Developmental cross talking in the ectomycorrhizal symbiosis: signals and communication genes

Authors

  • Francis Martin,

    Corresponding author
    1. Equipe de Microbiologie Forestière, Institut National de la Recherche Agronomique, Centre de Recherches de Nancy, 54280 Champenoux, France
    • Author for correspondence Francis Martin Tel: +33 383 39 40 89 Fax: +33 383 39 40 69 Email:fmartin@nancy.inra.fr

    Search for more papers by this author
  • Sébastien Duplessis,

    1. Equipe de Microbiologie Forestière, Institut National de la Recherche Agronomique, Centre de Recherches de Nancy, 54280 Champenoux, France
    Search for more papers by this author
  • Frank Ditengou,

    1. Equipe de Microbiologie Forestière, Institut National de la Recherche Agronomique, Centre de Recherches de Nancy, 54280 Champenoux, France
    Search for more papers by this author
  • Hubert Lagrange,

    1. Equipe de Microbiologie Forestière, Institut National de la Recherche Agronomique, Centre de Recherches de Nancy, 54280 Champenoux, France
    Search for more papers by this author
  • Catherine Voiblet,

    1. Equipe de Microbiologie Forestière, Institut National de la Recherche Agronomique, Centre de Recherches de Nancy, 54280 Champenoux, France
    Search for more papers by this author
  • Frédéric Lapeyrie

    1. Equipe de Microbiologie Forestière, Institut National de la Recherche Agronomique, Centre de Recherches de Nancy, 54280 Champenoux, France
    Search for more papers by this author

Summary

Development of ectomycorrhizas involves multiple genes that are implicated in a complex series of interdependent, sequential steps. Current research into ectomycorrhiza development and functioning is aimed at understanding this plant–microbe interaction in a framework of the developmental and physiological processes that underlie colonization and morphogenesis. After a brief introduction to the ectomycorrhizal symbiosis, the present article highlights recent work on the early signal exchange taking place between symbionts, and sketches the way functional genomics is altering our thinking about changes in gene expression during the early steps of the ectomycorrhiza development.

Introduction

One of the principal ecological niches of fungi is the inside of plant and animal tissues and cells. Although the majority of associations between fungi and eukaryotes is either commensal or, quite often, mutually beneficial, most studies of animal- and plant-associated fungi have dealt with those rare species that cause diseases. Our major interest focuses on ectomycorrhizal fungi, which, unlike pathogenic parasites, have been associated with their host plant for an evolutionary length of time. The ectomycorrhizal symbiosis has evolved repeatedly over the last 130–180 Myr (LePage et al., 1997) and has had major consequences for the diversification of both the mycobionts and their hosts (Hibbett et al., 2000). Ectomycorrhizal fungi mainly belong to the Basidiomycotina (agarics, bolets) (Fig. 1a), although many species are found within the Ascomycotina (truffles) and the switch between saprobic and mycorrhizal lifestyles probably happened convergently, and perhaps many times, during evolution of these fungal lineages (Hibbett et al., 2000). The first mycorrhizal associations must have been derived from earlier types of plant–fungus interactions, such as endophytic fungi in the bryophyte-like precursors of vascular plants (Selosse & Le Tacon, 1998). Ectomycorrhizal symbioses have a distinct host range allowing formation of ectomycorrhiza on a limited set of trees and shrubs (Smith & Read, 1997). Some of these are shown in Fig. 1(b). However, a given species of ectomycorrhizal fungus is usually able to establish a mutualistic symbiosis on a broad range of species, although highly specific interactions may occur (e.g. Rhizopogon vinicolor/Pseudotsuga mensiesii).

Figure 1.

The ectomycorrhizal symbiosis. (a) Boletus edulis, one of the numerous Basidiomycotina able to establish a mutualistic symbiosis with host trees (Photograph courtesy of F. Martin). (b) A seedling of Douglas fir (Pseudotsuga menziesii) colonized by the ectomycorrhizal basidiomycete Laccaria bicolor. The fungal mycelium has developed ectomycorrhizas on the root system and has produced a basidiocarp above ground (Photograph courtesy of P. Klett-Frey). (c) Short roots ensheathed by an ectomycorrhizal fungus. The yellow mantle covers the root tip and rhizomorphs (i.e. hyphal bundles) are extending in the medium (Photograph courtesy of J. Garbaye). (d) Transverse section of a Eucalyptus/Pisolithus ectomycorrhiza showing the external (EM) and internal (IM) mantles: the fungal hyphae have begun to penetrate between the epidermal cells of the root cortex (RC) to form the Hartig net (HN). Epidermal cells are radially enlarged. Extramatrical hyphae (EH) are exploring the medium (Photograph courtesy of G. Chevalier).

In temperate and boreal forests, up to 95% of the short roots form ectomycorrhizas. Ectomycorrhizas have a beneficial impact on plant growth in natural (Read, 1991) and agroforestry ecosystems (Grove & Le Tacon, 1993). Central to the success of these symbioses is the exchange of nutrients between the symbionts (Smith & Read, 1997). The fungus gains carbon from the plant while plant nutrient uptake is mediated via the fungus. In addition, the establishment of the symbiosis is required for the completion of the fungal life cycle (i.e. formation of fruiting bodies).

Differences in organogenetic programmes are reflected in variations in ectomycorrhizal root morphology, but overall there are pronounced developmental similarities suggesting that key developmental programmes are triggered in both symbionts. In all cases, the mycobiont must have the ability to recognize and become associated with its host, escape the host defense surveillance, and establish bi-directional nutrient transfers. Ectomycorrhiza development thus involves multiple genes that are implicated in a complex series of interdependent, sequential steps. Current research into ectomycorrhiza development and functioning is aimed at understanding this plant–microbe interaction in a framework of the developmental and physiological processes that underly colonization and morphogenesis. After a brief introduction to the ectomycorrhizal symbiosis, the present article aims to highlight recent work on the early signal exchange taking place between symbionts, and to sketch the way functional genomics is altering our thinking about changes in gene expression during the early steps of the ectomycorrhiza development.

The anatomy and development of ectomycorrhiza

Ectomycorrhizas are characterized structurally by the presence of a dense mass of fungal hyphae forming a pseudoparenchymatous tissue ensheathing the root (Fig. 1c). This is the Hartig net of intercellular hyphæ, characterized by labyrinthine branching and an outward network of hyphae prospecting the soil and gathering nutrients (Kottke & Oberwinkler, 1987). The mantle of fungal tissue surrounding the host lateral roots varies in composition from the characteristic pseudoparenchymatous tissue to a rather open-wefted arrangement of hyphae. Development of a mature mantle proceeds through a programmed series of events (Horan et al., 1988; Martin et al., 1997). Fungal hyphae originating from a soil propagule or an older mycorrhiza penetrate into the root cap cells and grow through them. Backwards from the tip the invasion of root cap cells proceeds inwards until the hyphae reach the epidermal cells (Chilvers, 1968; Feugey et al., 1999). The root cap tissue is progressively transformed by invasion and dismemberment of the cells into the inner layers of the mantle. Mantle formation therefore commences selectively in the apical root region (Horan et al. 1988). After attachment onto epidermal cells, hyphae multiply to form a series of layers. These aggregating hyphae, which could develop into a biofilm of several hundred µm thick, then differentiate to form the mature mantle. The hyphae in these structures are encased in an extracellular polysaccharide and proteinacous matrix. Air and water channels that allow the flow of nutrients into the symbiosis innervate these structures (Ashford et al., 1989). There is a structural (Paris et al., 1993) and physiological (Cairney & Burke, 1996) heterogeneity within the ectomycorrhizal mantle, and between the mantle and the inward and outward fungal networks. These morphological analyses of the infection process have shown a fairly complex influence of the root on the fungus, including a general growth stimulus, a trophic response directing hyphal growth inwards towards the plant tissues and a morphogenetic effect leading to compact hyphal mantle development. In addition to putative morphogens, the supply of nutrients, the presence of a physical support and the supply of O2 likely play a role in mantle formation (Read & Armstrong, 1972; Martin et al., 1999). On the other hand, fungal hyphae stimulate lateral root formation, dichotomy of the apical meristem in conifer species, and cytodifferentiation (radial elongation, root hair decay) of root cells (Fig. 1d) (Horan et al., 1988).

As stressed above, morphogenesis of ectomycorrhiza encompasses a series of complex and overlapping ontogenic processes in symbionts: switching off the fungal growth mode, initiation of lateral roots, aggregation of hyphæ, arrest of cell division in ensheathed roots, radial elongation of epidermal cells (Kottke & Oberwinkler, 1987; Horan et al., 1988; Feugey et al., 1999). Morphological differentiation is accompanied by the onset of novel metabolic organizations in fungal and plant cells leading to the finished functioning symbiotic organ (Martin & Botton, 1993; Cairney et al., 1989; Rygiewicz & Andersen, 1994; Hampp et al., 1995). The entire purpose of the complex programme of multicellular development in the ectomycorrhizal symbiosis is to extend the function of the root system (Read, 1991; Smith & Read, 1997). Extramatrical hyphae, the mantle and the intraradicular hyphal network are active metabolic entities that provide essential nutrient resources (e.g. phosphate, nitrogen) to the host plant. The contribution of essential nutrients is reciprocated by the provision of a stable carbohydrate-rich niche in the roots for the fungal partner, making the relationship a mutualistic symbiosis. This simple perspective does not do justice to the challenge of defining this symbiosis, a complex phenotype that is dependent on the interactions between a spectrum of different host and microbial molecules. This mutuality of host and microbe, and its implicit coevolutionary implications, are key concepts. The distinction between saprobic and mycorrhizal behaviour is not a sharp cut-off (Hibbett et al., 2000), but a continuum and must take into account: ecological factors; variations in host surveillance mechanisms; and evolutionary factors that determine the cost (i.e. fitness) of the symbiosis (Fitter, 1991). The ecological performance of ectomycorrhizal fungi is therefore a complex phenotype affected by many different traits and by environmental factors. Identification of the primary factors controlling the development of the symbiosis and its metabolic activity (e.g. nutrient acquisition) will open the door to understanding the ecological fitness of the ectomycorrhizal symbiosis.

Mutual signal exchange fires the developmental process

By contrast to some other plant–microbe interactions (Dénariéet al., 1996; Moller & Chua, 1999), the nature of the signalling molecules and the molecular basis of signal perception and transduction in mycorrhiza are unknown or ill-defined. Identifying the processes that regulate the information flow between mycorrhizal fungi and host root is an active research area. Host plants release into the rhizosphere critical metabolites that are able to trigger basidiospore germination (Fries et al., 1987), growth of hyphae towards the root (Horan & Chilvers, 1990) and the early developmental steps of mycorrhiza formation (Béguiristain & Lapeyrie, 1997; Salzer et al., 1997a; Ditengou & Lapeyrie, 2000). Molecules that control the interactions between symbionts can be classified as follows:

  • • tropism of hyphae for host tissues (rhizospheric signals);
  • • attachment and invasion of host tissues by hyphae (adhesins, hydrolases);
  • • induction of organogenetic programmes in both fungal and root cells (hormones and secondary signals);
  • • facilitating survival of the mycobiont despite plant defense responses;
  • • coordinating strategies for exchanging carbon and other metabolites (e.g. vitamins) for in planta colonization and for balancing growth of the soil fungal web with its role in gathering minerals from the soil.

Rhizospheric signals

Based on current knowledge of the molecules released in other plant–microbe interactions, the early plant host signals secreted into the rhizosphere can include flavonoids, diterpenes, hormones and various nutrients. As shown in Fig. 2, host-released metabolites, such as the flavonol, rutin, and the cytokinin, zeatin, strikingly modified hyphal morphology (Lagrange et al., 2001). When present in the growth medium at very low concentrations, rutin stimulated growth of Pisolithus (Fig. 2a), whereas the cytokinin, zeatin modified the hyphae branch angle (Fig. 2b). These rhizospheric molecules are therefore able to induce morphological changes similar to those observed during actual ectomycorrhizal development (Kottke & Oberwinkler, 1987; Horan et al., 1988). Root exudates and zeatin also trigger an enhanced accumulation of fungal molecules, such as hypaphorine, the betaine of tryptophan (Béguiristain & Lapeyrie, 1997; H. Lagrange & F. Lapeyrie, unpublished results). This fungal alkaloid is the major indole compound isolated from the ectomycorrhizal fungus Pisolithus (Béguiristain et al., 1995) and is produced in larger amounts by this fungus during mycorrhiza development (Béguiristain & Lapeyrie, 1997).

Figure 2.

The molecular cross-talk taking place in the rhizosphere of Eucalyptus globulus colonized by Pisolithus. Root exudates alter the morphology of the mycelium of Pisolithus. These morphological changes are induced by the flavonol, rutin, which stimulates fungal growth, expressed as colony diameter (a). Low concentrations of zeatin modify the hyphae branching angle (b) and the accumulation of the tryptophan betaine, hypaphorine (c). On the other hand, Pisolithus-secreted hypaphorine and Indole-3-acetic (IAA) trigger morphological changes (i.e. arrest of root hair elongation and stimulation of short root formation) in the root system (d). (After Béguiristain & Lapeyrie, 1997, Ditengou et al., 2000; H. Lagrange & F. Lapeyrie, unpublished).

Hypaphorine secreted by Pisolithus induces morphological changes in root hairs of Eucalyptus seedlings. Incubation in the presence of hypaphorine leads to a decreased rate of elongation and a transitory swelling of the apex of the root hair (Ditengou et al., 2000). Addition of indole-3-acetic acid (IAA) restores tip growth (Ditengou et al., 2000) indicating antagonistic effects of hypaphorine and auxin on the observed morphological changes. Development of root hairs is divided into cell fate determination, hair initiation, and hair elongation (Gilroy & Jones, 2000). The latter step is highly polarized, involving a narrow zone (< 10 µm) at the tip, where new membrane and cell wall are built from fusion of secretory vesicles (Geitmann & Emons, 2000). Growth in root hairs is associated with an apex-high cytosolic free Ca2+ gradient generated by a local Ca2+ influx at the tip, which probably has a crucial role in the regulation of vesicle fusion (Geitmann & Emons, 2000). Whether the hypaphorine-induced cytoskeleton changes are related to interactions with calcium channels, cofilin/actin-depolymerizing proteins and/or auxin signalling pathways (Ditengou et al., 2000) is currently under investigation.

Root hairs are a key site for microbial interactions (Peterson & Farquhar, 1996) and it has been suggested that interaction between the ectomycorrhizal fungus and root hairs may play a role in the symbiosis development (Thomson et al., 1989). However, this contention has been challenged (Horan et al., 1988). Mycorrhizal colonization is initiated in the root cap region and then propagates by an acropetal extension of root and fungal tissues (Chilvers, 1968; Horan et al., 1988). Only new epidermal cells, formed after fungal invasion of the apex, are involved in this infection process. The intercellular network of hyphae (i.e. the Hartig net) develops only between epidermal cells that undergo a radial elongation as a result of fungal contact. In such circumstances, root hairs will not emerge on a colonized root surface, suggesting that their development has been inhibited very early, either mechanically or chemically (e.g. through the action of hypaphorine), by the fungus (Ditengou et al., 2000).

Changes in hormonal balance

In addition to flavonoids, phytohormones, including auxins, cytokinins, abscisic acid and ethylene, are produced by ectomycorrhizal fungi (Gogala, 1991). Many studies indicate that changes in auxin balance are a prerequisite for mycorrhiza organogenesis (Rupp et al., 1989; Gay et al., 1994; Karabaghli-Degron et al., 1998; Kaska et al., 1999). As can be seen in Fig. 2, ectomycorrhizal fungi enhance proliferation of short roots (Carnero Diaz et al., 1996). In conifers, externally supplied fungal exudates, extracts, or synthetic auxins partially mimic the effect of mycelium in inducing root proliferation and dichotomous branching of lateral roots (Smith & Read, 1997). The presence of plant-derived tryptophan in the rhizosphere could be sufficient for ectomycorrhizal fungi to enhance the biosynthesis of fungal IAA (Rupp et al., 1989). However, recent studies have shown that tryptophan synthase is dispensable for auxin production by the nonmycorrhizal ascomycete Aspergillus nidulans and that auxin is not produced from indole, an intermediate of the tryptophan synthase reaction (Eckert et al., 2000). Pine inoculated with mutant strains of Hebeloma cylindrosporum that overproduce IAA generated an increased number of ectomycorrhizal roots (Gay et al., 1994) which presented a strikingly altered morphology: a multiseriate Hartig net (Gea et al., 1994). This confirms that some morphogenetic steps controlling development of mycorrhiza are regulated by fungal IAA. Inhibitors of polar auxin transport, such as 2,3,5-triiodobenzoic acid (TIBA), restrict the stimulation of lateral root formation and the colonization of the tap-root cortex of conifer seedlings by ectomycorrhizal fungi (Karabaghli-Degron et al., 1998). Pisolithus-secreted hypaphorine acts as a natural auxin antagonist (Ditengou & Lapeyrie, 2000). Whether tryptophan and/or other components of the plant exudates induces an IAA amplification loop in the rhizospheric mycelium, synthesis of IAA must be tightly controlled or compensated by other factors, since above a certain concentration, exogenously supplied IAA inhibits root development.

This brief review has attempted to shed light on recent advances in how the ectomycorrhizal symbionts respond to rhizospheric signals, including hormones, and how these various signals interact in order to execute the appropriate developmental responses (Fig. 2). The multiplicity of identified signals (auxins, alkaloids, cytokinins, phenylglycoside) has confirmed the complex network of cues controlling development of the symbiosis. These signals can act in a synergistic (rutin/zeatin) or antagonistic (IAA/hypaphorine) manner. Signal perception may culminate in the induction of downstream target gene products whose expression probably underpins physiological and/or development responses (short root proliferation, alteration of hyphal branching). Although many pieces of the puzzle remain to be elucidated, it seems inescapable that the cross-talk between rhizospheric metabolites, the hormonal balance and signalling networks involving Ca2+ play an important role in coordinating the execution of the appropriate responses and it will be fascinating to learn more about the multiple facets of this communication network.

Walking in the promised land: how to escape host surveillance

It is well established that a variety of fungal products can elicit inducible plant defense responses in both host and nonhost plants, and that such responses can also be triggered by plant products released during cell-wall degradation (Boller, 1995). It seems a reasonable assumption that nonspecific oligosaccharidic and proteinaceous elicitors are released during root colonization by ectomycorrhial fungi (Salzer et al., 1996; 1997c). Signal transduction pathways, related to the perception of microbial invasion of host plant tissues, include changes in protein phosphorylation status, modifications in ion fluxes, increase in cytosolic Ca2+ concentration, depolarization of the plasma membrane, production of reactive oxygen species (ROS), and alterations in gene expression (Yang et al., 1997). Post-translational modification of proteins by phosphorylation is a general mechanism in the reception/transduction of signals originating from pathogens (Yang et al., 1997) and symbionts (Pingret et al., 1998). Thus, it appears to be of prime importance to identify the protein kinases, protein phosphatases and corresponding protein substrates that are involved in the early steps of signal transduction in ectomycorrhiza.

Experimental evidence suggests that cell wall components (e.g. glucan and chitin fragments) of mycobionts bind to specific sites on cells from host trees (e.g. Picea abies) (Salzer et al., 1997c). Treatments of cells of Norway spruce (Picea abies) with elicitors released from the ectomycorrhizal fungus Hebeloma crustuliniforme induce K+ and Cl effluxes into the medium followed by a Ca2+ influx into the cells within a few minutes (Schwacke & Hager, 1992; Salzer et al., 1996). Phosphorylation of a 63-kDa and dephosphorylation of a 65-kDa protein were identified using in vivo phospholabeling as little as 4 min after elicitor addition (Salzer et al., 1997a; Hebe et al., 1999). After a few minutes, extracellular alkalinization and a transient accumulation of ROS were observed (Schwacke & Hager, 1992; Salzer et al., 1996; Salzer et al., 1997b). These initial signalling processes can also be triggered by the G-protein activator, mastoparan, and the protein phosphatase inhibitor, cantharidin and okadaic acid (Salzer et al., 1997a; Hebe et al., 1999). Similar elicitor-induced reactions are regarded as initial events of a hypersensitive response (HR) in plant–pathogen interactions (Yang et al., 1997; Heath, 2000), but the role of such rapid reactions in ectomycorrhizal interactions remains to be elucidated.

Specific and nonspecific elicitors seem to trigger signalling cascades involving protein kinases, protein phosphatases, and elements of the mitogen-activated protein (MAP) kinase pathway (Heath, 2000). Some nonspecific, broad-spectrum defenses are clearly mounted in hosts when the ectomycorrhizal fungus penetrates directly into the root and digests its way through the apoplastic space. For example, elicitors from the ectomycorrhizal fungi Amanita muscaria and H. crustuliniforme induce the enzymes chitinase and peroxidase in spruce (Picea abies) cells (Sauter & Hager, 1989; Salzer & Hager, 1993), whereas cell-free extracts of Pisolithus elicit induction of peroxidases and chitinases in E. globulus seedlings (Albrecht et al., 1994). Feugey et al. (1999) suggested that these induced defense responses may limit Hartig net formation. However, massive root colonization by the mycobiont does not induce the HR, a rapid death of cells at the infection site that is associated with pathogen limitation as well as with defense gene activation (Maleck & Dietrich, 1999). This suggests that ectomycorrhizal fungi suppress defense responses through yet unknown mechanisms. Rapid changes in protein phosphorylation are inactivated by extracellular spruce chitinases and β-1,3-glucanases (Salzer et al., 1997b) and cleavage of chitinous elicitors from H. crustuliniforme by these host chitinases prevents induction of K+ and Cl release, extracellular alkalinization and synthesis of H2O2 spruce cells (Salzer et al., 1997a). Elicitor-induced changes of wall-bound and secreted peroxidase activities in suspension-cultured cells of spruce are attenuated by auxins (Mensen et al., 1998), suggesting interactions between elicitor and auxin signalling cascades. However, it remains to be demonstrated that these signalling networks occur in intact ectomycorrhiza.

Genes of transduction networks in symbionts

Formation of the symbiotic organ requires temporally and spatially controlled activity of genes and proteins participating in the morphogenetic process. The proliferation of roots and fungal tissues, and the need to adapt to a rapidly changing environment (changes in pH, enhanced fluxes of nutrients, presence of ROS) require a variety of gene products. Sensing this novel environment and coordination of fungal and plant developments might involved signalling networks. Fungal genes, such as PF6.2 and ras from Laccaria bicolor (Kim et al., 1999; G. Podila, pers. comm.) and ras from Pisolithus (Fig. 3a), are induced before any physical contact (S. Duplessis & H. Lagrange, unpublished), confirming that diffusible elicitors are involved in the early steps of the ectomycorrhizal interaction.

Figure 3.

Phylogenetic analysis of Ras GTPases (a) and the α subunit of heterotrimeric G proteins (b) from yeasts and filamentous fungi. The unrooted trees were constructed with PAUP 4.0b4a programme using sequence alignments obtained with the MultAlin programme (Corpet, 1988). (b) Um-Gpa4 was used as an outgroup (Regenfelder et al., 1997). Ras and Gpα sequences of Pisolithus are in black boxes. The α subunits of heterotrimeric G proteins known as being implicated in virulence are in open boxes. Bootstrap values are for 1000 replications. Pt, Pisolithus tinctorius; Cc, Coprinus congregatus; Mg, Magnaporthe grisea; Cp, Cryphonectria parasitica; Nc, Neurospora crassa; An, Aspergillus nidulans; Um, Ustilago maydis; Pc, Pneumocystis carinii f. sp. carinii; Sp, Schizosaccharomyces pombe; Cn, Cryptococcus neoformans; Sc, Saccharomyces cerevisiae; Kl, Kluyveromyces lactis; Ca, Candida albicans; Sb, suillus bovinus; Cci, Coprinus cinereus; Le, Lentinula edodes; Fn, Filobasidiella neoformans var. neoformans; Af, Aspergillus fumigatus; Lb, Laccaria bicolor; Bf, Botryotinia fuckeliana; Ct, Colletotrichum trifolii; Mr, Mucor racemosus. (a) DDBJ/EMBL/GenBank accession numbers of Ras G proteins: Sb-Ras1p, AF250024; Cci-Ras, D13295; Le-Ras, D00742; Fn-Ras, AF294647; Af-Ras, L42299; An-Ras, U03025; Lb-Ras, AF034098; Bf-Ras1, U79558; Ct-Ras, AF044895; Mr-Ras1, M55175; Mr-Ras3, M55177. (b) DDBJ/EMBL/GenBank accession numbers of G protein α subunits: Cc-Gpa1, X68031; Mg-MagB, AF011341; Cp-Cpg-1, L32176; Nc-Gna1, L11453; An-FadA, U49917; Um-Gpa1, U85775; Mg-MagC, AF011342; Nc-Gna2, L11452; Pc-Pcg1, U30791; Um-Gpa2, U85776; Sp-Gpa1, M64286; Mg-MagA, AF011340; Cp-Cpg-2, L32177; Cn-Gpa1, U09372; Um-Gpa3, U85777; Sc-Gpa2, U18778; Kl-Gpa2, L47105; Sp-Gpa2, D13366; Ca-Cag1, M88113; Sc-Gpa1, M15867; Um-Gpa4, U85778.

Communication genes, such as heterotrimeric GTPases, ras, the Ca2+/calmodulin-dependent phosphoprotein phosphatase (calcineurin), and serine/threonine kinases are expressed in 4-day-old Pisolithus/Eucalyptus ectomycorrhiza (Voiblet et al., 2001); they represent about 13% of the cloned genes in this symbiosis (Table 1). The sequence of the Pisolithus gene that codes for the α subunit of the heterotrimeric GTPase, Pt-Gpα, is very similar to that of the gene encoding the fungal Gpα involved in pathogenesis (Fig. 3b; Alspaugh et al., 1998; Kahmann et al., 1999), suggesting that it probably plays a similar role in early signalling in the symbiotic interaction. The concentration of transcripts of Gpα was increased in 4 d-old E. globulus/Pisolithus ectomycorrhiza (Voiblet et al., 2001; S. Duplessis & F. Martin, unpublished). This up-regulation may partly result from changes in the concentrations of C and N as suggested by the the enhanced expression of this gene in mycelia grown on C-rich or N-depleted medium (S. Duplessis & F. Martin, unpublished). In addition to these changes in transcript concentrations, ectomycorrhizal development induces alteration in the phosphorylation status of protein (Salzer et al., 1997c). The signal transduction pathways involved in linking symbiont contact with morphogenetic changes to the cellular machinery remain to be elucidated. It appears that both transcriptional and post-translational regulations are involved in the differentiation processes that are essential for symbiosis development.

Table 1. Differential expression of genes of signalling pathways in ectomycorrhiza vs free-living partners. The columns represent: (1) the expressed sequence tag (EST) clone ID (2) the genomic origin (fungus or plant) (3) the ratio for the normalized hybridization values of transcripts expressed in the symbiotic tissues and in the free-living partners, and (4) the best database match (and corresponding species) (after Voiblet et al., 2001)
Clone IDOrganismRatioBest database match (species)
ud283Fungus4.0PWP2/Transducin (Saccharomyces cerevisiae)
9B9Plant3.4Calmodulin (Daucus carota)
8A7Plant2.7ADP-ribosylation factor (Oryza sativa)
8D10Fungus2.5SHP1 protein phosphatase (Schizosaccharomyces pombe)
11A6Fungus2.1GTP-binding protein, α subunit (Coprinus congregatus)
St92Fungus1.9Putative histidine kinase (Arabidopsis thaliana)
1D4Fungus1.8GTP-binding protein YPT1 (Neurospora crassa)
St114Fungus1.8Serine/Threonine protein kinase (S. cerevisiae)
5E9Fungus1.7Rab11D/ras-related protein (Lotus japonicus)
8D4Fungus1.5Lectin receptor-like protein (A. thaliana)
6C8Fungus1.5PtCPC2, Gβ/RACK-like protein (Pisolithus tinctorius)
EgPtdB24Fungus1.3GTP-binding protein GTB1 (Mus musculus)
7B5Fungus1.2Ras1p (Suillus bovinus)
St18Fungus1.2SNF1 (carbon catabolite derepressing) (S. cerevisiae)
8B8Plant1.0ADP-ribosylation factor (Vigna unguiculata)
8A1Fungus0.9Serine/Threonine protein kinase (A. thaliana)
11C5Fungus0.9Serine/Threonine protein kinase (Caenorhabditis elegans)
11D1Fungus0.9Calcineurine Β subunit (N. crassa)

Analysing transcription during early symbiosis development

The interaction between ectomycorrhizal fungi and roots is accompanied by a series of molecular changes in both the mycobiont and the host plant. These events result in altered expression of a number of proteins (Hilbert et al., 1991) and genes (Voiblet et al., 2001). There is now an ever-growing number of technologies available for conducting large-scale expression studies, including sequencing-based methods, such as the serial analysis of gene expression, and solid-support approaches, such as oligonucleotide and cDNA arrays (Strauss & Falkow, 1997; Bouchez & Höfte, 1998). Over the past 3 yr, expression profiles have been reported for processes ranging from fundamental cellular events such as cell cycle (Spellman et al., 1998) and oncogenesis (Alizadeh et al., 2000) to physiological challenges such a change in a nutrient source (Wang et al., 2000) or insect-induced wounding of leaves (Reymond et al., 2000). Transcriptional changes are observed in as many as 10–20% of all genes during a given process and only a fraction of these differences have been characterized previously. The simultaneous measurement of several hundreds of transcripts allows more precise differentiation between related patterns of change, making it likely that new topographical features of genetic regulation will be identified. Elucidation of the nature of genes differentially expressed during the development of ectomycorrhiza could help in understanding the molecular basis of the early events in plant–ectomycorrhizal fungus interaction.

By using differential screening of about 500 arrayed cDNAs, we found 65 SR genes differentially expressed during ectomycorrhizal mantle formation (Voiblet et al., 2001). The number of SR genes displaying similarity to genes involved in cell wall and membrane synthesis, stress defence response, protein degradation (in plant cells) and protein synthesis (in hyphae) suggests a highly dynamic environment in which symbionts are sending and receiving signals, are exposed to high levels of stress and are remodeling tissues. A striking result of these studies is the fact that all genes investigated are common to the nonsymbiotic and symbiotic stages. At the developmental stage studied, symbiosis development does not induce the expression of ectomycorrhiza-specific genes but, rather, a marked change in gene expression in the partners, suggesting that genes for vegetative development may have been recruited to function during mycorrhiza formation. A similar pattern has been shown for nodulation in alfalfa (Györgyey et al., 2000; Jiménez-Zurdo et al., 2000).

Conclusions

The nature of the signals released by the ectomycorrhizal symbionts, how these signals are transduced within the partners, and how these processes trigger the expression of symbiosis-regulated genes that assist in partner recognition and the formation of symbiotic tissues are only beginning to be understood. We can expect that new components of the transduction pathways will soon be identified, facilitating an understanding of the cross-talking between signalling networks. About 100 SR-genes have been identified in various ectomycorrhiza associations, the products of which may play a role in recognition and attachment of the mycobiont onto root surfaces, formation of the symbiotic interface, signalling networks, protein turnover, organogenesis, and novel symbiotic metabolism. However, many questions concerning the differentiation of plant and fungal symbiotic structures remain unanswered. Master regulatory genes (e.g. homeogenes) that may control morphogenesis in the symbiosis have not yet been isolated and the biochemical activities associated with several SR genes are still unknown. Elucidation of the underlying mechanisms may lead to ways to intervene in this process. There are two routes for further work. Firstly, ectomycorrhiza-regulated genes need to be studied in order to determine their function in the development of the symbiosis by using reverse genetics. Secondly, the transcriptional regulation of SR-genes needs to be analysed. What transcription factors regulate SR-gene expression, and how is elicitor/signal-dependent activation of these transcription factors achieved? The answers to these questions will provide further highlights into the signalling networks and early gene regulation processes involved in ectomycorrhiza development. Eventually, such studies will lead to a better understanding of plant–microbe interactions and evolution of plant–fungus associations.

Acknowledgements

This paper was presented at the 52nd Harden-New Phytologist Conference ‘Signalling in plants’ held October 18–22, 2000, at the Wye College in Wye, Kent, UK. F.M. would like to thank J. Gallon and C. Smith for their invitation to contribute to this meeting and the New Phytologist Trust for its financial support. S.D., H.L. and C.V. were supported by Doctoral Scholarships from the Ministère de l’Education Nationale, de la Recherche et de la Technologie. FAD was supported by a fellowship from the Government of Gabon. We also appreciated partial support from the Groupement de Recherches et d’Etude des Génomes, the INRA Collaborative Research Programmes in Microbiology, the European Commission INCO-DC Programme (contract number: ERBIC18CT-98319) and COST-EUROSILVA-WG3–2000.

Ancillary