Overexpression of the High-Affinity Nitrate Transporter OsNRT2.3b Driven by Different Promoters in Barley Improves Yield and Nutrient Uptake Balance

Improving nitrogen use efficiency (NUE) is very important for crops throughout the world. Rice mainly utilizes ammonium as an N source, but it also has four NRT2 genes involved in nitrate transport. The OsNRT2.3b transporter is important for maintaining cellular pH under mixed N supplies. Overexpression of this transporter driven by a ubiquitin promoter in rice greatly improved yield and NUE. This strategy for improving the NUE of crops may also be important for other cereals such as wheat and barley, which also face the challenges of nutrient uptake balance. To test this idea, we constructed transgenic barley lines overexpressing OsNRT2.3b. These transgenic barley lines overexpressing the rice transporter exhibited improved growth, yield, and NUE. We demonstrated that NRT2 family members and the partner protein HvNAR2.3 were also up-regulated by nitrate treatment (0.2 mM) in the transgenic lines. This suggests that the expression of OsNRT2.3b and other HvNRT2 family members were all up-regulated in the transgenic barley to increase the efficiency of N uptake and usage. We also compared the ubiquitin (Ubi) and a phloem-specific (RSs1) promoter-driven expression of OsNRT2.3b. The Ubi promoter failed to improve nutrient uptake balance, whereas the RSs1 promoter succeed in increasing the N, P, and Fe uptake balance. The nutrient uptake enhancement did not include Mn and Mg. Surprisingly, we found that the choice of promoter influenced the barley phenotype, not only increasing NUE and grain yield, but also improving nutrient uptake balance.


Introduction
Plants take up and use various N forms from soil, mostly the inorganic ions ammonium (NH 4 + ) and nitrate (NO 3 − ) [1][2][3]. Plant growth and development needs the presence of NO 3 − [4]. To adjust to the different concentrations of NO 3 − in the soil, plants have developed multiple nitrate uptake systems.

Phenotype of Transgenic Barley Lines during the Vegetative Growth Stage
Representative photographs of these transgenic barley lines during the vegetative growth stage are shown in Figure 2A and 2B. The Ubi-1 and Ubi-3 lines grew better than WT at the seeding stage ( Figure 2A). Correspondingly, the fresh weights of Ubi-1/3 were higher than the WT ( Figure 2C). By contrast, the RSs1-1/2/3 line had a better phenotype than the wild type in plant weight and tiller at the tillering stage ( Figure 2B). The fresh weights of the transgenic lines were also increased at the tillering stage, besides the RSs1-3 line not being marked ( Figure 2D). These results indicated that the rice nitrate transporter gene NRT2.3b can function in barley to increase growth. Moreover, the effect of the ubiquitin promoter driving strong gene expression in rapidly dividing cells is faster in the phenotype than in RSs1, which is a phloem-specific promoter (Figure 2A, B), or expressing OsNRT2.3b in the phloem of rice by in situ experiment [25,26].   (30 µg). WT: wild-type; Ubi-1/2/3 and RSs1-1/2/3 overexpression were performed with the ubiquitin promoter and phloem-specific promoter (RSs1) driven OsNRT2.3b gene, respectively, as below.

Phenotype of Transgenic Barley Lines during the Vegetative Growth Stage
Representative photographs of these transgenic barley lines during the vegetative growth stage are shown in Figure 2A,B. The Ubi-1 and Ubi-3 lines grew better than WT at the seeding stage ( Figure 2A). Correspondingly, the fresh weights of Ubi-1/3 were higher than the WT ( Figure 2C). By contrast, the RSs1-1/2/3 line had a better phenotype than the wild type in plant weight and tiller at the tillering stage ( Figure 2B). The fresh weights of the transgenic lines were also increased at the tillering stage, besides the RSs1-3 line not being marked ( Figure 2D). These results indicated that the rice nitrate transporter gene NRT2.3b can function in barley to increase growth. Moreover, the effect of the ubiquitin promoter driving strong gene expression in rapidly dividing cells is faster in the phenotype than in RSs1, which is a phloem-specific promoter (Figure 2A,B), or expressing OsNRT2.3b in the phloem of rice by in situ experiment [25,26]. the copy numbers of transgenic barley lines were identified by Southern blot ( Figure S1). The results show that Ubi-1/2/3 and RSs1-1/2 were one copy insertion lines and RSs1-3 was two copy insertion line. Each lane was loaded with an equal quantity of proteins (30 μg). WT: wild-type; Ubi-1/2/3 and RSs1-1/2/3 overexpression were performed with the ubiquitin promoter and phloem-specific promoter (RSs1) driven OsNRT2.3b gene, respectively, as below.

Phenotype of Transgenic Barley Lines during the Vegetative Growth Stage
Representative photographs of these transgenic barley lines during the vegetative growth stage are shown in Figure 2A and 2B. The Ubi-1 and Ubi-3 lines grew better than WT at the seeding stage ( Figure 2A). Correspondingly, the fresh weights of Ubi-1/3 were higher than the WT ( Figure 2C). By contrast, the RSs1-1/2/3 line had a better phenotype than the wild type in plant weight and tiller at the tillering stage ( Figure 2B). The fresh weights of the transgenic lines were also increased at the tillering stage, besides the RSs1-3 line not being marked ( Figure 2D). These results indicated that the rice nitrate transporter gene NRT2.3b can function in barley to increase growth. Moreover, the effect of the ubiquitin promoter driving strong gene expression in rapidly dividing cells is faster in the phenotype than in RSs1, which is a phloem-specific promoter (Figure 2A, B), or expressing OsNRT2.3b in the phloem of rice by in situ experiment [25,26].

Characterization of Transgenic Barley during the Reproductive Growth Stage
The phenotypes of all the barley lines were not significant different at the grain-filling stage. Grain-related indicators are the basic criteria for measuring crop yields. Thus, we measured the grain weight, N, and metal concentration in the seeds of all the barley lines. We found that the grain weight of the transgenic barley lines was increased relative to WT ( Figure 3A). Moreover, the Ubi-1/2/3 and RSs1-1/2/3 lines were approximately 60% and 40% higher than those of the WT ( Figure 3A). NUE in the transformants was higher than WT (ranging from 27% to 43%) ( Figure 3C). The total N content of Ubi-1/2/3 was increased by 33%, whereas the total N of RSs1-1/2/3 lines was only increased 19% and exhibited no statistically significant difference compared to WT ( Figure 3B). Furthermore, the Fe concentration in seeds was only increased in the Ubi-1/2/3 lines ( Figure 3D). Additionally, we analyzed other elements in the seeds of different lines. The Ubi-1/2/3 lines exhibited no significant differences in manganese ( Figure S2A) and magnesium ( Figure S2B) concentrations relative to WT. This result was similar to the overexpression of OsNRT2.3b with Ubi promoter in rice seeds [36], but the manganese concentrations in RSs1-1/2/3 seeds were significantly lower than those in the WT ( Figure S2A).

Characterization of Transgenic Barley during the Reproductive Growth Stage
The phenotypes of all the barley lines were not significant different at the grain-filling stage. Grain-related indicators are the basic criteria for measuring crop yields. Thus, we measured the grain weight, N, and metal concentration in the seeds of all the barley lines. We found that the grain weight of the transgenic barley lines was increased relative to WT ( Figure 3A). Moreover, the Ubi-1/2/3 and RSs1-1/2/3 lines were approximately 60% and 40% higher than those of the WT ( Figure 3A). NUE in the transformants was higher than WT (ranging from 27% to 43%) ( Figure 3C). The total N content of Ubi-1/2/3 was increased by 33%, whereas the total N of RSs1-1/2/3 lines was only increased 19% and exhibited no statistically significant difference compared to WT ( Figure 3B). Furthermore, the Fe concentration in seeds was only increased in the Ubi-1/2/3 lines ( Figure 3D). Additionally, we analyzed other elements in the seeds of different lines. The Ubi-1/2/3 lines exhibited no significant differences in manganese ( Figure S2A) and magnesium ( Figure S2B) concentrations relative to WT. This result was similar to the overexpression of OsNRT2.3b with Ubi promoter in rice seeds [36], but the manganese concentrations in RSs1-1/2/3 seeds were significantly lower than those in the WT ( Figure S2A). Seed size is an important agronomic trait; thus, we assessed seed morphology. The seeds of transgenic barley were larger than those of WT ( Figure 4A, B). The seed lengths of Ubi-1/2/3 and RSs1-1/2/3 lines were increased by 9% and 10%, respectively ( Figure 4A), and the seed widths of the Ubi-1/2/3 and RSs1-1/2/3 lines were increased by 24% and 27% compared with that of the WT ( Figure  4B). Seed size is an important agronomic trait; thus, we assessed seed morphology. The seeds of transgenic barley were larger than those of WT ( Figure 4A, B). The seed lengths of Ubi-1/2/3 and RSs1-1/2/3 lines were increased by 9% and 10%, respectively ( Figure 4A), and the seed widths of the Ubi-1/2/3 and RSs1-1/2/3 lines were increased by 24% and 27% compared with that of the WT ( Figure 4B). These results indicate that a high expression of OsNRT2.3b with a pH-sensing motif in barley enhances the seed length and width to increase grain weight. Meanwhile, the ubiquitin promoterdriven expression of OsNRT2.3b in barley resulted in an improved Fe concentration in seeds when compared with the phloem-specific expression of OsNRT2.3b.

Plant Development and Metabolism at Maturity
Physiological indexes are used to measure plant development and metabolism, and plant shoot growth is the basis for normal plant ontogenesis and crop yield. The transgenic lines exhibited an increased tiller number per plant ( Figure 5A) as well as increased shoot dry weight ( Figure 5B). The Ubi-1/2/3 line exhibited an increased nitrate concentration in shoot by 23%, and the RSs1-1/2/3 lines showed an increasing trend, but it was not statistically significant compared with WT ( Figure 5C). The stems of transgenic lines show a trend of increased total N content but it was not significantly different ( Figure 5D), which was contrary to the increased total N of seeds in transgenic lines ( Figure  3B).  These results indicate that a high expression of OsNRT2.3b with a pH-sensing motif in barley enhances the seed length and width to increase grain weight. Meanwhile, the ubiquitin promoter-driven expression of OsNRT2.3b in barley resulted in an improved Fe concentration in seeds when compared with the phloem-specific expression of OsNRT2.3b.

Plant Development and Metabolism at Maturity
Physiological indexes are used to measure plant development and metabolism, and plant shoot growth is the basis for normal plant ontogenesis and crop yield. The transgenic lines exhibited an increased tiller number per plant ( Figure 5A) as well as increased shoot dry weight ( Figure 5B). The Ubi-1/2/3 line exhibited an increased nitrate concentration in shoot by 23%, and the RSs1-1/2/3 lines showed an increasing trend, but it was not statistically significant compared with WT ( Figure 5C). The stems of transgenic lines show a trend of increased total N content but it was not significantly different ( Figure 5D), which was contrary to the increased total N of seeds in transgenic lines ( Figure 3B). These results indicate that a high expression of OsNRT2.3b with a pH-sensing motif in barley enhances the seed length and width to increase grain weight. Meanwhile, the ubiquitin promoterdriven expression of OsNRT2.3b in barley resulted in an improved Fe concentration in seeds when compared with the phloem-specific expression of OsNRT2.3b.

Plant Development and Metabolism at Maturity
Physiological indexes are used to measure plant development and metabolism, and plant shoot growth is the basis for normal plant ontogenesis and crop yield. The transgenic lines exhibited an increased tiller number per plant ( Figure 5A) as well as increased shoot dry weight ( Figure 5B). The Ubi-1/2/3 line exhibited an increased nitrate concentration in shoot by 23%, and the RSs1-1/2/3 lines showed an increasing trend, but it was not statistically significant compared with WT ( Figure 5C). The stems of transgenic lines show a trend of increased total N content but it was not significantly different ( Figure 5D), which was contrary to the increased total N of seeds in transgenic lines ( Figure  3B).   supply, the transgenic barley lines exhibited no significant differences compared to WT ( Figure 6A). In contrast, the uptake of 15 N-NO 3 − was higher influx in transgenic lines after five minutes. The nitrate influx rates of the Ubi-1/2/3 and RSs1-1/2/3 lines were higher than that of the WT (ranging from 38% to 53%) ( Figure 6B). These results indicated that the overexpression of the high-affinity nitrate transporter OsNRT2.3b improved NO 3 − uptake ( Figure 6B) into the plant ( Figure 5C), which resulted in an increased total N content in seeds ( Figure 3B) to further enhance the grain yield. Taken together, OsNRT2.3b as a rice nitrate transporter can improve nitrate acquisition and transport in barley, which is a favored nitrate crop. This result is consistent with the overexpression of OsNRT2.3b in rice. [25].

NH4 + and NO3 -Influx Rates in WT and OsNRT2.3b Transgenic Barley
The influence of OsNRT2.3b overexpression on the 15 N-NH4 + and 15 N-NO3 -influx of hydroponically grown barley from root into intact plant was determined. Under 0.2 mM 15 NH4 + supply, the transgenic barley lines exhibited no significant differences compared to WT ( Figure 6A). In contrast, the uptake of 15 N-NO3 -was higher influx in transgenic lines after five minutes. The nitrate influx rates of the Ubi-1/2/3 and RSs1-1/2/3 lines were higher than that of the WT (ranging from 38% to 53%) ( Figure 6B). These results indicated that the overexpression of the high-affinity nitrate transporter OsNRT2.3b improved NO3 -uptake ( Figure 6B) into the plant ( Figure 5C), which resulted in an increased total N content in seeds ( Figure 3B) to further enhance the grain yield.

The Effect of Different N Treatments on Gene Expression, and Total N, P, and Other Fe Concentration in Different Plant Parts
We investigated the effect of OsNRT2.3b overexpression on barley growth under different N treatments (0.2 mM NH4 + and 0.2 mM NO3 -). NH4 + and NO3 -treatments up-regulated the expression of OsNRT2.3b, which was mainly expressed in leaves and stem sheath ( Figure 7A, D). The expression of OsNRT2.3b in leaves of the Ubi-1/2/3 lines was increased 1.6-fold and 3.8-fold, compared to the RSs1-1/2/3 lines treated with 0.2 mM NH4 + and 0.2 mM NO3 -, respectively ( Figure 7A, D). Interestingly, the expression of OsNRT2.3b in all lines treated with 0.2 mM NO3 -was increased by more than 10-fold compared with 0.2 mM NH4 + treatment ( Figure 7A, D). NH4 + treatment did not affect the dry weight of the different plant parts ( Figure 7B) or the total N concentration in any of the lines ( Figure 7C). In contrast, NO3 -treatment increased the dry weight of leaves and sheaths in the transgenic lines (especially the Ubi-1/2/3 line), whereas the roots remained unaffected ( Figure 7E). NO3 -treatment exhibited a similar effect on the total N concentration ( Figure 7F). These effects of NO3 -treatment on dry weight and total N concentration corresponded with the expression of OsNRT2.3b under 0.2 mM NO3 -treatment ( Figure 7D).   Figure S4A, B). The barley NRT2 gene family requires a partner protein gene HvNAR2.3 to transport nitrate [37]. Interestingly, HvNRT2.1/2.2/2.3 and HvNAR2.3 were not up-regulated in barley roots ( Figure S4C). The elemental concentrations tested in different parts of all the barley lines showed no significant differences with the 0.2 mM NH4 + and NO3 -treatments ( Figure S5). From statistical analysis, NH4 + treatment did not affect the total N, P, or Fe in shoots and roots of any barley line (Table S3). However, the shoots of transgenic barley lines exhibited a higher total N than WT under 0.2 mM NO3 -treatment. On the other hand, RSs1-1/2/3 exhibited the highest total P and Fe in shoots (Table 1). In contrast, the amounts of total N, P, and Fe exhibited an opposite trend in roots (Table 1). These results suggest that more nutrients were transported to the shoots from the roots for plant growth in the transgenic lines.   Figure S4A,B). The barley NRT2 gene family requires a partner protein gene HvNAR2.3 to transport nitrate [37]. Interestingly, HvNRT2.1/2.2/2.3 and HvNAR2.3 were not up-regulated in barley roots ( Figure S4C). The elemental concentrations tested in different parts of all the barley lines showed no significant differences with the 0.2 mM NH 4 + and NO 3 − treatments ( Figure S5). From statistical analysis, NH 4 + treatment did not affect the total N, P, or Fe in shoots and roots of any barley line (Table S3). However, the shoots of transgenic barley lines exhibited a higher total N than WT under 0.2 mM NO 3 − treatment. On the other hand, RSs1-1/2/3 exhibited the highest total P and Fe in shoots (Table 1). In contrast, the amounts of total N, P, and Fe exhibited an opposite trend in roots (Table 1). These results suggest that more nutrients were transported to the shoots from the roots for plant growth in the transgenic lines.  Barley has seven candidate members of the NRT2 family (HvNRT2.1-2.7) and three partner proteins HvNAR2 [38]. Interestingly, OsNRT2.3b has a high sequence similarity to HvNRT2.5 in barley ( Figure S3). Furthermore, HvNRT2.5 has a pH-sensing motif similar to that identified in OsNRT2.3b [25] ( Figure S3B). The relative expression of HvNRT2.5 in transgenic barley roots was up-regulated under 0.2 mM NO 3 − condition, compared with the wild types ( Figure 8A). However, the relative expression of HvNRT2.5 in leaves and sheaths was not significantly different in all barley lines under 0.2 mM NO 3 − supply ( Figure 8A). HvNRT2.5 transcript was clearly up-regulated in the roots under NO 3 − supply ( Figure 8B). Simultaneously, we found that the HvNRT2.5 needs a partner protein HvNAR2.3 to transport nitrate under 0.25 mM NO 3 − , pH 5.5 supply when expressed in oocytes ( Figure 9A). The co-injection of HvNRT2.5 and HvNAR2.3 in oocytes did not significantly increase nitrate transport relative to water-injected controls under 10 mM NO 3 − , pH 5.5 treatment ( Figure 9B).
These data indicated that HvNRT2.5 belongs to a high-affinity nitrate transporter and it needs a partner protein HvNAR2.3 to transfer nitrate in oocytes. In further hydroponic experiments, barley was grown in 10 mM NH4 + /NO3 -. A 10 mM NH4 + and NO3 -supply resulted in an increased expression of OsNRT2.3b in the leaves of transgenic lines and a suppressed expression in sheaths and roots ( Figure S6A, D). The dry weights of the different plant parts ( Figure S6B, E) and the total N concentration ( Figure S6C, F) were not affected by 10 mM NH4 + and NO3 -. These results indicate that high N concentrations suppress the expression of the highaffinity nitrate transporter OsNRT2.3b to decrease N uptake and transfer.

The Characteristics of HvNRT2.5, a Gene Homologous to OsNRT2.3b
Barley has seven candidate members of the NRT2 family (HvNRT2.1-2.7) and three partner proteins HvNAR2 [38]. Interestingly, OsNRT2.3b has a high sequence similarity to HvNRT2.5 in barley ( Figure S3). Furthermore, HvNRT2.5 has a pH-sensing motif similar to that identified in OsNRT2.3b [25] ( Figure S3B). The relative expression of HvNRT2.5 in transgenic barley roots was upregulated under 0.2 mM NO3 -condition, compared with the wild types ( Figure 8A). However, the relative expression of HvNRT2.5 in leaves and sheaths was not significantly different in all barley lines under 0.2 mM NO3 -supply ( Figure 8A). HvNRT2.5 transcript was clearly up-regulated in the roots under NO3 -supply ( Figure 8B). Simultaneously, we found that the HvNRT2.5 needs a partner protein HvNAR2.3 to transport nitrate under 0.25 mM NO3 -, pH 5.5 supply when expressed in oocytes ( Figure 9A). The co-injection of HvNRT2.5 and HvNAR2.3 in oocytes did not significantly increase nitrate transport relative to water-injected controls under 10 mM NO3 -, pH 5.5 treatment ( Figure 9B). These data indicated that HvNRT2.5 belongs to a high-affinity nitrate transporter and it needs a partner protein HvNAR2.3 to transfer nitrate in oocytes.   HvNAR2.1, HvNAR2.2, HvNAR2.3, and CHL1 activity in nitrate uptake in Xenopus oocytes. Uptake of 15 NO3 − into oocytes injected with water, single cRNA, or mixtures as indicated. Oocytes injected In further hydroponic experiments, barley was grown in 10 mM NH4 + /NO3 -. A 10 mM NH4 + and NO3 -supply resulted in an increased expression of OsNRT2.3b in the leaves of transgenic lines and a suppressed expression in sheaths and roots ( Figure S6A, D). The dry weights of the different plant parts ( Figure S6B, E) and the total N concentration ( Figure S6C, F) were not affected by 10 mM NH4 + and NO3 -. These results indicate that high N concentrations suppress the expression of the highaffinity nitrate transporter OsNRT2.3b to decrease N uptake and transfer.

The Characteristics of HvNRT2.5, a Gene Homologous to OsNRT2.3b
Barley has seven candidate members of the NRT2 family (HvNRT2.1-2.7) and three partner proteins HvNAR2 [38]. Interestingly, OsNRT2.3b has a high sequence similarity to HvNRT2.5 in barley ( Figure S3). Furthermore, HvNRT2.5 has a pH-sensing motif similar to that identified in OsNRT2.3b [25] ( Figure S3B). The relative expression of HvNRT2.5 in transgenic barley roots was upregulated under 0.2 mM NO3 -condition, compared with the wild types ( Figure 8A). However, the relative expression of HvNRT2.5 in leaves and sheaths was not significantly different in all barley lines under 0.2 mM NO3 -supply ( Figure 8A). HvNRT2.5 transcript was clearly up-regulated in the roots under NO3 -supply ( Figure 8B). Simultaneously, we found that the HvNRT2.5 needs a partner protein HvNAR2.3 to transport nitrate under 0.25 mM NO3 -, pH 5.5 supply when expressed in oocytes ( Figure 9A). The co-injection of HvNRT2.5 and HvNAR2.3 in oocytes did not significantly increase nitrate transport relative to water-injected controls under 10 mM NO3 -, pH 5.5 treatment ( Figure 9B). These data indicated that HvNRT2.5 belongs to a high-affinity nitrate transporter and it needs a partner protein HvNAR2.3 to transfer nitrate in oocytes.   HvNAR2.1, HvNAR2.2, HvNAR2.3, and CHL1 activity in nitrate uptake in Xenopus oocytes. Uptake of 15 NO3 − into oocytes injected with water, single cRNA, or mixtures as indicated. Oocytes injected These data suggest that a low concentration of nitrate can increase OsNRT2.3b gene expression in barley more than a low ammonium treatment (Table S4, see Figure 2 in [36]), which in turn increases the transport of nutrients, and ultimately increases the grain yield in the transgenics. This may also be due to the up-regulation of HvNRT2.5 expression in the transgenic barley root, while increasing the nitrate uptake and overexpression of OsNRT2.3b in transgenic lines shoot further facilitates nitrate transport.

Discussion
Plant uptake of NO 3 − is carefully regulated depending on the N supply forms and N status of the plant [39]. In rice, it has been reported that the OsNRT2 gene family has key functions in the uptake of NO 3 − from soil and transport from root to shoot [10,25,26,40,41]. Normally, the OsNRT2 family (except OsNRT2.3b) requires the partner protein OsNAR2.1 for transporting nitrate [30]. OsNRT2.3b is one of the two transcripts with variation in expression and contains a pH-sensing motif that can regulate the transport activity under various N supply forms and improve NUE and grain yield [10]. However, it is not known whether the OsNRT2.3b transporter can exert a nitrate transport function in barley. In this study, the main objective was to demonstrate whether the overexpression of OsNRT2.3b driven by the strong promoter (ubiquitin) and the phloem-specific promoter (RSs1) in barley can enhance the transport of nutrients as well as improve NUE and grain yield. We found that the increased biomass phenotype appeared earlier in the strong promoter (ubiquitin) OsNRT2.3b lines when compared with the phloem-specific promoter (RSs1) barley lines ( Figure 2). The fresh weights of Ubi and RSs1 barley lines were higher than those of the WT (ranging from 28% to 77%) ( Figure 2C,D). Correspondingly, the expression of OsNRT2.3b at the RNA and protein level in transgenic ubiquitin-driven lines was higher than that in the phloem-specific promoter lines (Figure 1). The short-term 15 N-NO 3 − influx rate was increased in the lines with an overexpression of OsNRT2.3b, showing that increased levels of the rice transporter in barley can increase nitrate uptake compared with WT under an external 0.2 mM NO 3 − supply ( Figure 6). More nitrate was transferred from the root to the shoot in the transgenic barley. Therefore, the measured nitrate concentrations in shoots in transgenic barley were higher than those in WT ( Figure 5C). These data demonstrated that OsNRT2.3b expressed in barley is directly involved in nitrate uptake and transport, which also leads to change in phenotypes, seed morphology, and biomass in transgenic lines ( Figure 2, Figure 3, Figure 4). However, the phenotypes of transgenic lines exhibited no visible significant differences during the grain-filling period ( Figure 3A). One explanation is that when a critical plant size is established, the overexpression of OsNRT2.3b no longer influenced plant biomass, emphasizing the importance of N supply for establishing the vegetative phase of growth. Another explanation is that the phloem-specific overexpression of OsNRT2.3b just enriched sieve-tube nitrate, and the optimizing pH homeostasis function in phloem cells could increase the long-distance transport of total P and Fe to accumulate more biomass and grain yield during the later growth stages ( Table 1). The N content in the two transgenic lines is similar, and there was a comparable P/Fe content change ( Table 1). As for the reason why N content was similar in two types of transgenic plants, we think that even though the Ubi promoter was driving the expression in all the types of cells, the phloem-specific expression contributed a stronger functional influence on N transport in the plants. Indeed, the overexpression of OsNRT2.3b in rice can maintain a relatively low phloem sap pH, which enhanced the accumulation of total P and Fe in leaves [25]. Therefore, the overexpression of OsNRT2.3b driven by the phloem (RSs1) promoter can achieve a similar phenotypic effect to that provided by the strong expression of the Ubi promoter. Nevertheless, the detailed molecular mechanism remains to be revealed. The OsNRT2.3 gene had two transcripts: OsNRT2.3a and b. In rice, the expression of OsNRT2.3a occurred in roots and was induced by nitrate treatment, whereas the expression of OsNRT2.3b was relatively weak in roots, but it was abundant in leaves [25,42]. The expression of OsNRT2.3b is generally very low in WT rice [25]. Some HvNRT2 family members require the partner protein HvNAR2.3 to transport nitrate in barley, but not HvNAR2.1 [37]. Some papers have reported that the HvNRT2.1, HvNRT2.2, and HvNRT2.3 genes belong to the high-affinity nitrate transporter family [43,44]. The oocyte data shows that HvNRT2.5 also belongs to the high-affinity nitrate transporter family ( Figure 9A,B).
Interestingly, the expression pattern of OsNRT2.3b was affected by the external N supply forms ( Figure 7A,D). One explanation is that a lower expression of OsNRT2.3b in NH 4 + plants compared to NO 3 − plants, as shown in Figure 7, may result from another level of regulation by a micro RNA [45].
Additionally, the expression of OsNRT2.3b was increased in Ubi transgenic barley more than in the RSs1 transgenic barley lines under 0.2 mM NO 3 − supply (Table S4). The high-affinity nitrate transporter OsNRT2.3b was strongly expressed in the leaves of transgenic barley under high NH 4 + /NO 3 − supply and weakly expression in the sheath and root ( Figure S6) [46]. This may be due to feedback regulation from the external N form and concentration. The expression of other HvNRT2s in transgenic lines was altered only by 0.2 mM NO 3 − supply, but not NH 4 + ( Figure S4 and Figure 8) The up-regulation of HvNRT2.5 in the transgenic lines roots improved nitrate uptake, and the overexpression of OsNRT2.3b promoted more nitrate transport to the shoots in the transgenic lines. Transgenic technology has been successfully used to up-regulate the expression of OsNRT2.3b in barley. The characterization of these transgenic barley lines has demonstrated that this rice high-affinity nitrate transporter has a specific role in nitrate uptake and transport from root to shoot in barley.

Plant Materials
We amplified the OsNRT2.3b (AK072215) open reading frame (ORF) from cDNA, which was isolated from Oryza sativa L.ssp. Japonica cv. Nipponbare using the primers listed in Table S1. Then, OsNRT2.3b was inserted into the pG3 vector by the AscI and KpnI enzyme sites without GUS(β-glucuronidase), which was driven by the RSs1 promoter [47] and transferred into the plant expression vector pB211. OsNRT2.3b was also cloned into the pMD19-T vector (Takara Biotechnology company, Dalian City, Liaoning Province, China) and expression vector pTCK303 with a ubiquitin promoter. The positive vector was identified by restriction digest and DNA sequencing. Next, the binary vectors pUbiquitin-OsNRT2.3b and pRSs1-OsNRT2.3b were introduced into the Agrobacterium tumefaciens strain AGL1, which was used to transform the immature embryos of the spring barley (H. vulgare L.) cultivar 'Golden Promise' by the method of Bartlett et al. 2008 [48] and Harwood et al., 2009 [49]. The T0-generation transgenic plants were identified by selection on 50 mg/L hygromycin. Transgenic T1 plants could be identified from null segregates by germination of the seed on hygromycin (100 mg/L) containing agar. Three independent T2 generation transgenic barley lines containing each construct were used for further analysis.

Growth Conditions
Barley plants (Hordeum vulgare L. cv Golden Promise) after hygromycin selection were planted in a greenhouse with artificial illumination with 16/8 h light/darkness at 23 • C and 18 • C, respectively. Light had a photon flux density at plant level equal to 300 µmol m −2 s −1 and 60% relative humidity.
For pot experiments, 10 seeds of each transgenic line were grown in pots with barley mix compost, which included perlite and grit [48].
In the hydroponic experiments, healthy barley seeds were selected, surface sterilized with 30% NaClO for 30 min, rinsed with distilled water 3 times, and then soaked for 6 h at 25 • C. Then, the seeds were germinated on moistened filter papers in dark germination boxes and placed into a growth chamber (23/18 • C, day and night). Uniformly sized 10-day-old seedlings of all the transgenic barley lines were transplanted into plastic pots for hydroponic culture in greenhouse. These seedlings were grown in 1/4 Hoagland's nutrient solution, which was applied in the following proportions: 2 mM KNO 3 , 1 mM NH 4  In addition, the barley roots were aerated in the hydroponic experiment using aquarium air pumps.

Western Blotting
The biosynthesis and purification of anti-OsNRT2.3b rabbit monoclonal antibody were as previously described [25,26]. All tissues that were from the stems of all transgenic barley lines were homogenized and lysed in buffer containing 1% Nonidet P-40 and the protease inhibitors. Then, the lysates were clarified by centrifugation, and the protein concentration was measured by spectrophotometry with Bradford reagent in A595. Fifty micrograms of protein was boiled in gel-loading buffer and analyzed on 10% SDS-PAGE gels. After that, the protein was transferred to polyvinylidene difluoride membranes and hybridized by first antibody OsNRT2.3b (1:500) and Actin (1:5000) overnight at 4 • C. The membrane was incubated in an appropriate second antibody (1:5000), followed by chemiluminescence detection to check the protein brands.

Southern Blot Analysis
Transgene copy numbers were identified by Southern blot analysis. First, genomic DNA were extracted from the leaves of WT and transgenic barley lines. Then, DNA was digested by Hind Шand EcoR I. Second, the digested DNA was separated on 1% (w/v) agarose gels, transferred to a Hybond-N + nylon membrane, and hybridized using the hygromycin-resistant gene.

Analysis of Agronomic Traits
The agronomic traits were measured for all the transgenic barley at the seeding stage, anthesis stage, and grain-filling stage. The seed morphology of transgenic barley lines was also determined. Plant fresh weight, seed weight, tillering number, and seed length/width were measured. Detailed methods for measurement of these agronomic traits were described previously [53].

Measurements of Dry Weight, Nitrate, Total N, and Metal Ion Accumulation
WT untransformed and transgenic barley lines were harvested at the mature stage (n = 4) and dried at 105 • C for 30 min. Then, shoots were dried for 3 days at 75 • C. Barley from hydroponic experiments was divided into root, stem sheath, and leaves. Dry weight was measured as biomass. The total N accumulation was measured by using the Kjeldahl method [43] and was assessed in different plant parts by multiplying the N concentration by the corresponding biomass. The calculating method of NUE (NUE, g/g) was described in Chen et al. 2016 [46]. Dried samples were digested in concentrated HNO 3 at 120 • C until no nitrogen oxide gas was released. These samples were further digested with HClO 4 at 180 • C, at which point they become transparent. Then, samples were diluted with ultrapure water, and the concentration of metal elements in the digestion solutions were analyzed by using ICP-OES (Inductively Coupled Plasma Optical Emission Spectrometer) (iCAP6300).

Statistical Analysis
All data were analyzed using one-way analysis of variance (ANOVA). Different letters showed a significant difference between transgenic barley line and wild types. Statistically significantly differences at the p < 0.05 level (one-way ANOVA) were performed using IBM SPSS Statistics version 20 software (SPSS Inc., Chicago, IL, USA).

Conclusions
In conclusion, the present study has demonstrated that the overexpression of OsNRT2.3b in barley affects the transport of N, P, and Fe. Even when the promoters were different, the N/P/Fe contents were similar, showing analogous phenotypes in the transgenic plants. This finding may be due to N/P/Fe synergism in the plant, which was driven by the N use and pH balance in these transgenic plants. This synergistic relationship between nutrients is worthy of further investigation. We have demonstrated a novel mechanism to optimize the gene expression in plants in which a targeted promoter and not a strong promoter is used to drive gene expression. The overexpression of OsNRT2.3b driven by two types of promoter in barley affected the potential to enhance the transport of nutrients and improve NUE. Furthermore, the transgenic data suggest that HvNRT2.5 has an important role in NUE in barley.
Supplementary Materials: Supplementary materials can be found at http://www.mdpi.com/1422-0067/21/4/1320/ s1. Table S1. Primers used to amplify the OsNRT2.3b open reading frame and identification primers of hygromycin. Table S2. Primers used for quantitative real-time polymerase chain reaction. Table S3. The effect of 0.2 mM NH 4 + treatment on the distribution of total N, P, and Fe in shoots and roots. Table S4. Comparison of the expression of OsNRT2.3b in transgenic barley driven by different promoters under 0.2mM NH 4 + /NO 3 − treatment. Figure S1. Identification of transgenic barley lines by Southern blot. Figure S2. The seed concentrations of Mn and Mg in different barley lines. Figure S3. Homology analysis of NRT2s genes. Figure S4. The relative expression of HvNRT2s in different plants parts under 0.2mM NH 4 + and 0.2mM NO 3 − treatments. Figure S5. Phosphorus (P) and iron (Fe) concentrations in different plant parts under 0.2 mM NH 4 + and 0.2 mM NO 3 − treatments. Figure