Increased trehalose biosynthesis in Hartig net hyphae of ectomycorrhizas

Authors


Author for correspondence: Uwe Nehls Tel: +49 7071 297 7657 Fax: +49 7071 295 635 Email: uwe.nehls@uni-tuebingen.de

Summary

  • • To obtain photoassimilates in ectomycorrhizal symbiosis, the fungus has to create a strong sink, for example, by conversion of plant-derived hexoses into fungus-specific compounds. Trehalose is present in large quantities in Amanita muscaria and may thus constitute an important carbon sink.
  • • In Amanita muscaria–poplar (Populus tremula × tremuloides) ectomycorrhizas, the transcript abundances of genes encoding key enzymes of fungal trehalose biosynthesis, namely trehalose-6-phosphate synthase (TPS), trehalose-6-phosphate phosphatase (TPP) and trehalose phosphorylase (TP), were increased.
  • • When mycorrhizas were separated into mantle and Hartig net, TPS, TPP and TP expression was specifically enhanced in Hartig net hyphae. Compared with the extraradical mycelium, TPS and TPP expression was only slightly increased in the fungal sheath, while the increase in the expression of TP was more pronounced. TPS enzyme activity was also elevated in Hartig net hyphae, displaying a direct correlation between transcript abundance and turnover rate. In accordance with enhanced gene expression and TPS activity, trehalose content was 2.7 times higher in the Hartig net.
  • • The enhanced trehalose biosynthesis at the plant–fungus interface indicates that trehalose is a relevant carbohydrate sink in symbiosis. As sugar and nitrogen supply affected gene expression only slightly, the strongly increased expression of the investigated genes in mycorrhizas is presumably developmentally regulated.

Introduction

Ectomycorrhizal symbiosis is a mutualistic interaction between certain soil fungi and trees of boreal and temperate forests. It helps to overcome nutritional and carbohydrate limitations faced by the respective partners of symbiosis. The basis of this interaction is the supply of photoassimilates by the host plant.

When fungal hyphae recognize an emerging fine root of a compatible plant partner, they direct their growth towards it (Martin et al., 2001) and colonize the root surface, (often) forming a sheath or mantle of hyphae, which encloses the root and isolates it from the surrounding soil (Blasius et al., 1986). After or at the same time as sheath formation, fungal hyphae grow inside the infected fine root, forming highly branched structures in the apoplast of the rhizodermis (angiosperms) and in the root cortex (gymnosperms). This so-called ‘Hartig net’ generates a large surface area between the two partners (Kottke & Oberwinkler, 1987).

The hyphal networks of ectomycorrhizas (fungal sheath and Hartig net) have different functions (Harley & McCready, 1952; Harley & Smith, 1983; Kottke & Oberwinkler, 1987; Smith & Read, 1997). The Hartig net, which serves as an interface between plant and fungus, is adapted to the exchange of plant-derived carbohydrates for fungus-derived nutrients. The function of the fungal sheath, in contrast, is that of intermediate storage of the nutrients that are delivered by soil-growing hyphae and are intended for delivery to the Hartig net, and of the carbohydrates that are taken up by the hyphae of the Hartig net and are intended for transport towards the soil-growing mycelium.

In ectomycorrhizal symbiosis, up to 30% of total plant photoassimilates can be transferred to the fungus to enable its proliferation (Finlay & Söderström, 1992; Söderström, 1992). Sucrose, the major long-distance transport carbohydrate of most plants, is presumably exported into the apoplast and hydrolyzed by a plant-derived acid invertase (Lewis & Harley, 1965b; Salzer & Hager, 1991; Hampp & Schaeffer, 1999). The resulting monosaccharides are then taken up by the fungal partner (Lewis & Harley, 1965a; Palmer & Hacskaylo, 1970; Chen & Hampp, 1993).

The driving force for carbon allocation in vascular plants is consumption at the sink site. Mycorrhizas attract carbohydrates much more efficiently than nonmycorrhizal fine roots (Bevege et al., 1975; Cairney et al., 1989), indicating a strong sink created by fungal hyphae in symbiosis. As a consequence, fungal hexose uptake capacity is greatly increased in Hartig net hyphae (Nehls et al., 1998, 2001a; Wiese et al., 2000; Nehls, 2004). The findings of nuclear magnetic resonance (NMR) (Martin et al., 1988, 1994) and biochemical investigations (Hampp et al., 1995; Schaeffer et al., 1996; Kowallik et al., 1998) indicate that imported hexoses are utilized for ATP generation, amino acid biosynthesis (carbon skeletons), and the formation of carbohydrate storage compounds. When not limited by carbon supply, ectomycorrhizal fungi produce a series of fungus-specific sugars and sugar alcohols (Martin et al., 1985, 1987, 1988, 1998). Different pools of storage carbohydrates can be distinguished: oligosaccharides (trehalose), polyols (mannitol, arabitol and erythritol), and the long-chain carbohydrate glycogen.

In addition to its function as a carbon reserve, trehalose can also act as a stabilizer and protectant of proteins and membranes (Gadd et al., 1987) against heat (Bell et al., 1992), cold (Tibbett et al., 2002), and oxidative stress (Banaroudj et al., 2001).

In eukaryotes, trehalose can be generated by two different enzymatic routes. The combined activities of trehalose-6-phosphate synthase (TPS) and trehalose-6-phosphate phosphatase (TPP) catalyze irreversible trehalose formation from uridine diphosphate (UDP)-glucose and glucose-6-phosphate (generating trehalose-6-phosphate as an intermediate). Trehalose phosphorylase (TP) (also known as trehalose synthase) catalyzes the reversible phosphorolysis of trehalose into glucose-1-phosphate and glucose. At cellular pH values (between 6 and 7), however, the thermodynamic equilibrium constant favors the biosynthesis of trehalose and not its degradation (Eis et al., 1998).

One model system for the investigation of carbon partitioning in ectomycorrhizas is the symbiosis established between fly agaric (Amanita muscaria) and poplar (Populus tremula ×tremuloides). Like other ectomycorrhizal fungi (Martin et al., 1987, 1998; Ineichen & Wiemken, 1992), fly agaric uses trehalose as an intermediate storage pool for carbohydrates in symbiosis (Schaeffer et al., 1995; Wallenda, 1996; Hampp & Schaeffer, 1999), but compartmentation of trehalose biosynthesis between hyphae of the Hartig net and the hyphal sheet has not been investigated. We thus separated physically hyphal mantle and Hartig net hyphae and studied the expression of genes coding for the key enzymes of trehalose biosynthesis: TPS, TPP and TP. In addition, we assayed the enzymatic activity of TPS, and determined the respective pool sizes of trehalose. Our data show that trehalose biosynthesis is primarily associated with Hartig net hyphae, and that it is not under the control of carbon and nitrogen supply.

Materials and Methods

Biological material

Amanita muscaria (L. Fr.) Pers. strain CS83 was isolated from a fruiting body from Schönbuch, Germany (Schaeffer et al., 1995). Mycelia were grown in liquid culture or on Petri dishes for 2–16 d in modified Melin Norkrans (MMN; Marx, 1969) medium in the presence of glucose (up to 40 mm) as the carbon source and various nitrogen sources (amino acids were supplied as casein hydrolysate; N-Z Amine HD; Sigma, St Louis, MI, USA). Fungal mycelium grown in liquid culture was collected by filtration using a Büchner funnel under suction, washed twice with deionized water, frozen in liquid nitrogen, and stored at −80°C.

Populus tremula × tremuloides was used as the plant partner for mycorrhiza formation under axenic conditions according to Hampp et al. (1996), with MMN medium containing ammonium at a final concentration of 300 µm as the sole nitrogen source. Mycorrhizal and nonmycorrhizal fine roots and nonmycorrhizal fungal hyphae (extraradical mycelium) were harvested, frozen in liquid nitrogen, and stored at −80°C.

Expression analysis

Expression analysis was performed by quantitative reverse transcriptase–polymerase chain reaction (RT-PCR). Isolation of total RNA from samples of 80 mg fresh weight was carried out either according to Nehls et al. (1998) or using the RNeasy Plant Mini Kit (Qiagen GmbH, Hilden, Germany) according to the manufacturer's instructions. Aliquots of c. 0.4 µg of total RNA were treated with DNAse I (Invitrogen, Carlsbad, CA, USA) and used for first-strand cDNA synthesis in a total volume of 20 µl, containing 50 pmol oligo-d(T)18 primer (GE Healthcare Europe, Freiburg, Germany) and 200 U Superscript II RNAse H Reverse Transcriptase (Invitrogen) according to the manufacturer's instructions. After synthesis, 30 µl of 5 mm Tris/HCl, pH 8, was added and aliquots were stored at −80°C.

PCR was performed in a total volume of 20 µl using 10 µl of Q-PCR-Mastermix (ABgene, Epsom, UK), 0.5 µl of cDNA, and 10 pmol of each primer in a MyiQ real-time PCR system (Bio-Rad, Hercules, CA, USA). Specific primers for the constitutively expressed A. muscaria genes SCIV038 and SCIV192 (Nehls et al., 2001a) were used as reference (only the results for SCIV038 are shown, but similar results were obtained using SCIV192 as a reference). PCR was always performed in duplicate. At least three independent cDNA preparations were used for analysis. For the determination of PCR efficiency, dilution series of each gene were prepared and used as the PCR template. The corresponding PCR efficiencies, calculated using the MyiQ software package (version 1.0; Bio-Rad), were 91.5% for A. muscaria trehalose-6-phosphate synthase (AmTPS), 92.3% for AmTPP, 91% for AmTP, 93.8% for SCIVO38 and 89.3% for SCIVO192 primers, respectively.

Primers used for analysis

  • SCIV038: 5′-CTCATCTGCTCTCGTGC-3′ and 5′-GAGATAGCACCTCCACG-3′

  • SCIV192: 5′-AAGCAGACCGTTACCTATC-3′ and 5′-TCCTGGCTCTCTTACTCC-3′

  • AmTPS: 5′-GCTGTCACAATGGATGC-3′ and 5′-AGTGCACGTTATGTAAGG-3′

  • AmTPP: 5′-TGCATACCCAGGATTTG-3′ and 5′-GACACTCTACATATTCGTGC-3′

  • AmTP: 5′-AATGCTGTTGCTTGGCTG-3′ and 5′-CGGTCTATAAAGTACAACAC-3′

Measurement of TPS activity and trehalose content by fluorescence microscopy

Freeze-dried entire and dissected mycorrhizas (separated into fungal sheath and fine roots containing the Hartig net) as well as nondissected fine roots were cut into two or three pieces of about equal length with custom-made micro knives and weighed on glass-fiber balances (for production and calibration, see Lowry & Passonneau, 1972) in a conditioned room (40% humidity; 20°C).

The principles of measurement are described in Outlaw et al. (1985). The reaction cuvettes consisted of a 5-mm-thick Teflon tray with holes of 5 mm diameter. The holes were closed by a thin Teflon film (Hansa Tech, Kings Lynn, UK) at the lower end of the Teflon tray and fixed by insertion of Teflon tubing with an outer diameter identical to the diameter of the hole. The inner diameter of the tubing was 2.5 mm. To prevent evaporation of the assay solution, the wells were filled with 5 µl of purified light mineral oil (Sigma). Using glass constriction pipettes, 2 µl of assay solution was submerged in the oil. Subsequently, the tissue sample was pushed through the oil into the assay droplet by means of a tiny quartz fiber glued to a glass or wooden handle. The whole Teflon tray was then transferred to the stage of an inverted microscope (Diavert; Leitz, Bensheim, Germany). The objective lens (PL Fluotar, ×40/0.70 EF; Leitz) was focused to a layer above the Teflon membrane within the brightest area of the droplet, avoiding shadowing by the sample. The excitation light (Hg lamp, HBO 103 W/2; Leitz) passed through an excitation filter (330–380 nm) into the assay droplet. Emitted fluorescent light passed through a dichroic mirror (< 400 nm) before reaching a photomultiplier. The analog signal from the photomultiplier was digitized by an AD converter (Serial Box Interface; Vernier, Beaverton, OR, USA) and recorded using the computer program logger pro (version 2.1; Vernier).

TPS activity was assayed according to Vanderkammen et al. (1989; method 2). A total volume of 2 µl contained HEPES (50 mm, pH 7.6), 40 mm glucose-6-phosphate, 2 mm MgCl2, 0.6 mm NADH, 1.5 mm phosphoenolpyruvate, 2 U lactate dehydrogenase, 2 U pyruvate kinase, and 1.7 mm UDP-glucose. Trehalose-6-phosphate and UDP were generated by TPS in a first reaction from UDP-glucose and glucose-6-phosphate. UDP was then phosphorylated via phosphoenolpyruvate/pyruvate kinase and the resulting pyruvate reduced to lactate under stoichiometric oxidation of NADH. The background of the reaction was measured in samples without glucose-6-phosphate.

Trehalose was quantified as glucose after hydrolysis by neutral trehalase (Jones & Outlaw, 1981). The assay consisted of 50 mm 2-morpholino-ethane-sulfonic acid (MES) (pH 6.5), 6.2 mm MgCl, 2.1 mm ATP, 0.8 mm NADP, 0.8 U glucose-6-phosphate dehydrogenase (from yeast), and 1.2 U hexokinase (from yeast). In a first step, glucose present in the tissue was converted into 6-phosphogluconic acid. When this reaction was finished, 1 × 10−4 U of neutral trehalase (from porcine kidney; Sigma) was added, and trehalose-derived glucose determined. Fluorescence resulting from NADPH was quantified from standard curves. If not otherwise stated, chemicals and enzymes were from Roche (Mannheim, Germany).

Ergosterol, a marker for membranes of ectomycorrhiza-forming fungi, was determined by high-performance liquid chromatography (HPLC) according to Martin et al. (1990) and Wallenda et al. (1996).

Phylogenetic analyses of AmTPS and AmTP

The predicted full-length protein sequences of AmTPS and AmTP were used for phylogenetic analysis. For alignments, full-length sequences from other fungi detected by BlastX in the nonredundant database of the National Center for Biotechnology Information (NCBI) (http://www.ncbi.nlm.nih.gov/BLAST/) were used. To obtain more basidiomycete sequences, publicly available fungal genomic DNA sequences from Laccaria bicolor (http://genome.jgi-psf.org/Lacbi1/Lacbi1.home.html) and Phanerochaete chrysogenum (http://genome.jgi-psf.org/whiterot1/whiterot1.home.html) were screened for TPS and TS genes using Blast X and AmTPS or AmTP as a template.

Miscellaneous

DNA fragments were cloned into the pCR 2.1-TOPO vector (Invitrogen) and used for transformation of One-shot competent Escherichia coli (Invitrogen). Overlapping sequencing was performed using M13 universal and reverse primers (Stratagene, La Jolla, CA, USA) as well as gene-specific primers (Operon Biotechnologies, Huntsville, AL, USA) and the ABI PRISM™ Dye Terminator Cycle Sequencing Ready Reaction Kit (Applied Biosystems, Foster City, CA, USA) on an automated ABI 3100 sequencer (Applied Biosystems) according to the manufacturer's instructions.

For analysis of DNA and protein sequences, the program packages GeneJockey II (Biosoft, Cambridge, UK) and BioEdit (version 7.0; Hall, 1999) were used. Sequence data were compared with public libraries using BlastX (http://www.ncbi.nlm.nih.gov/blast; Altschul et al., 1997) and further analyzed using Clustal X (version 1.8; Thompson et al., 1997).

Results

cDNA and deduced protein sequences of AmTPS, AmTPP and AmTP

cDNA clones revealing strong homology to TPSs, TPPs, and TPs were identified by random sequencing of a P. tremula ×tremuloidesA. muscaria ectomycorrhizal cDNA library (Küster et al., 2006).

The AmTPS cDNA (accession no. AJ300447) has a length of 2311 bp and contains an open reading frame that could code for a protein of 740 amino acids with a molecular mass of 80 167 Da. The deduced protein sequence of AmTPS had highest homology to TPSs from Laccaria bicolor, Phanerochaete chrysosporum and Ustilago maydis (60, 56, and 49% identity, respectively). TPS proteins from basidiomycetes, filamentous ascomycetes and yeast-related organisms cluster together in separate branches (Fig. 1). TPS might thus be a useful tool for phylogenetic grouping of fungi.

Figure 1.

Dendrogram of the phylogenetic relationship of Amanita muscaria trehalose- 6-phosphate synthase (AmTPS) or trehalose phosphorylase (AmTP) and related fungal proteins. The phylogenetic relationship of the deduced protein sequences of AmTPS (a) and AmTP (b) and related fungal proteins was determined by multiple sequence alignment using Clustal X (version 1.6; using default parameters) and visualized with TreeView (version 1.6.6; http://taxonomy.zoology.gla.ac.uk/rod/rod.html). Proteins are indicated by the name of the organism and either their accession number or their unique identifier in public Joint Genome Institute (JGI) databases.

Southern blot analysis revealed no indication of additional TPS genes (data not shown), which makes it probable that A. muscaria, like P. chrysogenum, L. bicolor and U. maydis, contains only one TPS gene in its genome.

A cDNA fragment coding for an AmTPP (Accession no. AJ642355) was identified in the same expressed sequence tag (EST) project (Küster et al., 2006). As the identified clone does not contain the entire reading frame, no phylogenetic analysis was carried out.

Furthermore, a cDNA coding for an AmTP (Accession no. AJ643200) from A. muscaria was identified. AmTP has a length of 2310 bp and contains an open reading frame that could code for a protein of 696 amino acids with a molecular mass of 77 278 Da. The deduced protein sequence of AmTP had highest homology to TPs from L. bicolor and Pleurotus sajor-caju (76.1 and 74.6% identity, respectively), but protein identity was always above 72% within the homobasidiomycete clade. While TP proteins of homobasidiomycetes and most filamentous ascomycetes group together in separate clusters, TPs of Cryptococcus neoformans (heterobasidiomycete) and Phaeosphaeria nodorum (ascomycete) are clearly different and as distant from other fungal TPs as the protein of an archaebacterium (Pyrobaculum aerophilum). As for TPS, the genome of L. bicolor also contains only one TP gene.

AmTPS, AmTPP and AmTP expression is strongly increased in hyphae of the Hartig net

Symbiotic interaction with Populus tremula × tremuloides (Figs 2, 3) showed a 5- to 6-fold increase in AmTPS1 transcript abundances compared with hyphae of the extraradical mycelium isolated from the same agar plate as mycorrhizas (Fig. 2). Although it was less pronounced, there was also an increase in AmTPP expression in mycorrhizas in comparison to free-living hyphae (c. 2.5-fold; Figs 2, 4). The largest differences in gene expression were obtained for AmTP. For this gene, transcript abundances were 25-fold higher in mycorrhizas compared with free-living (nonmycorrhizal) hyphae.

Figure 2.

Differential expression of Amanita muscaria genes encoding trehalose-6-phosphate synthase (AmTPS), trehalose-6-phosphate phosphatase (AmTPP) and trehalose phosphorylase (AmTP) in the fungal sheath and the Hartig net of A. muscaria ectomycorrhizas. Total RNA was isolated from the extraradical mycelium (ERM) grown on the same agar plate as ectomycorrhizas (Myc), and ectomycorrhizas were dissected into fungal sheath (Sheath) and the remaining fine root containing Hartig net hyphae (HN). Expression analysis was performed by quantitative reverse transcriptase–polymerase chain reaction (RT-PCR) using gene-specific primers for AmTPS, AmTPP and AmTP and calibrated to the constitutively expressed fungal reference gene SCIV038 (Nehls et al., 2001a) according to Selle et al. (2005).

Figure 3.

Impact of glucose nutrition on expression of Amanita muscaria genes encoding trehalose-6-phosphate synthase (AmTPS), trehalose-6-phosphate phosphatase (AmTPP) and trehalose phosphorylase (AmTP) in mycelia of A. muscaria grown in axenic culture. Glucose was added to a final concentration of 40 mm to mycelia precultivated in the absence of glucose for 1 wk. Mycelial samples were taken after different incubation times and actual glucose concentrations in the medium were determined. Mycorrhizas (Myc) and extraradical mycelium (ERM) were collected from the same agar plates (containing no glucose) after 6 wk of coculture. Expression analysis was performed by quantitative reverse transcriptase–polymerase chain reaction (RT-PCR) using gene-specific primers for AmTPS, AmTPP and AmTP, and calibrated to the constitutively expressed fungal gene SCIV038.

Figure 4.

Impact of nitrogen (N) and glucose (G) nutrition on expression of Amanita muscaria genes encoding trehalose-6-phosphate synthase (AmTPS), trehalose-6-phosphate phosphatase (AmTPP) and trehalose phosphorylase (AmTP) in A. muscaria mycelia grown in axenic culture. Fungal mycelia were nitrogen-starved for 1 wk before addition of different nitrogen sources (samples were collected after 2 d of incubation). Expression analysis was performed by quantitative reverse transcriptase–polymerase chain reaction (RT-PCR) using gene-specific primers for AmTPS, AmTPP and AmTP, and calibrated to the constitutively expressed fungal gene SCIV038. Aa, amino acid mixture.

When ectomycorrhizas were separated into fungal sheath and remaining fine roots, containing the Hartig net (see Nehls et al., 2001a), increased expression of all three A. muscaria genes was specifically observed in the hyphae of the Hartig net (Fig. 2): the increase in the amounts of transcript within Hartig net hyphae was c. 18-fold for AmTPS, 3.5-fold for AmTPP and 4-fold for AmTP. Gene expression in the hyphae of the nonmycorrhizal extraradical mycelium from the same plates was about 2-fold lower for AmTPS and AmTPP and 10-fold lower for AmTP compared with the hyphae of the fungal sheath (Fig. 2).

Impact of sugar and nitrogen supply on AmTPS, AmTPP and AmTP expression

In light of the fact that A. muscaria genes involved in sugar/nitrogen uptake and metabolism have been shown to be regulated by carbohydrate and nitrogen supply (Nehls et al., 2001a; Nehls, 2004), the impact of external hexoses and different nitrogen sources on expression of AmTPS, AmTPP and AmTP was investigated in hyphae grown in liquid culture.

The addition of glucose to carbon-starved mycelium resulted in a c. 5-fold increase in expression of AmTPS and AmTP after c. 2 h (Fig. 3). However, even after 1 d, i.e. before the glucose content in the growth medium was significantly reduced (Fig. 3, top), the transcript abundances had declined c. 2-fold. Sugar supply had only a minor impact on the transcript abundances of AmTPP (Fig. 3).

Nitrogen depletion resulted in a c. 2-fold decrease in AmTPS expression and a 2- to 3-fold increase in AmTPP transcript abundance compared with optimal growth conditions (Fig. 4), and these findings were independent of the presence of a carbon source. In the presence of glucose, the addition of ammonium but not of amino acids resulted in a c. 2-fold lower transcript abundance of AmTP.

AmTPS activity and trehalose content in hyphae of the plant–fungus interface

To address the question of whether enhanced AmTPS expression in hyphae of the Hartig net also results in a discernible increase in TPS activity, the hyphae of the fungal sheath and the Hartig net were analyzed separately. Because only small amounts of dissected mycorrhizal material could be obtained, a sensitive indicator reaction (NADH fluorescence) coupled to a microscope-based detection system was used (Outlaw et al., 1985). Mycorrhizas were separated into the fungal sheath and the remaining fine roots (which, according to the method of Nehls et al. (2001a), contain the Hartig net), and freeze-dried. The Hartig net (the mycorrhiza without the hyphal mantle) contains, in contrast to the fungal mantle, only a small portion of fungal cells, so the ergosterol content was used for calibration of enzyme activity. TPS activities of 39.0 (mean value ± standard error; ± 3.1, n = 7) and 288.4 (± 30.4, n = 7) µKat g−1 ergosterol were determined for the fungal sheath and Hartig net, respectively, indicating a 7.4-fold higher TPS activity in the hyphae of the Hartig net. No TPS activity was observed in nonmycorrhizal fine roots.

In addition to gene expression and enzyme activity, trehalose content was determined. Trehalose could be detected in entire mycorrhizas and in the fungal sheath while it was below the detection limit in the Hartig net. The trehalose content in Hartig net hyphae was thus calculated as difference between the trehalose content in entire mycorrhizas and that of the fungal sheath.

Entire mycorrhizas, the Hartig net and the fungal sheath had ergosterol contents of 0.61 (± 0.05, n = 8), 0.36 (± 0.07, n = 7) and 2.03 (± 0.38, n = 7) mg g−1 dry weight (DW), respectively. Assuming comparable ergosterol contents in the hyphae of the fungal sheath and those of the Hartig net, entire mycorrhizas and the Hartig net (the mycorrhiza without the hyphal mantle) had fungal contents of 30 and 17.7%, respectively. This indicates a partitioning of hyphae, with 42% belonging to the fungal sheath and 58% to the Hartig net in intact A. muscariaP. tremula × tremuloides ectomycorrhizas.

Trehalose contents were determined as 168.7 (± 27.6, n = 9) and 365 (± 65.6, n = 9) µmol g−1 DW for entire mycorrhizas and the fungal sheath, respectively. As 17.4 and 12.6% of ectomycorrhizal DW is represented by the Hartig net and sheath hyphae, respectively, the trehalose content is c. 2.7 times higher in the hyphae of the Hartig net than in those of the fungal sheath (fungal sheath content per gram ectomycorrhizal DW: 365 × 0.126 = 46 µmol; Hartig net content per gram ectomycorrhizal DW: 168.7 − 46 = 122.7 µmol).

Discussion

In saprophytic ascomycetes, trehalose was initially thought to be a storage carbohydrate in addition to glycogen. However, in contrast to glycogen, trehalose accumulates in yeast (and other ascomycetes) only during the short lag phase after the addition of glucose to a culture, and not during exponential cell growth. At least for yeast, this makes a function as a storage compound rather unlikely (Wiemken, 1990). The reason trehalose does not accumulate during exponential yeast growth is probably that high trehalase activity accompanies trehalose biosynthesis (enhanced TPS/TPP gene expression). As a consequence, trehalose accumulates only in trehalase-deficient yeast cells (and under stress conditions), indicating that trehalose cycling is a physiologically important feature of fast yeast growth (Francois & Parrou, 2001).

In contrast to the situation in yeast and other saprophytic ascomycetes, trehalose accumulates in many ectomycorrhizal fungi growing on glucose and in symbiosis (Martin et al., 1988; Ineichen & Wiemken, 1992; Wallenda, 1996; Wallenda et al., 1996; Smith & Read, 1997; Rangel-Castro et al., 2002), independent of their classification as ascomycetes or basidiomycetes. In ectomycorrhizal fungi, which show a relatively slow cell division rate, the utilization of trehalose (and/or polyols) as a carbon store could therefore be a physiological adaptation to slow growth together with a continuous sugar supply from a host plant. In combination with enhanced uptake of hexoses (Nehls et al., 1998, 2001b; Wiese et al., 2000) and increased metabolic turnover (Schaeffer et al., 1996; Kowallik et al., 1998; Hampp & Schaeffer, 1999; Nehls et al., 2001a), transformation of plant-derived monosaccharides into fungus-specific metabolites such as trehalose (Martin et al., 1987, 1998; Ineichen & Wiemken, 1992) could thus be the basis for the creation of a strong carbon sink, which is necessary to allocate photoassimilates towards the mycorrhizal root.

‘Metabolic zonation’ and ‘physiological heterogeneity’ have been discussed as important concepts for a functional understanding of ectomycorrhizal symbiosis (Cairney & Burke, 1996; Timonen & Sen, 1998; Nehls et al., 2001b). In A. muscaria ectomycorrhizas, AmTPS and AmTPP expression was induced almost exclusively in the hyphae of the Hartig net. Here, increased AmTPS transcript abundances (18-fold) resulted in significantly higher enzyme activity (7.4-fold), indicating that TPS activity is presumably transcriptionally regulated in A. muscaria.

Together with that of AmTPS, AmTPP expression was also up-regulated. Because of the high trehalose background we were, however, not able to properly assay TPP activity, but it can be assumed that the latter is also increased.

As in the case of AmTPS and AmTPP, AmTP transcript abundances were much higher in Hartig net hyphae than in those of the fungal sheath. However, in contrast to AmTPS and AmTPP activity (exclusively generating trehalose), AmTP could work in both directions, in the biosynthesis and the phosphorolysis of trehalose (Wannet, 1999). As Hartig net hyphae are exposed to elevated apoplastic sugar concentrations (Nehls et al., 2001a), the greatly enhanced expression of all three genes indicates that AmTP is mainly active in the direction of trehalose biosynthesis (otherwise, a futile ATP-consuming glucose/trehalose cycle would be the result). In contrast to those of the Hartig net, the hyphae of the fungal sheath are not exposed to elevated apoplastic sugar concentrations and are thus dependent on the utilization of endogenous resources. As trehalose phosphorolysis would reduce the energy demand for trehalose utilization, it could be speculated that in hyphae of the fungal sheath AmTP catalyzes trehalose degradation to reduce the energy demand of these fungal cells. This speculation is supported by the observation that, in comparison to the extramycelial mycelium, expression of AmTPS and AmTPP was only slightly (2-fold) enhanced in the fungal sheath, while AmTP revealed 10-fold higher transcript abundances. A development-dependent regulation of fungal TP activity was demonstrated for Agaricus bisporus. In substrate mycelia, which were well supplied with external carbohydrates, TP catalyzed trehalose formation. However, in hyphal aggregates (developing fruit body primordia), TP degraded trehalose, yielding glucose and glucose-1-phosphate as carbon and energy sources (Wannet, 1999). Similar to the situation in A. bisporus, TP seems also to be developmentally regulated in A. muscaria.

In addition to gene expression and enzyme activity, trehalose content in the Hartig net was increased 2.7-fold in relation to the hyphae of the fungal sheath. This relatively small increase in pool size could be a result of rapid turnover and/or export towards the extraradical mycelium, a prerequisite for both hyphal growth and perpetuation of the sink function of mycorrhizal roots. Independent of the fate of trehalose, the data clearly show that the Hartig net is the primary site of trehalose biosynthesis in ectomycorrhizal symbiosis.

Differences in the apoplastic hexose concentrations of the Hartig net and the fungal sheath (Nehls et al., 2001a,b) as well as in their cellular nitrogen contents (Nehls et al., 2001a; Wipf et al., 2002; Javelle et al., 2004; Nehls, 2004) have been shown to act as signals regulating fungal physiological heterogeneity in ectomycorrhizas. This is obviously less relevant for the genes in this study. Compared with symbiosis, external glucose concentrations and the cellular nitrogen content had only a minor impact on the expression of any of the investigated genes and are thus obviously not the relevant signals. Evidently, signals other than sugar and nitrogen concentrations exist for the regulation and maintenance of physiological heterogeneity in fly agaric–poplar ectomycorrhizas.

In contrast to ectomycorrhizal symbiosis, where the expression of AmTPS and AmTPP increased simultaneously in the hyphae of the Hartig net, fungal growth on rich carbon and nitrogen sources in pure culture had opposite effects on the transcript abundances of the two genes. While AmTPS expression was slightly induced under these conditions, that of AmTPP was repressed. This can probably be explained by the finding that the substrate of TPP, trehalose-6-phosphate, acts as an important metabolic regulator in a number of fungi (Blazquez et al., 1993; Panneman et al., 1998; Foster et al., 2003). As increased AmTPS and reduced AmTPP activity would result in an increased cellular trehalose-6-phosphate content, fine-tuning of cellular carbon metabolism might occur via changes in the pool size of this compound in A. muscaria also.

In summary, in ectomycorrhizas formed between A. muscaria and P. tremula × tremuloides, expression of genes coding for proteins involved in trehalose biosynthesis (AmTPS, AmTPP and AmTP), as well as TPS activity and trehalose content, are increased specifically in hyphae located at the plant–fungus interface. This indicates that it is the Hartig net that is responsible for both the creation of the carbon sink and the redistribution of plant-derived carbohydrates. As AmTP can catalyze both biosynthesis and degradation of trehalose, and as its activity is under metabolite control, investigation of metabolite pools in the hyphae of the fungal sheath and the Hartig net will be the aim of future work.

Acknowledgements

This work was financed by the Deutsche Forschungsgemeinschaft as part of the focus program ‘Molecular Biology of Mycorrhiza’ (Ne 332/9-1). We are indebted to Margaret Ecke, Christopher Harvey, and Andrea Bock for excellent technical assistance and to Dr Nina Grunze for critical reading of the manuscript.

Ancillary