A Genetic Screen for Nitrate-Regulatory Mutants Captures the Nitrate Transporter Gene

Nitrate regulatory mutants ( nrg ) of Arabidopsis were sought using a genetic screen that employed a nitrate-inducible promoter fused to the yellow fluorescent protein marker gene YFP. A mutation was identified that impaired nitrate induction, and it was localized to the nitrate regulatory gene NLP7 , demonstrating the validity of this screen. A second, independent mutation ( nrg1 ) mapped to a region containing the NRT1.1 ( CHL1 ) nitrate transporter gene on chromosome 1. Sequence analysis of NRT1.1 in the mutant revealed a nonsense mutation that truncated the NRT1.1 protein at amino acid 301. The nrg1 mutation disrupted nitrate regulation of several endogenous genes as induction of three nitrate-responsive genes ( NIA1 , NiR and NRT2.1 ) was dramatically reduced in roots of the mutant after 2 hr treatment using nitrate concentrations from 0.25 mM to 20 mM. Another nrt1.1 mutant (deletion mutant chl1-5 ) showed a similar phenotype. The loss of nitrate induction in the two nrt1.1 mutants ( nrg1 and chl1-5 ) was not explained by reduced nitrate uptake and was reversed by nitrogen deprivation. Microarray analysis showed that nitrate induction of 111 genes was reduced and of 3 genes increased 2-fold or more in the nrg1 mutant. Genes involved in nitrate assimilation, energy metabolism and pentose-phosphate pathway were most affected. These results strongly support the model that NRT1.1 acts as a nitrate regulator or sensor in Arabidopsis. by qPCR. that for all three NiR NIA1 and NRT2.1 in the mutant after 1-2 hr of N-starvation.


INTRODUCTION
6 inception) mutants were originally identified in Lotus as being defective in bacterial recognition, infection thread formation and nodule primordia initiation (Schauser et al., 1999). NIN genes encode nuclear-targeted DNA binding proteins with bZIP domains containing a signature RWPxRK sequence. The Arabidopsis NLP7 gene was recently shown to encode a nucleartargeted protein that is needed for full nitrate induction of several nitrate-responsive genes (Castaings et al., 2009). NLP7 mutants have altered root growth (longer primary roots and more lateral roots) typical of N-starved plants and are more resistant to water stress.
The nitrate transporter gene NRT1.1 has also been implicated in nitrogen regulation. A transcriptome analysis using serial analysis of gene expression (SAGE) showed that about 300 genes were miss regulated in nrt1.1 mutant roots, and in particular, the NRT2.1 high-affinity transporter gene showed reduced ammonium repression in the nrt1.1 mutant (Munos et al., 2004). This result is consistent with the report that NRT1.1 mediates nitrate demand regulation of high-affinity nitrate uptake (Krouk et al., 2006). NRT1.1 also controls root colonization of nitrate-rich patches by a signaling pathway that may include ANR1 as both genes are expressed in similar tissues (especially root tips) and ANR1 derepression requires NRT1.1 function (Remans et al., 2006). A signaling role for NRT1.1 is also supported by the finding that nitrate reversal of glutamate inhibition of primary root growth requires NRT1.1 function (Walch-Liu and Forde, 2008;Forde and Walch-Liu, 2009). However, because NRT1.1 functions as a nitrate transporter, making it difficult to distinguish between regulatory and transport functions, it is still controversial whether NRT1.1 is a nitrate sensor or not.
To identify additional nitrate regulatory genes and mechanisms, we performed a forward genetic screen using a nitrate-regulated promoter fused to a YFP marker. Putative mutants that showed reduced nitrate induction of the marker gene were isolated and examined. Two independent mutations were mapped and sequenced and found to reside in the NRT1.1 and the A nitrate-inducible promoter (NRP) was fused to DNA encoding the yellow fluorescence protein (YFP) and transformed into Arabidopsis. Homozygous transgenic plants were generated and tested for nitrate-responsive YFP expression using fluorescence microscopy. Seedlings grown four days with 2.5 mM NH 4 -succinate (on agarose plates with no nitrate) were treated with 20 mM KNO 3 or 20 mM KCl (both with 2.5 mM ammonium succinate) for 16 hr then examined for YFP fluorescence. The nitrate-treated seedlings had much stronger root fluorescence than the chloride-treated controls ( Fig.   1A) indicating that YFP expression was induced by nitrate in these plants.
Homozygous transgenic plants were then EMS-mutagenized to produce M2 seedlings, of which approximately 35,000 were screened for low YFP fluorescence after nitrate treatment. Initially 68 seedlings with low fluorescence were identified. Retesting in the next generation recovered 6 seedlings. Two mutants Mut21 (nrg1) and Mut164 were selected for further analysis. An example of the reduced fluorescence phenotype observed in the mutants is shown for Mut21 ( Fig. 1C-D).

Identification of Mut164 as an allele of NLP7.
The Mut164 mutation was mapped to a 55 kb fragment demarcated by the genes At4G23930 and At4g24040. All 15 genes within this region were sequenced from the mutant. This analysis revealed a mutation (C to T) in the second exon of At4g24020 (NLP7) that converted proline at position 223 to a serine. Because NLP7 has been identified as a nitrate regulatory gene (Castaings et al., 2009), identification of Mut164 in our screen demonstrated that our strategy for identifying nitrate regulatory mutants was working.

Identification of Mut21 as an allele of NRT1.1.
The nrg1mutation responsible for the Mut21 phenotype was mapped to chromosome 1 in a region encompassed by BAC clones F12K11 and F20D23 (Fig. 2). This region contained the NRT1.1 (CHL1) gene. RNA transcript analysis by quantitative polymerase chain reaction (qPCR) using oligonucleotide primers to the 3' end of the transcript revealed that there was almost no detectable NRT1.1 transcript in the nrg1 mutant (data not shown). NRT1.1 genomic DNA was amplified and sequenced from nrg1. A mutation 8 was found that converted codon Q301 to a stop codon (Fig. 2). Thus, nrg1 is allelic to NRT1.1.

Nitrate induction of gene expression is defective in nrg1.
Our analysis of nrg1 showed that nitrate induction of the NRP-YFP transgene was greatly diminished. To determine if regulation of endogenous genes was similarly affected, nitrate regulation of several nitrate-inducible genes (NiR, NIA1, NRT2.1) was examined. A well-characterized nrt1.1 mutant (deletion mutant chl1-5, (Tsay et al., 1993;Munos et al., 2004)) was included in these experiments to verify that the Mut21 phenotype was due to the mutation in NRT1.1. Plants were grown for 5 days on agarose plates with 2.5 mM NH 4 -succinate as the sole nitrogen source then treated with 20 mM KNO 3 or 20 mM KCl in the presence of 2.5 mM ammonium succinate for two hours.
Root mRNA was prepared then analyzed by qPCR. Data in Fig. 3 show that nitrate induction of NiR, NIA1 and NRT2.1 in both nrg1 and chl1-5 was significantly reduced (by greater than 80%) compared to WT. Note that mM ammonium was present during these treatments, which explains the low level of nitrate induction of NRT2.1.

Nitrate induction of gene expression is restored by N deprivation in nrg1.
The virtual loss of nitrate-induced gene expression by nrt1.1 mutations was a surprise.
We have tested for such phenotypes in the past and found little difference between WT and nrt1.1 mutants (unpublished data). Recently, Hu et al., reported a 1.7-2.2 decrease in nitrate induction of NiR, NIA1 and NRT2.1 in chl1-5 mutants compared with WT (Hu et al., 2009), which is much less than what we observed (see Fig. 3). Upon comparison of experimental protocols, we noticed that our previous conditions included a N starvation pretreatment to enhance the nitrate response, which was not done in our current experiments with Mut21. To determine if the Mut21 phenotype is affected by N deprivation, the previous nitrate induction experiment, in which plants were exposed continuously to N in the form of ammonium (Fig. 3), was repeated except that seedlings were first N-deprived for 24 h before nitrate treatment. The results show almost no loss of nitrate induction in mutant plants ( Fig. 4) indicating that N starvation for 24 hr had restored nitrate induction in Mut21 and thus rendered the nitrate response NRT1.1independent.
To determine how long it takes to lose NRT1.1-dependent induction upon N starvation, a time course experiment was performed (Supplemental Figures 1-3). Plants were grown hydroponically on 2.5 mM ammonium succinate for seven days then Nstarved by transfer to the same media with no ammonium succinate. Plants were then treated with 1 mM KCl or KNO 3 for 30 min. Root were harvested, mRNA prepared and analyzed by qPCR. The data show that for all three genes tested (NiR, NIA1 and NRT2.1), nitrate induction began to recover in the mutant after 1-2 hr of N-starvation.
After 24 hr, nitrate induction in the mutant was almost as high as for wildtype plants.
The effect of N starvation on NRT1.1 expression was measured to determine if the loss of the Mut21 phenotype could be accounted for by a loss of NRT1.1 mRNA.
Over the first 8 hours of N starvation, the level of NRT1.1 mRNA increased about 1.6fold (Supplemental Figure 4). However, after 24 hr, the level dropped 4-fold. These results indicate that the loss of NRT1.1-dependent regulation during the first 8 hours of N starvation is not due to the loss of NRT1.1 expression (i.e. mRNA) and may be due to a post-transcriptional modification. At 24 hr, the drop in NRT1.1 mRNA was sufficiently large that it should contribute to the loss of the Mut21 phenotype.
The experiments described above cannot determine where it is the N deprivation in general or the loss of ammonium in particular that is responsible for the loss of the Mut21 phenotype. Including 5 mM ammonium during the 2 hr nitrate induction treatment of N-starved seedlings did not restore the Mut21 phenotype (data not shown).
Further experiments are needed to resolve this issue.

Loss of nitrate induction in nrg1 is not accounted for by impaired nitrate uptake.
Since NRT1.1 encodes a nitrate transporter, it is possible that the loss of nitrate induction in the nrt1.1 mutants is due to reduced nitrate uptake. To test this idea, nitrate induction of NiR in WT and the two nrt1.1 mutants were assayed at various concentrations of nitrate (0.25 -20 mM) in the presence of ammonium (Fig. 5). Nitrate induction was virtually abolished in both mutants at all nitrate concentrations tested under these conditions. Next, nitrate accumulation in whole seedlings was also measured after the same 2 hr treatments (Fig. 6) under the same conditions. Nitrate accumulation was lower in the mutants than the WT at all the concentrations of nitrate tested; however, the amount of accumulation was still substantial enough in the mutants (36-77% of WT) to support nitrate induction. For example, nitrate accumulation at 20 mM nitrate in the mutants is as much or more than in WT plants treated with 0.25 mM to 5 mM nitrate, yet nitrate induction is vanishing small in the mutants at 20 mM nitrate (Fig. 6). In fact, the amount of nitrate entering the plants under all concentrations tested is more than sufficient for induction, as uptake from solutions with only 2-5 μM nitrate is needed for strong induction (Wang et al., 2007). Thus, reduction in nitrate uptake cannot explain the loss of nitrate induction in the mutants.  Table 1).
The microarray data showed that 111 genes had lower induction ratios of 2-fold or more in nrg1 plants and only 3 genes had higher induction ratios of 2-fold or more in nrg1 plants (Supplemental Table 2). Many known nitrate-inducible genes including NiR  Table 3).

DISCUSSION
There has been mounting evidence that NRT1.1 functions not only as a nitrate transporter but also as a regulator. NRT1.1 expression is atypical for a root uptake transporter, being targeted to root tips, lateral root primordia and nascent shoot organs Because NRT1.1 functions as a nitrate transporter, the signaling defects described above could be explained by reduction of nitrate uptake into cells in which the nitrate sensor resides. In the reports described above, inhibition of nitrate uptake was found not to explain the nrt1.1 mutant phenotypes; however, in the earlier reports where bulk uptake into roots was measured, it was difficult to rule out the possibility that reduced nitrate uptake into select sensing cells in root tips could account for the effects. Results from Walch-Liu and Forde (Walch-Liu and Forde, 2008) provide additional insights because they found that a nonphosphorylatable form of NRT1.1, which retains low affinity but not high affinity uptake activity (Liu and Tsay, 2003), was not capable of nitrate reversal of glutamate inhibition of root growth (Walch-Liu and Forde, 2008). In our experiments, we measured gene expression in whole roots, which is not restricted to a small number of select cells in the root, so that measurements of nitrate uptake into seedlings should be more indicative of nitrate availability for induction. In our system, nitrate uptake in the mutants was more than sufficient to induce a nitrate response, yet induction was clearly impaired. was nitrate-free with 2.5 mM ammonium succinate as the nitrogen source. After incubation at 4ºC for two days, seedlings were grown at 25ºC with 24 hr light. Four-day old seedlings were then flooded with 12.5 ml of medium containing 20 mM KNO 3 and 2.5 mM ammonium succinate for 16 hr. Seedling were screened under a fluorescence microscope (Nikon Eclipse TE2000-U) and rescued. Putative mutants were selfed then rescreened. Confirmed mutants were backcrossed to the transgenic WT and made homozygous before analysis.
Growth and treatment conditions: For qPCR analyses and nitrate accumulation assays, seedlings were grown on vertical agarose plates as described above for 5 days with 2.5 mM ammonium succinate as the sole nitrogen source in plant growth medium (Wang et al., 2004). The seedlings were then flooded with 12.5 ml of plant growth medium (with 2.5 mM ammonium succinate) plus KNO 3 at various concentrations for 2 hr with agitation (60 rpm) under light. Roots were then collected for total RNA preparation (as described (Wang et al., 2003)). Control samples were prepared at the same time with the same concentration of KCl in place of KNO 3 .
For nitrate treatments without ammonia, seedlings were grown on plant growth medium with 2.5 mM ammonium succinate for 4 days then transferred to fresh agarose plates without nitrogen for 24 hr followed by flooding with nitrate containing plant growth media as described above except that there was no ammonium succinate in the liquid medium.
For the microarray analysis, plants were grown in aseptic hydroponics as described (Wang et al., 2007) for 7 days with modifications as follows: Seedlings were transferred to 100 ml of fresh medium with 2.5 mM ammonium succinate after 6 days of growth and continued incubation for 24 hr. Nitrate and control chloride treatments were initiated by adding KNO 3 or KCl to the growth media to yield 1 mM concentration then incubated for 30 min before harvesting roots.        Plants were grown and treated as described in legend to Fig. 3 except at day 4, plants were transferred to N-free medium for 24 hrs then treated with 20 mM nitrate or chloride for 2 hr with no added ammonium succinate. Root mRNA levels were determined by qPCR. Error bars represent standard deviation (n=3).  Seedlings were grown 5 days on agarose plates with 2.5 mM ammonium succinate then treated with various concentrations of KNO 3 or KCl for 2 hours in the presence of 2.5 mM ammonium succinate before roots were collected for RNA preparation. NiR mRNA levels were determined by qPCR. Error bars represent standard deviation (n=3).  Nitrate in media Nitrate concn in tissues, mM WT nrg1 chl1-5 WT nrg1 chl1-5 WT nrg1 chl1-5 WT nrg1 chl1-5 Figure 6. Nitrate accumulation.
Seedlings grown 5 days with 2.5 mM ammonium succinate were treated with various concentrations of KNO 3 (same as for Fig. 5) in the presence of 2.5 mM ammonium succinate for 2 h. Whole seedlings were then collected for nitrate assays as described in Materials and Methods. Error bars represent standard deviation (n=3).