The two strains of Phlebia radiata Fr. used in this study, P. radiata (wild) and its cellulase less mutant P. radiata Cel 26, were obtained from the collection maintained at the Department of Wood Science, Swedish University of Agricultural Sciences. P. radiata Cel 26 was produced by UV-irradiation of spores [3].

P. radiata (ATCC 64658) was cultivated in laboratory bioreactors using a modified version of the low-nitrogen media described by Fenn and Kirk [4] by omitting vitamins and mineral salts but including glucose (2.8 mM), guaiacol (0.8 mM) (Merck, Darmstadt, FRG) and Tween 80 (0.05% w/v, Atlas, Essen; FRG). For further details and the enzyme purification see Niku-Paavola et al. [5].

Antisera to laccase, MnP/LiP and LiP were raised in New Zealand white rabbits immunized monthly over a 6-month period by intracutaneous injections of the purified proteins (1.0 mg) emulsified in 1 ml of Freund's adjuvant (Difco Laboratories). The immunoglobulin (IgG) fraction of the sera was then prepared by ammonium sulphate precipitation using standard techniques and the sera purified by affinity chromatography. Antiserum to laccase showed a single band with laccase and no cross reactivity with purified MnP or LiP enzymes. LiP reacted with only LiP. One antibody cross reacted with both MnP and LiP (for simplicity termed MnP/LiP here) and was used for simultaneous labelling of both enzymes.

Small wood blocks (3.0×1.5×0.5cm) of birch (Betula verrucosa Ehr.) were placed in 125 ml Erlenmeyer flasks, containing moist (60% moisture content) garden loam and autoclaved. Flasks were subsequently inoculated with a fungal-spore mycelium suspension of P. radiata or P. radiata Cel 26 previously grown on 2.5% w/v malt agar plates. Flasks were then incubated at 22 °C with a relative humidity of 70%. Wood blocks were removed after two months incubation when weight losses were of the order of 45% (P. radiata wild) and 35% (P. radiata Cel 26) respectively. For TEM and immunocytochemical observations, small matchstick samples were removed from the wood blocks and processed as described below.

Wood samples were fixed directly for 3 h at room temperature (RT) in either: (A) 3% w/v glutaraldehyde containing 2% w/v fresh paraformaldehyde in 0.1-M Na-cacodylate (pH 7.2), or (B) 4% v/v paraformaldehyde containing 0.1% w/v glutaraldehyde in the same buffer. Samples fixed in fixative A were subsequently washed in buffer (3×15min) and post-fixed in 2% w/v OsO₄ in buffer for 1 h at RT. Thereafter samples were washed successively in buffer (2×15min) and distilled water (3×15min). Samples were then dehydrated in an ethanol series (20–70%, 10% steps, 10 min each) and infiltrated with London Resin (London Resin Co. Basingstoke, UK) and polymerised overnight at 60 °C. Samples fixed in fixative B were washed in buffer (4×15min) and thereafter dehydrated and infiltrated with resin as described above. Selected material was sectioned with a Reichert FC4 ultramicrotome and sections collected on nickel grids. Post-staining was conducted with 4% aqueous uranyl acetate (5 min) and lead citrate (15 min) [6]. Observations were made using a Philips CM12 STEM operated at various accelerating voltages.

Ultrathin sections for immunocytochemistry were treated as described by Daniel et al. [7,8]. Labelling was performed on osmicated and non-osmicated samples. Sections were post-immunolabelled using polyclonal antibodies to each of the three enzymes (laccase, MnP/LiP, LiP) using 5- and 15-nm gold-labelled secondary antibodies in single- or double-immunolabelling procedures. Controls included omission of the primary antibody stage, use of rabbit IgG instead of the antibody and the use of antibodies preadsorbed with their respective enzyme. In addition, labelling studies were also performed on undegraded wood.

For HR-Cryo-SEM, small samples (ca 1 mm²) were removed from wood blocks partially degraded by monocultures of the white-rot fungi and rapidly frozen in LN₂ slush (−210°C). Thereafter, the samples were transferred into an Oxford CT1500HF Cryo-system, fractured at −95 to −120°C, deep-etched for various periods of time (usually 2–3 min at −95 °C), coated with Au/Pd (ca 4 nm) using a Denton HR sputter coater and observed at low voltages (1.0–8.0 kV) and at −120°C using an Hitachi S-4500 FE-SEM.

Both Phlebia radiata (wild) and P. radiata Cel 26 hyphae colonized the wood structure and produced typical white-rot decay patterns; the former a simultaneous rot where all cell components (cellulose, lignin, hemicelluloses) were progressively removed from the cell lumen outwards and the latter, preferential attack where the lignin and hemicelluloses were selectively removed across the wall leaving the cellulose [2,3]. Using light microscopy and TEM, decay was seen to occur in either a thin advancing zone across the wall in which all the components were degraded (simultaneous attack) or was reflected by the development of a thick zone (i.e. thickness of the secondary wall) in which the cellulose remains (preferential attack); very few structural details of the degraded wood cell wall were however, evident.

In contrast to previous TEM and conventional SEM observations, HR-Cryo-FE-SEM and deep-etching provides a very different morphological view of the fibre wall decay process. Observations of fractured P. radiata wild degraded birch fibres showed two main morphological features: (1) the birch fibre cell wall was composed of aggregates (i.e. spaghetti-like structures) that were organized into concentric lamellae within the fibre S2 cell wall. During decay the lamellae were progressively separated and attack of the aggregates occurred (Fig. 1a and b); and (2) that the fungal hyphae growing within the cell lumina of fibres were in contact with the surrounding degraded walls, mediated by a complex extracellular “slime” matrix (Fig. 1a and d). The slime matrix had two morphological forms: a more fibrillar type was often organized into discrete concentric bands (2–5 per cell) that surrounded lumina hyphae and which radiated across the cell lumen and was in contact with regions of decay (Fig. 1a and c); and a second type, a ‘gel-like’ slime material present in areas of active wood cell wall decay or in regions where the fibre wall appeared to have been almost totally degraded (Fig. 1d and e). The gel-like material was also frequently arranged in radial patterns (Fig. 1d and e) and often appeared attached and integrated with slime material forming the concentric bands at sites of the previous location of the lumen wall (Fig. 1d, arrows).

HR-Cryo-FE-SEM observations on fractured and deep-etched birch cell walls degraded by P. radiata Cel 26 showed similar features in morphology as for the wild type, but differed in that the majority of the S2 wall material remained in the form of concentric orientated lamellae composed of discrete aggregates. Fig. 2a–d show different views of the remaining aggregates demonstrating both their concentric arrangement, spaghetti-like form and fairly uniform size (ca 40–60 nm); features similar to that reported previously for softwood kraft pulps (i.e. delignified fibres) following rapid freezing and cryo-observations or after application of replica techniques [9]. In addition, fractured radial sections gave indications for characteristic ‘droplet’ material lying in an ordered manner across the aggregates (Fig. 2e, arrows). The chemical composition of the droplet materials is currently unknown, but is thought to reflect remaining hemicelluloses or lignin associated with the aggregates. Their location and spatial distribution strongly indicated a natural occurrence rather than any type of artefact produced during preparation. In contrast, TEM sections through preferential degraded cell walls gave some indications for a concentric orientation of the wood components but no indication for aggregates as seen with cryo-fractures (Fig. 2f).

Using chemical fixation and conventional SEM preparation techniques (i.e. fixation, dehydration, critical point drying) shrinkage and coagulation of the slime resulted in the matrix lying closely adpressed to the lumen cell wall and no indication for a temporal distribution was noted (not shown). TEM observations of P. radiata wild degraded wood also showed the extracellular slime material to be arranged in a characteristic concentric fashion and in highly degraded wood fibres, the only remaining wood material was located in middle lamellae and middle lamellae cell corner regions, the slime matrix completely filling the cell lumina and engulfing the remaining wood cell wall (e.g., Fig. 3a and b). TEM also revealed that the slime material was fibrillar in form and that the concentric bands were also interconnected by slime material (Fig. 3a and b). Detailed observations showed the slime material in the thin concentric bands to have a tripartite-type structure (Fig. 3c, inset) similar to that produced by other fungi (e.g., both brown and white rots) during decay of wood [10,11]. Cryo-SEM further revealed the presence of numerous calcium oxalate crystals and deposits of manganese (confirmed by TEM energy dispersive X-ray microanalysis) in the cell lumina that were also embedded in the extracellular slime matrix.

Using immunogold cytochemistry and specific antibodies, the spatial distribution of the oxidative enzymes was determined. Observations revealed the association of the three enzymes (laccase, MnP/LiP, LiP) with the degraded birch wood cell-wall regions in P. radiata wild, as well as the extracellular slime material (Fig. 3c–e). Consistent with the known molecular size of the enzymes (42–64 kDa) and limited porosity of the undegraded wood cell walls, the ligninolytic enzymes were restricted to degraded wood cell wall regions (Fig. 3e) and did not penetrate into the unmodified (i.e. ‘unopened’) wood structure. Some differences were noted in the labelling pattern of the two fungi. Strong labelling of regions of active decay was shown for P. radiata wild while only weak labelling of highly degraded walls was noted for P. radiata Cel 26. TEM also revealed a periplasmic distribution (frequently with periplasmic vesicles) of the enzymes within the luminal hyphae from where these glycoproteins were presumably secreted into the cell lumina in close proximity of the wood cell wall, or possibly conveyed by the slime material onto the wood cell wall (Fig. 4a–f). The enzymes were also localized on the tripartite membranes within the cell lumen (Fig. 3c). Both laccase, MnP/LiP and LiP (not shown) were found localized in highly degraded middle lamellae (Fig. 3d) and middle lamellae cell corner regions distant from the original lumen hyphae indicating some type of transport or diffusion mechanism in operation. Labelling results were more intense for non-osmicated (fixative B) samples but the spatial distribution was similar with osmicated samples. All controls were negative.

The present observations show evidence for the involvement of extracellular slime materials during the degradation of wood fibres by P. radiata, particularly the wild type. Both HR-Cryo-FE-SEM with the rapid freezing of samples and TEM using conventional fixation and processing techniques showed evidence for the presence of slime materials during cell wall decay, while the immunocytochemical approach further showed the association of the enzymes with extracellular slime. The fact that the slime has a high molecular weight with substantial polysaccharide composition is consistent with its retention during the rapid freezing and deep-etching process where it presumably behaves as a cryo-protectant. The association of extracellular slime materials with fungal hyphae during wood decay has been known for many years and it has been variously implicated in wood decay processes [2,10–16]. Previous immuno- and cytochemical studies with the white-rot fungi Phanerochaete chrysosporium and Trametes versicolor [10] has shown the association of LiP and the H₂O₂ producing enzyme pyranose 2-oxidase [14] with extracellular tripartite membranes and slime lining the fibre lumen wall of partially degraded fibres [10,12] while laccase has also been shown localized in extracellular slime associated with Rigidoporus lignosus and Lentinus edodes hyphae [16–18]. More recent work has also shown enzymes associated with tripartite membranes in brown rot fungi (Daniel & Volc, unpublished). Previous NMR studies have also shown the extracellular slime of P. chrysosporium as a β-1,3–1,6-d-glucan [15,19], a polymer similar to that found in fungal cell walls.

To explain the decay features of P. radiata wild, the following working hypothesis is presented:

• (i) the extracellular slime is produced progressively by lumina hyphae and conveys the enzymes (in this case oxidoreductases) involved in the decay process onto the fibre wall (initial stages involving coating of the lumen wall);
• (ii) slime material accumulates on the wall and penetrates into the fibre cell wall following the native fibre architecture causing progressive decay across the wall; with simultaneous decay, cell wall thinning occurs, and with preferential attack selected decay of lignin and hemicelluloses occurs in the wall;
• (iii) enzymes associated with the slime carry out local decay of the fibre wall components causing both delamination of lamellae and aggregate mineralisation;
• (iv) other components associated with the slime (e.g., Mn(II) or oxalate) possibly provide a means for initial non-enzymatic attack and ‘opening’ of the fibre wall for enzymatic attack.

Such a process would be advantageous in that the wall degrading enzymes may be progressively transported and retained at sites of fibre wall attack thus overcoming the known problems for the penetration of high molecular weight proteins (such as enzymes) into the unmodified wood cell wall. In addition, as the fibre wall becomes more distant to the lumenal hyphae in advanced stages of decay (i.e. during middle lamella decay), the slime will provide the medium for retaining contact between the fungus and the fibre wall. Slime has several other advantages in that it can also retain moisture in the substrate and it can also act as a storage material representing a by-product of lignocellulose mineralisation [20].

Previous studies have also implicated slime as a major component of wood decay systems for a range of other wood decay fungi [10–18,20]. However, a perplexing problem in this study is why strong labelling of the enzymes has not so far been recognized in preferentially degraded fibre cell walls. Possibly this reflects a more highly developed non-enzymatic mechanism directed from lumen hyphae by preferential white-rot fungi like P. radiata Cel 26. However, it is also possible that the TEM approach with inherent resin embedding of samples is not the optimal method for showing this process. Interestingly our, cryo-fixed and FE-SEM observations show the wall of P. radiata Cel 26 preferentially degraded birch fibres to have a macrofibrillar aggregate structure, a feature which is consistent with the morphological cell wall structure of delignified fibres in which the cellulose is retained and most of the other components have been removed [21,22]. However this is not evident from TEM observations of the wall structure from delignified P. radiata Cel 26 fibres (Fig. 2f). Additional problems include shrinkage of the fibre wall material during resin embedding and polymerisation, and possibly inherent problems with steric hindrance during immunolabelling studies. It is noteworthy that cellulose within the fibre wall is considered to be arranged in aggregates in a concentric or possibly concentric-radial arrangements (see [2] for references) a feature that is reflected by the present cryo-observations. Lignin, which is expected to complement the cellulose organization of the fibre wall, has been shown elegantly by Ruel an co-workers [23] using permanganate staining to have a concentric orientation using scanning-transmission techniques (STEM) at very high magnifications, but the same pattern has not so far been recognized using specific lignin antibodies in TEM studies, an aspect which may also reflect the problems with the TEM technique. An aspect that also needs further study is to distinguish between the slime material actively secreted by the fungus (presumably the tripartite membranes) and that produced or transformed from the wood polymers during decay. Slime is produced by decay fungi in situations of low nitrogen and high glucose, and represents a way by the fungus for controlling nutritional conditions and the local microenvironment [20]. In this case, the enzymes responsible for glucan synthesis/hydrolysis are also high molecular weight proteins and should also be located at sites of fibre wall decay.