Introduction

Nitrate (NO3−) is the main nitrogen source for most plant species1. The uptake of nitrate by plant roots and its translocation to the other organs involve a set of membrane transport proteins with different functional properties: low, high or dual affinity; influx, efflux or bidirectional transport; pH dependencies; channels or transporters2. The different nitrate transporters are found within 5 protein families: NRT22; SLAC/SLAH3; ALMT4; ClC5 and NPF (NRT1/PTR Family)6. Members of this last family, NPF, display sequence and structure homologies with animal and bacterial PTR proteins7,8. The structure of NPF proteins comprises twelve transmembrane helices with cytosolic N- and C-terminal ends and with a long cytosolic loop between the 6th and the 7th transmembrane domain. The funding member of this family AtNPF6.3/NRT1.1/CHL1 is a nitrate transporter playing several roles in roots: it is involved in nitrate influx9 and translocation to the shoot10, but it is also a nitrate sensor11,12,13,14. Since the characterization of AtNPF6.3/NRT1.1, 11 other NPFs have been characterized as nitrate transporters in Arabidopsis6,15. The NPF family of Arabidopsis comprises 53 members and other substrates are also transported by NPF proteins such as glucosinolates16, histidine17, dipeptides18,19, ABA20,21 and auxin12. Up to now, protein sequences cannot be used to predict the substrate specificity of the uncharacterized NPFs.

In order to identify substrates of yet uncharacterized Arabidopsis NPFs, we have designed two functional screens in heterologous systems. The first one uses functional complementation of yeast strain defective in peptide uptake (Y06009, ΔPTR2 Δhis Δleu Δmet Δura). The principle is that the yeast strain deleted for its dipeptide transporter PTR219 is not able to grow with dipeptides as the sole source of nitrogen. Expression of plant dipeptide transporters will restore growth19. The second screen is functional expression in xenopus oocytes that has been used to identify several transporters22. It consists in injecting cRNA coding a protein with unknown function and screening for transport activity by supplying potential substrates in the external medium. This approach has been successfully employed to identify AtNPF2.10/GTR1 and AtNPF2.11/GTR2 as glucosinolate transporters16.

Here, we expressed 26 Arabidopsis NPFs in yeast and xenopus oocytes and screened for dipeptide and nitrate transport activity. We identified NPF5.5 (At2g38100) and NPF5.10 (At1g22540) as new nitrate transporters. We demonstrate that NPF5.5 is expressed in the embryo and affects its nitrogen content.

Results

To optimize the expression of NPFs in heterologous system we used their Coding DNA Sequence (CDS, from ATG to stop without any plant UTRs). Twenty-six CDS were either obtained from ABRC (in pENTR223) or amplified by PCR and cloned in a Gateway® entry vector (pENTR/D/TOPO or pDONR207 or pCR8-GW-TOPO). LR reactions were performed to transfer these CDS in destination vectors designed for inducible yeast expression (pYES-DEST52®) or xenopus oocyte expression, pGEM-GWC: this vector is based on pGEM-HE23 with a Gateway® cassette inserted between the 5′ and 3′ β-globin UTR from xenopus. NPFx-pGEM-GWC were linearized and used for in vitro cRNA synthesis (see methods for details).

The mutant yeast strain (Y06009, obtained from Euroscarf) used in this study is deleted for the only dipeptide yeast PTR transporter (ΔPTR2). This strain is also unable to synthetize His, Leu, Met and Ura (His−, Leu−, Met−, Ura−). Efficient NPF transformation with the pYES-DEST52 vector restored the auxotrophy for uracil. Therefore, standard growth of transformed yeast was achieved adding His, Leu and Met to the medium (Figure 1, black bars). As a consequence, the media used for the screen was the minimal media supplemented with His, Met and Leu in the form of Leu-Leu (Figure 1, green bars). Yeast can grow only if they are able to take up Leu-Leu as Leu source. Yeasts were individually transformed with 26 different NPFs. The only growing strains were those expressing the already known Arabidopsis dipeptide transporters: AtNPF8.1/PTR1 or AtNPF8.3/PTR224. No other NPF complementing the mutant phenotype could be identified among the 26 tests (Figure 1).

Figure 1
figure 1

Screen for dipeptide transport complementation by NPFs in S. cerevisiae.

Control (C, non-transformed) and NPF-transformed ΔPTR2 yeast mutants were grown for 20 hours on SC minimal media (complemented with uracil for control yeast) containing His, Leu and Met (black bars), SC minimal media containing Met (red bars) and SC minimal media containing His, Met and Leu-Leu (green bars). (inset) Growth kinetics in the 3 different media for yeasts expressing NPF8.1. Results from one experiment representative of 2–3 independent repeats.

Full size image

In the second screen, xenopus oocytes were injected with NPF coding cRNA and15NO3− accumulation studies were performed: oocytes were bathed in 30 mM K15NO3 for 2 hours and15N was quantified in oocytes (see methods). In this screen, several NPF-expressing oocytes display15N accumulation higher than non-injected ones (Figure 2). The positive controls, NPF6.3/NRT1.19, NPF1.1/NRT1.1215, NPF1.2/NRT1.1115 NPF2.9/NRT1.925 and NPF6.2/NRT1.426 display such behavior. Interestingly, expression of one other NPF induced a significant 15N accumulation, similar to the NPF6.3/NRT1.1 level: NPF5.5 (At2g38100). Another member of this subfamily, NPF5.10 (At1g22540), induce higher 15N accumulation than control oocytes, even if the accumulation level is lower than for NPF5.5 (Figure 2). This screen should be continued with the NPF that have not been tested but also with the one tested here but that do not display accumulation higher than wild-type. Indeed at least two proteins that have been shown to mediate nitrate influx into the oocytes were not identified in our screen: NPF2.13/NRT1.727 and NPF4.6/NRT1.228.

Figure 2
figure 2

Screen for nitrate transport activity of NPFs in xenopus oocytes.

Control (non-injected) and NPF-injected oocytes were bathed in 30 mM K15NO3 for 2 hours.15N accumulation is expressed as % of accumulation in NPF6.3/NRT1.1-expressing oocytes. Values are mean +/− SEM from 3 experiments. *** indicate significant difference with control oocytes at p < 0.001 (t-test).

Full size image

RT-PCR has been performed to demonstrate that NPF5.5 is expressed in roots and shoots29. Based on publicly available expression data (eFP browser, http://bbc.botany.utoronto.ca/efp30), we found that NPF5.5 is also expressed in the embryo. This expression is not dependent on developmental stage31. Quantitative PCR experiments confirmed the expression of NPF5.5 in the embryo (Figure 3b). To determine the role of NPF5.5, we isolated embryos from two homozygous KO lines (Salk_039337 and GK756F02, Figure 3a) and found that no full-length npf5.5 mRNA accumulated in these genotypes (Figure 3b). npf5.5-1 and npf5.5-2 mutant plants growth in greenhouse conditions is similar to WT plants. Total nitrogen content of the embryos, analyzed by MS, showed a 7–8% reduction in the two KO lines compared to Col-0 (Figure 3c).

Figure 3
figure 3

Characterization of npf5.5 KO lines.

(a) Schematic representation of NPF5.5 insertion lines. Blue arrows represent the sens and antisens oligonucleotides used for Q-PCR analysis. (b) Q-RT-PCR analysis were performed on mRNA isolated from embryo at the bent cotyledon stage. Values are mean +/− SEM from 3 experiments. *** indicate significant difference with control oocytes at p < 0.001 (t-test). (c) Nitrogen content in embryos of two NPF5.5 KO lines. The total nitrogen content was determined in embryos of two NPF5.5 KO lines (npf5.5-1 and npf5.5-2). Values are expressed in percent of N content decrease compared to total N content in WT embryos. The average was made on 5 batches of 75 embryos, n = 5.

Full size image

Neither full-length NPF5.5 cDNA nor 5′-ESTs sequences were present in databases, but data from high-density oligonucleotide arrays spanning the entire genome32 allowed us to speculate that a NPF5.5 exon is present before the predicted ATG. We therefore characterized this region and performed PCR on cDNA. Two different cDNAs with different in frame ATG were identified (Figure 4a and b). These two cDNAs were named NPF5.5a and NPF5.5b. They are also both different from the one used in the screen that was cloned from predicted sequences (named NPF5.5 predicted in Figure 4). Nitrate transport activity of the two proteins coded by these two mRNAs was measured in xenopus oocytes. Both NPF5.5a and NPF5.5b expressing oocytes accumulated more15N than control oocytes, demonstrating that both versions are nitrate transporters (Figure 4c).

Figure 4
figure 4

Alternative NPF5.5 transcripts.

(a) Schematic representation of NPF5.5 5′ genomic and RNA sequences. Blue arrows represent the sens oligonucleotides used for identification of NPF5.5a and b. (b) Amino-acid alignment of the N-terminal region of the 3 different NPF5.5 proteins. (c) Influx in control (non-injected) oocytes (white bar), in NPF6.3-expresssing oocytes (black bar) and in the 3 forms NPF5.5-expressing oocytes (grey bars) after 2 hours of incubation in 30 mM15NO3 solution. P: predicted sequence, a and b: 2 NPF5.5 forms. Influx in control oocytes (white bar), in NPF6.3-expresssing oocytes (black bar) and in NPF5.5-expressing oocytes (grey bars) after 2 hours of incubation in 30 mM15NO3 solution. *** and ** indicate significant difference with control oocytes at p < 0.001 and p < 0.005 respectively (t-test).

Full size image

Discussion

Because nitrogen is an essential nutrient for plants, we have developed functional screens in heterologous systems (yeast and xenopus oocytes) to identify new nitrogen transporters in Arabidopsis. In one family of nitrogen transporters, NPF (NRT1/PTR Family6), several proteins have been identified as dipeptides, amino-acids and nitrate transporters, three main forms of nitrogen for nutrition of the plants. We expressed in both yeast and xenopus oocytes 26 of the 53 Arabidopsis NPF CDS. While we were unable to identify new dipeptide (Leu-Leu) transporters in yeast, our screen in xenopus oocytes identified two as yet uncharacterized nitrate transporters: NPF5.5 and NPF5.10. While the NPF5 subfamily is the largest one in plants (663 members out of 2384 in 31 fully sequenced plant genomes6), only one member AtNPF5.2/PTR3 has been functionally characterized in Arabidopsis and shown to be a dipeptide transporter33. None of the NPF5 members (NPF5.1, 3–5, 7, 9–14, 16) tested in this study complemented yeast deficiency in Leu-Leu uptake. Nevertheless, NPF5.5 and NPF5.10 are the first nitrate transporters identified in this subfamily (Figure 2).

Human and animal feeding relies on organic nitrogen produced by plants. Plant seeds provide an important source of this nitrogen. Several nitrate transporters have been demonstrated to be expressed in the seed2. For example, NRT2.7 is a high-affinity vacuolar nitrate transporter, highly expressed in seeds and embryos and involved in nitrate accumulation in the seed34. NPF2.12/NRT1.6 is expressed in vascular tissues of the silique and in funiculus and is important for embryo development35. Several organic nitrogen (peptide) transporters are involved in nitrogen accumulation in embryos18,36.

NPF5.5 is expressed in different tissues within the plant29 but is the first NPF transporter gene so far showing an expression in the embryo. Transcriptomic analyses on embryos isolated at different stages showed that NPF5.5 level is rather stable during embryo maturation31. Its expression in the embryo (at bent cotyledon stage) has been confirmed by Q-RT-PCR (Figure 3b). To assess the role of NPF5.5 in the embryo, we have isolated two homozygous KO lines in which no full-length NPF5.5 mRNA was found (Figure 3b). Nitrogen content (as % of dry weight) was determined for WT and the 2 KO lines on 5 batches of 75 isolated embryos. Embryo nitrogen content is 7–8% lower in both KO lines compared to WT (Figure 3c), demonstrating an effect of NPF5.5 function on the embryo nitrogen content. Since NPF5.5 gene was knock out in all the plant, we cannot exclude that part of this phenotype involves expression of NPF5.5 in other tissues.

NPF5.5 is one of the Arabidopsis genes for which there is no cDNA sequence available in databases but we have identified two different mRNAs transcribed from this gene. These different forms most likely arise from different transcription initiation site (instead of alternative splicing). Both proteins encoded by these two different transcripts are nitrate transporters (Figure 4c). Further experiments are needed to identify the role of these different transcripts. Two non-exclusives kind of regulation could exist. The first one is a post-transcriptional regulation due to different 5′-UTR. The second one is a post-translational regulation involving the 22 amino acids extended N-terminus of NPF5.5b that contains phosphorylable residues (2 serines and 2 threonines).

Methods

Cloning in Gateway entry vectors

NPF CDS were either obtained from ABRC (cloned in pENTR223 for NPF2.9, 2.13, 4.3, 4.5, 5.1, 8.1 and 8.3) or cloned in pENTR/D/TOPO (for clones 2.8, 4.4, 5.3, 5.5, 5.9, 5.10, 5.14 and 6.3), pDONR207 (for clones NPF1.2, 4.6, 5.11, 5.16 and 6.2) or pCR8-GW-TOPO (for clones NPF1.1, 5.4, 5.7, 5.12, 5.13 and 8.5). Each clone was sequenced and compared to Col-0 genomic sequence. LR reaction was performed according to the manufacturer's instructions (Life Technologies), to clone the CDS in yeast expression vector (pYES-DEST52) or xenopus oocyte expression vector (pGEM-GWC).

The primers used for NPF5.5 variants identification are: NPF5.5a-s (ATGAGAGATATGGCGGCAGAG) and NPF5.5b-s (ATGACCAATATAACAGAGTATGATAAAGAA).

Expression in yeast

S. cerevisiae Y06009 strain (BY4741; MATa; ura3Δ0; leu2Δ0; his3Δ1; met15Δ0; YKR093w::kanMX4), obtained from Euroscarf and defective for dipeptides uptake was transformed with the cDNAs of the 26 NPF members, cloned in the yeast expression plasmid pYES-DEST52. Cells were transformed according to37 and plated for 3 days on YNB medium (yeast nitrogen base 1.7 g.l−1, (NH4)2SO4 5 g.l−1, D-glucose 20 g.l−1, agar 20 g.l−1) supplemented with the appropriate amino acids for transformant selection (Histidine, Leucine, Methionine). For uptake experiments, 5 ml cultures of cells were grown for two days in SC minimal medium (yeast nitrogen base 1.7 g.l−1, galactose 20 g.l−1, agar 20 g.l−1) supplemented with the appropriate amino acids, washed three times in water and resuspended in the adequate volume of water to have a OD 600 = 1. Ten μl of these cells were grown for 20 h in SC minimal medium supplemented with 20 μg.ml−1 of Methionine or 20 μg.ml−1 of Histidine, 60 μg.ml−1 of Leucine, 20 μg.ml−1 of Methionine or with 20 μg.ml−1 of Histidine, 20 μg.ml−1 of Methionine and Leucine-Leucine 200 μM.

Expression in xenopus oocytes

NPFx-pGEM-GWC were linearized and in vitro transcribed with mMessage mMachine T7 Ultra Kit following manufacturer protocole (Life Technologies). Oocytes were obtained and injected as previously described38. Briefly, 3 days after NPF cRNA injection, injected or control (non-injected) oocytes were incubated 2 hours in 2 ml of ND96 medium (pH 6.5) containing 30 mM of15N-nitrate (atom %15N abundance: 99,9%). Oocytes were then washed 5 times in 15 ml of ND96 medium (pH 6.5) at 4°C. Batches of 2 oocytes were then analyzed for total N content and atom %15N abundance by Continuous-Flow Mass Spectrometry, using an Euro-EA Eurovector elemental analyzer coupled with an IsoPrime mass spectrometer (GV instruments, Crewe, UK).

Plant material and growth

The Arabidopsis thaliana accession used is Columbia (Col-0). The npf5.5 mutants are SALK_039337 (npf5.5-1) and GABI_756F02 (npf5.5-2) and were obtained from NASC and GABI-kat, respectively.

Arabidopsis seeds were sown in pots (Neuhauss Humin substrate N2, Klasmann-Deilmann, Germany) and grown in greenhouse under long-day conditions (16 H light at 23°C and 8 H dark at 21°C, humidity is not controlled but is around 50%). When the external light is not sufficient, it is supplemented with sodium vapor lamps (150 μE.m−2.s−1). Plants are watered once or twice per week. Under these conditions, the seed to seed cycle is 6 to 8 weeks, depending on the season.

Relative gene expression

Embryos were isolated (at bent cotyledon stage) from developing siliques of Arabidopsis39. Frozen (−80°C) samples were homogenized for 1 min at 30 s−1 (Retsch mixer mill MM301; Retsch, Haan, Germany) in 2 mL tubes containing one tungsten bead (2.5 mm diameter). Total RNA was extracted from homogenized tissues by using TRIzol reagent (Life Technologies, Carlsbad, CA) according to the manufacturer's protocol. Three micrograms of RQ DNase (Promega, Madison, WI) digested total RNA was used to prepare cDNA by reverse transcription with M-MLV reverse transcriptase (Promega) and Gene expression was determined by quantitative real-time PCR (Roche Diagnostics, Mannheim, Germany) using LightCycler FastStart DNA Master SYBR Green I (Roche Diagnostics). Expression levels of NPF5.5 (F: GTACATTCTACTCGGTTCAATCACA and R: ACTCCCACGTTTAATAGAACCA) were normalized to expression levels of the CLATHRIN (At4g24550 F, AGCATACACTGCGTGCAAAG and R: TCGCCTGTGTCACATATCTC) and ACTIN (At3g18780.1 F: CTGAGGCTGATGATATTCAACC and R: ACACTGGGAAAAACAGCCC).

Total nitrogen content

Embryos were isolated (at bent cotyledon stage) from developing siliques of Arabidopsis39. Batches of 75 embryos were analyzed for total N content by Continuous-Flow Mass Spectrometry, using an Euro-EA Eurovector elemental analyzer coupled with an IsoPrime mass spectrometer (GV instruments, Crewe, UK).