Differential gene expression during pre-symbiotic interaction between Tuber borchii Vittad. and Tilia americana L.
Ectomycorrhizal formation is a highly regulated process involving the molecular reorganization of both partners during symbiosis. An analogous molecular process also occurs during the pre-symbiotic phase, when the partners exchange molecular signals in order to position and prepare both organisms for the establishment of symbiosis. To gain insight into genetic reorganization in Tuber borchii during its interaction with its symbiotic partner Tilia americana, we set up a culture system in which the mycelium interacts with the plant, even though there is no actual physical contact between the two organisms. The selected strategies, suppressive subtractive hybridisation and reverse Northern blots, allowed us to identify, for the first time, 58 cDNA clones differentially expressed in the pre-symbiotic phase. Sequence analysis of the expressed sequence tags showed that the expressed genes are involved in several biochemical pathways: secretion and apical growth, cellular detoxification, general metabolism and both mutualistic and symbiotic features.
KeywordsTuber borchiiPre-infection phaseSuppressive subtractive hybridizationExpressed sequence tags
Ectomycorrhizae are symbiotic associations of plant roots and filamentous fungi. Nutrients are exchanged between the two partners by new functional biochemical pathways in the ectomycorrhizal structure. The plant provides organic carbon compounds to the fungus (Salzer and Hager 1991; Chen and Hampp 1993; Nehls et al. 1998), while the fungus, in turn, provides access to nitrogen, phosphorus and other nutrients otherwise unavailable to the plant (Rousseau et al. 1994; Brun et al. 1995; Perez-Moreno and Read 2000). Furthermore, there is evidence that mycorrhizal plants are more likely to survive in unfavourable conditions because the ectomycorrhizal association significantly improves resistance to plant pathogens (Duchesne et al. 1989).
The development of a functional ectomycorrhiza requires a new genetic and biochemical procedure (Tagu et al.1993; Martin et al. 2001; Sundaram et al. 2001; Voiblet et al. 2001). First, the fungus needs to grow towards the host roots. In a second stage, it must envelop the radical apparatus of its host plant and, finally, it must infect the fine roots, allowing development of the symbiotic structure.
Several studies have been carried out on differently expressed fungal genes during the infection stage and the ectomycorrhizal phase (Burgess et al. 1995; Martin et al. 1997; Laurent et al. 1999), with a particular focus on carbon and nitrogen metabolisms (Salzer and Hager 1991; Chen and Hampp 1993; Nehls et al. 1998, 1999). In contrast, little is known about fungal differential gene expression during the pre-infection stages. The aim of the present study is to identify differentially regulated genes during the early interaction between plant and fungus in order to understand the molecular events which lead to the formation of ectomycorrhizae in Tuber borchii Vittad.
Tub. borchii is an hypogeous ascomycetous fungus belonging to the Pezizales order. It is capable of forming ectomycorrhizae on the fine roots of gymnosperms and angiosperms. Although Tub. borchii ascocarp is not as popular as Tub. magnatum Pico, it is important commercially in Mediterranean areas.
An in vitro system in which the two symbionts (Tilia americana L., Tub. borchii) interact without direct contact was compared with the same in vitro system containing only the fungus; and 58 clones were identified, representing fungal genes differentially expressed in the pre-contact phase.
Most of these genes are involved in secretory and apical growth processes, while others seem to be involved in infection processes. There are also some involved in mitochondrial metabolism, gene expression regulation and general metabolism.
Materials and methods
The biological material for the construction of the subtracted library consisted of vegetative mycelia of Tub. borchii (strain 10 RA) grown on Nylon membranes laid on solid MS/2 medium (Murashige and Skoog 1962) at pH 6.5 and 24°C, with 10 g/l of glucose as carbon source, for 30 days under 16 h of light provided by cool white fluorescent lamps. These conditions were used for the Driver sample. The Tester sample was prepared as follows: Tub. borchii mycelia were grown under the above-mentioned conditions, but in the presence of the host plant Til. americana, from which they were separated by a Nylon membrane.
Independent samples for reverse Northern blot experiments were prepared according to the method of Sisti et al. (1998). Plantlets of Til. americana were first micropropagated and then inoculated with tissue blocks of Tub. borchii mycelium in tube cultures (4 cm diameter × 30 cm height) filled with 22 ml of peat-moss vermiculite (1:30 v/v) and embedded with 10 ml of MS/2 medium, at pH 6.5 and 10 g/l of glucose (interaction tubes). All the cultures were maintained for 30 days (ectomycorrhizae take longer) in a culture room at 24±1°C under 16 h of light provided by cool white fluorescent lamps (3,500 lx). Tubes in which the mycelium was permitted to grow alone were prepared as control tubes.
Isolation of total RNA
Total RNA for subtraction and reverse Northern experiments was extracted using a Qiagen RNeasy kit according to the manufacturer’s instructions. A DNase (Ambion) digestion step was performed before all subsequent reactions.
Subtracted library construction and clone selection
RNA obtained from the Tester and Driver cultures was used for the PCR select cDNA subtraction experiment (Clontech) according to the manufacturer’s instructions, with the exception of the ligation efficiency test and the subtraction efficiency test, which were carried out through amplification of the actin gene (AF462034) with specific Tub. borchii primers (forward 5′-GAGATGAGGCCCAATCCAAAC-3′, reverse 5′-CCAGAATCCAAACGATACCGG-3′). The resulting fragments were inserted in pGEM vector II (Promega) and subsequently cloned using Escherichia coli XL1-Blue. A total of 115 recombinant bacterial clones, obtained by blue–white selection on plates containing isopropyl-β-D-thiogalactopyranoside/X-Gal, were selected and sequenced by a private laboratory. A total of 58 clones of reliable sequence were considered for further analysis, using the BLASTX program (National Center for Biotechnology Information; NCBI). Clones were considered significantly similar to known proteins for E-values lower than 1×10−4.
Dot blotting and reverse Northern experiments
cDNA inserts of purified plasmids from the subtracted library were amplified by PCR with 0.5 μM universal primers SP6 and T7, 200 μM each deoxynucleotide, 0.1 units of Taq DNA polymerase and buffer supplied by the manufacturer (Qiagen). The PCR was performed using the Applied Biosystem GeneAmp 9700 PCR system according to the following parameters: 94°C for 5 min and 30 cycles of 94°C for 0.5 min, 55°C for 0.5 min and 72°C for 1 min. All PCR products were checked on 1.2% agarose gel stained with ethidium bromide. Identical quantities of the PCR products were blotted on several Hybond-N+ Nylon filters using blotting solution (0.4 N NaOH) and the BioRad Bio-Dot apparatus. Further, serial dilutions of an internal standard (actin gene) were added in each filter.
The RNA extracted from the tubes (three control tubes, three interaction tubes) was used for the synthesis of labelled cDNA for probes. The cDNA was obtained using oligo(dT) primers and Clontech PowerScript reverse transcriptase according to the manufacturer’s instructions. One microlitre of each cDNA was employed for labelled probe synthesis by the Amersham random prime labelling system Rediprime II kit in the presence of 30 μCi of (32P)dATP. The six replicated membranes were probed at 60°C for 12 h in hybridisation solution (0.3 M sodium phosphate, pH 7.2, 1 mM EDTA, 1% bovine serum albumin, 7% SDS). The washing procedure was carried out twice in 2× SSC/0.1% SDS and once in 1× SSC/0.1% SDS.
Data acquisition and analysis
Membranes were exposed by the Biorad BI imaging screen cassette and subsequently analysed using the Biorad GS 250 molecular imager system. The raw images were imported with Gel-Pro Analyzer ver. 3.1 software and the intensity of each spot was calculated using the “outline close to the dot option”. All values were normalised by constitutive actin gene expression (serial dilution from well F1 to F3) and the results were imported with SPSS statistical package ver. 10.1 software in order to perform the Mann–Whitney U-test on all spots of control and interaction samples.
After 30 co-culture days, Tub. borchii hyphae were gently removed from the membrane surface, fixed with a drop of FAA (40% formaldehyde/70% ethanol/acetic acid, 5:90:5), washed and mounted. Microscopic observation was performed with a Leica DMLB light microscope equipped with a DC300F (charge-coupled device) camera.
Tub. borchii genes differentially expressed during the pre-symbiotic interaction between Tub. borchii Vittad. and Til. americana L. The genes are sorted by E-values
Best BLASTX database match
Ribonucleotide reductase (Emericella nidulans)
Centractin-like protein (Pneumocystis carinii)
GAS-2 homologue (Candida glabrata)
26 s proteasome regulatory subunit mts4 (Schizosaccharomyces pombe)
Histone H4 (Neurospora crassa)
Putative GDP-mannose pyrophosphorylase (Candida albicans)
Inorganic pyrophosphatase (Pichia pastoris)
Asparagine synthase Asn2p (Saccharomyces cerevisiae)
COX1-i1 protein (Yarrowia lipolytica)
Nuclear transport factor 2 (Aspergillus nidulans)
60S ribosomal protein L6, YL16-like (Saccharomyces cerevisiae)
Probable mRNA maturase aI5-alpha (Saccharomyces cerevisia)
Related to trichodiene oxygenase cytochrome P450 (Neurospora crassa)
Aspartic protease (Aspergillus oryzae)
Glyoxal oxidase (glx1) [Arabidopsis thaliana]
COI intron 9 protein (Podospora anserina)
Cytochrome P450 (Fusarium sporotrichioides)
Hypothetical protein (Schizosaccharomyces pombe)
Hypothetical protein (Schizosaccharomyces pombe)
Putative secreted protein (Streptomyces coelicolor A3(2))
Rho gdp dissociation inhibitor. (Schizosaccharomyces pombe)
Syntaxin binding protein 1, sec1 family (Schizosaccharomyces pombe)
Related to fluconazole resistance protein (FLU1) [Neurospora crassa]
DNA-binding protein amdA (Emericella nidulans)
Alpha-l-rhamnosidase A precursor (Aspergillus aculeatus)
Histidine-rich protein (Plasmodium lophurae)
Possible nuclear pore complex associated (Schizosaccharomyces pombe)
RNA-dependent RNA polymerase (Ophiostoma mitovirus 3a)
RNA-dependent RNA polymerase (Ophiostoma mitovirus 3a)
CG9682 gene product (Drosophila melanogaster)
Hypothetical protein (Schizosaccharomyces pombe)
Putative integral membrane protein (Streptomyces coelicolor A3(2))
Alpha/beta-gliadin precursor (Triticum aestivum)
Filament-associated late protein FALPE (Amsacta moorei entomopoxvirus)
CG9682-PA (Drosophila melanogaster)
Hypothetical protein XP_146970 (Mus musculus)
Actin cytoskeleton-assiociated (Schizosaccharomyces pombe)
Putative Myb-related transcription activator protein (Arabidopsis thaliana)
Cold-inducible RNA binding protein (Xenopus laevis)
Nar1p (Saccharomyces cerevisiae)
Myosin-IA (Acanthamoeba castellanii)
Intercellular adhesion molecule 1 (Rattus norvegicus)
Envelope glycoprotein (Human immunodeficiency virus type 1)
Unnamed protein product (Homo sapiens)
Hypothetical protein (Cytophaga hutchinsonii)
Hypothetical protein XP_119113 (Homo sapiens)
Hypothetical protein XP_096386 (Homo sapiens)
CG15406-PA (Drosophila melanogaster)
Hypothetical protein (Chlamydia trachomatis)
Hypothetical protein (Plasmodium falciparum 3D7)
The clones that showed the highest homology with known proteins may be involved in: (1) cellular organelle dynamics and cell wall construction [S102 (GAS-2 homologue protein), S32 (centractin-like protein), S53/2 (putative secreted protein), S97 (syntaxin binding protein 1), S43 (Rho GDP dissociation inhibitor, GDI), S22 (aspartic protease), S100 (putative GDP-mannose pyrophosphorylase)], (2) mitochondrial/microsomal metabolism and cellular detoxification processes [S38 (glyoxal oxidase), S41 (cytochrome P450), S71 (COI intron 9 protein), S67 (related to trichodiene oxygenase cytochrome P450), S76 (COX1-i1 protein)], (3) cellular signaling [S35 (inorganic pyrophosphatase), S4 (nuclear transport factor 2), S29 (DNA-binding protein amdA)], or (4) cell cycle accomplishment and general metabolism [S87 (asparagine synthase), S103 (ribonucleotide reductase), S93 (26S proteasome regulatory subunit mts4), S59 (histone H4), S56/2 (60S ribosomal protein L6), S28 (probable mRNA maturase aI5-alpha), S27 (alpha-l-rhamnosidase A precursor), S42 (histidine-rich protein). Other clones analysed showed less similarity to known genes and hence it was not possible to confirm their possible role in the cell (S2, S11, S56, S31, S75, S26, S73, S37, S68, S79, S82, S96, S48/2, S44, S63, S23, S53, S24, S112, S3, S17, S74, S35/2, S48, S46, S61).
These results lead us to believe that a complex series of molecular mechanisms are activated in the very first stages of ectomycorrhizal formation (pre-infection phase), before the plant and fungus actually make contact. We may gain insight into the interaction between Tub. borchii and its host plant by looking at three principal events taking place during this phase.
In particular, the S97 clone encodes for the protein sec1, a molecule responsible for binding syntaxin and involved in vesicle membrane fusion (Brummer et al. 2001; Peng and Gallwitz 2002). The Candida glabrata GAS-2 protein-like molecule, encoded by clone S102, is a homologue of GAS1p in Saccharomyces cerevisiae and is also a homologue of C. albicans PHR1 and PHR2 (Weig et al. 2001). GAS1p, PHR1 and PHR2 are involved in cell wall assembly, in that they code a 1,3-β-glucanosil transferase (Mouyana et al. 2000). Moreover, the apical growth events are consistent with the expression of clone S32, which is directly involved in cell nuclear migration towards the apical tips of hyphae (Xiang and Morris 1999; Hirozumi et al. 1999).
The second metabolic process may be represented by the enzymes involved in cellular detoxification processes, encoded by clones S41 and S81. In particular, S81 encodes for a protein related to the fluconazole resistance protein FLU1 in Neurospora crassa. This protein is an ABC transporter providing azole resistance to the organisms expressing the gene (Calabrese et al. 2000). Evidence of the activation of detoxification processes can also be found in the detection of azole molecules and other volatile organic compounds during the pre-infection stage in the Til. americana–Tub. borchii mycorrhizal synthesis system (Menotta et al. 2004). Clone S41 encodes for the protein cytochrome P450, which is probably also involved in cellular detoxification processes (Lamb et al. 1999; Seo et al. 2000). In fact, this protein is strongly induced by benzothiazole, one of the volatile organic compounds previously detected (Menotta et al. 2004; Suwanchaichinda and Brattsten 2002).
Finally, a part of the identified clones includes genes involved in the general metabolism of the mycelium, such as S59, S71, S76, S87, S103 and S27. In particular, S76 (which encodes for a protein with similarities to a COX1 protein) and S71 (which encodes for a COI intron 9 protein), may be involved in the protein turnover of mitochondria, suggesting an increase in mitochondrial activity in this stage. Their expression or over-expression during the early interaction of Tub. borchii with Til. americana strongly suggests that mycelial metabolism increases in the presence of the plant, probably as a response activated by mitogenetic processes.
Many studies have been carried out during the past few years concerning the molecular changes which take place in Tub. borchii ectomycorrhizae (Polidori et al. 2002) or in other ectomycorrhizal species (Tagu et al. 1993; Martin et al. 2001, Voiblet et al. 2001). In contrast, limited information is available regarding early interactions in other ectomycorrhizal species (Podila et al. 2002) and very little is known about the biochemical processes that are at work during the pre-infection stage in Tub. borchii just before the establishment of symbiosis. In the current study, 58 genes differentially expressed at this stage were detected in Tub. borchii and cultivated without contact with its symbiotic plant partner, Til. americana.
The reverse Northern experiments showed that genes selected by suppressive subtractive hybridization are actually expressed independently from the culture system tested. In fact, both the mycelia grown in ectomycorrhizal synthesis tubes and the mycelia grown on Nylon (see Materials and methods) suggest that the selected genes are not affected by the culture system. Furthermore, the pre-symbiotic interaction model provided to be a useful tool for studies of the pre-symbiotic phase of these organisms.
This evidence leads us to believe that a complex series of molecular mechanisms switches on during the very first stage of ectomycorrhizal formation, before the plant and fungus make contact. In particular, for the first time, we highlight the expression of several genes that may be involved in the apical growth of Tub. borchii towards the roots of its symbiotic plant, as shown by morphological analysis.
Proteins such as 60S ribosomal protein and cytochrome P450 (corresponding to clones S56/2 and S41, respectively) have been shown to be directly or indirectly involved in fungus–plant signaling in arbuscolar symbioses (Gianinazzi-Pearson et al. 2002). Furthermore, an improved mitochondrial metabolism was recently demonstrated in Gigaspora rosea (Tamasloukht et al. 2003). The results reported in this paper are consistent with those findings.
This evidence suggests that there may be a common expression mechanism in a wide range of species during the pre-symbiotic plant–fungus interaction, both in endo- and in ectomycorrhizal species.
In our findings, we have highlighted the expression of genes having a high homology with genes involved in the pathogenic and saprophytic interaction of fungi. In fact, clone S38 encodes for a glyoxal oxidase, an enzyme that is necessary for the oxidation activity of ligninolytic peroxidases in Phanerochaete chysosporium (Kersten et al. 1995; Cullen 1997). The ability to degrade lignin might be a necessary step for Tuber infection. Glyoxal oxidase is also expressed by the pathogenic fungi Fusarium oxysporum and Trichoderma atroviride. Finally, clone S102 encodes for a C. glabrata GAS-2 protein-like molecule. This protein has been shown to be necessary for virulence during the infection of host tissue (Weig et al. 2001; De Bernardis et al. 1998; Muhlschlegel and Fonzi 1997; Saporito-Irwin et al. 1995; Ghannoum et al. 1995).
In conclusion, the present study represents a first step towards gaining a better understanding of the molecular mechanisms at work in the initial phases of symbiosis in Tuber; and it attempts to highlight the early signal exchanges that occur between the two symbionts prior to actual contact. Several genes involved in various biochemical mechanisms, not yet cloned in the genus Tuber, were isolated. Even if further studies are required, in light of the results obtained herein and those reported in literature cited above, it can be asserted that, in the pre-contact phase, the fungus switches on a series of mechanisms, such as those to recognise and attack plant roots and defend the fungus from substances secreted by the plant. All of those mechanisms are accompanied by highly regulated metabolic modifications.
Studies on the molecular mechanisms at work in the pre-infection phase in Tub. borchii–Til. americana are an important part of research on symbiosis. Most studies have mainly focused on endo-mycorrhizal fungi and ectomycorrhizal basidiomycete fungi, whereas the present study concerns an ascomycete fungus.
Moreover, since some of the mechanisms hypothesised here are similar to those used by pathogenic or mycorrhizal fungi during the infection of host-plant tissue, there may be a common ancestral infection process that subsequently evolved into different mechanisms of interaction: symbiotic, pathogenic and all the intermediate stages. Considering recent advances in the study of fungal behaviour in plant–fungus interactions (Hall et al. 2003), we could hypothesise a more complicated relationship. In fact, many “ectomycorrhizal fungi” such as Tricholoma matsutake can switch from a symbiotic stage to a parasitic one, depending on host-plant species, general health of the host plant and pedo-climatic conditions.