INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Nitrogen loss in fertilized crop fields is a serious environmental problem because it contributes not only to decreased diversity of vegetation and greenhouse warming, but also to pollution of drinking water sources (Frink et al., 1999). In Asia, the staple crop is rice (Oryza sativa), 75% of which is harvested from irrigated lowlands. Unfortunately, due to the mixture and fluctuation of the oxidative and reductive environment in rice-planted submerged soil, rice cultivars usually recover less than 50% of the nitrogen fertilizer in irrigated fields (Vlek and Byrnes, 1986; Cassman et al., 1993). Denitrification and leaching of nitrate left in flooded soils are major contributing factors to nitrogen loss (Reddy, 1982; Buresh et al., 1989; Aulakh and Singh, 1996). Understanding the mechanisms of how rice takes up nitrate and assimilates it at the genetic level is a critical step toward alleviating the nitrogen loss problem.

Intense efforts, most notably by Hasegawa and coworkers (1992), have been directed to identify genetic mutants of rice defective in nitrate acquisition or assimilation. Mutants that are either resistant (Hasegawa et al., 1992; Sato et al., 1997) or hypersensitive (Hasegawa and Ichii, 1994; Hasegawa et al., 1995) to chlorate, a nitrate analog that is toxic when reduced, were isolated. In addition, mutants were found with reduced nitrate uptake activity (Hasegawa, 1996) or that exhibited nitrogen deficiency symptoms when grown with nitrate as the sole nitrogen source (Ichii et al., 1993). However, most of the rice mutants identified so far are either defective in nitrate reductase or in the proton-ATPase that provides the proton gradient for nitrate uptake (Hasegawa et al., 1992; Ichii et al., 1993; Hasegawa, 1996; Sato et al., 1997). One mutant, M605, might be the result of a defect in nitrate transport (Hasegawa and Ichii, 1994; Hasegawa et al., 1995), but it has not been further characterized. To date, no rice nitrate transporter genes have been isolated.

In comparison, recent years have seen a flurry of reports describing the molecular cloning of nitrate transporter genes in other higher plants (for review, see Von Wiren et al., 1997; Crawford and Glass, 1998; see Daniel-Vedele et al., 1998). These newly identified genes define two sequence-distinct nitrate transporter families called NRT1 and NRT2. Results from expression and functional studies obtained at this early stage assign NRT1 to low-affinity (millimolar nitrate) and NRT2 to high-affinity (micromolar nitrate) transporters. The one exception is CHL1 (AtNRT1), which has been shown to be a dual-affinity nitrate transporter (Wang and Crawford, 1998; Liu et al., 1999). Although higher plants comprise both low- and high-affinity nitrate transport systems (encoded by NRT1, NRT2, and other yet-to-be identified gene families), physiological studies of nitrate uptake in rice have focused primarily on the high-affinity uptake phase, using micromolar levels of nitrate (Sasakawa and Yamamoto, 1978; Youngdahl et al., 1982; Hasegawa and Ichii, 1994; Raman et al., 1995; Kronzucker et al., 1999). Few studies have investigated the kinetics of low-affinity nitrate uptake in rice.

We report the cloning and functional characterization of a nitrate transporter gene from rice, OsNRT1. The cloning of OsNRT1 was made possible by searching the rice expressed sequence tag database for homologs of CHL1 (Tsay et al., 1993). Functional characterization in Xenopus oocytes and in situ hybridization suggests that OsNRT1 encodes a constitutively expressed transport system for low-affinity nitrate uptake. OsNRT1 exhibits more functional properties in common with AtNRT1:2, a second Arabidopsis NRT1 gene reported recently (Huang et al., 1999), than with CHL1. The successful cloning of OsNRT1 may suggest a straightforward approach to cloning other rice nitrate transporter genes, intensifying molecular studies of nitrate uptake in rice. The properties of OsNRT1 further our understanding of the NRT1 family.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Plant Materials and Culture Conditions

Rice (Oryza sativa cv Nipponbare) was used in all of the experiments unless mentioned otherwise. Plants were grown hydroponically on iron mesh in modified Kimura B solution (Ehara et al., 1990) containing 180 µM (NH₄)₂SO₄, 109 µM KNO₃, 274 µM MgSO₄, 911 µM KH₂PO₄, 31 µM ferric citrate, 183 µM Ca(NO₃)₂, 2.5 µM H₃BO₃, 0.2 µM MnSO₄, 0.2 µM ZnSO₄, 0.05 µM CuSO₄, and 0.05 µM H₂MoO₄. The solution was refreshed every other day. Plants were grown under a 16-h light/8-h dark photoregime, 70% relative humidity, and a temperature of 28°C. Light was provided at approximately 250 µmol m² s¹.

Isolation and Sequencing of OsNT1 cDNA

The rice expressed sequence tag (EST) clone RICR2778 was identified by a homology search using the protein sequence of CHL1 (Tsay et al., 1993). The insert of RICR2778, obtained from the Rice Genome Center (Rice Genome Research Program, Japan), was subcloned into vector pBluescriptSK+ (Stratagene, La Jolla, CA) and sequenced by serial deletions of the cDNA in both directions (Sambrook et al., 1989) and by the dideoxy-chain termination method (Sanger et al., 1977) using the Sequenase 2.0 kit (United States Biochemical, Cleveland). The missing part of the cDNA at the 5' end was rescued by rapid amplification of 5' cDNA ends (5'-RACE) with two primers of RICR2778 (RC2, 5'-GAATTGTACAGTACTTCCCC-3', nt 473 to nt 492, and RC4, 5'-TTCTGAGAAGAGACTGGATCTGTCC-3', nt 589 to nt 613, in the reverse direction). The RC4 primer was used to synthesize the first strand of the cDNA using the 5'-RACE kits (Gibco-BRL, Gaithersburg, MD). This first strand was tailed with a stretch of 15 cytidines by terminal deoxytransferase. The new 5' sequence was then amplified by Pfu DNA polymerase (Stratagene) in a thermal cycler (Hybaid, Middlesex, UK) with the RC2 primer and an anchor primer provided in the kit. The obtained fragments were cloned into SmaI-restricted pBluescriptKSII and confirmed by sequencing. The full-length OsNRT1 clone, designated as pSK-OsNRT1F, was constructed by replacing the 5' end of pRICR2778 with the BamHI-NdeI fragment of the 5'-RACE product.

DNA and RNA Gel Blotting Analysis

Genomic DNA was isolated from rice tissues using urea extraction buffer (Sheu et al., 1996). Genomic DNA (10-15 µg) was digested with the restriction enzyme and fractionated in a 0.8% (w/v) agarose gel. For RNA gel blotting, total RNA was isolated from rice tissue with the TRIZOL reagent (Gibco-BRL, Grand Island, NY). Total RNA (15-25 µg) was fractionated in a 1% (w/v) agarose gel circulated with a 10 mM sodium phosphate buffer, pH 6.5. The DNA or RNA was then transferred overnight by capillary with 10× SSC (1.5 M NaCl and 0.17 M sodium citrate) to Hybond-N nylon membrane (Amersham International, Buckinghamshire, UK). The membrane was hybridized overnight with a ³²P-labeled, 1.7-kb, SalI-NotI-restricted OsNRT1 cDNA probe at 65°C in hybridization buffer containing 5× SSC, 0.1% (w/v) SDS, 5× Denhardt's solution, and 25 µg/mL fragmented salmon sperm DNA. The blots were also hybridized with ³²P-labeled exon 1 of the rice nitrate reductase 1 (nia1) gene (obtained by PCR amplification) (Choi et al., 1989) as a positive control for nitrate induction and a 1.4-kb EcoRI fragment of the rice actin 1 gene (McElroy et al., 1990) as an internal loading control. The blots were washed sequentially with 2× SSC and 0.1% (w/v) SDS for 30 min at room temperature; 2× SSC and 0.1% (w/v) SDS for 30 min at 65°C; 0.2× SSC and 0.1% (w/v) SDS for 30 min at 65°C.

Yeast Transformation and Growth Assays

The OsNRT1 cDNA fragment excised with BamHI at both ends from pSK-OsNRT1F was blunted with Klenow DNA polymerase and inserted into a BstXI-restricted, Klenow-filled pFL61 vector (Minet et al., 1992) to create pOsNRT1(S) (with OsNRT1 in the sense orientation) or pOsNRT1(AS) (with OsNRT1 in the antisense orientation) downstream of the phosphoglycerate kinase (PGK) promoter. pOsNRT1(S) and pOsNRT1(AS) were transformed into yeast strain PB1X-2A (MATa ura3-52 leu2-3 lys1-1 his4-32 ptr2::LEU2) grown in YEPG medium (1% [w/v] yeast extract, 2% [w/v] peptone, and 2% [w/v] Glc) (Perry et al., 1994) according to the method previously reported by Gietz and Schiestl (1995). Yeast transformants were selected on minimal medium composed of 2% (w/v) Glc, 1× yeast nitrogen base (YNB) without amino acids, and 2% (w/v) agar supplemented with 80 µM His and 80 µM Lys.

Growth assays were performed to assess the ability of yeast transformants containing pOsNRT1(S) and pOsNRT1(AS) to use the dipeptide (His-Leu) as a source of His in meeting auxotrophic requirements. The dipeptide medium consisted of minimal medium supplemented with auxotrophic requirements minus the amino acid components (His) of the added dipeptide (80 µM His-Leu). PB1X-2A and PB1X-2A (pYES2), a transformant containing the pYES2 vector, were used as negative controls, and PB1X-2A (pJP9) expressing the Saccharomyces cerevisiae peptide transporter PTR2p (Perry et al., 1994) was used as a positive control. Strains were grown in minimal medium (except for PB1X-2A which was grown in YEPG broth) at 30°C overnight and harvested by centrifugation. The cells were washed twice with sterilized distilled water, and were resuspended at a titer of 2 × 10⁸ cells mL¹. A 5-µL aliquot of each dilution, 2 × 10⁸, 2 × 10⁷, and 2 × 10⁶ cells mL¹, was applied to the dipeptide medium to achieve 10⁶, 10⁵, and 10⁴ cells, and incubated at 30°C. Growth of each strain was scored at 48 h.

Functional Analysis of OsNT1 Expressed in Xenopus Oocytes

The full-length OsNRT1 cDNA was subcloned into the oocyte expression vector pGEMHE containing the 5'-UTR and 3'-UTR of the Xenopus -globin gene (Liman et al., 1992) to enhance protein expression in oocytes. Capped mRNA was transcribed from the linearized plasmid in vitro using a kit (mMESSAGE mMACHINE, Ambion, Austin, TX). Oocytes were isolated and injected with 50 ng of cRNA as described previously (Tsay et al., 1993). Measurements were made in solutions of: (a) 230 mM mannitol, 0.3 mM CaCl₂, 5 mM 2-(N-morpholino)-ethanesulfonic acid (MES), and 10 mM Tris-MES at different ratios to yield the pH indicated, and (b) 220 mM mannitol, 0.3 mM CaCl₂, and 10 mM Tris-MES at the pH indicated plus HNO₃, Gly-Gly, His-Gly, Ala-His, Ala-Asp, or His at the concentration indicated. For the current-voltage (I-V) curve measurement, oocytes were clamped at 60 mV and assayed from 20 to 140 mV in 20-mV step for 300 ms each. For K_m measurement, oocytes were voltage-clamped to 60 mV or clamped at 60 mV and assayed from 30 to 150 mV in 30-mV step for 300 ms each and exposed to different concentrations of nitrate. Measurements were recorded with a 486-based microcomputer using the AXOTAPE and pCLAMP programs (Axon Instruments, Foster City, CA). The high- and low-affinity uptake assay of OsNRT1-injected oocytes determined by HPLC analysis was performed as described previously (Liu et al., 1999).

In Situ Hybridization

Plants were grown in modified Kimura B solution for 1 month, then shifted to a medium with 5 mM KNO₃ for 12 h before harvesting root tissues. The root tissues were fixed, dehydrated, embedded in Paraplast, and sectioned as described previously (Huang et al., 1999) with the following modifications: (a) in xylene:Paraplast mixture (3:1, 1:1, 1:3) for 1 h each instead of 30 min, (b) pure melted Paraplast at 62°C for 2 h three times.

Full-length OsNRT1 without the poly(A⁺)-tail was synthesized by PCR using primer RC7 (5'-CCGGATCCATGGACTCCTCATACC-3') and primer RC8 (5'-CCTCTAGAGCAACACAATTGTCC-3'), and cloned into SmaI-restricted pBluescript SII vector to give plasmid pOsNRT1(A), which was confirmed by sequencing. A ³³P-labeled OsNRT1 antisense RNA probe was synthesized using T7 RNA polymerase (Promega, Madison, WI) from EcoRI-linearized pOsNRT1(A), and a sense RNA probe was synthesized using T3 RNA polymerase (Promega) from XbaI-linearized pOsNRT1(A). Root sections were hybridized to hydrolyzed RNA probes and washed as described previously (Huang et al., 1999). Slides were exposed for 2 months.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Cloning and Sequence Analysis of OsNRT1

As described in the experimental procedures, a full-length OsNRT1 clone was constructed by rescuing, using 5'-RACE, the missing sequences of RICR2778, a rice EST clone that was initially identified by its homology with the sequence of the Arabidopsis nitrate transporter gene CHL1. The final clone obtained had an insert of 1,950 bp. Genomic Southern analysis indicated that OsNRT1 is a single-copy gene, because only one hybridized band was detected when the genomic DNA was digested with BglII, BamHI, EcoRI, or XhoI (Fig. 1, lanes 1-3 and 5), and two bands were found when the genomic DNA was restricted with HindIII (Fig. 1, lane 4), which has a cutting site in the cDNA of OsNRT1.

View larger version (51K): [in this window] [in a new window]   Figure 1.   Genomic DNA-blot analysis of OsNRT1 gene. Genomic DNA (10 µg), isolated from the rice cv Nipponbare, was digested with BglII, BamHI, EcoRI, HindIII, and XhoI, fractionated in a 0.8% (w/v) agarose gel, transferred, and hybridized with radiolabeled full-length (1.7 kb) OsNRT1 cDNA probes.

Sequencing the OsNRT1 cDNA clone revealed a 1,755-bp open reading frame for a 584-amino acid protein with a predicted molecular mass of 64 kD (Fig. 2A). Hydropathy analysis of the deduced amino acid sequence suggests that, similar to the predicted topologies of previously identified NRT1 proteins, OsNRT1 contains 12 putative transmembrane domains with a long hydrophilic loop separating the two groups of six transmembrane domains (Fig. 2B). OsNRT1 shares significant sequence identity (30%-50%) with members of a growing proton-dependent transporter family called PTR or POT (Paulsen and Skurray, 1994; Steiner et al., 1995) that comprise not only nitrate transporters from higher plants but also peptide transporters from bacteria, fungi, animals, and higher plants (Fig. 3). No significant homology with members of the NRT2 high-affinity nitrate transporter family (Daniel-Vedele et al., 1998) was detected.

View larger version (54K): [in this window] [in a new window]   Figure 2.   Sequence of OsNRT1 cDNA and hydropathy profile of OsNRT1. A, Nucleotide sequence of OsNRT1 cDNA and the deduced amino acid sequence of the OsNRT1 protein (accession no. AF140606). OsNRT1 cDNA (1,950 bp) contains an open reading frame starting at position 144. , Potential N-linked glycosylation sites. , Consensus phosphorylation sites. The putative transmembrane regions are underlined and numbered. B, Hydropathy analysis of the predicted sequence of OsNRT1. Hydropathy profile of the OsNRT1 protein was determined by the method of Kyte and Doolittle (1982) using a window of 19 amino acid residues.

View larger version (21K): [in this window] [in a new window]   Figure 3.   Phylogenetic tree of OsNRT1, CHL1, and its homologs in higher plants. The phylogram was created using the Growtree program of the GCG package (version 9.1, Genetic Computer Group, Madison, WI). CHL1 (AtNRT1) and AtNTR1:2 are nitrate transporters from Arabidopsis (Tsay et al., 1993; Huang et al., 1999), BnNRT1;2 is a nitrate/His transporter from B. napus (Muldin and Ingemarsson, 1995; Zhou et al., 1998), LeNRT1-1 and LeNRT1-2 are from tomato (Lauter et al., 1996), AtPTR2B is an Arabidopsis peptide transporter (Frommer et al., 1994; Rentsch et al., 1995; Song et al., 1996), and HvPTR1 is a barley peptide transporter (West et al., 1998).

Substrate Specificity of OsNRT1

To determine whether OsNRT1 encodes a nitrate transporter or a peptide transporter, in vitro-synthesized OsNRT1 complementary RNAs were microinjected into Xenopus oocytes. Two to 3 d after injection, these oocytes were incubated in a mannitol buffer of pH 7.4, voltage clamped at 60 mV, and then perfused with substrate solution at pH 5.5. The shift of pH from 7.4 to 5.5 elicited little current in both OsNRT1- and water-injected oocytes (bottom row of Table I), whereas the inward current change elicited by nitrate in OsNRT1 cRNA-injected oocytes was approximately 200 times larger than those observed in the water-injected controls (top row of Table I). In contrast, OsNRT1-injected oocytes and water-injected controls showed little or no inward current at 60 mV in the presence of the neutral dipeptide Gly-Gly, a positively charged dipeptide Ala-Lys, or a negatively charged dipeptide Ala-Asp.

The substrate specificity of OsNRT1 remained unchanged at lower membrane potentials. Inward currents elicited by nitrate in OsNRT1 cRNA-injected oocytes became larger at more negative membrane potentials, but little or no current was elicited by neutral or charged dipeptides at all of the membrane potentials tested (Fig. 4A). These data are indicative of OsNRT1 encoding an electrogenic nitrate transporter.

View larger version (17K): [in this window] [in a new window]   Figure 4.   The I-V difference relationship of OsNRT1 expressed in Xenopus oocytes. Currents presented here are the difference between measurements conducted in the presence and absence of substrate. Currents elicited at the end of 300-ms pulses were plotted as the function of the voltage. A, Voltage dependence of nitrate-elicited currents in OsNRT1 cRNA-injected oocytes. The OsNRT1-injected oocyte was exposed to 10 mM nitrate (), 10 mM dipeptides Gly-Gly (), 10 mM His-Gly (), 10 mM Ala-His (), and 10 mM Ala-Asp () at pH 5.5. The oocyte was voltage clamped from 60 mV to a voltage between 20 and 140 mV for 300 ms at 20-mV increments. Each data point represents the average current obtained from four independent oocytes isolated from three different frogs. B, His and nitrate uptake. OsNRT1-injected oocytes were exposed to 10 mM nitrate at pH 5.5 (), 10 mM His at pH 5.5 (), 10 mM nitrate plus 10 mM His at pH 5.5 (), and 10 mM His at pH 8.5 (). Each data point represents the average current obtained from three independent oocytes. C, pH dependence of the nitrate-elicited current in OsNRT1-injected oocytes. , Current elicited by nitrate at pH 5.5. , Current elicited by nitrate at pH 7.4. , Current elicited by nitrate in a water-injected oocyte at pH 5.5. , Current elicited by nitrate in a water-injected oocyte at pH 7.4. Shown here are average currents elicited in three OsNRT1-injected oocytes from three donor frogs.

To further confirm that OsNRT1 is not a peptide transporter, OsNRT1 was introduced into yeast mutant auxotrophic for His in both the sense (S) and antisense (AS) orientations. Figure 5 shows that PB1X-2A  (pOsNRT1 [S]) and PB1X-2A (pOsNRT1 [AS]) yeast transformants exhibited no growth on the dipeptide medium of His-Leu after 48 h of incubation, indicating that these two yeast transformants were unable to use His-Leu as a source of amino acid His. In contrast, PB1X-2A (pJP9), a positive control expressing the S. cerevisiae peptide transporter PTR2p (Perry et al., 1994), showed heavy growth in 48 h. The negative controls, PB1X-2A () and PB1X-2A (pYES2), on the same dipeptide medium exhibited no growth. A significant number of plant membrane proteins expressed in yeast function properly at the plasma membrane of yeast (for review, see Rentsch et al., 1998). OsNRT1 mRNA was expressed at a high level in the yeast transformant, as evidenced from RNA gel blotting analysis (data not shown). Therefore, it is likely that the observed no growth of PB1X-2A (pOsNRT1 [S]) on the dipeptide medium is due to the lack of dipeptide uptake of OsNRT1.

View larger version (147K): [in this window] [in a new window]   Figure 5.   Growth assays of OsNRT1-transformed dipeptide-auxotrophic mutant yeast. pOsNRT1(S) and pOsNRT1(AS) yeast transformants were grown for 48 h on dipeptide medium (His-Leu) to assess their ability to use dipeptide (His-Leu) as a source of amino acid. (), PB1X-2A negative control; pJP9 (PTR2p), pYES2 (Vector), pOsNRT1 (S), and pOsNRT1 (AS), PB1X-2A transformed with the plasmid of pJP9, pYES2, pOsNRT1 (S), and pOsNRT1 (AS), respectively.

In addition to nitrate and peptide, His uptake activity has been reported for members of the PTR family. For example, PHT1, isolated from rat brain, is a peptide and His transporter (Yamashita et al., 1997), and the Brassica napus nitrate transporter BnNRT1;2 exhibits significant His transport activity, especially at pH 8.5 (Zhou et al., 1998). However, at the membrane voltages tested here from 20 to 140 mV, oocytes injected with OsNRT1 cRNA did not respond to His at either pH 5.5 or 8.5 (Table I; Fig. 4B). Moreover, the currents elicited by nitrate at pH 5.5 in OsNRT1-injected oocytes were the same whether or not His was present in the external solution (Fig. 4B). Based on these analyses of OsNRT1 in Xenopus oocytes and yeast cells, we can conclude that OsNRT1 encodes a nitrate transporter with no (or little) peptide and His transport activity.

pH Dependence of Nitrate-Elicited Currents

When OsNRT1-injected oocytes were voltage-clamped at 60 mV and then exposed to 10 mM nitrate at different pH values for six independent OsNRT1-injected oocytes, the currents elicited by nitrate at pH 7.4 were only approximately 34% ± 17% of the currents elicited by nitrate at pH 5.5. The currents elicited at pH 8.4 were further reduced to 2% ± 14% of the currents elicited at pH 5.5. Similar results were obtained in the I-V measurements (Fig. 4C): 10 mM nitrate elicited larger currents at pH 5.5 than at pH 7.4 in OsNRT1-injected oocytes at each of the membrane potentials (from 20 to 140 mV) tested. The I-V curves of Figure 4C indicated that the currents elicited at pH 7.4 were approximately 30% to 40% of the currents elicited at pH 5.5. In contrast, similar treatments at pH 5.5 or 7.4 elicited no more than 5-nA currents in water-injected oocytes (Fig. 4C). The positive inward current of OsNRT1-injected oocytes in response to negatively charged nitrate and the pH dependence of the current elicited suggest that OsNRT1 functions as a proton-coupled nitrate cotransporter, with the ratio of proton to nitrate being larger than one.

Low-Affinity Nitrate Transport of OsNRT1

To determine the nitrate affinity of OsNRT1, OsNRT1-injected oocytes voltage-clamped at 60 mV were exposed to different concentrations of nitrate at pH 5.5. As shown in Figure 6A, the amplitudes of the inward positive currents elicited by nitrate at pH 5.5 were concentration dependent. For 14 oocytes tested, the K_m values obtained by fitting the current-concentration curves to the Michaelis-Menten equation were in the range of 7.2 to 12.6 mM, with an average of 9.1 ± 1.8 mM. At the membrane potentials tested, from 30 to 150 mV, the K_m values for nitrate measured at pH 5.5 were voltage independent, with an average of 8 to 9 mM (Fig. 6B).

View larger version (14K): [in this window] [in a new window]   Figure 6.   Kinetics of nitrate-evoked currents in OsNRT1-injected oocytes. A, Concentration dependence of nitrate-elicited currents in a single OsNRT1-injected oocyte. The oocyte was voltage-clamped at 60 mV, and the inward currents elicited by nitrate at pH 5.5 were plotted as a function of the external nitrate concentration. For this particular experiment, the Km value calculated by fitting to the Michaelis-Menten equation using the nonlinear least-squares method in the Origin 5.0 program (Microcal Software, Northampton, MA) was 7.3 ± 0.4 mM. The average Km calculated from 14 independent oocytes was 9.1 ± 1.8 mM. B, Voltage dependence of Km for nitrate determined at pH 5.5 in four OsNRT1-injected oocytes. Km values were determined by fitting I-V curves to the Michaelis-Menten equation at each voltage. Shown here are average values of Km calculated from four individual oocytes isolated from two different frogs.

We have recently shown that CHL1 is a dual-affinity nitrate transporter, but its Arabidopsis homolog AtNRT1:2 is a low-affinity nitrate transporter (Huang et al., 1999; Liu et al., 1999). Therefore, it was important to determine whether OsNRT1 could also function as a high-affinity nitrate transporter. The same batch of OsNRT1-injected oocytes was incubated with either 10 mM or 250 µM of nitrate for low- and high-affinity uptake measurement, respectively. The measurements were carried out with HPLC analysis, as described previously (Liu et al., 1999). The results showed that only low-affinity (Fig. 7A) not high-affinity (Fig. 7B) nitrate uptake activity was observed in OsNRT1-injected oocytes. In contrast, and confirming our previous findings (Liu et al., 1999), the same batch of oocytes injected with CHL1 exhibited significant activity in both low- and high-affinity phases of nitrate uptake. This indicates that like AtNRT1:2 of Arabidopsis, OsNRT1 is a low-affinity nitrate transporter with no apparent activities in high-affinity uptake.

View larger version (28K): [in this window] [in a new window]   Figure 7.   Nitrate uptake analysis of OsNRT1-injected and CHL1-injected oocytes. A, Low-affinity nitrate uptake activities of OsNRT1-injected and CHL1-injected oocytes determined with 10 mM NO3 at pH 5.5. Oocytes were incubated with 10 mM NO3 at pH 5.5 for 3 h, then assayed for the retained nitrate by HPLC. B, High-affinity nitrate uptake activities of OsNRT1-injected and CHL1-injected oocytes determined with 0.25 mM NO3 at pH 5.5. Oocytes were incubated with 0.25 mM NO3 at pH 5.5 for 3 h, then the amount of nitrate depleted from the medium was determined by HPLC. Each data point represents the average value obtained from the measurements of three batches, each consisting of five oocytes. Similar results were obtained with oocytes isolated from three different frogs.

Constitutive Expression of OsNRT1 in Rice Root

As shown in Figure 8, RNA gel blotting analysis indicated that OsNRT1 is a root-specific gene with little or no expression in the shoot. At time 0, when the plants were depleted for nitrate for 2 d, there was a low level of the mRNA of the leaf-specific nitrate reductase nia1. However, 4 h after the addition of nitrate, the nia1 mRNA level increased approximately 10-fold. After 8 h, the nia1 mRNA level experienced a negative feedback regulation and began to decrease. In contrast, the OsNRT1 mRNA level remained relatively stable before and after nitrate induction. This indicates that like Arabidopsis AtNRT1:2 (Huang et al., 1999) and tomato LeNRT1-1 (Lauter et al., 1996), OsNRT1 is a constitutive gene.

View larger version (43K): [in this window] [in a new window]   Figure 8.   Northern analysis of OsNRT1 expression. cv Nipponbare rice plants were grown in the Kimura B solution for 30 d, shifted to N-depleted Kimura B solution for 2 d, and then transferred to Kimura B solution containing 5 mM NaNO3, pH 5.5, for the times indicated. Total RNA (15 µg) was fractionated in a 1% (w/v) agarose gel, transferred to nylon membrane, and hybridized with radiolabeled DNA from OsNRT1, nia1, and Act1.

Accumulation of OsNRT1 mRNA in Root Epidermis and Root Hair

To determine the root tissue in which OsNRT1 is expressed, roots of 30-d-old rice were fixed, sectioned, and hybridized with ³³P-labeled sense and antisense transcripts of the OsNRT1 DNA. From the exterior to the core, rice roots contain one layer of epidermal cells, followed by one layer of exodermal cells, one layer of sclerolenchymal cells, several layers of cortical cells, and then the vascular cylinder (Fig. 9; Clark and Harris, 1981). As shown in Figure 9, B and D, a high density of silver grains (which appear yellow due to the colored filter), indicating the accumulation of OsNRT1 mRNA, was found predominantly in the epidermal cells and root hairs. As a control, similar sections hybridized with the sense probe showed only a background level of signal in all layers of the root (Fig. 9F).

View larger version (134K): [in this window] [in a new window]   Figure 9.   In situ hybridization of OsNRT1 to rice root sections. A to D, In situ hybridization of antisense OsNRT1 probe to a cross-section of rice root tissues. E and F, In situ hybridization of sense OsNRT1 probes to a cross-section of rice root tissues. A, C, and E, Bright-field microscopy. B, D, and F, Double exposures using a colored filter for the dark-field exposure, causing the OsNRT1 signals to appear yellow. ep, Epidermis; ex, exodermis; sc, sclerenchyma; co, cortex; as, air space (aerenchyma); hr, root hair; x, late metaxylem. Bars = 100 µm.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

In this study, a new member (and the first from rice) of the NRT1 nitrate transporter family was cloned. NRT1 is a subset of the PTR family that also includes proton-coupled oligopeptide transporters (Paulsen and Skurray, 1994; Steiner et al., 1995). Because chemically distinct compoundsnitrate, peptide, and Hishave been identified as substrates for different members of the PTR family, it was critical that we clarify the substrate specificity of the new member, OsNRT1.

As shown above, significant nitrate-induced currents indicative of uptake were observed in OsNRT1-injected Xenopus oocytes, whereas no apparent uptake activity was observed for various dipeptides or His (Fig. 4, A and B). Augmenting this oocyte uptake result, OsNRT1 could not complement a yeast dipeptide auxotrophic mutant, while the S. cerevisiae PTR2p and Arabidopsis AtPTR2B peptide transporters could (Perry et al., 1994; Steiner et al., 1994; Fig. 5). Furthermore, the dependence on membrane potential and pH of nitrate uptake activity (Fig. 4), in conjunction with expression of OsNRT1 in the rhizodermis (Fig. 9), are consistent with the notion that OsNRT1 is a proton-dependent electrogenic nitrate transporter directly involved in uptake. The functional assays of OsNRT1, together with those of other NRT1 genesCHL1 (Tsay et al., 1993; Huang et al., 1996) and AtNRT1:2 (Huang et al., 1999) from Arabidopsis and BnNRT1;2 from B. napus (Zhou et al., 1998)provide converging evidence that the NRT1 proteins function primarily as nitrate transporters. Interestingly, this is despite OsNRT1 being more similar in sequence to two peptide transporters than to the nitrate transporters (Fig. 3). Consequently, the substrate specificity of members of the PTR family cannot be inferred from sequence comparison, leaving future research to delineate the sequence and structure components underlying the distinct substrate specificities of this transporter family.

We conducted nitrate induction experiments with different rice cultivars, including two local Japonica cultivars (cv Taikeng 9 and cv Tainung 67) and two Indica cultivars (cv Taichung Sen 10 and cv IR-36). In all of the rice cultivars tested, OsNRT1 was expressed constitutively. In addition, in two wild rice species (Oryza perennis and Oryza officinalis), one hybridizing band was found in the RNA get blotting analysis using OsNRT1 as a probe, and the intensity of the band remained approximately constant before and after nitrate induction (data not shown). Several studies have shown that the nitrate uptake activities of rice are repressed when ammonium is also present in the nutrient medium (Youngdahl et al., 1982; Kronzucker et al., 1999). However, our data showed no repressed expression of OsNRT1 mRNA when ammonium was present in the assay solution (data not shown). Therefore, it might be that an ammonium inhibition effect on uptake by OsNRT1 does not occur at the transcriptional level. Alternatively, the observed repression of nitrate uptake by ammonium may be the result of inhibition of another nitrate transporter gene(s).

The oocyte uptake assays revealed significant activities of OsNRT1 in the low-affinity but not in the high-affinity range of nitrate concentrations. Kinetic measurements yielded a Michaelis-Menten K_m for nitrate of approximately 9 mM independent of the membrane potentials tested (Figs. 6 and 7; Table I). These data suggest that OsNRT1 encodes a constitutive component for low-affinity nitrate uptake. Recent studies (Tsay et al., 1993; Huang et al., 1996, 1999; Lauter et al., 1996) seem to indicate that the NRT1 family comprises both the constitutive and the nitrate-inducible component of the low-affinity nitrate transport system of higher plants. In addition, a striking functional difference between the two components is obvious in the case of Arabidopsis: Whereas CHL1, the nitrate-inducible component, is involved in both low- and high-affinity nitrate uptake (Wang and Crawford, 1998; Liu et al., 1999), AtNRT1:2 (previously referenced as NTL1 or NRT3), the constitutive component, exhibits only low-affinity uptake activities (Huang et al., 1999; Liu et al., 1999), as does OsNRT1 (Fig. 7). The resemblance between AtNRT1:2 and OsNRT1 extends to their tissue-specific expression patterns. Both are primarily expressed in root hair and epidermis, and this expression location is independent of root age (Fig. 9; Huang et al., 1999). In contrast, as the root matures, the great majority of CHL1 is expressed in inner layers, first in the cortex and then in the endodermis (Huang et al., 1996).

There are now several clones in the rice EST database that share higher sequence identity with CHL1 than does RICR2778, the clone from which OsNRT1 was derived. Some of these genes might encode proteins with the same properties of CHL1. On the other hand, the root structure of rice is distinctly different from that of Arabidopsis. Rice roots contain exodermis just external to one layer of sclerolenchymal cells. In addition, as rice root matures, its cortical cells develop into aerenchyma (Fig. 9A). These three structuresexodermis, sclerolenchyma, and aeren-chymaare all absent in the Arabidopsis root and they will restrict the apoplast diffusion of ions (Colmer and Bloom, 1998). The significant difference in root structure may suggest that orthologs of the nitrate transporter genes in different species will exhibit a significant variation in their physiological functions. Therefore, it will be particularly interesting to find out if a CHL1-like rice gene (if it exists) is similarly expressed in the inner cells (endodermis) of mature roots. In addition to OsNRT1, which is a member of the NRT1 family, a rice cDNA clone belonging to the NRT2 family has been reported (GenBank accession no. AB008519). Elucidation of the distinct roles of these nitrate transporter genes should yield considerable knowledge for genetic manipulations to lessen the demand on nitrogen fertilization.