Bread Wheat (Triticum aestivum L.) Grain Protein Concentration Is Related to Early Post-Flowering Nitrate Uptake under Putative Control of Plant Satiety Level

The strong negative correlation between grain protein concentration (GPC) and grain yield (GY) in bread wheat complicates the simultaneous improvement of these traits. However, earlier studies have concluded that the deviation from this relationship (grain protein deviation or GPD) has strong genetic basis. Genotypes with positive GPD have an increased ability to uptake nitrogen (N) during the post-flowering period independently of the amount of N taken up before flowering, suggesting that genetic variability for N satiety could enable the breakage of the negative relationship. This study is based on two genotypes markedly contrasted for GPD grown under semi-hydroponic conditions differentiated for nitrate availability both before and after flowering. This allows exploration of the genetic determinants of post-flowering N uptake (PANU) by combining whole plant sampling and targeted gene expression approaches. The results highlights the correlation (r² = 0.81) with GPC of PANU occurring early during grain development (flowering–flowering + 250 degree-days) independently of GY. Early PANU was in turn correlated (r² = 0.80) to the stem-biomass increment after flowering through its effect on N sink activity. Differences in early PANU between genotypes, despite comparable N statuses at flowering, suggest that genetic differences in N satiety could be involved in the establishment of the GPC. Through its strong negative correlation with genes implied in N assimilation, root nitrate concentration appears to be a good marker for evaluating instantaneous plant N demand, and may provide valuable information on the genotypic N satiety level. This trait may help breeders to identify genotypes having high GPC independently of their GY.


Introduction
Although grain yield (GY) remains the main breeding priority for wheat breeders, grain protein concentration (GPC) is another important priority, as GPC is a key element of wheat end-use value [1,2] and thus determines both the price paid to the farmer per tonne and also the opportunity for achieving export quality.However, the strongly negative relationship between GY and GPC [3,4] complicates the simultaneous improvement in these two traits, as increases in GY are generally detrimental to GPC.A classic agronomic strategy for obtaining high GY with adequate GPC in high-input agriculture is based on growing genotypes having strong yield potential, coupled with fertilisation practices that favour high GPC.In particular, delaying the last nitrogen (N) fertiliser application at heading, allows the low GPC, potentially associated with high GY, to be increased to an acceptable level.However, this approach is increasingly being questioned in a world where there is increasing pressure to reduce N inputs, both because of its rising cost and also because of a rising awareness of the environmental damage resulting from excessive N use in agriculture [5].
Monaghan et al. [6] proposed an alternative strategy for improving GPC without decreasing GY, thus breaking the negative GY-GPC relationship.This strategy originates from the observation that some genotypes exhibit robust deviations (whether positive or negative) from the regression of GPC on GY.This deviation, grain protein deviation (GPD), has a strong genetic basis [4,7] and has interesting potential as a breeding criterion.Based on this proposition, some studies have focused on identifying the physiological basis of genetic variability in GPD.In particular, Bogard et al. [8] showed that GPD was highly correlated with post-flowering N uptake regardless of plant N status at flowering.This led to the hypothesis that GPD could be conditioned by a putative genetic variability in the control of N satiety, that would allow some varieties to absorb more N after flowering.Testing of such a hypothesis demands a better understanding of post-flowering N uptake in wheat and its regulation by plant N status.
Grain N originates from two distinct sources: remobilisation from vegetative organs of N assimilated pre-flowering, and post-flowering uptake of N from the soil.The relative contributions to grain N of these two sources are negatively correlated [6,8,9], and are strongly dependent on environmental factors [10].Typically, for wheat under field conditions, postflowering N uptake contributes between 5 and 40% to total grain N [8,[11][12][13].The fraction depends also on growing context and, to a lesser extent, on genotype.A number of experiments have been carried out on wheat and barley under controlled conditions to study N uptake, including during the post-flowering period.These offer interesting understandings, such as of effects of N availability on the dynamics of biomass accumulation and N concentration [14], of N uptake dynamics under N-limiting conditions [15], of root N uptake capacity during the post-flowering period [16], and of the fate of the N taken up post-flowering [17].Nevertheless, little is known of the kinetics of post-flowering N uptake, or of the associated control mechanisms.New information is needed at both the whole plant scale and at the molecular scale.In the latter case, the conservation of the metabolic pathways associated with both N uptake and N assimilation in higher plants, offers useful tools for studying post-flowering N uptake at the molecular scale.
Several recent reviews provide an extensive view of known mechanisms of N uptake and assimilation [18][19][20][21].Uptake of N appears to be regulated mainly by plant demand [22,23].However, a major area of continuing discussion among authors centres on the nature of the signal for N satiety.This has been proposed to be: (a) NO 3 - [24][25][26] or (b) circulating amino-acids such as glutamine [27,28].The principal forms in which N is taken up from the soil by plants are as nitrate (NO 3 -) and as ammonium (NH 4 + ), with NO 3 -being preferred under aerobic conditions [29].In higher plants, NO 3 -uptake involves two distinct systems classified according to their affinities for NO 3 -.The low-affinity transport system (LATS) is predominant when N concentrations ([N]) in the medium are high, with a K m constant in the mM range.The high-affinity transport system (HATS) operates under low [N] conditions, with a K m in the μM range [19].
LATS is coded by the NRT1 gene family which has 53 known members in Arabidopsis thaliana while HATS is coded by the NRT2 gene family which has seven members in A. thaliana.Among the LATS family, AtNRT1.1 (CHL1) is the most studied, and both it and AtNRT1.2 are expressed in the root epidermis and participate in NO 3 -uptake [30].Meanwhile, AtNRT1.1 is a dual-affinity transporter being both constitutive and inducible, and it is involved in NO 3 -uptake under both high-and low-NO 3 -concentrations [31].Moreover, it has recently been shown that AtNRT1.1 can transport both NO 3 -and/or auxin depending on the external concentrations of NO 3 - [32].In rice, Oryza sativa, OsNRT1 is the homolog of AtNRT1.1 and has been cloned and characterised as a low-affinity NO 3 -transporter which is expressed constitutively in the root epidermis [33].Based on sequence homology approaches, a number of other putative NRT1 family members (homologous to the A. thaliana NRT1 family) have been found in a range of species including: Cucumis sativus, Oryza sativa, Zea mays, Sorghum bicolor and Brachypodium distachyon [34,35].Although homologues of most members of the A. thaliana NRT1 gene family probably exist in wheat, to our knowledge, only one has been studied [36].
The two main genes expressed in A. thaliana roots coding for HATS are AtNRT2.1 and AtNRT2.2.Of these, AtNRT2.1, in interaction with an AtNAR2 protein [37] accounts for about 75% of the high-affinity NO 3 -uptake [38].These genes are induced by low NO 3 -concentration ([NO 3 -]) and repressed by high [NO 3 -] and so belong to the group of inducible HATS.
The HATS NO 3 -uptake system is also implicated in plant responses to N starvation.Expression of AtNRT2.1 increases rapidly when NO 3 -is supplied to N-starved plants but later decreases if the provision of NO 3 -is maintained [39].Putative homologues of the main HATS coding genes in maize are ZmNRT2.1 and ZmNRT2.2.These were detected by the sequence homology approach [34] and were later studied under hydroponic conditions where they seem to play a key role in N uptake even at [NO 3 -] higher than 1 mM [40].The orthologue of AtNRT2.1 in wheat is TaNRT2.1, and this has been characterised and its expression profiles suggest that it too belongs to the inducible HATS [41].To our knowledge, TaNRT2.1 is the only member of the NRT2 gene family currently characterized in wheat.Following its uptake, NO 3 -is reduced to nitrite (NO 2 -) by nitrate reductase (NR) and the NO 2 -is subsequently reduced to ammonium by nitrite reductase (NiR) [42].A loop between glutamine synthetase 2 (GS2) and the two forms of glutamine 2-oxoglutarate amino transferase (GOGAT), NADH-GOGAT and Fd-GOGAT, finally integrate the N into amino acids [43].Genes involved in N reduction and assimilation have been investigated in wheat [44][45][46].The objective of our study was to provide a clearer view of post-flowering NO 3 -uptake dynamics in wheat in relation to the N status at flowering.To study the uptake of NO 3 -in detail, a semi-hydroponic approach under controlled conditions was established because of its susceptibility to environmental factors.Wheat plants (cv.Récital) were grown under four preflowering conditions with contrasting levels of [NO 3 -].At flowering, all conditions were altered to be NO 3 -non-limiting, allowing plants to fully exploit their NO 3 -uptake potentials.The use of 15 N pulse labelling allowed detailed recording of the dynamics of post-flowering NO 3 -uptake and relative quantification of the expression of the main genes involved in both NO 3 -uptake and assimilation and enabled links to be observed between gene expression and NO 3 -uptake.

Plant material
Selected wheat grains (55 to 60 mg) of cv.Récital were sown in germination trays filled with compost and placed in a heated greenhouse at 20°C for 14 days.After emergence, plants were vernalised for six weeks in a plant growth chamber (6°C, 8 h photoperiod, light intensity 350 μmol PAR m -2 s -1 ).After vernalisation, the roots were washed free of compost and plants were transplanted into PVC tubes (7 cm diam, 60 cm high) filled with a perlite:sand mixture (1:1, v: v) for semi-hydroponic culture.The lower end of each tube was closed with a perforated cap to contain the substrate while allowing excess nutrient solution to drain.Two plants were transplanted into each tube and tubes were placed vertically in eight containers (64 tubes per container, container area 0.49 m 2 ).The resulting planting density (260 plants m -2 ) is comparable to that in the field under local agronomic practices.Containers were placed in a growth chamber under a long-day photoperiod (16 h light at 20°C, 8 h dark at 18°C).Light intensity was 650 μmol PAR m -2 s -1 .Air temperature, humidity and light intensity at canopy level were recorded every 15 min using a CR1000 datalogger (Campbell Scientific, Logan, Utah, USA).Each tube was fitted with an automated microirrigation system, receiving one of four different nutrient solutions containing 1, 4, 7 or 10 mM NO 3 -(the four nutrient treatments are referred to as N1, N4, N7 or N10, respectively).The nutrient solution compositions were adapted from Castle and Randall [47] (S1 Table ).Two of the eight containers were irrigated with each of the nutrient solutions, with each tube receiving 66 ml of nutrient solution every three hours.The flowering date (Zadok's GS65) of each main stem was recorded.From flowering to maturity, all containers were irrigated with 66 ml every three hours of the same nutrient solution containing 10 mM of NO 3 -.Before each sampling, a subset of plants from each treatment (8 tubes per NO 3 -pre-flowering condition) were submitted to 24 h of pulse labelling with 15 N by watering with 66 ml every three hours of a 10 mM NO 3 -nutrient solution enriched in 15 N by 10%.

Sampling protocol
Eight sampling dates were chosen during the post-flowering period.The first destructive sampling took place one day after flowering (GS65) and the remainder at approximately GS65+ 125, 250, 400, 500, 600, 800 and 1200 degree-days (DD) base 0°C.Each sampling was made at the same time of day (1-2 h after the start of the light period).On each sampling date, 16 tubes (32 plants) from each of the four NO 3 -treatments were collected.Of these, four were used for physiological measurements and analysis of total nitrogen, four for gene expression and NO 3 analysis and the remaining eight for 15 N analysis.Each tube (two plants) was considered a biological replicate.

Physiological measurements and total nitrogen sample analysis
Physiological analyses were performed on four biological replicates.The processing of the two plants from each tube consisted first of counting the number of tillers and spikes per tube.The plants were then separated into six fractions; grains, chaff, stems (including leaf sheaths), green laminae, dry laminae, and roots.At stage GS65, in the absence of developed grains, ears were not separated into grain and chaff.All samples were oven dried to constant weight for 48 h at 80°C before dry weight (DW) measurement.Dry samples were ground to a fine powder using a ball mill and total nitrogen concentration was measured with the Dumas combustion method using a Flash EA 1112 Series CNS analyser (ThermoFisher Scientific, Illkirch, France).

Analysis of 15 N
Plants exposed to 15 N pulse labelling were oven dried for 48 h at 80°C before DW measurement and ground before shipping to the UC Davis Stable Isotope Facility (Davis, CA, USA) for 15 N abundance determination where samples were analysed using a PDZ Europa ANCA-GSL elemental analyser interfaced to a PDZ Europa 20-20 isotope ratio mass spectrometer (IRMS) (Sercon Ltd., Cheshire, UK).The 15 N analyses were performed on eight biological replicates.

Gene expression and NO 3 -concentration assay sample preparation
The same samples were used to assay gene expression and NO 3 -concentrations.Roots, stems and flag leaves of the two plants from each tube were pooled to form one biological replicate per organ, ground in liquid nitrogen, and stored at -80°C pending analysis.

NO 3 -concentration assays
Root, stem and flag leaf sub-samples (10 mg) were shipped to the Bordeaux INRA Metabolome Platform (https://www.bordeaux.inra.fr/umr619/RMN_index.htm; Bordeaux, France) for determination of NO 3 -concentrations using a spectrophotometry method as described in Cross et al. [48].The NO 3 -assays were performed on four biological replicates.

qRT-PCR experiments
Total RNA was extracted from three biological replicates of root, stem and flag leave samples from treatments N4 and N10 for the first seven sampling dates.Samples from the final sampling date were not extracted because of their advanced state of senescence.About 100 mg of frozen powder were used for each RNA extraction which was carried out using the Nucleomag 96 RNA kit (Macherey-Nagel, Düren, Germany) on the Biosprint 96 (Qiagen, Hilden, Germany).Total RNA samples were then purified with the NucleoSpin 96 RNA kit (Macherey-Nagel).The RNA quality was observed by migration in agarose gel.Reverse transcription of 1 μg of RNA was carried out with the iScript Select cDNA Synthesis kit (Bio-Rad, Richmond, California).The three steps-extraction, purification and reverse transcription-were carried out according to the manufacturer's instructions.Primer pairs were developed to quantify expression levels of TaNRT1 (GenBank AY587264), TaNRT2.1 (GenBank AF332214.1),TaGS2 (GenBank DQ124212.1),TaNR, TaNiR, TaFd-GOGAT and TaNADH-GOGAT (whose partial sequence coming from Boisson et al. [44]) (S2 Table ).All couples developed were generic and, thus, amplified three homeologous copies.The specificity of the primer pairs and their efficiency were validated between 85 and 100%.
The quantitative real-time PCR experiment (qPCR) was carried out on a LightCycler 480 system (Roche, Indianapolis, Indiana) with the LightCycler 480 SYBR Green 1 Master Kit.The qPCR program involved pre-incubation for 10 min at 95°C, followed by 45 cycles of amplification, each consisting of denaturation for 10 s at 95°C, followed by annealing for 15 s at 60°C and finishing with elongation at 72°C for 15 s.To ensure that single products were amplified, a melting curve was analysed at the end of each assay.Because of relatively low expression levels, the two NO 3 -transporter genes were analysed with the cDNA template diluted to 1/8, while for all other genes the cDNA template was diluted to 1/20.qPCR was carried out using a 15 μl reaction volume containing 5 μl diluted cDNA, 7.5 μl SYBR green mix, 0.75 μl each of 10 μM forward and reverse primers and 1 μl of water.Relative expression was determined using the ΔCT method corrected for primer efficiency [49].Results were normalised to the expression of two housekeeping genes, Ta54280 and Ta54948, selected from Paolacci et al. [50] (S2 Table ) whose expression stability had already been validated under our experimental conditions.

Statistical analyses
Results were analysed after conversion to a per-square-meter basis, using tube surface area.All statistical analyses were carried out using R v2.15.1 [51], and graphics were drawn using Sigma-Plot v8.0.

Nitrogen nutrition index
The N nutrition index (NNI) was calculated at flowering as the above-ground N concentration divided by the critical plant N concentration.The critical plant N concentration is defined as the minimum N concentration needed for maximum growth rate [52].The critical N concentration was calculated using the equation of the critical N concentration curve for cereals proposed by Justes et al. [53].

Effects of pre-flowering NO 3 -treatments on plant morphology, biomass and N concentration
Four levels of NO 3 -in the nutrient solutions were used until flowering (1, 4, 7 and 10 mM, respectively referred to as N1, N4 N7 and N10).These led to highly contrasting tiller numbers, biomasses and N concentrations at flowering (Table 1).Tiller number increased significantly with rising [NO ].With regard to the critical nitrogen dilution curve developed for wheat [53], only N1 appeared to be limited by N status with NNI less than 1.This index was developed for field crops, so it should be used carefully under our conditions, nevertheless it clearly indicates a differential effect for the four [NO 3 -] treatments used in our study.
After flowering, all plants were exposed to the same, non-limiting 10 mM NO 3 -nutrient solution.At maturity, the DW hierarchy between treatments was conserved (Table 2) although there was no significant difference between N7 and N10.The range of DW was fairly consistent with values of 1701, 2966, 4637 and 4822 gDW m -2 for N1, N4, N7 and N10, respectively.For plant [N] at maturity, despite no significant differences between N1, N4 and N10, there did seemed to be a trend for plants having been exposed to high pre-flowering N, to have less plant [N] at maturity.Plant [N] at maturity was 2.1 and 2.2%DW for the high-NO 3 -treatments (N7 and N10, respectively) and 2.4 and 2.5% DW for the low-NO 3 -treatments (N1 and N4, respectively).Interestingly, the two groups of treatments described for plant [N] stood out for the agronomic variables of grain yield, grain N concentration (grain [N]) and total grain N.There were no significant differences in grain yield either between N1 and N4 (690, 1051 g m -2 , respectively) or between N7 and N10 (1828 and 1944 g m -2 , respectively).Probably due to the semihydroponic growth conditions, grain yields were high (equivalent to 69 and 194 q ha -1 for the two extreme treatments).Grain [N] values were 3.1 and 2.9%DW for N1 and N4, respectively and 2.5%DW for N7 and N10, respectively.There were strong and highly significant opposite-going effects of pre-flowering NO 3 -treatment on grain yield and grain [N].Grain yield and grain [N] were therefore negatively correlated (r 2 = 0.99, p = 0.002).Total grain N (Table 2) was lower than total post-flowering N uptake (calculated by difference between total plant N at maturity, Table 2, and total plant N at flowering, Table 1), indicating that there was no net remobilisation of N from vegetative parts to the grains under our semi-hydroponic conditions.Harvest index lay between 0.39 and 0.44, which is close to values commonly observed under field conditions.Starting from significantly different levels of plant [N] at flowering, plants from the four [NO 3 -] treatments quickly converged showing comparable [N] by GS65+300DD (Fig. 1).Convergence occurred between flowering and GS65+300DD, and led to plant [N] values between approximately 2.1 and 2.5%DW for all treatments.Interestingly, the convergence between treatments was reflected not only in an increase in plant [N] in the low-N pre-flowering treatments (N1 and N4) but also in a decrease in the high-N pre-flowering treatments (N7 and N10).During this phase, plants from N1 and N4 exhibited [N] increases of 59 and 20%, respectively, while plants from N7 and N10 exhibited [N] decreases of 13 and 18%, respectively.From GS65+300DD until maturity, plant [N] remained relatively stable for all treatments, varying between 2.0 and 2.7%DW.

NO 3 -uptake during the post-flowering period
The 15 N pulse labelling allows precise measurement of the dynamics of plant NO 3 -uptake during the post-flowering period.Fig. 2A shows that all treatments had strong NO 3 -uptakes at flowering,  -uptake showed a rapid decrease from flowering until a date between GS65+300DD and GS65+400DD where uptake stopped.NO 3 -uptake then recovered rapidly to a substantial level between 0.2 and 0.5 gNO 3 -m -2 day -1 depending on [NO 3 -] treatment and then slowly declined until maturity.In addition to NO 3 -uptake dynamics, these results also show that NO 3 -uptake continued until late during the post-flowering period when plants were exposed to a non-limiting level of NO 3 -in the nutrient solution.
When NO 3 -uptake is normalised to plant DW, it is easier to interpret differences in NO 3 -uptake between the four [NO 3 -] treatments (Fig. 2B).Although in terms of temporal dynamics, NO 3 -uptake looks very similar either in absolute terms (Fig. 2A) or when normalised by DW (Fig. 2B), the latter representation shows that plants from N1 had strong NO 3 -uptakes per unit DW (Fig. 2B).As expected, NO 3 -uptake was strongly impacted by plant N status at flowering.

Post-flowering dynamics of [NO 3 -] and NO 3 -transporter gene expression in roots
The gene expression and NO 3 -concentration measurements were carried out only on N4 and N10 plants because these represent a best compromise between a distinct N effect and non-  ] in the roots were not significantly related to NO 3 -uptake (data not shown).
In the two members of the main families of low and high affinity NO 3 -transporters, TaNRT1 and TaNRT2.1,post-flowering gene expression dynamics provides valuable information that increases our understanding of NO 3 -uptake.For TaNRT2.1, the relative expression level at flowering depended on the pre-flowering NO 3 -treatment (Fig. 4A) with its expression in N4 roots being twice that in N10 roots.Both treatments then exhibited a sharp decrease in expression of root TaNRT2.1 between flowering and GS65+200DD.After that, the difference in expression between the two treatments became small, both showing an increase after 200DD and a further decrease between GS65+400DD and GS65+600DD.The last measurement point shows an unexpected increase in TaNRT2.1 expression for both treatments but the more pronounced increase in N4 led to significant differences between N4 and N10.
The relative expression dynamics of TaNRT1 in roots (Fig. 4B) differed from that of TaNRT2.1.First, expression of TaNRT1 was low at flowering with no significant difference in relative expression levels between N4 and N10 at that time.After that, expression in N4 plants was consistently lower than in N10 plants, except at GS+400.Second, the expression patterns between the two genes are extremely different, with TaNRT1 showing an increase in expression from flowering to GS65+400DD and then a decrease which was more pronounced in N4 than in N10.This was followed by an increase during the final measurement period in N4 which came into balance with expression in N10.
The level of NO 3 -uptake per unit DW was significantly and positively correlated with TaNRT2.1 expression in N4 (r 2 = 0.707, p = 0.018; Fig. 5A) but not in N10 (r 2 = 0.011).The variation in both variables was extremely small in the latter treatment relative to those in N4 plants.Conversely, TaNRT1 expression was not significantly correlated with NO 3 -uptake (r 2 = 0.21 and 0.33 for N4 and N10, respectively; Fig. 5B).A strong negative correlation was observed between root [NO 3 -] and TaNRT2.1 expression for N4 (Fig. 6A), while no significant relationship was detected for N10.Results from N10 plants were significantly impacted by three unexplained outliers originating from the GS65 +400DD sampling date.Without these, the relation between the two variables would have been comparable under the two treatments.
In order to focus on a developmental phase when the grain sinks for N were weak, the relation between root NO 3 -content and TaNRT2.1 expression was also observed during the period between flowering and GS65+300DD.Here, the negative relationship between TaNRT2.1 expression and root [NO 3 -] was highly significant for both treatments (r 2 = 0.94 and 0.82 for N4 and N10, respectively, Fig. 6B).

NO 3 -reduction and assimilation network
To provide a biologically coherent description, correlations between the expressions patterns of the major genes involved in NO 3 -uptake, the reduction and assimilation networks were examined pairwise according to the path of NO 3 -assimilation.Root samples from N4 and N10 were observed independently.In the same way as for [NO 3 -], results presented for gene expression are based solely on root samples.This specific focus on roots was again retained for clarity, enabled by high positive correlations (0.61 < r < 0.81) observed between roots and flag leaves for all observed genes except TaNADH-GOGAT (S3 Fig) .Expression dynamics in stems during the postflowering period were however correlated to those observed in roots for TaNR only (S4 Fig).
There was a significant positive correlation between the expression patterns of the highaffinity NO 3 -transporter family member TaNRT2.1 and the nitrate reductase TaNR in both N4 and in N10 in roots (Fig. 7A).The lower r 2 observed for N10 is explained mainly by the three outliers.Conversely, there was no obvious link between the expression of the low-affinity NO 3 -transporter family member TaNRT1 and TaNR either in N4 or in N10 (Fig. 7B).Subsequently, there was a strong positive correlation between TaNR and the nitrite reductase (TaNiR) expression patterns (Fig. 7C).Correlations were significant in both N4 and N10 between these two contributors to N reduction.Continuing to follow the path of N in the plant, significant positive correlations were observed between the expressions of TaNiR and glutamine synthetase 2 gene (TaGS2) in the two treatments (Fig. 7D).Lastly, correlations were observed between the expressions of TaGS2 and two forms of glutamine oxoglutarate aminotransferase (GOGAT)-ferredoxin-dependent GOGAT (TaFd-GOGAT; Fig. 7E) and NADH-dependent GOGAT (TaNADH-GOGAT; Fig. 7F).At this level of N assimilation,    expression patterns were positively and significantly correlated both in N4 and N10.The overall observation of correlations between the relative expression patterns of all these central network genes suggests a coordinated regulation of nitrogen metabolism with, probably, a limited involvement of TaNRT1 under our experimental conditions.

Discussion
The objectives of this study were to provide a detailed vision of post-flowering N uptake and to identify possible control mechanisms, using a prospective approach based on correlations between physiological traits, NO 3 -uptake, NO 3 -concentrations and expression level of key genes of the N metabolism.Clearly, the present work does not provide any functional dissection of the interaction between these physiological processes but highlights strong correlations between variables during post-flowering N uptake on wheat plants in realistic conditions in term of plant size, canopy structure and grain sink strength.

Evaluation of growing conditions
Growing wheat in controlled hydroponic conditions allows nutrient availability to be closely regulated.Nevertheless, it is difficult to obtain nutrient availabilities fully comparable with those under field conditions.Previous studies [14][15][16][17]54] have developed fertilisation protocols for use in hydroponic culture that allow field N conditions to be mimicked with constant availability of growth-limiting levels of N. Here, a more classical approach was used to drive growth during the pre-flowering period.This involved four distinct A further objective of the experimental setup was to apply NO 3 -at excess levels during the post-flowering period in order to study the effects of N status at flowering on NO 3 -uptake under non-limiting conditions.Although 10 mM [NO 3 -] in a nutrient solution is usually considered to be non-limiting for growth in Arabidopsis [21,55,56], and is usually the highest level N treatment in hydroponic wheat [36,57], it is difficult to be certain that this level was strictly non-limiting as we did not include treatments with even higher levels of [N].Barneix [58] stated in a review that high N uptake rates, lead to high levels of N nutrition, but they inhibit leaf senescence and the remobilisation of N to the grain.In our study, the absence of significant net remobilisation of N from vegetative parts during the post-flowering period (the amount of N taken up was higher than the total in the grain for all treatments) suggests that NO 3 -was indeed available in a sufficient quantity.This scenario differs from the 60 to 95% of N remobilisation usually observed in the field [8,11,13].Therefore, it seems reasonable to infer that the NO 3 -treatments used here led not only to highly contrasting stands at flowering but they further allowed the post-flowering growth to be non-limiting with regard to NO 3 -.

Post-flowering control of NO 3 -uptake
Under our semi-hydroponic, non-limiting NO 3 -conditions, plants exhibited positive NO 3 -uptake until late in grain filling.Similar observations have been reported for barley cultured hydroponically under N-limiting conditions [15] and for wheat [14,16,17,54], as well as for wheat under semi-controlled conditions, more similar to the field [59].In these studies, N uptake also occurred until close to grain maturity, showing that these plants do not have an intrinsic physiological incapacity to take up N during grain filling.Compared to the NO 3 -uptake rates in the present study, the lower rates generally observed in the field are thus unlikely related to an intrinsic incapacity of the plant but instead to some unfavourable conditions that affect either the availability of N in the soil or the plant's demand for it.
In their hydroponic study on wheat, Oscarson et al. [16] showed that under N-limiting conditions, N uptake gradually decreases throughout the post-flowering period.However, under non-limiting conditions, we show that NO 3 -uptake is far from constant from flowering to maturity.This leads to the hypothesis that NO 3 -uptake is regulated by factors other than NO 3 availability.Information on N uptake dynamics in wheat under non-limiting conditions is scarce, particularly during the post-flowering period.Both Imsande and Touraine [60], and Feil [22] in a review based on both controlled conditions and field experiments hypothesised that N uptake reflects the internal demand of the crop and is not determined only by the external [N].This idea also finds support in some more recent experimental studies [61,62] as well as in assessments of N uptake arising from wheat crop simulation models.These models, despite some calculation differences, calculate N uptake as governed by a balance between the plant's needs to support growth and the availability of N in the soil [63].Thus, under nonlimiting N conditions, N uptake is potentially fully governed by the plant's demand for N for growth.
If the present study confirms earlier findings on post flowering N uptake in wheat, it also provides a clearer view of the small-scale dynamics of the process.Post-flowering NO 3 -uptake dynamics can be divided into three distinct phases.The first of these is from flowering to sometime between GS65+300DD and GS65+400DD (depending on treatment).Here NO 3 -uptake declines considerably from its level at flowering.Next, in the second phase, NO 3 -uptake almost ceases for a period of about 100DD.And last, in the third phase, a recovery of NO 3 -uptake occurs which persists to maturity.Interestingly, each phase coincides with a physiological evolution in development, supporting the idea that NO 3 -uptake could be strongly controlled by a growth demand for N. The general decline in NO 3 -uptake during the first phase is synchronous with a reduction in the sink strength of plant growth.Indeed, immediately after flowering, there are few new sinks for N as the photosynthetic apparatus is already fully developed, whereas the spikes are not yet into a phase of rapid biomass increase.The main biomass increment occurring during this phase is driven by post flowering stem elongation.The stem is a structural tissue that is relatively poor in N compared with, say, a leaf.Both the upper internode and the ear peduncle are known to continue extension for approximately one week after anthesis [64].Although a common tendency in NO 3 -uptake decrease was observed after flowering in the four NO 3 -treatments, there were quite marked differences in NO 3 -uptake levels between treatments.Thus, immediately after flowering when NO 3 -availability became non-limiting, NO 3 -uptake ranked inversely with the level of pre-flowering NO 3 -treatment.Also, NO 3 -uptake rate at flowering and its dynamics during the first phase immediately afterwards, are in line with a convergence towards a common value for [N].This observation supports the idea that NO 3 -uptake could be regulated also by plant N status.The high levels of NO 3 -uptake at this time observed in the low-NO 3 -treated plants, was associated with an increase in total plant [N] and this mainly through N enrichment of the existing organs.This phenomenon is a familiar response of plants re-supplied with N after a period of N starvation.Under such conditions increased N uptake rates have been reported in both wheat [65] and barley [24].Furthermore, the plant [N] convergence between treatments was not due solely to increased plant [N] in the low-NO 3 -treatments but also to decreased plant [N] in the high-NO 3 -treatments.This was linked mainly with a greater increase in stem biomass in the high-NO 3 -treatments than in the low-NO 3 treatments (data not shown).The latter was probably because of a higher cover density resulting in more light competition among the plants [52].Because stems are poorer in N than leaves, the relative change in proportion between leaf and stem coupled with a low NO 3 -uptake rate resulted in a decrease in [N] at the whole plant level.
In contrast to the observations of Oscarson et al. [16], NO 3 -uptake in our study stopped completely for a period of about 100°Cdays around GS65+300DD to GS65+400DD.Physiologically, this phenomenon coincided with the absence of a significant sink for N.This is because the vegetative parts of the plant were fully developed, while the reproductive parts (the grains) had not yet started active filling.Furthermore, plant [N] was already extremely high at this stage in all treatments.
Still in agreement with the hypothesis of the control of N uptake by the N growth demand, the NO 3 -uptake resumed when the grains started to fill.If grain storage protein filling starts around GS65+240DD, then its maximal rate occurs from about GS65+300DD to GS65 +600DD [66].In absence of any net N remobilisation from the vegetative parts, this phase of renewed NO 3 -uptake was the main source of N for grain filling.Using 15 N pulse labelling in a hydroponic study, Oscarson et al. [17] showed that where N was limiting, a temporary increase in NO 3 -availability late during grain filling resulted in a preferential allocation of newlyassimilated N to the grain.The authors concluded that 15 N taken up during grain filling was rapidly incorporated in the mobile N pool of the plant and was allocated partly to tissue maintenance but mainly to grain filling.Surprisingly, our different treatments which resulted in highly-contrasting grain yields instead had closely-similar NO 3 -uptake rates during this third phase.This suggests that NO 3 -uptake was slowed in the high-NO 3 -treatments although these, at least potentially, had greater sink demand for N.As NO 3 -was in excess, the ability of plants to assimilate N probably reached its limits under such extreme conditions.

Relationships between plant NO 3 -uptake and the expression of N metabolism genes
The relative expressions of the seven main genes involved in NO 3 -uptake and N assimilation were followed for the N4 and N10 treatments.These genes are hypothesised to be representative of N network activity and composed of genes coding NO 3 -transporters, and key enzymes involved in NO 3 -reduction and assimilation.Relative expression dynamics were followed in the roots during the post-flowering period with the double purpose (a) of better understanding their behaviour and interrelationships for plants of different N status under non-limiting conditions, and (b) of observing the relationship between 15 N-based measurements NO 3 -uptake and the expression the N network genes.Although the results presented here for genes expression were focused on root samples, the conclusions of the study should not be biased for two main reasons.Firstly, a large part of N reduction and assimilation occurs in roots during grain filling in wheat (60 to 75% depending on N fertilization level) [67].Secondly, the high positive correlations observed for all studied genes between roots and flag leaves (S3 Fig) , except TaNADH-GOGAT, suggest that NO 3 -reduction and assimilation processes could be subjected to a common regulation in these two major organs for N metabolism.The significant correlation between NO 3 -uptake and TaNRT2.1 relative expression in N4 plants during the post-flowering period suggests an important role for this high-affinity nitrate transporter in NO 3 -uptake.Nevertheless, the absence of a correlation in N10 plants, suggests that the relation is not linear at constant high NO 3 -availability.Conversely, no correlation between NO 3 -uptake and TaNRT1 relative expression could be detected, suggesting a limited role for this gene under our conditions.Moreover, our results show that TaNRT2.1 is expressed at higher levels than TaNRT1 despite the high [NO 3 -] (10 mM) in the nutrient solution.Comparable results have been found in Arabidopsis grown in 10 mM NO 3 -supplied for 5 min with 6 mM NO 3 -, where AtNRT2.1 was expressed at higher levels than AtNRT1.1 both in the plant's vegetative and reproductive stages [21].Similarly, in maize, ZmNRT2.1 and ZmNRT2.2 were expressed at higher levels than ZmNRT1.1 and ZmNRT1.2throughout the life cycle with both 0.5 and 2.5 mM NO 3 -, and ZmNRT2.1 expression patterns showed strong similarities to the N uptake dynamic [40].In wheat, Wang et al. [36] also showed that TaNRT2.1 is always expressed at higher levels than TaNRT1 in NO 3 -ranges comparable to ours.These results suggest that TaNRT2.1 participates in NO 3 -uptake in a substantial way, even at high [NO 3 -].
Corroborating correlations with NO 3 -uptake, TaNRT2.1 relative expression also correlated with TaNR relative expression in roots (Fig. 7A) as well as in flag leaves (S5 Fig) .The presumed major role of TaNRT2.1 in post-flowering NO 3 -uptake is supported by its consistent expression dynamics with the rest of the network.Conversely, the expression dynamics of TaNRT1 was not correlated with TaNR and, therefore neither with the rest of the network.TaNRT2.1 thus appears to have a major role in NO 3 -uptake, especially during periods of rapid uptake under non-limiting NO 3 -conditions.However, the correlation between expression and uptake was not perfect, suggesting that TaNRT2.1 is complemented by other NO decrease.This hypothesis for NO 3 -uptake regulation fits with the assumption that NO 3 -uptake is controlled by plant growth rate under non-limiting conditions.Study of the relative expressions of the key N network genes reveals their apparent common regulation during post-flowering period, with strong correlations between the various components.Root [NO 3 -] would seem to be a good candidate for the role of transmitting the plant N status signal to the NO 3 -transporters.Although relative gene expression studies should be complemented by enzymatic activity assays, because additional regulation may occur at the post-transcriptional level, such as for NR [68], cytosolic GS1 [69], or the root ammonium transporter AMT1.1 [70], our results still seem to indicate a major role for TaNRT2.1 transporter in post-flowering NO 3 -uptake under non-limiting NO 3 -conditions.Plants have shown the ability to quickly regulate their [N] under non-limiting NO 3 -conditions when previously NO 3 -starved, coinciding with increases in the relative expressions of key N metabolism genes.It is possible that this approach may allow identification of genetic variability in the ability to quickly assimilate NO 3 -when NO 3 -is available.A large capacity for NO 3 -assimilation during short favourable periods during the post-flowering growth phase may be a determinant element for increasing grain [N].

Hypotheses for grain N content limitation
The main justification of studying the physiological basis of post flowering N uptake is its strong link with wheat grain [N], a major agronomic trait.Following the work of Monaghan et al. [6], Bogard et al. [8] demonstrated that genetic variability in grain [N] was strongly related to post-flowering N uptake capacity, regardless of plant N status at flowering.Thus, one objective of this work was to open tracks in the understanding of the complex regulations of post-flowering N uptake and N assimilation.
Here, high levels of grain yield were associated with decreased grain [N].This observation is classic, since it is the basis of the extremely well known negative relationship between grain yield and grain protein [4,22,71].It is, however, less intuitive under our conditions of nonlimiting NO 3 -availability during grain filling.
In an ear-halving experiment with four cultivars of contrasting yield potential, Martre et al. [72] showed that grain [N] was mainly source-limited as grain [N] increased in halved ears, and as grain N per ear was nearly constant.More precisely, they stated that if yield is clearly limited by sink size, then their ear-halving experiment suggests that grain N filling is relatively more limited by N source.Following this hypothesis, several potential processes limiting N grain filling can be identified, from root N uptake to protein synthesis within the grain, passing through N assimilation and N transport to ears.Our study suggests that NO 3 -uptake can reasonably be dismissed from this series, because of the high post-flowering availability of NO 3 -.In addition the results obtained by Oscarson et al. [16] show that intrinsic root N uptake capacity was never a limiting factor under hydroponic conditions.Grain protein synthesis capacity and N transport into the ear can also be eliminated.First, there is no obvious reason for a higher capacity for protein synthesis in grains under a low-NO 3 -treatment than in those under a high-NO 3 -treatment, based on the fact that in our study, the grains of the four treatments had comparable individual masses (S4 Table ).Second, using wheat ears grown in liquid culture, Barlow et al. [73] show that ear capacity to transport N compounds and grain capacity to synthesise protein were not limiting elements.This is supported by the conclusions of Martre et al. [72] which state that the capacity of the sink to synthesise proteins does not regulate grain N accumulation.We therefore conclude that, the most likely hypothesis to explain the decrease in grain [N] with increasing yield, is a limitation operating at the level of N assimilation, with a possible saturation of the assimilation pathway.
In our study, increased grain yield was principally related to higher grain numbers per square meter (S4 Table ), and these were mainly conditioned by pre-flowering NO 3 -availability.
Grain growth rate has been shown to be independent of N availability [74].The independence between grain development and N availability implies that N demand by the grain during the post-flowering period depends mainly on a plant yield potential established before flowering and therefore varies with the availability of N before flowering.Thus, plants from the low-NO 3 -treatments developed a relatively small yield potential, and filled a low number of grains under later high-NO 3 -conditions after flowering.Conversely, plants from the high-NO 3 -treatments developed a relatively high yield potential, and filled a higher number of grains under the later high-NO 3 -conditions.In the latter case, this necessarily implies a reduced share of available N per grain, thus leading to lower grain [N].Our results do not allow a precise conclusion regarding the main metabolic process(es) responsible for source limitation.Nevertheless, Jenner et al. [75] proposed that a single, or a few, of the amino-acids necessary for grain protein synthesis may limit the global N flux to the grain.For example proline is a major component of storage proteins but represents only a small fraction of the total amino-acid pool in the plant and this can be limiting.In addition, Howarth et al. [74] showed that glutamine was the principal amino acid accumulated in the grain, and that its accumulation occurred in the first seven days of grain filling.It may therefore be hypothesised that this key amino-acid may be limiting in the context of a strong demand over a short period of time.

Conclusion
This study based on a semi-hydroponic culture in controlled conditions of wheat cv.Récital shows that post-flowering NO 3 -uptake is controlled by N status at flowering in early stages following flowering.Latter in the cycle, during the active grain development phase, NO 3 -uptake appeared to be regulated by N demand for growth, although this study did not allow to functionally demonstrate this hypothesis.TaNRT2.1 seems to play a major role in NO 3 -uptake, with expression patterns in both N4 and N10 treatments that were positively correlated with those of the main genes involved in NO 3 -reduction and assimilation.This study also shows that root [NO 3 -] could play an important role in the regulation of TaNRT2.1 expression.These findings were obtained on wheat plants at a developmental stage and with a canopy structure that are both meaningful for the underlying agronomic question asked.We believe these results form a good working basis for future research on genetic variability associated with the control of post-flowering N uptake with a long-term goal of forcing a break in the negative relationship between grain protein and grain yield.
3 -].Values lay between 522 tillers m -2 for N1, and 1371 tillers m -2 for N10, indicating a strong morphological response to rising N. Tiller number increased linearly with [NO 3 -] (r 2 = 0.99, p = 0.005; data not shown), with an average slope of 92.5 tillers m -2 per mM of [NO 3 -].Plant biomass ranged from 393 gDW m -2 for N1 to 1702 gDW m -2 for N10.Even though the biomass difference between plants at flowering from N4 to N7 was not significantly different (p = 0.168), a linear regression across the four [NO 3 -] treatments vs plant biomass did reveal a significant relationship (r 2 = 0.95, p = 0.026; data not shown).This trend shows that all treatments lay within a range where [NO 3 -] was limiting for plant growth.Plant N concentration (plant [N]) at flowering increased strongly and linearly with nutrient [NO 3 -] (r 2 = 0.99, p = 0.003; data not shown).The average slope was 0.183 percentage points of plant [N] per mM of NO 3 -.All the above trends are supported by a calculation of NNI at flowering (Table1), which shows a significant linear response to NO 3 -(r² = 0.99, p = 0.005; data not shown).The average increase was 0.22 points per mM of [NO 3 - with levels between 0.36 and 1.08 gNO 3 -m -2 day -1 depending on [NO 3 -] treatment.After flowering and regardless of [NO 3 -] treatment, NO 3

NO 3 -]
Also at flowering, NO 3 -uptake per unit DW was considerably larger for plant exposed to the low [treatments pre-flowering.Relative to N10, daily fluxes of NO 3 -uptake were 5.7-, 5and 2.4-times higher for N1, N4 and N7, respectively.Furthermore, the normalisation of NO 3 uptake per unit DW allows better comparisons of the different periods of the temporal dynamic.Indeed, the direct effect of NO 3 -uptake on plant [N] was much stronger at flowering than later during the post-flowering period (Fig.2B).Lastly, the exact stage of NO 3 -uptake recovery seems to differ between the [NO 3 -] treatments.The most obvious trend is that plants exposed to higher pre-flowering [NO 3 -], restarted uptake earlier than those exposed to lower pre-flowering [NO 3 -].

Fig 2 .
Fig 2. Plant NO 3 -uptake dynamic during post-flowering period for the four NO 3 -treatments per square meter (A) or per unit DW (B).Data were obtained by substitution of the previous nutrient solution by an identical one but labelled 10% 15 N nutrient solution 24 h before sampling.Values are the means of eight biological repetitions ± 1 standard error for N1 (open circles), N4 (grey triangles), N7 (dark grey squares) and N10 (black diamonds).doi:10.1371/journal.pone.0120291.g002

Fig 4 .
Fig 4. Relative expression patterns of TaNRT2.1 (A) and TaNRT1 (B) genes in roots from flowering (GS65) to GS65+800DD for two contrasting NO 3 -treatments (N4 and N10).Values are the means of three biological repetitions ± 1 standard error for N4 (grey triangles) and N10 (black diamonds).Quantification was performed by qRT-PCR.Relative expression values were calculated using the ΔCT method corrected for primer efficiency, using Ta54280 and Ta54948 as internal controls.doi:10.1371/journal.pone.0120291.g004

Fig 5 .
Fig 5. Relations between plant NO 3 -uptake and relative expression levels of TaNRT2.1 (A) and TaNRT1 (B) in roots from flowering (GS65) to GS65+800DD for two contrasting NO 3 -treatments (N4 and N10).Uptake data were obtained by substitution of previous nutrient solution by an identical one but labelled 10% 15 N nutrient solution 24 h before sampling.Values are means of eight biological repetitions for NO 3 -uptake and of three biological repetitions for gene expression for N4 (grey triangles) and N10 (black diamonds).Correlations are based on seven post-flowering sampling dates.Statistical analyses were by the Pearson correlation test.Gene expression quantification was performed by qRT-PCR.Relative expression values were calculated using the ΔCT method corrected for primer efficiency, using Ta54280 and Ta54948 as internal controls.doi:10.1371/journal.pone.0120291.g005

Fig 6 .
Fig 6.Relations between roots NO 3 -concentration and relative expression level of TaNRT2.1 in roots for two contrasting NO 3 -treatments (N4 and N10) from flowering (GS65) to GS65+800DD (A) and from flowering (GS65) until NO 3 -uptake stop (respectively GS65+400DD and GS65+300DD for N4 and N10) (B).Values are the three individual biological repetitions for seven post-flowering sampling dates for N4 (grey triangles) and N10 (black diamonds) (A) or respectively four and three sampling dates for N4 and N10 (B).Gene expression quantification was performed by qRT-PCR.Relative expression values were calculated using the ΔCT method corrected for primer efficiency, using Ta54280 and Ta54948 as internal controls.Statistical analyses were by the Pearson correlation test.doi:10.1371/journal.pone.0120291.g006

NO 3 - 3 -]
treatments and generated canopies of markedly contrasting structure.The level of [NO 3 -] in the nutrient solutions strongly influenced pre-flowering plant development, with increments of [NO increasing tiller density linearly by about 90 tillers m -2 mM -1 , biomass by about 140g m -2 mM -1 and [N] by 0.2 percentage points mM -1 .

Table 1 .
Tiller number, plant dry weight, plant N concentration, total plant N and NNI at flowering for the four NO 3 -treatments.Values are the means of four biological repetitions ± 1 standard error (SE).Statistically non-significantly different groups (Tukey multiple comparisons, p < 0.05) are labeled with the same lowercase letter.NNI = Nitrogen Nutrition Index.

Table 2 .
Plant dry weight, plant N concentration, grain yield, grain N concentration and harvest index at maturity for the four NO 3 -treatments.
Values are the means of four biological repetitions ± 1 standard error (SE).Statistically non-significantly different groups (Tukey multiple comparisons, p < 0.05) are labeled with the same lowercase letter.doi:10.1371/journal.pone.0120291.t002