• Ectomycorrhizal interactions established between the root systems of terrestrial plants and hyphae from soil-borne fungi are the most ecologically widespread plant symbioses. The efficient uptake of a broad range of nitrogen (N) compounds by the fungal symbiont and their further transfer to the host plant is a major feature of this symbiosis. Nevertheless, we far from understand which N form is preferentially transferred and what are the key molecular determinants required for this transfer.
• Exhaustive in silico analysis of N-compound transporter families were performed within the genome of the ectomycorrhizal model fungus Laccaria bicolor. A broad phylogenetic approach was undertaken for all families and gene regulation was investigated using whole-genome expression arrays.
• A repertoire of proteins involved in the transport of N compounds in L. bicolor was established that revealed the presence of at least 128 gene models in the genome of L. bicolor. Phylogenetic comparisons with other basidiomycete genomes highlighted the remarkable expansion of some families. Whole-genome expression arrays indicate that 92% of these gene models showed detectable transcript levels.
• This work represents a major advance in the establishment of a transportome blueprint at a symbiotic interface, which will guide future experiments.
In natural ecosystems, the main resources acquired by plants are: light and CO2, through photosynthesis in the leaves; and mineral nutrients and water, through root and mycorrhiza uptake (i.e. arbuscular mycorrhizal (AM) or ectomycorrhizal (ECM)). These symbioses play a critical role for plant nutrient use efficiency in natural ecosystems that are usually characterized by nutrient limitation. Nutrient transport, namely hyphal absorption from soil solution, nutrient transfer from the fungus to plant and carbon movement from plant to fungus, is a key feature of mycorrhizal symbioses. Phosphorus, nitrogen and carbohydrates are considered to be the main nutrients transferred by the mycorrhizal symbiosis, although supply of water and trace elements can also be very important under certain conditions (Smith & Read, 1997).
In forest soils, N is present as mineral sources (, ) and/or as organic compounds such as amino acids, peptides and proteins (Marschner, 1995). Ammonium is a ubiquitous intermediate in nitrogen metabolism and one of the major nutrients for plants and microorganisms. Soil concentrations of the poorly mobile ammonium ion are generally lower than those of nitrate, yet ammonium is often preferred as a source of N because of its smaller assimilation cost (Marschner, 1995; von Wirén et al., 2000). Based on both [15N] labelling and mass balance data, it was demonstrated that hyphal acquisition contributes 45% to total plant N uptake under N deficiency (Jentschke et al., 2001). Disrupting the external mycelium from ectomycorrhiza greatly decreased [15N] ammonium uptake by birch seedlings (Javelle et al., 1999). Therefore, external hyphae can be considered as the ammonium-absorbing structure of ectomycorrhizal roots.
In addition, a number of observations support the view that ectomycorrhizal fungi have the ability to degrade macromolecular nitrogen and furthermore to take up and assimilate the products of hydrolytic degradation (small peptides, amino acids) (Chalot & Brun, 1998; Wallenda & Read, 1999; Nasholm & Persson, 2001). The fungal partner can also use animal-origin N and transfer it to the host tree (Klironomos & Hart, 2001). Indeed, even if plants seem to have abilities to exude proteolytic enzymes and to take up proteins and protein degradation products without assistance from other organisms (Paungfoo-Lonhienne et al., 2008), they seem to rely largely on their symbionts for organic N uptake under natural conditions: several studies have demonstrated that plant biomass and plant N content were higher for mycorrhizal plants grown on organic N as sole N source compared with nonmycorrhizal ones (Abuzinadah & Read, 1986; Abuzinadah et al., 1986; Tibbett & Sanders, 2002; Wallander, 2002). All of the major hydrolytic enzymes involved in mobilization of N from organic compounds have been detected in ericoid and some in ectomycorrhizal fungi (Read & Perez-Moreno, 2003). Furthermore, several gene families encoding secreted proteases have been identified in the ectomycorrhizal fungus Laccaria bicolor genome sequences and are expressed in ectomycorrhizal and fruiting body mycelia (Martin et al., 2008). Uptake of amino acids was investigated in the mycorrhizal fungi Paxillus involutus (Chalot et al., 1996) and Amanita muscaria (Nehls et al., 1999), which demonstrated their ability to absorb a variety of amino acids. Data indicate that ectomycorrhiza had higher capacities for glutamate uptake compared with nonmycorrhizal roots (Plassard et al., 2000) and that mycorrhiza formation induced a profound alteration of the metabolic fate of exogenously supplied glutamate (Blaudez et al., 2001). However, the extent of this potential to use organic N under natural conditions is not well known (Nasholm & Persson, 2001).
The physiological processes involved in the further transfer of N within the symbiotic tissues are still poorly understood. Net transport of nutrients from the soil solution to the aboveground part of the plant will be the result of three transport components: one located at the soil–fungus interface and the two others located at the symbiotic fungus–plant interface (fungal cell–apoplast, apoplast–plant cell). For most mycorrhizal types, carbon derived from photosynthesis is then transferred from plant to fungus, followed by translocation to the growing margins of the extraradical mycelium and to developing spore or fruiting bodies. The large surface of apoplastic interface between plant and fungal membranes is the site of solute exchange between the two symbionts. Nutrients must therefore pass several membrane barriers and the apoplastic interface before their assimilation by plant and fungi cells. Membrane transporters are the molecular determinants of transport in and out of the cell or organelles and thus key players in nutrient uptake and exchange mechanisms. Their regulation patterns are essential in adapting to changes in soil nutrient quantity and/or quality. Furthermore, it has been suggested that when a fungus colonizes a plant, the nature and dynamics of nutrient exchange determine the outcome of their interaction (parasitic or symbiotic) (Divon & Fluhr, 2007). Plant and fungal cells must be ‘reprogrammed’ to fulfil the tasks of massive nutrient transfer.
During the past 10 yr, only 12 proteins involved in the transport of N-related compounds have been fully characterized in ectomycorrhizal fungi (Table 1). Given the variety of N-compound transporters described in animals, plants and yeasts (Transport Classification Database TCDB, http://www.tcdb.org/), these 12 proteins characterized in mycorrhizal fungi are far from representing the genetic potential of a fungal organism such as L. bicolor. Characterization of many more fungal transporters is needed, and expected.
Table 1. Proteins involved in the transport of nitrogen (N)-related compounds characterized in ectomycorrhizal fungi
Indeed, with the genome of L. bicolor now available (Martin et al., 2008), we have the unique opportunity to identify and present a global view on proteins involved in transport of N compounds in an ectomycorrhizal fungus. This genomic view should shed light into the symbiotic way of life and physiology of L. bicolor and would allow us to hypothesize on the forms of N taken up from the soil and transferred to the plant symbiont. Furthermore, transporters are expected to be efficient metabolic checkpoints for emergence of complex anabolic or catabolic pathways. The gain or loss of transporters is an essential step in the differentiation of large intracellular metabolic pathways. It is reasonable to assume that it generally results from physiological selection of diverging duplicated genes during the adaptation to new ecological niches (Kellis et al., 2003; De Hertogh et al., 2006), or to different fungal ways of life (parasitic, symbiotic or saprobic). Further analyses are now needed to point out specific gain or losses of functions related to symbiotic behaviour.
The major aim of the present study is to establish a repertoire of gene models potentially involved in the transport of N compounds in the ECM fungal model, L. bicolor. We focused our analysis on genes encoding transporters of either inorganic or organic forms of N (ammonium, nitrate, urea, amino acid, peptide and amino acid derivatives (i.e. nucleobase, allantoate and polyamine)). These candidate genes belong to several multigene families (Table 2). The L. bicolor repertoire was compared with the gene contents of four other sequenced genomes of saprobic (Phanerochaete chrysosporium and Coprinopsis cinerea) or parasitic (Ustilago maydis and Cryptococcus neoformans) basidiomycetes. The expansion or reduction of gene model number of L. bicolor transporter families compared with these four other species is discussed. Furthermore, the genome of Saccharomyces cerevisiae was included in this comparative study, since it was the first eukaryotic genome to be fully sequenced, becoming the ‘workhorse’ for functional genomic analysis (systematic individual gene disruption). In addition, the transcript levels for all the genes included in this repertoire were analysed by large-scale expression analysis of fungal or symbiotic tissues.
Table 2. Families of nitrogen (N)-related compound transporters analysed in this study
Transporters are classified according to TC classification (Saier, 2000).
The ammonia channel transporter (Amt) family
The YAAH/ATO Family
The nitrate/nitrite porter (NNP) family
The solute:sodium symporter (SSS) family
The amino acid–polyamine–organocation (APC) superfamily
The l-type amino acid transporter (LAT) family
The amino acid/choline transporter (ACT) family
The yeast amino acid transporter (YAT) family
The amino acid/auxin porter (AAAP) family
The drug:H+ antiporter-1 (DHA1) family
The drug:H+ antiporter (DHA2) family
The mitochondrial carrier (MC) family
The oligopeptide transporter (OPT) family
The proton-dependent oligopeptide transporter (POT) family (also called the PTR (peptide transport) family)
The nucleobase:cation symporter-1 (NCS1) family
The nucleobase:cation symporter-2 (NCS2) family
The purine transporter, AzgA (AzgA) family
The drug/metabolite transporter (DMT) superfamily
The mitochondrial carrier (MC) family
The concentrative nucleoside transporter (CNT) family
The equilibrative nucleoside transporter (ENT) family
Allantoate and allantoin
The anion:cation symporter (ACS) family
The Drug:H+ antiporter-1 (DHA1) family
Materials and Methods
Free-living mycelium of L. bicolor S238N was grown on cellophane-covered agar plates containing modified Pachlewski medium (pH 5.5) according to Di Battista et al. (1996) before harvesting the proliferating hyphal tips at the periphery. Ectomycorrhizal materials collected were either Laccaria–Douglas fir and Laccaria–Populus trichocarpa (INRA clone 101-74) ectomycorrhiza grown in glasshouse or Laccaria–Populus tremula × alba (INRA clone 717-1B4) ectomycorrhiza grown in vitro, as described previously (Kohler et al., 2006; Martin et al., 2008). Fruiting bodies of L. bicolor S238N were collected below Douglas fir seedlings grown in a glasshouse. Two stages of fruiting bodies were harvested: an early stage (stage 1–2 of six) and pileus and stipe from the mature stage (stage 6 of six) (see the Supporting Information, Fig. S1).
In silico genome automatic annotation and manual curation
blast, Advanced Search and Gene Ontology tools at the JGI Laccaria genome database (http://genome.jgi-psf.org/Lacbi1/Lacbi1.home.html) and the INRA Laccaria database (http://mycor.nancy.inra.fr/IMGC/LaccariaGenome/) were used to identify gene models (putative genes) involved in N-compound transport in the draft genome of L. bicolor S238N-H82. Additional searches were performed to probe the Laccaria genome database with the use of a range of sequences of N-compound transporters and genes available from fungi at NCBI GenBank (http://www.ncbi.nlm.nih.gov), Transport Classification database (TCDB; http://www.tcdb.org) and TransportDB (http://www.membranetransport.org; Ren et al., 2007). All gene models detected were inspected manually, and the automatically selected best gene model of the JGI Laccaria genome database was modified if necessary. The manually annotated gene sequences were aligned and verified using the programmes clustalx (version 1.83.1) (Jeanmougin et al., 1998). Each curated homologue was further used for blast search at the JGI, YeastDB (http://www.yeastgenome.org/) and Broad-MIT Institute (http://www.broad.mit.edu) databases to check for similar genes in other fungi, including C. cinerea, C. neoformans, U. maydis, P. chrysosporium and S. cerevisiae. Throughout the paper, protein IDs were used to identify these gene models.
Whole-genome oligoarray analyses
Tissues were immediately frozen in liquid nitrogen and RNA extraction was carried out using either the RNeasy Plant Mini Kit (Qiagen, Courtaboeuf, France) or a cetyltrimethylammonium bromide (CTAB)-based extraction protocol for Douglas fir ectomycorrhizal root tips (Peter et al., 2006). Total RNA preparations were amplified using the SMART PCR cDNA Synthesis Kit (Clontech, Mountain View, CA, USA) according to the manufacturer's instructions.
The L. bicolor whole-genome expression array manufactured by NimbleGen Systems Limited (Madison, WI, USA) was used for expression analyses according to Martin et al. (2008). Single dye labelling of cDNAs, hybridization procedures, data acquisition, background correction and normalization were performed at the NimbleGen facilities (NimbleGen Systems, Reykjavik, Iceland). Average expression levels were calculated for each gene from the independent probes on the array and were used for further analysis. Probes with less than six nucleotide difference for closely related gene models (risk of cross-hybridization) were eliminated before the calculation of the average expression level. To estimate the signal background and the resulting cut-off level for significant expression, the mean intensity of 30 000 random probes present on the microarray was calculated. Gene models with an expression higher than threefold the cutoff level were considered as transcribed. Transcripts with a change of more than threefold in transcript level between two conditions were considered as significantly differentially expressed. The complete expression dataset is available as series (accession number # GSE9784) at the Gene Expression Omnibus at NCBI (http://www.ncbi.nlm.nih.gov/geo/).
Gene expression data presented in Fig. 2 were obtained as follows: expression levels mentioned for ectomycorrhiza corresponded to the mean value of expression in mycorrhiza of L. bicolor–Douglas fir (2 repetitions), mycorrhiza of L. bicolor–P. trichocarpa, produced in the glasshouse and L. bicolor–P. tremula × alba produced in vitro; values of expression in fruiting bodies corresponded to the mean value between stage 1, 2 and 6 fruiting bodies; finally, expression in free-living mycelium corresponded to the mean of 3-wk-old mycelium for two repetitions.
Full-length amino acid sequences of L. bicolor and the four other saprobic (P. chrysosporium and C. cinerea) or parasitic (U. maydis and C. neoformans) basidiomycete transporters, as well as characterized sequences of S. cerevisiae and other fungi transporters, were aligned by clustalw and imported into the Molecular Evolutionary Genetics Analysis (mega) package version 4 (Tamura et al., 2007). Phylogenetic analyses were conducted using the neighbour-joining (NJ) method implemented in (mega), with the pairwise deletion option for handling alignment gaps, and with the Poisson correction model for distance computation. Bootstrap tests were conducted using 1000 replicates. Branch lengths (drawn in the horizontal dimension only) are proportional to phylogenetic distances. Based on the same data, phylogenetic trees were reconstructed using different algorithms (Maximum Parsimony, Maximum Likelihood, Minimum Evolution) to confirm topology and bootstrap values. Analyses were performed using mega 4 software or online tools (http://www.phylogeny.fr; Dereeper et al., 2008).
Results and Discussion
The ammonia channel transporter (Amt) family (TC #1.A.11) This family of transporters is spread throughout the whole tree of life (Andrade & Einsle, 2007). In silico analysis highlighted a remarkable expansion of ammonium transporter (Amt) encoding gene models in the L. bicolor genome sequence (eight gene models), compared with saprobic or parasitic basidiomycetes (Table 3), and to the three Amt genes characterized in Hebeloma cylindrosporum (Javelle et al., 2003b). The phylogenetic analysis showed that putative Amt members can be grouped in six clusters, reflecting the classification previously established for the ascomycete ammonium transporters (Monahan et al., 2006), with the addition of cluster VI (Fig. 1). Laccaria bicolor sequences grouped only in clusters I and IV, which associated basidiomycete and ascomycete sequences contrary to the other clusters specific to either ascomycetes or basidiomycetes. Cluster I grouped five L. bicolor sequences, which might be orthologues of HcAMT3, and several other characterized low affinity transporters. This cluster also included one orthologue from each saprobic basidiomycete but, interestingly, no orthologue from the parasitic ones (Table 3) and was therefore clearly expanded in L. bicolor. Cluster IV contained three L. bicolor gene models (Table 3) and previously characterized high-affinity transporters. Lacbi1:188643 and Lacbi1:255304 might be orthologues of HcAMT2 of H. cylindrosporum and Lacbi1:190906 is closely related to HcAMT1.
Table 3. Number of transporter gene models in Laccaria bicolor, Coprinopsis cinerea, Phanerochaete chrysosporium, Ustilago maydis and Cryptococcus neoformans
A challenging question concerns the physiological role of multiple transporters, with likely distinct functional and regulatory properties, in the context of ammonium retrieval by mycelia and mycorrhiza. Transcript profiling analyses were then used to investigate gene expression in more detail. These analyses showed detectable transcript levels for six L. bicolor ammonium transporter genes, in all tissues tested (free-living, fruiting body and mycorrhizal mycelium). However, the expression of Lacbi1:255304 was very weak (Fig. 2). A cross-hybridization between oligonucleotide probes and the sequences of three closely related paralogues, Lacbi1:254389, Lacbi1:310393 and Lacbi1:313221 did not allow to distinguish their specific expression on microarrays. However, sequencing of reverse transcriptase polymerase chain reaction (RT-PCR) products suggested that Lacbi1:313221 was the only gene expressed among these three paralogues (data not shown). We considered therefore that the transcripts level detected on arrays correspond to this gene.
Lacbi1:300932 and Lacbi1:188643 were constitutively expressed in all tissues. These genes products could therefore ensure a basal level of ammonium uptake, as already demonstrated for HcAMT3 (Javelle et al., 2003b). Lacbi1:190906, Lacbi1:313221 and especially Lacbi1:331747 were highly expressed in ectomycorrhiza, while their expression in free-living mycelium was reduced to a basal level. Furthermore, Lacbi1:190906 was also highly expressed in fruiting bodies. We have shown by RT-PCR that this gene was upregulated by nitrogen starvation and downregulated by ammonium (data not shown), suggesting increased ammonium uptake capacities of the N-starved fungal cell. This result is consistent with work on the high-affinity transporters HcAMT1 and HcAMT2 (Javelle et al., 2001, 2003a,b).
The higher expression in ectomycorrhiza combined with the expansion of the Amt family in L. bicolor could suggest higher ammonium uptake capacities in fungal cells of mycorrhizal mycelium and therefore a key role of ammonium in the ectomycorrhizal symbiosis. Indeed, several studies highlighted the fact that ammonium taken up into the ectomycorrhizal fungal hyphae is incorporated to the key N donors glutamate and glutamine through either the glutamine synthetase/glutamate synthase pathway (GS/GOGAT) or glutamate dehydrogenase (GDH), before transfer to the plant (Smith & Read, 1997; Chalot et al., 2006). Ammonium has been also proposed for N form exchanged at the symbiotic interface (Chalot et al., 2006). There, ammonium retrieval by fungal hyphae would interfere with a function in N export. Interestingly, Willmann et al. (2007) showed that in functional ectomycorrhiza, the transcript level of the high-affinity ammonium transporter AMT2 of A. muscaria was decreased in both hyphae networks (sheath and Hartig net), while extraradical hyphae revealed strong gene expression. Our results did not show any repressed Amt gene in ectomycorrhiza; however, entire mycorrhizal tips were used for expression analysis. Independent harvest of L. bicolor extraradical hyphae, as well as further separation of Hartig net and fungal sheath mycelium of L. bicolor mycorrhiza, by laser capture microdissection (LCM) microscopy, will allow the expression analyses of Amt transporter genes in these microcompartments. Alternatively, the higher expression of Amt family members, as described in the present study, may be explained by increased capacity for vesicles loading. Diffusion of NH3 or active transport of via an Amt-mediated system into acidified vesicles would indeed ensure compartmentation of excess ammonia. The ammonia-loaded vesicles could then move via microtubules to the symbiotic membrane where vesicles would fuse with the plasma membrane and release ammonia into the interfacial apoplast (Chalot et al., 2006).
The YAAH/ATO family (TC #9.B.33) In the L. bicolor genome, two gene models were found as putative orthologues of yeast Ato protein-encoding genes, showing no expansion of this family compared with the four other sequenced basidiomycetes (Table 3). Several studies in S. cerevisiae, whose genome contains three YAAH/ATO members localized on plasma membrane, suggested that yeast Ato (Ammonia (Ammonium) Transport Outward) proteins could act as ammonium/H+ antiporters, extruding ammonium from yeast cells and importing protons (Řičicováet al., 2007). Many prokaryotes and basal eukaryotes (mainly fungi) but no plants or animals carry several homologues of these proteins.
One of these L. bicolor gene models, Lacbi1:295862, was constitutively expressed and was one of highest expressed genes of the genome (Fig. 2), while Lacbi1:295502 was upregulated in free-living mycelia (Fig. 2). The RT-PCR analyses confirmed a strong expression of Lacbi1:295862 in free-living mycelium, independently of the N physiological status while Lacbi1:295502 was repressed in N-starved mycelium (data not shown). These transcript levels in L. bicolor also suggest that the two ATO proteins may play a different role in ammonium export in ectomycorrhizal mycelium as well as in free-living mycelium. However, it cannot be excluded that they are not directly pumping ammonium out of the cells: they may be involved either in transport of other, yet unidentified substrates, or in transfer of a regulatory signal.
The nitrate/nitrite porter (NNP) family (TC #2.A.1.8) Nitrate transporters of the NNP family allow internalization of nitrate by an energy-dependent uptake process, before reductive assimilation trough the action of nitrate reductase and nitrite reductase. In agreement with previous findings, a single nitrate transporter gene was identified in the L. bicolor genome. This gene clustered with nitrate reductase and nitrite reductase, which participate in a common metabolic pathway. This is a well-described gene cluster in ascomycetes (Siverio, 2002) and in the basidiomycete H. cylindrosporum (Jargeat et al., 2003). Nitrate transporters are generally induced by and repressed by reduced N sources such as ammonium and glutamine. According to transcript profiling data, this gene was highly expressed in both ectomycorrhiza and free-living mycelium (Fig. 2). This family is remarkably conserved with respect to gene number per genome: in each of the five basidiomycete genomes, a single NNP encoding gene was found. Obviously, a strong selection pressure against gene duplication and gene loss exists for this transporter family among basidiomycetes. Interestingly, both L. bicolor and H. cylindrosporum have multiple ammonium transporters and a single nitrate permease: ammonia is arguably the most important inorganic N source for ectomycorrhizal fungi. However, it has been demonstrated that L. bicolor is able to grow on nitrate as sole N source (Yamanaka, 1999).
The solute:sodium symporter (SSS) family (TC#2.A.21) Urea represents an important nutrient in forest soils. A number of mycorrhizal fungi, including L. bicolor, have the capabilities to use urea as a sole N source (Yamanaka, 1999; Morel et al., 2008). Similarly to other basidiomycetes, two putative proteins involved in the transport of urea were found in L. bicolor genome (Table 3). Since most of the fungi possess several active urea transporter-coding sequences, one may wonder about the physiological significance of urea transport in fungi compared with plants, which possess a single urea transporter. Moreover, large-scale expression analysis carried out on L. bicolor mycelium and mycorrhiza revealed that both genes were expressed. Lacbi1:315056, which presents the highest sequence homology to PiDUR3, a urea transporter-encoding gene of P. involutus, was highly expressed in the plant-associated mycelium (Fig. 2). These results did not agree with the upregulation of PiDUR3 in the extraradical hyphae which are the absorbing structure of the ectomycorrhizal association between P.involutus and B. pendula. This may be explained by different N status mycelia since PiDUR3 gene has been shown to be sensitive to N catabolic repression (Morel et al., 2005, 2008). Both L. bicolor gene models were expressed in fruiting body mycelium, suggesting an important role of corresponding proteins in fruiting bodies. Several studies showed that L. bicolor forms reproductive structures after urea application (Sagara 1992, 1995). Urea may serve as a osmotically favourable form of nitrogen reserve in fruiting bodies, which can be used for protein synthesis, cell wall and other N compounds needed to build up the fruiting body, as already suggested for Agaricus bisporus (Wagemaker et al., 2005, 2006). The general importance of the urea cycle to N assimilation in ECM fungi (Martin et al., 1994) has also been demonstrated (Morel et al., 2005; Wright et al., 2005). Urea formation is indeed the result of purine catabolism and ornithine cycle activities and urea release accompanies the degradation of compounds like arginine, agmatine, allantoin and allantoic acid. In fungal cells, urea may also be broken down in ammonium by urease activity. Microarray analyses showed that the urease coding gene was upregulated in the ectomycorrhizal mycelium of L. bicolor (Martin et al., 2008), a result of interest since recent works suggested that ammonia released from urea breakdown could pass to the host plant cells (Chalot et al., 2006).
Amino acid transport
The amino acid–polyamine–organocation (APC) superfamily (TC #2.A.3) Most of the fungal amino acid transporters have been classified into the APC superfamily (Saier et al., 1999; Jack et al., 2000; van Belle & André, 2001; Wipf et al., 2002b; Chang et al., 2004). They mediate the transfer of a broad spectrum of amino acids with overlapping specificities. The APC superfamily was shown to be expanded in L. bicolor (29 gene models), compared with the saprobic or parasitic basidiomycetes (Martin et al., 2008) (Table 3). Interestingly, the number of gene models in L. bicolor was similar to that of S. cerevisiae (Andre, 1995; Chang et al., 2004). The expansion of L. bicolor APC gene models could be related to higher capacities to use organic nitrogen (amino acids) and to the dual trophic abilities (symbiotic vs saprobic) of L. bicolor (Martin et al., 2008). The phylogenetic analysis showed that APC putative members could be grouped in four clusters, reflecting the four families already described for S. cerevisiae APC transporters superfamily (Fig. 3).
The amino acid/auxin porter (AAAP) family (TC #2.A.18) The AAAP family was recently included in the APC superfamily (Chang et al., 2004) and contained only three L. bicolor members, which is less than other basidiomycetes (Table 3, Fig. 3). The three L. bicolor members were detected only in cluster III, which included the yeast Avts (Avt2-7) known to mediate amino acid transport into and out of vacuoles (Russnak et al., 2001). The three AAAP genes of L. bicolor were constitutively expressed in all tissues tested (Fig. 2) and may therefore represent the genetic potential for amino acid sequestration into vacuole of fungal hyphae, a process that has already been supposed to be a key determinant in N transfer between rhizomorph hyphae (Smith & Read, 1997).
The L-type amino acid transporter (LAT) family (TC #2.A.3.8) The LAT family included six L. bicolor gene models, which was similar to the C. cinerea gene models number but represented an expansion compared with other basidiomycetes (Table 3). The family was divided into three distinct clusters according to bootstrap values (Fig. 3). Cluster I grouped one orthologue of each of the five basidiomycetes, all closely related to the S. cerevisiae MUP3 gene which encodes a low-affinity methionine permease. Cluster II comprised one L. bicolor gene (Lacbi1:249579), which grouped with A. fumigatus MUP gene, but included neither P. chrysosporium nor U. maydis gene models. Interestingly, cluster III, which included the S. cerevisiae MUP1 gene, an l-methionine high-affinity permease also involved in cysteine uptake, comprised several paralogues of L. bicolor and saprobic fungi.
Four of the LAT genes (Lacbi1:229402, Lacbi1:249579, Lacbi1:166470 and Lacbi1:250926) were highly and constitutively expressed in all tissues tested, whereas the two other LAT genes (Lacbi1:298603 and Lacbi1:298666) were upregulated in fruiting bodies (Fig. 2). It is tempting to hypothesize that L. bicolor LAT gene models may encode putative methionine uptake systems, as demonstrated for the distantly related orthologues of S. cerevisiae. Thus, expansion of gene models number as well as the expression analyses may point to an important role of methionine in the physiology of L. bicolor. Mansouri-Bauly et al. (2006) proposed that L. bicolor may improve sulphate supply to the host-plant, and in return, plant would provide the fungus with reduced sulphur. We could speculate that some of L. bicolor LAT genes could be involved in plant-origin methionine import.
The amino acid/choline transporter (ACT) family (TC #2.A.3.4) The ACT family included nine L. bicolor and P. chrysosporium gene models, which represented an expansion compared with other basidiomycetes (Table 3). This family was subdivided into two distinct clusters and few orphan genes (Fig. 3). In cluster I, three genes of L. bicolor grouped with S. pombe thiamine uptake transporter recently characterized (Vogl et al., 2008). Laccaria bicolor gene expression analyses (Fig. 2) revealed functional specialization of mycelia, Lacbi1: 252280 being upregulated in mycorrhiza, Lacbi1:185515 in free-living mycelium and Lacbi1:185557 in both free-living mycelia and fruiting body. Cluster II was represented by the S. cerevisiae genes BIO5 (encoding a biotin transporter), HNM1 (encoding a high-affinity choline and ethanolamine permease) and several genes encoding gamma-aminobutyric acid (GABA) permeases. This cluster grouped six gene models of L. bicolor (Fig. 3). Two paralogues, Lacbi1:191715 and Lacbi1:311288 were closely related to Neurospora crassa (NCU07175.1) gene, which might be related to a GABA permease, but unfortunately, we were unable to obtain expression data for these two closely related genes because of cross-hybridization. Four other genes (Lacbi1:234289, Lacbi1:244309, Lacbi1:142738 and Lacbi1:254728) grouped with the S. cerevisiae UGA4 gene, which encodes a vacuolar membrane permease involved in the transport and utilization of GABA (Andre et al., 1993). Expression analysis revealed that Lacbi1:234289 and Lacbi1: 244309 were constitutively expressed in all tissues tested, while Lacbi1:142738 was 10 times upregulated in fruiting bodies and Lacbi1:254728 was 25 times upregulated in mycorrhiza and fruiting bodies. These results suggest functional divergence of these four L. bicolor gene models, which might have specialized and important functions related to GABA transport in symbiosis and fruiting body physiology. Previous results showed indeed that substantial [15N] enrichment in GABA occurred in P. involutus mycelium fed with  (Finlay et al., 1989). Accumulation of GABA was also observed during assimilation in the mycelium of free-living L. bicolor and excised beech (Fagus sylvatica) ECM root tips (Martin et al., 1986, 1994).
The yeast amino acid transporter (YAT) family (TC #2.A.3.10) The YAT family included 11 L. bicolor gene models and displayed thus significant expansion compared with other basidiomycete genomes (Table 3). Comparison of primary sequences gave rise to three clusters (Fig. 3). Cluster I comprised one L. bicolor gene (Lacbi1:143007) which grouped with the S. cerevisiae gene AGP2. This gene encodes the major polyamine permease in yeast, responsible for high-affinity spermidine transport and contributing to a substantial fraction of total putrescine uptake. Lacbi1:143007 was constitutively expressed in each mycelia type (Fig. 2). Cluster II comprised four gene models of L. bicolor (Fig. 3). Two of them (Lacbi1:157448 and Lacbi1:301980) were closely related to the A. fumigatus LYP gene, encoding a putative lysine-specific permease. The other paralogues (Lacbi1:143243 and Lacbi1: 296817) grouped with general amino acid permeases of P. chrysogenum and A. fumigatus as well as several S. cerevisiae characterized genes. Owing to the high sequence similarity, only expression of Lacbi1:296817 was established and revealed that this gene was constitutively expressed. Cluster III included six genes of L. bicolor (Fig. 3). One of them (Lacbi1:182249) grouped with the S. cerevisiae DIP5 gene, which encodes a dicarboxylic amino acid permease and mediates mostly high-affinity and high-capacity transport of l-glutamate and l-aspartate, but also glutamine, asparagine, serine, alanine and glycine. This gene had orthologues in each of the other fungal genomes and was highly expressed and furthermore upregulated in mycorrhizas (×4.5) and fruiting bodies (×3) (Fig. 2). This cluster included also several characterized genes. Lacbi1:300838 grouped with the A. muscaria general amino acid permease that has been characterized as a high-affinity amino acid transporter with a broad substrate spectrum (Nehls et al., 1999). Low gene expression in ectomycorrhiza and amino acid uptake data indicated two main functions for AmAap1: uptake of amino acids from soil for fungal nutrition, and prevention of an amino acid loss by hyphal leakage in the absence of a suitable N source. Interestingly, the L. bicolor orthologue of AmAAP1 (Lacbi1:300838) was highly upregulated in ectomycorrhizal (×15.4) as well as in fruiting body (×20) mycelium. The transporter encoded by this gene might be a key marker of amino acid transport within ectomycorrhizal and fruiting body tissues. Finally, four L. bicolor paralogues closely related to the H. cylindrosporum general amino acid permease HcGAP1 were found in cluster III. HcGAP1 encodes a high-affinity general amino acid permease with a putative role in organic N uptake from the soil (Wipf et al., 2002a). Expression analyses revealed that two of the paralogues, Lacbi1:308595 and Lacbi1:308597, were constitutively expressed at low levels whereas the two others, Lacbi1: 245329 and Lacbi1:188015, were upregulated in mycorrhizal and fruiting body mycelia, thus likely playing an important role in symbiosis and fruiting body.
We have pointed out the remarkable expansion of the YAT multigenic family and according to the gene expression, revealed that several of these genes might be key determinants in ectomycorrhiza functioning. Indirect evidence from [15N]-labelling experiments supports the view of amino acid transfer from the fungus to the host (either glutamine or alanine in ECM and arginine in AM) (Smith & Read, 1997; Jin et al., 2005; Chalot et al., 2006). Unfortunately, several S. cerevisiae characterized genes (i.e. GNP1, encoding a glutamine permease; AGP1, encoding a low-affinity amino acid permease with broad substrate range, involved in uptake of asparagine, glutamine, and other amino acids; CAN1, encoding a plasma membrane arginine permease) grouped together and did not allow us to highlight L. bicolor orthologues. Further functional characterizations are now needed to point out key markers of exchanges.
The drug:H+ antiporter-1 (DHA1) family (TC #2.A.1.2) The DHA1 subfamily comprises 26 gene models in L. bicolor, and is expanded in all five basidiomycete genomes (Table 3) compared with the 15 genes found in S. cerevisiae. Two large clusters can be distinguished as described previously (Gbelska et al., 2006).
Cluster I was divided into two subclusters (Fig. 4). Known amino acid transporters were found in cluster I.1, which is represented by the multidrug-resistance transporter of S. cerevisiae Aqr1. Interestingly, Aqr1 is located in multiple internal-membrane structures and cell surface and is apparently involved in excretion of amino acids present in high concentrations in the cytosol (Velasco et al., 2004). Furthermore, expression of another yeast member of this cluster (YBR043c/QDR3) is under N control. Three L. bicolor genes were found within this subcluster and one gene, Lacbi1:297564, was upregulated in free-living mycelium (×24), whereas the two other genes (Lacbi1:301091 and Lacbi1:315406) were constitutively expressed (Fig. 2). Ahmad et al. (1990) showed that L. bicolor mycelium in pure culture assimilated more N than needed for growth and exceed-N was released in external medium as free amino acids. Laccaria bicolor AQR1 orthologues may play such a role in amino-acid excretion in free-living mycelium grown in presence of excess N. Furthermore, active transporters excreting amino acids at the plasma membrane could also be important in functional ectomycorrhiza.
Major expansion of the DHA1 family in L. bicolor corresponded to subcluster I-2 and will be discussed later according to putative substrate transported (see polyamine section).
The drug:H+ antiporter (DHA2) family (TC. 2.A.2.3) The L. bicolor genome contained nine members of the DHA2 subfamily, which is less than those found in C. neoformans and P. chrysosporium genomes (Table 3). Phylogenetic analyses distinguished two different clusters (Fig. 5). Cluster I corresponded to the S. cerevisiae ATR1 (YML116w) drug resistance gene. Three genes from L. bicolor belonged to this cluster and, interestingly, the genome of C. neoformans, a human pathogen, displayed a large number of genes belonging to this drug-resistance cluster. Lacbi1:298212 was upregulated in ectomycorrhiza (×30). Such a gene might be useful to deal with toxic plant compounds and/or pathogenesis-related (PR) proteins, which genes were upregulated early in symbiotic interaction (Duplessis et al., 2005). Orthologues of the yeast VBA1/2 genes, encoding vacuolar proteins involved in transport of basic amino acids (Shimazu et al., 2005) grouped within the same cluster (cluster II) and may be of interest for amino acid sequestration in vacuoles of mycorrhizal fungi. This cluster did not display differences in gene numbers of each of the five basidiomycetes and may thus represent the minimum equipment for vacuolar amino acid transport in fungi. One of the L. bicolor genes (Lacbi1:315452) was up-regulated in fruiting bodies (×16.5), whereas Lacbi1:315438 and Lacbi1:236212 were highly expressed in ectomycorrhiza and free-living mycelium (Fig. 2). This suggests a functional divergence among these three genes.
The mitochondrial carrier (MC) family (TC #2.A.29) Among the 36 L. bicolor genes models classified as members of the MC family, five were identified as putative amino acid transporters. Their common function is to provide a link between mitochondria and cytosol. Transcript profiling showed no differential expression between tissues. However, two of the five putative mitochondrial amino acid transporters were highly expressed (Fig. 2). Lacbi1:301012 seemed to be an orthologue of the S. cerevisiae carnitine transporter gene CRC1 (Palmieri et al., 1999) and Lacbi1:187395 may be an orthologue of the S. cerevisiae aspartate/glutamate antiporter gene ScAGC1. ScAgc1 plays a role in ornithine synthesis by importing glutamate into mitochondrial matrix, and is essential for aspartate supply from mitochondria to cytosol, required for urea synthesis (Cavero et al., 2003). In addition, two orthologues of the N. crassa ARG13 and S. cerevisiae ARG11 genes (Catoni et al., 2003) were identified (Lacbi1:153024 and Lacbi:232799), whereas only one gene model was found in the four other basidiomycete genome sequences. The ScArg11 protein exports ornithine from mitochondria to cytosol and is required for arginine biosynthesis (Crabeel et al., 1996). Hence, to produce arginine, glutamate must pass from the cytosol to the mitochondrial matrix, where it is converted to ornithine. The resulting ornithine must pass from the matrix into the cytosol before it can be transformed to arginine. In yeast, ScAgc1 and ScArg11 indeed provide a link between the mitochondrial and cytosolic parts of arginine biosynthetic pathway, and thus are essential components of urea cycle. Chalot et al. (2006) proposed that inorganic N could be taken up, assimilated and converted into arginine in external fungal cells of ectomycorrhizal fungi. Arginine may further be translocated to internal fungal cells, where it would be broken down, releasing urea and ornithine. Urea could then be broken down in ammonium and pass to the host cells as mentioned earlier. Mitochondrial transporters may therefore play a key role in N transfer from mycorrhizal fungi to host plant.
The proton-dependent oligopeptide transporter (POT) family (TC #2.A.17) and the oligopeptide transporter (OPT) family (TC #2.A.67) The L. bicolor genome comprises two PTR (or POT family) genes, Lacbi1:252700 and Lacbi1:301981, which might be the orthologues of H. cylindrosporum HcPTR2A and HcPTR2B respectively. Previous studies showed that HcPTR2B was constitutively expressed independently of the N-source whereas HcPTR2A was strongly expressed during N-deficient conditions or in the presence of a secondary N source. The authors suggested that peptide uptake in H. cylindrosporum was regulated by mechanisms sensing both extracellular and intracellular N sources (Benjdia et al., 2006). By contrast, both L. bicolor transporter genes were constitutively expressed in fungal tissues, with Lacbi1:252700 highly expressed (Fig. 2). Thus, they may be involved in the constitutive uptake of peptides as suggested for HcPtr2B.
The genome analysis led to the identification of eight putative L. bicolor OPT genes, grouping in three clusters. Fungal homologues were not grouped by species, with the exception of some OPT members from C. albicans that grouped in a subcluster (I.2) (Fig. 6). Interestingly, saprobic fungi showed higher number of OPT gene models compared with parasitic ones (Table 3). An expansion of L. bicolor, P. chrysosporium and C. cinerea OPT gene number was clearly shown in a distinct subgroup (I.4). These three fungi, known to have capacities to degrade organic matter, may therefore have the capacities to take up end products of this degradation, across the cellular membrane in an energy-dependent manner. After uptake, the internalized peptides can be rapidly hydrolysed by peptidases and used as a source of amino acids, N or carbon. Expression analyses revealed different functional profiles of L. bicolor OPT gene models (Fig. 2). Interestingly, two of them are highly and specifically upregulated in fruiting body mycelium (Lacbi1:312906 and Lacbi1:190652) while two others (Lacbi1:310913 and Lacbi1:250173) were upregulated in both fruiting body and ectomycorrhizal mycelium (Fig. 2).
These results suggest that all fungal tissues tested may have capacities for peptide uptake either for fungal nutrition or internal recycling. It has been previously shown that nutrient mobilization from natural organic substrates (animal necromass) is a key function of the vegetative mycelium of mycorrhizal systems (Perez-Moreno & Read, 2000, 2001; Klironomos & Hart, 2001). In addition, L. bicolor belongs to a group of fungi that form reproductive structures on soils where decomposition of animal wastes occurred, which are thus rich in organic N (Sagara, 1995). Furthermore several gene families encoding secreted proteases have been identified in L. bicolor genome sequences and are expressed in ectomycorrhizal and fruiting body mycelia (Martin et al., 2008). This evidence indicates that peptide transport may be of importance for N cycling in forest soils.
Amino acid derivatives transport
Nucleobase transport Nucleobase (purine and pyrimidine) transport is a widespread phenomenon in prokaryotes and eukaryotes (Cabrita et al., 2002). Fungal nucleobase-specific transporters belong to three families, probably constituents of a single superfamily: the nucleobase:cation symporter-1 (NCS1) family (TC #2.A.39), the nucleobase:cation symporter-2 (NCS2) Family (TC #2.A.40), and the purine transporter, AzgA (AzgA) family (TC #2.A.1.40) of the MFS superfamily.
Phylogenetic analyses revealed that the NCS1 family was divided in two distinct clusters, with no distinction between ascomycete and basidiomycete sequences (Fig. 7). Cluster I grouped eight L. bicolor gene models and the four characterized S. cerevisiae genes encoding the purine–cytosine permeases (FCY), which mediate the active transport of adenine, hypoxanthine, guanine and cytosine into the cell (Ferreira et al., 1997). It also grouped TPN1, the yeast plasma membrane vitamin B6 transporter gene (Stolz & Vielreicher, 2003). This cluster showed a great expansion in L. bicolor genome compared with other basidiomycetes (Table 3). Transcript profiling analyses showed that only five of the eight gene models were expressed. Lacbi1:254256, Lacbi:248955, Lacbi1:310067 and Lacbi1:236267 were constitutively expressed while Lacbi1:245872 was upregulated in fruiting bodies (Fig. 2). This result could be of interest since the transporters of the NCS1 family absorb molecules involved in urea formation through purine degradation and, as already mentioned, fungi may accumulate substantial amounts of urea in their fruiting bodies. However, purine degradation accounts for only a minor part in urea synthesis in A. bisporus fruiting bodies (Wagemaker et al., 2005). By contrast, the undeniable upregulation of this transporter shows that the purine pathway could be of higher importance in L. bicolor fruiting bodies.
Cluster II contained a single L. bicolor gene model, which is less than for other basidiomycetes (Table 3). There was no C. cinerea sequence in this cluster. Characterized members of this cluster are the S. cerevisiae genes FUR4 (coding a uracil permease), FUI1 (coding a uridine permease located in the plasma membrane) NRT1 (a nicotinamide riboside transporter gene), THI10 (a thiamine transporter gene) and DAL4 (an allantoin permease gene) (Wagner et al., 1998; Belenky et al., 2008).
Sequences encoding transporters of NCS2 and AzgA families (one of each) were also found in the L. bicolor genome (Table 3, Fig. 7).
Nucleoside transport Nucleoside transporters are specialized integral membrane proteins that mediate cellular uptake and release of nucleosides and nucleoside analogue drugs, important for energetic metabolism of the cell. Four different protein families possess members involved in the transport of nucleosides (Table 2). These are presented and discussed in the Supporting Information Text S1, Nucleoside transport.
Allantoate and allantoin transport
The anion:cation symporter (ACS) family (TC #2.A.1.14) The ACS family from the MFS superfamily may transport either anions (e.g. phosphate) or organic anions (e.g. allantoate and nicotinate) in symport with protons or other cations (e.g. Na+). Pao et al. (1998) showed that proteins specific for different but structurally related substances are found within the same cluster. This family was highly expanded in all basidiomycetes (Table 3) compared with the nine genes found in S. cerevisiae genome.
To avoid biased group distribution owing to extended N- or C-termini, only the MFS domain was taken for the phylogenetic tree reconstruction. Three clusters were identified within this family (data not shown). Cluster I is represented by the S. cerevisiae genes FEN2, SEO1 and VHT1. Four L. bicolor genes belonged to this cluster and one gene (Lacbi1:248936) was upregulated in ectomycorrhiza (×5.3) (Fig. 2). Cluster II, represented by S. cerevisiae allantoate/ureidosuccinate permease gene DAL5, YCT1 and YLR004c, contained a variable number of members (Table 3). Interestingly, one L. bicolor gene (Lacbi1:187852) was upregulated in ectomycorrhiza (× 6.7) and fruiting bodies (×7.9) (Fig. 2) while the other (Lacbi1:151898) was upregulated in fruiting bodies (×12), suggesting functional divergence. Since DAL5 in yeast was shown to be subjected to N catabolism repression, over-expression of one DAL5 orthologue in ectomycorrhiza could indicate the N-starved status of ectomycorrhizal tissues. Accordingly, a DAL5 gene sharing strong homology with the yeast DAL5 was found to be upregulated in ectomycorrhizal symbiosis between P.involutus and B. pendula (Morel et al., 2005).
Cluster III was represented by the S. cerevisiae TNA1 (YGR260w) gene encoding a plasma membrane permease with high affinity for nicotinic acid (mainly used by the cells for NAD synthesis) or structurally related compounds (Klebl et al., 2000). Expression analyses revealed differential expression of these gene models according to mycelia type (Fig. 2). This suggests a neofunctionalization of these genes (one of the duplicates retains the ancestral functions while the other evolves to perform a novel function) of these L. bicolor genes.
The drug:H+ antiporter-1 (DHA1) family (TC #2.A.1.2) In cluster II of the DHA1 family, the TPO (polyamine transporters) genes dominate (Fig. 4). S. cerevisiae TPO1 to TPO4 genes all confer resistance to polyamine by detoxifying excess spermidine and putrescine (Albertsen et al., 2003). These transporters were originally proposed to be localized in the yeast vacuolar membrane (Tomitori et al., 1999, 2001), but more recent results have localized them in the plasma membrane (Sickmann et al., 2003). One L. bicolor orthologue of the yeast TPO2/3 genes was upregulated in fruiting bodies (Lacbi1:251253). Within subcluster II-1, two genes of L. bicolor were highly expressed (Fig. 2), one in ectomycorrhiza and free-living mycelium (Lacbi1:252635), and another one in free-living mycelium and fruiting bodies (Lacbi1:235328).
Among the DHA1 family, the major expansion in L. bicolor corresponded to subcluster I-2, which contained 13 out of the 26 L. bicolor genes (Fig. 4). This subcluster included yeast YCR023c. The biological role of YCR023c remains undetermined. However, it physically interacts with a Ptk2, a serine–threonine kinase involved in both ion transport across the plasma membrane and spermine and putrescine uptake (Ptacek et al., 2005). Thus, it is tempting to speculate that genes belonging to this subcluster may have a role in polyamine uptake. Interestingly, one gene (Lacbi1:312917) was upregulated in fruiting body mycelium; three (Lacbi1: 158313, Lacbi1:311884 and Lacbi1:333006) were highly expressed in ectomycorrhiza, whereas two others (Lacbi1: 296932 and Lacbi1:158313) displayed high levels of expression in both fruiting bodies and free-living mycelium.
From these genomic data, we can hypothesize that each of L. bicolor mycelia types may have a better ability for polyamine (or related compound such as putrescine) transport. However, Yamanaka (1999) showed that L. bicolor is unable to grow on media containing amines (ethylenediamine or putrescine) as sole N sources. In forest ecosystems, previous studies suggested that L. bicolor do not utilize amines directly, but are likely to use ammonium produced either by their microbial decomposition or chemically released from soil organic matter under alkaline conditions (created by the presence of amines).
Hibbett et al. (2000) suggested that mycorrhizal symbionts have evolved repeatedly from saprotrophic precursors, but also that certain ectomycorrhizal fungi have retained the genes for enzymes that can degrade plant cell walls, which may have facilitated repeated evolutionary reversals to saprotrophy. Sagara et al. (1992) also showed that, in Japan, L. bicolor is a facultative symbiont and may behave as a saprotroph in the soil. One can speculate on the specific aspects of each of these two trophic behaviours (saprobic vs symbiotic). No ectomycorrhiza-specific gene has been found so far (Duplessis et al., 2005), which suggests that ontogenic and metabolic programmes leading to functional symbiosis are driven by changes in the organization of gene networks and in gene expression in both partners (Duplessis et al., 2005). The coevolutionary implication is that the ectomycorrhizal symbiosis is not only a juxtaposition of the genetic potentialities of the two partners but leads indeed to novel metabolic patterns in hyphae and plant cells, based, among others, on increased uptake capacity of the root system and effective exchanges between symbionts (Hibbett et al., 2000; Duplessis et al., 2005).
Recent studies on mycorrhizal interactions have highlighted some key transporters for nutrient uptake or metabolite transfer at the biotrophic interface. New N, phosphorus or carbon transporters have been identified either from mycorrhizal fungi (H. cylindrosporum, A. muscaria, Tuber borchii, Glomus versiforme, Glomus intraradices) or from their host plants (Oryzae sativa, Medicago truncatula, Picea abies, P. trichocarpa) (Harrison et al., 2002; Selle et al., 2005; Chalot et al., 2006; Javot et al., 2007b; Muller et al., 2007). Many of these transporters displayed differential expression depending on nutrient concentration; however, only a few of them (endomycorrhiza-specific Pi transporter (Javot et al., 2007a), A. muscaria hexose transporter (Nehls et al., 1998) and P. trichocarpa ammonium transporters (Selle et al., 2005; Couturier et al., 2007) have been shown to be specifically upregulated in symbiotic tissues, indicating a putative role in symbiotic exchanges at the plant–fungal interface. Precise function and localization of most of these transporters in fungal tissues remain largely unknown and require further investigation. The nature of released nutrients or assimilated metabolites into the apoplast interface by the fungus remains unclear and the answer to whether amino acids or ammonium are the preferentially transferred N form have still not been elucidated. The nature of the N form transferred to plant and mechanisms involved in this transfer through the apoplast will likely arise from molecular approaches coupled with biochemical techniques.
Existing data, mostly deriving from studies based on targeted analytical approaches, lead only to partial views of biotrophic interface functioning. The present analysis aimed at establishing a repertoire of gene models that currently represent the genetic potential for transport of N-related compounds in L. bicolor. We have revealed the presence of at least 128 gene models encoding putative transport protein for N-containing compounds in the genome of L. bicolor. Among them, 118 genes (92%) showed detectable transcript levels. A comprehensive physiological analysis of the current dataset is not within the scope of the present study, but we would like to highlight some of the striking results from this genome survey which would be helpful for ectomycorrhizal and fruiting body characterization of functional compartmentation and key markers identification for exchanges.
In silico analysis of sequenced genomes of symbiotic (L. bicolor), saprobic (P. chrysosporium, C. cinerea) or parasitic (U. maydis and C. neoformans) basidiomycetes have enabled us to point out some remarkable expansions of L. bicolor transporter-encoding genes families (e.g. AMT (Fig. 1), YAT and LAT (Fig. 3), NCS1 (Fig. 7)), mostly by specific expansion of gene model number in several clusters of these families. These members are likely to play a key role in the symbiotic way of life compared with parasitic or saprobic ones. We may also assume that physiological constraints produced by the nature of the available growth substrates may have driven the emergence of paralogue transporters. Furthermore, these paralogues may play an important role in functional specialization of multicellular tissues differentiated in ectomycorrhiza or fruiting body. We also observed some clusters containing a majority of symbiotic and saprobic fungal sequences and only a minority of parasitic fungal ones (e.g. OPT; Fig. 6). Some L. bicolor gene models do not cluster with characterized genes, and could therefore correspond to new molecular determinants. This enrichment in N transporter encoding genes may point out a specialization of L. bicolor during the coevolution with its plant host. Symbiotic relationship is indeed mostly built on the ability of the fungus to transfer nutrient or assimilated metabolites, especially N, to the host.
Expression analyses have allowed us to point out a number of:
• constitutively expressed genes, which might play a fundamental role in basal transport processes
• upregulated or/and highly expressed genes in ectomycorrhiza, which may be important for mycorrhiza function at either import of N compounds at the soil–fungus interface or export of N compounds at the fungus–plant interface
• upregulated or/and highly expressed genes in the fruiting body, which may be important for fruiting body development, functioning and internal nitrogen recycling
• repressed genes in the fruiting body compared with free-living mycelium
• repressed genes in ectomycorrhiza compared with free-living mycelium. The latter genes may be involved in uptake of N compounds by the free-living mycelium as well as in the prevention of N compound losses by hyphal leakage
We showed that L. bicolor has the genetic potential for both mineral and organic N compound utilization. The expression of a part of this genetic potential will further define the functions of the different mycelia types and will emphasize the dual trophic abilities (symbiotic vs saprobic) of L. bicolor. The expansion of the AMT family, in addition to the putative role of these transporters in the symbiosis, could also be linked to the place of L. bicolor in ecological successions. L. bicolor is indeed described as a member of ‘ammonia fungi’– a group of fungi that occur after decomposition of urine, faeces or dead bodies has occurred, playing thus an important role in waste material removal (Sagara, 1992, 1995). Physiological studies on those fungi suggested that ammonia is the key substance for their growth, even if they readily use nitrate as sole N source (Yamanaka, 1999).
Our results highlighted a wide range of amino acid and peptide transporters involved either in uptake or internal recycling. L. bicolor has been shown to have little/no ability to grow on organic N sources such as amines, proteins or some amino acids (Yamanaka, 1999). However, expression patterns of these transporter genes could indicate that they may play a role in fungal physiology.
Although putative amino acid and ammonium exporters encoding genes (ScATO and ScAQR1 orthologues) did not display mycorrhiza-specific expressions patterns, they were expressed in mycorrhiza and we could not exclude an important role of these exporters at the symbiotic interface. However, vesicle loading proposed by Chalot et al. (2006) could also be an important mechanism in L. bicolor. Direct analysis of apoplast fluid (Solomon & Oliver, 2001) together with tracer experiments could be an appropriate tool to elucidate transfer between fungal and root cells. Detailed analysis of individual compounds using specific isotope analysis may also be a key to further address these important questions (Evans, 2001). Furthermore, the increasing application of molecular and postgenomic methodologies (large-scale oligoarrays), as well as technological improvements with higher spatial resolution (LCM technology), is now available for nutrient, transcript or protein detection within apoplast interfaces (Lohaus et al., 2001; Balestrini et al., 2007). Together with the present gene repertoire, this will allow further establishment of the biotrophic ‘transportome blueprints’ (transport proteins and compounds) at symbiotic interfaces.
E.L. was supported by a PhD fellowship from the ‘Ministère de l’Enseignement supérieur et de la Recherche’. We thank the US DOE Joint Genome Institute and the Broad Institute for access to the L. bicolor and C. cinerea genome sequences before publication. We also thank Marie-Pierre Oudot Le Secq and Benoît Hilselberger for their assistance in the bioinformatic analysis, and two anonymous reviewers for helpful comments on the earlier drafts of this manuscript. The transcript profiling analysis was supported by the EVOLTREE network of excellence.