Nitrate is an essential element for plant growth, both as a primary nutrient in the nitrogen assimilation pathway and as an important signal for plant development. Low- and high-affinity transport systems are involved in the nitrate uptake from the soil and its distribution between different plant tissues. By an in silico search, we identified putative members of both systems in the model legume Lotus japonicus. We investigated, by a time course analysis, the transcripts abundance in root tissues of nine and four genes encoding putative low-affinity (NRT1) and high-affinity (NRT2) nitrate transporters, respectively. The genes were sub-classified as inducible, repressible and constitutive on the basis of their responses to provision of nitrate, auxin or cytokinin. Furthermore, the analysis of the pattern of expression in root and nodule tissues after Mesorhizobium loti inoculation permitted the identification of sequences significantly regulated during the symbiotic interaction. The interpretation of the global regulative networks obtained allowed to postulate roles for nitrate transporters as possible actors in the cross-talks between different signalling pathways triggered by biotic and abiotic factors.
As sessile organisms, plants have developed sophisticated mechanisms to ensure appropriate adaptations to constantly changing environmental conditions. One example is the uptake of nitrate from soil, a critical process controlled by complex regulatory networks underlying both external and internal cues that modulate the uptake capacity in accordance with plant nutrient demand and soil nitrate availability. Plant roots have three different uptake systems to cope with low or high NO3- concentrations in soil, the high-affinity (constitutive and inducible HATS) and low-affinity NO3- uptake systems (LATS). The LATS generally has a larger capacity than does the HATS, and mediates nitrate uptake under high-nitrate environments (Crawford & Glass 1998), while HATS appears to play a major role in nitrate uptake when nitrate concentrations in the soil are very low (<250 µm; Crawford & Glass 1998).
Two families of NO3- transporters, NRT1 and NRT2, are generally thought to be responsible for LATS and HATS for both nitrate uptake and its distribution within the plant (Forde 2002; Miller et al. 2007). The only known exceptions are represented by the Arabidopsis thaliana and Medicago truncatula proteins AtNRT1.1 (CHL1) and MtNRT1.3, which display both low and high affinities (Liu, Huang & Tsay 1999; Morère-Le Paven et al. 2011). NRT1 members share sequence similarity with peptide transporters (PTR) and are considered part of the NRT1(PTR) superfamily. However, NRT1 and PTR members are functionally distinct as they act specifically in the transport of either nitrate or peptides, respectively. No significant sequence homology is found between the NRT1(PTR) and NRT2 family members. NRT1 and NRT2 are further distinguished from each other as the latter may require for nitrate transport activity an additional component, called NAR2 (Tsay et al. 2007).
A complete functional characterization of several A. thaliana nitrate transporters has recently revealed specialized roles in the control of nitrate flux from soil to root tissues and in different steps of nitrate flux throughout the whole plant body. AtNRT1.1, AtNRT1.2, AtNRT2.1, AtNRT2.2 and AtNRT2.4 are known to be involved in nitrate uptake from the soil solution into root cells (Tsay et al. 1993; Wang, Liu & Crawford 1998; Cerezo et al. 2001; Filleur et al. 2001; Little et al. 2005; Orsel et al. 2006; Li et al. 2007; Kiba et al. 2012). Regarding the nitrate movement between different plant tissues, NRT1.5 mediates nitrate efflux from root cells and is responsible for loading nitrate into the xylem for root-to-shoot nitrate transport (Lin et al. 2008); AtNRT1.8 is responsible for nitrate retrieval from xylem sap (Li et al. 2010); AtNRT1.7 is involved in the remobilization of nitrate from older to younger leaves (source to sink tissues) mediated by the phloem (Fan et al. 2009); AtNRT1.9 and AtNRT2.4 mediate nitrate loading into the root and shoot phloem, respectively (Wang & Tsay 2011; Kiba et al. 2012); AtNRT1.6 is involved in the transport of NO3- from maternal tissue to developing embryos (Almagro, Lin & Tsay 2008).
Consistent with their functional specialization, regulation of transporter expression in response to different signals as well as patterns of tissue localization varies among members of both NRT1(PTR) and NRT2 gene families. These genes are differentially responsive to NO3- and can be assigned to nitrate-constitutive, nitrate-repressible and nitrate-inducible categories (Okamoto, Vidmar & Glass 2003; Wang et al. 2003; Cai et al. 2008). Furthermore, they show different responses to diurnal cycle, pH, temperature, exogenous ammonium and nitrogen starvation conditions (Fan et al. 2009; Feng et al. 2011; Kiba et al. 2012). In the cases of AtNRT1.1 and AtNRT1.8 genes, a detailed molecular characterization showing a strong up-regulation in response to signals such as auxin and cadmium, respectively, has been reported (Guo, Wang & Crawford 2002; Li et al. 2010).
Plant adaptations to environmental modifications include the capacity to respond to changes in soil nutrient availability by modulating their root system developmental plan. NO3- is able to trigger signalling pathways modulating lateral root development both systemically and locally. Accumulation of NO3- (or its metabolites) in the shoot is responsible for triggering an inhibitory pathway leading to a block in root elongation soon after emergence from the primary root. In contrast, low, external local NO3- levels lead to the stimulation of mature secondary root elongations (Zhang & Forde 1998). The latter mechanism, responsible for preferential root colonization of NO3--rich patches of the soil, was shown to involve the nitrate transporters NRT1.1 and NRT2.1 (Little et al. 2005; Remans et al. 2006). NRT1.1 senses a wide range of nitrate concentrations in the soil through a phosphorylation switch at Thr101 (Ho et al. 2009). More recently, NRT1.1 has been reported to act as an auxin transport facilitator defining a possible mechanism for connecting nitrate and auxin signalling pathways, mediating secondary root elongation (Krouk et al. 2010).
Legumes are one of the largest family of flowering plants that occupy about 15% of earth's arable surface, providing 27% of the world's primary crop production (Graham & Vance 2003). Their capacity to establish a mutualistic symbiotic interaction with rhizobia, generating ammonia de novo from nitrogen gas (N2), makes legumes the major natural N-provider to the ecosystem (roughly 200 million ton of N per year; Kinkema, Scott & Gresshoff 2006). Symbiotic nitrogen fixation in legumes takes place in specialized organs called root nodules, which develop as a result of a series of signal exchanges between plant root cells and rhizobia in the soil (Oldroyd & Downie 2008). Although legume nodules and lateral roots are structurally and developmentally distinct, several analogies can be identified (Ferraioli et al. 2004). As in the case of secondary root development, auxin and cytokinin play major roles in nodule formation and development (Ding & Oldroyd 2009). Similarly, nitrate availability in the soil is known to strongly affect nodule formation as well as development and functioning (Bisseling, van den Bos & van Kammen 1978; Carroll & Mathews 1990; Fujikake et al. 2003; Barbulova et al. 2007; Fey & Vessey 2009). Specifically, the formation of root nodules occurs only when mineral nitrogen in the soil is limiting for plant growth. As in the case of secondary root developmental control, the effect of nitrate on nodule formation is exerted through both local and systemic controls (Omrane & Chiurazzi 2009; Mortier, Holsters & Goormachtig 2012). Finally, high ammonium nitrate concentrations in the media can even affect the competence of legume plants for nodulation before rhizobia inoculation, probably by modifying the N nutritional status of the plants (Omrane et al. 2009).
We present here the first molecular characterization of NRT1(PTR) and NRT2 families in a legume plant. The profile of expression of a subset of the L. japonicus genes identified by an in silico search has been analysed in different conditions to obtain a sub-classification based on the different type of responses to specific biotic and abiotic signals. This analysis also allowed the identification of genes that are specifically expressed during the symbiotic interaction with Mesorhizobium loti, suggesting an unexpected involvement of this class of transporter on nodule organogenesis and function.
MATERIALS AND METHODS
Plant and bacterial material
The Lotus japonicus ecotype GIFU B-129-F12 was used for all the described experiments. The M. loti strain R7A was used for the inoculation experiments and was grown in liquid TY-medium supplemented with rifampicin (20 mg L−1). The M. loti nifH- strain was kindly provided by Dr Clive Ronson (University of Otago, New Zealand) and was grown in the same medium supplemented with rifampicin and gentamicin (10 mg L−1).
Plant growth conditions
Lotus seeds surface sterilized for 20 min in 25% commercial bleach and 0.1% Triton, were germinated on H2O agar and after 5 d seedlings were transferred on Gamborg derivative B5 solid media (Duchefa, G0209.0025, Haarlem, the Netherlands). The media had the same composition as B5 media, except that (NH4)2SO4 and KNO3 were omitted and replaced with 1 mm glutamine. Additions of 10 µm indole-3-acetic acid (IAA) or 6-benzylaminopurin (BAP), or KNO3 at different concentrations were also made as indicated in the text. In the case of M. loti inoculation, N sources were completely omitted. The defined media contained Gamborg B5 vitamins (Duchefa, G0415.0250) and the pH was adjusted to 5.7 with 2-(N-morpholino) ethanesulfonic acid (MES; Duchefa, MI503.0250). Plants were cultivated in a growth chamber with a light intensity of 200 µmol m−2 s−1 at 23 °C with a 16 h/8 h day/night cycle.
The growth conditions and the M. loti inoculation procedure (107 cells per root tip) for the in vitro nodulation assay have been reported elsewhere (Barbulova et al. 2005).
The prom-CM0826.370-gusA construct was obtained by PCR amplification of genomic DNA fragment using two specific oligonucleotides: 5′-GCGTCGACCAACCGCAATTTAATTTGAATT-3′ (containing the SalI site); and 5′-CGGGATCCAGAAAAACAACAACAAAACCAT-3′ (containing the BamHI site). The purified PCR fragment was SalI-BamHI double digested and ligated into the pBI101.3 binary vector to obtain a translational fusion with the gusA reporter gene (Jefferson 1987). The resulting construct was introduced in Agrobacterium rhizogenes MSU440 strain (Martirani et al. 1999) by electroporation.
Total RNA was prepared from roots of Lotus plants by following the procedure reported by Kistner & Matamoros (2005). Total RNA (0.5 µg) was annealed to oligo-dT and reverse transcribed by using RT (Qiagen, 205311, Valencia, CA, USA) to obtain cDNA. Preliminary RT-PCR trials were first performed to determine the specificities of the gene primers and the presence of gene expression (amplicon size, sequence and approximate expression levels among the samples). All amplified fragments were inserted in pCR2.1 cloning vector (Invitrogen, Carlsbad, CA, USA) and confirmed by sequencing. Then, the number of cycles exploited for a linear range of gene amplification in the RT-PCR reactions was experimentally determined on the adjusted cDNA, for all the L. japonicus genes tested (Supporting Information Fig. S1). The ubiquitin (UBI) gene (AW719589) was used as an internal standard. Non-reverse transcribed RNAs were also included in each PCR reaction to confirm the absence of genomic DNA. Reaction products were electrophoresed on an agarose gel and images of stained DNAs were acquired. Densitometry values were measured using the Image Quant Software 5.2 (Molecular Dynamics, Carlsbad, CA, USA) and Typhoon 9200 Imager (GE Healthcare, Piscataway, NJ, USA). Two independent plant cultivations for each time course experiment were conducted and all samples were analysed three times by this technique with very consistent results. The sequences of all primers used in this work, the expected amplicon sizes and the annealing temperatures followed in the PCR reactions are listed in Supporting Information Table S2.
Histochemical GUS analysis
Histochemical staining of whole plant material was performed as described by Jefferson (1987). For sections, whole roots were stained and then fixed with 4% paraformaldehyde, 0.25% glutaraldehyde in 50 mm KPO4 buffer, 5 mm EGTA, 10 mm DTT, pH 7.2 and stored at 4 °C. The tissues were then washed with 50 mm KPO4 buffer pH 7.2. Fixed tissue segments were embedded in 6% agar and 60 µm sections obtained with a vibratome (Leica VT1000S, Wetzlar, Germany). The sections were analysed with a light microscope by means of dark- and bright-field optics.
Retrieval of L. japonicus NRT1(PTR) and NRT2 genes
The evolutionary histories were inferred using the neighbour-joining method (Saitou & Nei 1987). The evolutionary distances were computed using the Poisson correction method (Zuckerkandl & Pauling 1965) and are in the units of the number of amino acid substitutions per site. The analyses involved 93 and 15 amino acid sequences for NRT1(PTR) and NRT2 families, respectively. All positions containing gaps and missing data were eliminated. Evolutionary analyses were conducted in MEGA5 (Tamura et al. 2007).
Identification, structural genomic organization and evolutionary relationships of the L. japonicus NRT1(PTR) and NRT2 members
A remarkable feature of the NRT1(PTR) family is the number of NRT1(PTR) genes in higher plants. Arabidopsis and rice have 53 and 80 NRT1(PTR) members, respectively, that have been classified in four subgroups by phylogenetic analysis (Tsay et al. 2007). We searched for NRT1(PTR) members in the L. japonicus genomic database (see Materials and Methods) and identified 37 NRT1(PTR) genes (Supporting Information Table S1). Our search was not extended to partial sequences identified in the EST and genome databases, showing significant homology with the NRT(PTR) family. Hydrophobicity analyses performed with different servers (Hofmann & Stoffel 1993; Claros & von Heijne 1994; Tusnàdy & Simon 2001), predicted in 30 out of the 37 Lotus NRT(PTR) members, the presence of 12 transmembrane TM spanning regions with a large hydrophilic loop between TM 6 and 7, which is a feature of the NRT1(PTR) proteins in higher plant (Tsay et al. 2007; Supporting Information Table S1). L. japonicus NRT1(PTR) genes are found on five of the six chromosomes, the lone exception being chromosome 5. Most were found on chromosomes 1, 2 and 4. At least six gene clusters were identified that co-localize to the same contigs with duplicated paralogous genes and a minimal intergenic region estimated to be 2182 bp (chr1.0295.970−chr1.0295.980).
Currently, no sequence feature has been identified that distinguishes NRT1(PTR) family nitrate transporters from those that transport peptides. To identify those L. japonicus members most likely to act as nitrate transporters, we conducted a phylogenetic analysis that included the A. thaliana and L. japonicus NRT1(PTR) members along with characterized NRT1(PTR) nitrate transporters (Fig. 1). So far, 11 NRT1(PTR) genes from A. thaliana (Tsay et al. 2007), and single genes from Brassica napus (Zhou et al. 1998), Oryza sativa (Lin et al. 2000) and M. truncatula (Morère-Le Paven et al. 2011), have been proven to encode nitrate transporters in Xenopus oocyte functional studies. Most of these NRT1 proteins belong to subgroup I (Tsay et al. 2007). The phylogenetic tree was constructed by using the MEGA5 package (Tamura et al. 2007) and the obtained grouping was taken as an insightful clue to start a preliminary characterization of the L. japonicus NRT1(PTR) genes (Fig. 1). The L. japonicus members are found in all four subgroups previously identified. Nine L. japonicus members with higher level of homology with the NRT1 characterized proteins were further analysed (boxed in Fig. 1).
In the case of the NRT2 family, our search of the L. japonicus genome database led to the identification of four putative members. This number is consistent with previous studies in O. sativa and A. thaliana where four and seven putative NRT2 genes have been found (Tsay et al. 2007). Hydrophobicity analysis predicted the presence of 12 transmembrane TM spanning regions in all the four Lotus NRT2 members (Supporting Information Table S1). The four genes were distributed on three chromosomes: CM0001.20 on chromosome 1, CM0649.30 and CM0649.40 on chromosome 3, and CM0161.180 on chromosome 4. A phylogenetic tree of NRT2s was created after alignment of the NRT2 members from A. thaliana, O. sativa and the four L. japonicus sequences identified here (Fig. 2). The Lotus paralogs CM0649.30 and CM0649.40 are located in the same clade as the AtNRT2.1 and AtNRT2.2 proteins (Fig. 2). The 1593 bp cDNA sequences of the CM0649.30 and CM0649.40 genes, located 3800 bp apart from each other in a tail-to-tail orientation, share 95% of nucleotide identity. This tandem gene arrangement, with the same three exons structure, is also found in the A. thaliana genome, where the AtNRT2.1 and AtNRT2.2 genes share 90% nucleotide identity and are located 900 bp apart from each other. Finally, the same tandem configuration, but not gene structures, is also found in the rice genome for the OsNRT2.1 and OsNRT2.2 genes, where an intervening coding sequence, not observed in A. thaliana and L. japonicus, separates the two NRT2 genes (Cai et al. 2008). Therefore, we decided to name the L. japonicus sequences CM0649.40 and CM0649.30, LjNRT2.1 and LjNRT2.2, respectively.
Expression in response to nitrate
The members of the NRT families play different roles in nitrate flux from the soil throughout the plant body (Cerezo et al. 2001; Fan et al. 2009; Li et al. 2010; Wang & Tsay 2011). Many NRT genes show preferential expression in root tissue. Root developmental responses to the nitrate signal mediated by NRT proteins have been reported (Little et al. 2005; Remans et al. 2006; Li et al. 2010). As a first step to characterize members of the two L. japonicus transporter families, we analysed their response to nitrate provision in time-course experiments. Two-week-old Lotus plants, grown on B5/2 derivative solid media (Gamborg, Miller & Ojima 1968) supplemented with 1 mm glutamine as sole N source, were transferred at the time T0 to the same medium or in the presence of 0.01, 0.1, 1, 2 mm nitrate concentrations and root RNA was extracted at 10, 24 and 48 h. Gene expression was analysed by a semi-quantitative RT-PCR method using gene specific primers that recognized the 3'UTR regions (Supporting Information Table S2). Amplification products were confirmed by sequencing analysis. It was possible to detect expression of all 13 genes analysed in Lotus roots, although PCR conditions (number of cycles) to provide a linear range of gene amplification and a quantitative analysis of gene transcripts had to be modified in some cases, suggesting different levels of expression in roots (Supporting Information Fig. S1). A rough estimation based on these parameters indicated a higher level of expression in roots for the genes CM0247.130 and LjB20H09.30 among the NRT1/PTR sequences, whereas the NRT2 gene, CM0001.20, was apparently expressed at a lower level compared with the others (Figs 3 & 4).
The time-course experiment shown in Fig. 3 allowed the classification of the nine tested NRT1/PTR genes in three groups, namely nitrate inducible, nitrate repressible and nitrate constitutive, and as expected, most of the genes are included in the latter group as they did not respond significantly to the presence of nitrate in the growth medium (Okamoto et al. 2003; Wang et al. 2003). In particular, only the CM608.1210 gene was strongly and rapidly induced (3.5- to fivefold increase) in the whole range of KNO3 concentrations analysed (Fig. 3). The amount of transcript reached a peak by ten hours of nitrate provision and then gradually decreased, but it was still significantly higher after 48 h in the presence of nitrate. An opposite profile of expression was obtained with the CM0021.3040 gene that was repressed by 10 h of nitrate provision (1.75 to 2.9-fold decrease) and then resumed its normal level of expression (Fig. 3). In the other seven cases, a constitutive pattern of expression was observed during the whole time course (Fig. 3).
In the case of the NRT2 family, two out of the four genes analysed showed a strong nitrate-inducible profile of expression (Fig. 4). Consistent with the reports on the four O. sativa NRT2 genes, none of the L. japonicus members showed a nitrate-repressible profile of expression in roots (Cai et al. 2008; Feng et al. 2011). LjNRT2.1 was rapidly induced (2.7 to 3.8-fold increases) after 10 h of nitrate provision and then expression gradually decreased, but was still induced after 48 h of nitrate exposure (Fig. 4). CM0161.180 showed the strongest degree of induction as the amount of transcript increased up to 5.6-fold by 10 h of low nitrate provision (0.01 and 0.1 mm) and a 3.2 to 3.6-fold increase was maintained until 48 h. Interestingly, high nitrate provision (1 and 2 mm) did not induce CM0161.180 expression at the same level. At these higher nitrate levels, there was a transient peak of expression (3 to 3.7-fold increases) at 10 h of nitrate provision but at later time points expression returned to pre-exposure levels.
Expression in response to auxin and cytokinin
Cross-talks between the nitrate and auxin signalling pathways involved in developmental processes such as secondary root elongation and nodule organogenesis have been reported (Zhang & Forde 2000; Okamoto et al. 2009; Krouk et al. 2010; Reid, Ferguson & Gresshoff 2011). In the case of cytokinin, an increase of nitrate reductase (NR) activity and remobilization of nitrate was recently reported in wheat leaves after BAP addition (Criado et al. 2009). Furthermore, both auxin and cytokinin exogenous treatments might modulate the expression of some genes of the A. thaliana NRT1/PTR and NRT2 families (Guo et al. 2002; Tian, Uhlir & Reed 2002; Brenner et al. 2005; Kiba et al. 2011). In order to analyse the response of the L. japonicus NRT1(PTR) and NRT2 genes to exogenous auxin and cytokinin treatments, we first checked the profile of expression of L. japonicus orthologues of auxin and cytokinin responsive genes: CM0250.160, orthologue of the A. thaliana At2g14960.1, an IAA-amido synthetase gene of the GH3 auxin responsive family; and CM0778.210, orthologue of the A. thaliana At4g29740.2, a cytokinin oxidase gene. Plants were grown on 1 mm gln condition as previously described for the nitrate treatment and transferred after 2 weeks on the same medium plus 1 and 10 µm IAA or BAP for a time-course analysis (T0, T6h, T12h T24h and T48h). The CM0250.160 and CM0778.210 Lotus genes were quickly and specifically induced by auxin or cytokinin treatments (Fig. 5) at both concentrations (data not shown) and therefore only the 10 µm condition was used for the analysis of the NRT(PTR1) and NRT2 gene responses.
As for the nitrate treatment, auxin and cytokinin responses were classified as: inducible, repressible and constitutive. In four out of nine NRT1(PTR) genes, we observed a transcript decrease in response to the IAA treatment (Fig. 5). In the case of CM608.1210 and CM0021.3040, there was a gradual decrease (threefold) of the level of expression up to 48 h from the beginning of auxin provision, whereas this reduction was stable for CM0118.580 (about twofold) and transient in the case of LjB20H09.30. In the latter case, the quick decrease of the transcript level (about threefold) was followed by a strong sixfold increase between 24 and 48 h, which was also significant if compared with the T0 level (Fig. 5). Expression analysis of CM0608.1290 showed a unique profile, with a quick, strong (about threefold) and stable induction observed after the addition of 10 µm IAA (Fig. 5). Finally, CM247.130, CM0170.40, CM0826.350 and CM0826.370 did not show significant changes of the level of expression in response to auxin provision (Fig. 5).
In the case of cytokinin treatment, the pattern of gene responses appeared more variable. Three out of the nine genes analysed showed a transient profile of expression. CM0021.3040 showed a strong and quick decrease at T6h (4.2-fold) followed by a gradual increase up to the basal level of expression by 48 h of cytokinin provision. CM0826.370 had a similar profile with a gradual decrease of the transcript level up to 24 h, followed by a strong increase (fivefold) that was even significant when compared with the T0 level (Fig. 5). CM0608.1290 showed a profile of expression that was specular to that of CM0826.370, as the amount of transcript initially increased (almost threefold) up to 24 h of BAP treatment, to return to the basal level of expression 24 h later (Fig. 5). LjB20H09.30 and CM0247.130 showed a stable decrease of the amount of their transcripts, with a repressed state that was maintained for the whole time of treatment (Fig 5). Finally CM0608.1210, CM0170.140 and CM0118.580 showed a constitutive response to cytokinin (Fig. 5).
The NRT2 genes responded to the auxin treatment in different ways (Fig. 6). LjNRT2.1 showed a slight and quick increase of the amount of transcript up to 24 h with a return to the basic level of expression by 48 h of IAA provision (Fig. 6). LjNRT2.2 showed a quick slight increase of the transcript level, followed by a gradual decrease with a 2.5-fold reduction by 48 h. On the other hand, CM0001.20 showed an opposite profile of expression in response to auxin as a peak of expression was observed at 48 h from the auxin provision (2.3-fold increase; Fig. 6). Concerning CM0161.180 (CM0161.180 in release 1.0; http://www.kazusa.or.jp/lotus/release1), we observed a slight and gradual decrease of the transcript accumulation up to 24 h with a return to the basal level of expression 24 h later (Fig. 6).
Cytokinin provision repressed at different level the expression of the whole NRT2 genes family and this was particularly strong in the case of LjNRT2.2 mRNA that became almost undetectable after 24 and 48 h of BAP treatment (Fig. 6).
In order to analyse the profile of expression of the NRT1(PTR) and NRT2 gene families during the L. japonicus/M. loti symbiotic interaction, we performed a time-course experiment with extraction of total RNA from roots (1 and 3 d.p.i.) and nodules (12 and 28 d.p.i.) after M. loti inoculation. As internal marker to evaluate the fidelity of the experimental conditions, we followed the expression of the LjNIN gene that, as expected, was strongly induced in roots early after M. loti inoculation and reached the peak of expression in young nodules (Fig. 7). All the members of the NRT1(PTR) family, except CM0608.1290 and CM0826.370, did not show any modulation in the profile of expression (Supporting Information Fig. S2). CM0608.1290 and CM0826.370 showed a clear-cut and specific increase of the amount of transcripts in the nodular tissue and this was particularly significant in both young and mature nodules for the CM0826.370 gene (5.3 and 4-fold).
On the other side, three out of the four NRT2 genes appeared to be responsive to the symbiotic interaction process. LjNRT2.1 was severely down-regulated in young and mature nodules when compared with root tissue (2.2 and 3.6-fold), whereas a doubling of the LjNRT2.2 transcript amount was observed in root tissue at 24 h post-inoculation (Fig. 7). Finally, the CM0001.20 gene showed a 2.5-fold induction in roots at 72 h post inoculation, followed by a burst of expression in nodular tissue (Fig. 7). However for the CM0001.20 gene, the signal of the T0 sample was extremely weak and therefore the exact value of the nodule/root ratio, although significant, must be considered with caution.
Spatial pattern of the CM0826.370 promoter activity
In order to confirm RT-PCR analysis and to gain more insight in the possible role played by the CM0826.370 gene, we analysed its promoter-dependent spatial expression. A 2 kb PCR-amplified fragment covering the 5′ regulatory region was cloned in the pBI101.3 binary vector to obtain translational fusions with the gusA reporter gene. Lotus composite plants obtained upon transformation with Agrobacterium rhizogenes were used to analyse the expression of the translational fusion in a transgenic hairy root system. The pattern of GUS activity in hairy roots confirmed the induction of expression in young and mature nodular tissues obtained with the RT-PCR assay (Fig. 8a). Blue staining was not detected with this method in roots, even after 12 h of incubation at 37 °C, in agreement with the very low level of basal expression of the CM0826.370 gene in root tissue, indicated by the high number of PCR cycles needed to detect a consistent band of amplification in the semi-quantitative analysis (Fig. 3). Furthermore, this spatial analysis of expression permitted to obtain additional information on the expression of the CM0826.370 gene during early nodule developmental steps. As shown in Fig. 8b, GUS staining was detected in nodules primordia when cortical cells division activity just started to form nodule bumps. In mature nodules, blue staining was detected in the central zone where nitrogen fixation takes place and in the outer cortex (Fig. 8c), hence suggesting a role of the CM0826.370 gene throughout the nodule development and functioning steps. We also analysed the spatial profile of gusA expression after inoculation with an M. loti nifH- strain that can trigger a normal nodule developmental programme by invading cortical cells but does not fix nitrogen. After infection with the nifH- strain of Lotus plants transformed with the pr-CM0826.370-gusA construct, an increased number of white nodules unable to fix N could be detected at 5 weeks post-inoculation, but despite the lack of nitrogen fixation, the pattern and intensity of GUS activity in young nodules was identical to that elicited by the wild-type M. loti strain (data not shown). In other words, although we cannot exclude the existence of a post-transcriptional mechanism of CM0826.370 mRNA regulation in nodules, its expression in young nodules seems not to be dependent by N fixation activity.
A wide analysis of the NRT gene families expression profiles has been already described in A. thaliana and O. sativa that are representative of two plant habitats where nitrate availability can be completely different (Orsel, Krapp & Daniel-Vedele 2002; Okamoto et al. 2003; Feng et al. 2011), but it was never reported for legumes. The processes of nodule formation, development and functioning can be strongly affected by nitrate that might act both as a nutrient and a signal, making the regulation and function of NRT transporters and of their possible involvement in the mechanisms underlying these regulatory pathways worthy of investigation.
Thirty seven members of the NRT1(PTR) superfamily and four members of the NRT2 family have been identified by our in silico search in the L. japonicus genomic database. To gain more insight in the possible role played by the selected members (Fig. 1), we performed a molecular characterization that outlined a general regulatory network of these genes, suggesting possible links to different signalling pathways controlled by nitrate, auxin, cytokinin and bacterial symbiont (Table 1). We observed an extremely various picture of regulation of expression and consistently with previous results reported in A. thaliana, even pair of genes tandemly clustered may exhibit different patterns of expression (e.g.: LjNRT2.1 versus LjNRT2.2).
Table 1. Regulatory network of the analysed genes in response to nitrate, IAA, BAP and M. loti inoculation
To date, many data monitoring gene expression in response to NO3- provision in plants have been published, exploiting non-identical experimental conditions and cues to induce gene expression (nitrate concentrations, range of time course, etc.). Nevertheless, our analysis confirmed previous data reported for A. thaliana nitrate transporters where only the AtNRT1.1 gene, coding the dual affinity nitrate transporter CHL1, was shown to be induced by nitrate in roots (Okamoto et al. 2003; Wang et al. 2003). CHL1 is known to function in nitrate uptake from the soil and to play a key role in the sensing of nitrate concentrations (Tsay et al. 1993; Wang et al. 1998; Ho et al. 2009). The increase of the AtNRT1.1 mRNA has been observed with a wide range of NO3- concentrations (from 0.05 to 10 mm; Wang et al. 1998, 2000; Filleur & Daniel-Vedele 1999), consistently with the results we obtained for CM0608.1210 (Fig. 3). CM0608.1210 shares 35% identity with CHL1 and belongs to the same NRT1(PTR) subgroup I (Fig. 1). However, a strict correlation with the role played by CHL1 seems to be excluded by the different type of behaviour in response to auxin (Guo et al. 2002), as CM0608.1210 expression is down-regulated by exogenous addition of IAA (Fig. 5). Another example of nitrate transporter of the NRT1(PTR) family, showing a nitrate-dependent induction of expression in roots, is represented by the B. napus NRT1.2 gene that was characterized as a low-affinity nitrate/histidine transporter (Zhou et al. 1998). A further biochemical and genetic characterization will be essential to define the functional role of the L. japonicus CM0608.1210 gene.
The nitrate responses of the four L. japonicus NRT2 genes analysed in this work were not homogenous (Fig. 4). In particular, the two paralogs, LjNRT2.1 and LjNRT2.2, did not share the same nitrate-inducible profile of expression. Only LjNRT2.1 showed a quick increase of the transcript level in the whole range of nitrate concentrations utilized (Fig. 4), suggesting a different functional evolution of these two genes. In A. thaliana, NRT2.1 and 2.2 are both involved in high-affinity nitrate uptake, although a dominant role of AtNRT2.1 has been demonstrated (Filleur et al. 2001; Li et al. 2007). However, when AtNRT2.1 is mutated, AtNRT2.2 mRNA levels are increased threefold to compensate the functional loss of AtNRT2.1 (Li et al. 2007), and therefore we cannot exclude that a nitrate-inducible LjNRT2.2 profile of expression could be revealed only in an Ljnrt2.1 genetic background. CM0161.180 showed the major transcript increase in response to nitrate provision that was much stronger in the presence of low nitrate concentrations (Fig. 4). A similar behaviour with increases of the NRT2 transcript levels in roots of plants grown under limiting nitrate conditions when compared with plants grown under non-limiting nitrate conditions has been reported in A. thaliana (Orsel et al. 2002; Krouk, Tillard & Gojon 2006).
The analysis of the profiles of expression in the presence of exogenous IAA or BAP is particularly interesting as a cross-talk between proteins involved in sensing/transport and auxin/cytokinin signalling pathways has been clearly established in plant nutrition (Malamy 2005; Kiba et al. 2011). In general, the patterns of expression of the Lotus NRT1(PTR) genes reported in Fig. 5 indicated a non-immediate response to the addition of 10 µm IAA or BAP when compared with the auxin and cytokinin L. japonicus responsive genes utilized as internal controls. The exceptions are represented by CM0608.1290, CM0118.580, LjB20H09.30 and CM0021.3040 genes. The profile of expression of the CM0608.1290 is peculiar among the genes analysed in this work, as it is the only one showing a positive response to either auxin or cytokinin. A significant increase of the amount of transcripts, by comparing the level of expression with the untreated plants, was also observed for the CM0826.370 and LjB20H09.30 genes at 48 h of cytokinin and auxin treatments, respectively (Fig. 5).
To our knowledge, there are no reports about an auxin-dependent regulation of the NRT2 genes expression, except for a slight NAA-dependent increase observed for the OsNRT2.4 mRNA (Feng et al. 2011). Our analysis indicated a differentiated response to auxin provision of LjNRT2.1 and LjNRT2.2 genes, and a strong level of induction after IAA provision for the CM0001.20 gene (Fig. 6). The general picture of a repressible profile of expression observed for all the four NRT2 genes in the presence of BAP is consistent with some reports in literature showing a quick and strong down-regulation of AtNRT2.1, AtNRT2.3 and AtNRT2.6 genes in Arabidopsis seedlings exposed for 120 min to 5 µm BAP (Brenner et al. 2005). Although the presented data are too preliminary to make any assumption about the biological meaning of the observed modulation of the transcript levels, it is interesting to note that for most of the time points tested, the LjNRT2 genes present an opposite type of response to IAA and BAP, with higher level of transcripts in the presence of auxin (Fig. 6). High auxin/cytokinin ratios have been implicated in the control of the organogenesis programmes modelling root architecture, where an auxin maximum, which can be counteracted by cytokinin, has been demonstrated to be crucial for the establishment of the fate of the lateral root founder cell (Fukaki & Tasaka 2009). Therefore, considering that most of the plant NRT2 genes are preferentially expressed in roots, our reported patterns of response to phytohormones could suggest a link to the root hormones signalling pathways.
The phytohormone-dependent transcriptional regulation observed for some of the NRT(PTR) and NRT2 genes might have an intriguing intersection with the patterns of expression observed in roots and nodules upon M. loti inoculation (Fig. 7). Auxin and cytokinin play a major role in nodule formation and development by exerting opposite effects. Auxin transport in indeterminate nodulation is quickly and transiently blocked just above the site of nodule initiation (Mathesius et al. 1998), while cytokinin appears to promote the early pathway inducing cell divisions in the cortex that leads to nodule primordial formation ahead of the upcoming infection thread (Murray et al. 2007; Tirichine et al. 2007; Ding & Oldroyd 2009). The semi-quantitative analysis reported in Fig. 7 indicated an early induction of the LjNRT2.2 and CM0001.20 genes in L. japonicus roots inoculated with M. loti. Furthermore, analysis of the spatial localization of the promoter activity of the CM826.370 gene indicated an early expression in nodule primordium (Fig. 8b), whereas no detectable expression was observed in the root epidermal cells that respond earlier to the rhizobia inoculation by perceiving the Nod factor (Oldroyd & Downie 2008). It is certainly difficult to assume a role for low- or high-affinity nitrate transporters in nodulation, but an intriguing possibility, consistent with the observed early patterns of expression during nodule organogenesis, would predict an involvement in the high and/or low nitrate signalling pathways affecting nodule formation in a positive or negative way, respectively (Barbulova et al. 2007; Fey & Vessey 2009; Omrane & Chiurazzi 2009; Harris & Dickstein 2010). This would not be in contradiction with the nitrate-independent profiles of expression of CM0826.370, LjNRT2.2 and CM0001.20 observed in Lotus roots (Figs 3 & 4). In the case of CM0826.370, such a role in the signalling pathways governing the nodule formation process could be also played through an action as transporters of the CLE peptides, which have been recently reported to be involved in the auto-regulation mechanism controlling the nodule number (Okamoto et al. 2009; Harris & Dickstein 2010; Mortier et al. 2010). In particular, the CM0826.370 activation by cytokinin in the Lotus roots (Fig. 5) and its spatial expression in nodule primordia (Fig. 8) could be consistent with its action in the cytokinin-dependent nodule cortical pathway. Interestingly, a cytokinin-dependent induction of expression for the M. truncatula CLE13 in roots with a topology of promoter activity largely overlapping the one observed for CM0826.370 have been recently reported (Mortier et al. 2010, 2012). Recently, a putative nitrate transporter encoded by the M. truncatula LATD/NIP gene has been reported to be involved in nodule and roots development (Veereshlingam et al. 2004; Harris & Dickstein 2010; Yendrek et al. 2010). Expression of the LATD/NIP gene is strongly induced by cytokinin and inhibited by auxin in M. truncatula root tips. However, LATD/NIP is expressed throughout the plant and it is spatially confined to meristematic regions and surrounding cells of primary and secondary roots as well as nodule organs (Yendrek et al. 2010). The expression of CM0826.370 in the central region of mature nodules, where N fixation takes place, suggests its involvement also during the functioning steps (Fig. 8). It is known that an efficient symbiotic interaction is supported by a complicated network of nutrient exchanges between the two partners, mainly a flux of reduced carbon and other nutrients from the plant for reduced nitrogen from the bacteroids (Ludwig et al. 2003). A direct involvement of a nitrate transporter in these nutrient exchanges was never reported. The only indirect evidence comes from the identification in L. japonicus, by a proteomic approach, of a putative nitrate transporter localized on the peribacteroid membrane (PBM), which is the structural and functional interface between the legume plant and the rhizobia (Wienkoop & Saalbach 2003). This protein has been identified by our in silico search as the CM0295.980 member, found in the subgroup III of the L. japonicus NRT1(PTR) superfamily (Fig. 1) and was not included in the molecular characterization reported here. Interestingly, the spatial profile of expression of the prom-826.370-gusA in mature nodules overlaps that of the ammonium transporter L.jAMT1.1 (D'Apuzzo et al. 2004; Rogato et al. 2008). An involvement in the network of nutrient exchanges between the symbiotic partners occurring in mature nodules has been proposed for Alnus glutinosa AgDCAT1 protein (Jeong et al. 2004). AgDCAT1 represents an exception among members of the NRT1(PTR) family, which transport nitrate or di-tri-peptides because it might transport dicarboxylate (Jeong et al. 2004). AgDCAT1 is expressed specifically in nodules at the interface between the plant cell and the bacteria, and it was suggested to be involved in the supply of intracellular bacteria with dycarboxylates as carbon sources (Jeong et al. 2004).
In conclusion, we presented here a molecular characterization of the L. japonicus NRT1(PTR) and NRT2 gene families, showing an extremely diversified picture of expression patterns in responses to different signals, suggesting a multifaced range of specialized functions and connections to different signalling pathways for these proteins. We also identified, for some members, a profile of expression that was significantly regulated during the L. japonicus/M. loti symbiotic interaction indicating an involvement in different stages of nodule organogenesis and functioning. These data provide a sound foundation for future experiments that will help to elucidate the specific roles of each transporter in legume plants.
This work was supported by two grants from the Italian Ministry of Education: (1) Progetti di Rilevanza Nazionale, PRIN 2008, Prot. 2008WKPAWW and (2) Progetto CISIA, ‘Integration of Knowledge for Sustainability and Innovation in the Agrofood Made in Italy’; and and one grant from National Council of Research, Agrofood Department (Award for Funding Research of Excellence). We thank the facility of Integrated Microscopy of the IGB for the optical microscopy analysis. The authors also wish to thank Sara Salvia, Aurora Quattromani and Dr Kevin Roberg for technical assistance and help with the manuscript, respectively.