• External salinization can affect different steps of nitrogen (N) metabolism (ion uptake, N assimilation, and amino acid and protein synthesis) depending on the inorganic N source.
• Here, we assessed the net uptake of N supplied as nitrate or ammonium and N assimilation (combining metabolite analyses with molecular biological approaches) in grey poplar (Populus × canescens) plants grown under saline (75 mm NaCl) and control conditions.
• The specific (µmol N g−1 dry weight fine roots h−1) and total plant (µmol N per plant h−1) N net uptake rates, total plant N content, total plant biomass and total leaf protein concentration were reduced under saline conditions when plants were supplied with ammonium. In both nutritional groups, salt treatment caused pronounced accumulation of soluble N compounds in the leaves. The mRNAs of genes coding for enzymes catalyzing rate-limiting steps of both proline synthesis and degradation (delta-1-pyrroline-5-carboxylate synthase and proline dehydrogenase) as well as for NADH-dependent glutamate synthase were accumulated under saline conditions.
• Whereas under control conditions the plant N status seemed to be superior when ammonium was supplied, the N balance of ammonium-fed plants was more severely affected by salt stress than that of plants supplied with nitrate. Possible metabolic implications of stress-related accumulation of particular amino acids are discussed.
Excessive amounts of salt in the soil, most commonly NaCl, can have detrimental effects on plant growth and productivity (Boyer, 1982). These effects are mainly produced by salt-induced osmotic and salt ion-specific stresses, which affect the major plant metabolic activities. In addition, under hyperosmotic conditions the balance between synthesis and deactivation of reactive oxygen species (ROS) is disturbed, thereby enhancing intracellular oxidative stress as a secondary effect (for a review, see Parida & Das, 2005).
Primary and secondary salt stress signals trigger multiple plant responses to enhance adaptation to this stressful environment (Parida & Das, 2005). The perception, transduction and final effectors of salt signals have been intensively studied (Xiong et al., 2002b).
Microarray analyses and screening of knockout mutants, such as salt-oversensitive (sos) models, led to the identification of numerous salt-responsive genes (Gong et al., 2001; Xiong et al., 2002a). Proteins coded by these genes are involved, for example, in the regulation of salt uptake and transport (Na+/H+ antiporters), osmoprotection (delta-1-pyrroline-5-carboxylate synthase (P5CS)), ROS scavenging (superoxide dismutase), photosynthesis (chlorophyll-binding proteins), and cellullar signalling (calmodulin-binding proteins) (Wincov, 1998; Munns, 2005). Modification of these processes can improve ion and macromolecule homeostasis, detoxification reactions and plant growth.
In addition, several classes of N-containing compounds, mainly amino acids, quaternary ammonium compounds and polyamines, accumulate in plant tissues in response to excess salt stress (Mansour, 2000). Enrichment of plant tissues in free proline, an amino acid with a secondary amino group, is a typical plant response to osmotic stress (Parida & Das, 2005), and a strong relationship between proline accumulation and stress tolerance has been suggested (Nanjo et al., 1999). Upon osmotic stress, proline is synthesized from glutamate in two successive reactions, catalyzed by P5CS and delta-1-pyrroline-5-carboxylate reductase (P5CR). The pathway of proline degradation consists of two reactions, catalyzed by proline dehydrogenase (PDH) and delta-1-pyrroline-5-carboxylate dehydrogenase (P5CDH) (Hare & Cress, 1997). The processes leading to proline synthesis are up-regulated at the transcriptional level in dehydrated plants, and proline accumulation is achieved by both induction of proline synthesis and inhibition of proline degradation (Kiyosue et al., 1996). In Arabidopsis thaliana, reciprocal regulation of P5CS and PDH has been suggested to be a key mechanism in the adaptation of proline content during and after osmotic stress (Peng et al., 1996; Hayashi et al., 2000).
Glutamate is also the precursor of γ-amino butyric acid (GABA), a nonprotein amino acid, which accumulates in response to a wide range of environmental stresses. Upon salt exposure, elevated concentrations of Ca+ activate calmodulin-dependent glutamate decarboxylase, which catalyzes the formation of GABA from glutamate (for a review, see Kinnersley & Turano, 2000).
Many plants can utilize both nitrate and ammonium as an N source, but preferences for inorganic N sources differ among plant species (Crawford & Forde, 2002).
Populus is an increasingly used model for functional genomics in trees (Sterky et al., 2004). As the first sequenced woody plant (http://genome.jgi-psf.org/Poptr1/Poptr1.home.html), with a large number of genetic maps and a growing expressed sequence tag (EST) collection, poplar is an ideal subject for investigating the molecular background of stress-dependent physiological traits and for comparison of stress tolerance mechanisms between woody and herbaceous species (e.g. A. thaliana). The genus Populus comprises c. 30 species, with a wide range of tolerance to environmental stresses (Chen et al., 1997; Hare & Cress, 1997; Taylor, 2002). However, investigations on genetic adaptation and metabolite profiling upon salt exposure have been performed mainly with the relatively salt-tolerant Populus euphratica (Brosche et al., 2005; Ottow et al., 2005). Here we describe the effects of short-term salt treatment on N net uptake, accumulation of nitrogenous compounds and transcriptional regulation of proteins involved in N reduction and assimilation in the salt-sensitive species Populus×canescens (syn. Populus tremula×alba) (Hartmann et al., 2004). We assessed both the specific ammonium or nitrate net uptake rate per gram fine root dry weight (DW) and the whole-plant N net uptake rate. In addition, we determined concentrations (on a g DW basis) of N, proteins and amino compounds in source leaves, sink leaves and fine roots as well as total N contents on a tissue or whole-plant basis. Plants were kept either on ammonium or on nitrate in order to characterize the effect of the N source on salt sensitivity. The poplar hybrid used in this study is naturally occurring in floodplain forests (Willmanns, 1984), which are characterized by high temporal variability of N concentrations and available inorganic N form (nitrate/ammonium), in both cases as a consequence of periodic flooding. The experiments were performed in special glasshouses (solar domes; Brüggemann & Schnitzler, 2002), in order to ensure close to natural irradiation and climate conditions.
Materials and Methods
The experiments were performed with the wild-type grey poplar hybrid (Populus×canescens (Aiton) Sm. syn. P. tremula × alba), INRA clone 717 1B4, supplied by Picoplant Pflanzenvertrieb und -verkauf (Oldenburg, Germany). Seedlings were cultivated in a glasshouse in pots containing moistened balls (2–6 mm in diameter) of burned clay (Leca Ton; Leca, Lamstedt, Germany). Plants were irrigated with Hoagland solution (Hoagland & Arnon, 1950), which after 2 wk was replaced by modified Long-Ashton medium (Hewitt, 1966), containing two sources of mineral N, namely nitrate and ammonium (0.5 mm KNO3, 0.5 mm NH4Cl, 0.9 mm CaCl2, 0.3 mm MgSO4, 0.6 mm KH2PO4, 42 µm K2HPO4, 10 µm iron-ethylenediaminetetraacetic acid (Fe-EDTA), 2 µm MnSO4, 10 µm H3BO3, 7 µm Na2MoO4, 0.05 µm CoSO4, 0.2 µm ZnSO4 and 0.2 µm CuSO4, pH 5.5). After another 2 wk, plants were divided into two groups and grown for a further 5 wk on modified Long-Ashton medium, containing a single source of inorganic N (either 1 mm NH4Cl or 1 mm KNO3).
The salt stress experiment was performed in special glasshouses (solar domes) on the top of Mt Wank (1780 m above sea level), close to Garmisch-Partenkirchen (Germany), described in detail by Brüggemann & Schnitzler (2002). The photosynthetic photon fluence rate (PPFR) (400–800 nm) inside the solar domes was 75–85%, ultraviolet (UV)-A radiation was 70–75% and UV-B radiation was approx. 70% of outside values. According to Brüggemann & Schnitzler (2002), the elevated location minimized the ‘greenhouse effect’ of the solar domes, providing air temperatures and irradiation close to natural conditions at the valley level.
Each group of plants (pretreated with ammonium or nitrate medium) was split further into two subgroups, of which one was subjected to the salt treatment and the other was kept on the same medium as during the pretreatment without salt addition, thus serving as a control. Salt stress was increased gradually by applying 25, 50 and finally 75 mm of NaCl on the first, third and fifth days of treatment, respectively. Media were changed every second day during salt exposure to avoid pH changes and nutrient depletion. Two weeks after starting the salt treatment, plants were transferred to 1-l glass pots containing 900 ml of aerated nutrient solution, identical to the medium used before (modified Long-Ashton with or without 75 mm NaCl). After 72 h of preincubation, plants were transferred for 2 h into a fresh solution, in which the N source was 15N-labeled (either 1 mm 15NH4Cl or 1 mm K15NO3). The division into groups and subgroups (nitrate ± NaCl and ammonium ± NaCl) was maintained until harvest. Six tissues (fine roots, coarse roots, sink leaves, source leaves, bark and wood) were harvested from each plant.
Determination of tissue N and 15N contents and calculation of nitrate net uptake rates
The enrichment in 15N, resulting from 2 h of root incubation in the medium containing either K15NO3 or 15NH4Cl, and total N contents were measured in different oven-dried and homogenized plant tissues (fine and coarse roots, sink and source leaves, bark and wood). Aliquots of 1–2 mg were transferred into tin capsules (Type A; Thermo Quest, Egelsbach, Germany) and were injected into an elemental analyser (NA 2500; CE Instruments, Milan, Italy) coupled via a Conflo II Interface (Finnigan MAT GmbH, Bremen, Germany) to an isotope ratio mass spectrometer (Delta Plus; Finnigan MAT GmbH, Bremen, Germany). Specific 15N net incorporation (µmol g−1) was calculated using the following equation:
( Eqn 1)
(15Nt and 15Nc, the 15N abundance (atom percentage) in 15N-treated and control plants (without 15N application); [N], the total N concentration (g N g−1 DW); MW, the molecular weight of 15N (g mol−1)). The calculation of total 15N incorporation per tissue was based on the specific 15N incorporation and the respective biomass; total 15N incorporation per plant was computed by summing the total 15N incorporation of all tissues. Specific nitrate net uptake rates were calculated as the sum of 15N accumulation in the plants over time and based on the fresh weight (FW) of fine roots.
Extraction and determination of amino compounds and ammonium
Soluble N compounds were extracted and analyzed as described by Gessler et al. (1998). Tissues (fine roots, and source and sink leaves) were ground under liquid N with mortar and pestle. Aliquots of 100 mg of the frozen powder were homogenized in a mixture of 1 ml of methanol:chloroform (7 : 3) and 0.2 ml of Hepes buffer (5 mm ethyleneglycoltetraacetic acid (EGTA), 20 mm Hepes and 10 mm NaF, pH 7). Homogenates were incubated on ice for 30 min. Water-soluble amino compounds were extracted twice with 0.6 ml of distilled water. The aqueous phases were combined and freeze-dried (Alpha 2–4; Christ, Osterode, Germany). The dried material was dissolved in 1 ml of 0.2 m lithium citrate buffer (pH 2.2) directly before analyses. Amino compounds were separated and detected by an automated amino acid analyzer (Biochrom®; Pharmacia LKB, Freiburg, Germany) described in detail by Gessler et al. (1998).
Total proteins were extracted from fresh plant material (leaves or roots) ground under liquid N with mortar and pestle. A volume of 1 ml of extraction buffer (50 mm Tris-HCl, pH 8.0, 0.1% weight/volume (w/v) sodium dodecyl sulfate (SDS) and 1.4 µl of β-mercaptoethanol) was added to 50 mg of sample. Suspensions were shaken vigorously for 10 min at 4°C and centrifuged at 12 000 g, 4°C, for 15 min. Supernatants were transferred to new tubes. Pellets were washed once again with 500 µl of extraction buffer. After vortexing and centrifugation the upper phase was added to the previously collected supernatants. Protein solutions were purified using prepacked Sephadex™ G-25 columns (NAP™-25; Amersham Biosciences, Uppsala, Sweden). For column equilibration and sample elution, 25 mm Tris-HCl, pH 8.0, was used. Protein concentrations in the eluants were determined by Bradford's test, using a commercial protein assay (Bio-Rad Laboratories GmbH, München, Germany), with bovine serum albumin as a standard.
Isolation and expression analysis of poplar genes
Total RNA was extracted from 80 mg of powdered plant material (sink leaves and fine roots) using the Plant RNeasy kit, according to the manufacturer's instruction (Qiagen, Hilden, Germany). Partial cDNAs for genes involved in N assimilation and proline metabolism from Populus×canescens were cloned by reverse transcriptase–polymerase chain reaction (RT-PCR) amplification of total RNA from poplar roots and leaves with oligonucleotide primers derived from conserved domains. The functions of cloned fragments were inferred from sequence similarity searches (blast; National Centre of Biotechnology Information). Gene names and their abbreviations, accession numbers and similarities to homologous sequences are presented in Table 1. Additional verification of GS isoforms (cytosolic or plastidic) as well as GOGAT forms (ferredoxin-dependent or NADH-dependent) was performed by comparison of protein sequences predicted from particular cDNA fragments with characterized GS or GOGAT sequences from different species (Table 2).
Table 1. Relationships between identified poplar genes and their homologs
GenBank acc. no.
Closest homologue (acc. no.)
acc. no., accession number; PcFd-GOGAT, Populus × canescens ferredoxin-dependent glutamate synthase; PcGS1, Populus × canescens cytosolic glutamine synthetase; PcGS2, Populus × canescens plastidic glutamine synthetase; PcNADH-GOGAT, Populus × canescens NADH-dependent glutamate synthase; PcNR, Populus × canescens nitrate reductase; PcP5CS, Populus × canesceνσ delta-1-pyrroline-5-carboxylate synthase; PcPDH, Populus × canescens proline dehydrogenase.
Table 2. Comparison of protein sequences deduced from cloned Populus × canescens (a) glutamine synthetase (PcGS) and (b) glutamate synthase (PcGOGAT) fragments with homologous sequences of different species
For expression analysis, total RNA was isolated from fine roots or sink leaves of treated and control plants and aliquots of 0.75 µg were subjected to reverse transcription, using Super Script™ II RNase H (Invitrogen, Karlsruhe, Germany). To ensure internal control of the reaction, the housekeeping gene (β-tubulin; GenBank accession number AY353093) was amplified simultaneously in one tube with the gene of interest. For accurate discrimination of PCR products, gene-specific primers were designed to obtain a product size at least 40 bp different from that of the amplified β-tubulin fragment (370 bp). The following oligonucleotides were used for specific amplifications (* from Dluzniewska et al., 2006):
• Populus × canescens nitrate reductase (PcNR): 5′-ATCATCGGATCGGAGAGTTGG* and 5′-GACGGTGCTAGTTGGCGTATAG;
• Populus × canescens cytosolic glutamine synthetase (PcGS1): 5′-CCGTGATATTGTTGACTCCCAC* and 5′-GGCGCAGACCAAGCTTCTC*;
• Populus × canescens plastidic glutamine synthetase (PcGS2): 5′-GCGAAGTGACAGATTTAGGATTGC and 5′-ATCAGGTGGGTCCCAGTGTG;
• Populus × canescens ferredoxin-dependent glutamate synthase (PcFd-GOGAT): 5′-AACCCAAAGGCATCAGACTCAG and 5′-AGTAAAGCAGGTCCATCCCAAG;
• Populus × canescens NADH-dependent glutamate synthase (PcNADH-GOGAT): 5′-GGTGTTGTGGATATTCCTCCTG and 5′-TCAGATGCGGCGACAACCC;
• Populus × canescens delta-1-pyrroline-5-carboxylate synthase (PcP5CS): 5′-GTCATGTGTACGTGGATAAGTCTG and 5′-GCCCTCTGTCCACCAAATAAAG;
• Populus × canescens proline dehydrogenase (PcPDH): 5′-CAGAACCATTAACCCCGCAAG and 5′-GATCGCTGCCGAATATGTCAAG;
• Populus × canescensβ-tubulin: 5′-TAAATCCGGTCACCGATTTC* and 5′-GAACCGAATCAATCAACTCC*.
The parameters of multiplex reactions were determined experimentally for each amplified fragment, using mRNA of untreated samples as the template. In particular, the concentrations of primers were selected to obtain sufficient amounts of both amplicons and to ensure that primers would not limit the reactions. The number of cycles was chosen to ensure that both products were clearly visible on the agarose gel but stayed in the exponential phase of amplification (26–29 cycles). PCR fragments were separated by agarose gel electrophoresis and stained with ethidium bromide. Gels were visualized under UV light; images were taken using a gel documentation system, and quantified with Molecular Analyst® (Bio-Rad Laboratories, Hercules, CA, USA). Relative expression was calculated as the ratio between the quantified abundance of the amplified gene of interest and that of β-tubulin.
The data obtained in the experiments were subjected to analyses of variance (ANOVAs) and multiple range tests (Duncan) by ANOVA, and to Pearson product moment correlation. All statistic analyses were performed with Statistical Product and Service Solutions (SPSS) for Windows, version 13.0 (SPSS Inc., Chicago, IL, USA).
Salt treatment reduced total plant biomass when plants were supplied with a medium containing ammonium as an N source (Table 3). This reduction was mainly a result of decreased biomass accumulation in source leaves and coarse roots. In nitrate-fed plants, salinity caused a reduction of fine and coarse root biomasses; however, this effect did not significantly influence the total plant DW.
Table 3. Effect of the salt treatment on plant dry weight, total plant nitrogen (N) net uptake and total N content of different tissues of poplar (Populus × canescens) plants
Plants were kept on medium containing either 1 mm nitrate or 1 mm ammonium as an N source. The final NaCl concentration in the salt treatment was 75 mm. Data shown are means ± standard deviations of five replicates. The two analysis of variance (ANOVA) columns show the significance levels for differences between salt-treated and control plants for a given N nutrition form (nitrate or ammonium); ns, nonsignificant at P = 0.05; a–c, values with the same letter indicate homogenous groups (P < 0.05) when comparing all four treatments (nitrate control, nitrate plus salt, ammonium control and ammonium plus salt).
’Whole plant’ refers to the sum of leaf, root and stem biomasses or N contents.
Specific ammonium net uptake (µmol N g−1 DW fine roots h−1) by the fine roots was significantly higher than nitrate net uptake in the control treatment, but no significant difference was observed with salt exposure (Fig. 1). As a consequence of salt exposure, specific nitrate uptake did not change significantly, whereas specific ammonium net uptake was reduced significantly by 54%. Total plant ammonium net uptake (µmol N per plant h−1) was reduced to an even greater extent under saline conditions, whereas total nitrate net uptake was not significantly different between salt-treated and control plants (Table 3).
In nitrate-supplied plants, the total N concentration did not show salt-induced responses in any plant tissue (Fig. 2a). Total N contents of different organs changed to a lesser extent compared with plants fed ammonium; the largest reduction was in root total N content (Table 3).
Decreased ammonium net uptake in salt-treated plants resulted in decreased N content in source leaves and roots, which also affected the total plant N content (Table 3). However, as a consequence of the reduced plant DW, the N concentration in these tissues did not change (Fig. 2b). In young sink leaves of plants fed ammonium, the total N content did not differ between salt-treated poplars and controls (Table 3); the N concentration in this tissue was even slightly increased after salt treatment, whereas total N content and concentration in the fine roots were not affected by the salt treatment. (Fig. 2b).
The concentration of total nonprotein amino N (NPAN) was comparable in sink and source leaves between nitrate and ammonium-supplied controls (Fig. 2). Total NPAN concentrations in both leaf types ranged between approx. 19.0 and 33.44 µmol N g−1 DW when trees were not exposed to salt. In fine roots, total NPAN concentrations were significantly (P = 0.01) higher in ammonium-supplied (51.6 µmol N g−1 DW) than in nitrate-supplied (27.9 µmol N g−1 DW) control plants.
Irrespective of the N source, salt exposure resulted in a significant increase in total NPAN concentrations in sink and source leaves, but not in fine roots (Fig. 2). The increase in foliar total NPAN concentrations was, however, more pronounced in plants provided with ammonium. In sink and source leaves of ammonium-supplied plants, total NPAN increased 5.4- and 2.9-fold, respectively, whereas the concentration increased 3.9- and 2.4-fold, respectively, in nitrate-treated plants. Nonproteinogenic amino compounds contributed 23% (sink leaves) and 29% (source leaves) to the increase in total NPAN in ammonium-fed plants; in nitrate-supplied trees the contribution of this group of amino compounds was < 15%.
The total protein concentration increased significantly under salt treatment in sink leaves and fine roots of nitrate-supplied poplar but remained unchanged in source leaves. In contrast to nitrate-supplied plants and unlike total NPAN concentrations, protein concentration decreased in source and sink leaves of ammonium-treated plants under salt exposure.
In order to obtain more detailed information on salt-induced changes of N metabolism, the nine most abundant amino compounds plus ammonium are shown in Fig. 3. In a comparison of the profiles of amino compounds in leaves between nitrate- and ammonium-fed control plants, some differences became apparent: whereas in both source and sink leaves of ammonium-supplied plants serine was the most abundant amino compound, contributing approx. 26% of total NPAN, glutamate was most abundant (c. 25%) in nitrate-fed poplar plants. In fine roots, N source-related differences in the NPAN profile were not observed and asparagine was the dominant amino compound.
As a result of salt exposure, serine became the most abundant amino compound in sink and source leaves of both N nutrition treatments. In sink leaves of nitrate-supplied poplar, alanine, GABA, glutamine, proline and threonine additionally increased significantly by factors of 5.3, 21.1, 5.1, 5.2 and 4.2, respectively, when roots were treated with salt. In ammonium-fed plants, alanine, arginine, asparagine, GABA, glutamine, proline and threonine were enriched up to 39-fold. In source leaves, the pattern of the relative change in amino acid concentration was comparable to that for sink leaves under both N regimes, although the absolute increase was much lower. In ammonium-treated plants, the glutamate concentration actually decreased significantly as a consequence of salt treatment. In both sink and source leaves of ammonium-treated plants, the ammonium concentration increased in the salt treatments, an effect not observed in the nitrate-supplied plants.
In the roots of nitrate-fed poplar, aspartate, serine and threonine increased as a result of salt exposure. In ammonium-supplied poplar, an increase in concentration of particular amino compounds was not observed, whereas alanine, GABA and glutamine concentrations decreased. In contrast to the leaves, root proline was below the limit of detection.
Expression of cloned genes was analyzed by relative quantitative RT-PCR in the tissues of salt-treated and control plants.
Neither sink leaf nor root PcNR mRNA responded to the 2-wk salt treatment (Fig. 4). Salt-stressed plants did not show significantly different transcript abundances of PcNR to those observed in the tissues of controls (i.e. untreated plants, which were fed with the same N source, either nitrate or ammonium).
Genes for six enzymes involved in metabolism of amino acids were analyzed in the sink leaves of poplar plants; four of them are involved in the GS/GOGAT cycle (cytosolic glutamine synthetase (GS1), plastidic glutamine synthetase (GS20, ferredoxin-dependent glutamate synthase (Fd-GOGAT), and NADH-dependent glutamate synthase (NADH-GOGAT)), and the remaining two catalyze rate-limiting reactions in the proline assimilation and degradation pathways (P5CS and PDH, respectively).
Populus × canescens glutamine synthetase (PcGS) mRNA abundance was not affected by the salt treatment; genes for both cytosolic and plastidic isoforms of GS (PcGS1 and PcGS2, respectively) showed transcription levels comparable to those observed in controls. Different reactions to salt were observed for the expression of genes for different isoforms of GOGAT: PcFd-GOGAT transcription was not affected by NaCl treatment, whereas mRNA for PcNADH-GOGAT showed 2- and 4-fold increases in plants kept on media with nitrate and ammonium, respectively.
The mRNAs of the genes coding for the two enzymes known to be involved in downstream regulation of proline concentration upon osmotic stress, PDH and P5CS (Kiyosue et al., 1996), accumulated in the leaves of salt-stressed poplars. Expressions of these genes were 3–5 times higher after salt exposure, compared with controls (Fig. 4).
Significant interactions between salinity and N source were observed for specific (µmol N g−1 DW fine roots h−1) and whole-plant (µmol N per plant h−1) N net uptake. Salt treatment reduced only ammonium net uptake; in the plants kept on nitrate medium, differences in N net uptake rates between salt-exposed and untreated plants were not observed. Reduced ammonium uptake under saline conditions was also observed in Puccinellia phryganodes (Henry & Jefferies, 2003) and Triticum aestivum (Hawkins & Lewis, 1993). Long-term exposure to salinity leads to competitive inhibition of ammonium uptake (Bradley & Morris, 1991). It appears that the reduction of N uptake in plants grown in the solutions containing ammonium as a solely N source can be attributed to an increased Na+:ammonium ratio in the medium.
The reduction of ammonium net uptake in salt-treated poplars resulted in decreased whole-plant N content and plant DW compared with controls. Salt induced a reduction of source leaf biomass, whereas the DW of sink leaves was not altered by salt exposure (Table 3). This observation can be explained by the fact that NaCl is accumulated mainly in older, longer transpiring leaves, where it reaches high, toxic concentrations (Munns, 2005). The increased total N concentrations (Fig. 2b) in sink leaves of plants fed ammonium and exposed to NaCl, together with the reduction of total N contents in source leaves (Table 3), suggests that under saline conditions N was translocated from older tissues to the newly expanding leaves. In addition to disrupted N acquisition, reduced biomass may be a consequence of the toxic effects of ammonium, which accumulates in the leaves, probably as a result of enhanced protein degradation after salt treatment. It was also demonstrated, that the combined effect of salinity and ammonium medium impaired uptake of other ions, such as potassium (K+), calcium (Ca2+) and magnesium (Mg2+), leading to ion imbalances (Ashraf & Sultana, 2000; Al-Mutawa & El-Katony, 2001). The observation of Ehlting et al. (2006) that salt exposure of poplar decreased K+:Na+ ratios by more than one order of magnitude in leaves and roots is in agreement with the assumption of ion imbalances.
In contrast to our observations, tomato (Lycopersicon esculentum) and rice (Oryza sativa) (Abdelgadir et al., 2005) as well as Plantago maritima (Rubinigg et al., 2005) showed reduced nitrate net uptake when exposed to increased NaCl concentrations. Rubinigg et al. (2005) concluded that a lower N demand for growth and neither competitive nor noncompetitive inhibition of nitrate transport systems was responsible for the observed salt effect. In our experiment, neither total plant biomass nor whole-plant N content differed between controls and NaCl-treated plants grown with nitrate as the N source. As a consequence, we have to assume that no changes in whole-plant N demand occurred, which might explain the unchanged nitrate net uptake rates between salt-treated and control plants.
Although specific nitrate net uptake rates were not altered under saline conditions, the DW of fine roots was dramatically reduced. This alteration, which was not observed in leaves, might be explained as a consequence of the direct contact with medium with a high NaCl concentration. However, in many other plants, leaves showed stronger salt stress effects, whereas root tissues were relatively resistant (Munns, 2002).
Induction of NR expression and activity is dependent on the presence of substrate and on the flux of inorganic N into organic compounds (Samuelson et al., 1995; Crawford & Forde, 2002). Reduced NR activity in salt-treated Bruguiera parviflora was accompanied by a decrease of both nitrate uptake and total N content (Parida & Das, 2004). Similarly, nitrate was the most significant regulator of NR expression and activity in salt-stressed Zea mays (Baki et al., 2000). As described above, in grey poplars kept on nitrate as the sole N source, neither N net uptake nor plant total N content or tissue-specific N concentration was affected by salinity. This is consistent with a lack of alteration in the expression of PcNR in the tissues of salt-treated plants (Fig. 4a). NR activity was not measured in this study; however, salt stress-induced post-translational modifications seem to play a less important role in the regulation of NR activity than transcriptional regulation (Kaiser & Huber, 2001).
The large accumulation of amino compounds in the leaves of salt-exposed plants was the most remarkable response of N metabolism of grey poplar to the saline environment. In plants fed ammonium, the increased pool of soluble amino compounds can be attributed to enhanced proteolysis in leaves, because protein concentrations decreased in both sink and source leaves. The strong increase in total NPAN concentration in sink leaves might also have been attributable to changes in whole-plant N partitioning. In plants fed nitrate, N assimilation was not inhibited by salt stress; because of the reduction in root biomass (Table 2), proteins were accumulated (Fig. 2a) to concentrations that might exceed the needs of the roots, and thus the elevated amino acid pool in the leaves might at least partially originate from degradation of proteins in the roots and subsequent transport of amino compounds via the xylem in the acropetal direction.
Accumulation of free amino acids in plant tissues under salinity is a frequently reported response to salt stress, being observed in both salt-tolerant and salt-sensitive species and genotypes. The roles of most of the amino compounds accumulated remain unclear. It is likely that highly concentrated amino acids contribute to osmoprotective processes. In addition to raising cytosolic osmotic pressure they protect folded macromolecule structures, serve as N and energy sources and mitigate oxidative stress by scavenging ROS (Mansour, 2000).
The accumulation of GABA was the most remarkable change in the NPAN profiles of salt-stressed poplars. GABA is a highly soluble, nontoxic compound, and exhibits higher cryoprotective and hydroxyl-radical scavenging activity than proline (Smirnoff & Cumbes, 1989; Shelp et al., 1999). Arabidopsis thaliana knockout mutants of succinic semialdehyde dehydrogenase, one of the enzymes of the GABA synthesis pathway, exhibit oversensitivity to heat and UV as well as increased production of hydrogen peroxide (Bouchéet al., 2003). Thus, the strong increase of GABA concentration in the leaves observed in our experiment underlines the role of this amino acid in protection against osmotic and salt-dependent oxidative stresses.
The marked accumulation of serine in the leaves of NaCl-treated plants, yielding the highest abundance of all analysed amino compounds, might be at least partially attributable to increased photorespiration in salt-exposed poplars. Upon salt or drought stress, stomata are generally closed to prevent excessive water loss; this can result in a low intercellular pCO2:pO2 ratio and, as a consequence, an increased rate of oxygenation reaction of ribulose bisphosphate (Wingler et al., 2000). Serine is an intermediate of the photorespiratory pathway produced in the mitochondria during the recycling of glycolate-2-P to glycerate-3-P (Douce & Neuburger, 1999). In addition to its role in the photorespiratory pathway, we may also assume serine accumulation to play a role in osmoprotection, as already observed in animal models (Patrick & Bradley, 2000).
Although the leaf proline concentration increased as a result of salt treatment, the contributions of other proteinogenic and nonproteinogenic amino compounds, such as GABA, serine, alanine and arginine, to the increased total leaf NPAN pools were much more significant. Moderate accumulation of proline is a characteristic metabolic response to osmotic stress in glycophytes. Comparative analyses of salt-sensitive A. thaliana and salt-tolerant Thellungiella halophila, subjected to identical salt exposure, revealed 2- to 3-fold higher proline accumulation in the leaves of the tolerant species (Kant et al., 2006). Moreover, the concentration of free proline in T. halophila was also much higher than that detected in A. thaliana during normal growth conditions (Taji et al., 2004; Kant et al., 2006). The phreatophyte Populus tremula, which is also considered to be salt tolerant, accumulated 5 µmol proline g−1 FW (Jouve et al., 2004), whereas the maximal concentration of proline in Populus × canescens observed in our experiment in young leaves of ammonium-fed plants was 0.2 µmol proline g−1 FW (both values for plants not treated with NaCl). Thus, the concentration of free proline in plant tissues seems to be a physiological trait which distinguishes salt-sensitive and salt-tolerant species and genotypes.
The processes controlling the concentration of proline in dehydrated plants are subject to strong transcriptional regulation, and this applies to both proline synthesis and proline degradation (Kiyosue et al., 1996). For Populus species, none of the genes controlling proline synthesis or/and degradation has been identified and characterized until now. Ours is the first study in which partial sequences of poplar P5CS and PDH genes (PcP5CS and PcPDH) have been cloned and used for functional characterization.
The PcP5CS gene in poplar sink leaves was up-regulated by salt exposure in both N nutrition regimes and reached an expression level more than 3 times higher than that of the control plants. A salt-induced increase in transcription of P5CS was observed, for example, in A. thaliana (Strizhov et al., 1997), L. esculentum (Fujita et al., 1998), Medicago truncatula (Armengaud et al., 2004) and Medicago sativa (Ginzberg et al., 1998). Some studies repudiate the importance of transcriptional induction of P5CS for proline accumulation, suggesting that other mechanisms, such as increased proline translocation between tissues, can regulate intracellular proline concentrations (Fujita et al., 1998). Poplar leaf proline concentration correlated positively with PcP5CS expression (for all plants in all treatments: Pearson's correlation coefficient R = 0.846, P = 0.01, n = 20). Thus, we conclude that the reduction of glutamate to glutamic-γ-semialdehyde, catalyzed by P5CS, is heavily involved in the regulation of proline synthesis in poplar leaves. The transcript abundance of a second enzyme of the proline production pathway, P5CR, was not measured in this study. Although some studies reported salt-dependent induction of P5CR transcription and activity, the reduction of delta-1-pyrroline-5-carboxylate to proline seems not to be rate limiting in proline synthesis (Szoke et al., 1992). The induction of PcP5CS and the accumulation of free proline in poplar leaves were accompanied by a 4–5-fold induction of PcPDH expression, depending on the N form supplied to the plants. Simultaneous induction of both proline synthesis and proline degradation, accompanied by moderate proline accumulation in plants exposed to salinity, excludes the possibility of a meaningful role of this amino acid in the osmotic adjustment of poplar cells.
During osmotic stress, accumulation of the reduced form of pyridine nucleotides (NADPH) takes place (Hare & Cress, 1997). In both reduction reactions, leading to proline synthesis from glutamate, NADPH is used as a cofactor. Thus, proline synthesis can mitigate stress-dependent disturbance in the redox balance (Hare & Cress, 1997). Proline degradation, in contrast, is a process achieving a high energy yield (39 molecules of ATP). As proposed by Atkinson (1977), proline can serve as an important energy source, used in stressed cells for activating adaptive metabolic pathways.
Simultaneous induction of both the P5CS and PDH genes is a rather unusual observation. In most species, proline degradation was found to be repressed by hyperosmolarity, while rapid accumulation of PDH transcript was observed upon relief of stress and during plant recovery (Kiyosue et al., 1996; Verbruggen et al., 1996; Hare et al., 1999). It was proposed that proline accumulated during stress conditions serves as an important energy source for use during plant recovery (Hare & Cress, 1997). Simultaneous increases in proline synthesis and oxidation, as observed for Populus×canescens upon salt exposure in the present study, may thus indicate a plant reaction towards disturbed redox and energy balances.
Several studies indicate that under saline conditions increased activities of enzymes involved in the GS/GOGAT cycle ensure a high glutamate concentration, which is required for proline synthesis (Berteli et al., 1995; Lutts et al., 1999; Silveira et al., 2003). In contrast to the findings of these studies, in the present study neither PcGS nor PcFd-GOGAT was found to be activated in sink leaves, whereas PcNADH-GOGAT reacted strongly to salt exposure. The relatively low concentration of the GOGAT product glutamate in leaf tissues can be explained by a substantial consumption of this amino acid for the synthesis of GABA and/or, to a lesser extent, proline. Accumulation of glutamine (as observed here upon salt treatment and which might originate from enhanced protein degradation and photorespiratory ammonium reassimilation) can induce PcNADH-GOGAT expression. Moreover, the present results support the theory that, at high glutamine concentrations, GOGAT and not GS takes over the control of the GS/GOGAT cycle (Baron et al., 1994). It is not clear why salt exposure induced the NADH-dependent isoform of GOGAT, which plays a minor role in leaves under normal conditions, but is the dominating isoform in nonphotosynthetic tissues (Suarez et al., 2002). We might speculate that in the sink leaves the NADH-dependent isoform is involved in the synthesis of glutamate from glutamine, which arrives via the phloem or xylem from other tissues where nitrogen remobilization is assumed to take place under salt stress. Glutamine is the most important phloem- and xylem-mobile transport compound in many tree species (Gessler et al., 2004) and thus should be relocated when changes in whole-plant N partitioning occur. As proposed for proline synthesis, activation of the NADH-dependent isoform of GOGAT might also result from the high NADH:NAD+ ratio in salt-exposed tissues.
In conclusion, in salt-sensitive Populus×canescens, salt treatment significantly affected plant biomass, N acquisition, N metabolite accumulation and mRNA abundances for enzymes involved in amino acid metabolism, depending on N source. Whereas under control conditions the plant N status (e.g. whole-plant N net uptake and whole-plant N content; Table 3) seemed to be superior when ammonium was supplied, reduced N net uptake under salt treatment, associated with reduced plant N content as well as reduced total plant biomass, was found only when ammonium was the sole N source applied to the plants. Transcriptional regulation of the PcNADH-GOGAT, PcPDH and PcP5CS genes and alterations in leaf amino acid profiles seem to be part of plant response for osmotic and oxidative stresses, driven by the unfavourable high-salinity conditions.
This work was financially supported by the Deutsche Forschungsgemeinschaft (DFG), research group ‘Poplar Research Group Germany’ (PRG), under contract No. Re 515/20-1. We thank Stanislav Kopriva for helpful comments and fruitful discussion.