Nitrate transport capacity of the Arabidopsis thaliana NRT2 family members and their interactions with AtNAR2.1


Author for correspondence:
Anthony D. M. Glass
Tel: +1 604 822 4847


  • Interactions between the Arabidopsis NitRate Transporter (AtNRT2.1) and Nitrate Assimilation Related protein (AtNAR2.1, also known as AtNRT3.1) have been well documented, and confirmed by the demonstration that AtNRT2.1 and AtNAR2.1 form a 150-kDa plasma membrane complex, thought to constitute the high-affinity nitrate transporter of Arabidopsis thaliana roots. Here, we have investigated interactions between the remaining AtNRT2 family members (AtNRT2.2 to AtNRT2.7) and AtNAR2.1, and their capacity for nitrate transport.
  • Three different systems were used to examine possible interactions with AtNAR2.1: membrane yeast split-ubiquitin, bimolecular fluorescence complementation in A. thaliana protoplasts and nitrate uptake in Xenopus oocytes.
  • All NRT2s, except for AtNRT2.7, restored growth and β-galactosidase activity in the yeast split-ubiquitin system, and split-YFP fluorescence in A. thaliana protoplasts only when co-expressed with AtNAR2.1. Thus, except for AtNRT2.7, all other NRT2 transporters interact strongly with AtNAR2.1.
  • Co-injection into Xenopus oocytes of cRNA of all NRT2 genes together with cRNA of AtNAR2.1 resulted in statistically significant increases of inline image uptake over and above that resulting from single cRNA injections.


Nitrate uptake by plant roots from soil solution is mediated by members of three gene families, namely the NitRate Transporter families (NRT1 and NRT2) and the Nitrate Assimilation Related family (NAR2) (see reviews by Miller et al., 2009; Dechorgnat et al., 2011; Tsay et al., 2011). Full activity of the inducible high-affinity nitrate transport system (IHATS) in roots of Arabidopsis thaliana requires expression of two transporters, namely, NRT2.1 and NRT2.2, both of which are members of the major facilitator superfamily (MFS). Evidence for this assertion is based upon the high degree of correlation between physiological activity of the IHATS and transcript abundance of AtNRT2.1 and AtNRT2.2 in wild-type (WT) lines and the reduction of IHATS in mutants disrupted in AtNRT2.1 and or AtNRT2.2 (Zhuo et al., 1999; Filleur et al., 2001; Okamoto et al., 2003; Li et al., 2007). Nevertheless, it was first shown in Chlamydomonas reinhardtii that the NRT2.1 and NRT2.2 genes require simultaneous expression of another gene called NAR2 in order to express high-affinity nitrate transport (Quesada et al., 1994). In A. thaliana also, disruption of NAR2.1 (NRT3.1), in a T-DNA insertional mutant Atnar2.1-1, caused an almost complete loss of inducible high-affinity nitrate influx (Okamoto et al., 2006; Orsel et al., 2006). This observation is now readily explained by the demonstration that AtNRT2.1 and AtNAR2.1 form part of a 150-kDa plasma membrane (PM) complex (Yong et al., 2010) that was suggested to be a tetramer consisting of two subunits each of AtNRT2.1 and AtNAR2.1. Further, it was proposed that the 150-kDa complex is the functional high-affinity nitrate transporter in A. thaliana. Interestingly, this complex, as is AtNRT2.1, is absent from PM preparations derived from NAR2.1 mutants (Wirth et al., 2007; Yong et al., 2010), suggesting that AtNRT2.1n only exist in the root PM in association with AtNAR2.1. Membrane interactions between AtNRT2.1 and AtNAR2.1 were clearly indicated earlier by heterologous expression in the yeast two-hybrid and Xenopus oocytes systems. Only when oocytes were injected with both AtNRT2.1 and AtNAR2.1 RNA was nitrate uptake into oocytes detected (Orsel et al., 2006). In addition to A. thaliana, the two-component nitrate uptake system of NRT2 and NAR2 has been demonstrated to be present in other plant species such as barley (Hordeum vulgare) and rice (Oryza sativa) (Glass, 2009; Ishikawa et al., 2009; Feng et al., 2011). Yeast two-hybridization showed that OsNAR2.1 interacted not only with OsNRT2.1/OsNRT2.2, but also with OsNRT2.3a (Feng et al., 2011). In a study of NRT2/NAR2 interactions in barley, it was suggested that the C-terminus of HvNRT2.1 is possibly involved in binding to the HvNAR2.3 central region and that the Ser463 present in the HvNRT2.1 C-terminus plays a role in the binding ability (Ishikawa et al., 2009).

In addition to AtNRT2.1 and AtNRT2.2, there are five other members of the NRT2 gene family in A. thaliana, NRT2.3 to NRT2.7 (Orsel et al., 2002; Okamoto et al., 2003). Quantitatively, AtNRT2.1 and AtNRT2.2 are responsible for c. 80% of IHATS in A. thaliana (Li et al., 2007). The remaining 20% is probably attributable to expression of a constitutive high-affinity transport system (CHATS) and CHL1 (NRT1.1). Although a role for AtNRT2.7 in seed nitrate accumulation has been proposed (Chopin et al., 2007) and expression patterns of the remaining NRT2 genes have been documented (Orsel et al., 2002; Okamoto et al., 2003), their functions and possible interactions with NAR2 are unknown.

In the present study, therefore, we have used heterologous expression in the yeast two-hybrid system and transient expression in A. thaliana leaf protoplasts to look for possible interactions between AtNAR2.1 and the other NRT2 genes (AtNRT2.2 to AtNRT2.7) and to localize the demonstrated interactions in planta. To explore the physiological functions of the NRT2 polypeptides we have used transient expression of AtNRT2.1 to AtNRT2.7 (plus or minus AtNAR2.1) and inline image uptake into Xenopus oocytes.

Materials and Methods

Membrane yeast-two-hybrid screening for interactions of members of the AtNRT2 gene family with AtNAR2 as bait

Membrane yeast-two-hybrid (Y-2-H) screening for interactions of the AtNRT2 gene family with AtNAR2 was performed using the DUALmembrane kit from Dualsystems Biotech AG (Schlieren, Switzerland). cDNA of AtNAR2.1 was cloned into Y-2-H bait vector pTMBV4 using XbaI and StuI restriction sites in-frame with C-ubiquitin. Correct expression of the AtNAR2.1 bait construct was confirmed by co-expression with control plasmid Alg5-NubI (positive control- WT N-terminal ubiquitin) and Alg5-NubG (negative control- mutated N-terminal ubiquitin). By transforming the bait strain with an empty prey vector pDL2Nx, it was determined that addition of 5mM 3-amino-1,2,4-triazole (3-AT) to minimal media was sufficient to decrease the sensitivity of the histidine biosynthesis HIS3 reporter gene. cDNAs of all AtNRT2 genes except for AtNRT2.2 and AtNRT2.7 were cloned into the pDL2Nx prey vector using BamHI and EcoRI restriction sites, in-frame with N-ubiqutin. AtNRT2.2 and AtNRT2.7 were cloned into the pDL2xN prey vector using BamHI/EcoRI and BamHI/ClaI restriction sites, respectively. Sequences of oligonucleotide primers used to clone all seven members of the NRT2 family and AtNAR2.1 are shown in Supporting Information Table S1. Polyethylene glycol (PEG)-mediated transformation of yeast strain DSY-1 with bait and prey constructs was performed according to the Dualsystems Biotech manual. Transformation efficiency was checked on Synthetic Defined (SD) plates (0.2% w/v Difco™ yeast nitrogen base without amino acids (BD Biosciences, San Diego, CA, USA) and ammonium sulfate, 0.5% w/v ammonium sulfate, 0.1% w/v dropout mix, 2% w/v dextrose and 2% w/v agar) without leucine and tryptophan amino acids in the dropout mix. Screening for interaction of NRT2 proteins with AtNAR2.1 was achieved on SD plates without leucine, tryptophan and histidine, and by assay of β-galactosidase activity as per the manufacturer’s manual. A bait dependence test was performed to exclude false positives by co-transforming NRT2 clones with control bait pMBV-Alg5 provided with the DUALmembrane kit.

Arabidopsis thaliana (L.) Heynh. leaf protoplast isolation, transfection and confocal fluorescence imaging

Interaction of NRT2 proteins with AtNAR2.1 was investigated in vivo using the split-YFP method and transient expression in A. thaliana protoplasts (Citovsky et al., 2006). cDNA of AtNAR2.1 was fused in-frame to the N-terminal half of YFP in pSAT1A-nEYFP-N1 (XhoI/BamHI restriction sites). cDNAs of all NRT2 genes were fused in-frame with the C-terminal half of YFP using the pSAT4A-cEYFP-N1 vector. PCR-amplified cDNAs of AtNRT2.1, 2.2, 2.3, 2.5 and 2.7 were inserted into pSAT4A-cEYFP-N1 using XhoI and BamHI restriction sites, while for AtNRT2.4 and 2.6 cloning EcoRI and KpnI restriction sites were used. All primer sequences are provided in Table S2. Negative controls were plasma membrane ATP-binding cassette (ABC) transporters ABCG11 and ABCG12 fused to C-YFP and N-YFP halves, respectively, provided by McFarlane et al. (2010). In addition, co-transfection with complementary empty vector was used as a control. Arabidopsis thaliana leaf protoplasts were isolated and transfected with purified plasmid DNA according to the protocol by Tiwari et al. (2006). In brief, 1 g of leaves was cut into <1-mm strips and incubated in enzyme solution (1% (w/v) Cellulase Onozuka R10, 0.25% (w/v) Macerozyme R10 (Plant Media, bioWorld, Dublin, Ohio, USA), 400 mM mannitol, 10 mM CaCl2 and 5 mM MES, pH 5.7) for 90 min in darkness on a rotary shaker at 50 rpm. After protoplasts were released into solution, 1/3 volume of 200 mM CaCl2 was added. Next, protoplasts were collected by centrifugation at 180 g for 3 min and resuspended at 3 × 105 protoplasts ml−1 in Mg-manitol solution (400 mM mannitol, 15 mM MgCl2 and 4 mM MES, pH 5.7). 200 μl of protoplasts was transfected with 10 μg of plasmid DNA and an equal volume of PEG solution (40% (w/v) 3350 PEG, 100 mM CaCl2 and 400 mM mannitol, pH 10) at room temperature for 20 min. Protoplasts were recovered by centrifugation and resuspended in 1 ml of WI solution (500 mM mannitol, 20 mM KCl and 4 mM MES, pH 5.7). After 18 h of incubation in darkness at room temperature, protoplasts were visualized using a Spinning Disk PerkinElmer UltraView VoX Microscope (equipped with Leica DMI6000 inverted microscope and Hamamatsu 9100-02 CCD camera) and Volocity software (PerkinElmer, Waltham, Massachusetts, USA).

AtNAR2.1 and AtNRT2 gene family cloning and cRNA synthesis for Xenopus oocyte injections

cDNAs of all AtNRT2 genes (NRT2.1, NRT2.2, NRT2.3, NRT2.4, NRT2.5, NRT2.6 and NRT2.7) and AtNAR2.1 were amplified using High Fidelity Phusion polymerase (Finnzymes, Vantaa, Finland). Primer sequences included Gateway® Technology (Invitrogen, Grand Island, New York, USA) recombination sites (Table S3). Amplified cDNAs were combined with donor vector pDONR221 (Invitrogen, Grand Island, New York, USA) using BP Clonase II (Invitrogen) in BP recombination reaction to obtain entry clones. The entry clones were sequenced using M13 forward and reverse primers. Destination clones were prepared by LR recombination reaction utilizing pDONR221 clones, pGEMHE destination vector and LR Clonase II enzyme mix (Invitrogen). The pGEMHE vector contains 5’ and 3’ untranslated sequences of the β-globin gene from Xenopus laevis to increase the translation efficiency of heterologous RNA (Liman et al., 1992). pGEMHE clones were digested with NheI enzyme, and 1 μg of digested purified DNA was used to synthesize cRNA with the AmpliCap™ T7 High Yield Message Maker Kit (Epicentre, Madison, WI, USA). The integrity of cRNA was determined on a formamide denaturing Tris-Acetate-EDTA (TAE) agarose gel (Masek et al., 2005). Concentrations of cRNA were measured using the RiboGreen® RNA kit (Molecular Probes; Invitrogen), and adjusted to 500 ng μl−1.

Xenopus oocyte harvesting and injections

Xenopus oocytes were harvested according to the protocol by Hill et al. (2005). Ovaries were surgically removed from a X. laevis frog, and oocytes defoliculated in a 50-ml Falcon tube in Ca-free Ringers solution with 1.7% (w/v) collagenase and 0.05% (w/v) trypsin inhibitor, for 90 min with rotation. After digestion, oocytes were washed three times in hypotonic buffer (100 mM K2HPO4 and 0.1% (w/v) BSA, pH 6.5), and then incubated for 10 min in fresh hypotonic buffer. Oocytes were then washed two times in Ca-free Ringers solution, and once in Ca-Ringers solution. Thereafter, oocytes were stored in Ca-Ringers solution supplied with 50 μg ml−1 tetracycline, 100 units ml−1 penicillin, 100 μg ml−1 streptomycin, and 5% (w/v) heat-inactivated horse serum (Sigma Aldrich, USA). Healthy-looking oocytes (stage V–VI) were selected for micro-injection of cRNA. Glass microcapillaries (Drummond Scientific, Broomall, PA, USA) were pulled using a Narishige puller (Narishige, Tokyo, Japan) on heat settings of 11.83 and 9. Capillary tips for injection were ground at a 45°-angle on Microgrinder EG-400 (Narishige). Selected oocytes were injected with 50 μl of RNAse-free water or cRNA using a Nanoject II Auto-nanoliter injector (Drummond Scientific). 25 ng was used for single-gene cRNA injection, and 50 ng of cRNA was used for injection of NRT2 cRNA together with AtNAR2.1 (cRNA mixed 1 : 1 ratio).

Uptake of K15NO3 in Xenopus oocytes

Oocytes were incubated for 1 d after injection in Ca-Ringers solution (96 mM NaCl, 2 mM KCl, 0.6 mM CaCl2, 1 mM MgCl2 and 5 mM MES, pH 7.5) supplied with 50 μg ml−1 tetracycline, 100 units ml−1 penicillin, 100 μg ml−1 streptomycin, and heat-inactivated horse serum (Sigma Aldrich) to promote protein expression before the uptake experiments. During the second day of incubation, Ca-Ringers was supplied with antibiotics only. Twenty oocytes were selected per treatment and washed once in N-free MBS uptake solution (96 mM NaCl, 2 mM KCl, 2 mM CaCl2, 1 mM MgCl2 and 5 mM MES; pH 6.5) for 5 min before the experiment. After the washing solution had been removed, 5 ml of MBS uptake solution with 500 μM K15NO3 (> 98 atom %15N; Sigma Aldrich) was added and healthy oocytes were incubated for 12 h at 18°C. Thereafter, oocytes were briefly washed three times in 5 ml of ice-cold N-free MBS, and single oocytes were transferred into tin capsules (8 × 5 mm; SerCon, Cheshire, UK), dried at 50°C for 2 d and pressed in the tin capsules, and the isotope ratio was measured using SerCon Isotope Ratio Mass Spectrometer IRMS 20-22. Values obtained represent 15N enrichment compared with the standard atmospheric 15N : 14N ratio (delta 15N air).


Membrane yeast-two-hybrid interactions

Yeast-two-hybrid heterologous expression was used to assess possible interaction between AtNAR2.1 and all members of the A. thaliana NRT2 family. AtNAR2.1 was fused to the C-terminal half of ubiquitin using a bait vector, and each of the NRT2 genes was fused to the N-terminal half of ubiquitin in a prey vector. Yeast was transformed with AtNAR2.1 as bait and each of the NRT2 genes as prey individually. In addition, negative control bait vector (pMBV-Alg5) transformation was used for each of the NRT2 constructs. High transformation efficiency was confirmed on minimal nutrient media without leucine (for bait vector) and tryptophan (for prey vector), as shown in Fig. 1(a,b). Screening for putative interaction was achieved using minimal nutrient media without leucine, tryptophan and histidine. Fig. 1(a) shows that NRT2 constructs did not interact with the negative control bait plasmid, while AtNRT2.1, 2.2, 2.3, 2.4, 2.5 and 2.6 showed significant growth when co-transformed with AtNAR2.1 as bait on minus His media (Fig. 1b). These results in the yeast system were confirmed by positive β-galactosidase assays (Fig. 1c). By contrast, AtNRT2.7, co-transformed with AtNAR2.1 as bait, failed to grow on media without histidine or give a positive β-galactosidase assay (Fig. 1b,c).

Figure 1.

Heterologous expression and screening for interactions with AtNAR2.1 in the yeast two-hybrid system. (a) Expression of NRT2 genes with negative control bait pALG5; growth of yeast on selective media without leu and trp (left panel), and without leu, trp and his (right panel). (b) Expression of NRT2 genes with AtNAR2.1 bait; growth of yeast on selective media without leu and trp (left panel), and without leu, trp and his (right panel). (c) X-gal assay on yeast colonies expressing AtNAR2.1 and NRT2 genes. NAR, NitRate Assimilation Related family; NRT, NitRate Transporter families; NUBI, N-terminal ubiquitin; NUBG, mutated N-terminal ubiquitin.

Transient in planta interactions between AtNAR2.1 and AtNRT2 genes in Arabidopsis protoplasts

The interactions of genes from the AtNRT2 family with AtNAR2.1 and possible localization of the interaction were investigated by transient in vivo expression of split YFP-labeled AtNAR2.1 and NRT2 genes in A. thaliana leaf protoplasts. Fig. S1 shows bright field and fluorescence images of protoplasts transformed with one of AtNRT2.2-cEYFP, AtNRT2.3-cEYFP, AtNRT2.4-cEYFP, AtNRT2.5-cEYFP, AtNRT2.6-cEYFP and AtNRT2.7-cEYFP together with the ABCG12-nEYFP vector as a negative control. Only a very small amount of fluorescence was detected in the control protoplasts, mainly attributable to intrinsic fluorescence of the chloroplast (Fig. S1). However, protoplasts transformed with AtNRT2.2-cEYFP to AtNRT2.6-cEYFP together with AtNAR2.1-nEYFP exhibited strong fluorescence localized mainly in the PM (Fig. 2). Protoplasts transfected with AtNRT2.7-cEYFP and AtNAR2.1-nEYFP failed to show strong fluorescence (Fig. 2), indicating poor interaction of AtNRT2.7 with AtNAR2.1. It should be noted that in a previous paper (Yong et al., 2010) we demonstrated strong interactions between AtNRT2.1 and AtNAR2.1 in this protoplast system.

Figure 2.

Bright field (left) and confocal fluorescence images (right) of Arabidopsis thaliana protoplasts transfected with NRT2.2-2.7 genes fused to cEYFP and NAR2.1-nEYFP. NAR, NitRate Assimilation Related family; NRT, NitRate Transporter families.

Uptake of K15NO3 into X. laevis oocytes

Controls for the putative interactions were provided by injecting healthy oocytes with either water or with 25 ng of cRNA encoding each of AtNAR2.1, AtNRT2.1, AtNRT2.2, AtNRT2.3, AtNRT2.4, AtNRT2.5, AtNRT2.6 and AtNRT2.7. To test for functional interactions between AtNAR2.1 and AtNRT2 proteins, oocytes were also injected with mixtures of AtNAR2.1 cRNA and cRNA of each of the NRT2 genes. After 2 d of cRNA expression, injected oocytes were incubated in 0.5 mM K15NO3 for 12 h. Uptake of K15NO3 was evaluated through measurement of 15N enrichment of oocytes by isotope ratio mass spectrometry, and expressed compared with standard atmospheric 15N : 14N ratios (delta 15N air). Compared with water controls, single injections of AtNRT2.2, AtNRT2.3, AtNRT2.4, AtNRT2.5, and AtNRT2.6 all produced statistically significant (< 0.05) increases in 15N accumulation (Table S4). However, these increases were relatively small (average value of 56%) compared with those elicited by co-injection of AtNAR2.1 together with NRT2s. All of these gave statistically significant increases (< 0.05), with an average increase of 180%, when compared with the corresponding single NRT2 injections (Fig. 3). Likewise, co-injection of AtNAR2.1 together with NRT2s significantly increased 15N accumulation (< 0.05), with an average increase of 195%, when compared with 15N accumulation associated with a single NAR2.1 injection (Table S4). Co-injection of AtNRT2.1 with AtNAR2.1 gave the greatest increase of 15N accumulation (532%), followed by AtNRT2.5 (334%), while AtNRT2.3 (40%) and AtNRT2.4 (48%) gave the lowest increments (Table S4). Values presented are from a single experiment, with at least eight oocytes per cRNA injection. Experiments were repeated three times and showed similar values.

Figure 3.

K15NO3 uptake into Xenopus oocytes. Gray bars, oocytes injected with water, cRNA of NAR2.1 alone, and single NRT2 genes; black bars, oocytes injected with a mixture of NAR2.1 and individual NRT2 gene cRNA. 15N enrichment per oocyte is expressed as delta 15N compared with standard atmospheric 15N : 14N ratio. Values are average of = 8 ± SD. Asterisks signify that co-injection of NRT and NAR2 produced significantly higher 15N uptake (P < 0.05) than single injections of NAR2 alone. NAR, NitRate Assimilation Related family; NRT, NitRate Transporter families.


The plant NRT2 family of nitrate-nitrite porters (NNPs) belongs to the major facilitator superfamily of transporters. They all have 12 predicted transmembrane regions (TMRs) and a central cytoplasmic loop between TMR6 and TMR7, which is relatively small in A. thaliana, with 21 amino acids, and substantially longer in Aspergillus nidulans, with 91 amino acids (Forde, 2000). Both polypeptide ends of the NRT2s are located at the cytosolic side of the PM (reviewed in Saier et al., 1999; Forde, 2000; Law et al., 2008). A requirement for two distinct polypeptides belonging to the NRT2 and NAR2 families was first demonstrated in Chlamydomonas reinhardtii (Quesada et al., 1994). Confirmation of this finding was provided by Zhou et al. (2000a), who demonstrated that only when CrNAR2 was co-expressed with CrNRT2.1 was high-affinity nitrate uptake obtained in Xenopus oocytes. Likewise, a requirement for two distinct polypeptides (NRT2.1 and NAR2 homologs) was subsequently demonstrated in A. thaliana (Okamoto et al., 2006; Orsel et al., 2006), rice (Araki & Hasegawa, 2006; Cai et al., 2008; Yan et al., 2011), barley (Tong et al., 2005; Ishikawa et al., 2009), wheat (Triticum aestivum) (Cai et al., 2007) and the moss Physcomitrella patens (Tsujimoto et al., 2007). Evidence presented in a recent study of the 150-kDa AtNAR2.1/AtNRT2.1 complex localized in PM preparations from roots of A. thaliana suggests that this complex is the functional unit responsible for high-affinity nitrate influx (Yong et al., 2010). It was also suggested that this functional unit is probably a tetramer consisting of two subunits each of the two polypeptides. It appears, therefore, that this two-component high-affinity nitrate transport system for nitrate uptake from soils is universal among plants. Nevertheless, this is not the case in A. nidulans, in which NRTA, the NRT2.1 homolog, was able to generate nitrate currents in the Xenopus system in the absence of NAR2 (Zhou et al., 2000b). The reason for this difference is unclear as the predicted polypeptide sequence for the A. nidulans high-affinity transporter, NRTA, is sufficiently similar to the A. thaliana and barley high-affinity transporters that degenerate primers based upon the A. nidulans and C. reinhardtii sequences were used to identify the plant homologs (Trueman et al., 1996; Zhuo et al., 1999). One distinct difference between the plant and fungal transporters is the presence of a much longer central loop in A. nidulans. Another difference is that there appears to be no NAR2 homolog in A. nidulans (Yong et al., 2010). Is it possible that the functions carried out by NAR2 in plants are mediated by the large central loop of the fungus?

The results of earlier studies using the yeast two-hybrid system had established that AtNAR2.1 interacts with AtNRT2.1 (Orsel et al., 2006). The results of the present yeast two-hybrid study establish that AtNAR2 also interacts with AtNRT2.2, AtNRT2.3, AtNRT2.4, AtNRT2.5, and AtNRT2.6 based upon growth on minus histidine media and the β-galactosidase test. The apparent failure of AtNRT2.7 to interact with AtNAR2.1 represents an exception to the generality that all of the A. thaliana NRTs interact with AtNAR2.1. The in planta assays of the association between AtNAR2.1 and AtNRT2.2-2.7, by means of the split YFP-labeled AtNAR2.1 and NRT2 genes in A. thaliana leaf protoplasts, confirmed the results of the yeast two-hybrid assays. Thus, in this in vivo expression system using A. thaliana protoplasts, fluorescence of the split YFP was recovered after co-transfection of all NRT2s with AtNAR2, except that AtNRT2.7 gave only a very weak fluorescence signal (Fig. 2). Based upon these observations, AtNRT2.7 may be unique among AtNRT2 transporters. Further support for this hypothesis is provided by the following reports.

  • AtNRT2.7 was shown to be a tonoplast transporter (Chopin et al., 2007).
  • Of all AtNRT2s, AtNRT2.7 shows the lowest amino acid sequence similarity with AtNRT2.1 (Orsel et al., 2002).
  • A recent phylogenetic study of NRT genes demonstrated that AtNRT2.7 is the most divergent of all NRT2 genes and that there are no NRT2.7-like genes in genomes of sequenced grasses or in poplar (Populus trichocarpa) (Plett et al., 2010).

The results of the present Xenopus assays demonstrate that co-injection of AtNAR2.1 significantly increased 15N accumulation by all NRT2s (< 0.05) when compared with values obtained by single NRT2 injection (Table S4). The largest increases were obtained by co-injection of AtNAR2.1 with AtNRT2.1 (532% increase) and with AtNRT2.5 (334% increase), while AtNRT2.3 gave the smallest increase (40% increase). The average increase for all NRT2s was 180%. Likewise, compared with a single NAR2.1 injection, all NRT2s showed statistically significant increases (< 0.05) of 15N accumulation (average increase 195%) when co-injected with AtNAR2.1. Interestingly, single injections of AtNAR2.1 and all NRT2s (except for AtNRT2.1) increased 15N accumulation, compared with the water control, although the increases were small (average value 55%) compared with the average value obtained from co-injection (255%). Our findings with regard to single injections of NRTs confirm an earlier report by Chopin et al. (2007) that single injections of AtNR2.7 into Xenopus oocytes (in the absence of AtNAR2.1) caused a statistically significant increase of 15N accumulation compared with a water control. However, the authors did not assay the effect of co-injecting AtNAR2.1. Nevertheless, the present findings do not concur with those of other studies. For example, compared with the water control, the barley HvNRT2.3 failed to increase 15N accumulation when injected without HvNAR2.3, while co-injection did increase 15N accumulation (Tong et al., 2005). Similarly, single injections of CrNAR2 (Zhou et al., 2000a) or AtNAR2.1 and AtNRT2.3 (Orsel et al., 2006) failed to increase nitrate accumulation compared with water controls. The reason for these differences may be species-based or technical; for example, in the Zhou et al. study oocytes were assayed after 3–6 d, while oocytes were incubated for only 16 h before assaying 15N uptake in the Orsel et al. study, compared with 48 h in the current study.

In summary, the results of the yeast two-hybrid and A.thaliana protoplast experiments establish that all NRT2s, except AtNRT2.7, interact strongly with AtNAR2.1. The findings for the Xenopus oocyte system reveal that all NRT2 polypeptides (except AtNRT2.1) can sustain a small nitrate flux. However, when co-expressed with AtNAR2.1, this is very substantially increased compared with single NRT2 injections. The enhancement of inline image uptake by oocytes expressing AtNRT2.7 together with AtNAR2.1 may appear to be contradictory to the apparent absence of interaction indicated by the yeast two-hybrid and A. thaliana protoplast assay. However, nitrate uptake by AtNRT2.7, in a single injection, was not significantly higher (< 0.05) than the water control, while when co-injected with AtNAR2.1 it gave the lowest absolute value for 15N accumulation.


We gratefully acknowledge financial support in the form of a UBC graduate fellowship and travel award to Z.K., financial support from NSERC to A.D.M.G. and an Australian Research Council Linkage Grant (LP0776635) to B.N.K. We would also like to thank the UBC BioImaging facility for help with confocal microscopy and Dr Nelly Panté’s lab at UBC for assistance with the Xenopus oocyte work.