Hyphal cell walls from the plant pathogen Rhynchosporium secalis contain (1,3/1,6)-β-d-glucans, galacto- and rhamnomannans, (1,3;1,4)-β-d-glucans and chitin


G. B. Fincher, School of Agriculture, Food and Wine, and the Australian Centre for Plant Functional Genomics, University of Adelaide, Waite Campus, Glen Osmond, South Australia 5064, Australia
Fax: +61 8 8303 7102
Tel: +61 8 8303 7296
E-mail: geoff.fincher@adelaide.edu.au


A procedure has been developed for the isolation of cell walls from the hyphae of the causal agent for barley leaf scald, Rhynchosporium secalis (Oudem) J.J. Davis. Based primarily on monosaccharide linkage analysis, but also on the limited use of linkage-specific glucan hydrolases and solvent fractionation, the walls consist predominantly of (1,3/1,6)-β-d-glucans, (1,3;1,4)-β-d-glucans, galactomannans of (1,2;1,6)-Manp residues and (1,5)-galactofuranosyl [(1,5)-Galf] side chains, rhamnomannans of (1,6)-Manp residues and rhamnopyranosyl [(1,2)-Rhap] side chains, and chitin; the walls also contain ≈23% (w/w) protein. Electron microscopy shows the presence of distinct inner and outer wall layers. Treatment of wall preparations with guanidine hydrochloride dissolves the outer layer and enables separate analysis of the inner and outer walls. The insoluble, inner wall layer is composed of (1,3/1,6)-β-d-glucans, galacto- and rhamnomannans, (1,3;1,4)-β-d-glucans and chitin, whereas the soluble outer wall material contains a high proportion of rhamnomannan, and smaller proportions of galactomannan, (1,3;1,4)-β-d-glucan and (1,3/1,6)-β-d-glucan with only trace levels of chitin. It was confirmed by immunochemical and enzymatic analysis that at least a portion of the (1,3;1,4)-β-d-glucan component of the inner wall exists as a (1,3;1,4)-β-d-glucan. The analyses not only provide information that is important for a complete understanding of the interactions between R. secalis and barley, but they also identify potential targets for the development of fungicides or resistant transgenic barley varieties.


alcohol insoluble


















inner wall






outer wall










The imperfect fungus Rhynchosporium secalis (Oudem.) J.J. Davis is the causal organism of leaf scald or leaf blotch, which is an important fungal disease of barley (Hordeum vulgare L.) and can cause severe yield losses in cooler and semi-humid cropping regions around the world [1]. Despite the apparent absence of a sexual state, recent analyses of DNA structure and gene sequences indicate that R. secalis is a member of the ascomycetes group of fungi [2–4].

Following spore germination on the surface of the leaf, cuticular penetration occurs and a subcuticular fungal mycelium is established [1,5,6]. Very few hyphae penetrate into the underlying mesophyll layer of the leaf, until the epidermal cells collapse and the mesophyll cells in contact with the mycelium are destroyed [6]. Conidia are produced on short conidiophores that protrude from the subcuticular mycelium through holes in the cuticle [5–7]. Throughout this process, the R. secalis does not appear to degrade plant cell walls and does not form haustoria within individual host cells. The latter observations suggest that fungal toxins might be involved in destruction of the epidermal and mesophyll cells, and hence in disease development [6]. A potent toxic peptide, designated NIP1, has been purified from R. secalis and characterized [8].

Comparisons of the interactions of R. secalis with susceptible and resistant cultivars of barley indicate that resistance to infection can be expressed at several levels. Inhibition of spore germination and germ tube growth on the leaf surface is often associated with noncompatible interactions [6]. Penetration of the fungus through the cuticular layer might be physically inhibited by the rapid production of relatively large, callosic papillae around the penetration pegs [7,9], although more recent work suggests that callosic papillae do not impede fungal penetration when Arabidopsis leaves are infected with powdery mildew fungi [10,11]. Further, inhibition of subcuticular mycelium development through a gene at the RrsI locus might represent another point at which resistance is expressed [6]. Chitinases and (1,3)-β-d-glucan endohydrolases are present in the leaves of both resistant and susceptible cultivars of host plants, and are assumed to slow hyphal growth through degradation of fungal cell wall polysaccharides. In response to fungal attack, the activities of both plant enzymes increase more rapidly and to higher levels in resistant cultivars [12,13]. Additional pathogenesis-related proteins might also play a role in resistance. For example, thaumatin-like PR5 proteins selectively inhibit fungal growth in transgenic wheat [14].

Against this background, it is clear that physical contact between the cell walls of the invading fungus and the walls of the host plant represents an important early event in the infection process. Furthermore, the rapid degradation of either plant or fungal cell walls can be a key determinant of the success or otherwise of infection, because both fungal penetration of host cells and certain plant responses are targeted to degradation of the walls of the opposing organism [15–17]. In some plant–pathogen interactions, fragments of fungal cell wall polysaccharides, including (1,3;1,6)-β-d-oligoglucosides and chito-oligosaccharides, elicit further defence responses in the host plant that is under attack [18–21]. Regarding the interactions between R. secalis and barley, there is a considerable body of information on the composition of the cell walls of the host barley plant [22], but little or no information on the cell walls of the fungus. Here, we have developed a procedure for the isolation and fractionation of hyphal cell walls from R. secalis. The wall preparations have been examined by electron microscopy and analysed using chemical and enzymic techniques. In particular, permethylation analysis and linkage-specific enzymes have been used to identify major polysaccharide constituents of the fungal walls.


Isolation and fractionation of cell walls

The procedure used to isolate the R. secalis walls is summarized in Fig. 1. Prior to cell lysis and cell wall isolation, fungal mycelia were dispersed in aqueous buffer. Sonication was used to assist in the dispersal of the mycelial mat. Cell lysis was effected in two steps. First, the dispersed mycelia were subjected to a freeze–thaw treatment, and second, the suspension was passed through a French pressure cell, in which high-pressure shear forces are applied to break open cells. Any large, unbroken clumps of mycelial cells were retained on nylon mesh with a 280 μm linear pore size following wet filtration of the suspension. The cell wall material that passed through the nylon mesh was thoroughly washed and dried, but microscopic examination showed the presence of adventitious protein or membranous particles adhering to the walls (data not shown). Adherent particles, which were presumably of cytoplasmic origin, were removed by successive treatment with phenol and chloroform/methanol. These treatments would also be expected to inactivate any wall-bound enzymes that might alter constituent polysaccharides. The yield of the resultant total cell wall preparation was ≈0.35 g from 10 g fresh weight of mycelium, or 3.5%, although it must be remembered that the wall preparation procedure cannot be considered quantitative.

Figure 1.

 Procedure for the isolation of R. secalis cell walls and wall fractions. Treatments of the hyphae and subsequent wall fractions are indicated in the order in which they were performed. The major steps in the process were aimed at reducing the level of cytoplasmic contamination of the wall preparation without degrading integral wall components, and are described in detail in the text.

The total wall preparation was subsequently extracted with 6 m guanidine hydrochloride, which removed the outer layer of the wall. Approximate yields of the guanidine hydrochloride-soluble extract and the residual, insoluble wall preparation were 0.3 and 3%, respectively, based on the fresh weight of the fungal mycelium. Again, it should be noted that the fractionation procedure was not likely to be quantitative, although the cumulative yields of the two fractions compared with the total wall fraction were close to 90%.

Morphological appearance of R. secalis cell walls

TEM of intact hyphal cells before cell lysis and wall isolation revealed that the hyphal walls consisted of two readily distinguishable layers (Fig. 2). The thicknesses of the outer and inner layers were 0.2–1.2 and 0.1–0.35 μm, respectively. Although at the hyphal tips both layers appeared thicker than in the mature subapical wall (Fig. 2B), this effect might be explained by apparent distortions caused by glancing sections, which are often observed when tangential sections are taken of elongated, thin and flexible objects, particularly where the rigidity and other physical characteristics of the sectioned biological specimen vary.

Figure 2.

 TEM of R. secalis hyphae. Sections through the mycelial mat, showing adjacent hyphae (A) and higher magnification sections of individual hyphal tips (B,C), in which the inner (IW) and outer (OW) cell wall layers are clearly evident. There is some evidence for additional layers within the inner wall sections (B,C). These hyphae were used for the preparation of cell walls and wall fractions as described in Fig. 1.

SEM of the total cell wall preparation showed that the isolated wall fragments retained the tubular morphology of hyphal cells, and further confirmed that the walls of R. secalis consisted of at least two layers (Fig. 3). The outer layer of isolated walls was ‘rough’ or ‘rippled’ in appearance and showed considerable surface relief, whereas the inner surface of the walls appeared generally smooth (Fig. 3A). Following extraction of the total wall preparations with 6 m guanidine hydrochloride, the walls lost the ‘rippled’ relief on their outer surfaces; both inner and outer surfaces of the residual walls were smooth in appearance (Fig. 3C). Thus, through a combination of SEM and TEM, it has been possible to confirm the presence of the two wall layers, to obtain a surface view of the walls and to monitor the presence of adhering contaminants during the wall preparation process.

Figure 3.

 SEM of isolated cell walls and wall fractions from R. secalis. (A) A ‘crude’ hyphal cell wall fragment after disintegration and washing with buffer, showing the two IW and OW layers (arrows) and the rough surface of the outer wall layer. (B) A hyphal wall fragment, again showing the two wall layers (arrows) and the rough surface of the outer layer. (C) Inner wall layer following guanidine-HCl extraction, showing the removal of the rough outer wall layer. At this stage there was no evidence for two wall layers, and the surfaces of all walls had a smooth appearance.

Composition of the cell wall fractions

The protein contents of the three cell wall preparations were calculated from amino acid analyses (Table 1). The total wall preparation contained ≈23% (mol·mol−1) protein, the inner wall fraction ≈11.5% (mol·mol−1) protein and the outer guanidine hydrochloride-soluble fraction contained ≈37% (mol·mol−1) protein. The most abundant amino acid residues in the proteins were Ala, Gly, Glx, Asx and Ser, and no major differences were detected between proteins in the total wall and the two wall fractions (Table 1). Uronic acid contents were 2–3% (w/w) for the three wall preparations. In view of the difficulties imposed by contaminating glycogen on permethylation analysis of cell wall polysaccharides, particular care was taken to measure glycogen content. The total wall preparation contained < 2% (w/w) glycogen, and the inner residual layer after guanidine hydrochloride extraction contained 0.1% (w/w) glycogen. There was insufficient material to measure glycogen in the guanidine hydrochloride-soluble outer wall layer material.

Table 1.   Amino acid compositions (mol%)a of wall-associated proteins.
Amino acid residueWholeCell wall preparation
Inner layerOuter layer
  1. Excludes Trp. Glx = Glu + Gln; Asx = Asp + Asn.


Analysis of cell wall polysaccharides from R. secalis

The monosaccharide compositions of the total cell wall preparation were determined directly by acid hydrolysis (Table 2). The most abundant monosaccharide in the total wall preparation was glucose (Glc), with lower levels of mannose (Man), galactose (Gal), rhamnose (Rha), N-acetyl-d-glucosamine (GlcNAc) and N-acetyl-d-galactosamine (GalNAc).

Table 2.   Monosaccharide composition (mol%) of whole cell wall preparations from R. secalis. tr, < 1 mol%.
SugarWhole cell walls

The linkage positions of individual monosaccharides within the polysaccharides were determined by permethylation analysis (Table 3). The Glc-containing polysaccharides of the total cell wall preparation consisted predominantly of (1,3)-linked, (1,4)-linked and terminal Glc residues, although significant branching or substitution through C(O)6 was also evident (Table 3). Only trace amounts of (1,6)-linked Glc residues were detected. The trace levels of (1,4,6)-linked Glc residues confirmed that there was little, if any, contamination of the wall preparations with glycogen. Thus, glycogen accounts for < 2% of the total wall preparation and < 0.1% in the inner fraction, indicating that it is a minor contaminant and therefore that intracellular material is also a minor contaminant. Very small amounts of the polysaccharides with (1,3)- or (1,3,6)-linked Glc residues were extracted with guanidine hydrochloride into the ‘outer wall’ fraction; most of these residues were associated with the guanidine hydrochloride insoluble, inner wall fraction (Table 3).

Table 3.   Linkage composition (mol%) of isolated cell walls and wall fractions. tr < 0.5 mol%. Values in bold correspond to the sum of linkages for each monosaccharide including trace values.
MonosaccharideDeduced linkageInner layerOuter layerWhole cell walls (experimental)Whole cell walls (theoretical)a
  1. a Theoretical value, determined by addition of 90% of inner and 10% of outer mol% values, to represent relative yields. b Terminal rhamnose is deduced from 1,5-di-O-acetyl, 2,3,4-tri-O-methyl rhamnitol, etc. c Not detected.


In the Man-containing polysaccharides (1,2)-linked Manp residues predominated, although there were also significant levels of terminal, (1,6)-, (1,3,6)-, (1,2,3)- and (1,3)-linked Manp residues (Table 3). Although the Manp residues from the permethylation analyses were found in similar proportions in the inner wall layer, the total proportion of Manp was greater in the outer wall than in the inner wall fraction. Galf residues were relatively abundant in the polysaccharides of the total wall preparation, where they accounted for 7% (mol·mol−1) of wall polysaccharides, but significant levels of (1,3)- and terminal galactopyranosyl (Galp) residues were also present (Table 3). Galf and Galp residues were determined from the mass spectra of partially methylated alditol acetates. Based on the spectra, t-Gal and 3-Gal must be assigned Galp. Whereas 4-Galp and 5-Galf have similar spectra, the presence of 5-Galf and absence of 4-Galp in other ascomycetes strongly suggests that the derivatized sugar detected was 5-Galf. Terminal Rhap and (1,4)-linked GlcNAcp were also detected (Table 3). The Rhap residues were more abundant in the outer wall than the inner wall, although the 4-GlcNAcp was most abundant in the inner wall fraction (Table 3).

The alcohol insoluble residue (AIR) of hyphae yielded 7.5% (w/w) of acid-insoluble crystalline material. This was composed of terminal (4 mol%), 3- (26 mol%) and 4-linked Glcp (14 mol%), and terminal (1 mol%) and 4-linked GlcNAcp (55 mol%) residues. AIR was used rather than inner and outer fractions because of the large amount of material required to determine crystalline wall material and (1,3;1,4)-β-d-glucan content.

The (1,3;1,4)-β-d-glucan content was determined to be ≈1% (w/w) of the hyphal AIR using the commercial assay kit. This was consistent with the amount estimated by comparison of the HPAEC profile of lichenase-digested Rhynchosporium AIR and barley flour (Fig. 4). The HPAEC profile revealed that the (1,3;1,4)-β-d-glucan structure consisted primarily of the trisaccharide, G4G3G, with small amounts of the disaccharide, laminaribiose (G3G), and the tetrasaccharide, G4G4G3G. Immunogold labelling of hyphae with the (1,3;1,4)-β-d-glucan-specific antibody (Fig. 5) showed clear and obvious labelling of the inner cell wall, with little or no labelling of the outer wall.

Figure 4.

 HPAEC profile of the (1,3;1,4)-β-d-glucan endohydrolase digests of R. secalis cell walls. A no-enzyme control sample was run to show that the oliogosaccharides were generated from enzyme treatment of the sample. The retention times of G3G, G4G3G and G4G4G3G were determined from the chromatography of the standards, laminaribiose (G3G), laminaritriose (G3G3G) and a digest of (1,3;1,4)-β-d-glucan-containing barley flour (G4G3G and G4G4G3G).

Figure 5.

 TEM of R. secalis wall fragments labelled with the (1,3;1,4)-β-d-glucan-specific mAb. Sections were directly labelled with the antibody (A) or labelled after pre-treatment with a (1,3;1,4)-β-d-glucan specific endohydrolase (B). Pretreatment with the enzyme greatly diminished labelling, confirming the detection of (1,3;1,4)−β-glucan. IW and OW layers are indicated.

Estimation of wall polysaccharide composition

The overall polysaccharide composition of the total wall preparation (Table 4) was estimated from the permethylation data in Table 3, based on a number of assumptions from previously published structural analyses of fungal wall polysaccharides, particularly those from the ascomycetes. Galactomannan was estimated by addition of monosaccharide linkages based on a ratio of 5-Gal:2-Man:6-Man:2,3-Man:2,6-Man of 5 : 1 : 1 : 1 : 1, as described by Fontaine et al. [23] for the alkali-insoluble galactomannan isolated for Aspergillus fumigatus. In this case, all the 5-Galf residues contribute to this polysaccharide and the appropriate Manp residues are summed based on the predicted ratio. (1,3;1,4)-β-d-Glucan was estimated by addition of 4-Glcp and 3-Glcp according to the ratios predicted from the HPAEC analysis of lichenase-digested AIR, where 4-Glc:3-Glc is estimated at 2 : 1. Any remaining 4-Glcp not accounted for by this estimate was attributed to 4-linked glucan. Rhamnomannan/mannan was estimated by addition of all Rhap linkages and all remaining Manp linkages. Glucan was estimated from the sum of all remaining Glcp linkages and chitin was estimated from the 4-GlcNAc content of the wall. Over 80 mol% of the wall preparations and wall fractions could be accounted for through these estimations (Table 4).

Table 4.   Polysaccharide composition (mol%) of R. secalis cell walls and wall fractions, deduced from their linkage compositionsa.
PolysaccharideInner layerOuter layerWhole cell walls
  1. Estimate based on the methylation data from Table 3, as described in the Results.



When intact hyphae from the causal agent of barley leaf scald, R. secalis, were examined by TEM prior to wall isolation, two quite distinct wall layers were evident (Fig. 2). The layers were in the range 0.1–1.2 μm in lateral walls, but became much thicker at hyphal tips (Fig. 2). The outer layer was relatively electron dense, but appeared to be comprised of somewhat fluffy, loosely packed material. The inner layer of the cell wall was relatively electron transparent, but denser and, in some sections, appeared to be multilamellate (Fig. 2B,C). Similar ultrastructural features have been observed in the walls of the yeast Saccharomyces cerevisiae and in filamentous fungi [24,25].

A procedure was developed for the isolation and fractionation of hyphal cell walls from the R. secalis fungus (Fig. 1). Our initial objectives for the wall isolation were to develop a method that was rapid, simple and resulted in a wall preparation with acceptably low levels of glycogen and other contaminating cytoplasmic material, such as protein. The fungal mycelium was therefore subjected to a preliminary freeze–thaw treatment before disruption of the cells in a French pressure cell. The homogenate was sonicated to dislodge adherent cytoplasmic material, which required further SDS, phenol and chloroform/methanol extractions for its complete removal (Fig. 1). Thus, adherent cytoplasmic particles were detected on the inner wall surface before phenol and chloroform/methanol treatments, but little or no adventitious material could be seen on the inner surfaces of the final, total cell wall preparation. Some cold water-soluble material may have been lost from the walls during isolation, through the use of aqueous buffers to wash the walls at certain stages of the preparative procedure (Fig. 1).

SEM of the isolated total wall preparation also suggested that there were two layers in the unfractionated walls (Fig. 3A,B), consistent with the distinct appearance of two wall layers in sections of the intact fungal cells revealed by TEM (Fig. 2B,C). The inner surfaces of the walls were smooth in appearance, but the outer surfaces were characterized by a rippled relief, in some cases with numerous globular protrusions (Fig. 3A).

Treatment of the isolated walls with 6 m guanidine hydrochloride appeared to dissolve most if not all of the rippled, outer wall layer. The residual wall following guanidine hydrochloride extraction, which accounted for ≈85% by weight of the total wall preparation, was smooth on both its internal and external surfaces (Fig. 3), contained a single layer and appeared to comprise the inner layer of the original wall preparation. Thus, the distinct bilamellate appearance of the total cell walls was lost after guanidine hydrochloride extraction. It was concluded from these observations that the outer wall layer was removed during guanidine hydrochloride extraction, and that the guanidine hydrochloride-soluble material was therefore likely to be enriched in outer wall components. However, it must be emphasized that the guanidine hydrochloride extraction was unlikely to be quantitative or completely specific for the outer wall. Thus, although the two wall fractions were likely to be enriched in inner and outer wall components, they would almost certainly contain components from both layers. In addition, the yield of guanidine hydrochloride-soluble and -insoluble material only accounted for 95% (w/w) of the total walls, and it is likely that some material was lost during the ethanol precipitation of the guanidine hydrochloride-soluble fraction.

Amino acid analyses of the total wall, inner wall residue and guanidine hydrochloride-soluble proteins were used to estimate the protein contents of the intact walls and the ‘inner’ and ‘outer’ layers. The guanidine hydrochloride-soluble material consisted of ≈37% (w/w) protein, which was considerably higher than the 11.5% in the inner wall residue and the 23% found in the total wall preparation. The high protein content of the guanidine hydrochloride-extracted material was consistent with the higher electron density of the outer wall layer observed by TEM (Fig. 2). The outer layers of walls from yeast and other fungi are known to be enriched in mannoproteins [26], glycosylphosphatidylinositol-anchored proteins [27], Pir proteins [28], agglutinins [29], hydrophobins [30] and a range of enzymes [31,32]. Despite these apparent differences in protein distribution across the wall, amino acid compositional analyses revealed no distinctive or unusual features between the proteins in the three samples, and amino acid compositions were quite similar in the total walls and in the two guanidine hydrochloride fractions (Table 1).

Permethylation analysis of the walls allowed the identification of Glc-, Man- and Gal-containing polysaccharides in the walls (Table 3). The 4-linked GlcNAcp residues are likely to be components either of chitin, or of tightly bound wall glycoproteins, because they are found predominantly in the guanidine hydrochloride-insoluble wall fraction. The presence of terminal, 3- and 3,6-linked Glcp residues is consistent with the presence of branched or substituted (1,3/1,6)-β-d-glucans. The low levels of 6-Glcp are consistent with observations by Bernard and Latge [33], in which the ascomycete, A. fumigatus, contains a branched 3,6-glucan but not a 6-linked glucan of the type found in walls of S. cerevisieae. Similarly, the linkage composition of the Man-containing polysaccharides in the total wall preparation is consistent with the presence of (1,2;1,6)-α-d-mannans and (1,3;1,6)-α-d-mannans. Whether the mannans are phosphorylated is not known at this stage, although some are likely to be covalently linked to proteins to form mannoproteins [13,34]. Some of the Man-containing polysaccharides are potentially rhamnomannans, similar to those from the ascomycete Sporothrix schenckii [35], for which the (1,6)-Manp backbone is substituted with a Rhap–(1,2)-Rhap side chain at C(O)3. Based on the presence of 5-Galf, 2-Manp, 6-Manp, 2,3-Manp and 2,6-Manp, other mannans are predicted to consist of the galactomannan-type, which are a component of the alkali-insoluble walls of A. fumigatus [23]. The 4-linked Glcp residues may originate from cellulose, however, this is unlikely because there is no convincing evidence for the presence of cellulose in ascomycete species. However, members of the ascomycetes do contain glucans with alternating (1,3)- and (1,4)-Glcp linkages in the form of (1,3;1,4)-α-d-glucan (nigeran) and the less commonly identified (1,3;1,4)-β-d-glucan [23]. Immunogold labelling with the (1,3;1,4)-β-d-glucan-specific antibody and the results of the commercial assay for (1,3;1,4)-β-d-glucan support the presence of (1,3;1,4)-β-d-glucan in the wall of R. secalis. Furthermore, oligosaccharide profiles of the (1,3;1,4)-β-d-glucan-specific endohydrolase digests suggest that the glucan is primarily composed of cellobiose repeats interrupted by a single (1,3)-β-Glcp residue rather than alternating (1,3)- and (1,4)-Glcp residues.

Perhaps the most important information arising from compositional analyses of the type undertaken here on the R. secalis walls is the identification of particular wall polysaccharides and the estimation of their approximate levels in the walls. It must be acknowledged that permethylation analysis does not allow discrimination between α- and β-linkages, and that considerable caution should be exercised in the deduction of overall polysaccharide composition from methylation data. However, the availability of structural analyses for other fungal wall polysaccharides [23,35–38] can facilitate these estimations, at least for some wall polysaccharides. Alkaline extraction procedures have frequently been used for the isolation and analysis of fungal wall components, but this procedure was avoided here because of the potential for alkaline β-elimination of O-linked sugars in wall glycoproteins, alkaline peeling of wall polysaccharides such as (1,3)-β-d-glucans, partial degradation of wall proteins [39] and removal of alkali-labile non-glycosyl substituents, such as O-acetyl groups, which are commonly found on wall polysaccharides.

Within these interpretative constraints, we suggest that the cell wall polysaccharides of the fungal pathogen R. secalis contain ≈23 mol% branched or substituted (1,3/1,6)-β-d-glucans, ≈26 mol% (1,3;1,4)-α/β-d-glucans, ≈18 mol% galactomannans, ≈17 mol% rhamnomannans and ≈6 mol% chitin (Table 4). In making these calculations, the values for the glucan components of the wall are likely to be the least accurate, because of the complexity of these polysaccharides in fungal walls and because no information on anomeric configuration, other than for the (1,3;1,4)-β-d-glucans, was obtained here.

Whether these polysaccharides are interconnected [40] is not known at this stage. However, the partitioning of different components into the outer, guanidine hydrochloride-soluble phase and the inner, insoluble phase can give some insight as to the organization of the R. secalis cell wall. Chitin is restricted to the inner wall layer, which is also rich in (1,3/1,6)-β-d-glucan, galacto- and rhamnomannans and (1,3;1,4)-α/β-d-glucans, whereas the major component of the outer wall are the rhamnomannans.

In summary, the data obtained here for hyphal cell walls of the fungal pathogen R. secalis indicate that the wall is a bilayered structure with an inner electron-transparent and carbohydrate-rich layer and an electron-dense, protein- or glycoprotein-rich outer layer. Given that the fungal cell walls are the structures that first come into contact with the host, eliciting phytoalexin production in plants and antibodies in animals, information on the composition, disposition and fine chemical structure of constituent polysaccharides, glycoproteins and proteins is necessary for a complete understanding of plant–pathogen interactions. In the case of R. secalis infection of barley, the fungal walls differ dramatically from those of the barley host, and therefore represent an obvious target for fungicides, either through specific inhibition of the biosynthesis of key fungal wall components or through interference with wall assembly and modification [41].

Materials and methods

Culture conditions

The isolate of R. secalis used here came from the culture collection of the Department of Plant Science, University of Adelaide. Stock cultures were maintained on Lima bean agar at 18 °C. The fungi were cultured in 200 mL bottles containing 100 mL of 1% (w/v) yeast extract and 1% (w/v) malt extract, pH 5.5. The shape of the culture bottle was similar to that of oblong Fernbach flasks. The bottles were inoculated with 0.5 × 4.5 cm2 mycelial sheets from cultures grown on the Lima bean agar plates, and were incubated statically at 18 °C for 120 h.

Preparation of cell walls

All steps in the wall preparation were performed at 4 °C and all aqueous buffers contained 1 mm phenylmethanesulfonyl fluoride to inhibit proteolytic enzymes. Fungal mycelium (10 g) was washed three times with 500 mL of 50 mm Tris/HCl buffer, pH 7.0, (neutral buffer) on nylon mesh (pore size 280 μm), resuspended in 10 mm sodium acetate buffer, pH 5.0, containing 1% (w/v) SDS and 10 mm NaCl, and sonicated for 5 min (intensity 4, pulse 10 s; Soniclean; Transtek Systems, Thebarton, Australia). The sonicated material was again washed three times with 500 mL of 0.05 mm Tris/HCl buffer, pH 7.0, on the nylon cloth and frozen in liquid nitrogen. The frozen tissue was thawed and resuspended in 10 mm sodium acetate buffer, pH 5.0, containing 1% (w/v) SDS and 10 mm NaCl, and homogenized using a French pressure cell (SLM Aminco Instruments, Rochester, NY, USA). In our hands, the French pressure cell was the most effective technique for reproducibly breaking open large quantities of the extremely resilient, tough mycelial material. About 80% of cells were disrupted with this method. Manual methods, such as grinding under liquid nitrogen, were ineffective and did not yield reproducible and sufficient quantities of the walls. The broken cells were filtered over the 280 μm nylon mesh to remove large mycelial fragments and intact cells. The filtrate was centrifuged at 2500 g for 30 min. The insoluble material was washed several times with the neutral wash buffer, and subsequently on glass microfibre filters (GF/A 42.5 mm; Whatman, Maidstone, UK) with ice-cold ethanol until the supernatant become clear. The material was resuspended in 500 mL of the neutral buffer and is referred to as the ‘crude cell wall’ fraction.

To remove adventitious cytoplasmic proteins, the wall preparation was mixed gently in phenol saturated with 50 mm Tris/HCl buffer, pH 7.0, for 1 h at 4 °C. Walls were recovered by centrifugation and washed on a glass filter three times with ice-cold ethanol and three times with acetone, and suspended in chloroform/methanol (1 : 1 v/v) for 1 h at 4 °C. The wall suspension was filtered on a glass filter, washed with acetone and dried under vacuum at 22 °C. This preparation is referred to as the ‘total cell wall fraction’.

The total cell wall preparation (0.05 g) was extracted for 24 h in 10 mL ice-cold 6 m guanidine hydrochloride containing 0.1 m CaCl2. The wall residue was recovered by centrifugation and washed with 50 mm Tris/HCl buffer, pH 7.0, repeatedly to remove guanidine hydrochloride. The residue of the walls after guanidine hydrochloride extraction was resuspended in 15 mL 1% (w/v) SDS, heated at 100 °C for 10 min and washed free of SDS solution with neutral buffer over nylon cloth. SDS treatment was shown to remove adventitiously associated cytoplasmic proteins from the wall preparation. The residual wall material was extracted again with ice-cold 6 m guanidine hydrochloride, washed progressively with absolute ethanol and acetone, and dried under vacuum at 22 °C. This preparation is referred to as the ‘inner cell wall fraction’, on the basis of its morphological similarity with the inner layer of the total wall preparation.

The 6 m guanidine hydrochloride-soluble fraction from the walls was precipitated with an equal volume of absolute ethanol. Precipitated material was collected by centrifugation, washed three times with 50% (v/v) ethanol and three times with absolute ethanol, and dried. This preparation was designated the ‘outer cell wall fraction’, on the basis of the disappearance the outer wall layer during 6 m guanidine hydrochloride extraction, although it is possible that it contained some material of inner wall origin.

Transmission electron microscopy

Samples of fungal wall material were fixed in 4% (v/v) paraformaldehyde/1.25% (v/v) glutaraldehyde in 50 mm NaCl/Pi, pH 7.0, containing 4% (w/v) sucrose. The samples were post-fixed in 2% (w/v) osmium tetroxide, pH 7.2, dehydrated through an ethanol series and embedded in LR White resin. Sections (95 nm) were transferred to gold grids, post-stained with 2% (w/v) aqueous uranyl acetate and observed in a Philips CM 100 transmission electron microscope at an accelerating voltage of 75 kV.

Immunoelectron microscopy

Fragments of hyphae generated by grinding in liquid nitrogen were fixed in 2.5% (v/v) glutaraldehyde in NaCl/Pi (pH 7.2), dehydrated in a graded ethanol series and embedded in LR white resin. Ultrathin sections (80 nm) were cut on a Leica Ultracut R microtome (Leica Microsystems, Wetzlar, Germany) using a diamond knife and collected on 100-mesh formvar-coated gold grids. The grids were preincubated in blocking buffer (1% w/v BSA in NaCl/Pi) for 30 min at room temperature and placed in a 1 : 500 dilution of (1,3;1,4)-β-d-glucan monoclonal antibody (IgG, kappa light) [42] (Biosupplies Australia) in blocking buffer for 1 h at room temperature and overnight at 4 °C. The grids were washed three times in NaCl/Pi and twice in blocking buffer and incubated in a 1 : 20 dilution of an anti-mouse secondary IgG conjugated to 18 nm colloidal gold (Jackson ImmunoResearch, West Grove, PA, USA). Grids were washed in NaCl/Pi and several times in H2O before post-staining in 2% (w/v) aqueous uranyl acetate. As a control, some sections were pretreated with a purified Bacillus subtilis (1,3;1,4)-β-d-glucan endohydrolase prior to antibody labelling. The sections were viewed on a Philips (Eindhoven, The Netherlands) BioTwin transmission electron microscope and images captured with a Gatan multiscan digital camera.

Scanning electron microscopy

Samples were dehydrated with acetone and dried under vacuum at 22 °C. The dried samples were placed onto metal stubs and coated with carbon and gold. These were examined in a Philips EM 300 scanning electron microscope at an accelerating voltage of 10 kV.

Monosaccharide linkage analysis

Freeze-dried cell walls (≈0.1 mg) were swollen in 100 μL dimethylsulfoxide with sonication for 50 min and held at room temperature for 18 h. Methylation analyses were subsequently performed using the NaOH/CH3I method of Ciucanu and Kerek [43], as described by McConville et al. [44], except that the samples were taken through the methylation procedure twice. Partially methylated polysaccharides were hydrolysed with 90% (v/v) formic acid for 1 h at 80 °C, followed by hydrolysis with 2.5 m trifluoroacetic acid for 5 h at 100 °C. Samples were reduced with 1 m NaBD4 in 2 m NH4OH for 18 h at 4 °C and acetylated with acetic anhydride. Partially methylated alditol acetates were separated on a Chrompack Capillary Column CP-Sil 5 CB LB/MS (Varian, Australia) using a Hewlett-Packard 6890 Series GC System and a Hewlett-Packard 5973 mass selective detector (Agilent Technologies, Palo Alto, CA, USA). All samples were analysed in duplicate.

Monosaccharide analysis

The total wall fraction was hydrolysed with 72% v/v H2SO4 for 1 h at room temperature and, following dilution to 1 m, was further hydrolysed at 100 °C for another 3 h. Monosaccharides were analysed, in duplicate, as their alditol acetates on a Chrompack Capillary CP-Sil 5 column under the same conditions described above [45]. All samples were analysed in duplicate.

Determination of (1,3;1,4)-β-d-glucan

AIR of Rhynchosporium was prepared by collecting hyphae on GF 120 glass fibre paper (Whatman) and washing on the filter with water. The hyphae were frozen in liquid N2 and ground with a mortar and pestle. The ground material was extracted with 80% (v/v) ethanol three times. During the ethanol washes, the AIR was filtered over a 250 μm sieve and the filtrate collected for analysis.

The AIR was treated with acetic acid/nitric acid in water according to Updegraff [46] to determine the amount and composition of crystalline polysaccharides. The crystalline material was measured by weight after the acid treatment and analysed for monosaccharide linkage composition as described above.

(1,3;1,4)-β-d-Glucan content was estimated using a commercial glucan detection kit (Megazyme International, Wicklow, Ireland). The same enzyme (lichenase) used in the kit was used to release oligosaccharides for profiling by HPAEC. The AIR (≈10 mg) was suspended in 1 mL of sodium phosphate buffer (pH 6.5) and 2 U of enzyme added. The mixture was incubated at 50 °C with shaking for 2 h and centrifuged to pellet the insoluble material. The oligosaccharide-containing supernatant (10 μL) was injected onto a CarboPac PA1 column using a Dionex ICS 3000 system equipped with an autosampler (Dionex). Oligosaccharides were eluted at 1 mL·min−1 with a sodium acetate gradient from 50 mm in 0.2 m sodium hydroxide to 350 mm in 0.2 m sodium hydroxide over 15 min. A no-enzyme treated control, a lichenase digest of barley flour, laminaribiose and laminaritriose were included as controls under the same chromatographic conditions.

Other analyses

The amino acid compositions of wall-associated protein were determined by J. Lahnstein (Nucleic Acid and Protein Chemistry Unit, University of Adelaide, Australia). Uronic acids were estimated colorimetrically by the procedure of Galambos [47] and glycogen content of the wall preparations was determined using a total starch assay kit (Megazyme International).


The work was supported by grants from the Australian Research Council (to GBF and MH) and the Grains Research and Development Corporation (to AB and GBF). We thank Hugh Wallwork and Felicity Keiper for providing the fungal isolate and for helpful advice.