• Thalli of Cetraria islandica (Icelandic moss) and Evernia prunastri (oak moss) contain considerable amounts of the linear (1→3), (1→4)-β-glucan lichenin (lichenan), which different proportions of linkage groups in either species.
• The immunocytochemical location of lichenan within the thallus is reported using a monoclonal antibody (barley anti-(1→3), (1→4)-β-glucan antibody) and low-temperature scanning electron microscopy (LTSEM) techniques.
• The antibody labelled ultrathin sections of C. islandica, but not of E. prunastri. In C. islandica, lichenin was located in the extracellular matrix of the peripheral cortex and in a thick outer wall layer of medullary hyphae. LTSEM of fully hydrated and desiccated thalli of C. islandica showed that both the cortical extracellular matrix and the outer wall layer of medullary hyphae shrink dramatically during drought stress, indicating that these are major sites of water storage. A mycobiont-derived, hydrophilic wall surface layer prevents the thallus interior from becoming water-logged at full hydration.
• The LTSEM observations and immunocytochemical data strongly suggest that lichenin is a structural compound, rather than a storage product, with important functions in thallus water relations.
Lichen thalli, the symbiotic phenotype of lichen-forming fungi in association with their photobiont, are known to contain considerable amounts of water-soluble glucans. Best investigated are lichenin (lichenan), a hot-water-soluble linear (1→3) (1→4)-β-glucan, and isolichenin (isolichenan), a cold-water-soluble (1→3) (1→4)-α-glucan, both with different proportions of linkage groups in diverse taxa of lichens (Gorin et al., 1988; Stone & Clarke, 1992). These polysaccharides are usually assumed to be storage products, probably of the fungal partner, but their precise location within the thallus is still unknown.
As reviewed by Stone & Clarke (1992), linear (1→3) (1→4)-β-glucans have an interesting distribution among organisms. They were detected not only in a range of lichen-forming ascomycetes, but also in the Poaceae, that is wild and cultivated grasses, where they occur in various amounts as cell wall components in the seeds, stems and leaves. Among cereals the highest quantities of (1→3) (1→4)-β-glucan were found in the endosperm of barley (Hordeum vulgare L.). Since it is not degraded during fermentation, this glucan has a positive effect on the palatability of beer and, due to its high viscosity, stabilizes the foam; therefore brewers are interested in quantifying its contents. For this purpose a (1→3) (1→4)-β-glucan-specific monoclonal antibody was raised in the laboratory of Bruce Stone; it was successfully used for in vitro quantification and in situ immunocytochemical location of (1→3) (1→4)-β-glucan in barley (Meikle et al., 1994). As this monoclonal antibody is now commercially available (BIOSUPPLIES Australia, Parkville, Australia) it was tested in the present study on ultrathin sections of the two species of macrolichens with the chemically best known lichenins: Cetraria islandica (L.) Ach. (‘Icelandic moss’) with an approx. 3 : 7 proportion of (1→3) and (1→4) linkage groups, and Evernia prunastri (‘oak moss’, the ‘mousse de chêne’ in perfume industry), with an approximately 7.5 : 2.5 proportion of (1→3) and (1→4) linkage groups (Stone & Clarke, 1992).
Materials and Methods
Cetraria islandica samples were collected by S. Ott in a forest area near Mullsjö (Brännasen), Västergötland, South Sweden, in 1990 and kept in the dry state at −20°C in our laboratory. Evernia prunastri samples were freshly collected on twigs of witch hazel (Hamamelis mollis), in the Botanical Gardens of the University of Zürich, Zürich, Switzerland.
Scanning electron microscopy (conventional SEM)
Specimens were chemically fixed in the vapour of a 4% solution of osmium tetroxide, dehydrated in an ascending series of acetone, critical point dried, mounted on specimen stubs, sputter-coated with gold, and examined in a Cambridge Stereoscan (Cambridge, UK) at 20 kV.
Maceration technique for conventional SEM studies
The hydrophilic, gelatinous extracellular material of cortical and medullary hyphae was removed by means of a maceration technique, adapted from Anglesea et al. (1982). Specimens were incubated in a saturated solution of the commercially-available washing powder ARIEL (manufactured by Procter and Gamble (Geneva, Switzerland); with Bacillus subtilis-derived protease) for 36 h at 37°C, with a brief sonication after 24 h, followed by six water washings of 30 min each; chemical fixation in FAA (formaline 40%, acetic acid and alcohol, 3 : 1 : 6) was followed by two water washings of 10 min each. Specimens were subsequently incubated in 1% periodic acid for 2 min, washed five times for 5 min, 5% potassium hydroxide for 30 min, washed five times for 5 min, 1% acetic acid for 2 min, followed by three washings of 20 min each. The macerated specimens were dehydrated in an ascending series of acetone, critical point dried, mounted on specimen stubs and, to improve their conductivity in the electron beam, exposed to the vapour of a 4% solution of osmium tetroxide for 20 min before sputter coating.
Low temperature scanning electron microscopy (LTSEM)
In our laboratory the BioRad cryotrans system (Watford, UK) is interlinked with a Hitachi S 4000 (field emission) scanning electron microscope (Tokyo, Japan). Fully hydrated samples were mounted with white glue (KONSTRUVIT Geistlich, Schlieren), desiccated samples with water-free conductive paste. Both were cryoimmobilized in subcooled liquid nitrogen (LN2), freeze-fractured, sputter-coated with an alloy of gold and palladium and examined on the cold table of the LTSEM at −178°C at an accelerating voltage of 20 kV.
Specimen preparation for transmission electron microscopy (TEM)
The desiccated thalli of C. islandica were allowed to rehydrate for at least 15 h. Under a dissecting microscope, which was kept in a hood, small thallus fragments were dissected in a drop of the aqueous fixative into blades that were < 0.2 mm thick and < 1 mm long. These were transferred to vials containing 1.5% acrolein and 1.25% glutaraldehyde in phosphate buffer, pH 7.1, incubated for 4 h at room temperature, washed 3 times for 20 min in buffer, and postfixed in 2% buffered osmium tetroxide. After dehydration in a graded series of acetone the specimens were infiltrated with either a 1 : 1 mixture of the epoxy resins EPON and Spurr’s (all compounds from PLANO (W. Plannel, GmbH, Wetzlar, Germany), or in the methacrylate UNICRYL (British Biocell). Flat embedding was performed between teflonized microscopy slides.
Immunolabelling at the TEM level
Ultrathin sections, approx. 90 nm thick, were mounted on Parlodium-coated TEM grids and sequentially incubated at room temperature in (1) phosphate buffer saline (PBS, 0.5% NaCl) containing 1% bovine serum albumin (BSA) for 30 min; (2) the monoclonal antibody (mouse antibarley (1→3) (1→4)-β-glucan; BIOSUPPLIES Australia, Parkville) at a protein concentration of 10 µg ml−1; (3) five washings in PBS/0.2% Tween 20; (4) two washings in PBS/1% BSA; (5) the gold-conjugated secondary antibody (goat-antimouse, with 30 nm gold granules; British BioCell) at a 1 : 20 dilution for 1 h; (6) three washings in PBS/0.2% Tween 20; (7) two washings in PBS/1% BSA; (8) two washings in distilled water. As controls, sections were incubated in either the secondary antibody only, or in the monoclonal antibody which had been preincubated for 30 min with barley (1→3) (1→4)-β-glucan (0.1 mg ml−1; BIOSUPPLIES Australia). All sections were stained with 2% uranyl acetate for 15 min at 30°C, followed by one washing with distilled water, and lead mixture according to Sato (1967) for 7 min, followed by two washings with distilled water. Sections were examined in a Hitachi H 7000 TEM at an acceleration voltage of 60 kV.
Defined areas (e.g. outer wall layer, cell wall proper, protoplast) were cut out of photographic prints and quantified gravimetrically with a Mettler PC 440 delta range balance (Mettler Instruments, Greifensee, Switzerland).
An anatomical peculiarity of C. islandica is the thick, gelatinous extracellular matrix of the upper and lower cortex; it amounts to > 80% of cortical biomass, as estimated on the basis of morphometric measurements, whereas the fungal cell wall proper and the protoplast make up less than 10% each. Upon maceration of the specimen, that is the removal of the gelatinous extracellular material, the cortical hyphae are seen as a system of branched and partly anastomosing hyphae (Fig. 1d–f), the massive extracellular matrix being their secretion product. Very thick walls, as seen in freeze-fractured specimens in LTSEM (Figs 1h, 4d–e) or in ultrathin sections in TEM preparations (Figs 2a–c, 3c–d), are a typical feature of medullary hyphae of C. islandica, the protoplast accounting for approximately 14 ± 2% of hyphal biomass. Three different zones are distinguishable in these fungal cell walls: (1) the cell wall proper (approx. 13 ± 1% of hyphal biomass); (2) the thick, electron-transparent outer layer (approximately 73 ± 1% of hyphal biomass); and (3) a very thin, proteinaceous surface layer which is responsible for wall surface hydrophobicity. Crystalline secondary metabolites and oxalate crystals were often seen adhering to and embedded within this water-repellent wall surface layer (Fig. 1i). In hyphae that contact the cellulosic walls of Trebouxia cells the outer layer was found to be less prominent (Fig. 3c–d). Due to the very thick extracellular gelatinous matrix of the cortex and outer wall layer of medullary hyphae the specimens were extremely difficult to infiltrate with methacrylate or epoxy resins and thus difficult to section.
LTSEM studies on water relations in C. islandica confirmed former reports by Scheidegger (1994), indicating that free water does not accumulate intercellularly, but is confined to the symplast and to the apoplast, the latter being exceptionally massive, as shown in freeze-fractured LTSEM preparations (Fig. 4d–e) and ultrathin sections (Figs 2a–c, 3c–d). Even in the fully hydrated state the thallus interior, that is the algal and medullary layers, remained gas-filled due to very significant water-repellency of the wall surfaces (Figs 1g–h, 4d–e). Liquid droplets of unknown origin (exhudates or possibly method-dependent artefacts) were occasionally seen on algal or hyphal surfaces in the thallus interior (Fig. 4d–e). Due to the high water repellency of this wall surface layer, which is also reflected in the contact angle of the droplet with the surface, the formation of a continuous water film on hyphae or algal cells is prevented.
Immunocytochemical tests using the barley anti-(1→3) (1→4)-β-glucan antibody were performed on ultrathin sections of Unicryl (methacrylate) and Epon-Spurr (epoxy resin) embedded specimens. The labelling pattern was the same on both types of sections, but the epoxy resin mixture had better sectioning properties. In C. islandica a strong labelling was associated with electron-transparent areas interspersed with more electron-opaque areas in the extracellular gelatinous matrix of the cortex and more prominent in proximity to the hyphal wall proper (Fig. 3b). A lower labelling intensity was evident at the periphery of the cortex where the extracellular matrix appeared less heterogenous. A very heavy labelling was found in the thick, electron-transparent outer wall layer of the aerial hyphae in the thallus interior (Figs 2a, 3c–d), whilst no labelling was recorded on the protoplast of the mycobiont on the cell wall, the degrading mother cell wall and protoplast of the Trebouxia photobiont (Fig. 3c–d). No reaction was found in control sections with either the preincubated primary antibody (Fig. 2b) or the secondary antibody alone (Fig. 2c).
By contrast to the positive reaction observed in C. islandica the barley anti-(1→3) (1→4)-β-glucan antibody did not recognize the (1→3) (1→4)-β-glucan of Evernia prunastri, the other parmeliacean species investigated. In this species no labelling was detectable in the cortex or in the medullary hyphae, although the ultrastructural organization was substantially similar to that observed in C. islandica (Fig. 3a).
In comparative studies of freeze-fractured LTSEM preparations of fully hydrated and desiccated specimens dramatic changes in the thickness of the cortical layer and the diameter of aerial hyphae in the gas-filled thallus interior were observed (Fig. 4a–e). During desiccation Trebouxia cells shrivelled and were strongly deformed (Fig. 4a–c), a well-known event in lichen photobionts (Brown et al., 1987; Honegger & Peter, 1994; Honegger, 1995; Scheidegger et al., 1995; Honegger et al., 1996). The fungal cells shrank and, in response to the enormous volume changes in their protoplast, underwent cytoplasmic cavitation (Fig. 4a–c), characteristic features of lichen-forming fungi (Honegger, 1995; Scheidegger et al., 1995; Honegger et al., 1996). During rehydration the diameter of thick-walled hyphae of the medullary and algal layer increased by approximately 120%, the major changes being restricted to the cell wall. The diameter of the protoplast increased by only about 20% (the cytoplasmic gas bubble being included in this measurement) and the wall thickness by approx. 250% as calculated from LTSEM images (Fig. 4a–e). As the outer wall layer contributes 72–75% of the hyphal diameter it is primarily this zone that undergoes the most dramatic volume changes in response to fluctuations in the water content of the thallus. During drought stress-induced shrinkage the proteinaceous wall surface layer folded up, often forming longitudinal plicae along the hyphae (Fig. 4c).
As predicted by B. Stone (pers. comm.) the commercially available barley anti-(1→3) (1→4)-β-glucan antibody recognized the lichenin of C. islandica as this presents similar proportions of linkage groups as the barley glucan (Stone & Clarke, 1992). The fact that E. punastri lichenin, with its different proportions of (1→3) and (1→4) linkage groups than in C. islandica lichenin, was not recognized indirectly shows that this particular antibody binds to a glucan complex larger than a single (1→3) (1→4)-β-glucosidic linkage group.
Most likely linear (1→3) (1→4)-β-glucans in the thalli of other members of the Parmeliaceae have a similar localization as shown in C. islandica, most taxa so far investigated ultrastructurally having more or less prominent outer wall layers in the aerial hyphae in the thallus interior; examples are Usnea spp. (Chervin et al., 1968; Malachowski et al., 1980), Parmelia tiliacea (Honegger, 1984, 1986) and Hypogymnia physodes (Fiechter & Honegger, 1988).
As one among other glucans of the thick outer layer of the cell wall lichenin likely contributes substantially to the very special physico-chemical properties of the leathery thallus of ‘Icelandic moss’. Barley glucan and lichenin are opaque in the dry state and glass-clear, translucent in the wet state, as is the cortex of the thallus, a prerequisite for light transmission to the algal layer. Both glucans are highly hydrophilic, absorb water rapidly and, by doing so, swell remarkably. The same applies to the extracellular gelatinous matrix of the cortex and to the hyphal walls, especially the outer wall layer in hyphae of the medullary and algal layers, which shrink reversibly during desiccation (Fig. 4a–e). As a hot-water (45–60°C)-soluble compound lichenin is prone to be washed out of the outer periphery of the cortex; the low labelling observed in this particular area (Fig. 3b) supports this assumption.
The close association of a thick, hydrophilic and a thin, hydrophobic wall layer in aerial hyphae of the thalline interior of C. islandica and other Parmeliaceae is particularly intriguing. Both wall layers play key roles in thallus water relations and, during wetting and drying cycles, in canalizing fluxes of solutes from the thalline periphery to the algal layer and vice versa. For a long time lichen thalli were assumed to fill up with intercellular water at high levels of hydration, a more or less complete hindrance to gas exchange (Honegger, 1991). Such flooding of the thalline interior would take place when the hyphal wall surfaces were composed of hydrophilic glucans such as lichenin and colleagues. In all macrolichens so far investigated ultrastructurally a thin, proteinaceous wall surface layer was found to spread from the contacting hyphae over the wall surface of the photobiont, thus sealing the apoplastic continuum with a hydrophobic coat. This proteinaceous wall surface layer prevents the thalline interior from becoming waterlogged at high levels of hydration and thus is essential for the functioning of the symbiotic relationship (Honegger, 1991, 1997, 2001). Thin, hydrophobic wall surface protein layers seem to be an ubiquitous feature in aerial structures of non-lichenized and lichenized fungi. In several taxa they were shown to be composed of hydrophobins, a class of proteins with very peculiar structural and biochemical properties (Wessels, 1999, 2000; Wösten & Wessels, 1997). The hydrophobic wall surface layer of C. islandica has not been isolated and biochemically characterized as yet, nor is its fine structure known in detail. However, hydrophobins have been characterized at the molecular and structural level in other lichen-forming ascomycetes (Scherrer et al., 2000), and ultrastructurally similar protein coats (so-called rodlet layers) are known from Parmeliaceae and a wide range of other lichen-forming ascomycetes (Honegger, 1984, 1985, 1986; Fiechter & Honegger, 1988, 1991, 2001 unpublished).
Combined with results from LTSEM observations, our immunocytochemical data strongly suggest that lichenin is primarily a structural element of the fungal wall, with important functions in thalline water relations, rather than a storage compound of lichen-forming ascomycetes.
Our sincere thanks are due to Bruce Stone for manyfold help and advice, to Nick Carpita for valuable information, to Verena Kutasi and Jean-Jacques Pittet for their assistance with the artwork, and to the Swiss National Science Foundation for generous financial support (grant Nr. 31–52981.97 to R.H.).