•Legumes form a symbiotic interaction with bacteria of the Rhizobiaceae family to produce nitrogen-fixing root nodules under nitrogen-limiting conditions. We examined the importance of glutathione (GSH) and homoglutathione (hGSH) during the nitrogen fixation process.
•Spatial patterns of the expression of the genes involved in the biosynthesis of both thiols were studied using promoter-GUS fusion analysis. Genetic approaches using the nodule nitrogen-fixing zone-specific nodule cysteine rich (NCR001) promoter were employed to determine the importance of (h)GSH in biological nitrogen fixation (BNF).
•The (h)GSH synthesis genes showed a tissue-specific expression pattern in the nodule. Down-regulation of the γ-glutamylcysteine synthetase (γECS) gene by RNA interference resulted in significantly lower BNF associated with a significant reduction in the expression of the leghemoglobin and thioredoxin S1 genes. Moreover, this lower (h)GSH content was correlated with a reduction in the nodule size. Conversely, γECS overexpression resulted in an elevated GSH content which was correlated with increased BNF and significantly higher expression of the sucrose synthase-1 and leghemoglobin genes.
•Taken together, these data show that the plant (h)GSH content of the nodule nitrogen-fixing zone modulates the efficiency of the BNF process, demonstrating their important role in the regulation of this process.
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.
The symbiotic interaction between legumes and soil bacteria rhizobia leads to the development of a new organ, the root nodule, wherein the differentiated bacteria, the so-called bacteroids, fix atmospheric nitrogen (N). This biological process represents a fundamental source of ammonia for the normal growth and development of most legumes. Thus, symbiotic N2 fixation is important at the economical and agricultural level, limiting the use of fertilizer for the N nutrition of plants.
The partnership between Medicago truncatula and Sinorhizobium meliloti is a good model with which to study the establishment of the symbiotic interaction (Barker et al., 1990; Cook et al., 1997). The molecular mechanisms involved in the recognition of each symbiont by the other have been particularly extensively studied during the last decade (for reviews, see Jones et al., 2007; Oldroyd & Downie, 2008; Masson-Boivin et al., 2009). Entrance of the bacteria into root hairs occurs via the formation of infection threads and in parallel a new meristem, the nodule primordium, is initiated from root cortical cells. In Medicago, the persistent activity of this nodule meristem results in the development of elongated indeterminate nodules. During the first few weeks, many cells deriving from the nodule meristem are infected by bacteria released from infection threads (infection zone II), leading to the development of a large area of N fixation (zone III) and to a maximum N-fixing capacity. From the fifth week of development, symbiotic equilibrium begins to be disrupted and a senescent zone IV is formed at the proximal end of the nodule. This phenomenon mainly occurs before plant flowering and the onset of leaf senescence.
Reactive oxygen species and antioxidants are known to be implicated in nodule metabolism (Chang et al., 2009). Root nodules have a stronger antioxidant defence capacity than the parental roots: higher amounts of antioxidant molecules (ascorbate, reduced glutathione (GSH) and homoglutathione (hGSH)) and higher activities of antioxidant enzymes such as superoxide dismutase, catalase, peroxidases and ascorbate-glutathione cycle enzymes (glutathione reductase, monodehydroascorbate reductase, dehydroascorbate reductase and ascorbate peroxidase) (Dalton et al., 1986; Evans et al., 1999; Becana et al., 2000; Matamoros et al., 2003). The roles of redox balance in the orchestration of the nodulation process have been reviewed previously (Puppo et al., 2005; Marino et al., 2009). In legumes, there is a strong positive correlation between nitrogenase activity and nodule ascorbate and (h)GSH contents (Dalton et al., 1993; Matamoros et al., 2003; Groten et al., 2005, 2006). Two major low-molecular-weight thiols, GSH and homoglutathione hGSH, are synthesized in Medicago truncatula roots. The pathway for GSH and hGSH synthesis involves two ATP-dependent steps. In the first reaction, γ-glutamylcysteine synthetase (γECS) catalyzes the formation of γ-glutamylcysteine from glutamate and cysteine and in the second reaction glycine or β-alanine is added to the C-terminal site of γ-glutamylcysteine by GSH synthetase (GSHS) or hGSH synthetase (hGSHS), respectively (Frendo et al., 1999, 2001). (h)GSH deficiency induced by pharmacological and transgenic approaches inhibits root nodule formation in M. truncatula, indicating that GSH and hGSH play a crucial role in this process (Frendo et al., 2005). Moreover, GSH synthesis deficiency in the bacteroid strongly impairs biological nitrogen fixation (BNF) (Harrison et al., 2005). However, to our knowledge, there is no report clearly showing that the modification of the plant (h)GSH pool affects BNF activity in the mature nodule.
The objective of our study was to determine the importance of (h)GSH in the functioning nodule by analyzing transgenic nodules with decreased or increased (h)GSH content in the nodule nitrogen-fixing zone. The promoter of Nodule Cysteine Rich family member 001 (NCR001), a member of the nodule-specific cysteine-rich protein family expressed in the nitrogen-fixing zone (Mergaert et al., 2003), was chosen to build genetic constructs allowing the modification of (h)GSH content in the nitrogen-fixing zone. The increases and decreases in (h)GSH content produced were correlated with significant increases and decreases in BNF, respectively. Moreover, BNF modifications were associated with changes in nodule gene expression level. Finally, the γECS-deficient nodules were smaller than control nodules, indicating that γECS deficiency disturbs nodule development.
Materials and Methods
Plant material, bacterial strains and growth conditions
Medicago truncatula Gaertn. ecotype Jemalong J5 and Sinorhizobium meliloti RCR2011 strain were grown as described in Frendo et al. (2005). The Agrobacterium rhizogenes strain Arqua1 was used for the induction of hairy roots (Quandt et al., 1993). The induction of transgenic hairy roots was performed using two different protocols. To test the nodulation efficiency of the different constructs, the in vitro method for inducing hairy roots on M. truncatula described by Boisson-Dernier et al. (2001) was used. For all other experiments, a modified protocol deriving from that of Vieweg et al. (2005) was used. The seedlings were planted in sand and fertilized with the culture medium described by Rigaud & Puppo (1975). All the plants used in this study were grown in a 16-h light (24°C): 8-h dark (20°C) photoperiod. The wild-type primary root was removed from the plant and one transgenic root per composite plant was selected by fluorescence detection of the T-DNA-expressed green fluorescent protein (GFP). Each biological repetition consisted of 15–20 composite plants harboring one transgenic root obtained from independent transformation events. Three different biological repetitions were analyzed for each construct. Composite plants expressing MtγECS-RNAi and control-RNAi constructs were analyzed 5 wk post inoculation. Composite plants overexpressing MtγECS and SmγECS and control overexpression constructs were analyzed 7 wk post inoculation.
GSH and hGSH quantification
Thiols were extracted with hydrochloric acid, derivatized with monobromobimane and quantified after separation on reverse-phase HPLC as described by Fahey & Newton (1987). Commercial GSH (Sigma) and γ-EC (LGC Standards, Molsheim, France) were used as standards. The hGSH used as a standard was synthesized by Neosystem (Strasbourg, France).
Nitrogenase activity assay
N2 fixation activity was determined by C2H2 reduction, using a gas chromatograph (Agilent GC6890N; Agilent Technologies, Massy, France). Nodulated plantlets were incubated at 25°C, in rubber-capped tubes containing 10% C2H2 in air (Hardy et al., 1968). The acetylene reduction assay was performed on single transgenic roots corresponding to one transformation event.
Construction of overexpression and RNAi-γECS vectors
Sinorhizobium meliloti γECS and M. truncatula γECS open reading frames were amplified by PCR using respectively Sm-γECS-F and Sm-γECS-R primers and Mt-γECS-F and Mt-γECS-R primers (Supporting Information Table S1). The two sequences were verified by sequencing and inserted into pDONRTM207 (Invitrogen). Finally, a multisite gateway destination vector, pKM43GWD (VIB, Ghent, Belgium), was used for overexpression constructs, pMtNCR001 was inserted as the 5′ element using the entry vector pENTL4L1-pMtNCR001, cECS inserts were obtained from pDONR-MtcECS or pDONR-SmcECS, and pENTR-R2-T35S-L3 was inserted as the 3′ element (VIB, Ghent, Belgium) using Gateway Technology (Invitrogen). Respectively, expression vectors were named pKM43GWD-MtγECS and pKM43GWD-SmγECS (Fig. S1). A control plasmid expressing GUS under the control of pMtNCR001, pKM43GWD-GUS, was obtained in the same way with PENTR-gusTM (Invitrogen) as an insert donor in the multisite gateway (Fig. S1).
For the RNAi construct, the cauliflower mosaic virus (CaMV) 35S promoter (P35S) in the pK7GWIWG2D(II),0 vector (VIB) was replaced by the MtNCR001 promoter (Mergaert et al., 2003). Following the nomenclature used for these binary vectors (http://gateway.psb.ugent.be), we named our construct pK7GWIWG5D(II), where 5 denotes the promoter pMtNCR001. SacI and SpeI restriction sites were added to pMtNCR001 by a PCR amplification with the pMtNCR001SacI-F and pMtNCR001SpeI-R primers, using pENTL4L1-pMtNCR001 as a template. The resulting 2634-bp PCR product was subcloned into the pGEM-T vector (Promega). The insertion of this promoter was performed in three sequential subcloning steps. First, a 2472-bp SacI-P35S:ccdB:intron-MluI from the pK7GWIWG2D(II),0 vector was subcloned into a modified ΔEagI pGEM-T vector without a SpeI site. Secondly, p35S was replaced by pMtNCR001 inserted into the SacI-SpeI sites. Finally, pK7GWIWG5D(II) was obtained by insertion of the SacI-pMtNCR001:ccdB:intron-MluI cassette back into the original pK7GWIWG2D(II),0 vector.
For the RNAi construct, a 532-bp PCR product was obtained by a PCR amplification of MtγECS with MtγECS-RNAi primers. The PCR product was cloned in a pDONRTM207 vector (Invitrogen). Then the MtγECS fragment was inserted into pK7GWIWG5D(II) using Gateway Technology (Invitrogen), and named pK7GWIWG5D(II)-MtγECS (Fig. S1). The control RNAi vector was built with a lacZ insert. A 294-bp ClaI-EcoRV fragment was digested from the pKOK5 plasmid (Kokotek & Lotz, 1989) and cloned into pENTR4. Then the lacZ insert was inserted into pK7GWIWG5D(II) using Gateway Technology (Invitrogen), and named pK7GWIWG5D(II)-lacZ (Fig. S1).
For light microscopy, nodules were fixed in 1% glutaraldehyde and 4% formaldehyde in 0.05 M phosphate buffer (pH 7.2), washed, dehydrated, embedded in Technovit 7100 (Kulzer Histo-Technik, Heraeus France, Villebon, France) according to the manufacturer’s instructions, and stained with toluidine blue (Van de Velde et al., 2006). Nodules were cut on a microtome to produce 5-μm slices and mounted.
Total RNA was extracted with Trizol (Invitrogen). cDNA was synthesized with the Omniscript RT kit (Qiagen) following the manufacturer’s protocol using 2 μg of total RNA. Quantitative RT-PCR was performed using a DNA Engine Opticon 2 Continuous Fluorescence Detection system (MJ Research, Biorad, Marnes-la-Coquette, France) and a qPCR MasterMix Plus for SYBR® green I (Eurogentec, Angers, France). In each reaction, 5 μl of 100-fold-diluted cDNA and 4.5 pmol of each primer (sequences used are described in Table S1) were used in 15 μl. The initial denaturing time was 10 min, followed by 40 PCR cycles of 95°C for 15 s and 60°C for 1 min. The specificity of the amplification was confirmed by the presence of a single peak in a dissociation curve at the end of the PCR reaction. Data were quantified using Opticon Monitor 2 (MJ Research) and normalized with the 2−ΔΔCT method (Livak & Schmittgen, 2001). Two constitutively expressed genes, Mtc27 and the 40S ribosomal protein S8 gene, were the endogenous controls used for the 2−ΔΔCT normalization method (Van de Velde et al., 2006). Two RNA samples corresponding to nodules from three to four transgenic roots were analyzed by qRT-PCR for each biological experiment. Each qRT-PCR reaction for each of the three biological experiments was performed in triplicate.
Promoter fusion and histochemical localization of GUS activity
A 1.5-kb DNA fragment upstream of the starting ATG of γecs (Medtr4g149470.1), gshs (Medtr7g139670.1) and hgshs (Medtr7g139660.1) was amplified by PCR using gene-specific primers (Table S1) with the following restriction sites: EcoRI/XhoI for γECS and BamHI/NotI for GSHS and hGSHS. The DNA fragments were cloned into the pGEM-T vector (Promega) according to the manufacturer’s instructions and verified by sequencing. Using the restriction sites, the fragments were subcloned into the entry vector pENTR4 (Gateway Cloning System; Invitrogen) and then fused to the GUS reporter gene in the destination vector pkGWFS7 (Karimi et al., 2002). The identity, integrity and orientation of the promoter fragment were checked by sequencing. The destination vector carrying the gus fusion was electroporated into the A. rhizogenes strain Arqua1. Medicago truncatula roots were transformed with A. rhizogenes by the method described previously (Boisson-Dernier et al., 2001). Transformed roots were selected on kanamycin plates and transferred onto modified Fahraeus medium as described previously (Frendo et al., 2005). Transgenic roots were stained with GUS assay buffer as described by Hemerly et al. (1993). Roots from at least 20 plants from three biological experiments were examined.
Our data were reported as mean ± standard error. The significance of the results was assessed using the Mann–Whitney nonparametric test which allows the comparison of small quantitative samples. When the number of samples was enough to define a normal distribution (30 samples), the results were assessed using Student’s t-test.
Localization of the expression of γECS, GSHS and hGSHS in nodules
The spatial expression patterns of γECS, GSHS and hGSHS were investigated in nodules by promoter:GUS fusion analysis. The 1.5-kb regions upstream of MtγECS, MtGSHS and MthGSHS were isolated using the available M. truncatula genomic data (http://www.medicago.org/genome and http://www.ncbi.nih.gov/) and fused to the uidA gene, and a histochemical analysis of GUS activity was preformed in 3-wk-old transgenic nodules.
GUS expression from the pMtγ-ECS:uidA fusion was mainly observed in the meristematic zone (zone I), the infection zone (zone II) and the beginning of the fixation zone (zone III) of the nodule (Fig. 1a). Nodule sections confirmed the localization of the GUS staining (Fig. 1b). pMtγECS:uidA expression was also detected in the root vascular tissue. Analysis of the spatial localization of MtGSHS expression, using the pMtGSHS:uidA fusion, showed that GUS activity was mainly located in vascular tissue and the cortex of the nodule (Fig. 1c). Nodule sections showed that GUS staining was also present in nodule zones II and III (Fig. 1d). The expression of the pMthGSHS:uidA fusion was detected in the vascular tissue and in the cortex of the nodule (Fig. 1e). Nodule cross-sections revealed that GUS staining was also present in zone I (Fig. 1f). γECS expression was higher during the infection process, during the differentiation of bacteroids and at the beginning of the fixation process. Moreover, GSHS expression may have been higher than hGSHS expression in the infection and nitrogen-fixing zones.
Production of chimeric constructs allowing depletion and accumulation of γECS in the nodule nitrogen-fixing zone
Involvement of (h)GSH in the first step of nodule formation has already been demonstrated (Frendo et al., 2005). In order to investigate the importance of (h)GSH in the nitrogen fixation process, we modified the (h)GSH pool in zone III using transgenic approaches. A promoter able to drive strong, nodule zone III-specific expression was selected to express our overexpression and RNAi constructs. The promoter of NCR001, a member of the nodule-specific cysteine-rich protein family, was chosen for its specific activity in nodule zone III (Mergaert et al., 2003). The spatial activity pattern of the MtNCR001 promoter (pMtNC001) was verified in nodules by promoter:GUS fusion analysis. The 2.6-kb region upstream of MtΝCR001 was fused to the uidA gene, and histochemical analysis of GUS activity was performed in 3- and 6-wk-old nodules (Fig. S2a). In 3-wk-old nodules, the expression of the pMtNCR001:uidA fusion was restricted to the nitrogen-fixing zone. Similarly, in 6-wk-old nodules, GUS activity was only detected in nodule zone III. However, stronger GUS staining was detected in nodule zone III near zone II. The threshold cycle (CT) value of pMtNCR001, which is indicative of the gene transcription efficiency, was verified by qRT-PCR analysis and compared with that of other genes expressed in nodules (Fig. S2b). The MtNCR001 and leghemoglobin (LEG) gene CT values were similar, showing that MtNCR001 transcripts are strongly accumulated in nodules. The high activity of pMtNCR001 allows us to build genetic constructs allowing the modulation of (h)GSH in the nitrogen-fixing zone without affecting the initiation of nodule development.
Effect of (h)GSH content modulation on BNF and expression of genes involved in nodule metabolism
To determine the importance of (h)GSH in root nodule functioning, RNAi-silenced lines for the MtγECS gene in nodule zone III were analyzed for their BNF and the expression of nodule-specific genes. Composite plants with transgenic roots expressing a γECS-RNAi construct under the control of pNCR001 were obtained. As a control, composite plants with transgenic roots expressing a bacterial β-galactosidase (lacZ) RNAi construct under the control of pNCR001 (control-RNAi plants) were generated. To verify that this γECS-RNAi construct did not alter the nodulation efficiency of the transgenic roots, the number of nodules was analyzed at 7 and 14 days post inoculation (dpi) and compared with the number of nodules on control-RNAi roots. No significant difference was observed between the two populations, indicating that the γECS-RNAi construct does not affect the early steps of the nodulation process (Fig. S3). To investigate the specificity and efficiency of the γECS-RNAi construct, total GSH and hGSH concentrations were quantified in leaves, roots and nodules of composite γECS-RNAi plants and compared with the GSH and hGSH concentrations of control-RNAi composite plants (Fig. 2a). As expected, the root and leaf (h)GSH pools of γECS-RNAi plants were similar to those of control-RNAi plants. By contrast, the GSH and the hGSH concentrations were reduced respectively by 45% and 37% in transgenic γECS-RNAi nodules compared with the control-RNAi nodules. This result shows that the RNAi-γECS construct is able to efficiently reduce the (h)GSH pool in the nodule.
To study the effect of (h)GSH nodule deficiency on BNF, acetylene reduction activity (ARA) was quantified for γECS-RNAi and control-RNAi roots at 4 wk post S. meliloti infection. The ARA was expressed per nodule number (ARAn) and per mg of fresh weight (ARAfw) (Fig. 2b). The ARAn was significantly reduced, by 25%, in γECS-RNAi nodules compared with the control-RNAi nodules. This result shows that (h)GSH deficiency has a significant impact on BNF per nodule. By contrast, the ARAfw was similar in γECS-RNAi nodules and control-RNAi nodules. The alteration of ARAn without a significant change in ARAfw suggests that there is a modification of the ratio fresh weight of nodule:number of nodules for γECS-RNAi nodules compared with the control-RNAi nodules. To test this hypothesis, the ratio fresh weight:number of nodules was calculated for the transgenic nodule samples (Fig. 3a). In parallel, the size of the nodules was estimated by microscopic analysis (Fig. 3b). The weight of γECS-RNAi nodules was significantly lower than that of control-RNAi nodules. This lower weight of the γECS-RNAi nodules was associated with their smaller size (Fig. 3b,c). Taken together, these results show that (h)GSH deficiency affects not only BNF but also the development of the nodule.
In order to determine whether a strong (h)GSH deficiency may significantly alter the ARAfw, the data obtained for γECS-RNAi nodules were split into two groups as a function of their (h)GSH content. The first group represented 75% of the samples and showed the highest content of (h)GSH among the γECS-RNAi nodules (H group) (Fig. 4a). The second group represented 25% of the samples and showed the lowest content of (h)GSH among the γECS-RNAi nodules (L group). The (h)GSH content and the ARAfw of the two groups were compared with those of control-RNAi nodules. The (h)GSH content was respectively 30% and 75% lower in the H and L groups than in control-RNAi nodules. In the H group, the ARAfw was not significantly modified in comparison with the ARAfw observed in control-RNAi nodules, suggesting that a 30% reduction of the (h)GSH content does not affect significantly BNF per nodule fresh weight, the specific BNF (Fig. 4b). By contrast, the ARAfw of the L group was significantly reduced, by 64%, compared with both control-RNAi nodules and the H group. This result shows that a 75% depletion of (h)GSH significantly affects the specific BNF in nodules.
In order to determine whether the lower BNF efficiency observed in L group nodules could be correlated to lower expression of genes associated with nodule functioning, expression analysis of the sucrose synthase-1 (SUS1), LEG, NCR001 and thioredoxin S1 (TrxS1) genes was performed using qRT-PCR (Fig. 5). SUS1 is involved in the regulation of carbohydrate supply in nodules and is regulated at the transcriptional level (Baier et al., 2007; Marino et al., 2008). LEG is a specific zone III gene marker which is crucial for nitrogen fixation efficiency (Mergaert et al., 2003; Ott et al., 2005) and LEG concentration is associated with nitrogen fixation efficiency (Gogorcena et al., 1995, 1997; Ott et al., 2005; Marino et al., 2006; Lopez et al., 2008). As mentioned previously (Mergaert et al., 2003), NCR001 is a nodule-specific gene marker expressed in the nitrogen-fixing zone. TrxS1 is a thioredoxin specifically expressed during the interaction between M. truncatula and S. meliloti (Alkhalfioui et al., 2008). The expression of γECS, GSHS and hGSHS, which are involved in the (h)GSH synthesis pathways, was also analyzed to investigate the effect of lower expression of γECS on the (h)GSHS expression level. hGSHS, SUS1 and NCR001 transcript levels were similar in the γECS-RNAi and control-RNAi nodules. By contrast, transcript levels corresponding to γECS, LEG and TrxS1 were significantly lower in the γECS-RNAi than in the control-RNAi nodules. The lower expression of TrxS1 and LEG shows that the lower (h)GSH level in γECS-RNAi is also correlated with lower expression of marker genes linked to nodule development and functioning. Surprisingly, the GSHS transcript level was significantly higher in the γECS-RNAi nodules than in the control-RNAi nodules, suggesting that down-regulation of (h)GSH metabolism induces GSHS expression.
To investigate the effect of an increased (h)GSH concentration in the nitrogen-fixing zone, transgenic composite plants overexpressing M. truncatula γECS cDNA (MtγECS) and S. meliloti γECS cDNA (SmγECS) under the control of pNCR001 were constructed and analyzed. SmγECS was used to avoid the potential post-traductional regulation of the plant γECS which may reduce the efficiency of the gene overexpression. The levels of (h)GSH and BNF were measured using ARA assays in 7-wk-old nodules. The analysis of control and γECS-overexpressing transgenic lines (80 control lines, 60 MtγECS-overexpressing lines and 35 SmγECS-overexpressing lines) allowed us to detect higher GSH content in γECS-overexpressing nodules compared with control nodules. BNF was measured in 40 γECS-overexpressing composite plants showing a 30% higher nodule GSH level compared with control nodules (Fig. 6). In both Mt/SmγECS-overexpressing nodules, GSH content was much higher than in control nodules. By contrast, no difference was observed in hGSH content. These results show that γECS overexpression in nodule zone III significantly modifies GSH content without affecting the hGSH content. The higher GSH content was correlated with a significantly higher ARA detected in γECS-overexpressing nodules compared with control nodules (Fig. 6a,b). This result indicates that a higher GSH level has a positive effect on BNF efficiency.
In order to assess whether the higher BNF efficiency observed in Mt/SmγECS-overexpressing nodules could be correlated with higher expression of genes induced in nodule functioning, expression analysis of SUS1, LEG, NCR001 and TrxS1 was performed using qRT-PCR (Fig. 7). Transcript levels corresponding to GSHS, hGSHS and TrxS1 were similar in Mt/SmγECS-overexpressing and control plants. By contrast, transcript levels corresponding to SUS1, LEG and NCR001 were higher in γECS-overexpressing nodules than in the control. Thus, overexpression of γECS in nodules allows higher nitrogen fixation activity, correlated with higher expression of genes involved in nodule functioning.
In this study, the expression localization of the genes involved in the (h)GSH pathway was examined and the physiological importance of (h)GSH content in BNF was studied by investigating the physiological and molecular characteristics of nodules presenting modified (h)GSH contents. The differential spatial expression patterns observed for genes involved in (h)GSH synthesis suggest that GSH and hGSH synthesis could be differentially regulated in the different nodule tissues. The γECS expression level was higher in zone I, zone II and the beginning of zone III. The higher γECS expression level observed in zones I and II may be linked to the meristematic activity in zone I and to the modifications of nodule cell size and metabolism in zone II to allow the settlement of the bacteria into the plant cell. Indeed, (h)GSH accumulation has been observed in meristematic cells and in trichome cells in which endoreduplication occurs as in nodule zone II (Sanchez-Fernandez et al., 1997; Gutierrez-Alcala et al., 2000; Vinardell et al., 2003). The higher GSHS and hGSHS expression levels observed in the parenchyma may be important for the efficiency of the O2 diffusion barrier through the ascorbate/glutathione cycle. Indeed, Dalton et al. (1998) proposed that O2 entrance is regulated by adjustment of respiratory activity, with ascorbate peroxidase functioning as a scavenger of H2O2 generated by respiration. These results suggest that aerobic respiration in the parenchyma may restrain the diffusion of O2 into the nodule (Becana et al., 2000). The accumulation of superoxide and H2O2 in the nodule parenchyma observed in pea (Pisum sativum) nodules strengthens this hypothesis (Groten et al., 2005). Spatial expression pattern analysis suggested that in nodule zone III GSHS expression is higher than that of hGSHS. Furthermore, the finding of a higher nodule GSH content in zone III γECS-overexpressing nodules but no change in the hGSH concentration strengthens this hypothesis. These results are in agreement with the GSHS and hGSHS enzymatic activities detected in pea nodule zone III (Matamoros et al., 1999). Finally, the significantly higher expression of GSHS in γECS-RNAi than in the RNAi-control nodules shows that GSHS expression is affected by (h)GSH deficiency, whereas modification of hGSHS expression was not detected in the same conditions. Taken together, our results are in favor of tissue-specific differential regulation of GSH and hGSH synthesis in M. truncatula nodules. This differential regulation of GSH and hGSH metabolism pathways suggests that these two low-molecular-weight thiols may play different roles in M. truncatula nodules. Multiple publications have already shown the difference of GSH/hGSH accumulation during seedling development or heavy metal treatment (Frendo et al., 1999; Harrison et al., 2003). Similarly, differential regulation of GSH and hGSH metabolism by nitric oxide (NO) has been observed in M. truncatula roots (Innocenti et al., 2007). Thus, the high NO content detected in nodule zone III (Baudouin et al., 2006) may also play a role in the differential GSH/hGSH metabolism in the nitrogen-fixing zone.
The development of genetic tools with which to modulate the level of (h)GSH specifically in nodule zone III represents a step forward in elucidating the importance of these molecules in the nitrogen fixation process without disturbing the initial steps of the symbiotic interaction. The correlation between γECS expression level, GSH content and ARA for both RNAi and overexpressing transgenic nodules demonstrates that (h)GSH plays a crucial role in the regulation of BNF. The higher expression of SUS1, LEG and NCR001, which are development and functioning nodule markers, in γECS-overexpressing nodules corroborates the link between GSH content and BNF. In parallel, LEG and TrxS1 mRNA levels were lower in the γECS-RNAi than in control-RNAi nodules. The regulation of the SUS1, NCR001 and TrxS1 genes in γECS-overexpressing and γECS-RNAi nodules suggests that increased or decreased (h)GSH content modify different metabolism processes, resulting in differential gene expression responses. Among the different nodule marker genes, LEG expression appears to be correlated with both increases and decreases in (h)GSH content and BNF. LEG, which is involved in the bacteroid oxygen supply, has an important role in supporting energy production via the respiration process and is needed for nitrogenase activity (Ott et al., 2005). The modulation of the LEG expression level in γECS transgenic nodules suggests that the respiration rate and energetic metabolism are modified by (h)GSH content, with higher rates in γECS-overexpressing nodules and lower rates in γECS-RNAi nodules. Similarly, the higher SUS1 expression level in γECS-overexpressing nodules also suggests that higher carbon nutrition is needed in these conditions, as SUS1-sucrose breakdown delivers the energy required for BNF via hexose supply to nodule metabolism (Baier et al., 2007). The expression analysis of NCR001 and TrxS1, which are transcriptionally regulated in nodule zone III and during nodule development, respectively, showed that NCR001 and TrxS1 transcript accumulation is regulated under our conditions. To our knowledge, the role of this thioredoxin in nodule formation and functioning has not been defined. Our results showing that lower (h)GSH content and lower BNF efficiency are associated with lower TrxS1 expression demonstrate that its expression is impaired under altered nodule functioning conditions. In parallel, higher (h)GSH content and higher BNF efficiency were associated with significantly increased NCR001 transcript accumulation. NCR peptides have been shown to be involved in differentiation of bacteria into symbiosomes (Van de Velde et al., 2010). However, the high number of NCR peptides (> 300) and their various expression localizations in nodules suggest additional roles for these peptides. Our results suggest that the expression of NCR001 is not only regulated by nodule development but also modified by physiological conditions in the nodule. Nevertheless, modification of the NCR001 and TrxS1 expression could be due to multiple reasons as changes in (h)GSH accumulation, nodule development and BNF efficiency are associated with these modifications. Further experiments are needed to clarify the regulation of these nodule-specific genes under our conditions.
Our results demonstrate that plant (h)GSH content modulates BNF and support a role of (h)GSH in the regulation of this biological process, as already established for the bacterial symbiotic partner S. meliloti (Harrison et al., 2005). In plants, GSH is involved in both gene transcriptional regulation via GSH-regulated transcription factors and post-traductional protein regulation by glutaredoxins (Rouhier et al., 2008; Foyer & Noctor, 2009). Thus, (h)GSH is involved in both the regulation of gene expression and the post-traductional regulation of enzyme activities. In leaves, carbon metabolism is highly regulated by redox state via thioredoxins and glutaredoxins (Meyer et al., 2009; Kotting et al., 2010). In nodules, plant carbon metabolism is crucial to support both plant and microsymbiont metabolic requirements. Thus, modification of (h)GSH content may change both glutathionylation and the cysteine redox state of the proteins regulating their activities. This redox regulation is potentially illustrated by the modification of starch metabolism during the symbiotic interaction (Harrison et al., 2005; Redondo et al., 2009).
Moreover, the development of γECS-RNAi nodules was affected, suggesting that the deficiency of (h)GSH in nodule zone III may modify nodule growth. The reduction of nodule growth has previously been associated with deficiency in nitrogen fixation (Mitra & Long, 2004; Ott et al., 2009). Thus, the reduction of nodule growth under (h)GSH depletion may be linked to modification of BNF efficiency. Alternatively, (h)GSH may be directly involved in nodule growth regulation. Indeed, GSH has been shown to regulate auxin transport and root growth (Bashandy et al., 2010; Koprivova et al., 2010).
In conclusion, we report that (h)GSH concentration regulates BNF efficiency in nodules and a deficiency in (h)GSH impairs nodule growth. Whether these modifications in nitrogen fixation and nodule development are directly or indirectly linked to the (h)GSH content still needs to be determined. Transcriptomic, proteomic and metabolomic analysis of transgenic plants expressing the γECS-chimeric constructs will be useful to answer these questions.
We thank Janice de Almeida Engler and Gilbert Engler for their help during microscopy analysis. We also thank Peter Mergaert and Alain Puppo for providing the NCR001 promoter and for critical reading of the manuscript, respectively. We are indebted to the symbiosis team members for numerous nodule harvests. We gratefully acknowledge Stephanie Piardi for plant care. S.E.M. is a recipient of an IMAGEEN fellowship from the European Communities and a fellowship from the Tunisian government (bourse d’alternance).