The high-affinity poplar ammonium importer PttAMT1.2 and its role in ectomycorrhizal symbiosis

Authors


Author for correspondence: Uwe Nehls Tel: +49 7071 2977657 Fax: +49 7071 295635 Email: uwe.nehls@uni-tuebingen.de

Summary

  • • One way to elucidate whether ammonium could act as a nitrogen (N) source delivered by the fungus in ectomycorrhizal symbiosis is to investigate plant ammonium importers.
  • • Expression analysis of a high-affinity ammonium importer from Populus tremula× tremuloides (PttAMT1.2) and of known members of the AMT1 gene family from Populus trichocarpa was performed. In addition, PttAMT1.2 function was studied in detail by heterologous expression in yeast.
  • • PttAMT1.2 expression proved to be root-specific, affected by N nutrition, and strongly increased in a N-independent manner upon ectomycorrhiza formation. The corresponding protein had a KM value for ammonium of c. 52 µm. From the seven members of the AMT1 gene family, one gene was exclusively expressed in roots while four genes were detectable in all poplar organs but with varying degrees of expression. Ectomycorrhiza formation resulted in a strong upregulation of three of these genes.
  • • Our results indicate an increased ammonium uptake capacity of mycorrhized poplar roots and suggest, together with the expression of putative ammonium exporter genes in the ectomycorrhizal fungus Amanita muscaria, that ammonium could be a major N source delivered from the fungus towards the plant in symbiosis.

Introduction

Plant species differ greatly in their capacity to utilize particular nitrogen (N) forms, which contributes to their unique spatial and temporal distribution (Kronzucker et al., 1997; Min et al., 1999). Poplar, a member of temperate forest ecosystems, is equally capable of growing at low and high concentrations of both NO3 or NH4+ (Min et al., 1999). Nitrogen influx studies in Populus tremuloides revealed saturable, high-affinity importers (HATS) operating at external N concentrations below 500 µm, and low-affinity importers (LATS) which are active at higher N concentrations (Min et al., 2000). However, none of these transporters has yet been analysed at the molecular level.

In nonpolluted forest ecosystems N is restricted mainly to the litter layer. Plants, however, usually have a very limited capacity to utilize complex substances, which are typical for this organic layer (Perez-Moreno & Read, 2001a,b). Furthermore, the amount of free N in soil water is limited owing to reduced bacterial nitrification activity and the tight association of ammonium to humic substances (Yu et al., 2002). This limitation of easily available N sources in forest soils has led to a variety of adaptations in trees. One is the formation of ectomycorrhizas, an association of fungal hyphae of certain ascomycetes or basidiomycetes with fine roots of trees of temperate and boreal forests (Smith & Read, 1997). The essence of ectomycorrhiza function is the bidirectional exchange of soil-derived nutrients (e.g. N) delivered by the fungus for plant-derived carbohydrates across an apoplastic plant/fungus interface, the so-called Hartig net (Smith & Smith, 1990). The plant particularly benefits from the improved nutrition (e.g. with N and phosphate) in symbiosis (Read et al., 1991). While N uptake from soil by fungal hyphae is relatively well understood, N export at the plant–fungus interface is still a matter of debate. According to literature, amino acids are most commonly believed to be the N source delivered by fungal hyphae at the Hartig net, but ammonium is also discussed (Smith & Read, 1997; Javelle et al., 2004).

One way of obtaining more information about the putative N source that is delivered by fungal hyphae is the investigation of plant N importers. In this report we focused on the expression analysis of one family of high affinity ammonium transporters (AMT1) from poplar.

Materials and Methods

Biological material

Poplar growth and ectomycorrhiza formation in a Petri-dish system Populus tremula × tremuloides and Populus trichocarpa cuttings were rooted in MS-medium (Sigma, St Louis, MO, USA) under sterile conditions. 6-wk-old plants were transferred into Petri dishes containing sugar-free MMN medium (Hampp et al., 1996) such that the root system was inside and the shoot outside of the Petri dish. Both plant root systems inoculated with Amanita muscaria to form ectomycorrhizas and uninoculated control plants were grown in small plastic glasshouses for additional 6 wk at 18°C using a 12 h photoperiod with 100 µE m−2 s−1 illumination. Fully developed leaves (12 wk old), stems, fine roots, and mycorrhizas were isolated around noontime, frozen in liquid nitrogen, and stored at −80°C.

Growth conditions of glasshouse experiments Populus tremula× tremuloides and P. trichocarpa plants were grown in plastic pots (12 × 10 × 10 cm) filled with a mixture of Perlite, sand, and commercial potting soil (Floradur Type 1; Floragard, Oldenburg, Germany; 1 : 1 : 1; v : v : v) as a substrate. Plants were watered daily with tap water and fertilized every 2 wk with 100 ml of a nutrient solution containing 6 g l−1 of a complete fertilizer (Hakaphos Blau; Bayer, Leverkusen, Germany). Plants were kept under long-day conditions (16 h light exposure) with day and night temperatures of 20 ± 3°C and a relative humidity of 60 ± 10%, respectively. Sink leaves (3 wk old), fully developed source leaves (4–6 months old), stems (basal 10 cm), second-order main roots and late-order fine roots were isolated around midday, frozen in liquid nitrogen and stored at −80°C.

Nitrogen nutrition experiments  For treatments with different N sources, 6-wk-old Populus tremula × tremuloides plants were grown for further 6 wk in a Petri dish system and transferred into scintillation vials containing 20 ml MMN medium such that the root system was inside and the shoot outside. The medium was aerated several times during the growth period that followed. To obtain N-starved plants, a preincubation for 3 d in a medium without any N source was performed before the nutrition experiment. After exchange for fresh MMN medium containing the appropriate N concentration, plants were incubated for up to 4 d (18°C, 12 h day/night periods with 100 µE m−2 s−1 illumination) with an exchange of the solutions twice a day (morning and late afternoon). Plants were harvested (if not described otherwise) around midday and leaves, stems and roots were collected and frozen in liquid nitrogen and stored at −80°C.

Isolation of ectomycorrhizas and nonmycorrhized fine roots from field-grown poplar plants  Field samples were collected from the root systems of 2-yr-old P. tremula × tremuloides plants after washing the roots with tap water to remove soil particles. Ectomycorrhizas and nonmycorrhizal fine roots were collected under a binocular using DuPont no. 5 forceps (Roth, Karlsruhe, Germany) and stored in RNAlater (Qiagen, Hilden, Germany) for up to 2 wk before RNA extraction.

Northern blot analysis

Total RNA was isolated according to Nehls et al. (1998). Northern blot analysis was performed according to Nehls et al. (1998) using 8 µg total RNA of nonmycorrhized poplar roots grown in axenic culture. Hybridization was performed using a labelled probe from the 3′-UTR (untranslated region) of the PttAMT1.2 cDNA. Signal intensities were quantified using a flatbed scanner (Scanmaker IIHR; Microtek, Neuss, Germany) and the program package nih image (version 1.62; http://rsb.info.nih.gov/nih-image).

Semiquantitative reverse transcriptase polymerase chain reaction (RT-PCR)

Aliquots of c. 1 µg total RNA were treated with DNAse I (Invitrogen, Groningen, the Netherlands) according to the manufacturer's instructions and used for first-strand cDNA synthesis in a total volume of 20 µl, containing 50 pmol oligo-d(T)18-primer (Amersham Pharmacia Biotech, Braunschweig, 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 were added and aliquots were stored at −80°C.

For further polymerase chain reaction (PCR) experiments, the amount of first-strand cDNA in the different samples was calibrated such that all samples showed identical signal intensities after amplification with specific primers for 18S rRNA (5′-ACGCTCTGGATACATTAGC-3′ and 5′-TCCACCAACTAAGAACGGC-3′), the nearly constitutively expressed poplar gene PttJip1 (Grunze et al., 2004) and a poplar actin gene (Langer et al., 2004), followed by agarose gel electrophoresis and ethidium bromide staining. The PCR was performed using 2 U Goldstar Taq (Eurogentec, Seraing, Belgium) in a total volume of 25 µl containing 5% dimethyl sulfoxide (cycling conditions: 1 min at 91°C; 30 s at 55°C; 30 s at 72°C). Cycle numbers were chosen such that amplification was always in the linear range.

PttAMT1.2-specific primers (5′-GTCACTACAACATTGGCTGG-3′ and 5′-GGCATCAAATTTGATCACAC-3′) were used for PCR (the reverse primer was selected from the 3′-UTR of the gene).

Aliquots of 6 µl of the PCR products were separated on 2% agarose gels and stained with ethidium bromide. Gels were visualized under UV light (312 nm) and images were taken using a video documentation system (Gel Doc 2000; Bio-Rad, Munich, Germany). Signal intensities were quantified using the program package nih image (version 1.62; http://rsb.info.nih.gov/nih-image). All PCRs were replicated at least three times using first strand cDNA of at least two independent RNA preparations.

To quantify differences in gene expression of a given gene in two different samples, dilution series of the first strand cDNA of the sample revealing the higher transcript level were prepared and PCR reactions were performed together with the undiluted second sample using gene-specific primers. The dilution that revealed the same signal intensity as the undiluted sample was then used to calculate the expression ratio.

Real-time PCR

First-strand cDNA synthesis was performed as described for RT-PCR. Real-time PCR was performed using 10 µl Q-PCR-Mastermix (containing Sybr green and fluorescein; ABgene, Hamburg, Germany), 0.5 µl cDNA, and 10 pmol of each primer in a MyiQ Real time PCR system (Bio-Rad, Hercules, CA, USA). Specific primers for 18S rRNA, actin primers (Langer et al., 2004) and the nearly constitutively expressed poplar gene PttJip1 (Grunze et al., 2004) were used as references. The PCR was always performed in triplicates together with dilution series of the reference genes. Three different cDNA synthesis reactions of at least two different biological replicates were used for analysis.

Primers used for analysis (the reverse primer was chosen to target the 3′-UTR of the genes) were

  • PoptrAMT1.1a: 5′-GTTCTTCATAAGCTTAAGC-3′; 5′-CAAGCATATTGAAGCACC-3′

  • PoptrAMT1.1b: 5′-TTCGCAAGCTCAAGCTC-3′; 5′-CATACAACATTTATCGAC-3′

  • PoptrAMT1.2a: 5′-GATGAGATGGCAGGCATG-3′; 5′-GTGGTGAGTTTCGATGAGG-3′

  • PoptrAMT1.2b: 5′-CGTGCCTTGTCAAATCC-3′; 5′-GTTTGCGAAGGAGACG-3′

  • PoptrAMT1.3: 5′-CATCCTCGCTTCTACG-3′; 5′-CCAAAGCGCCAATGATAC-3′

  • PoptrAMT1.4a: 5′-CAAGGCTCTTAGCTGCAC-3′; 5′-CTATCCATTTCCCACTGC-3′

  • PoptrAMT1.4b: 5′-GACTGCTTATGGGAGGTG-3′; 5′-TGCAGGGCTAAGACAACG-3′

Heterologous expression of PttAMT1.2 in a yeast mutant

An EcoRI/XhoI digested cDNA fragment containing the entire PttAMT1.2 coding region was inserted into the EcoRI/XhoI-digested yeast expression vector pDR196 (Rentsch et al., 1995). The Saccharomyces cerevisiae triple mep mutant 31019b (Marini et al., 1997) was transformed with the PttAMT1.2 expression construct according to Gietz & Woods (2002). Uptake experiments of 14C-labelled methylamine and competition experiments with NH4+ were performed according to von Wiren et al. (2000). The Km and Ki values were calculated using the hyper-software (John Easterby's Software, http://www.liv.ac.uk/~jse).

Construction of the phylogenetic tree

The protein alignment was constructed using clustalx (Thompson et al., 1997). Ambiguous alignment positions were excluded from the phylogenetic analysis. To estimate phylogenetic relationships, the alignment was analysed using a Bayesian approach based on Markov chain Monte Carlo (MCMC), as implemented in the computer program mrbayes 3.1 (Ronquist & Huelsenbeck, 2003). This approach allows one to estimate the posterior probabilities that sequence groups are present in the true tree, given the sequence alignment.

We ran two independent MCMC analyses, each involving four incrementally heated chains over two million generations, starting from random trees and assuming a percentage of invariable alignment sites with γ-distributed substitution rates of the remaining sites. Rather than specifying an amino acid substitution model we allowed the Markov processes to sample randomly from the substitution models implemented in mrbayes. Trees were sampled every 100 generations resulting in an overall sampling of 20 000 trees per run, from which the first 4000 trees of each run were discarded (burn in). The remaining 12 000 trees sampled in each run were pooled and used to compute a majority rule consensus tree to get estimates for the posterior probabilities. Branch lengths were averaged over the sampled trees. Stationarity of the process was controlled using the tracer software (Rambaut & Drummond, 2003), version 1.2.1.

Miscellaneous

DNA was isolated from gel pieces using the NucleoSpin gel extraction kit (Macherey and Nagel, Düren, Germany). DNA fragments were cloned into the pCR.2.1-TOPO vector (Invitrogen) and used for transformation of One-shot competent E. coli (Invitrogen). Overlapping sequencing was performed using M13 universal and reverse primers (Stratagene, Heidelberg, Germany) as well as gene specific primers (Invitrogen) and the ABI PRISM Dye Terminator Cycle Sequencing Ready Reaction Kit (Applied Biosystems, Foster City, CA, USA) on an automated sequencer ABI 3100 (Applied Biosystems) according to the manufacturer's instructions.

For analysis of DNA and protein sequences, the program package genejockey ii (1998, P. I. Taylor, Cambridge, UK) was used. The DNA sequence information was compared with publicly available sequence information using blastx (NCBI, http://www.ncbi.nlm.nih.gov/blast; Altschul et al., 1997) and further analysed using clustalx (1.8, NCBI; Thompson et al. 1997).

Nitrate and ammonium contents in soil were determined according to Brettar et al. (2002).

Results

Phylogenetic relationship of PttAMT1.2

A cDNA (PttAMT1.2, Accession No. AJ646889), coding for a high-affinity ammonium transporter, was obtained from an expressed sequences tag (EST) project (U. Nehls, unpublished) using fully developed P. tremula × tremuloides/A. muscaria ectomycorrhizas (Nehls et al., 2001) as a source. The predicted protein has a length of 507 amino acids with a calculated molecular mass of 54 043 Da, revealing the highest similarity to AMT1.2 proteins of Arabidopsis thaliana and Lycopersicon esculentum (80% and 72% identity, respectively).

To compare the phylogenetic relationships of the deduced PttAMT1.2 protein with other poplar ammonium transporters, the genome sequence (Brunner et al., 2004) of P. trichocarpa (v1.0; available at http://genome.jgi-psf.org/Poptr1/Poptr1.home.html) was screened for homologous sequences using blastn and tblastn. Members of two gene families (AMT1 and AMT2, Sohlenkamp et al., 2002) encoding putative high-affinity ammonium transporters were identified in the genome. In this contribution we focused on members of the AMT1 gene family.

For phylogenetic analysis, the deduced protein sequences of PttAMT1.2 (P. tremula × tremuloides) and all P. trichocarpa genes of the AMT1 gene family as well as all known A. thaliana and L. esculentum sequences were used (Fig. 1).

Figure 1.

Phylogenetic relationships between the deduced protein sequences of PttAMT1.2 (Populus tremula × tremuloides) and all Populus trichocarpa, Arabidopsis thaliana and Lycopersicon esculentum ammonium transporters. Majority-rule consensus of 32 000 trees from two independent Markov chain Monte Carlo (MCMC) analyses, sampled after stationarity had been reached. In each of the two analyses four incrementally heated chains were run over 2 million generations, starting from random trees and assuming a percentage of invariable alignment sites with γ-distributed substitution rates of the remaining sites. Branch lengths were averaged over the sampled trees and are scaled in terms of expected numbers of amino acid changes per site. Decimal numbers are estimates for the posterior probabilities that the corresponding bipartition is correct given the sequence data. For details see the text. GenBank Accession Nos. of genes used for tree formation: AtAMT1.1, NC003075; AtAMT1.2, AF083036; AtAMT1.3, AF083035; AtAMT1.4, T0452; AtAMT1.5, AP000382; LeAMT1.1, X92854; LeAMT1.2, 3023281; LeAMT1.3, AF118858; BnAMT1.3, AF306518. Populus trichocarpa sequences according to http://genome.jgi-psf.org: PoptrAMT1.1a, eugene3.00081611; PoptrAMT1.1b, fgenesh1_pg.C_LG_X000434; PoptrAMT1.2a, eugene3.00130493; PoptrAMT1.2b, grail3.0085006901; PoptrAMT1.3, fgenesh1_pg.C_LG_IX001127; PoptrAMT1.4a, grail3.0021033901; PoptrAMT1.4b, fgenesh1_pg.C_LG_II002592.

PttAMT1.2 clusters together with two P. trichocarpa homologues (PoptrAMT1.2a and PoptrAMT1.2b) and AtAMT1.2 from Arabidopsis. The sequences of this cluster are related to LeAMT1.2 from L. esculentum. The AMT1.2 subfamily does not appear as a supported group and includes also PoptrAMT1.3 from P. trichocarpa and LeAMT1.3 from Lycopersicon, respectively. However, the last two sequences are united by a long branch, and their inclusion in the AMT1.2 cluster is not well supported. The genes of the AMT1.1 subfamily cluster together but, significantly, include a well-supported group containing BnAMT1.3 from Brassica napus and AtAMT1.3 and AtAMT1.5 from Arabidopsis. The three genes of the AMT1.4 subfamily, PoptrAMT1.4a, PoptrAMT1.4b from Populus trichocarpa and AtAMT1.4 from Arabidopsis, form a well-supported cluster.

PttAMT1.2 encodes a high-affinity ammonium importer

In order to investigate whether the PttAMT1.2 gene encodes a functional ammonium importer, the entire coding region of the cDNA was cloned into the yeast expression vector pDR196 (Rentsch et al., 1995). The S. cerevisiae triple mep mutant 31019b (Marini et al., 1997) was transformed with the PttAMT1.2 expression construct. Only transformants expressing the PttAMT1.2 cDNA in sense orientation were able to grow on agar plates containing 0.5 mm ammonium as sole N source.

As ammonium uptake of PttAMT1.2 could not be determined directly in yeast, the kinetic properties of the transporter were investigated by uptake experiments using radioactively labelled methylamine (Fig. 2a). A KM value of 83.1 ± 5 µm was determined for this artificial substrate. Inhibition of methylamine uptake by ammonium addition revealed a Ki of 52 ± 4 µM, revealing that PttAMT1.2 has a similar affinity to both methylamine and ammonium and encodes a high-affinity ammonium importer (Fig. 2b).

Figure 2.

Functional analysis of PttAMT1.2 by heterologous expression in yeast. The entire coding region of PttAMT1.2 was expressed in a yeast mutant defective in high-affinity ammonium import. Yeast cells were transformed either with the expression vector alone or the vector containing PttAMT1.2. (a) Yeast cells were incubated with increasing methylamine concentrations (10–175 µm) containing the same specific activity (40 kBq [14C]-methylamine µmol−1). The rate of methylamine uptake (pmol min−1) was calculated from the radioactivity incorporated by yeast cells expressing PttAMT1.2 (squares) or containing only the empty expression vector (triangles) and used to calculate the KM value. (b) Yeast cells expressing PttAMT1.2 were incubated with 100 µm methylamine and increasing concentrations (0–250 µm) of ammonium sulphate. The inhibition of methylamine uptake (%) was measured and used to calculate the Ki value for ammonium from the Hanes plot (insert).

Expression profile of PttAMT1.2 in P. tremula ×tremuloides organs

The expression of PttAMT1.2 was investigated in expanding leaves (about half the size of mature leaves), mature leaves, stems, lateral roots (second-order without fine roots) and fine roots of 6-month-old P. tremula × tremuloides plants grown in the greenhouse (Fig. 3) as well as 3-month-old plants grown in a Petri dish system (data not shown) by semiquantitative RT-PCR, resulting in comparable expression profiles. While the transcript level of PttAMT1.2 in lateral roots was twice that of fine roots, gene expression was hardly detectable in leaves and stems of poplar plants.

Figure 3.

Expression profile of PttAMT1.2 in organs of poplar plants grown in pot culture in a greenhouse. Total RNA was isolated around midday from different organs of Populus tremula × tremuloides plants grown for 6 months in pot culture in a greenhouse. The RNA was used for semiquantitative reverse transcriptase polymerase chain reaction using PttAMT1.2-specific primers. The signal intensities of a constitutively expressed poplar actin gene (PtACT2; Langer et al. 2004) as well as PttJip1 (Grunze et al., 2004; data not shown) and 18S rRNA (data not shown) were used as references.

Impact of nitrogen nutrition on PttAMT1.2 expression

To investigate the impact of N nutrition on the expression of the root-specific PttAMT1.2 gene, 3-month-old poplar plants were pregrown in Petri dishes containing 400 µm ammonium as N source. The plants were transferred into aerated liquid medium containing different ammonium or nitrate concentrations and further incubated for up to 4 d (with a medium exchange twice a day). Where mentioned, N starvation was performed for 3 d before the nutrition experiment.

PttAMT1.2 expression was affected by N nutrition (Fig. 4). Ammonium concentrations between 10 µm and 100 µm resulted in a fourfold higher transcript level compared with a concentration of 2 mm N (ammonium or nitrate) in the growth medium.

Figure 4.

Impact of nitrogen source and concentration on PttAMT1.2 expression. Roots of 3-month-old Populus tremula × tremuloides plants were incubated in aerated liquid MMN (Hampp et al., 1996) medium for 4 d (harvest around noontime) containing different ammonium or nitrate concentrations: (1) 10 µm; (2) 100 µm; (3) 500 µm; (4) 2 mm. Isolated total RNA was used for Northern blot hybridization using the 3′-UTR (untranslated region) of the PttAMT1.2 cDNA. 28S/18S rRNA was used as loading control.

The addition of 100 µm ammonium to N-starved poplar plants also resulted in a fourfold increase of PttAMT1.2 expression within 6 h (Fig. 5). Gene expression remained stable during the investigated period and only a slight diurnal regulation of the transcript level (twofold lower expression during the night compared with daytime) could be observed.

Figure 5.

Kinetic of PttAMT1.2 expression with respect to nitrogen (N) induction. Roots of 3-month-old Populus tremula × tremuloides plants, pregrown for 3 d in MMN (Hampp et al., 1996) medium in the absence of any N source, were incubated in aerated liquid MMN medium containing 100 µm ammonium for up to 2 d. The expression rate of PttAMT1.2 was determined by semiquantitative reverse transcriptase polymerase chain reaction using gene specific primers. The signal intensities of a constitutively expressed poplar actin gene (PtACT2; Langer et al. 2004) as well as PttJip1 (Grunze et al., 2004, data not shown) and 18S rRNA (data not shown) were used as references. Harvest after 13 h was at night while all other samples were collected during daytime (with illumination).

Ectomycorrhiza formation and PttAMT1.2 expression

Two different sources of mycorrhizas were used for this study; mycorrhizas obtained under sterile conditions in a Petri dish system (4-month-old P. tremula × tremuloides or P. trichocarpa) and mycorrhizas from a field (2-yr-old P. tremula ×tremuloides; P. trichocarpa plants are not available in nature in Europe).

Mycorrhizas obtained from the Petri dish system were exclusively formed with A. muscaria as partner, a fungus that, under natural conditions, is mainly found in well-established forest ecosystems.

The field sample contained mycorrhizas formed with several different ectomycorrhizal fungi, typical for natural conditions. However, c. 70% of the isolated mycorrhizas from the field contained a Hebeloma species, which are common ectomycorrhizal basidiomycetes present in young plantations as well as in established forest ecosystems. The determination of the fungal partners present in isolated mycorrhizas (data not shown) was performed by morphotyping according to Agerer (1987–93) and sequencing of DNA fragments obtained from isolated genomic DNA of single mycorrhizas amplified by internal transcribed spacer (ITS) primers (Carnero Diaz et al., 1997).

Another difference between plants obtained from the Petri dish system and the plantation was the available N source. Ammonium (200 µm or 3,8 mg kg−1) was the sole N source used in the Petri dish system, while soil samples obtained from the field revealed an ammonium content of 2.9 mg kg−1 soil and a nitrate content of 4.25 mg kg−1.

In both the Petri dish and the field experiment mycorrhiza formation of P. tremula × tremuloides resulted in a fourfold increase of PttAMT1.2 expression (Fig. 6). This enhanced gene expression in mycorrhizas was independent of N nutrition, since at the given ammonium concentration, PttAMT1.2 expression is maximal in nonmycorrhized fine roots.

Figure 6.

Impact of ectomycorrhiza formation on the expression of PttAMT1.2. Total RNA was isolated from (last order) fine roots and mycorrhizas of 3-month-old Populus tremula × tremuloides plants grown with and without Amanita muscaria in a Petri dish system (containing ammonium as sole nitrogen source) and from 2-yr-old, field-grown plants. The expression of PttAMT1.2 was determined by semiquantitative reverse transcriptase polymerase chain reaction using gene-specific primers. The signal intensities of a constitutively expressed poplar actin gene (PtACT2; Langer et al. 2004) as well as PttJip1 (Grunze et al., 2004; data not shown) and 18S rRNA (data not shown) were used as references.

Expression profiles of AMT1 genes in P. trichocarpa and impact of ectomycorrhiza formation

To get a broader overview on the impact of ectomycorrhiza formation on the expression of poplar ammonium importer genes, specific primer pairs for all seven members of the AMT1 gene family of P. trichocarpa were designed.

Real-time PCR was performed to investigate the gene expression patterns in different organs of P. trichocarpa (Fig. 7).

Figure 7.

Expression profiles of Populus trichocarpa ammonium transporter genes in organs of plants grown in pot culture in a greenhouse. Total RNA was isolated around midday from different organs (see key within figure) of P. trichocarpa plants grown in pot culture for 6 months in a greenhouse. The RNA was used for real time polymerase chain reaction using gene-specific primers. The signal intensities were calibrated according to a constitutively expressed poplar actin gene (Langer et al., 2004) and 18S rRNA.

The transcript levels of two genes, PoptrAMT1.2a and PoptrAMT1.4b, were below the detection limit. Like its closest homologue PttAMT1.2 (P. tremula × tremuloides), PoptrAMT1.2b revealed a root-specific expression pattern, showing a twofold stronger expression in lateral compared with fine roots.

PoptrAMT1.1a and PoptrAMT1.3 were mainly expressed in leaves, revealing a comparable transcript level in expanding and mature leaves. PoptrAMT1.3 expression was 22-fold lower in stems and 11-fold lower in roots. PoptrAMT1.1a revealed a 2.5-times lower expression in stems and a sixfold lower expression in roots. PoptrAMT1.1b and PoptrAMT1.4a were expressed at a comparable level in most poplar organs investigated. While the transcript level of PoptrAMT1.4a was threefold lower in fine roots, PoptrAMT1.1b expression was about twofold higher in main roots and 1.5 times higher in mature leaves.

With respect to fine roots, a comparable high transcript level was obtained for PoptrAMT1.2b, PoptrAMT1.3 and PoptrAMT1.4a while PoptrAMT1.1b expression was 1.5 times lower and PoptrAMT1.1a was more than five-times lower.

Mycorrhiza formation resulted in a ninefold increase for the root-specific PoptrAMT1.2b gene, a fivefold increased transcript level for PoptrAMT1.3 and a sevenfold increase for PoptrAMT1.4a (Fig. 8). The expression levels of PoptrAMT1.1a and PoptrAMT1.1b were slightly (twofold) reduced in mycorrhized plants.

Figure 8.

Impact of ectomycorrhiza formation on the expression of Populus trichocarpa ammonium transporter genes in fine roots. Total RNA was isolated from (last order) fine roots and mycorrhizas of 3-month-old P. trichocarpa plants grown with and without Amanita muscaria in a Petri dish system (containing ammonium as sole nitrogen source). The RNA was used for real time polymerase chain reaction using gene specific primers. The signal intensities were calibrated according to a constitutively expressed poplar actin gene (Langer et al., 2004) and 18S rRNA.

Discussion

To enable host plant nutrition in symbiosis, soil-growing hyphae of ectomycorrhizal fungi take up N (in organic or inorganic form), assimilate it as amino acids and transfer part of this N towards the plant partner. However, mycorrhized root tips themselves are also capable of taking up N. Nevertheless, with the exception of nitrate (Ek et al., 1994), all N sources investigated to date were taken up and assimilated by fungal hyphae before their transfer towards the plant (Melin & Nilsson, 1952; Finlay et al., 1989; Rygiewicz & Andersen, 1994; Wallenda & Read, 1999). While N uptake by fungal hyphae has been investigated extensively (for reviews see Javelle et al., 2004; Nehls, 2004), the chemical nature of the N excreted by fungal hyphae at the plant/fungus interface is still a matter of debate.

Based on three lines of evidence it is commonly believed that amino acids are transferred from the fungus to the plant in the Hartig net (Smith & Read, 1997; Chalot & Brun, 1998; Javelle et al., 2004). First, ectomycorrhizal fungi excrete amino acids under certain conditions (France & Reid, 1983; Plassard et al., 1993). Second, tree roots are able to take up amino acids (Persson & Nasholm, 2001). Third, when glycine (double-labelled with 15N and 13C) was applied to the humus layer, double labelled amino acids were found in mycorrhized plants (Näsholm et al., 1998).

However, other investigations question the general role of amino acids as a plant N source in symbiosis for several reasons: First, the efflux of the amino acid analogue aminoisobutyric acid (the pH effect on amino acid efflux was never investigated) is minimal at low pH values (Chalot & Brun, 1998). Since the apoplastic pH of the Hartig net is rather low (3.1–4; Nenninger & Heyser, 1998), a strongly reduced amino acid excretion capacity of fungal hyphae at the plant/fungus interface can be supposed. Second, in a series of reports Hodge et al. (1998, 1999, 2000) studied plant and microbial N acquisition using different double (13C, 15N)-labelled N sources (i.e. lysine, urea; algal amino acid mixture; lyophilized algal cells). Unlike the results of Näsholm et al. (1998) enrichment in 15N but not in 13C was recorded in plant tissues, indicating that organic N sources are usually metabolized by microbes before their transfer to the plant. These results may thus indicate that the observed presence of double-labelled glycine in mycorrhized plants (Näsholm et al., 1998) might be the result of excess N during glycine feeding, resulting in direct amino acid uptake (via the apoplast of the fungal sheath) by the plants without fungal metabolic participation. A similar pathway was been shown for nitrate uptake by mycorrhized plants (Ek et al., 1994). Third, in plant–bacterial interactions (e.g. rhizobia, Tate et al., 1998; free living N2-fixing bacteria of the rhizosphere, Becker et al., 2002) mainly ammonium (but not amino acids) is excreted by symbiotic bacteria and high-affinity plant ammonium importers are induced (similar to what we observed in poplar ectomycorrhizas).

These results, together with the strongly increased ammonium uptake capacity of mycorrhized poplar roots (this contribution), indicate that ammonium could be a major source of fungus-derived N in symbiosis. This hypothesis is further corroborated by the finding that ammonium is exported by intraradical hyphae in arbuscular mycorrhizal symbiosis (Govindarajulu et al., 2005).

How could a net export of ammonium by fungal hyphae of the Hartig net happen? First, ammonium has to be provided through enzymatic degradation presumably of amino acids in fungal hyphae at the plant/fungus interface. Second, ammonium must permeate the plasma membrane towards the common apoplast. One possibility is the diffusion of NH3 across the plasma membrane (Yip & Kurtz, 1995). The greatly enlarged surface area of the fungal hyphae at the Hartig net (Kottke & Oberwinkler, 1987) would favour this NH3 leakage from hyphae in symbiosis. Upon entering the apoplast, the low pH (Nenninger & Heyser, 1998) would immediately lead to protonation of NH3, preventing a return by diffusion. However, whether NH3 leakage is sufficiently fast to enable plant N nutrition is questionable (Burgstaller, 1997).

In addition to NH3 leakage, ammonium efflux proteins, as described for S. cerevisiae (Guaragnella & Butow 2003), could be responsible for fungal ammonium export at the plant–fungus interface. Genes encoding proteins, that are homologous to supposed ammonium export proteins of S. cerevisiae are expressed in the ectomycorrhizal fungus A. muscaria (Accession Nos. AJ644726 and AJ644793; U. Nehls, unpublished). In addition, certain aquaporins have been reported to be capable of ammonium transport (Jahn et al., 2004) and could thus mediate ammonium leakage out of fungal hyphae as well.

After entering the common apoplast of the plant/fungus interface, the re-import of ammonium into fungal hyphae is prevented since the expression of fungal ammonium importer genes is repressed in a N-dependent manner (Javelle et al., 2003; A. Selle, unpublished). Finally, the strongly increased uptake capacity of plant root cells (enabled in poplar by PttAMT1.2/PoptrAMT1.2b, PoptrAMT1.4a and PoptrAMT1.3) generates a sink that favours further ammonium leakage out of the fungal hyphae.

In summary, our results indicate that ammonium (perhaps in addition to amino acids) could be a major source of fungus-derived N in symbiosis. Investigations of fungal genes involved in amino acid degradation and ammonium assimilation in mycorrhizas are in progress to further prove our working model.

Acknowledgements

We are indebted to Margret Ecke, Andrea Bock and Susanne Schwager for excellent assistance. This work was financed by the Deutsche Forschungsgemeinschaft (Ne 332/4-2 and Ne 332/9).

Ancillary