, the 'wall' is very much an active metabolic compartment, despite being external to the plasma membrane. Thus, in recent years there has been a tendency to use the more general term 'extracellular mat- rix'. In this chapter both expressions are used interchangeably. 5.2.2 Cell wall architecture When viewed using various electron microscopic protocols, the walls of most plant cells can be shown to consist of distinctive rodlike or fibrillar structures embedded in seemingly amorphous material. The visible structures are microfibrils of cellulose (Fig. 5.2) and they are largely crystalline. The microfibrils are embedded in material which is generally termed the matrix of the cell wall. The cellulose microfibrils are not normally arranged at CARBOHYDRATE METABOLISM: STRUCTURAL CARBOHYDRATES 209 Figure 5.5 Electron micrograph of a cell wall with secondary thickening, showing brief helicoidal interruptions (h, dotted lines) between thick uniform layers with longitudinally orientated microfibrils (Ig). ml, middle lamella, pr, reversion point. Sclerenchyma cell from the stem oi Aristolochia clematitis. (x 21000) Reproduced from Roland et al. (1987), with permission. Examples of such processes (delignifications) are acid chlorite and acid sulfite treatments: both oxidize the phenolic rings and convert the lignin into water-soluble derivative. Acid sulfite treatment is part of the paper-making process. To isolate and study the structures of the carbohydrates of highly lignified tissues, such as woods and crop straws, it is virtually essential to delignify the tissue, to give an almost lignin-free starting material, known as 'holocellulose\ Non-lignified cell walls are not normally subjected to delignification. On the basis of extractability, three principal groups of plant cell wall carbohydrates can be distinguished: the pectins, the hemicelluloses and cellulose. The pectins are extracted using aqueous solutions containing substances, such as ammo- nium oxalate or EDTA, which are capable of chelating Ca^"^ or other divalent metal cations. The hemicelluloses are more difficult to remove from the cell wall material, requiring fairly concentrated solutions of sodium or potassium hydroxide. The residue after extraction of the pectins and the hemicelluloses is rich in cellulose, and is often termed alpha-cellulose. This term, which originated in the paper-making industry, is somewhat confusing since the main component of the ^//?/7^-cellulose fraction is cellulose, a polymer composed entirely of beta-linked glucose residues! To avoid such confusion the term alkali-insoluble residue will be used throughout this section. Fortunately, the independently developed micro- scopic and extraction-based classifications of the carbohydrate components of the cell wall are compatible one with the other. Microscopic examination of tissues undergoing traditional extraction confirms that the pectic and hemicellu- lose fractions derive from the matrix of the cell wall, whereas the alkali-insoluble residue clearly includes the microfibrillar component. The pectin fraction, the hemicellulose fraction and alkali-insoluble residue also differ from each other in their molecular structures, which are discussed in section 5.3. Although it is intuitively obvious that the microfibrillar component of the cell wall will differ in its biological function from the matrix components, it is becoming increasingly apparent that the pectin and hemicellulose frac- tions may also have different roles to play in wall functionality. Current functional wall models are discussed in section 5.4. 5.3 STRUCTURES AND INTERACTIONS OF PLANT CELL WALL POLYSACCHARIDES This section outlines the main structural features of the macromolecular components of the pectin and hemicellulose fractions of the wall, as well as of cellulose, the polysaccharide which makes up the bulk of the alkali-insoluble residue. 210 J.S. GRANT REID It should be noted that there is no such thing as 'the typical cell wall'. Wall material isolated from tissues dominated by cells with thick secondary walls (e.g. wood) is very rich in both cellulose and the hemicellulose fraction, but contains very little or no pectin. On the other hand cell wall material isolated from the flesh of many fruits is particu- larly rich in the pectin fraction. In general the structures of the hemicellulose polymers present in primary cell walls are different from those in secondary cell walls. Cell walls (primary or secondary) of plants from different major taxo- nomic groupings can also differ greatly with respect to the relative proportions and the macro- molecular compositions of their hemicellulose and pectin fractions. Also the molecular structures of the polymeric carbohydrates of plant cell walls are often obtained after their isolation from the cell wall, and that isolation process may itself have brought about structural changes, possibly including unsuspected ones. For example, delignification will oxidize not only lignin but also phenolic substituents attached to carbohydrate polymers. Also alkaline extraction procedures will hydrolyze ester hnkages. This may result in the loss of acetyl substituents attached to the hydroxyl groups of carbohydrate polymers, or in the breakage of any ester-linkages cross-linking differ- ent macromolecular components of the cell wall. Accidental hydrolysis or other degradation during extraction is also capable of removing labile features from a polymer, or of decreasing its molecular weight. Despite the above cautionary words, it is pos- sible to draw general conclusions concerning the structures of plant cell wall polymers, and this is attempted below. Nonetheless modern cytochemi- cal studies, particularly on primary cell walls, have revealed differences in cell wall molecular compo- sition between different cells in a tissue and even between different areas of the wall of a given cell (Knox et al., 1990). Such differences are potentially of great relevance to the understanding of cellular differentiation and recognition. 5.3.1 Cellulose From the structural viewpoint cellulose is probably the best-understood of all the carbohydrates of the plant cell wall (Nevell &c Zeronian, 1985). It is easily obtained in a relatively pure state by exhaustive extraction with alkali of the hemicellu- losic fraction of the cell wall, and can be subjected to structural determination by standard chemical methods. Cellulose obtained in this way from most tissues yields over 90% D-glucose on acid hydro- lysis, the remainder being small amounts of other sugars, notably xylose, galactose and mannose. Since some tissues, such as cotton hairs, yield cellulose which gives essentially 100% glucose on hydrolysis it is usually assumed that the non- glucose sugars arise from incomplete removal of hemicellulose polymers. Structural analyses of cellulose show that the primary structure of the molecule is very simple: a long, monotonous, linear sequence of D-glucose residues joined together by (3(1 ^f^ 4) glycosyl linkages (Fig. 5,6). The average degree of poly- merization (number of glucose residues in a chain) of cellulose from secondary cell wall sources is high, in the region of 10000, whereas the limited number of studies that have been carried out on primary cell wall celluloses indicate a lower value (c. 2000-6000). When considering the molecular weights of carbohydrate polymers it must, of course, be borne in mind that the molecular weight is an average value. Ideally the measured average value will reflect the peak of the natural distribu- tion of chain lengths. Often it will also reflect accidental degradation during isolation and prepa- ration of the sample for analysis by a physico- chemical technique. No simple comparative procedure such as sodium dodecyl sulfate (SDS)- gel electrophoresis (used for proteins) is available for carbohydrate polymers. In the cell wall and after purification free of the pectin and hemicellulose fractions, cellulose exists as rod-like, highly insoluble fibrils. In the cell wall they are known as microfibrils. Viewed using the electron microscope, the microfibrils are tubular structures, usually elliptical in cross-section with diameters ranging from about 5 to 30 nm (Figs. 5.2 and 5.6b). Much, although not all, of the cellulose within the microfibril exhibits crystalline order when subjected to X-ray crystallographic analysis. The remainder is amorphous or 'paracrystalline' (Fig. 5.2). Within the individual chains, the glucose units (in their energetically preferred chair con- formations) are arranged in a spiral or helical pattern in which each residue is rotated at 180° relative to the preceding and following ones (Fig. 5.2). This arrangement can also be predicted using minimum-energy calculations, and is con- firmed by X-ray crystallography. Two forms of crystalline cellulose, which may be differentiated by X-ray crystallography, are known: cellulose I and cellulose II. In its native state cellulose always exists in the cellulose I form. However, if native cellulose is dissolved and recrystallized (regener- ated), or even simply treated with strong alkali CARBOHYDRATE METABOLISM: STRUCTURAL CARBOHYDRATES 211 (a) R R R R R G l c l-»4Glc l->4Glc l-*4Glc l-»4Glc l->4Glc 1-* (b) ^*^"''^^^^, Cellulose structure and organization, (a) Chemical structure, (b) Electron micrograph showing microhbnls of cellulose. Reproduced from Roelofsen (1965), with permission. without dissolution (mercerization) it is converted into cellulose II. Cellulose II is the thermodyna- mically stable form of cellulose, and cannot be reconverted into cellulose I. The existence of native cellulose in the less thermodynamically stable (metastable) form is a strong indication that the biosynthesis of cellulose involves a process of orientation of the nev^ly formed (1 -^ 4)-/?-D- glucan chains v^hich results in their crystallization in the less thermodynamically stable cellulose I form (Haigler, 1985). There is cytochemical evidence (Chanzy & Henrissat, 1985) that in cellulose I the parallel glucan chains all run in the same direction, v^hereas the X-ray evidence has been interpreted to indicate that the glucan chains in cellulose II are arranged in an alternating fashion. However, there is still some controversy over this point, and the molecular arrangements differentiating cellulose I and cellulose II cannot be held to be fully understood (Delmer, 1987). Cellulose is present almost universally in the cell walls of higher plants, constituting the microfi- brillar phase. It accounts for about 30-40% by weight of the cell walls of woody tissues which are dominated by secondary cell walls. If lignin is neglected this is over 50% of the wall carbohydrates. In primary cell walls cellulose accounts for only about 20% of the wall. 5.3.2 Hemicelluloses The definition of 'hemicellulose' often causes communication difficulties between plant scien- tists. To some, the 'hemicellulose fraction' is the material removed from cell wall material by alkaline extraction. To others a 'hemicellulose' is a cell wall polymer with a particular type of molecular structure and perhaps a particular notional function within the cell wall. In fact, if the hemicellulose polymers are differentiated clearly from the pectic polymers, which may be largely removed from cell wall material by prior extraction with neutral solutions of divalent metal chelators, then the two usages of the word 'hemicellulose' are compatible. After removal of the pectic polymers, alkaline extraction of cell walls does yield a range of carbohydrate polymers with structural features in common, and which may well play a role in wall architecture which is distinct from that of both cellulose and pectin. In this chapter the alkaH-soluble post-pectin wall 214 J.S. GRANT REID (b) Hemicelluloses from primary cell walls Structural information on the hemicellulose poly- mers of the secondary cell wall was gathered over several decades. The accumulation of sufficient bulk plant tissue for large-scale hemicellulose extractions presented no real problem, and suffi- cient quantities of purified hemicelluloses for conventional structural analyses (methylation ana- lysis, periodate oxidation, partial hydrolysis etc.) could be obtained. By contrast tissues with cell walls which are almost exclusively primary (growing tips, elongation zones etc.) could not easily be harvested in bulk and in any case yielded a very much smaller proportion of wall material. Thus, up to the mid 1970s, there was little or no direct experimental evidence of the chemical structures of the polymers present in the hemi- cellulose fractions of primary cell walls. Primary cell wall analysis was virtually restricted to analyses of sugar residue composition (Meier & Wilkie, 1959; Thornber & Northcote, 1962). In the mid 1970s the structural analysis of primary cell walls became practically feasible. Rapid, efficient protocols for the total methylation of very small amounts of carbohydrate polymers were developed, as well as new separation and identification procedures for partially methylated alditol acetates. Accordingly it became possible to carry out methylation analysis on milligram amounts of purified polysaccharides or even intact cell walls. Studies on plant pathogenic micro-organisms were beginning to yield enzymes capable of hydrolyzing specific glycosyl linkages within the polymers of the cell wall. Suspension cultures of plant cells, which have walls which, at least superficially, resemble primary cell walls were recognized as possible model systems for the investigation of primary cell wall structure. Most importantly, pioneering work, notably that carried out at the University of Colorado by P. Albersheim and colleagues, brought all of these new approaches to bear on the problem of the molecular structure of the plant primary cell wall (Talmadge et al., 1973; Bauer et al.^ 1973; Keegstra et al.^ 1973). The primary cell wall is now known to contain cellulose, pectin and hemicellulose polymers some of which are quite distinct in structure from the hemicelluloses of the secondary wall. Many primary cell walls also contain significant quantities of structural protein, and small amounts of phenolic materials, some bound covalently to the carbohydrate polymers (McNeil et aL, 1984; Fry, 1988). Structural studies have been carried out on the primary cell walls of a relatively large number of plant species, most of them angiosperms of some economic importance in temperate countries. The available data indicate two distinct types of primary cell walls, which differ in the composi- tions of their matrix components. The primary cell walls of dicotyledonous plants and those mono- cotyledonous plants that are not grasses have cell walls which are rich in both pectin and hemi- cellulose. Such primary cell walls have been described as type 1 (Carpita & Gibeaut, 1993). The primary cell walls of the grasses (Gramineae), termed type 2 primary cell walls (Carpita & Gibeaut, 1993) contain very little pectin, and proportionally more polymers which, on the basis of their alkali-solubility can be classified as hemicellulose. Furthermore, the relative propor- tions and the molecular structures of the hemi- celluloses in walls of the two types also differ (McNeil et aL, 1984). (i) Type 1 primary cell walls The hemicellulose fractions from type 1 primary cell walls normally contain xyloglucan as the principal component, alongside a smaller amount of a xylan. The xyloglucan (Hayashi, 1989), which accounts for about 20% of the dry weight of the cell wall, is remarkable in the degree of precision and order in its structure. It is best considered as a backbone with successive layers of substitution. The backbone is a chain of /?(!—> 4)-linked D-glucopyranosyl residues identical to that of cellulose, but considerably shorter. The backbone carries single xylopyranosyl substituents which are joined to it by a ( l —> 6)-linkages. Most of these substituents (apparently all of them in some cases) are arranged in a very regular fashion along the backbone: three consecutive glucosyl residues carry substituents while the fourth does not (Fig. 5.9). Thus most of the xyloglucan chain may be considered to be composed of a series of Glc4Xyl3 repeating units (Fig. 5.9). The subunits may be further substituted. Of the three side-chain xylose units on the Glc4Xyl3 repeating unit, the two farthest from the non-reducing end of the subunit may carry D-galactopyranosyl residues Hnked /?(! -^ 2) to xylose (Fig. 5.9). Of these, one may bear a further substituent, namely an L-fucopyranose residue linked a(l -^2) to galac- tose (Fig. 5.9). The precise arrangement of substituents along the backbone of the xyloglucan molecule indicates a high degree of metabolic control over the biosynthetic process. There is evidence that xyloglucan may be partially acety- lated in its native state (York et al., 1988). CARBOHYDRATE METABOLISM: STRUCTURAL CARBOHYDRATES 215 xyll B ia 6 FFucll la 2 (Gall) 13 2 xyll ia R B 6 rFucii rFucii la la 2 2 (Gall) (Gall) 13 13 2 2 Xy]Jl x y l l Xy l l X y l l \la la la la B B 6 B B 6 B B 6 B Gi lc l -^4Glc l ->4Glc l -*4Glc l ->4Glc l ->4Glc l -*4Glc l ->4Glc l ->4Glc l ->4Glc l -^4Glc l 6 6 6 ta Ta Ta X y l l x y l l x y l l 2 2 2 ti3 W t/3 (Gall) (Gall) (Gall) Figure 5.9 Structural features of primary cell wall xyloglucans. Note the concentric layers of substitution and the Glc4Xyl3 based structural subunits (see text). The glucan backbone is in bold type, and the xylose substituents in normal type. {Gal), possible galactose substituent. [Fuc], possible fucose substituent. The existence of the structural subunit based on Glc4Xyl3 was detected by hydrolysis of xyloglu- cans using a purified endo-l3(l —^ 4)-D-glucanase, which hydrolyzed the backbone chain at the glycosyl linkage following the unsubstituted glu- cosyl residue (Fig. 5.9), thus releasing the subunit oligosaccharides (Hayashi &: Maclachlan, 1984). This illustrates the value of enzymes of known specificity in the investigation of plant cell wall polysaccharide structures. Also present in the hemicellulose fraction from type 1 primary cell walls is a hemicellulose which is structurally very similar to the xylan-type hemicelluloses of secondary walls. It has a /?(! -^ 4)-linked D-xylan backbone which carries a range of different substituents attached to C-2 and/or C-3 of backbone xylose. (ii) Type 2 primary cell walls The hemicellulose fractions from type 2 primary cell walls are very rich in polysaccharides of the xylan type. They are accompanied by smaller amounts of a xyloglucan, and a mixed-linkage /?(! ^ 3, 1 -^ 4)-glucan. Although the xylans of primary cell walls have been studied to a lesser degree than those from secondary walls it is clear that the former are the more complex in their molecular structures. A (3(1 -^ 4)-linked D-xylan backbone is fairly heavily decorated by a variety of short side chains, the most important of which are illustrated in Fig. 5.8(b). The site of attachment of the side chains is to C-2 and/or C-3 of the backbone xylose residues. The acidic nature of the majority of the side chains is perhaps of significance, since pectin (see section 5.3.3), which provides most of the acidic functions in the type 2 primary cell wall, is present only in very low amounts in the type 2 primary wall. Although the primary cell walls of the grasses do contain xyloglucans in small amounts, these polysaccharides are apparently less regular in their substitution pattern than the xyloglucans in type 1 primary cell walls. The degree of xylose substitu- tion of the glucan backbone is lower, and there are fewer of the distinct Glc4Xyl3-based structural units which dominate the structures of the xyloglucans present in type 1 cell walls. Some of the xylose residues are galactose-substituted, but very few, if any, of these carry a fucosyl substituent (Hayashi, 1989). A characteristic hemicellulose of the type 2 primary cell wall is the mixed-linkage /?(! -^ 3, 1 -^ 4)-glucan (Stone & Clarke, 1992). The back- bone is linear, in the sense that it carries no side branches. However, two linkage-types, (1 -^ 3) and (1 —^4), are present in the backbone, and the overall shape or conformation of the molecule depends on their relative proportions and distribu- tion (Fig. 5.10). In fact the arrangement of the two linkage types within the molecule shows a high degree of order. In the first place the relative proportions of the two types of linkages within the molecule are apparently constant (ca.2.2). Sec- ondly, the distribution of the two linkage types within the molecule is non-random. The (1 -^ 4)- Hnkages do not occur singly, but in blocks. The blocks of (1 ^ 4)-linkages are separated by (1 -^ 3)-hnkages, which only occur singly. The molecule is therefore a series of /3(1 —> 4)-linked structural domains or blocks separated by single /3(1 —> 3)-linkages. Most of the structural blocks 216 J.S. GRANT REID 4 G l c l ->4Glcl-^3Glc 1^4Glc l ->4Glcl-»3Glc l->4Glc l-»4Glc 1 - > 4 G 1 C 1 - > 3 G 1 C 1 - » 4 ~ Figure 5.10 Structural features of mixed-linkage /3(1 —> 3, 1 —̂ 4)-glucan. contain two or three (1 —̂ 4)-linked glucose resi- dues, but smaller proportions of longer blocks are present (Stone & Clarke, 1992). As in the case of xyloglucans the structural regularity of the mixed-linkage ^^-glucan molecule was detected with the help of an enzyme which hydrolyzed the molecule in a specific and pre- dictable way. In this instance the enzyme was an unusually specific /3-glucanase ('lichenase') from the bacterium Bacillus subtilis. This enzyme hydrolyzes /?(! -^ 4)-glucosyl linkages, but only if they immediately follow a /?(! ^ 3)-glucosyl link. Thus digestion of a /3(1 ^ 3, 1 —• 4)-glucan with the B. subtilis lichenase releases /^-linked gluco-oligosaccharides all of which are (1 —• 4)- linked except for the linkage nearest to the reducing end, which is (3{\ —> 3) (Fig. 5.10). Each originated from a structural block, and the number of sugar residues in each is equal to the number of /?(! ^ 4)-linked glucose residues in the block. By analyzing the relative proportions of all the saccharides released, the relative numbers of structural blocks of different sizes within the glucan molecule may be determined. 5.3.3 Pectin The definition of ^pectin' is almost as problematic as that of 'hemicellulose'. To the food manufac- turer (or consumer) pectin is a natural fruit polysaccharide which is used, in jams for example, because of its ability to gel in the presence of high concentrations of sugar. The commercially impor- tant pectins originate in the cell walls of some fruits (citrus fruits and apples), the primary cell walls of which are particularly rich in them. In fact pectin appears to be present universally in primary cell walls, and it is a major constituent of primary cell walls of type 1. It is also present in the middle lamella between cells of all types. The term 'pectin' encompasses a complex group of polysaccharides, some of which may be struc- tural domains of larger, more complex molecules. This is, however, not certain. In addition to 'pectin' the terms 'the pectic substances' and 'the pectic polysaccharides' are in common use. In this chapter 'pectin' or 'the pectin fraction' will be used to describe the entire chelator-soluble polysacchar- ide fraction from a given type of cell wall. The term 'pectic polysaccharide' will be used for a single structural entity derived from or present in the pectin fraction, whether it exists as an inde- pendent macromolecule or as a major structural domain of a larger one. Pectin is acidic, containing a high proportion of D-galacturonic acid residues, most of which are present in a linear backbone. In some pectic polysaccharides the backbone consists almost entirely of D-galacturonic acid residues, in the pyranose ring form, joined together by a{\ —̂ 4)- linkages. The carboxylic acid groups of some of the galacturonic acid residues are esterified with methanol (Fig. 5.11). Pectic polysaccharides which have this structure, predominantly or exclusively, are known as 'homogalacturonans\ The term is possibly misleading, because the chain seldom consists exclusively of galacturonic acid and methylgalacturonate residues. A number of L-rhamnose (a 6-deoxyhexose) residues are nor- mally interspersed within the chain. The linkage from D-galacturonic acid to L-rhamnose is a ( l —̂ 2), and the linkage from rhamnose to the following galacturonic acid is a(\ -^ 4) (Fig. 5.11). Other pectic polysaccharides have a much higher proportion of rhamnose residues in the backbone, often to the extent of an alternating arrangement of galacturonic acid and rhamnose residues. Such rhamnose-rich pectic polysaccharides are called 'rhamnogalacturonans' (Fig. 5.12a) (McNeil et al., 1984). The rhamnose residues in rhamnogalactur- onans, and probably also the homogalacturonans, can serve as anchorage points for side chains attached to the backbone. Consequently, because of the variable frequency of the rhamnose residues, there are structures within pectin that are much more highly branched than others. The highly branched regions are often referred to as the 'hairy regions' (de Vries et ai, 1982), and the less highly branched regions as the 'smooth regions'. a a a a a a ^MeGalUAl-*4i»feGalUAl-»4J«feGalUAl->2Rhal-»4HeGalUAl-»4/feGalUAl-*4J»feGalUAl Figure 5.11 Structural features of the homogalacturonan component of pectin. MeGalUA, methyl ester of galacturonic acid. CARBOHYDRATE METABOLISM: STRUCTURAL CARBOHYDRATES 219 deposited continuously, to keep pace with the increase in wall surface area. Carbohydrate components already present within the wall undergo rearrangement, both passively as a result of growth-related changes in cell dimensions, and actively to permit and even to control the amount, the rate and the direction of growth. Since the realization that cell growth is dependent on metabolic processes going on within the wall, there has been an urgent research interest in determining which wall components undergo modification, and how such modifications in turn influence the mechanical properties of the cell wall. It has also become clear that changes in the mechanical properties of primary cell walls accompany many developmental processes, nota- bly fruit-ripening, and that the texture and quality of most of our vegetable foods is intimately bound-up with the mechanical properties of the plant primary cell wall. Cell growth, fruit ripening and texture are three important parameters over which agronomists would like to exercise control in crop plants. Yet the successful manipulation of these parameters in a systematic manner will depend upon research progress in two related areas. A primary requirement is an understanding of how individual cell wall components interact to determine the mechanical properties of the cell wall. Also the eventual structural modification within the plant of individual cell wall components requires an understanding, at the molecular level, of the metabolic pathways which bring about their biosynthesis and metabolic turnover. Current ideas on the supramolecular inter- actions of the carbohydrate components of the primary cell wall are introduced below. Bio- synthesis and turnover are covered in the following sections. Cellulose Microfibril Figure 5.15 Potential linkages between xyloglucan and cellulose. Reproduced, with permission, Hayashi (1989). 5.4.1 Cellulose-hemicellulose interactions There is no evidence for covalent linkages between cellulose, the microfibrillar component of the cell wall, and any of the carbohydrate or other components of the matrix. On the other hand, there is good evidence that certain of the hemicelluloses of the matrix undergo strong non-covalent binding to cellulose. The best- documented example of this is the interaction between xyloglucan, the main hemicellulose poly- saccharide of the type 1 primary cell wall, and cellulose. The total extraction of xyloglucan from cell walls is very difficult, requiring very high concen- trations of alkali, or other reagents capable of disrupting hydrogen bonds. Once isolated, xylo- glucan binds specifically to cellulose in vitro (Hayashi et al., 1987), and can be removed from the surface of the cellulose microfibrils only by treatments which disrupt hydrogen bonds. There can be little doubt that at least some of the xyloglucan present in the primary cell wall of type 1 is bound to cellulose. It is possible, given the known distance between cellulose microfibrils within the wall and the estimated length of a xyloglucan molecule, that a single xyloglucan molecule could bind to more that one cellulose microfibril, effectively tethering microfibrils together (Fig. 5.15). Micrographs of onion cell walls show clear intermicrofibril bridges (McCann et al., 1990), which are visible after the removal of pectin but not after the extraction of xyloglucan (Fig. 5.16). Thus it is possible that the hydrogen- bonding interaction between cellulose and xylo- glucan is the basis of a network within the native primary cell wall, known as the xyloglucan- cellulose network. There is insufficient xyloglucan in the type 2 primary cell walls of grasses for a xyloglucan-cellulose network to be important. Hemicelluloses present in the type 2 primary cell wall (mixed-linkage glucan and xylan) have, however, been shown to hydrogen-bond to cellu- lose, although less strongly than xyloglucan. It is possible that a network involving the bridging of cellulose microfibrils by these molecules exists in the type 2 wall. 5.4.2 Pectin interactions There is little direct evidence that the pectin in the plant primary cell wall is covalently linked to hemicellulose or to cellulose, or that it is strongly hydrogen-bonded to other components. On the 220 J.S. GRANT REID Figure 5.16 Electron micrograph of onion cell walls. Fast-freeze-deep-etch rotary-shadowed replica technique, following removal of most of the pectin from the wall. Note the thin bridges between thicker cellulose microfibrils. Bar = 200nm. Reproduced from McCann et al. (1990), with permission. other hand pectin undergoes self-interaction to the extent that it may itself form the basis of a network within the primary cell wall. It is well known that many pectins, in vitro, form gels in the presence of calcium ions (Jarvis, 1984). The formation of a gel by an otherwise soluble polymer implies the formation of inter- molecular junction zones (Fig. 5.17), and there is general agreement that in the case of pectin these take the form of an interaction between the negatively charged homogalacturonan domains (or smooth regions) and the divalent, positively charged calcium ions. There is good evidence that calcium-pectin crosslinks exist also in the type 1 primary cell wall. Calcium levels in the wall are high enough to support such interac- tions, treatment of cell walls with calcium chelators results in loss of wall strength, and histochemical studies with antibodies indicate that such structures are present within the wall S G ^ polymer molecule in solution 3fc j""cl'0" zone Figure 5.17 Simplified representation of the formation of a gel network, following the formation of junction zones between polymer molecules in solution. S, polymer molecules in solution. G, gel network. CARBOHYDRATE METABOLISM: STRUCTURAL CARBOHYDRATES 221 (Liners et al., 1992). For these reasons it is supposed that at least some of the pectin within the primary cell wall exists as a crosslinked network, the pectin network. Other types of crosslinkages are known to be possible within the pectin network, for example ester crosslinks and crosslinkages between the phenolic substi- tuents which are known to be attached in low amounts to pectin. These will not be discussed here since direct evidence for them at the present time is scant. The type 2 primary cell walls of the grasses contain very little pectin, and do not contain an equivalent molecular complex. 5.4.3 Primary cell wall models Much of the pioneering work on the structures of the carbohydrate components of the primary cell wall was done in the laboratory of Peter Albersheim and coworkers at the University of Colorado, and it was that group which put forward the first tentative function model of the primary cell wall (the type 1 wall as defined here) (Keegstra et al., 1973). The Albersheim model (Fig. 5.18) was the first of its kind and the clear inspiration for more recent versions, it incorpor- ated a framework of cellulose microfibrils linked to a xyloglucan-pectin-protein matrix by the xyloglucan-cellulose hydrogen-bonding interac- tion described above. The xyloglucan, pectin and protein components were depicted as being linked together via xyloglucan-pectin and pectin-protein covalent Hnkages (Fig. 5.18). No such covalent linkages have been generally confirmed (but see, Qi et al., 1995) and more recent cell wall models are based on the hypothesis of independent networks within the cell wall. The type 1 cell wall is envisaged to be composed of the cellulose- xyloglucan network, the pectin network, plus in some cases a protein network, all contributing independently to the overall mechanical and other properties of the cell wall. The principal network within the type 2 cell wall may be considered to be cellulose-xylan. The networks within the type 1 cell wall were differentiated clearly by Carpita & Gibeaut (1993) (Fig. 5.19), although the represen- tation is not to scale. The model by McCann &: Roberts (1991) (Fig. 5.20) for the type 1 cell wall of onion has the cellulose microfibrils and the re- maining matrix-filled spaces between them, scaled correctly relative to the overall thickness of the primary wall. s = — cellulose elementary fibril I I III III - xMoglucan 4 4 4 • ^^" protein with arabinosjl tetrasaccharides A A A # 4 A ~ glycosidically attached to the h>droxyproline ^ ^ ^ ^ ^ ^ residues rhamnogalacturonan main chain of the pectic polysaccharide arabinan and 4-linked galactan side chains of the pectic polymer 3,6-linked arabinogalactan attached to serine of the wall protein total pectic polysaccharide unsubstituted seryl residues of the wall protein Figure 5.18 The Albersheim model for polymer interactions in the plant primary cell wall. From Keegstra et al. (1973), with permission. 224 J.S. GRANT REID Figure 5.21 Structure of cyclic diguanylic acid. G, guanine. Reproduced, with permission, from Delmer (1987). In A. xylinum, the high-energy glucose donor for cellulose biosynthesis is UDP-glucose (UDP- Glc). The enzyme complex responsible for the synthesis (cellulose synthase) is tightly bound to the bacterial ceil membrane. Careful detergent treatment of the membranes results in their dispersion and the formation of a solution containing protein/detergent/lipid micelles which retain low cellulose synthase activity. A high level of activity is restored by adding the effector molecule cyclic diguanylic acid (Fig. 5.21), which was isolated from A. xylinum extracts and is believed to be an activator of the enzyme complex in vivo (Ross et al., 1987). The cellulose synthase of A. xylinum was partially purified by a technique known as product entrapment, which has proved useful for the puri- fication of soluble or detergent-solubilized enzymes which catalyze the formation of an insoluble product. When the insoluble cellulose fibers formed by the detergent-solubilized A. xyli- num cellulose synthase are collected by centrifuga- tion, the synthetic enzymes remain associated with the cellulose and can be partially separated from contaminating proteins (Fig. 5.22) (Lin & Brown, 1989). incubate Ky spin Vli/ isolate pellet >- >m^ Km^ 1. Detergent-solubilised enzyme preparation plus NDP-sugar substrate 2. Suspension of insoluble polysaccharide product 3. Product pelleted by centrifugation 4. Isolated pellet may contain polysaccharide synthase enzyme(s) specifically bound to product Figure 5.22 Diagrammatic representation of product entrapment. The catalytic subunit of A. xylinum cellulase synthase was identified positively by a further technique called photoaffinity labeling, which has been used to identify a variety of membrane- associated enzymes or receptors (Lin et al., 1990). A radiolabeled, photoreactive analog of a sub- strate or other ligand is mixed at low concentra- tion with the membrane preparation, and the mixture is irradiated with ultraviolet light to convert the photoactivable reagent into a chemical species so reactive that it will bind covalently to any molecule in its vicinity. Reagent molecules tightly bound to a protein will react with it, thus radiolabeling it and permitting its localization on SDS-gels by autoradiography. Non-bound reagent molecules react with water. The use of 5'-azido- UDP-glucose (Fig. 5.23), a UDP-Glc photoanalog, in experiments of this type allowed the positive identification of the catalytic subunit of A. xyli- num cellulose synthase, eventually permitting the cloning and characterization of the genomic sequence encoding the protein (Saxena et al., 1990). There is considerable ultrastructural evidence that plant cellulose synthesis occurs at the plasma membrane of the cell, and that rosette-like structures on the plasma membrane may be the catalytic centers (Fig. 5.24). Plasma membrane preparations, or indeed any mixed membrane preparations from higher plants, catalyze at best only extremely small amounts of /3(1 —̂ 4)-linked glucan from UDP-glucose, which may possess some of the properties of cellulose (Okuda et al., 1993). Much glucan is formed, but it is mainly /3(1 -^ 3)-linked. f3{\ -^ 3)-Linked glucan is not a normal constituent of the plant cell wall, but it is formed by many plant cells as a response to injury, and is usually called 'callose\ Thus any breakage of the plant cell membrane 'switches' its biosyn- thetic machinery to the formation of wound callose. The essential completeness of the switch- over, and other circumstantial evidence, has fostered the idea that perhaps the enzymes responsible for the synthesis of callose and cellulose are one and the same, and that cell breakage triggers a change in the transfer specifi- city from (1 ^ 4) to (1 ^ 3), mediated possibly by the elimination of the membrane potential and changes in ion-concentration following cell dis- ruption. Should this be true, current efforts in several laboratories throughout the world to identify the enzymes involved in callose synthesis would have enhanced significance. An obvious approach to the identification of the catalytic subunit of higher plant cellulose synthase is to probe plant cDNA libraries using the CARBOHYDRATE METABOLISM: STRUCTURAL CARBOHYDRATES 225 t t f t Add 5'-azido- UDP-GIc 1 labelled with P-32 (*) ^̂ ^ * * Solubilised proteins UV-irradiate. AzIdo-UDP-GIc reacts with, and binds covalently to, any nnolecule In very close proximity. Thus UDP-Glc-binding proteins are labelled with P-32 3 Precipitate proteins and separate by SDS-PAGE < > Gel Autoradiogram P-32 labelled protein (-•-•) probably binds UDP-GIc Figure 5.23 Diagrammatic representation of use of photoaffinity labeling to identify a UDP-glucose binding protein. sequence of the A. xylinum cellulose synthase catalytic subunit. At the present time, however, no success has been reported, indicating that the two sequences may not be highly homologous. Nor has the addition of cyclic diguanylic acid to plant membrane preparations been reported to enhance cellulose synthesis in vitro. Thus, at the time of writing, the biosynthesis of the most abundant pure substance on earth remains something of an enigma. 5.5.2 Hemicelluloses and pectin Although the matrix polysaccharides are more complex structurally than cellulose, attempts to obtain their biosynthesis in vitro from sugar nucleotide precursors have been accompanied by some success. The biosynthesis in vitro of several hemicelluloses, notably glucomannan (section 5.3.2a), xylan (sections 5.3.2a and b), xyloglucan (section 5.3.2b) and mixed linkage /3(1 -^ 3, 1 —» 4)-glucan (section 5.3.2b) has been reported. Of the pectin structural domains, only galactan (section 5.3.3) biosynthesis has been demonstrated convincingly. The enzymes catalyzing the formation of matrix polysaccharides from sugar nucleotide precursors are tightly bound to membranes. The subcellular origin of the active membranes has not always been determined, but when it has, Golgi mem- branes have always been indicated. This is in contrast to cellulose and callose biosynthesis, which are plasma membrane associated. To detect the biosynthesis of cell wall matrix polysaccharides in vitro, tissues are normally homogenized, and partially purified membrane preparations are obtained by centrifugation pro- tocols. The membrane preparations are incubated with appropriate sugar nucleotide precursors, labeled (usually with ^^C) in the sugar moiety, and labeled high-molecular-weight material is separated from non-incorporated precursor and from any low-molecular-weight labeled products (Fig. 5.25). In experiments such as this, the identification of the high-molecular-weight labeled material is of the utmost importance. For example, in an experiment to investigate the incorporation Cell Wall Face W / EP (outer) Face PF (inner) Face Rosettes Figure 5.24 Stylized drawing of structures revealed by freeze-fracture of plasma membranes of certain algae and of higher plants. The 'rosettes' are believed to be terminal complexes associated with the ends of growing cellulose microfibrils (MF). Reproduced, with permission, from Delmer (1987). 226 J.S. GRANT REID 1 v«y h i I.I.I i w Plant tissue Membrane-bound Pellet, containing enzyme preparation polysaccharide* e.g. ...GIc1-4Glc1-4Glc1-4Glc... Structure of polysaccharide* 1. Homogenisation followed by differential centrifugation 2. Incubation with sugar nucleotide labelled with C-14(*) in the sugar part (e.g. UDP-GIc*). Precipitation of labelled polysaccharide, and removal of remaining labelled nucleotide by centrifugation and washing pellet 3. Structural studies on labelled polysaccharide (methylation analysis or enzymatic analysis Figure 5.25 Schematic representation of methods used to demonstrate cell wall polysaccharide biosynthesis in vitro. of label from UDP-[*'^C]glucose, several different hemicelluloses (glucomannan, xyloglucan, mixed- linkage glucan), as well as callose or even cellulose are all possible products. The interpretation of some of the earlier work on matrix polysaccharide biosynthesis is difficult, because it was carried out before techniques for the characterization of very small amounts of polysaccharides became avail- able. In recent work, labeled polysaccharides formed in vitro have increasingly been subjected to full methylation analysis or to fragmentation using enzymes of known substrate specificity. Glucomannan (a secondary cell wall hemicellu- lose (section 5.3.2a)) and mixed-linkage glucan (a hemicellulose of type 2 primary cell walls (section 5.3.2b)) biosynthesis are used below as examples of the modern approach to product characterization. (a) Glucomannan biosynthesis The biosynthesis of glucomannan, the principal hemicellulose of the secondary cell walls of coniferous woods, provides a good example of the importance of characterizing polymeric products formed in vitro from sugar nucleotide precursors (Dalessandro et ai, 1986). Membrane preparations from cambial cells (i.e. cells actively differentiating into wood fibers with secondary walls) of pine catalyzed the incorporation of label from GDP-[^^C]mannose into a labeled polymeric product which, on acid hydrolysis, released not only labeled mannose, but also labeled glucose. Similarly when labeled GDP-mannose was the precursor a polymeric product was formed which yielded both labeled glucose and mannose on hydrolysis. On methylation analysis the product clearly contained glucose and mannose residues linked together in a linear fashion by (1 ^^ 4)- linkages as in glucomannan. Thus the catalytic membranes from pine cambium must contain an enzyme capable of interconverting GDP-mannose and GDP-glucose (a GDP-mannose-2-epimerase) and glycosyltransferases capable of catalyzing the transfer of mannose and glucose residues from GDP-mannose or GDP-glucose to elongate a growing glucomannan chain (Fig. 5.26). The membrane-bound enzymes have not yet been isolated and purified, and little is known about the control of the ratio mannose/glucose in vivo. D= Incubate with GDP-Man* or GDP-GIc* 2. Pellet polysaccharide* Acid hydrolysis Methylation analysis Membrane-bound enzyme from pine cambium Labelled pellet ..Glc*1 -4Man*1-4Man*1-4.. Structure: Glucomannan labelled in glucose and mannose residues Figure 5.26 Schematic representation of methods used to demonstrate secondary cell wall glucomannan biosynthesis in vitro. (Dalessandro et al., 1986.) CARBOHYDRATE METABOLISM: STRUCTURAL CARBOHYDRATES 229 levels at which the Man/Gal ratio is controlled (Edwards et ai, 1989, 1992; Reid et al., 1995). Two mechanisms are involved, the biosynthetic process, and a post-depositional modification. In fenugreek and guar, the Man/Gal ratios (1.1 and 1.6, respectively) are determined by the biosyn- thetic mechanism alone. On the other hand in senna, the Man/Gal ratio (3.3) is estabhshed by the controlled removal of galactose by a-galactosidase action from a primary biosynthetic product (Man/ Gal ratio = 2.3) during late seed development (Edwards et al., 1992). Post-depositional modifica- tion may be of much wider importance in the control of the structures of other cell wall polysaccharides than has been supposed hitherto. In the case of pectin, the removal of methyl ester groups by pectin methylesterase action is known to occur. Two tightly membrane-bound enzymes are involved in the biosynthesis of galactomannans. A GDP-mannose-dependent (3(1 -^ 4)-D-mannosyl- transferase (or mannan synthase) is responsible for elongating the mannan backbone towards the non- reducing end by catalyzing the transfer of mannosyl residues one at a time from GDP-mannose. A highly specific UDP-galactose-dependent a-D-galactosyl- transferase catalyzes the transfer of a galactosyl residue to an acceptor mannose residue at or near the non-reducing end of the growing D-mannan chain (Fig. 5.29). In vitro, in the presence of labeled GDP-mannose alone (no UDP-galactose), the pro- duct is labeled /?(! ^^ 4)-D-mannan. In the presence of GDP-mannose and UDP-galactose, the product is a galactomannan. In the presence of labeled UDP- galactose alone, no labeled galactomannan is formed. Some degree of control of the Man/Gal ratio of the in vitro galactomannan can be obtained by varying the in vitro concentrations of the two sugar nucleotides; by lowering the GDP-mannose concentration at constant (saturating) UDP-galac- tose, higher degrees of galactose substitution may be obtained. Thus the formation of the galactosyl side chains in galactomannans is strictly dependent on simultaneous elongation of the mannan chain, whereas the mannan chain can be elongated independently of the formation of side chains (Edwards ^f^/., 1989). In the investigation of galactomannan biosynth- esis, as in the case of /5(1 —> 3, 1 -^ 4)-glucan, enzymatic hydrolysis was crucial to the identifica- tion and characterization of labeled in vitro products of biosynthesis. The enzyme in this case was an endo-/3(l -^ 4)-D-mannanase from Asper- gillus niger. This enzyme recognizes a sequence of five mannosyl residues in a mannan chain, hydrolyzing between residues 3 and 4 numbered from the non-reducing end (Fig. 5.30) (McCleary & Matheson, 1983). Galactosyl substitution of the mannan chain in galactomannans affects the action of the enzyme in a very well-defined fashion, substituents at residues 2 and/or 4 of the recognition sequence preventing hydrolysis. Thus galactomannans with higher degrees of substitu- tion are, in general, hydrolyzed to a lesser extent than galactomannans with lower degrees of substitution. However, the pattern of substitution of the galactomannan also influences both the degree of hydrolysis, and the quantitative distribu- tion of galactomannan oligosaccharide fragments released in the enzyme digests. Galactomannan biosynthesis by the experimental model depicted in Fig. 5.29 has been computer- simulated with the inbuilt assumption of a second- order Markov-chain. That is to say, it has been assumed that the probability of obtaining galactose- substitution at the acceptor mannosyl residue (labeled a in Fig. 5.29) is influenced by the pre- existing state of substitution at only the nearest and the second-nearest neighbor mannosyl residues (labeled 1 and 2 in Fig. 5.29). On this assumption Figure 5.29 Enzyme interaction in galactomannan biosynthesis. A GDP-mannose- dependent mannan synthase catalyzes the elongation of the mannan backbone towards the non-reducing end (arrow). A UDP-galactose-dependent galactosyltransferase catalyzes the transfer of a galactosyl residue to an acceptor mannose residue (a) at or near the end of the growing mannan chain. 1,2, nearest and second-nearest neighbor mannosyl residues to the acceptor residue (see text). M- .•^ a — M- G 1 - M - 1 2 — M- 1 — M ̂ ^ G \ M ^ M ^ a 1 2 M Chain-elongation towards non-reducing end Possible galactose-acceptor mannose residue Nearest-neighbour mannose residue Second nearest-neighbour mannose residue 230 }.S. GRANT REID G a l G a l G a l Ga l 1 2 3 4 5 i i i i Man->Man->Man-*Man->Man-»Man-*Man-*Man-»Man^Man->Man->Man-»Man t t T 1 2 3 4 5 G a l G a l Gal 1 endo-13-mannanase (does not cut) Figure 5.30 Action on galactomannans of an endo-/?(l Matheson, 1983). endo-13-mannanase (cuts) • 4)-D-mannanase from Aspergillus niger (McCleary & there are four independent probabilities of obtain- ing galactose substitution at the acceptor mannose: POO^PIOJPOI andPii (Fig. 5.31). Any numerical set of four probabilities will then suffice for the computer- generation of a hypothetical galactomannan with a uniquely described distribution of galactose resi- dues along the mannan backbone and a defined Man/Gal ratio. The action of the A. niger mannanase has also been computer-simulated, so that the input of a probability tetrad into the computer will generate not only a unique galacto- mannan structure but the relative amounts of the various manno- and galactomanno-oligosacchar- ides which would be released on the exhaustive hydrolysis of that galactomannan using the A. niger mannanase. Fragmentation, using the A. niger mannanase, of labeled galactomannans generates the relative amounts of labeled manno- and galactomanno-oligosaccharides as experimental data. Thus by inputting an experimental data set derived from the enzymatic fragmentation of a labeled galactomannan biosynthesized in vitro, the computer algorithm will generate a unique probability tetrad. This was done for galactoman- —Jfai]-»Man-^an— Poo t Gal Gal i —Man^fMan-fHan- Pio Gal I —Hai]-»Man-Maii- t Gal Figure 5.31 Illustration of the four possible states of substitution at the nearest and second-nearest neighbor mannosyl residues to the galactosylacceptor mannosyl residue (italicized). PQO, Pio? Poi and Pn are the corresponding probabilities of obtaining galactose substitution at the acceptor mannosyl residue. nans, covering a wide range of Man/Gal ratios formed in vitro using the enzymes from the high-, medium- and low-galactose species. To allow comparison of probability-sets derived from galactomannans with different Man/Gal ratios all probability sets were scaled linearly, with Poo being ascribed the arbitrary value 1.00. When this was done it was observed that all probability sets from a single species were closely similar, yet quite different from those obtained from the other two species. This confirmed the correctness of the second-order Markov-chain model, and demon- strated clearly that the specificities of the galacto- mannan-forming transferase enzymes were different in the three species. The transfer specifi- cities, defined mathematically in the probability sets, may each be considered to define a maximum Man/Gal ratio for a galactomannan synthesized according to the transfer rules thus defined. This is the Man/Gal ratio of the galactomannan formed when the highest of the four probabilities within a set is set to 1.00 and the other three are rescaled correspondingly in a linear fashion. When this was done for the fenugreek, guar and senna consensus probability sets, the maximum galactose-content predicted for each species was either identical with or only very slightly higher than that of the primary product of the biosynthetic machinery. This was remarkable, indicating not only that the distribution of substituents on the main chain of the galactomannan, but also the total amount of substitution is determined by the specificities of the membrane-bound transferase enzymes (Reid et al., 1995). Although galactomannans are not present gen- erally in cell walls, most hemicelluloses have a non-regular distribution either of different residues within a backbone or substituents along a linear backbone. It seems likely that the application of Markov-chain statistics in these cases also, will reveal the operation of precisely defined statistical rules generated by the specificities of highly organized membrane-bound transferase com- plexes. CARBOHYDRATE METABOLISM: STRUCTURAL CARBOHYDRATES 231 5.6 METABOLIC TURNOVER OF STRUCTURAL COMPONENTS The plant cell wall is an extracellular matrix and, as such, might be expected to undergo only slow metabolic turnover. It is unlikely that the lignified walls of non-living cells such as vessel elements and fibers are turned over. On the other hand, living cells might be expected to possess the metabolic machinery to rearrange the molecular structures of their (mainly primary) cell walls. There is ample evidence that they do. The cell wall storage polysaccharides of seeds, for example (section 5.5.2c) are completely mobihzed after germination, and the enzymatic mechanisms involved have in several cases been described in detail (Reid, 1985b). This section highlights two incidences where the enzymatic modification of a particular cell wall matrix component is associated with, and believed to control, developmental events, namely xyloglucan turnover in the context of cell elongation, and pectin turnover in the context of fruit ripening. It draws the reader's attention finally to the steadily increasing evidence that oligosaccharide fragments, originating almost certainly from the hydrolysis in vivo of cell wall polysaccharides, are powerful regulatory and signaling molecules in plants. 5.6.1 Xyloglucan turnover and cell growth The process of cell wall expansion (Taiz, 1984) is one in which the wall of the cell extends in response to the internal pressure (turgor) of the cell's contents. Although turgor pressure is necessary for growth to occur, there is much evidence that, in higher plants at least, cell expansion is controlled by metabolic events occurring within the cell wall itself. In other words, cell expansion occurs as a result of cellular control over wall extensibility. The changeover from a non-extensible to an extensible cell wall is often called 'loosening'. The biochemical mechanism of loosening (which must, of course, be reversible) has occupied cell wall biochemists for some years, and there is a great deal of circum- stantial evidence that, in type 1 primary cell walls at any rate, xyloglucan modification may be involved. Xyloglucan has been associated with cell expansion following the demonstration that it is strongly hydrogen-bonded to cellulose (section 5.3.2b), and especially since the realization that a xyloglucan-cellulose network may exist within the primary cell wall (section 5.4.3). Initially it was suggested that wall loosening might be brought about by the controlled weakening of the hydro- gen bonds between cellulose and xyloglucan, thus allowing relative movement of the cellulose microfibrils. Change in wall pH was suggested as a possible agent to induce weakening of the hydrogen bonds, since the growth-promoting substance auxin is known to bring about wall acidification by stimulating the active transport of hydrogen ions out of the cytoplasm. In the absence of evidence for the direct weakening of cellulose- xyloglucan binding by naturally occurring pH changes within the wall (Valent 6c Albersheim, 1974), attention turned to the possibility that glycosyl linkages within xyloglucan molecules might be broken, possibly reversibly by trans- glycosylase action (Albersheim, 1975). Xyloglucan was further implicated in auxin- induced growth by the observation that auxin- induced elongation of stem sections of peas and other dicotyledonous plants was accompanied by xyloglucan turnover and solubilization (Labavitch 8c Ray, 1974). Xyloglucan solubilization occurred as a response to auxin even when growth was prevented by lowering cell turgor using osmoti- cally active substances. Auxin treatment also increased the levels of the enzyme endo- l3{\ -^ 4)-D-glucanase (Fan 6c Maclachlan, 1966) (often incorrectly called 'cellulase'). On purifica- tion (Byrne et al., 1975), this enzyme was shown to hydrolyze xyloglucan in preference to cellulose (Hayashi et al., 1984), the final products of xyloglucan hydrolysis in vitro being the Glc4Xyl3-based structural subunits depicted in Fig. 5.9. The possible involvement of the endo- /3(1 ^ 4)-D-glucanase in the control of growth was further promoted by the observation that one of the structural subunits of xyloglucan, the nonasaccharide Glc4Xyl3GalFuc (Fig. 5.9) was able, at nanomolar concentrations, to inhibit auxin-induced extension of pea stem sections (York et al., 1984). Was this a natural feedback control mechanism? It was recognized that endo-(3{\ -^ 4)-D-gluca- nase action alone could not control cell expansion, since wall-loosening is reversible, and the gluca- nase was capable of breaking glycosyl linkages but not re-forming them. More recently an enzyme has been discovered which breaks internal linkages in the xyloglucan backbone reversibly. This enzyme, xyloglucan ^wJo-transglycosylase (or XET), cata- lyzes the breakage of the intersubunit glycosyl linkage in xyloglucans, freeing the portion of the polymer towards the reducing end and transferring the portion of the molecule towards the non-reducing end onto the non-reducing end of a xyloglucan or xyloglucan oligosaccharide 234 J.S. GRANT REID lyase cleaves glycosyl linkages, not by hydrolysis, but by 1,2-elimination, producing oligosaccharides with an unsaturated sugar derivative at the non- reducing end. Both the normal and the unsaturated oligosaccharides are active elicitors. The response is chain-length dependent, the optimum being at about 10-15 sugar residues depending on the plant and the response. In this instance, a 'structural' plant cell wall polysaccharide carries encoded information. A further example of cellular signaling by oligosaccharides derived from 'structural' cell wall polysaccharides was alluded to in the context of xyloglucan metabolism and cell expansion (section 5.6.1). Since the initial observation that the xyloglucan oligosaccharide Glc4Xyl3GalFuc (Fig. 5.9) inhibits auxin-induced elongation of pea stem segments (York et al., 1984), other effects of xyloglucan oligosaccharides on plant develop- ment have been documented (Darvill et al., 1992; Aldington & Fry, 1993). The structural require- ments for the inhibition of auxin-induced growth in pea stem segments have been investigated, and the structural motif Glc2GalFuc at the reducing terminus of the molecule singled out as essential to the effect (McDougall & Fry, 1989). Other oligosaccharides lacking this motif, however, were shown to be weak promoters of growth (e.g. Glc4Xyl3Gal, Fig. 5.9) in the same system (McDougall & Fry, 1990). It is not known for certain whether or not xyloglucan oligosaccharides are regulatory mole- cules in vivo, although they have been detected in culture filtrates of plant cells. If they have a regulatory role, and if they are generated only during the turnover of xyloglucan in the cell wall, then it might be expected that the pathways forming and removing them would be highly regulated. Two enzymes capable of forming the Glc4Xyl3-based oligosaccharides from xyloglu- can are known: endo-(i{\ —^ 4)-D-glucanase and XET (section 5.6.1). An a-L-fucosidase capable of removing the outer fucose residues from xyloglucan oligosaccharides has been character- ized (Augur et al., 1993), as has a /3-galactosidase which catalyzes the removal of the subsequent layer of galactosyl residues (Edwards et al,, 1988). The further degradation of the remaining Glc4Xyl3 core depends on the co-operative inter- action of two enzymes. A xyloglucan-specific a-D- xylosidase (O'Neill et ai, 1989; Fanutti et ai, 1991) catalyzes the removal of only the xylosyl residue attached to the non-reducing terminal glucose, and cannot act again until this has been removed by the action of /3-glucosidase. In all, three rounds each of a-xylosidase and /3-glucosi- dase hydrolysis are required for the total hydro- lysis of the oligosaccharide to glucose and xylose (Fig. 5.33). The term 'oligosaccharin has been coined for those cell wall-derived oligosaccharides which, when applied externally, modify plant develop- ment or metabolism (Darvill et al,, 1992; Alding- ton & Fry, 1993). No oligosaccharin receptors have yet been described, and their status as regulatory molecules in vivo is still uncertain. 5.7 CONCLUSIONS It was mentioned at the beginning of this chapter that plant structural polysaccharides constitute the most important renewable resource on earth, and that their direct use as industrial raw materials is inevitable. It should now be clear that they are of economic and commercial importance for other reasons. This chapter has highlighted current efforts to clarify the links between the chemical structures of plant cell wall carbohydrates, their interactions, cell wall ultrastructural organization, and tissue mechanical and textural properties. The practical goal of exercising control over the mechanical properties of crop plants and the textures of fruits and vegetables adds purpose to the excitement of current research in the area of plant 'structural' carbohydrates. X X X i i 1 4 2 G->G-K5-*G -> X + G->G-K3->G -• X + G T T X X X i 1 + G^^G^^ •* 2X t X X 1 + G + GMiM^ i2 X 2 1 i 3X + 4G 4- 3X + 2G + GM^ <- 2X + 2G + G-»G Figure 5.33 Co-operative action of hydrolytic enzymes in the final stages of xyloglucan oligosaccharide degradation. 1, action of xyloglucan oligosaccharide-specific a-xylosidase. 2, action of /3-glucosidase. CARBOHYDRATE METABOLISM: STRUCTURAL CARBOHYDRATES 235 REFERENCES Albersheim, P. (1975) Sci. Am. 232, 81-95. Aldington, S. & Fry, S.C. (1993) Adv. Bot. Res. 19,1-101. Augur, C , Benhamou, N., Darvill, A. & Albersheim, P. (1993) Plant } . 3, 415-426. Bauer, W.D., Talmadge, K.W., Keegstra, K. & Alber- sheim, P. (1973) Plant Physiol. 51, 174-184. Becker, M., Vincent, C. & Reid, J.S.G. (1995) Planta 195, 331-338. Bird, C.R., Smith, C.J.S., Ray, J.A., Moureau, P., Bevan, M.J., Birds, A.S., Hughes, S., Morris, P.C., Grierson, D. & Schuch, W. (1988) Plant Mo/. Biol. 11, 6541-662. Brady, C.J. (1987) Annu. Rev. Plant Physiol. 38, 155-178. Brady, C.J., MacAlpine, G., McGlasson, W.B. &: Ueda, Y. (1982) Aust. J. Plant Physiol. 9, 171-178. Brown, R.M. Jr, Haigler, C.H., Suttie, J., White, A.R., Roberts, E., Smith, C., Itoh, T. & Cooper, K. (1983) / . Appl. Polymer Sci.: Appl. Polymer Symp. 37, 33-78. Byrne, H., Christou, N.V., Verma, D.P.S. & Mac- lachlan, G.A. (1975) / . Biol. Chem. 250, 1012-1018. Carpita, N.C. & Gibeaut, D.M. (1993) Plant J. 3, 1-30. Chanzy, H. & Henrissat, B. (1985) FEBS Lett. 184, 285-288. Cook, J.A., Fincher, G.B., Keller, F. & Stone, B.A. (1980) In Mechanisms of Saccharide Polymeriza- tion and Depolymerization (J.J. Marshall, ed.), pp. 301-315. Academic Press, New York. Crookes, P.R. & Grierson, D. (1983) Plant Physiol. 72, 1088-1093. Dalessandro, G., Piro, G. 8c Northcote, D.H. (1986) Planta 169, 564-574. Darvill, A., Augur, C , Bergmann, C , Carlson, R.W., Cheong, J.-J., Eberhard, S., Hahn, M.G., Lo, V.M., Marfa, V., Meyer, B., Mohnen, D., O'Neill, M.A., Spiro, M.D., van Halbeek, H., York, W.S. & Albersheim, P. (1992) Glycobiology 2, 181-198. Delmer, D.P. (1987) Annu. Rev. Plant Physiol. 38, 259-290. de Silva, J., Jarman, CD., Arrowsmith, D.A., Stronach, M.S., Chengappa, S., Sidebottom, C. & Reid, J.S.G. (1993) Plant J. 3, 701-711. de Vries, J.A., Rombouts, F.M., Voragen, A.G.J. & Pilnik, W. (1982) Carbohydr. Polym. 2, 25-33. Edwards, M., Dea, I.C.M., Bulpin, P.V. & Reid, J.S.G. (1986) / . Biol. Chem. 261, 9489-9494. Edwards, M., Bowman, Y.J.L., Dea, I.C.M. &c Reid, J.S.G. (1988) / . Biol. Chem. 263, 4333-4337. Edwards, M., Bulpin, P.V., Dea, I.C.M. & Reid, J.S.G. (1989) Planta 178, 41-51. Edwards, M., Scott, C , Gidley, M.J. & Reid, J.S.G. (1992) Planta 187, 67-74. Fan, D-F. & Maclachlan, G.A. (1966) Can. ]. Bot. 44, 1025-1034. Fanutti, C , Gidley, M.J. & Reid, J.S.G. (1991) Planta 184, 137-147. Fanutti, C , Gidley, M.J. &: Reid, J.S.G. (1993) Plant J. 3, 691-700. Fischer, R.L. & Bennett, A.B. (1991) Annu. Rev. Plant Physiol. Plant Mo/. Biol. 42, 675-703. Fry, S.C. (1988) The Growing Plant Cell Wall. Long- man, Harlow, 333 pp. Fry, S.C, Smith, R.C, Renwick, K.F., Martin, D.J., Hodge, S.K. & Matthews, K.J. (1992) Biochem. J. 282, 821-828. Gordon, R. & Maclachlan, G. (1989) Plant Physiol. 91, 373-378. Grierson, D., Slater, A., Spiers, J. 6c Tucker, G.A. (1985) Planta 163, 263-271. Haigler, CH. (1985) In Cellulose Chemistry and its Applications (R.P. Nevell & S.H. Zeronian, eds), pp. 30-83. Horwood, Chichester. Hanna, R., Brummell, D.A., Camirand, A., Hensel, A., Russell, E.F. 8c Maclachlan, G.A. (1991) Arch. Biochem. Biophys. 290, 7-13. Hayashi, T. (1989) Annu. Rev. Plant Physiol. Plant Mol. Biol. 40, 139-168. Hayashi, T. & Maclachlan, G. (1984) Plant Physiol. 75, 596-604. Hayashi, T. & Matsuda, K. (1981) /. Biol. Chem. 256, 11117-11122. Hayashi, T., Marsden, M.P.F. & Delmer, D.P. (1987) Plant Physiol. 83, 384-389. Hayashi, Y., Wong, Y-S. & Maclachlan, G. (1984) Plant Physiol. 75, 605-610. Henry, R.J. & Stone, B.A. (1982) Plant Physiol. 69, 632-636. Jarvis, M.C (1984) Plant Cell Env. 7, 153-164. Keegstra, K., Talmadge, K.W., Bauer, W.D. & Alber- sheim, P. (1973) Plant Physiol. 51, 188-196. Knox, J.P., Linstead, P.J., King, J., Cooper, C & Roberts, K. (1990) Planta 181, 512-521. Labavitch, J.M. 8c Ray, P.M. (1974) Plant Physiol. 53, 669-673. Lin, E.G. &c Brown, R.M. Jr (1989) In Cellulose and Wood - Chemistry and Technology (C Schuerch, ed.), pp. 473-492. Wiley, New York. Lin, E.G., Brown, R.M., Drake, R.R. &C Haley, B.E. (1990) ; . Biol. Chem. 265, 4782-4784. Liners, F., Thibault, J.F. &c van Cutsem, P. (1992) Plant Physiol. 99, 1099-1104. Lloyd, CW. (1984) Int. Rev. Cytol. 86, 1-51. McCann, M.C & Roberts, K. (1991) In The Cyto- skeletal Basis of Plant Growth and Form (CW. Lloyd, ed.), pp. 109-129. Academic Press, London. McCann, M.C, Wells, B. & Roberts, K. (1990) / . Cell Sci. 96, 323-334. McCleary, B.V. & Matheson, N.K. (1983) Carbohydr. Res. 119, 191-219. McDougall, G.J. 6c Fry, S.C (1989) Plant Physiol. 89, 883-887. McDougall, G.J. 8c Fry, S.C. (1990) Plant Physiol. 93, 1042-1048. McNeil, M., Darvill, A.G., Fry, S.C. &C Albersheim, P. (1984) Annu. Rev. Biochem. 53, 625-663. McQueen-Mason, S.J. &C Cosgrove, D.J. (1993) Plant Physiol. 102, 122 (Meeting Abstract). McQueen-Mason, S.J., Durachko, D.M. &: Cosgrove, D.J. (1992) Plant Cell 4, 1425-1433. 236 J.S. GRANT REID McQueen-Mason, S.J., Fry, S.C., Durachko, D.M. & Cosgrove, D.J. (1993) Planta 190, 327-331. Medford, J.I., Elmer, J.S. & Klee, H.J. (1991) Plant Cell 3, 359-370. Meier, H. & Wilkie, K.C.B. (1959) Holzforschung 13, 177-182. Nevell, R.P. & Zeronian, S.H. (1985) Cellulose Chem- istry and its Applications. Horwood, Chichester. Neville, A.C. & Levy, S. (1985) In Biochemistry of Plant Cell Walls (C.T. Brett & J.S. Hillman, eds), pp. 99-124. Society for Experimental Biology Seminar Series 28, Cambridge University Press, Cambridge. Nishitani, K. & Tominaga, R. (1992) / . Biol. Chem. 1^1, 21058-21064. Okazavs^a, K., Sato, Y., Nakagawa, T., Asada, K., Kato, I., Tomita, E. & Nishitani, K. (1993) / . Biol. Chem. 268, 25364-25368. Okuda, K., Li, L., Kudlicka, K., Kuga, S. & Brown, R.M. Jr (1993) Plant Physiol. 101, 1131-1142. O'Neill, R.A., Albersheim, P. & Darvill, A.G. (1989) /. Biol. Chem. 264, 20430-20437. Parker, M. (1984) Protoplasma 120, 224-232. Preston, R.D. (1974) In Dynamic Aspects of Plant Ultrastructure (A.W. Robards, ed.), pp. 256-309. McGraw^-Hill, London. Qi, X., Behrens, B.X., West, P.R. & Mort, A.J. (1995) Plant Physiol. 108, 1691-1701. Reid, J.S.G. (1985a) In Biochemistry of Storage Carbohydrates in Green Plants (P.M. Dey & R.A. Dixon, eds), pp. 265-288. Academic Press, London. Reid, J.S.G. (1985b) Adv. Bot. Res. 11, 125-155. Reid, J.S.G., Edwards, M., Gidley, M.J. & Clark, A.H. (1995) Planta 195, 489-495. Roelofsen, P.A. (1965) Adv. Bot. Res. 2, 69-149. Roland, J.C., Reis, D., Vian, B., Satiat-Jeunemaitre, B. & Mosiniak, M. (1987) Protoplasma 140, 75-91. Ross, P., Weinhouse, H., Aloni, Y., MichaeU, D. & Weinberger-Ohana, P. (1987) Nature 325, 279-281. Ryan, C.A. & Farmer, E.A. (1991) Annu. Rev. Plant Physiol. Plant Mol. Biol. 42, 651-674. Saxena, I.M., Lin, F.C. & Brown, R.M. Jr (1990) Plant Mol. Biol. 15, 673-683. Stevenson, T.T., Darvill, A.G. & Albersheim, P. (1988) Carbohydr. Res. 182, 207-226. Stone, B.A. & Clarke, A.E. (1992) Chemistry and Biology of (l-^3)-/3-glucans. La Trobe University Press, Victoria, Australia. Taiz, L. (1984) Annu. Rev. Plant Physiol. 35, 5S5-657. Talmadge, K.W., Keegstra, K., Bauer, W.D. & Alber- sheim, P. (1973) Plant Physiol. 51, 158-173. Thornber, J.P. & Northcote, D.H. (1962) Biochem. J. 82, 340-346. Timell, T.E. (1964) Adv. Carbohydr. Chem. 19,247-302. Timell, T.E. (1965) Adv. Carbohydr. Chem. 20,409-483. Valent, B.S. & Albersheim, P. (1974) Plant Physiol. 54, 105-108. Vian, B. & Roland, J.C. (1987) New Phytol. 105, 345-357. Vian, B., Reis, D., Mosiniak, M. & Roland, J.C. (1986) Protoplasma 131, 185-199. York, W.S., Darvill, A.G. & Albersheim, P. (1984) Plant Physiol. 75, 295-297. York, W.S., Gates, J.E., van Halbeek, H., Darvill, A.G., Albersheim, P., Tiller, P.R. &c Dell, A. (1988) Carbohydr. Res. 173, 113-132. Wilkie, K.C.B. (1979) Adv. Carbohydr. Chem. Biochem. 36, 215-264.