•Adaptation of Medicago truncatula to local nitrogen (N) limitation was investigated to provide new insights into local and systemic N signaling.
•The split-root technique allowed a characterization of the local and systemic responses of NO3− or N2-fed plants to localized N limitation. 15N and 13C labeling were used to monitor plant nutrition. Plants expressing pMtENOD11-GUS and the sunn-2 hypernodulating mutant were used to unravel mechanisms involved in these responses.
•Unlike NO3−-fed plants, N2-fixing plants lacked the ability to compensate rapidly for a localized N limitation by up-regulating the N2-fixation activity of roots supplied elsewhere with N. However they displayed a long-term response via a growth stimulation of pre-existing nodules, and the generation of new nodules, likely through a decreased abortion rate of early nodulation events. Both these responses involve systemic signaling. The latter response is abolished in the sunn mutant, but the mutation does not prevent the first response.
•Local but also systemic regulatory mechanisms related to plant N status regulate de novo nodule development in Mt, and SUNN is required for this systemic regulation. By contrast, the stimulation of nodule growth triggered by systemic N signaling does not involve SUNN, indicating SUNN-independent signaling.
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.
Although soil nitrate content may, in some agricultural areas, show large, temporary increases, possibly as a result of water pollution (Vitousek et al., 1997), under natural conditions mineral nitrogen (N) generally limits plant growth because of spatial and temporal fluctuations in its availability in the soil. Nitrogen is acquired in a variety of forms by most plant species. Among them, legumes acquire N indirectly from atmospheric N2, through endosymbiosis with N2-fixing bacteria that involves the formation of a specific symbiotic organ (nodules) on roots. Dinitrogen-fixing bacteria export ammonium to host the plant, while the plant supplies bacteria with photosynthesis-derived carbon metabolites (Kouchi & Yoneyama, 1984). Nevertheless, even in legumes, because of the high energy cost of N2 fixation (Voisin et al., 2003) and the high sensitivity of nodules to environmental factors (Sprent et al., 1988; Serraj et al., 1999; Salon et al., 2001), N acquisition is often spatially and temporally limited.
NO3−-fed plants can compensate for a local N limitation and maintain their N status by stimulating NO3− acquisition by roots exposed elsewhere to mineral N (Forde, 2002). A short-term adaptation response is mediated by a rapid increase in both the capacity and the affinity of the uptake systems present in these roots. High-affinity NO3− transporters involved in these responses were identified in Arabidopsis (Cerezo et al., 2001; Filleur et al., 2001). In the long term, systemic adaptation to N deficit may also involve changes in root architecture to increase soil exploration and root proliferation in nitrate-rich zones (Zhang & Forde, 1998; Forde & Lorenzo, 2001; Remans et al., 2006). In the case of N2-fixing plants, the suppression of the bacterial capacity to fix N2 induced by an Ar/O2 treatment is correlated not only to a reduction of N2 acquisition but also to a local inhibition of rhizobial nodule and root growth (Singleton & van Kessel, 1987). This effect has been related to a decrease of the allocation of plant photosynthates toward inefficient roots and nodules. However, to our knowledge it is not known whether there is, in this case, as in the case of NO3− as a N source, a whole-plant compensatory response to local N limitation involving long-distance signaling.
In a previous study we have shown that the three main pathways involved in the acquisition of N (NH4, NO3 and N2) in Medicago truncatula (Mt) are regulated by the N status of the whole plant (Ruffel et al., 2008). The uptake activities of the three pathways are repressed by systemic signals when a high concentration of downstream N metabolites accumulates in the plant. Nevertheless, despite the common feature of these regulations, transcriptomic studies have shown that the gene networks targeted by these systemic signals are highly dependent upon the N source present in the root environment. Evidence of differential responses of the pathways of N acquisition to N limitation came from short-term local N starvation. Among the three N sources, only NO3−-fed plants were found to be able to compensate rapidly for a local N limitation by increasing the N uptake of the roots exposed elsewhere to NO3−. By analogy with the regulation of NO3− acquisition in Arabidopsis (Lejay et al., 1999; Gansel et al., 2001), this up-regulation was explained at both the physiological and molecular levels by the release of a systemic repression exerted by downstream N metabolites (Ruffel et al., 2008). The absence of equivalent responses in NH4+-fed and N2-fed plants was not the result of a carbon metabolite limitation (Ruffel et al., 2008) and suggested that the abolition of systemic repression does not allow the pathways involved in the acquisition of these sources to adapt rapidly to such constraints. Taken together these data suggested that the three acquisition pathways have contrasting abilities to mediate a rapid adaptation of the plant to N limitation. However, because only short-term responses to N limitation were investigated, the question of developmental changes that may also contribute to plant adaptation was not addressed.
Nodule formation and function require as a prerequisite low mineral N supply, whereas the addition of high concentrations of nitrate or ammonium inhibits the activity of pre-existing nodules (Streeter & Wong, 1988; Carroll & Mathews, 1990; Ferguson & Mathesius, 2003;Barbulova et al., 2007). Variants/mutants of alfalfa and Lotus japonicus which develop spontaneous nodules in the absence of rhizobia display similar nodule development responses to the amount of mineral N supply (Truchet et al., 1989; Tirichine et al., 2006), indicating a strong plant basis for this regulation. Evidence suggesting that control of nodule development by plant N status may involve distinct pathways involving nitrate itself and products of its assimilation has been reported (Carroll & Mathews, 1990; Barbulova et al., 2007). A transcriptomic approach has identified N-regulated transcripts potentially associated with the competence of L. japonicus plants to perform the nodulation program (Omrane et al., 2009). Supernodulating mutants with increased nodule number were isolated in several legume species, including Mt (review by Kinkema et al., 2006; Oka-Kira & Kawaguchi, 2006). These mutants are impaired in the ‘autoregulation of nodule number’ (AON) which is the systemic feedback repression of nodulation by the pre-existing nodules (Kosslak & Bohlool, 1984; Caetano-Anollés & Gresshoff, 1991). Different complementation groups may display this phenotype, but until now most of the mutations identified at the molecular level belong to one group. The corresponding gene, isolated in various legume species, (L. japonicus HAR1, pea SYM29, soybean NARK, and Mt SUNN), encodes a LRR-protein kinase closely related to the Arabidopsis CLAVATA 1 protein (Krusell et al., 2002; Nishimura et al., 2002; Searle et al., 2003; Schnabel et al., 2005). The current model suggests that following the perception of bacterial Nod factor, a root-derived signal is emitted to the shoots that respond by translocating a secondary signal to the roots via the phloem, which ultimately regulates nodule number (Oka-Kira & Kawaguchi, 2006). The phloem signal is lacking in supernodulating mutants (Carroll et al., 1985; Delves et al., 1992; Jiang & Gresshoff, 2002). Interestingly, the AON defect in the Mt sunn mutants and in other related mutants of soybean and Lotus has been shown to be associated with the ability of roots to form nodules in the presence of high concentrations of mineral N, conditions where nodule formation is inhibited in the wild-type, suggesting that the AON pathway shares common elements with the control of nodule development by whole-plant N status (Kinkema et al., 2006 and references therein). Recently, Nod factor/nitrate-induced CLE genes identified in L. japonicus have been proposed as candidates for the root-derived signal driving systemic regulation of nodulation upstream of HAR1 (Okamoto et al., 2009), thus giving further support to the hypothesis of cross-talk between the two pathways (Gresshoff et al., 2009).
In this study, we have characterized the ‘long-term’ adaptation strategy of Mt N2-fixing plants to a localized suppression of N acquisition and compared this with the strategy of the NO3−-fed plant. To mimic a local N stress, half of the root samples were deprived of N2 (Ar/O2 instead of air) in split-root systems. Systemic effects of treatments were investigated on the other halves of the root systems that remained supplied with the N source (air). The duration of the treatment (up to 18 d) was intended to reveal any significant changes in the process of functional nodule formation, in contrast to earlier short-term studies (Singleton & van Kessel, 1987; Ruffel et al., 2008). Plant growth, root growth, nodule growth and nitrogen fluxes were measured, allowing us to distinguish between the dynamics of the responses affecting the structures involved in N acquisition (root, nodule) and their N retrieval activities and to characterize the functional significance of these responses with regard to plant nutrition. Transgenic plants expressing the reporter gene pMtENOD11-GUS were used to investigate the effect of N limitation on early nodulation events. Photoassimilate allocation during N starvation was monitored by 13C isotopic labeling. The role of the AON pathway in the adaptation of N2-fixing plants to N deficit was further evaluated using the Mt sunn-2 mutant.
Materials and Methods
Plant growth conditions and experimental design
Seeds of Mt genotype A17, hypernodulating sunn-2 mutant (1 back-cross; Schnabel et al., 2005), and transgenic line L416 expressing the pMtENOD11-GUS construct (Charron et al., 2004) were scarified in H2SO4 99% (8 min), rinsed with water, placed on blotting paper in Petri dishes and cold-treated at 5°C (12 h). Seeds were then germinated in the dark. After 4 d, once primary roots were c. 3–4 cm long, the root tips were cut, which promoted root branching. As soon as plants had produced c. 10 secondary roots (3-wk-old plants), they were placed in hydroponic culture tanks (Barker et al., 2006). The plants were raised in a glasshouse with temperatures of 20 : 22°C (night : day) and a photoperiod of 14 h, and supplementary artificial light was provided when necessary to complement photosynthetic active radiation. Hydroponic culture tanks contained nutrient solution aerated permanently with an aquarium pump (Schemel & Goetz, Offenbach am Main, Germany). Plants were supplied with a nutrient solution containing mineral N (5 mM KNO3, 1 mM KH2PO4, 1 mM MgSO4, 0.25 mM K2SO4, 0.25 mM CaCl2, 30 μM H3BO3, 10.6 μM MnSO4, 0.7 μM ZnSO4, 3.2 μM CuSO4, 1 μM Na 2MoO4, 0.05 mM Na-Fe-EDTA, pH 6.8). For experiments with nodulated plants, roots of plants at the eighth leaf stage were inoculated and grown for two additional weeks with the Sinorhizobium meliloti strain 2011 in the same nutrient solution without mineral N, renewed every week. Typically, nodules appeared at 4–6 d after inoculation and were fully functional after 2 wk. Nitrogen deprivation treatments were achieved by removing mineral N from the nutrient solution in the case of NO3−-fed plant, whereas in the case of N2-fixing plants, the air was replaced by an Ar : O2 (80 : 20) mixture.
The amounts of C entering the plant through photosynthesis and N being reduced through symbiosis on whole intact plants were quantified with 13CO2 and 15N2 (Warembourg & Roumet, 1989; Voisin et al., 2003). Labeling experiments were conducted in a gas-proof chamber, allowing control of the shoot and root atmospheres using Dasylab software (SM2i, Villiers St Frédéric, France). The environmental parameters of this chamber were as follows: 8 : 16 h, light : dark cycle, 22°C, 60% hygrometry, 400–600 μmol photon m−2 s−1 (bottom and top of canopy, respectively) photosynthetic active radiation, and 350 μl l−1 CO2 concentration. Light was provided by four 400 W metal halide lamps (Mazda MAIH, APPRO5-21850, Saint Apollinaire, France) on each side of the enclosure and two 1000 W high-pressure sodium vapor lamps (Osram Nav-T, APPRO5-21850, Saint Apollinaire, France) above it. Aeroponic split-root systems were inserted in the labeling chamber 3 d before experiments for acclimation of plants. Shoots were exposed to a 13C-enriched atmosphere (5%13C/12C) for 7.5 h; the air CO2 concentration was continuously measured with an infrared gas analyzer (Ciras, PP Systems, Montigny le Bretonneux, France) and maintained by automatic CO2 injection using massic flow meters (Fluxmatec-93600, Aulnay sous Bois, France). Symbiotic N fixation was measured by enriching the root atmosphere with 15N2 (5%15N/14N) for 24 h. Isotopic compositions of gases in the labeling chamber were controlled by mass spectrometry throughout the experiment. Plants were harvested at the end of the experiment, and separated into shoots, roots, and nodules. Dry matter was determined after oven drying at 80°C (48 h). 15N and 13C content was analyzed using a continuous-flow isotope ratio mass spectrometer (Sercon, Crewe, UK) coupled to a C-N elemental analyzer (Thermo Electron NC2500, Courtaboeuf, France). The quantities of C (QC) and N (QN) in shoot, roots or nodules acquired during the labeling period were calculated as follows:
ECorgan and ENorgan are the 13C and 15N abundances, respectively, of the plant organ of the labeled plant or nonlabeled control plant; ECsource and ENsource are the 13C and 15N enrichment, respectively, of the labeled shoot and root atmosphere; DM is dry matter; and %C and %N are the percentages of C and N in dry matter (w/w), respectively.
Analysis of nodule and root development and growth
To analyze the response of nodule and root growth to local deprivation of individual plants, relative increments of root and nodule biomass present in compartments of the split-root systems (Fig. 1) were estimated. Nitrogen limitation was generated by removing the N source on one side of roots of split-root plants (see Fig. 1). At the end of the treatment, nodules and roots were collected and their dry weights were determined. The initial nodule and root biomass present before the initiation of the N-limitation treatment on the plant were estimated nondestructively by image analysis: a preliminary experiment (Supporting Information, Fig. S1) was conducted on nodulated plants similar to those used at the initiation of the treatments to establish the relationships between values estimated from image analysis (root/nodule surface, apparent nodule number) and values measured by destructive methods (root dry weight, nodule dry weight, nodule number). Images of roots were obtained using an Epson GT 15000 scanner (Seiko-Epson Corporation, Nogano, Japan) and analyzed (ARCGIS 9 software; ESRI, Redlands, CA, USA) to determine root/nodule surface and the number of visible nodules. To detect early nodulation events, transgenic L416 roots were stained for β-glucuronidase (GUS) activity as in Lagarde et al.(1996). Nodulation events were scored on high-resolution images and assigned to three classes according to their developmental stage (1, root infection sites, including those associated with a nodule primordium; 2, young nodules associated with a visible root outgrowth; 3, well-protruding nodules with a spherical or elongated shape (see Figs S2, 2e,f in Journet et al., 2001).
Statistical analyses were performed using the GLM procedure of SAS (SAS Institute, 1987). Means were compared using the Student–Newman–Keuls test at the 0.05 probability level.
N2-fixing plants display long-term compensatory responses that enhance N2 acquisition in response to N limitation
To detect ‘long-term’ adaptive responses that may involve changes in root and/or nodule development, experiments were conducted with either NO3−-fed or N2-fixing plants up to 14 d of N limitation. The systemic responses to N limitation were investigated by comparing the untreated roots of N-limited plants that remained supplied with N, with the roots of control plants homogeneously supplied with N during the treatment (Figs 1,2a). In NO3−-fed plants, removing the N source from half of the root system did not result in any overall growth reduction after 14 d (Fig. 2b), with little impact on plant total N concentration (Fig. 2c). This indicates that the untreated roots efficiently increased their NO3− uptake, fully compensating for the suppression of NO3− acquisition by treated roots. This adaptive response to compensate for the N deficit is explained by both the rapid up-regulation of NO3− transport systems (Ruffel et al., 2008) and a specific stimulation of growth of the untreated roots as compared with both treated roots of N-limited plants and roots of control plants (Fig. 2d).
Conversely, growth and total N concentration (Fig. 3a,b) of N-limited plants, were reduced by 20% as compared with the N2-fixing control plants. Taking into account the fact that control plants acquired c. 50% of their biomass and N content during the last 14 d, the N limitation of half of the root system resulted in c. 40–50% reduction of growth and N acquisition. This agrees with, and extends, our previous results showing a lack of up-regulation of N2 fixation after 4 d of treatment (Ruffel et al., 2008). However, measurements of N intake at the end of the 14-d period showed that the untreated roots of N-limited plants displayed a 70% increase in net 15N2 acquisition as compared with their counterparts in control plants (Fig. 3c). This showed that an up-regulation of N acquisition by the plant also occurred in untreated roots of N2-fixing, N-limited plants, probably with a marked delay and not as efficiently as for NO3−-fed plants, to maintain plant growth under local N limitation. Because this response was observed in the untreated roots of N-limited plants, we concluded that it was the result of a systemic signaling exerted by the plant N status.
The systemic response of N2-fixing plants to N2 limitation specifically involves nodule growth and development
Specific N2-fixation activity was calculated as the ratio of 15N2 assimilation per unit of nodule biomass. Specific N2-fixation activity was similar for roots of control plants (C roots in Fig. 1, 131 ± 11 μmol N g−1 nodules h−1) and untreated roots of N-limited plants (L roots in Fig. 1, 150 ± 7.5 μmol N g−1 nodules h−1) exposed to the same local environment during the treatment. Hence specific N2-fixation activity was not instrumental in the up-regulation of N, N acquisition by the plant observed after 14 d of localized N starvation (Fig. 3c).
Indeed, this response resulted entirely from a strong stimulation of nodule growth as compared with control plants observed after only 11 d of treatment (Fig. 4a). In the untreated roots of N-limited plants, nodule growth accelerated markedly, while it almost ceased in the Ar/O2-treated roots. This indicates that the N-limitation treatment triggered both local and systemic signals, modulating nodule development to favor N acquisition by the untreated side of the root system, still supplying plants with N2 fixation products; although, to a lesser extent, a similar response was also observed on roots, suggesting that local and systemic signals are also regulating root development (Fig. 4b).
The stimulation of nodule growth in the untreated root side of N-limited plants may have resulted from the enlargement of already developed nodules and/or the initiation and development of new nodules. Well-protruding nodules (class 3; Materials and Methods) were scored in all plants to determine whether any change in numbers occurred in response to N limitation (Fig. 5). After 11 d of N limitation, little change in nodule number was observed in both sides of the root system of control and N-limited plants (Fig. 5a). Consequently, as mean nodule weight in both control plants and untreated roots of N-limited plants increased (Fig. 5b), the increment of nodule biomass during this period resulted almost exclusively from the enlargement of pre-existing nodules. However, after 18 d of local N starvation, a significant increase (+ 20%) in nodule number was observed on the untreated roots of the N-limited plants as compared with their treated roots and roots of control plants (Fig. 5a). This suggests that a prolonged N limitation period stimulated nodule formation specifically in roots of N-limited plants that remained exposed to N2. Hence the overall adaptive response of nodule growth to N limitation in Mt (Fig. 4a) may be the sum of two successive processes: a growth stimulation of pre-existing nodules followed by an increased generation of new nodules.
Nodule formation includes successive phases where a subset of the early symbiotic structures may be arrested in their development. Therefore, the late stimulation of nodule formation in response to N limitation (Fig. 5a) might be the result of either an increased initiation rate of infection/nodulation events or a decrease in the abortion rate of such early events. To distinguish between these two hypotheses, we used transgenic Mr plants expressing the β-glucuronidase (GUS) reporter gene under the control of the promoter of the early nodulin MtENOD11 gene (Journet et al., 2001; Charron et al., 2004) to score early nodulation events (root infection sites possibly associated with underlying nodule primordia; class 1; Materials and Methods). The frequency of early events detected by this method was expressed as a ratio of the total number of infection sites and nodules present.
Four days after the initiation of the treatment, the relative numbers of early nodulation events were similar in roots of control plants and N-limited plants (Fig. 6). After 11 d, a lot fewer early nodulation events were scored on the treated side of N-limited plants as compared with the other roots, showing that the treatment itself may locally affect nodule initiation. Although the relative number of early nodulation events at day 11 was similar for the untreated roots of N-limited plants and the roots of control plants(Fig. 6), these roots displayed markedly different rates of nodule formation between days 11 and 18 (Fig. 5). Thus, nothing indicates that the higher frequency of well-protruding nodules observed in the untreated roots of N-limited plants at day 18 resulted from enhanced nodule initiation 1 wk earlier. Instead, this would support the hypothesis that a reduction of abortion rate of early nodulation events is responsible for this developmental response.
Change in C assimilate partitioning is an early response to local N starvation
The response of nodule growth observed after 11 d in N-limited plants, with a strong stimulation on the untreated side and complete arrest on the treated side (Fig. 4a), suggested that a major change in C assimilate partitioning occurred in these plants. To detect early changes in C flow triggered by local N limitation, both control and N-limited plants were exposed for 1 d to 13CO2 atmosphere 4 d after the beginning of the N-limitation treatment. The organ sink strength was evaluated by the ratio of 13C allocated to the organ by its biomass. Interestingly, a significant change in C allocation within the plant was revealed early after the beginning of the N-limitation treatment (Fig. 7). In both control and N-limited plants, nodules were the plant organ displaying the highest sink strength. Nodule sink strength was highest in the untreated roots of N-limited plants (Fig. 7), indicating that suppressing N2 fixation locally for 4 d benefited the roots still able to fix N2 for photosynthate allocation. This change in C allocation occurred before any significant change in nodule development and/or activity was detected.
The sunn-2 mutant is impaired in the systemic control of nodule formation by plant N status
As the increase of nodule biomass in the plant response to N limitation is derived in part from a stimulation of nodule formation by a whole-plant N deficit (Fig. 5a), the question of the control of AON pathway by the N status of the plant was investigated. Split-root experiments (Fig. 8) were designed to compare the ability of roots belonging to either N-sufficient or N-limited wild-type and sunn-2 mutant plants to form nodules in the absence of mineral N using an experimental system similar to that in Ruffel et al. (2008). Both types of genotypes had half of their root system transferred to a N-free nutrient solution (−N roots), and the other half supplied with mineral N (+N roots): high N (10 mM NH4NO3) for N-sufficient plants and a low N (0.5 mM NO3−) for N-limited plants. A mixed N source was chosen for the high N treatment to ensure that whole plants were fully N-sufficient, as NH4NO3 is a more efficient N source than NO3− or NH4+ alone (Krouk et al., 2006 and references therein).
In wild-type Mt plants, high local N provision in N-sufficient plants triggered a strong systemic inhibition of nodule formation in both +N and −N roots (Fig. 8, Table S1), demonstrating that nodule formation was dramatically feedback-regulated by systemic signals related to the whole-plant N status. However, nodule formation was clearly more strongly inhibited in +N roots (no visible nodules) than in −N roots (few small and white nodules, Fix- phenotype), confirming that a local inhibitory effect of NH4NO3 10 mM on nodule formation also exists. Comparison of the size and the color of the nodules present in the −N roots of N-sufficient and N-limited plants suggested that plant N status inhibited not only nodule formation but also nodule function. Interestingly, the inhibition of nodule formation by plant N status was no longer observed in the sunn-2 mutant, which had many visible nodules in both the +N and −N roots of either N-limited or N-sufficient plants (Fig. 8, Table S1). The SUNN pathway was indeed strongly involved in the systemic control of nodule formation by the N status of the whole plant. Nevertheless, careful comparison of nodules on the −N roots of N-sufficient plants and N-limited plants indicated that N status of the plant also had a strong effect on the size and the color (small and white nodules in N-sufficient plants vs large pink nodules and N-sufficient plants) of the sunn-2 nodules. Thus other mechanisms, regulating nodule development and function in relation to the plant N status, but independent of SUNN, remained efficient in the mutant.
The sunn-2 mutant is able to respond to N deficiency by rapid increase in N2 fixation
We have shown previously (Ruffel et al., 2008) that nodule N2 fixation-specific activity is under the control of a systemic repressive signal related to the N status of the plant. Nevertheless, attempts to relieve this systemic repression by preventing N2 fixation on one-half of the split-root system (Ar/O2 treatment) failed to up-regulate N2 fixation-specific activity in the other half of the root system, both in short-term (4 d; Ruffel et al., 2008) and long-term (14 d) experiments. One hypothesis to explain this paradox was that nodule N2 fixation-specific activity was already fully de-repressed and at its maximum capacity in control plants, and thus could not be further increased by the N-limitation treatment. Interestingly, the sunn-2 mutant offered the opportunity to directly test this hypothesis. Indeed, although hypernodulating mutants form many more nodules than wild-type plants, this is generally not associated with increased N2 acquisition, because of lower specific N2-fixation activity in these mutants than in the wild-type (Carroll et al., 1985; Schuller et al., 1988). Thus, it is possible that N2-fixation activity is not at its maximum in the sunn-2 mutant, and therefore could respond to the N-limitation treatment. To test this prediction, sunn-2 and wild-type N2-fixing plants were subjected to N limitation for 4 d by treatment with Ar/O2 exposure on half of their root system, while the other half was flushed with a mixture of 15N2/O2, in an experiment similar to the one described in Fig. 1. As expected, the specific N2-fixation activity of sunn-2 nodules was much lower than that of the wild-type in control conditions (Fig. 9a). After 4 d of N limitation, both specific N2-fixation activity and total N intake in the untreated side of the root system remained unchanged in wild-type plants (Fig. 9a,b). However, in the same conditions (Fig. 9b), the sunn-2 genotype increased specific N2-fixation activity in the untreated roots (+ 60%) and increased whole-plant total N acquisition (+ 30%). These results indicated that sunn-2 mutant plants trigger a quicker response of N2 fixation than the wild-type to the variations of the whole-plant N status, because they are able to de-repress specific N2-fixation activity, while wild-type plants cannot. As a consequence, this supports the hypothesis that specific N2-fixation activity was indeed already fully de-repressed in control wild-type plants. This also shows that the systemic N signaling pathway controlling nodule N2 fixation-specific activity is active in the sunn-2 mutant, and therefore that this pathway is independent of the SUNN gene.
The ability of NO3−-fed plants to react to N deprivation of one portion of its root system by increasing the NO3− uptake of the other roots has been documented in several species (Burns, 1991; Lainéet al., 1995). It involves the up-regulation of NO3− uptake systems (Cerezo et al., 2001; Gansel et al., 2001) and the stimulation of root growth (Drew, 1975; Robinson, 1994; Zhang & Forde, 1998; Zhang et al., 1999). Following our previous report concerning NO3− uptake systems (Ruffel et al., 2008), this work showed that this general scheme holds true for M. truncatula. When NO3− is the N source, the suppression of N acquisition in half of the roots is efficiently compensated by the other half, and results in unaltered growth of the whole plant. Equivalent adaptation was not found in N2-fed legume plants since the local suppression of N2 fixation in split-root Mt plants did not result, after either 4 d (Ruffel et al., 2008) or 14 d (this study), in any increase in specific N2-fixation activity in the untreated roots. Nevertheless, in the long term, N2-fixing plants compensated partially for the detrimental effect of local N limitation by stopping nodule growth in treated roots and by stimulating nodule development in untreated roots. Local reduction of nodule growth and rhizobia proliferation in response to Ar/O2 treatment has been reported in soybean, and has been interpreted as a ‘sanction mechanism’ exerted by the plant on an inefficient symbiotic partner (Singleton & van Kessel, 1987; Kiers et al., 2003). In this study we showed that this local response was associated with a systemic stimulation of nodule development in roots of N-limited plants still exposed to N2. However, the delay required for the full activation of new nodules generates a temporary N deficit, and plant growth restriction.
In the untreated roots of N-limited plants, the responses of nodule development to systemic N signaling involved two different and successive processes. First, the mean nodule weight of pre-existing nodules increased. Second, new nodules were generated. We used transgenic MtENOD11-GUS plants to investigate more precisely the effect of such N limitation on early nodule development. The increase in nodule number observed after 18 d was not associated with a previous stimulation of de novo early nodulation events, suggesting that other steps, downstream of root infection/nodule initiation, are likely to be targeted by N signaling. Therefore, the local and systemic effects induced by the same Ar/O2 treatment are likely to regulate different steps of nodule development, suggesting that different pathways are involved in these two responses.
The mechanisms involved in the regulation of nodule development as a function of the N status of the whole plant remain to be characterized. In our study, increased carbon allocation towards efficient nodules before any growth response became observable appeared as a fast response to local N limitation in N2-fixing plants. These data are consistent with previous studies in soybean, indicating that Ar/O2 treatments triggered a reduction of C allocation towards inefficient nodules (Singleton & van Kessel, 1987). Although nodules constitute a relatively small compartment of the plant in terms of biomass, they display one of the highest sink strengths for carbon. The availability of carbon metabolites and their utilization as a key factor governing nodule growth and activity is a very popular hypothesis (review by Vance & Heichel, 1991 and references therein). Our data provide some support for this hypothesis, suggesting that developmental responses to N limitation may be triggered by modified carbon allocation within the plant. However, whether the stimulated carbon allocation to nodules is part of the signaling mechanisms involved in the developmental responses, or is an early consequence of the initiation of these responses remains unknown.
To date, AON involving the SUNN gene in Mt (Schnabel et al., 2005) is the unique signaling mechanism identified as responsible for a systemic control of nodulation. AON and N signaling may be connected because the inhibition of nodule formation by high mineral N concentration is suppressed in hypernodulating mutants (Kinkema et al., 2006). This repression operates at early stages of nodulation (Malik et al., 1987; Heidstra et al., 1994), and some of the early steps of the Nod factor signaling cascade are blocked by high NO3− (Heidstra et al., 1997). Whether the AON pathway is involved in a direct inhibition of nodulation by mineral N or by organic N metabolites is still a matter of debate (Caetano-Anolles & Gresshoff, 1990; George & Robert, 1991; Kuppusamy et al., 2004; Tirichine et al., 2006). Differential inhibitory effects of NO3− and ammonium on nodulation in L. japonicus (Barbulova et al., 2007) suggested that NO3− and NH4+ may act on different regulatory mechanisms. The involvement of AON in the regulation of nodule development by the N status of the plant has been poorly investigated, especially in Mt. In this paper, the effect of a local high mineral N provision on distant nodule formation and growth was demonstrated, indicating that not only local but also systemic regulatory mechanisms related to the plant N status are regulating nodule development in Mt. This observation does not confirm the statement of Cho & Harper (1991) that, in soybean, the site of N application primarily controls the site of nodulation inhibition. However, careful analysis of Cho & Harper’s data reveals that a systemic effect of N application was also present in the soybean experiment but was not discussed in the paper; local application of 5 mM NO3− to one side of a split-root system did reduce twofold the number of nodules present on the untreated roots at the other side (Table S1, cf. lines 1 and 3).
The role of SUNN in nodule development responses to systemic repression triggered by high N supply was investigated. Although Barbulova et al. (2007) recently hypothesized that inhibition of early steps of nodulation by high concentrations of mineral N in L. japonicus might be mediated by HAR1-dependent and independent pathways, they were not able to separate local and systemic mechanisms and did not address the question of nodule growth, since L. japonicus nodules are determinate. The outcome of our Medicago experiments is that the nodule growth response to systemic signaling of plant N status involves both SUNN-dependent and SUNN-independent processes. The SUNN pathway is clearly required for the systemic repression of nodule formation in response to high N supply. However, the stimulation of nodule growth triggered by systemic N signaling does not involve SUNN, as the increase of nodule size in the untreated roots of N-limited plants was maintained in the mutant. We also showed that the systemic regulation of the N2 fixation-specific activity by the N status of the plant is independent of SUNN, since the sunn-2 mutant was able to up-regulate this activity in response to localized N starvation. The sunn-2 mutant remained able to sense its N status and to trigger several adaptive responses to N limitation, such as increase in nodule growth and stimulation of N2-fixation activity.
A low specific N2-fixation activity has been recorded in hypernodulating mutants (Schuller et al., 1988). In this paper we found that, in sunn-2 plants, this activity may be up-regulated in response to systemic signaling triggered by local N limitation. This observation suggests that the low specific N2-fixation activity in sunn results from a repression by systemic N signaling, which compensates for the de-regulated nodule proliferation. Therefore, the absence of up-regulation of this activity in control wild-type plants in response to N limitation is likely to be because it has reached its maximum capacity rather than because of a lack of control of this activity by plant N status. Why nodule N2-fixation capacity would be so tightly limited in wild-type plants remains to be determined. The flux of carbon metabolites delivered to the nodules has been determined to be a limiting factor for N2 fixation (Bacanamwo & Harper, 1997; Salon et al., 2001).
However, other factors have to be involved, as stimulated carbon allocation to untreated roots of N-limited plants is not associated with an increase in nodule-specific N2-fixation activity. Interestingly, previous studies have pointed out that M. truncatula A17 plants inoculated with the classical Sm2011 generally used by the scientific community are N-stressed when symbiotic N2 fixation is the main N source (Moreau et al., 2008), and that several other Rhizobium strains may promote higher fixation activity when associated with Medicago (Mhadhbi et al., 2005). Whether plants associated with these strains display the same or different responses as when associated with Sm2011 deserves to be investigated. This may help to clarify the role of the microsymbiont in the modulation of N2 fixation in response to N limitation.
We would like to thank Richard Thompson for a critical and constructive reading of the manuscript. This work was supported by the Sixth Framework Programme Grain Legume Integrated Project of the European Union (postdoctoral grants to SF and SR), by the AgroBI incitative action of INRA.