Characterization of nitrite uptake in Arabidopsis thaliana: evidence for a nitrite-specific transporter



  • Nitrite-specific plasma membrane transporters have been described in bacteria, algae and fungi, but there is no evidence of a nitrite-specific plasma membrane transporter in higher plants. We have used inline image to characterize nitrite influx into roots of Arabidopsis thaliana.
  • Hydroponically grown Arabidopsis mutants, defective in high-affinity nitrate transport, were used to distinguish between nitrate and nitrite uptake by means of the short-lived tracers inline image and inline image. This approach allowed us to characterize a nitrite-specific transporter.
  • The Atnar2.1-2 mutant, lacking a functional high-affinity nitrate transport system, is capable of nitrite influx that is constitutive and thermodynamically active. The corresponding fluxes conform to a rectangular hyperbola, exhibiting saturation at concentrations above 200 μM (Km = 185 μM and Vmax = 1.89 μmol g−1 FW h−1). Nitrite influx via the putative nitrite transporter is not subject to competitive inhibition by nitrate but is downregulated after 6 h exposure to ammonium.
  • These results signify the existence of a nitrite-specific transporter in Arabidopsis. This transporter enables Atnar2.1-2 mutants, which are incapable of sustained growth on low nitrate, to maintain significant growth on low nitrite. In wild-type plants, this nitrite flux may increase nitrogen acquisition and also participate in the induction of genes specifically induced by nitrite.


Nitrite (inline image) is a form of inorganic nitrogen (N) that is widely available in soil and aquatic environments. Although soil nitrite concentrations are typically low compared to nitrate, under some conditions they may become elevated (Riley et al., 2001). inline image is produced as a result of the oxidation of inline image by bacteria such as Nitrosomonas and accumulates in aerated soil at elevated pH due to disruption of the second step of nitrification. It is also produced from nitrate in soils and aquatic systems through plant and microbial reduction of nitrate by the enzyme nitrate reductase. While the reduction of nitrate to nitrite by NAD+ can occur anaerobically, the large energy requirement for nitrite reduction to ammonium results in disruption of this second step of nitrate assimilation whenever environmental conditions limit metabolism. Thus, for example, low iron (Fe) availability, and low light may lead to nitrate reduction and subsequent nitrite excretion by phytoplankton (Collos, 1982). Likewise, when nitrite reduction is blocked in the nitrite reductase (niiA5) mutant of Aspergillus nidulans, nitrate uptake is exactly balanced by an equivalent nitrite efflux (Wang et al., 2007b). Significant nitrite can also accumulate at lower pH values in water-logged, poorly aerated soils (Lee, 1979). For example, anoxic conditions in rice paddy fields often cause significant increases in inline image concentration, as a consequence of incomplete denitrification (Samater et al., 1998). Morard et al. (2004) reported that tomato plants grown under anaerobic conditions are able to utilize nitrate for ‘nitrate respiration’ and excrete nitrite. inline image is also abundant in oceans at the base of the euphotic zone where it accumulates as a result of nitrification during summer and excretion by phytoplankton during winter (Meeder et al., 2012). Its concentration in oceans normally ranges from 10 to 400 nM, but can reach up to 4500 nM (Lomas & Lipschultz, 2006). Nitrite availability in soil varies greatly as influenced by conditions discussed above. Burns et al. (1995) reported nitrite concentrations in fertilized grassland soil in Ireland ranging from 0 to 2.7 μg N g−1. Likewise, in agricultural soil in Kansas significant amounts of nitrite were found near groundwater amounting to 0.16 mM (Jones & Schwab, 1993). Uwah et al. (2009) reported above 200 μg g−1 in soil samples from two areas in Nigeria, while nitrite ranged from 0.01 to 0.14 mM in soil samples from Santiago del Estero, Argentina (Lopez Pasquali et al., 2007).

The presence of inline image in nutrient solution can have adverse effects on plant growth (Phipps & Cornforth, 1970; Lee, 1979; Samater et al., 1998; Zsoldos et al., 2001; Ezzine et al., 2011). Its toxicity is, however, more pronounced at high concentrations and low pH (Bancroft et al., 1979), possibly as a result of free nitrous acid (the conjugate acid of inline image) permeation, in addition to plant inline image anion uptake (Zentmyer & Bingham, 1956). However, at more modest concentrations (< 1 mM), it has been reported on numerous occasions that inline image can also be taken up as an alternative N source if available in soil solution. Criddle et al. (1988) reported that wheat (Triticum aestivum) seedlings take up significant amount of nitrite, and Zsoldos et al. (1993) found that wheat seedlings take up inline image faster than inline image. Brinkhuis et al. (1989) measured high rates of inline image uptake in the brown seaweed Laminaria japonica, with a very strong initial rate that stabilizes after 1 h. In summary, inline image uptake has been examined in many species of plants, algae, fungi and bacteria.

Nitrite is an intermediate in the nitrate assimilation pathway, being reduced to inline image by nitrite reductase activity in plastids (Crawford, 1995). Given that nitrate reduction by nitrate reductase occurs in the cytosol, there is therefore an obvious requirement for nitrite transport into the chloroplasts. Recent findings indicate the existence of nitrite-specific transporters in chloroplasts of Arabidopsis thaliana and Cucumis sativa (Sustiprijatno et al., 2006; Sugiura et al., 2007; Ferrario-Mery et al., 2008), and in chloroplasts of Chlamydomonas reinhardtii (Rexach et al., 2000; Mariscal et al., 2004).

The uptake of nitrite from the external environment may occur through nitrate–nitrite bispecific transporters, examples of which are NRT2.1/NAR2 in C. reinhardtii (Galvan et al., 1996), NARU in Escherichia.coli (Jia et al., 2009), NRTA and NRTB in Aspergillus nidulans (Wang et al., 2008), or by inline image-specific transporters as reported in C. reinhardtii (Galvan et al., 1996), E. coli (Jia & Cole, 2005; Jia et al., 2009), Hansenula polymorpha (Serrani & Berardi, 2005), A. nidulans (Wang et al., 2008; Unkles et al., 2011), Neurospora crassa (Gao-Rubinelli & Marzluf, 2004) and Nostoc ANTH (Bhattacharya et al., 2002). Nitrite transporters in A. nidulans (NitA) and E. coli (NirC) belong to a family of formate–nitrite transporters. Nitrite transport system III in C. reinhardtii is coded by the NRT2.3 gene, a member of the Major Facilitator Superfamily (MFS), and functions independently of the NAR2-like gene which is essential for nitrate uptake by NRT2.1 or NRT2.2 (Rexach et al., 1999; Fernandez & Galvan, 2008). NAR2 is also part of a two-component high-affinity nitrate transporter in plants (Okamoto et al., 2006; Orsel et al., 2006). The inducible high-affinity nitrate transport system (iHATS) in Arabidopsis includes MFS proteins such as AtNRT2.1 and AtNRT2.2 that are capable of nitrate uptake only when present as part of a molecular complex with the smaller protein encoded by AtNAR2.1 (Yong et al., 2010).

Nitrite uptake by higher plants has been investigated extensively, but results are inconclusive regarding the existence of nitrite-specific transporters. Ibarlucea et al. (1983) found that nitrite uptake in barley (Hordeum vulgare) seedlings was inducible, and followed Michaelis–Menten kinetics. Jackson et al. (1974b) reported that the addition of nitrite inhibited induction of nitrate uptake in Triticum vulgare, whereas the reciprocal effect on nitrite uptake was not elicited by nitrate, suggesting the possibility of a separate nitrite transport system. Similarly, De La Haba et al. (1990) showed that ammonium inhibited uptake of nitrate, but had no effect on the uptake of nitrite in sunflower (Helianthus annuus). Other reports, however, support the idea of a dual nitrate/nitrite transport system because the two ions have similar properties, and research undertaken with barley showed that nitrate and nitrite ions mutually inhibit uptake competitively (Aslam et al., 1992; Siddiqi et al., 1992).

Nevertheless, there is no conclusive evidence of a nitrite-specific plasma membrane transporter in higher plants. It is generally accepted that nitrite uses nitrate transporters belonging to the MFS nitrate–nitrite porter family, such as AtNRT2.1 in Arabidopsis. The availability of several Arabidopsis mutants defective in high-affinity nitrate transport has enabled us to characterize nitrite uptake in Arabidopsis using inline image to measure nitrite influx. This approach allowed us to clearly distinguish between nitrate and nitrite uptake, and provides evidence for the existence of a nitrite-specific transporter in roots of Arabidopsis.

Materials and Methods

Plant material and growth conditions

Arabidopsis thaliana (L.) Heynh plants (wild-type ecotype Wassilewskija, and knock-out mutant lines Atnrt2.1 (Salk_141712), Atnrt2.2 (Salk_043543) and Atnrt2.1-nrt2.2 (Salk_035429) (Li et al., 2007), and Atnar2.1-2 (Okamoto et al., 2006)) were grown hydroponically under nonsterile conditions as described previously (Zhuo et al., 1999; Okamoto et al., 2003). Three to four seeds were sown into 1.5-cm plastic cylinders filled with acid-washed sand and fitted into floating styrofoam platforms. The platforms floated in plastic containers filled with 7 l of nutrient solution (1 mM KH2PO4, 0.5 mM MgSO4, 0.25 mM CaSO4, 20 μM Fe-EDTA, 25 μM H3BO3, 2 μM ZnSO4, 2 μM MnSO4, 0.5 μM CuSO4, 0.5 μM Na2MoO4 and 1 mM NH4NO3). Solutions were aerated continuously by means of aquarium stones and the pH of solutions was maintained c. 6 by adding powdered CaCO3. Nutrient solutions were completely replaced once a week. Plants were grown for 4 wk, and then deprived of N for the fifth week. To induce iHATS plants were next transferred to solution containing 1 mM KNO3 or KNO2 for 6 h. Growth conditions in the growth room were 8 h of light (100 μmol m−2 s−1 at plant level) and 16 h of dark, at corresponding temperatures of 24 and 22°C, respectively, and a relative humidity c. 70%. In the experiment where pH effects on nitrite uptake were measured, 5 mM 2-(N-morpholino) ethanesulfonic acid (MES) was used as a buffering agent. For growth on MS agar plates, Arabidopsis seeds were sterilized in 1% bleach (plus 0.01% Tween 20) for 15 min, and left for 3 d in sterile water at 4°C for imbibition. Seeds were then sown on half strength solid N-free MS salts media (pH = 6, 0.8% w/v agar), supplemented with 0.25 mM KNO2 or 0.25 mM KNO3. The plates were kept in a vertical position and plants grown for 2.5 wk under the same conditions as described above.

inline image and inline image isotope synthesis

13N-nitrate was generated by proton irradiation of water at the cyclotron facility (Tri-University Meson Facility), University of British Columbia as described earlier (Siddiqi et al., 1989). This radioactive nitrate was used as the source material to generate 13nitrite following the method of McElfresh et al. (1979). Trace quantities of hydrogen peroxide, added to the water target to promote an oxidizing environment for the generation of inline image, were removed by the addition of 1 ml commercial catalase enzyme (2 g l−1) (Sigma Aldrich) because the reduction of nitrate to nitrite by a cadmium column is compromised by the presence of hydrogen peroxide. 13Nitrate was then passed twice through the cadmium column prepared according to McElfresh et al. (1979). This procedure generated > 96% 13N-nitrite as determined by passing the column eluate through an HPLC with a gamma detector in series with the column. Passage through the column resulted in replacement of the 13N-nitrate peak by a peak corresponding to 13N-nitrite. After the cadmium reduction, the column eluate was treated with 100 μl 2N KOH and boiled for 2 min to remove any contaminating inline image. The pH was brought back to neutral by addition of 10% H2SO4 (v/v).

inline image and inline image influx measurements

Nitrate influx, using inline image, was measured as described earlier (Zhuo et al., 1999; Okamoto et al., 2003). The basic components of the solution for pre-treatment, influx and desorption were the same as those of the growth media, except that low concentrations of KNO3 or KNO2 replaced NH4NO3 (exact concentrations are given in each figure). Before measuring 13N influx, plants were pretreated for 5 min with solution containing the same concentration of nitrate or nitrite as the influx solution, and then transferred for 5 min into the influx solution, which was labelled with 13N. After the influx period, roots were desorbed with nonlabelled solution (identical to pre-treatment solution) for 2 min to desorb the radioactive isotope from the apoplast. Plant tissue was immediately harvested, roots were spun at low speed for 20 s to remove excess solution, and thereafter gamma emission was measured using a gamma-counter (MINAXI Auto-Gamma 5000 series; Packard Instruments, Downers Grove, IL, USA). Along with plant tissue samples, samples of influx solutions were counted using the gamma counter, and values used for calculation of 13N content in tissue. Each sample was counted twice to correct for possible 18F contamination. The root tissue was weighed after measuring emission to calculate influx rates.

Temperature coefficient determination and use of metabolic inhibitor

The temperature coefficient (Q10) was determined by incubating 6-wk-old Atnar2.1-2 plants in uptake medium, containing 100 μM KNO2 labelled with 13N-nitrite, at 10 and 23°C, according to standard procedure for 5-min influx measurement, as described above. Q10 values were calculated from the equation Q10 = (R2/R1)(10/T2-T1), where R1 and R2 are influx rates at 10°C (T1) and 23°C (T2), respectively. Detailed protocols are described in (Glass et al., 1990).

In order to evaluate the effect of metabolic inhibition on nitrite uptake, 10 μM of the protonophore carbonyl cyanide m-chlorophenyl hydrazone (CCCP) (Sigma Aldrich) was included in the pre-treatment and influx solution containing 100 μM KNO2. inline image influx was measured according to the procedure described above.

Statistical analysis

All treatments included at least five replicates, and experiments were repeated at least twice. ANOVA calculations and multiple t-test comparisons were carried out using GraphPad Prism v6 (GraphPad Software Inc., La Jolla, CA, USA). The same program was used for direct fitting of curves using the Michaelis–Menten equation or linear fitting.


Kinetics of nitrite uptake in wild-type and mutants defective in nitrate transport

We used wild-type (WT), and previously characterized mutants defective in high-affinity nitrate transport, to examine nitrite uptake in the low concentration (i.e. high-affinity) range. WT and all mutant genotypes were capable of significant 13N-nitrite influx in this concentration range. WT had the highest influx, while mutants showed lower capacities for nitrite uptake (Fig. 1); double mutant Atnrt2.1-nrt2.2 had the lowest influx, followed by Atnrt2.1 and Atnrt2.2. Direct fit of 13N-nitrite influx data using the Michaelis–Menten equation provided high r2 values (Table 1), and allowed estimation of kinetics parameters Km and Vmax (Table 1).

Table 1. Estimates of Km and Vmax parameters for inline image influx in different Arabidopsis thaliana genotypes based on direct fit using the Michaelis–Menten equation
GenotypeKm (μM)Vmax (μmol g−1 FW h−1)
  1. Mean ± SE, = 5–10.

Wild-type125.5 ± 32.66.44 ± 0.86
Atnrt2.2 44.85 ± 7.24.73 ± 0.27
Atnrt2.1 106.8 ± 27.34.58 ± 0.57
Atnrt2.1-nrt2.2 37.9 ± 51.96 ± 0.09
Atnar2.1-2 185 ± 491.89 ± 0.27
Figure 1.

inline image influx at different concentrations of KNO2, in different genotypes of Arabidopsis thaliana. (a) Wild type (WT); (b) Atnrt2.2; (c) Atnrt2.1; (d) Atnrt2.1-nrt2.2, fitted line is a direct Michaelis–Menten fit (± SD of five replicates).

High-affinity nitrate influx into roots of the Atnar2.1-2 mutant is virtually absent at low external nitrate concentration, exhibiting 5% or less of WT flux and demonstrating a linear pattern of concentration response (Okamoto et al., 2006; Orsel et al., 2006). By contrast, inline image influx in the Atnar2.1-2 mutant in the concentration range 10 to 250 μM was substantial, and followed Michaelis–Menten kinetics (Fig. 2), with Km = 185 ± 45 μM and Vmax = 1.89 ± 0.27 μmol g−1 FW h−1. We also measured inline image and inline image influx at 100 μM nitrate and nitrite, respectively, in mutants and WT in the same experiment to reduce potential variation in plant growth and other variables that, in separate experiments, might make direct comparisons more difficult. Compared to WT, nitrate influx in the Atnrt2.1-nrt2.2 double mutant was reduced by 52%, while nitrite influx was reduced by only 15% of WT values. Nitrate influx in the Atnar2.1-2 mutant was only 5% of WT values, while nitrite influx remained at 60% of WT (Table 2).

Table 2. Influx of inline image and inline image at concentrations of 100 μM KNO3 and KNO2, respectively, in different Arabidopsis thaliana genotypes
 inline image influx (μmol g−1 FW h−1)% reduction of WTinline image influx (μmol g−1 FW h−1)% reduction of WT
  1. ± SE of five replicates; different letters indicate significant difference < 0.05, t-test within a treatment.

Wild-type (WT)5.25 ± 0.39a03.69 ± 0.40a0
Atnrt2.1-2.2 2.69 ± 0.07b483.20 ± 0.55a15
Atnar2.1 0.24 ± 0.05c951.90 ± 0.37b40
Figure 2.

inline image influx at different concentrations of KNO2, in the Atnar2.1-2 mutant of Arabidopsis thaliana, fitted line is a direct Michaelis–Menten fit (± SD of 10 replicates).

Effect of induction by nitrite/nitrate and pH on inline image influx of the Atnar2.1-2 mutant

inline image influx of the Atnar2.1-2 mutant was measured at 100 μM KNO2 after N-starved plants were induced for 3–12 h with 1 mM KNO2 or KNO3. There was no significant effect of induction by either nitrite or nitrate on influx of inline image (Fig. 3). Because nitrous acid is an uncharged molecule that might diffuse across the plasma membrane, we evaluated the effects of pH on inline image influx in the Atnar2.1-2 mutant from 100 μM KNO2. Nitrite influx increased substantially as pH was lowered from 6 to 4. Above pH 6 (from 6 to 8) there was no significant effect on influx (Fig. 4). The standard pH of the nutrient solution used in all experiments was c. 6.5.

Figure 3.

Induction of inline image influx in the Atnar2.1-2 mutant of Arabidopsis thaliana by 1 mM nitrate or nitrite (+ SD of five replicates). (*, P < 0.05).

Figure 4.

Effect of pH on inline image influx from 100 μM KNO2 in the Atnar2.1-2 mutant of Arabidopsis thaliana (+ SD of five replicates). 6-WT signifies WT plants at pH 6.

Nitrate as a competitor of inline image influx

Based upon previously reported competitive inhibition of nitrite uptake by nitrate and vice versa, it has been suggested that nitrate and nitrite use the same transporters for entry through the plasma membrane. To determine the effect of nitrate addition on nitrite influx, we used inline image to measure nitrite influxes in WT and the Atnar2.1-2 mutant in the presence and absence of nitrate. Figure 5(a) shows Lineweaver–Burk plots of nitrite influx based on four different concentrations of nitrite in WT, with and without 250 μM KNO3. The intersection of plot lines at the y-axis indicates competitive inhibition of nitrite uptake by nitrate. By contrast, Lineweaver–Burk plots of nitrite influx in the Atnar2.1-2 mutant are parallel, and almost aligned (Fig. 5b), showing virtually no effect of nitrate on nitrite influx in this mutant.

Figure 5.

(a) Lineweaver–Burk plots of inline image influx (V is influx rate = μmol g−1 FW h−1) from different concentrations (S is substrate concentration, in μM) of KNO2 in Arabidopsis thaliana wild-type (WT) in the presence and absence of 250 μM KNO3; (b) Lineweaver–Burk plots of inline image influx (V = μmol g−1 FW h−1) from different concentrations (S, in μM) of KNO2 in Atnar2.1 in the presence and absence of KNO3. Circles, -KNO3; squares, +KNO3. Symbols represent the means of five replicates ± SD.

Effect of temperature, ammonium and metabolic inhibitor on inline image influx

It is now accepted that iHATS nitrate uptake is thermodynamically active. We used 13N-labelled nitrite to determine the effect of temperature reduction and a metabolic inhibitor on influx in Atnar2.1-2 plants. Plants were grown and induced according to the standard procedure, and subjected to the following conditions: reduced temperature during the influx period, 6 h pretreatment with 1 mM NH4H2PO4 or 10 μM CCCP (a protonophore) for 5 min in the pre-treatment solution with 100 μM KNO2, before incubation in the tracer-labelled influx solution. Nitrite uptake rates were lower at lower temperature, and were used to calculate Q10 coefficients (Table 3), that varied from 1.72 to 2.14, according to the temperature range examined. The effect of short exposure to the protonophore was even more pronounced, diminishing influx of inline image in the Atnar2.1-2 mutant from 3.86 μmol g−1 FW h−1 in control, to 1.25 μmol g−1 FW h−1 in CCCP-treated plants (Table 4). Six hours of ammonium treatment decreased nitrite influx dramatically, from 3.8 to 0.8 μmol g−1 FW h−1.

Table 3. Calculated temperature coefficient (Q10) values for influx of inline image in the Arabidopsis thaliana Atnar2.1-2 mutant, at 100 μM KNO2
Temperature range (˚C)Q10 factor
  1. Plants induced for 6 h with 1 mM KNO3.

Table 4. Effect of ammonium treatment and protonophore carbonyl cyanide m-chlorophenyl hydrazone (CCCP) addition on influx of inline image in Arabidopsis thaliana Atnar2.1-2, at 100 μM KNO2
TreatmentInflux (μmol g−1 FW h−1)
  1. Plants induced for 6 h with 1 mM KNO3 (± SE of five replicates).

Control3.86 ± 0.36
CCCP1.25 ± 0.08
Ammonium (6 h)0.84 ± 0.15

Comparison of nitrate and nitrite as N sources for growth of WT and Atnar2.1-2

A prediction arising from the presence of an independent nitrite transporter is that at low concentrations, growth of Atnar2.1-2 on nitrite should be superior to that on nitrate. Figure 6(a–c) confirms this prediction: Fig. 6(b) shows that on low nitrate Atnar2.1-2 established virtually no growth, whereas growth on nitrite was substantial, although less than that of WT (Fig. 6a). Fresh weights of Atnar2.1-2 grown on 250 μM nitrate or nitrite, were 15% and 45%, of WT values, respectively (Fig. 6c).

Figure 6.

Growth of Arabidopsis thaliana wild-type (WT) and Atnar2.1-2 mutant on ½ Murashige and Skoog (MS) media plates supplemented with 0.25 mM KNO2 (a) and 0.25 mM KNO3 (b); (c) plant FW of Atnar2.1-2 mutant grown on 0.25 mM KNO2 and 0.25 mM KNO3 expressed as% of WT. Columns represent the means of fifteen replicates ± SD.


Nitrite is an important N source reported to be utilized by many organisms, including bacteria (Bhattacharya et al., 2002; Jia et al., 2009), fungi (Schloemer & Garrett, 1974; Wang et al., 2008), phytoplankton (Gabas et al., 1981; Cresswell & Syrett, 1982; Sivasubramanian & Rao, 1988; Abdel-Basset & Ali, 1995), algae (Brinkhuis et al., 1989; Galvan et al., 1991) and higher plants (Jackson et al., 1974a; Criddle et al., 1988; Zsoldos et al., 1993). A recent study by Wang et al. (2007a) showed that nitrite is also a potent signal for N metabolism transcriptome regulation in Arabidopsis roots, uniquely inducing significant numbers of nitrate-inducible genes (in wild-type (WT) Arabidopsis) that were not induced in NR mutants. It was concluded, therefore, that a significant number of the genes (apparently induced by nitrate) were actually induced in response to nitrite, produced as a result of nitrate reduction. Nevertheless, the importance of nitrite as a nutrient for plants has been overlooked, and it has mainly been studied as a toxic agent (Lee, 1979; Samater et al., 1998; Zsoldos et al., 2001; Ezzine et al., 2011). The toxicity associated with nitrite is prevalent under conditions of low pH and high concentrations (Zsoldos et al., 1995; Ezzine et al., 2011), conditions that favour conversion of inline image to nitrous acid (HNO2). However, these conditions are not widespread in the environment and therefore its importance as a plant nutrient requires greater emphasis.

Physiological measurements of nitrate/nitrite competition at the uptake level have suggested that nitrate and nitrite share the same transport system, based upon observed competitive inhibition of nitrate uptake by nitrite (and vice versa) in barley (Aslam et al., 1992; Siddiqi et al., 1992). Therefore, it might be considered that a distinct (unique) nitrite transporter would be redundant. Yet in particular cases (e.g. Chlamydomonas reinhardtii and Aspergillus nidulans) where it was possible to completely eliminate nitrate uptake, growth on nitrite was still possible and subsequent studies identified specific genes encoding nitrite transporters that were incapable of nitrate carriage (Rexach et al., 1999; Wang et al., 2008). In A. nidulans the responsible nitrite transporter (NitA) is a member of the FNT group which is distinct from the NrtA and NrtB nitrate transporters that belong to the Nitrate-Nitrite Porter (NNP) family, as do the Arabidopsis high-affinity nitrate transporters (AtNRT2.1 and AtNRT2.2). Unfortunately, nitrate uptake in the Arabidopsis Atnrt2.1-nrt2.2 double mutants is not completely eliminated (retaining c. 40% of WT nitrate influx, as shown in Table 2 and in Filleur et al., 2001 and Li et al., 2007), so these mutants would not provide the appropriate context in which to identify a unique nitrite transporter. In place of Arabidopsis NRT2 mutants we selected to employ Atnar2.1 mutants, in which nitrate influx is reduced to c. 3–5% of WT values (Okamoto et al., 2006; Orsel et al., 2006; Yong et al., 2010). In previous biochemical studies of high-affinity nitrate influx by AtNRT2.1, it was demonstrated that in T-DNA mutants of AtNAR2.1, despite the presence of AtNRT2.1 mRNA, the corresponding protein was absent from plasma membrane (PM) preparations (Wirth et al., 2007; Yong et al., 2010). Indeed, Yong et al. (2010) showed that AtNRT2.1 and AtNAR2.1 form a 150-kDa PM complex thought to consist of two sub-units each of the two polypeptides. This complex is absent from Atnar2.1-2 mutant plants. Therefore, this mutant proved to be the most suitable genotype for further investigations of nitrite influx as there are no functional nitrate transporters to mask the contribution of a putative nitrite-specific transporter(s).

Figure 1(a) shows that nitrite influx in WT is substantial, of the order of that reported for nitrate influx. Mutants disrupted in NRT2.2 (Fig. 1b), NRT2.1 (Fig. 1c) and NRT2.1/NRT2.2 (Fig. 1d) exhibit reduced nitrite influx that is quantitatively consistent with substantial nitrite transport via the high-affinity nitrate transporters. This inference is supported by the form of the Lineweaver–Burk plot (Fig. 5a) in which it is demonstrated that in WT plants nitrate reduced nitrite influx competitively.

The highest nitrite influx reduction was observed in the Atnar2.1-2 mutant, in which virtually all high-affinity nitrate influx is eliminated (Fig. 2). It is noteworthy that in this mutant inline image influx conforms to a rectangular hyperbola, and the r2 for regression was significantly higher than for a linear fit to the data. Supporting Information Fig. S1 shows four genotypes together for comparative purposes. As these experiments (Figs 1, 2) were performed separately for each genotype, we measured influx at a single concentration in order to examine different genotypes side-by-side, to better compare uptake of nitrite vs nitrate in WT, double and Atnar2.1-2 mutants. Because of the short half-life of 13N (t0.5 = 9.96 min) it is not possible to accommodate large numbers of treatments. Reduction of both nitrite and nitrate influx was the highest in the Atnar2.1-2 mutant (Table 2). Nevertheless, while Atnar2.1-2 plants retained only 5% of WT nitrate influx, nitrite influx was retained at 60% of WT (Table 2). This finding signifies the existence of an additional transport mechanism for nitrite, independent of the AtNAR2.1 gene. Likewise, in A. nidulans the NitA gene encodes a nitrite-specific transporter that appears to function independently of any NAR2-like polypeptide (Unkles et al., 2011), and also the nitrite transport system III in C. reinhardtii (coded by CrNRT2.3) which is independent of the CrNAR2 gene (Rexach et al., 1999). Kotur et al. (2012) reported that all NRT2 transporters in Arabidopsis, with the exception of AtNRT2.7, require NAR2.1 for functional nitrate transport. Nitrite fluxes that were lower in Atnar2.1-2 than in the double mutant Atnrt2.1-nrt2.2 (Table 2) suggest that despite the presence of a nitrite-specific transporter, in WT plants nitrite may be absorbed by both nitrate–nitrite transporters and this nitrite-specific transporter. However, because soil nitrate concentration typically exceeds that of nitrite, the former might be competitively inhibited, whereas the latter could function independently of nitrate. This may be an important consideration with respect to the induction of nitrite-inducible genes (Wang et al., 2007a).

Incubation of Atnar2.1-2 plants in nitrate or nitrite for 0, 3, 6 or 12 h before influx experiments, demonstrated that the putative HATS nitrite-specific transporter is not upregulated by those treatments; that is, it is not inducible (Fig. 3). This provides another difference between nitrite influx via AtNRT2.1, whose expression is increased several fold by exposure to nitrate (data not shown), and the putative nitrite-specific transporter. Incubation in 1 mM nitrate for 12 h decreased nitrite influx significantly, possibly due to feedback by ammonium or other nitrogen metabolites that might act as signals for uptake regulation, similarly to the regulation of the nitrate HATS (Zhuo et al., 1999; Vidmar et al., 2000; Nazoa et al., 2003).

Nitrite influx in Atnar2.1-2 mutants increased between pH 6 and 4 (Fig. 4). In part, this effect might be explained by permeation of HNO2 (pKa = 3.4) across the plasma membrane at low pH. Yet despite a 100-fold increase of nitrous acid concentration between pH 6 and pH 4 13N influx increased only six-fold (Fig. S2). At pH values above 6 > 99.9% of inline image is in the anionic form and hence its entry across the plasma membrane is a metabolically dependent flux, demanding the participation of membrane transporters. In all other experiments influx media were maintained at pH c. 6.5, ensuring that virtually no HNO2 permeation would contribute to measured 13N influx. Nevertheless under low pH conditions, as, for example, in forest soils, nitrous acid permeation may be significant.

Owing to their similar characteristics, nitrate and nitrite are known to compete for the binding site of nitrate/nitrite porters, and exhibit competitive inhibition of uptake (Aslam et al., 1992; Siddiqi et al., 1992). In the present experiments with Arabidopsis, we have observed similar results in WT, where the addition of 250 μM nitrate decreased and inhibited nitrite influx competitively (Fig. 5a), suggesting that both ions are using the same transporter, most probably AtNRT2.1, the major contributor to iHATS nitrate uptake (Li et al., 2007). By contrast, in the Atnar2.1-2 mutant, lacking either NRT2.1 or NRT2.2 activity, nitrate was without effect on nitrite influx (Fig. 5b), suggesting the operation of a nitrite-specific transporter that is incapable of nitrate transport.

Studies of the energetics of nitrate uptake revealed active transport mechanisms, suggested to take the form of a symport of two protons with one inline image (reviewed in Crawford & Glass, 1998). The temperature coefficient (Q10) is a quotient defining the ratio of a reaction at toC/t-10°C. The observed values (Table 3), that are significantly higher than 1, suggest that nitrite influx at pH 6.5 is thermodynamically active, unlike passive processes that have Q10 values close to 1. Clarkson & Warner (1979) reported that nitrate uptake in ryegrass is very sensitive to temperature. Likewise, Glass et al. (1990) have reported high Q10 values for nitrate uptake in barley. In addition, the three-fold reduction of nitrite influx into roots of Atnar2.1-2 plants in the presence of the protonophore CCCP, known to disrupt the proton gradient and inhibit ATP synthesis, is consistent with the metabolic dependence of influx rather than the result of a passive permeation of nitrous acid (Table 4). Ammonium is well documented to inhibit nitrate uptake. For example, Lee & Drew (1989) reported that inhibition was evident within 3 min of ammonium application due to its direct effects on nitrate transport. In addition, ammonium may reduce nitrate influx through effects operating via glutamine at the transcriptional level (Vidmar et al., 2000; Nazoa et al., 2003). In the case of nitrite uptake, findings on the effects of ammonium have been controversial. On the one hand, De La Haba et al. (1990) concluded that ammonium had no effect on nitrite uptake in sunflower. Ibarlucea et al. (1983), on the other, reported that ammonium diminished nitrite uptake in barley. In the present study, 6-h ammonium treatment of Atnar2.1-2 plants reduced nitrite influx from 3.8 to 0.8 μmol g−1 FW h−1. This provides an additional argument against passive diffusion of nitrite or nitrous acid across the PM, and supports the proposal of a distinct nitrite transport system that is downregulated by ammonium. The importance of the distinct nitrite transport system in Arabidopsis is evident from the comparative growth of the Atnar2.1-2 plants on low nitrate and nitrite (Fig. 6a–c). It has been shown that Atnar2.1-2 mutant is incapable of growth on low nitrate (250 μM) as sole source of N (Fig. 6b; Okamoto et al., 2006; Orsel et al., 2006; Yong et al., 2010). The mutant stops growing after seed N reserves are depleted, and fails to develop the first true leaves, while cotyledons become yellow. However, this mutant grows successfully on low nitrite as a sole N source (Fig. 6a). Although smaller than WT, young Atnar2.1-2 plants maintain 45% of WT weight on nitrite, while reaching only 15% of WT weight on nitrate media (Fig. 6c). In summary, for the first time in higher plants, we have provided the following evidence of the existence of a nitrite-specific transporter:

  • The Atnar2.1-2 mutant, lacking a functional iHATS for nitrate influx, is, nevertheless, still capable of significant nitrite influx (60% of WT) that conforms to Michaelis–Menten kinetics. While the Atnar2.1-2 plants cannot sustain growth on low nitrate, this nitrite uptake allows them to grow on low nitrite as sole N source.
  • Unlike nitrate influx, this putative nitrite-specific influx is not inducible but constitutive.
  • Nitrite influx by this nitrite-specific transporter is unaffected by nitrate competition, and the putative nitrite transporter is incapable of nitrate uptake.
  • Nitrite influx by means of the nitrite-specific transporter is an active process, and is subject to down-regulation by ammonium.


The authors gratefully acknowledge financial support for this project in the form of an NSERC Discovery Grant (0570) to ADMG and a University of British Columbia graduate fellowship to ZK. The authors gratefully acknowledge the assistance of the University of British Columbia (TRIUMF) cyclotron facility for provision of 13N. The authors would like to thank Arabidopsis Biological Resource Center for provision of T-DNA Arabidopsis insertional mutants.