A high affinity fungal nitrate carrier with two transport mechanisms.

We have expressed the CRNA high affinity nitrate transporter from Emericella (Aspergillus) nidulans in Xenopus oocytes and used electrophysiology to study its properties. This method was used because there are no convenient radiolabeled substrates for the transporter. Oocytes injected with crnA mRNA showed nitrate-, nitrite-, and chlorite-dependent currents. Although the gene was originally identified by chlorate selection there was no evidence for transport of this anion. The gene selection is explained by the high affinity of the transporter for chlorite, and the fact that this ion contaminates solutions of chlorate. The pH-dependence of the anion-elicited currents was consistent with H(+)-coupled mechanism of transport. At any given voltage, currents showed hyperbolic kinetics with respect to extracellular H(+), and these data could be fitted with a Michaelis-Menten relationship. But this equation did not adequately describe transport of the anion substrates. At higher concentrations of the anion substrates and more negative membrane voltages, the currents were decreased, but this effect was independent of changes in external pH. These more complicated kinetics could be fit by an equation containing two Michaelis-Menten terms. The substrate inhibition of the currents could be explained by a transport reaction cycle that included two routes for the transfer of nitrate across the membrane, one on the empty carrier and the other proton coupled. The model predicts that the substrate inhibition of transporter current depends on the cytosolic nitrate concentration. This is the first time a high affinity nitrate transport activity has been characterized in a heterologous system and the measurements show how the properties of the CRNA transporter are modified by changes in the membrane potential, external pH, and nitrate concentration. The physiological significance of these observations is discussed.

Nitrate is an important nitrogen source for many organisms ranging from bacteria and cyanobacteria to fungi and plants. Nitrate carriers for transport across the plasma membrane must be present in many different cell types. In fungi, nitrate transport has been demonstrated in Penicillium chrysogenum (1) and Neurospora crassa (2), and the presence of H ϩ /nitrate symport activity was shown in the plasma membrane of cells of Candida utilis (3) and Emericella (Aspergillus) nidulans (4). In N. crassa, the nitrate transport system was induced by the external supply of NO 3 Ϫ and the K m was 0.25 mM, whereas the nitrite system had a K m of 86 M (5). Electrophysiological characterization of nitrate transport in intact hyphae of N. crassa showed that high affinity transport was sensitive to both membrane voltage and external pH (6). Nitrate uptake had complex kinetics in E. nidulans but the K m was estimated as 200 M (7).
The crnA mutants of E. nidulans were first isolated as a class of chlorate-resistant strains that were able to utilize nitrate as the sole nitrogen source (8). The crnA mutants were subsequently found to be defective in nitrate uptake at the conidiospore and young mycelial stages (7). The crnA gene has been cloned (9) and belongs to the Major Facilitator superfamily of membrane transporters, being in the same family (nitrate-nitrite porter, NNP) 1 as NarK, the nitrite efflux system from Escherichia coli and the NRT2 high affinity nitrate transporters from algae and higher plants (10). A related nitrate transporter from the yeast Hansenula polymorpha has also been isolated (11). Genes encoding high affinity nitrate transporters from the alga, Chlamydomonas reinhardtii have been isolated and complementation of mutant strains has shown that two genes (Nar2 and either NRT2;1 or NRT2;2) are necessary for a functional nitrate transport system (12). In Chlamydomonas, nitrate and nitrite are transported by different specific transport systems and also by a bispecific transporter (13). The injection of a mixture of Nar2 and Nrt2;1 mRNA into Xenopus oocytes gave nitrate transporter activity, but Nar2 mRNA was also toxic to these cells (14). Recently, evidence for a mammalian H ϩ /NO 3 Ϫ cotransport mechanism has also been described (15).
To date there are no reports of the detailed characterization of high affinity nitrate transporters in a heterologous expression system, although members of the NRT1 family of low affinity transporters have been studied in Xenopus oocytes (16,17). As in Chlamydomonas, a high affinity nitrate transport system requires two gene products; we tested if crnA, the first nitrate transporter gene to be isolated, encodes a fully functional uptake system. This is important because many homologs of the NNP family have recently been cloned by sequence homology to crnA (18), but the function of these cannot easily be demonstrated. The plant family members are of fundamental importance to plant biology and agriculture because nitrogen supply is the main factor limiting growth and yield of crops. Functional characterization of the NNPs is hindered by the lack of convenient tracers for the transported substrates. 13 N-labeled nitrate and nitrite are not generally available, and 36 Cl-labeled chlorate is not commercially produced or reliably produced. However, electrophysiological techniques can be used to assay for nitrate transport. Many plant and mammalian transporters have been successfully expressed and subsequently characterized in Xenopus oocytes, but there is only one report of a fungal membrane protein being expressed, the yeast ␣-factor receptor (19). Here we have been able to demonstrate and characterize the high affinity nitrate and nitrite proton cotransport activity of the CRNA protein expressed in oocytes.

EXPERIMENTAL PROCEDURES
Construction of a Full-length crnA cDNA-Unless otherwise stated, all standard molecular biology protocols were performed (20). A fulllength crnA cDNA clone was generated using a unique internal SphI site located near the middle of the coding region to join a 5Ј-fragment generated by reverse transcriptase-PCR with a 3Ј-fragment obtained from the partial crnA cDNA clone pSTA1500 (9). The cDNA template for reverse transcriptase-PCR was synthesized by reverse transcription of poly(A) ϩ RNA from nitrate-grown E. nidulans using avian myeloblastosis virus reverse transcriptase. The PCR reaction was performed using PfuI polymerase (Stratagene) according to the supplier's instructions. The primers (5Ј-GTCGAGTTTGGATCCAACTTC-3Ј and 5Ј-ACG-GCCATGGAATTCACACCT-3Ј) were designed to amplify the cDNA sequence corresponding to nucleotides 1237-2929 of the genomic sequence (GenBank™/EBI accession no. M61125), and each contained one nucleotide substitution introduced to create a restriction site (BamHI and EcoRI, respectively) for cloning purposes. The PCR product was digested with BamHI and EcoRI and cloned into pYES2 (Invitrogen) to give pCRNA1. A clone containing the crnA sequence downstream of the SphI site was generated by subcloning the 1-kb SphI-XbaI fragment from pSTA1500 into pUC19 to generate pCRNA2. The HindIII-SphI fragment from pCRNA1, containing the crnA sequence upstream of the SphI site, was then transferred into the corresponding sites in pCRNA2 to give pCRNA5. The sequence of the resultant full-length crnA cDNA in pCRNA5 was determined using custom primers and a T7 sequencing kit (Amersham Pharmacia Biotech), and comparison with the published genomic crnA sequence (9) showed it to be correct. The sequence has been entered into the GenBank™/EBI database under the accession no. U34382. To obtain a version of the crnA cDNA clone suitable for in vitro transcription and oocyte expression, the insert from pCRNA5 was excised using HindIII and XbaI and cloned into the HindIII and SpeI sites of pXE3 to give pCRNA-R1. The transcription vector pXE3 contains a 75-bp poly(A) tail cloned into the XbaI and NotI sites of pBluescript SK (Stratagene) and lacks a region of the polylinker from KpnI to HincII that might interfere with translation. 2 After linearizing pCRNA-R1 by digestion at a unique NotI site immediately 3Ј to the poly(A) tail, full-length crnA mRNA was synthesized using a T7 RNA transcription kit (Ambion). Purification of the mRNA and confirmation of its size and integrity by gel electrophoresis were performed according to protocols provided by the supplier of the in vitro transcription kit.
Oocyte Preparation, Injection, and Electrophysiology-Oocytes were prepared and treated as described previously (17). Healthy oocytes at stage V or VI (21) were chosen for injection with 50 nl of crnA mRNA (1 ng/nl), or 50 nl of diethyl pyrocarbonate (DEPC)-treated water as control. Injections were performed as described previously (17), and experiments were performed 4 -7 days after injection.
The anion-elicited currents were assayed in oocytes using the twoelectrode voltage-clamp method and using the pCLAMP software 6.0 (Axon Instruments, Foster City, CA) as described previously (17). All electrophysiological measurements were made in a saline containing (in mM) 116 NaCl, 2 KCl, 2 CaCl 2 , 1 MgCl 2 , 5 HEPES, pH 7.2. For experiments to vary external pH (pH o ), between pH 7 and 8, HEPES was used but for the more acidic saline MES was added in place of HEPES. The pH was adjusted by the addition of 5 M NaOH solution. Experiments were performed in a 0.5-ml plexiglass chamber, which was perfused continuously with saline at a rate of 2 ml min Ϫ1 throughout the recording. Only oocytes that had resting potentials more negative than Ϫ30 mV in saline at pH o 7.2 were used for voltage-clamp experiments.
For steady-state current measurements, the oocyte membrane potential was clamped at Ϫ50 mV. From this value the membrane was pulsed to a range of different test potentials for 120 ms from ϩ10 to Ϫ180 mV with either 40 or 20 mV incremental steps, followed by a 1-s interpulse interval at the holding potential (V H ). The membrane currents reached steady-state within 50 ms of clamping, and the mean values were calculated from the last 20 ms at each clamping voltage. The anion-elicited currents were used to obtain current-voltage difference curves (I-V) as described previously (17). Oocytes were treated with these anions for less than 1 min to minimize the accumulation of nitrate within the cell. As the transporter may have a very high affinity for nitrate, we measured the concentration of nitrate present in nitrite solutions using a standard assay method (22). These measurements showed that nitrate was present at less than 1%, even in nitrite solutions that had been vigorously aerated for 1 h. In all experiments, the oocytes were allowed to adjust for at least 5 min after changing the external pH before any treatments were applied.
Determination of Kinetic Parameters for the Anion Substrates-At any given membrane potential, steady-state currents measured as a function of external substrate concentration were fit to single Michaelis-Menten functions (23) by a non-linear least squares method using Sigmaplot software (Jandel Scientific, Erkrath, Germany). This function did not fit some of the data, and a more complicated model that could describe cis-or substrate-inhibition was required. In this model, the uptake of substrate (S) can be described as the sum of two apparently independent Michaelis-Menten processes, which can be written as Equation 1, where the superscripts I and II define the two phases, each of which is characterized by a saturating velocity i max and a K m . This type of analysis has been used to describe enzyme reactions that involve two substrates.

RESULTS
Substrate Specificity of CRNA-To obtain the crnA mRNA necessary for oocyte expression studies, we needed a full-length crnA cDNA sequence in a suitable transcription vector. However, the only crnA cDNA clone available (pSTA1500, Ref. 9) was incomplete at the 5Ј-end, whereas the cloned crnA gene (in pSTA4) was also unsuitable for our purposes because it contained three introns (9). We therefore used RT-PCR to generate the missing 5Ј part of the cDNA sequence and combined this with a segment of the pSTA1500 insert to create a full-length crnA cDNA clone, pCRNA5, which was then fully sequenced to establish its authenticity (see "Experimental Procedures"). Because of an error in assigning the precise location of one of the introns, crnA was originally reported to encode a polypeptide of 483 amino acids with 10 transmembrane domains (9). The crnA sequence in pCRNA5 was confirmed as encoding a polypeptide of 507 amino acids and 12 predicted transmembrane domains (see GenBank™/EBI accession no. U34382).
Transporter activity in oocytes injected with the crnA mRNA was examined electrically using the two-electrode voltage clamp technique. Oocytes injected with mRNA for the truncated form of CRNA did not show any nitrate-elicited currents (data not shown). Fig. 1A shows the I-V difference curves obtained for membrane voltages (V m ) between 0 and Ϫ160 mV for oocytes injected with full-length crnA mRNA. Subtracting the I-V relationship obtained in the absence of substrate from that obtained in its presence generated these curves. Both nitrate and nitrite (applied as the sodium salt at a concentration of 0.1 mM and at pH o 7) elicited currents in oocytes, which had been injected with crnA mRNA. These currents increased, as the membrane potential became more negative and showed saturation kinetics as the concentration increased. When nitrite and nitrate were supplied together at saturating concentrations, the currents did not increase, suggesting that the two anions bind to and are transported at the same site on the protein (data not shown).
In initial experiments it was found that chlorate also elicited currents in the crnA-injected oocytes. However, when the chlorate solution was prepared a few minutes before the experiment, no current could be measured at any membrane voltage (Fig. 1A). To test the possibility that some of the chlorate was being converted to chlorite during storage of the solution and that it was actually chlorite that was being transported, we treated an oocyte expressing CRNA with a freshly prepared solution of sodium chlorite. This treatment elicited current similar to that obtained with NO 3 Ϫ , confirming that ClO 2 Ϫ was being transported (data not shown).
No significant currents were obtained when crnA-injected oocytes were treated with similar concentrations of several other possible substrates, including bicarbonate, sulfate, cyanate, histidine, and the dipeptide His-Leu (data not shown). The ability of CRNA to transport Cl Ϫ was also tested, by replacing the Cl Ϫ in the bathing solution with gluconate whereas leaving the cation concentration unchanged. Nitrate, but not Cl Ϫ , could evoke currents in crnA-injected oocytes in this modified saline, showing that CRNA was unable to transport chloride. Water-injected control oocytes showed no significant current when treated with any of these ions at the same concentration. Replacing the bathing saline with a solution in which the sodium was replaced with choline tested the sodiumdependence of the NO 3 Ϫ -elicited current. The magnitude of the nitrate-elicited currents in a crnA-injected oocyte in each saline were then measured and compared. The experiment showed that the currents were not significantly different and so indicate no role for Na ϩ in NO 3 Ϫ transport by CRNA. Fig. 1B shows an experiment where pH o was 7 and the NO 3 Ϫ concentration was varied from 1 to 200 M to generate a family of I-V curves. The steady-state currents were measured as a function of voltage for each NO 3 Ϫ concentration. These nitrateelicited currents became larger at more negative membrane potentials and appeared to saturate at an external NO 3 Ϫ concentration between 40 and 60 M. Fig. 2A shows the relationship between the nitrate transport activity of CRNA and the external NO 3 Ϫ concentration from 1 to 150 M at pH o 7 for seven different membrane voltages. These results confirmed that CRNA was a high affinity NO 3 Ϫ transporter, with the transport activity saturating around 40 M, but the data only fit a Michaelis-Menten function at the lower membrane voltages. A similar pattern was observed for the other anion substrates, and an example for ClO 2 Ϫ but with no fitted line is shown in Fig. 2B.
Voltage-dependence of i max and K m for Protons-Steady-state NO 3 Ϫ -elicited currents were measured as a function of voltage and pH o . At a fixed external NO 3 Ϫ concentration (200 M) and at any given voltage the data from the I-V difference curve fit a single Michaelis-Menten function (Fig. 3A). Fig. 3B shows the voltage-independence of the K m H values calculated from these fits. The values for K m H were also voltage-dependent, changing from 0.14 M (pH 6.85) at 0 mV to 0.004 M (pH 8.4) at Ϫ180 mV (Fig. 3B). The i max H values also obtained from these lines were voltage-dependent, increasing from Ϫ26 nA at 0 mV to Ϫ63 nA at Ϫ180 mV (data not shown).
Inhibition by Anion Substrates-When an oocyte expressing CRNA was treated with nitrate concentrations from 1 M to 20 mM and the concentration was plotted on a log scale, the substrate inhibition becomes very clear (Fig. 4). At more negative voltages and at NO 3 Ϫ concentrations larger than 80 M, the current mediated by the transporter decreased, showing an inhibition of transport activity by NO 3 Ϫ . For example, at Ϫ150 mV the current in 20 mM NO 3 Ϫ is decreased by 50% when compared with that obtained at 100 M NO 3 Ϫ . Whereas the equivalent percentage inhibition at Ϫ70 mV is 20%, and at Ϫ30 mV, there was no inhibition by higher concentrations of nitrate. The data shown in Fig. 4 could be described by a more complicated model that included two additive Michaelis-Menten functions (24). From fits of this model to these data points the K m and i max values for each function (I and II) could be determined at each membrane voltage (see "Experimental Procedures"). The mean value of K m I NO3 was 20.5 M and was independent of changes in membrane voltage in the range from Ϫ160 to Ϫ70 mV, whereas at less negative membrane voltages, K m I NO3 increased to 88 M at Ϫ30 mV. The voltage-dependence of the K m I and K m II values was also determined for two other substrates, NO 2 Ϫ and ClO 2 Ϫ . The kinetic parameters were calculated from fits to families of different NO 2 Ϫ and ClO 2 Ϫ concentrations at pH o 7. For both substrates, the parameters i max NO2 and i max ClO2 were voltage-dependent, increasing as the membrane potential became more negative, whereas in contrast, both K m ClO2 and K m NO2 were voltage-independent (data not shown). When compared with nitrate, the values of K m I were 3and 4-fold higher for NO 2 Ϫ and ClO 2 Ϫ respectively. The effect of changing pH o on the nitrate-elicited currents was measured in oocytes injected with crnA mRNA. The nitrate-elicited currents increased as the pH o was decreased from 8 to 6, but at pH o 5.5 this effect saturated showing no further increase in current (Fig. 3A). At pH o 5.5, when the mRNA-injected oocyte was treated with a range of nitrate concentrations, the substrate-inhibition of current was again demonstrated (Fig. 5A). To compare the results obtained at different pH o , the nitrate-elicited currents were normalized to the maximal current obtained and this is shown as the percentage substrate inhibition (Fig. 5B). Fig. 5B shows a comparison of the percentage nitrate inhibition at pH o 5.5 and 7. The substrate inhibition of current was not effected by a change in pH o from 7 to 5.5, and similar profiles were obtained at intermediate pH o values (data not shown).
Kinetic Model for the Anion-Substrate Inhibition-Three fundamental observations define the anion-inhibition of CRNA activity. 1) It increases with higher concentrations of anion substrate. 2) It is independent of changes in external pH (over the range 8 to 5.5), and 3) it is voltage dependent, increasing at more negative voltages. This information can be used to build a reaction cycle model of CRNA-mediated transport. A decrease in the inward current can occur by three possible mechanisms, a change in the amount of available carrier in the membrane, a decrease in the influx of protons or an increase in anion influx. There are other options that involve changes in efflux (decreased for protons and increased for anions), but these can be discounted because they are thermodynamically unfeasible. The practicability of each of these three mechanisms will be considered.
As the substrate-inhibition of CRNA does not depend on pH o , but increases with the inwardly directed chemical gradient for anions, the most likely mechanism to achieve this result is an increased anion influx. The only mechanism that can account for both of these observations is that nitrate must bind before protons, and some transfer of the anion across the membrane can occur in this form (see Fig. 6). There are usually two recognized steps at which the transmembrane charge transfer can occur. These are the translocation of either the loaded carrier (positive charge) or the unloaded carrier (negative charge). A model involving the binding and translocation of nitrate via a negatively charged empty carrier is thermody- namically unlikely. Particularly, as the substrate-inhibition is voltage-dependent increasing at more negative membrane voltages. The charged form of the carrier should be the loaded form with all three ions, two protons and one nitrate ion, bound for translocation. At high external concentrations of nitrate, the anion binds first and before the proton binding. Some translo-cation across the membrane occurs driven by the chemical gradient of nitrate (Fig. 6, dashed line). In other words, nitrate binding to CRNA happens before the proton binding and nitrate slippage through the carrier only occurs at high external concentrations. This step will be independent of the proton gradient because the membrane translocation of nitrate does not require a proton. Although the negative membrane voltage will tend to oppose nitrate translocation into the oocyte, there is a large chemical gradient driving nitrate transport. The nitrate slippage increases at more negative membrane voltages because the proton-coupled charge translocation step increases at more negative voltage providing more unloaded carrier for this process to occur. In support of this aspect of the model the substrate inhibition, relative to the maximal current, increases at more negative voltages; for example, 50% nitrate inhibition at Ϫ150 mV and 20% at Ϫ70 mV (see Fig. 3).

DISCUSSION
The NO 3 Ϫ transporter CRNA of E. nidulans has been heterologously expressed in Xenopus oocytes enabling the first electrophysiological characterization of a fungal carrier protein using this expression system. Although there have been in vivo electrophysiological studies on high affinity nitrate transport activity in Neurospora (6), these measurements could not exclude the possibility that the nitrate transport processes being analyzed were mediated by more than one protein. Oocytes injected with mRNA for a truncated form of the protein did not show any nitrate-elicited currents. These results suggest that this deletion altered the transport activity or the membrane targeting of the protein.
In the green algae Chlamydomonas, genetic analysis has established that the NRT2;1 and NRT2;2 genes, which are homologous to crnA and which encode the membrane proteins responsible for high affinity nitrate and nitrite transport in this organism, require the activity of an additional gene (Nar2) to specify functional transport systems (12,13). Preliminary experiments confirmed that the Chlamydomonas NRT2;1 transporter is not functional when expressed in oocytes, and nitrate transport activity was only measured after co-injection of Nar2 mRNA (14). Possible roles for Nar2 include acting as a second subunit for the NRT2 transporter or an involvement in trafficking of the NRT2 protein from the endoplasmic reticulum to the plasma membrane. In the oocyte experiments reported here, functional high affinity NO 3 Ϫ transport was achieved with a single protein. The ability to obtain functional expression of crnA in oocytes without the assistance of a second gene suggests that there may be important differences between the algal and the fungal nitrate transporters, despite the sequence homology between the crnA and NRT2 genes. One striking distinction between the predicted structures of the fungal and the algal nitrate transporters of this family is in the arrangement of the hydrophilic domains within the primary sequence of the proteins. The fungal proteins lack a large C-terminal domain that is found in the algal (and higher plant) NRT2 proteins and instead have a large central loop between transmembrane domains 6 and 7 (10,18).
The CRNA transporter expressed in oocytes can also transport nitrite and chlorite, but with a lower affinity than nitrate. The K m for NO 3 Ϫ of the carrier expressed in oocytes was less than the value obtained for E. nidulans cells in vivo, for example 200 M (7). As CRNA has a high affinity for NO 3 Ϫ , the nitrite-elicited current could have resulted from contamination of the NO2 Ϫ solution with nitrate. To check for this possibility we measured how much nitrate could be present in a solution of nitrite and attempted to favor this oxidation reaction by vigorously aerating a nitrite solution. Even after vigorous aeration for 1 h, less than 1% of the nitrite had been converted to ؊ cotransport by CRNA. The main feature of this cycle is that the two processes that can transfer charge across the membrane are the translocation of the loaded CRNA and the proposed "slippage" of the partially loaded carrier (represented by a dashed arrow). Note that both these processes will result in the actual flux of nitrate from outside into the cell, but they give currents of opposite sign. nitrate. To explain the apparent differences in K m NO2 and K m NO3 would require 15% nitrate contamination in the nitrite solutions, a figure that is larger than we could measure. This result shows that, like one of the nitrate transport systems in Chlamydomonas (13), CRNA can also transport nitrite. Although a specific transporter for nitrite has not been demonstrated in E. nidulans, the identification of mutations resulting in hypersensitivity to nitrite implies that it may exist (8,25,26). The original crnA mutation was selected by growth on chlorate-containing medium at mM concentrations (7), but results reported here show that the selection was actually for ClO 2 Ϫ resistance (Fig. 2B). The presence of M contaminating concentrations of chlorite in the original chlorate solutions leading to isolation of crnA.
The activity of CRNA in oocytes was strongly dependent on the membrane potential, both in providing a direct energy source for transport (with current increasing, as membrane voltage became more negative) but also in changing some of the kinetic properties of the carrier. It was found that K m H but not K m NO3 was voltage-dependent. One of the main features previously described for in vivo high affinity nitrate transport, both in Neurospora and in plants (6,27), was the voltage-dependence of the transport system. To demonstrate this voltagedependence in vivo, it was necessary to clamp the membrane voltage at values more negative than Ϫ200 mV. However this was not possible in the oocyte experiments because at values more negative than Ϫ180 mV there was activation of an endogenous chloride channel in the oocyte plasma membrane (28), which then dominated the I-V curve making it difficult to determine the carrier current. Nonetheless, it has been possible to demonstrate the voltage sensitivity of plant carriers expressed in Xenopus oocytes (16,29,30). In this work and for AtSUC1 the Arabidopsis sucrose transporter (30) there was a similar 1.8-fold increase in the current as membrane voltage was increased from Ϫ50 to Ϫ150 mV. In contrast, there was almost no voltage-sensitivity of the nitrate cotransport in this range of membrane voltage for Neurospora; only at voltages more negative than Ϫ150 mV did the voltage-sensitivity become obvious (6). In Arabidopsis root hair cells over the same range of membrane voltages (-50 to Ϫ150 mV) the nitrateelicited current increased by 1.8-fold; only at more negative voltages did the effect of membrane voltage become even larger (27). These measurements show that the membrane potential has an important role in regulating CRNA activity, and so any environmental changes that influence this parameter will have effects on nitrate transport.
This model for the reaction cycle of CRNA (Fig. 6) must be very sensitive to the concentration gradient of anion substrate (nitrate) across the plasma membrane. In the oocyte experiments, the cells were previously incubated in a solution containing no nitrate, and the exposure times to nitrate were deliberately minimized. Increasing cytoplasmic concentrations of nitrate can test the model, because the predicted response should be a decrease in the substrate inhibition. There is indirect evidence for this model because the point at which the onset of the substrate inhibition occurs varies slightly from oocyte to oocyte (compare Figs. 2A, 4, and 5, A-B). This would be consistent with differing internal concentrations of nitrate in each of these oocytes. To test this model, the pattern of anion-substrate inhibition of CRNA transporter activity should be compared for an oocyte that was preincubated in zero nitrate solution and then transferred to nitrate-containing saline. This type of experiment is technically difficult to perform because associated with the nitrate accumulation in the oocyte, there is an acidification of the cytosol that complicates the interpretation of the experiment.
The kinetic model developed here to describe the substrate inhibition has important physiological consequences for the activity of the high affinity nitrate uptake system. The substrate inhibition of CRNA was assayed as a decrease in the measured current, but the actual flux of anions was not determined. When the cytosolic pool of nitrate is depleted, the concentration gradient alone can drive uptake into the cell. In contrast to plants (31), fungi have small vacuolar stores of nitrate to maintain the cytosolic nitrate pools, so the concentration may be depleted more readily. An evolutionary advantage is provided by a nitrate uptake system that requires less energy input by the cell. For example, when a sudden flush of nitrate supply occurs in the environment, there can be a large increase in the external nitrate concentrations. This transport mechanism proves an energetic advantage to the fungus by uncoupling the proton gradient from nitrate uptake. Nitrate influx is driven by the concentration gradient minimizing the requirement for the more energetically expensive coupling to proton fluxes, and yet the same protein can mediate uptake by cotransport at very low external nitrate concentrations. It remains to be seen if these two transport mechanims are general features of the NNP family or are only found in the fungal members (type II) that are defined by having a large hydrophilic central loop (18).
In conclusion, we have demonstrated that a single protein can function as a high affinity nitrate transport system although the carrier has two different mechanisms for achieving nitrate membrane translocation into the cell.