• Ammonium and nitrate are the prevalent nitrogen sources for growth and development of higher plants. Here, we report on the characterization of the ammonium transporter (AMT) family in the perennial species Populus trichocarpa.
• In silico analysis and expression analysis of AMT genes from poplar was performed. In addition, AMT1;2 and AMT1;6 function was studied in detail by heterologous expression in yeast.
• The P. trichocarpa genome contains 14 putative AMTs, which is more than twice the number of AMTs in Arabidopsis. In roots, the high-affinity AMT1;2 strongly increased upon mycorrhiza formation and might be partly responsible for the high-affinity ammonium uptake component measured in poplar. Transcript level for the high-affinity AMT1;6 was strongly affected by the diurnal cycle. AMT3;1 was exclusively expressed in senescing poplar leaves. Remarkably AMT2;1 was highly expressed in leaves while AMT2;2 was mostly expressed in petioles. Specific expression of AMT1;5 in stamen and of AMT1;6 in female flower indicate that they have key functions in reproductive organ development in poplar.
• The present study provides basic genomic and transcriptomic information for the poplar AMT family and will pave the way for deciphering the precise role of AMTs in poplar physiology.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Ammonium is a primary source of nitrogen (N) for both perennial and annual species. It is taken up from the soil by ammonium transporters through the plasma membrane of root cells (Kaiser et al., 2002) and incorporated into glutamine via glutamine synthetase (GS) present in the cytoplasm and plastids, the biochemistry of which has also been described in perennial species (Suarez et al., 2002). Ammonium is also produced within plants either by reduction of nitrate and nitrite obtained from the soil, or by catabolism of endogenous amino compounds. For example, the photorespiratory nitrogen cycle generates a large amount of ammonium in leaf mitochondrias that is subsequently transported to chloroplasts for reassimilation by GS, although this statement has now to be interpreted with caution, given the recent finding that the nuclear encoded GS GLN2 may also be addressed to the mitochondria (Taira et al., 2004).
The biochemistry and molecular biology of ammonium transport in plants has been extensively studied and recent comprehensive reviews are available (Schjoerring et al., 2002; Loque & von Wiren, 2004). Physiological studies on ammonium uptake by plant roots has provided evidence for the existence of a low affinity nonsaturable transport system (LATS), which operates in the millimolar concentration range and a high-affinity transport system (HATS), which operates in the submillimolar concentration range. The HATS exhibits saturation kinetics, energy dependence, and leads to depolarization of the plasma membrane electrical potential (Ludewig et al., 2002).
More recently, a second subfamily of ammonium transporters (AMT2) with distinct biochemical features, has been identified in several plants such as A. thaliana (Sohlenkamp et al., 2002), L. japonicus (Simon-Rosin et al., 2003) and O. sativa (Suenaga et al., 2003). Plant AMT2 family members are more closely related to ammonium transporters from prokaryotes than they are to plant AMT1. While plant members of the AMT1 subfamily are preferentially expressed in roots (with the exception of LeAMT1;3 (von Wiren et al., 2000b), LjAMT1;1 and LjAMT1;2 (D’Apuzzo et al., 2004), AtAMT2 showed a higher level of gene expression in shoots compared with roots (Sohlenkamp et al., 2002) and OsAMT2;1 presented a constitutive expression in shoots and roots (Suenaga et al., 2003). These distinct expression patterns may support the fact that individual members of the AMT family would function not only in ammonium uptake in roots, but also in ammonium recycling during leaf senescence or photorespiration (Howitt & Udvardi, 2000; von Wiren et al., 2000a). Multiple forms of ammonium transporters in higher plants allow a greater regulatory flexibility and organelle-, cell-, tissue- or organ-specialization, and enable cells to take up ammonium over a wide range of concentrations (D’Apuzzo et al., 2004).
As exemplified above, most studies have focused on a few annual species (A. thaliana, L. esculentum, L. japonicus and O. sativa), thus available information concerning N transport in perennial woody models is much more limited, particularly at the molecular level. The perennial life style of trees implies specific physiological traits that have been recently discussed (Suarez et al., 2002; Bhalerao et al., 2003; Andersson et al., 2004; Sterky et al., 2004). Related to N nutrition, one of these traits is the mobilization and storage in perennial tissues of the N present in leaves during autumn, which is remobilized at the beginning of the next growing season (Suarez et al., 2002). Thus, transport mechanisms for nitrogenous compounds are likely to be different from those used by annual species, and probably rely on perennial-specific transport functions.
Taking advantage of the shotgun sequenced Populus trichocarpa (Nisqually 1) genome (Tuskan et al., 2006), we present here the expression analysis of AMT members from the perennial tree species poplar and extend the analysis of five AMT members to include the functional characterization of three AMT1 and two AMT2 members by heterologous expression in yeast.
Materials and Methods
Plant growth and media composition
Populus tremula × alba (clone INRA 717 1B4) cuttings were cultivated as described by Kohler et al. (2003) in a growth chamber for 6 wk with a light intensity of 80 µmol m−2 s−1 and a day–night temperature regime of 16 h at 24°C at 80% relative humidity and 8 h at 18°C at 60% relative humidity. For further experiments (N regime, light regime), P. tremula× alba plants were grown hydroponically for 2 months in the following nutrient solution: 0.75 mm CaCl2·2H2O, 4.7 mm KNO3, 5.15 mm NH4NO3, 0.375 mm MgSO4·7H2O, 0.3 mm KH2PO4, 100 µm H3BO3, 5 µm KI, 100 µm MnSO4 H2O, 30 µm ZnSO4 H2O, 1 µm Na2Mo4·2H2O, 0.1 µm CuSO4·5H2O, 0.1 µm CoCl2, 100 µm FeSO4·7H2O. The pH was adjusted to 5.6–5.8 with KOH. The nutrient solution was renewed every week. When needed, plants were then transferred to a N-free nutrient solution. To maintain K+ at a constant concentration KCl (2.36 mm) and K2SO4 (1.17 mm) were substituted for KNO3.
All experiments were carried out using the P. tremula × alba poplar clone, except when specific tissues were examined. Thus, fruits (harvested 21 April 2004), buds, stamens and female flowers (harvested 29 March 2005) were collected on 25-yr-old trees (Populus trichocarpa species) grown outdoors under natural conditions at the University campus.
Populus tremula × alba (clone INRA 717 1B4) cuttings were cultivated in a growth chamber for 2 wk as already described and then transferred onto Petri dishes, coinoculated with the ectomycorrhizal fungus Paxillus involutus (Batsch) Fr, in a modified MS solution (Murashige & Skoog, 1962), supplemented with 0.1% glucose and 1.5% agar. Paxillus involutus was pregrown for 10 d on agar medium containing modified Melin-Norkrans medium, as described by Selle et al. (2005). Mycorrhizal and nonmycorrhizal control roots were harvested 1 month after contact.
Cloning of AMTs cDNAs, yeast transformation and [14C]MA uptake
Predicting coding sequences corresponding to AMT1;1, AMT1;2, AMT1;3, AMT1;4, AMT1;5, AMT1;6, AMT2;1, AMT2;2, AMT3;1 and AMT4;5 were amplified by polymerase chain reaction (PCR) from cDNA generated from same tissues used for reverse transcriptase polymerase chain reaction (RT-PCR) studies. The fidelity of PCR amplifications was verified by sequencing. The amplification products were cloned into pGEM-T Easy vector (Promega, Madison, WI, USA) and, after NotI digestion, inserted into the yeast expression vector pFL61 (Minet et al., 1992). The yeast strain 31019b (MATa mep1Δ mep2Δ::LEU2 mep3Δ::KanMX2 ura3; (Marini et al., 1997)) was transformed with pFL61 harboring the cDNA sequences of AMTs. Yeast transformants were selected on N-free minimal medium supplemented with 1 mm glutamine and 3% glucose. Yeast complementation tests were performed as described previously (Javelle et al., 2003). Methylammonium uptake assays were conducted under previously described experimental conditions (Gazzarrini et al., 1999; Montanini et al., 2002).
Total RNA extraction was performed with the RNeasy Plant Mini kit (Qiagen, Darmstadt, Germany) from approximately 100 mg of frozen tissues of poplar. To remove contaminating genomic DNA, the samples were treated with DNAse I (Qiagen). To obtain cDNA, 1 µg of total RNA was annealed to oligo-dT primers (Promega) and reverse transcribed using Reverse Transcriptase (Qiagen) at 37°C for 1 h. Every reaction was set up in three replicates. For each AMT, the PCR program was as follows: 94°C for 3 min and 37 cycles of 94°C for 30 s, 58°C for 45 s, and 72°C for 1 min 45 s. AMT fragments were gel-purified and sequenced to ensure accuracy and specificity. The whole set of AMT genes (14 genes) was tested by RT-PCR in every experiment carried out but only AMTs detected are retained in figures for greater clarity. A cDNA fragment corresponding to the constitutively expressed ubiquitin gene was amplified simultaneously (28 cycles) and used as a control. Cysteine protease (CP) gene was amplified (35 cycles) and used as control of the senescing state of leaves. The sequences of the gene-specific oligonucleotides designed in the nonconserved 5′ and 3′ regions of the genes were used for RT-PCR and are listed in Table 1.
Table 1. Primers used for reverse transcriptase polymerase chain reaction analysis
Amino acid extraction and analysis
Amino acids were extracted twice from 10 to 20 mg dried plant tissues with 300 µL 70% (v : v) cold ethanol. The samples were dried using a Reacti-Therm. Heating Module (Pierce, Rockford, IL, USA) and resuspended in 400 µl 0.1 n HCl. Amino acids and standards were then purified on a Dowex 50WX-8 cation ion exchange column (Sigma-Aldrich, St Louis, MO, USA) and aliquots of purified samples were transferred to microvials, dried in a Reacti-Therm Heating Module (Pierce) and derivatized according to Javelle et al. (2003). Gas chromatography and mass spectrometry (GC-MS) analysis was performed as described previously (Javelle et al., 2003).
Soluble sugars extraction and analysis
Soluble sugars were extracted from approx. 100 mg of fresh plant tissues, with 1.1 ml 70% (v : v) cold methanol. The extracts were centrifuged to get supernatant for assay of soluble sugars by the anthrone solution colorimetric method (Miller, 1959). The sugar content was determined by measuring the absorbance at 620 nm. These values were then regressed against readings from a set of standard solutions of glucose.
Full-length amino acid sequences were aligned by clustalw and imported into the Molecular Evolutionary Genetics Analysis (mega) package version 3.1 (Kumar et al., 2004). Phylogenetic analyses were conducted using the neighbor-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. The accession numbers or gene models of version 1.1 of the P. trichocarpa genome database used were: PtrAMT1;1 (Poptr1_1 : 804848), PtrAMT1;2 (Poptr1_1 : 665333), PtrAMT1;3 (Poptr1_1 : 565016), PtrAMT1;4 (Poptr1_1 : 799507), PtrAMT1;5 (Poptr1_1 : 645545), PtrAMT1;6 (Poptr1_1 : 804509), PtrAMT2;1 (Poptr1_1 : 802015), PtrAMT2;2 (Poptr1_1 : 808726), PtrAMT3;1 (Poptr1_1 : 175190), PtrAMT4;1 (Poptr1_1 : 754260), PtrAMT4;2 (Poptr1_1 : 260884), PtrAMT4;3 (Poptr1_1 : 205629), PtrAMT4;4 (Poptr1_1 : 806873), PtrAMT4;5 (Poptr1_1 : 424130). Other accession numbers were: AtAMT1;1 (At4g13510), AtAMT1;2 (At1g64780), AtAMT1;3 (At3g24300), AtAMT1;4 (At4g28700), AtAMT1;5 (At3g24290), AtAMT2;1 (At2g38290), LeAMT1;1 (P58905), LeAMT1;2 (O04161), LeAMT1;3 (Q9FVN0), LjAMT1;1 (Q9FSH3), LjAMT1;2 (Q7Y1B9), LjAMT1;3 (Q70KK9), LjAMT2;1 (Q93 × 02), OsAMT1;1 (Os04g43070), OsAMT1;2 (Os02g40710), OsAMT1;3 (Os02g40730), OsAMT2;1 (Os05g39240), OsAMT2;2 (Os01g61550), OsAMT2;3 (Os01g61510), OsAMT3;1 (Os01g65000), OsAMT3;2 (Os02g34580), OsAMT3;3 (Os03g62200), OsAMT4;1 (Os12g01420). BnAMT1;2 (AAG28780), PttAMT1;2 (AJ646889).
Cloning and in silico analysis of the AMT family
The first whole-genome sequence and assembly for a tree species, namely that of P. trichocarpa was recently reported (Tuskan et al., 2006). The JGI P. trichocarpa gene search mode revealed the existence of 14 AMT gene models. As suggested by a phylogenetic tree based on protein multiple sequence alignment, six sequences from P. trichocarpa genome fell in the subgroup AMT1 and eight in the subgroup AMT2. Populus thus possesses a higher number of AMT genes in both subfamilies than Arabidopsis or rice. Populus AMT2 as well as rice AMT2 members further divided into three clades (Fig. 1).
Genes encoding proteins from different sub-branches markedly differ in their exon–intron structure: for example, the intron size and location of splicing sites are roughly conserved in genes in each AMT2 subclade, while AMT1 genes contain no intron with the exception of LjAMT1;1 (Salvemini et al., 2001).
Analysis of the assembled genome revealed relatively recent whole-genome duplication shared among all modern taxa in Salicaceae. A second, older duplication appears to be shared with the Arabidopsis lineage (Tuskan et al., 2006). These duplicated genes originated through very recent small-scale gene duplications and one relatively recent large-scale gene duplication event (Sterck et al., 2005). A detailed analysis of duplication events for the AMT members revealed that AMT1;1 and AMT1;4 derived from a common ancestor through an ancient duplication event, and that AMT2;2 and AMT2;1 derived from a common ancestor through a recent duplication event. Interestingly, AMT2;1 and AMT2;2 showed specific expression patterns, with AMT2;1 mostly expressed in leaves and AMT2;2 mostly expressed in petioles (Fig. 2).
Figure 1 presents tentative signature sequences for the proteins of the AMT1 and AMT2 plant subfamilies, based on the existing AMT signature (Saier, 1998). Logo representation of the AMT signature was constructed at the web interface program weblogo (Crooks et al., 2004) on the basis of the alignment of 39 plant AMT members. These sequences may serve as identification motifs specific for the two subfamilies, and when scanned against the GenBank database, retrieved only members of each subfamily. These sequences can be used to identify additional members of the two subfamilies in other plant species as they are being sequenced.
Analyses carried out with the transmembrane (TM) prediction program THMM (http://www.CBS.dk) suggest that the poplar AMT family proteins are likely to have 9–11 TM domains as do the other AMT plant members (Schwacke et al., 2003). In silico analysis accompanied by experiments using fusion proteins, indicate that the majority of the Mep/Amt proteins contain 11 membrane-spanning helices, with the N-terminus on the exterior face of the membrane and the C-terminus on the interior (Marini & Andre, 2000; Thomas et al., 2000).
Table 2 shows that the AMT gene and AMT protein sequences share a high degree of identity between Populus species, with nucleotide and protein identities of at least 97.7% and 96.9%, respectively. This facilitates in silico and RT-PCR analyses and comparisons across Populus species.
It is also worth noting that most AMT genes are sporadically distributed across the poplar genome, except AMT1;4, AMT1;5 and AMT4;1 which are located on linkage group (LG) II. The AMT1;4 locus stretches from position 23601578 to 23603065 while the AMT1;5 locus extends from position 23599283–23600819 on poplar LG II, thus appearing as tandem repeats (see Supplementary Material, Fig. S1).
All transcript levels were measured in all experiments described later. Expression levels for AMT1;3, AMT1;4, AMT4;1, AMT4;2, AMT4;3, and AMT4;4 were undetectable in any organs examined so far in poplar tissues. By contrast, the other AMT genes showed contrasting expression patterns as examined in the following text.
Effects of various N regimes
In order to investigate the impact of N regime on the regulation of AMT transcript levels, total RNAs and amino acids were extracted from roots and leaves of plants transferred in a N-free medium for various lengths of times, and further resupplied with N for 8 h. The concentration of total amino acids, measured by GC-MS, in roots and leaves decreased by five- and three-fold, respectively, during 9 d of N starvation (Fig. 3a). In N-fed poplar (grown for 2 months in N-supplied medium), asparagine was the predominant amino acid, accounting for 62% and 33% in roots and leaves, respectively, followed by glutamine in roots (Fig. 3b) and glutamate in leaves (Fig. 3c). The glutamine concentration decreased by 26-fold after 9 d of N-starvation in roots (Fig. 3b), whereas asparagine decreased by 25-fold in leaves (Fig. 3c). The amino acid pools were rapidly reconstituted after 8 h of N resupply in roots, while in leaves the asparagine levels did not completely reached the +N level.
AMT1;2 transcripts strongly increased in roots under a short period of N starvation, but declined after 9 d of N starvation and increased again upon N resupply (Fig. 4a). It is worth noting that expression of AMT1;2 was restricted to roots, whereas AMT1;1 transcripts were detected in shoot and root tissues (Fig. 2). AMT1;1 transcripts were undetectable under N-sufficient condition in roots, increased under N starvation, and remained weakly expressed during nitrogen resupply (Fig. 4a). In leaves, AMT1;1 was weakly expressed and poorly affected by N starvation (Fig. 4b). Similarly, AMT1;5 and AMT2;1 were only weakly expressed under these experimental conditions (data not shown). AMT1;6 transcripts declined under N starvation, compared with a N-sufficient growth condition, but remained low after N resupply (Fig. 4b), which might be linked to the poor increase in asparagine level at that time. All other AMTs were also analysed but were not detected under the conditions used in this experiment.
Effects of light regimes
Figure 5 shows diurnal variations in AMT expression levels in roots and leaves of plants collected at 4-h intervals in poplar. Our data indicate that AMT1;1, AMT1;2, AMT1;6 and AMT2;1 transcript levels exhibit diurnal rhythms, the largest diurnal changes in transcript levels being found for AMT1;2 in root (Fig. 5a) and AMT1;6 in leaf (Fig. 5b). AMT1;1 and AMT1;5 were only weakly expressed in leaves throughout the time course of the experiment (not shown). When light was switched off, the soluble sugar content in roots was at the highest level (98 mg equivalent glucose g−1 FW at 22 : 00 h) and decreased to 69 mg equivalent glucose g−1 FW after 8 h of dark regime. In leaves, the soluble sugar content was, respectively, 445 and 368 mg equivalent glucose g−1 FW at 22 : 00 h and 06 : 00 h. In addition, expression of a leaf asparagine synthetase gene (ASN2) was followed and showed a diurnal variation, similar to AMT1;6 transcript level, although at a lesser extent.
AMT expression during senescence
The expression of the AMTs was studied in young, mature and senescing leaves. The senescing state of the leaves was confirmed by the parallel amplification of a cysteine protease gene (Bhalerao et al., 2003; Andersson et al., 2004), which was detected only at the latter stage (Fig. 6). AMT1;5 and AMT1;6 showed increased expression levels in senescing leaves compared with young leaves. By contrast, AMT3;1 was exclusively expressed in senescing leaves (Fig. 6).
AMT expression in reproductive organs
Stamens exhibited a strong expression level for AMT1;5 transcripts, whereas female flowers showed expression for AMT1;6 (Fig. 7a). Amino acid pools in these tissues also reached the highest levels ever measured in poplar tissues, especially in stamens (Fig. 7b). We might therefore conclude that AMT1;5 and AMT1;6 are not under the control of N catabolic repression in these poplar tissues. This also indicates that another as yet unidentified mechanism allowed for full derepression of AMT1;5 and full repression of AMT1;6 in stamens, and the reverse in female flowers.
Effects of mycorrhizal inoculation
Figure 8 shows variations in AMT expression levels of roots inoculated with the ectomycorrhizal fungus Paxillus involutus, collected after 1 month of growth after contact. Our data indicate a substantial increase of the root AMT1;2 transcript level, whereas the other root expressed AMTs did not show any significant transcript variation (not shown).
In silico analysis of ESTs
An in silico analysis of poplar expressed sequence tags (ESTs) was carried out using the GenBank EST databases, from which 69 ESTs corresponding to AMTs from poplar could be retrieved. This analysis confirmed the broad distribution of AMT1;1 and AMT2;1 transcripts and the root distribution of AMT1;2 transcripts. The lack of transcript for AMT1;4 and for some AMT4 subclade members is also confirmed. Expressed sequences tags corresponding to AMT1;6 were mostly found in EST database from leaves (Table 3), which agreed with the fact that AMT1;6 was mostly expressed in shoot tissues (Figs 2, 3, 4, 5 and 6).
Table 3. In silico analyses of ammonium transporter (AMT) genes in poplars (Populus)
Number of occurrences in EST database
EST tissue localization
RT-PCR product (present work)
EST, expressed sequences tag; RT-PCR, reverse transcriptase polymerase chain reaction.
Number of occurrences and tissue localizations of AMT ESTs retrieved from the EST GenBank database (Altschul et al., 1997). Tissues were as follows: B, bud; Ba, bark; C, cambium; CC, cell culture; F, fruit; FB, floral bud; FC, female catkin; FF, female flower; L, leaf; P, petiole; R, root; S, stem; SL, senescing leaf; St, stamen; X, not determined.
3C, 11L, 2SL, 2S, 1Ba, 1R, 1X
R, L, F, FF, P, B
R, L, St
L, F, FF, P
1R, 2CC, 2FC, 1FB, 2SL, 1L
L, FF, P
4L, 1B, 2Ba
Although detected by RT-PCR, no EST was retrieved for AMT1;5 and AMT2;2, probably because there were no EST databases for the specific tissues, i.e. stamens (AMT1;5). Conversely, ESTs were found for AMT4;2 in floral bud databases, for AMT3;1 in poplar bark database (Ralph et al., 2006) and for AMT4;5 in poplar cultured cells. These are tissues that we have not investigated in the present study but this opens new perspectives for the search of AMT4 subgroup transcripts. For AMT3;1, transcripts of which were detected specifically in senescing leaves (Fig. 6), four ESTs were retrieved from a leaf EST database, although the development state was not provided in the database. AMT1;3 was not detected by RT-PCR, although we found five ESTs in leaf databases. However these studies were of P. euphratica (three ESTs (Brosche et al., 2005)) or systemically wound-induced leaf (two ESTs; Christopher et al., 2004) – specific conditions that were not tested in the present study.
Functional expression of AMTs in a yeast mutant defective in NH4+ uptake
Yeast has provided a genuine heterologous expression system for characterizing many plant nutrient and metabolite transporters. The yeast strain 31019b is defective in the three endogenous NH4+ transporters (mep1, mep2 and mep3) and is unable to grow on medium containing 1 mm NH4+ as the sole N source (Marini et al., 1997). Transformation with the yeast expression vector pFL61 bearing the AMT1;2, AMT1;5, AMT1;6, AMT2;1 or AMT2;2 coding sequences under the control of the constitutive yeast phosphoglycerate kinase gene promoter (Minet et al., 1992), conferred the ability of 31019b to grow at 1 mm NH4+ as the sole N source (Fig. 9). Yeast transformed with empty plasmid (pFL61) was used as a negative control. Thus, all five genes encode functional NH4+ transporters. By contrast, the other three-expressed poplar AMTs (AMT1;1, AMT3;1 and AMT4;5) were unable to complement the Mep deficiency in the yeast strain 31019b. This inability might result from poor targeting to the yeast plasma membrane. Two intron-free AMTs, for which transcripts were not detected (AMT1;3 and AMT1;4), were cloned using their genomic sequence and used to transform the triple Mep mutant, but were unable to complement the growth deficiency on low ammonium (data not shown). Cells expressing AMT1; 5 grew less than the control strain on 10 mm ammonium (Fig. 9), suggesting that AMT1;5 expression was toxic to yeast cells.
We further tested the sensitivity of yeast cells to increasing methylammonium concentrations (Fig. 10). Cells expressing AMT2;1, or AMT2;2 retained the ability to grow on methylammonium, therefore indicating that methylammonium was not taken up by these cells. As found for other AMT2 family members (Sohlenkamp et al., 2002), it seems that poplar AMT2;1, or AMT2;2 have a particularly low affinity for methylammonium, whereas AMT1 family members can readily take up methylammonium.
Kinetics studies of AMT1;2 and AMT1;6 transporters expressed in yeast
To determine differences in substrate affinities between the AMT proteins, we used 14C-labeled methylammonium as a substrate analog and measured short-term uptake in transformed yeast Δmep strain 31019b. Only members of the AMT1 subfamily showed methylamine uptake capacities (Sohlenkamp et al., 2002). Among the AMT1 members, expression of AMT1;2 and AMT1;6 conferred the ability to take up 14C-labeled methylammonium in the range of 0.03–3 mm whereas 31019b transformed with the vector alone (Fig. 11a) did not take up significant amounts of 14C-labeled methylammonium in the same concentration range. Kinetic parameters were calculated from Lineweaver–Burk plots. AMT1;2 displayed a higher affinity (Km = 45 µm) and a lower capacity (Vm = 0.349 ± 0.007 nmol min−1 108 cell−1) for methylammonium than AMT1; 6 (Km = 85 µm; Vm = 1.058 ± 0.024 nmol min−1 108 cell−1). Since the affinities for methylammonium do not necessarily reflect the affinity for ammonium, we performed inhibition studies with varying NH4+ concentrations. The Ki values were deduced from Dixon plots, and a 50% inhibition was observed at 11 and 16 µm for AMT1;2 and AMT1;6, respectively (Fig. 11b). Inhibition studies indicated that these two NH4+ transporters possess a much higher affinity for NH4+ than for methylammonium.
An extended AMT family in the poplar genome
Analysis of genomes of Arabidopsis and rice revealed the presence of six and 10 AMT genes, respectively. In the poplar genome, 14 AMT genes were identified, and assigned to the AMT1 subfamily (six genes) or to the AMT2 subfamily (eight genes). Interestingly, poplar possesses a much greater number of AMT2 genes than Arabidopsis and a greater number of AMT1 genes than rice. Although a few poplar AMT transcripts were not detectable in any tissue tested in this work, the expression patterns of the expressed members strongly suggest specific features related to the peculiar physiology of a perennial and mycorrhizal tree. Trees differ fundamentally from annual plant species in that they are adapted to survive on a long time-scale. The most obvious manifestation of this is the development of wood, or secondary xylem, from the vascular cambium. This secondary meristem is essential for tree growth and development and in providing support for a tall structure. Taken with caution, a comparison of the number of AMT genes in different species might suggest that plant species from different environments or with different life style organize ammonium transport with a fairly unequal number of ammonium transporters.
The root specific AMT1;2 is highly regulated by N availability and mycorrhization
AMT1;2 was specifically and highly expressed in roots (Figs 4 and 5). Determination of affinity constants for 14C methylammonium uptake by AMT1;2 in yeast and transport inhibition by NH4+ revealed that this protein is a high-affinity ammonium transporter (Km NH4+ = 11 µm), indicating that it might be partly responsible for the HATS ammonium uptake component measured in Populus species (Min et al., 2000). The downregulation of AMT transcript in N-fed P. tremula × alba root is consistent with the downregulation of HATS ammonium influx measured in P. tremuloides (Min et al., 2000). Indeed, AMT1;2 transcript levels in poplar (Fig. 4a) negatively correlated with root glutamine concentrations (Fig. 3b), as already demonstrated for AtAMT1;1 (Rawat et al., 1999) and LeAMT1;1 (von Wiren et al., 2000b). AMT1;2 seems to represent the functional ortholog of AtAMT1;1 and LeAMT1;1. However, prolonged incubation in a N-deficient growth medium resulted in a decrease of the AMT1;2 transcript level, probably owing to the appearance of proteolysis-derived ammonia. In addition, AMT1;2 was overexpressed in P. involutus mycorrhizal roots. Indeed, expression of its closest ortholog PttAMT1;2 from Populus tremula × tremuloides was increased fourfold in the ectomycorrhizal symbiosis between P. tremula× tremuloides and Amanita muscaria, suggesting an increased ammonium uptake capacity of mycorrhizal poplar roots (Selle et al., 2005). Recent studies on the arbuscular mycorrhizal fungus Glomus intraradices demonstrated that ammonium is exported by intraradical hyphae in arbuscular mycorrhizal symbiosis (Govindarajulu et al., 2005), suggesting that ammonium could be a major source of fungus-derived N in mycorrhizal symbiosis (Chalot et al., 2006).
In roots, four other AMT genes, AMT1;1 (Fig. 4), AMT1;5 (Table 3), AMT2;2 (Fig. 2) and AMT4;5 (Table 3) were expressed, reflecting a potential role in covering the N demand through uptake of soil NH4+ in various N regime. Although the kinetic parameters were not determined for this subset of genes, we might hypothesize that some of them could be involved in the LATS component of ammonium influx (Min et al., 2000), or in the retrieval of leaked ammonium (von Wiren et al., 2000a). However, they were also expressed in other poplar tissues such as petiole (AMT1;1 and AMT2;2) or stamen (AMT1;5) and therefore a specific function in root uptake can be ruled out.
The shoot-specific AMT1;6 and AMT3;1 are strongly upregulated during senescence
Several lines of evidence indicate that transport of NH4+, and hence leaf-expressed AMTs, might be of particular importance for N nutrition and metabolism in leaves. First, NH4+ can be imported from the atmosphere into leaf cells through stomata (Husted & Schjoerring, 1996). Second, significant concentrations of NH4+ have been measured in the xylem (Rawat et al., 1999), indicating that a considerable amount of N is translocated to shoots in this form. Third, the photorespiratory N cycle generates a large amount of NH4+ in leaf mitochondria that is subsequently transported to chloroplasts for reassimilation by GS, implying that the expression of AMT genes can be important to ensure the recycling of NH4+ during photorespiration (D’Apuzzo et al., 2004). Finally, the AMTs might also be recruited during the senescing process.
Expression of AMT1;6, AMT2;1 and AMT3;1 was exclusively detected in shoot tissues (Table 3). AMT1;6 is the closest ortholog of LeAMT1;3 in tomato (von Wiren et al., 2000b) and both genes form a distant clade in the AMT1 subfamily (Fig. 1). AMT1;6 complemented the triple ΔMep yeast strain (Fig. 9) and had a high affinity constant (Km NH4+ = 16 µm) comparable to that of AMT1;2 (Fig. 11b). It was suggested that LeAMT1;3 might be linked to the synthesis of asparagine as a N storage form in darkness or to the deamination of glutamate by glutamate dehydrogenase (von Wiren et al., 2000b). Consistent with this observation, AMT1;6 transcripts declined under N starvation in poplar (Fig. 4b), which was characterized by a decrease of the asparagine pool (Fig. 3c). In Arabidopsis, asparagine synthetase is encoded by a small gene family (ASN1, ASN2 and ASN3) and it has been shown that ASN2 expression correlates with ammonium metabolism (Wong et al., 2004). Indeed, the light induction of ASN2 is ammonium dependent and ASN2 transcripts decreased after transfer to ammonium-free growth medium. It has been suggested that the physiological role of ASN2 may be related, directly or indirectly to the recapture of lost N resources under stress conditions. However, the hypothesis that LeAMT1;3 is involved in the transport of photorespiratory NH4+ was not supported. AMT1;6 expression was strongly affected by the diurnal cycle (Fig. 5b) and correlated well with the sugar content. This is in good agreement with a recent report, which indicated that sugars make a major contribution to the diurnal changes of gene expression, while light, N metabolites and water deficit may make smaller contributions (Blasing et al., 2005).
Expression of AMT1;5, AMT1;6 and AMT3;1 increased with the state of maturation of leaves and highly correlated with the expression of the marker gene, cysteine protease whose implication in senescence has been revealed (Bhalerao et al., 2003) (Fig. 6). However, in contrast to AMT1;5 and AMT1;6, AMT3;1 expression was only detected in senescing leaves. Ammonium release during senescence has been documented (Mattsson & Schjoerring, 2003) and the specific expression of the glutamine synthetase gene NtGLN 1;3 was observed in senescing leaves of Nicotiana tabacum (Brugiere et al. 2000). In addition to the putative function of AMT2 members during leaf senescence (Howitt & Udvardi, 2000; von Wiren et al., 2000a; van der Graaff et al., 2006), our data suggest that AMT1 members might also be recruited to ensure ammonium assimilation during the process of leaf senescence.
It is also worth noting the contrasting expression patterns of AMT1;5 and AMT1;6: both were fully expressed in senescing leaves (Fig. 6), but the expression of AMT1;5 in flowers was specific to stamens and that of AMT1;6 specific to female flowers (Fig. 7a). The higher expression of AMT1;5 in stamens (Fig. 7) is in good agreement with data on its closest ortholog AtAMT1;4. Indeed, to complement the poplar data, we extracted genevestigator (Zimmermann et al., 2004) (http://www.genevestigator.ethz.ch) data for the Arabidopsis AMT gene family, which indicated that AtAMT1;4 was solely expressed in male flower but not in female flower parts. AMT1;5 and AMT1;6, might have important and specific roles in reproductive organ development in the dioecious plant, poplar. Further functional analysis is required to establish the precise role of AMTs and their correlation to plant reproduction mechanisms.
The two AMT genes AMT2;1 and AMT2;2 have been partly subfunctionalized during poplar evolution
AMT2;1 and AMT2;2 restored growth of the triple Mep mutant (Fig. 9) but were unable to take up methylamine, as demonstrated for other AMT2 members (Simon-Rosin et al., 2003). The single members of the AMT2 subfamily in Arabidopsis and L. japonicus, as well as the closest ortholog in rice (OsAMT2;1) were expressed in all organs (Sohlenkamp et al., 2002; Simon-Rosin et al., 2003; Suenaga et al., 2003). AtAMT2 was suggested to play a significant role in transferring ammonium between the apoplast and symplast of cells throughout the plant. Although expression of AtAMT2 in shoots responded little to changes in root N status, transcript levels in leaves declined under high CO2 conditions (Sohlenkamp et al., 2002). The LjAMT2;1 gene was found to be constitutively expressed throughout Lotus plants particularly in all major tissues of nodules. LjAMT2;1 would be implicated in the recovery of ammonium lost from nodule cells by efflux. A similar role may be fulfilled in other organs, especially leaves, which liberate ammonium during normal metabolism (Simon-Rosin et al., 2003). Unlike other plant species, this set of AMT genes in poplar displayed specific expression pattern, AMT2;2 being mostly expressed in petiole and AMT2;1 being highly expressed in leaves (Fig. 2). AMT2;2 and AMT2;1 arose from a recent duplication event (Sterck et al., 2005; Tuskan et al., 2006), which, in addition, resulted in a functional specialization (this work). The release of duplicate copies from constraints enables the evolution of new functions (neofunctionalization) or the loss of function (formation of a pseudogene), as observed for MADS-box genes (Irish & Litt, 2005). Alternatively, duplicated gene copies can potentially diverge in their roles, retaining different subfunctions of the original gene. The distribution and tissue localization of this subset of AMT2 genes in various plant species (AtAMT2;1, LjAMT2;1, OsAMT2;1, AMT2;1 and AMT2;2) suggest that the two poplar genes have been partly subfunctionalized during poplar evolution.
The high expression of AMT1;1 and AMT2;2 in petiole tissues is intriguing (Fig. 2). Much controversy exists about whether or not NH4+ is translocated in the xylem from roots to shoots. It was demonstrated that NH4+ might indeed constitute a significant part of the N translocated from the roots to the shoot in the xylem. A fundamental requirement for ammonium translocation is the presence of transport systems capable of loading NH4+ into the xylem and of subsequently moving NH4+ from the leaf apoplast solution into the leaf cells (Schjoerring et al., 2002). Furthermore, significant NH4+ contents were measured in the phloem sap of ammonium-fed tobacco plants, which was correlated with a higher glutamate dehydrogenase activity and polypeptide content in stems (Terce-Laforgue et al., 2004). Considering the frequency of plasmodesmata along the palisade cell/bundle sheath cell/companion cell and paraveinal mesophyll cell/bundle sheath cell/companion cell routes, Russin & Evert (1985) concluded that Populus deltoides would rather behave as a symplastic loader, suggesting that N compounds may also use the symplastic route for loading. In that case, active transport would not be required during these transfer steps. The compartmentalization of the ammonia assimilatory pathway in the vascular tissue, especially in perennial plants (Paczek et al., 2002), to better control carbon and N allocation at particular stages of plant development and/or under certain physiological conditions, needs to be better assessed.
The analysis of the AMT expression patterns in different organs of poplar show that expression of a given transporter can be very specific or can overlap with other genes, thus covering most tissues tested so far in poplar. The cell specificity of the individual members still remains to be demonstrated. The present data could show that multiple AMT genes with different spatial expression and transcriptional regulation allow the plant to respond differentially to varying nutritional conditions in the environment. Specific features of the perennial lifestyle plant, such as a high potential for ammonium remobilization during senescence or upon ectomycorrhization, together with subfunctionalization of AMT members, have been highlighted. The elucidation of the AMT gene family may present a comprehensive foundation for future studies on the ammonium nutrition of perennial plants.
B.M. was supported by a postdoctoral fellowship and J.C. by a PhD fellowship from the ‘Ministère déléguéà l’Enseignement supérieur et à la Recherche’. We thank Prof. Simone Ottonello (University Parma, Italy) for providing facilities for yeast uptake experiments and Dr Pascal Frey and Veronica Pereda (INRA Nancy) for providing leaf material. Financial support from the IFR 110 (Génomiques, Ecophysiologie et Ecologie fonctionnelle) is gratefully acknowledged.