Hydrophobins in the lichen symbiosis

The ‘invention’ of hydrophobins (Wessels, 2000) may have been a critical step in allowing the evolution of plant–fungus symbioses. It is already known that these specialised hydrophobic proteins – only discovered in the late 1980s – have a key role in making hyphal wall surfaces water repellent. Recently, in addition, preliminary investigations have suggested that they might be involved with Hartig net formation in the interaction between ectomycorrhizal fungi and compatible tree hosts (Mankel et al., 2000). In this issue, however, a much wider role in symbiosis is demonstrated – two papers from Rosemary Honegger’s group in Zürich (see pp. 175–184 and 185–195) describe a new role for hydrophobins which is of fundamental importance to the maintenance of the lichen symbiosis. It appears that they prevent waterlogging of key airspaces within lichen thalli, allow channelling of water and nutrients between the symbiotic partners and, in effect, isolate the photobiont, making it dependent on the mycobiont for its supply of water and nutrients. The lichen symbiosis of a fungal ‘mycobiont’ and a photosynthetic algal and/or cyanobacterial ‘photobiont’ partner is one of the best examples of two or more organisms living successfully together. Vegetation dominated by lichens covers approximately 8% of the earth’s land surface and the special attributes of the symbiosis allows lichens to survive and grow in some of the most extreme environments on earth (Purvis, 2000). Of particular relevance is the ability of lichens to survive repeated cycles of drying and wetting, with many lichens able to maintain photosynthesis despite wide changes in thallus water content (Nash, 1996). The work described in this issue explores how successful gas exchange and water flow, vital to the functioning of the lichen symbiosis, can be maintained in internally stratified lichens. These have a multilayered structure with an outer gelatinous cortex, an upper algal layer in which the photobiont partner is entrapped by fungal hyphae, and an inner gas-filled medulla (Fig. 1). They constitute approx. 25% of all lichens (Honegger, 1991). It has long been observed that the inner thallus appears to contain unwettable compounds that resist water entry, thereby preventing waterlogging of the lichen and so maintaining the thallus ‘inner atmosphere’ (Lange et al., 1997). This is vital to the functioning of the lichen symbiosis as it permits gas exchange for the photobiont and thereby continued photosynthesis. The aerial hyphae of the medulla and those surrounding the photobiont have also been observed to have water-repellent properties, with the hydrophobic cell wall surface layer often covering both the mycobiont and photobiont. This is thought to be highly significant with regard to translocation of solutes as it may provide an apoplastic continuum between the partners (Honegger, 1991). Thus, water and dissolved nutrients taken up by the gelatinous outer cortex might then be passively translocated to the photobiont within the apoplastic space of the fungal cell wall with no leakage (Fig. 1), although there is little experimental evidence and symplastic transport has also been proposed (Crittenden, 1989). Key to these functions is the hydrophobicity of the cell wall surface, but the biochemical nature of this has remained unclear until now. Diagram illustrating hydrophobin distribution (bold lines) in lichen thallus. Adapted from Wessels (2000). One suggestion has been that certain lichen secondary metabolites, chemical compounds produced as a result of the symbiosis, were responsible for conferring the hydrophobic properties. However, when the roles of the lichen substances lecanoric and diploschistesic acids were examined in the soil lichen Diploschistes muscorum it was shown that thalli remained water repellent and viable even after the removal of the majority of these hydrophobic compounds (Lange et al., 1997). This involved the harsh procedure of rinsing dry thalli in water-free acetone, again emphasizing the remarkable survival properties of lichens. These findings led the authors to suggest that hydrophobicity may arise from structural features of the lichen thallus and/or the presence of hydrophobic proteins. It is becoming rare for entirely new classes of proteins to be discovered, and so there was considerable interest when the ‘hydrophobins’ were discovered by Wessels and coworkers during the late 1980s and 1990s (Wessels, 1997). These small (90–150 amino acids), secreted proteins are characterised by the following properties (Wessels, 2000; Wösten, 2001): Hydrophobicity. Eight cysteine residues at conserved sites in the amino acid chain. Ability to self-assemble into an amphipathic film at hydrophilic–hydrophobic interfaces. They appear to occur exclusively in mycelial fungi and have a variety of important roles such as involvement with the formation of aerial hyphae and fruit bodies, coating of spore surfaces to ensure hydrophobicity, and allowing adhesion to hydrophobic surfaces. These properties are clearly of great significance to fungal plant pathogens (Wessels, 1997; Wösten, 2001). The reason for their late discovery is that they form highly insoluble aggregates in SDS solutions, and so they were not identified by traditional protein extraction techniques – specialist procedures are required (Wessels, 1997). The first clues that the hydrophobicity of thalli and hyphal cell walls might be due to hydrophobins came from detailed electron-microscopy work. Such studies, many also pioneered by Honneger and coworkers, revealed the presence of a proteinaceous, water repellent layer on the hyphal wall with a distinct semicrystalline rodlet pattern (Honegger, 1991). Significantly, such mosaics of bundles of rodlets in a thin (5–10 nm) layer are characteristic of the self-assembling films formed by hydrophobins (Wessels, 1997; Wösten, 2001). Among the species studied were the widespread foliose lichen Xanthoria parietina– a key specimen being derived from an asbestos roof near Rosemary Honegger’s institute in Zürich (i.e. close to hand!) (Fig. 2). Freeze-fracture electron microscopy again revealed the presence of a rodlet layer covering the fungal and algal wall surface (Scherrer et al., 2000). A major breakthrough was then made with the partial cloning of a gene ‘XPH1’ from X. parietina that produced a deduced protein with the characteristics of a hydrophobin (Scherrer et al., 2000). This gene was cloned by N-terminal sequencing of a hydrophobic protein extract coupled with an RT-PCR approach. This provided the first evidence that hydrophobins are produced by lichen forming-fungi. The foliose lichen Xanthoria parietina showing thallus with sexual reproductive ‘apothecia’ in the centre. Scale bar indicates 0.5 cm. Photo courtesy of P. D. Crittenden. The results reported in this issue provide the first direct evidence of the location of hydrophobins in the lichen thallus, which is of great significance in assessing their role. The results also provide important evidence that hydrophobins are produced by a wide taxonomic range of lichen mycobionts. In the first paper, Scherrer et al. completed the cloning of XPH1 from X. parietina and then went on to study the expression of the gene using Northern blotting. They found that the gene was only expressed when the mycobiont was growing symbiotically, with no expression detected in the mycobiont alone in axenic culture. They then performed in situ hybridisation experiments and demonstrated that there was differential expression of the gene in different parts of the lichen thallus, as illustrated in a series of beautiful photographs. There was high expression of the hydrophobin in the medullary hyphae and those hyphae in contact with the algal layer. However, no signal was detected in the hydrophilic cortex or certain reproductive structures. These results are highly significant as the hydrophobins were detected precisely in those regions where they would be required to provide a hydrophobic coating to prevent the interior of the thalli becoming waterlogged (Honegger, 1991). Hydrophobins were also only produced during the symbiosis, with presumably no need for their expression in the mycobiont alone. One surprise was that only very weak expression of the gene was detected in the algal region, where a rodlet layer consistent with the presence of hydrophobin protein had already been observed in microscopy studies. The authors suggested that this might be related to a concentration effect, with rodlet layers as seen around the algal cells only formed at low hydrophobin concentrations – indicating that hydrophobins were produced by the hyphae of the medulla and then transported to the algal layer. Whatever the explanation, the observation of the rodlet layer around the algal cells would be consistent with hydrophobins having an important role to play in maintaining the apoplastic continuum between the partners (Honegger, 1991). In the second paper, Trembley et al. studied the expression of hydrophobin genes in the taxomomically unrelated basidiolichen Dictyonema glabratum. Lichens formed by basidiomycete fungi are very rare, constituting only 0.4% of all lichens (Honegger, 1991). Despite their taxomonic differences, a common feature with X. parietina is that gas-filled spaces are present in the photobiont layer of D. glabratum even when the lichen is water-saturated. This is significant for D. glabratum as it undergoes daily cycles of desiccation and hydration. Also, the walls of the hyphae lining air cavities are covered by a hydrophobic, rodlet-patterned layer indicative of hydrophobins. In previous investigations, Trembley et al. (2002) identified three hydrophobin genes from D. glabratum, termed ‘DGH1’, ‘DGH2’ and ‘DGH3’, which shared 54–66% amino acid identity. Lichen samples had this time been collected from a Costa Rican rain forest. In the work described in this issue, Trembley et al. then proceeded to perform in situ hybridisation experiments to determine the expression of the three genes within the lichen thallus. They found differential expression by hyphae according to age, location and particular hydrophobin. All three hydrophobins were expressed in young, early stratifying edges of fruitbodies. However, as fruit bodies became more clearly divided into distinct layers a different pattern was evident. Hydrophobins DGH1 and DGH2 were expressed by hyphae in the photobiont layer and aerial hyphae in the uppermost parts of the lower stratum. By contrast, DGH3 was strongly expressed in the hyphae of the boundary layer between the upper cortex and the photobiont layer. Taken as a whole, these results were again consistent with hydrophobins being detected precisely in those regions where they would be required to provide a hydrophobic coating to prevent the interior of the thalli becoming waterlogged and maintain apoplastic continuum (Honegger, 1991). The fact that DGH1, DGH2 and DGH3 had different distributions suggests that they might have different functional roles; it is pertinent to note that other basidiomycete species also produce multiple hydrophobins which appear to have different functional roles (Wösten, 2001). In a final development in the second paper, Trembley et al. investigated the cellular location of the main hydrophobin DGH1 within thalli of D. glabratum. This was achieved by raising antibodies against DGH1 which were then labelled for use in electron microscopy studies. It was found that the protein was present at high levels in the electron dense outer layers of fungal cell walls lining the gas-filled spaces in the photobiont layer, and in hyphal walls of the aerial hyphae in the gas-filled lower stratum. Labelling was also evident in the transition zones between the two strata at points of interhyphal contact. All these results are beautifully illustrated by a further series of electron micrographs. The cellular distribution of the secreted hydrophobins is again strongly consistent with a role for the hydrophobins in maintaining a water repellent coating for hyphae at key positions within the lichen thallus. It has been reported since 1895 onwards that the interior of many lichen thalli are water repellent and that this is important to the functioning of the symbiosis (Lange et al., 1997). Honegger and coworkers have now provided clear evidence – through a series of well executed experiments, combining the diverse disciplines of biochemistry, molecular biology and light/electron microscopy – that hydrophobins have a key role in providing this property. Ideally the role of hydrophobins would now be tested by gene knockout, and although molecular studies with lichen-forming fungi are still at an early stage, advances have been made and gene cloning is at least feasible (Gargas et al., 1995; Murtagh et al., 2000; Dyer et al., 2001; Miao et al., 2001). Further work is also required to investigate the importance of hydrophobins in nonstratified lichens, which are formed by the majority of lichen-forming fungi (Honegger, 1991).