The tissue and developmental expression patterns of maize nitrogen transport genes requires defining to help understand the underlying mechanisms that influence nitrogen uptake and redistribution in maize plants, both of which are key traits in improving nitrogen utilisation of cereal crops. In the context of nitrate transporters, ZmNPF7.10 was the only gene to have a ubiquitous expression pattern independent of the age of the plant; however, its transcript levels were found to be quite low (Fig. 1f). Two members of the NPF7 family (AtNPF7.3, AtNPF7.2) have been previously described in A. thaliana. AtNPF7.3/AtNRT1.5 encodes a low-affinity NO3− transporter located in the root pericycle cells near the xylem and is responsible for xylem loading of NO3− . AtNPF7.3/AtNRT1.5 is also linked to the tolerance to cadmium, drought and salt stress as knock-out mutant plants displayed higher resistance to these abiotic stresses . AtNPF7.2/AtNRT1.8 also encodes a low-affinity NO3− transporter  expressed within xylem parenchyma cells. AtNPF7.2/AtNRT1.8 is involved in the efflux of NO3− from xylem vessels. Together, AtNPF7.3/AtNRT1.5 and AtNPF7.2/AtNRT1.8 are believed to work in concert with each other to load and unload NO3− from the xylem, respectively. The maize NPF7 family contains a total of 12 members  but only two members were analysed in our study. ZmNPF7.12 was not expressed (data not shown) and ZmNPF7.10 was found to be ubiquitously expressed across tissues and developmental stages (Fig. 1f). Given the similar orthology between the A. thaliana and the maize genes, the function of ZmNPF7 genes in xylem loading/unloading is likely. A more detailed analysis of the NPF7 family in maize would help better understand the control of NO3− loading in the xylem.
The putative HATS transporter genes ZmNRT2.1, ZmNRT2.2 and ZmNRT3.1A displayed similar root-specific expression profiles (Fig. 1 g, h and j respectively). Moreover, their root-specific expression was higher at R1 than at the V7 stage, reaching 139-, 243- and 687- fold levels of the control genes, respectively. These data confirmed previous results on the dominant localisation of ZmNRT2.1 and ZmNRT2.2 transcripts in maize roots [16, 18, 27]. The co-localised expression of ZmNRT3.1A, ZmNRT2.1 and ZmNRT2.2 tend to confirm the function of ZmNRT3 in maize as described by Lupini et al., 2016 . The authors showed a functional interaction of ZmNRT2.1 with ZmNRT3.1A in regulating NO3− uptake along the root axis of maize. The two-component NO3− uptake system of NRT2-NRT3 has been demonstrated to be also present in other plant species such as A. thaliana [21, 36], barley  or rice . We followed the expression profile of a second NRT3 gene, ZmNRT3.1B, and detected a higher expression in old leaves than in roots (Fig. 1k). However, ZmNRT3.1B expression is negligible compared to its homologue ZmNRT3.1A, which was expressed at a 100-fold lower level (Fig. 1 j and 1 k).
ZmNRT2.5, another putative HATS gene, was found expressed in all organs at low levels at the V7 stage (Fig. 1i). At the reproductive stage, no transcript could be detected in roots or silks. Although at R1, the gene was expressed in the leaves, cobs and tassels and its expression was significantly higher in the husk leaves (Fig. 1i). ZmNRT2.5 was the only putative NO3− transporter found expressed at high levels in the husk leaves. As husk leaves play a central role in the distribution of nitrogen during the grain filling period , the role of ZmNRT2.5 in this process needs to be examined further.
The recently characterised genes ZmNPF6.4, coding a low-affinity NO3− transporter , and ZmNPF6.6, coding a high affinity NO3− transporter , displayed similar expression patterns (Fig. 1c and d respectively). Both genes were mainly expressed in the roots, although transcripts could be detected in other organs but at a significantly lower level. Both gene expression patterns were higher at the vegetative stage than the reproductive stage. We expect this anomaly was due to the reduced nitrogen uptake capacity of maize after flowering . Indeed, post-silking only 35–55% of the grain nitrogen originates from nitrogen uptake, the rest provided from the pre-existing nitrogen stored before silking in leaves and stems [6, 38]. The root expression pattern of the two genes is in line with the results found previously for ZmNPF6.4 and ZmNPF6.6, two of the main NO3− root transporters [22, 29].
ZmNPF6.8 was the only gene found with a specific expression pattern in old leaves for both vegetative and reproductive stages (Fig. 1e). In senescing leaves, programmed degradation of leaf proteins are an important source of remobilised nitrogen used to supplement growing organs, including grains or newly formed leaves [3, 38, 39]. The fact that ZmNPF6.8 encodes a putative NO3− transporter expressed in source organs makes it an important target gene to further explore leaf nitrogen remobilisation. Another avenue where ZmNPF6.8 may play an important role is in the transport of polyamines, a class of low molecular weight aliphatic polycations. In A. thaliana, the mutant line sper3–3 shows an increased tolerance to toxic levels of polyamines . The corresponding gene, AtNRT1.3/AtNPF6.4, a close orthologue to ZmNPF6.8, was found to be expressed in leaves, stems and flowers. Tong et al., 2016  concluded that the transport or metabolism of polyamines is associated with the NO3− transport activities in the parenchymal tissues of A. thaliana shoots.
ZmNPF6.2 transcripts were detected mainly in leaves and gamete-producing organs (silk and tassel) (Fig. 1a). Its leaf expression was higher at R1 stage compared to V7. ZmNPF6.2 is an orthologue of AtNPF6.2/AtNRT1.4 [13, 15]. In A. thaliana, AtNPF6.2/AtNRT1.4 encodes a NO3− LATS transporter expressed in the petiole and the adjacent part of the midrib of the leaf . This low-affinity NO3− transporter may be involved in the regulation of leaf NO3− homeostasis. Given the homology and expression pattern similarities between AtNPF6.2 and ZmNPF6.2, it is possible that this transporter carries out the same function in both the plant species. The ZmNPF6.2 homologue, ZmNPF6.3 , was also expressed in the silks and tassels as well as in the roots but its expression was around 10-fold lower than ZmNPF6.2 (Fig. 1b and a respectively).
The two putative NH4+ LATS transporter genes were constitutively expressed in our experiments. ZmAMF1.1 and ZmAMF1.2 expression was similar in all the organs independent of the growth stage (Fig. 2h and i, respectively). Moreover, both the genes had comparable levels of transcripts. The ubiquitous expression of AMF1 genes in maize was surprising. Contrary to the nitrate transporters ZmNPF6.6 and ZmNPF6.8 that were specifically expressed in the roots and old leaves, respectively, (Fig. 1d and e), ZmAMF1.1 and ZmAMF1.2 seem to be present in every organ tested (Fig. 2h and i). The function of these two genes and their respective protein activities require further investigation.
Three genes of the AMT1 family were found expressed in our experiment. ZmAMT1.1A was constitutively expressed at V7. Transcripts were also found at R1 and in all organs except the cobs where their levels were minor (Fig. 2a). Our results are in accordance with those previously described by Gu et al.  where they detected ubiquitous ZmAMT1.1A expression at the seedling and silking stages and 15 days post pollination . ZmAMT1.1A expression patterns is conserved in other plant species, including rice (OsAMT1.1) and sorghum (SbAMT1.1) [23, 42]. Contrary to ZmAMT1.1A, its closest homologue, ZmAMT1.1B showed specific expression only in the silks (Fig. 2b). This particular pattern had already been seen by Gu et al. (2013), which showed an enhanced expression of ZmAMT1.1B in the immature ear at the silking stage . A third member of the AMT1 family, ZmAMT1.3, was found specifically expressed in the roots independently of the growth stage of the plant (Fig. 2c). A similar pattern has been described at the seedling stage  . However, in the reproductive stage, ZmAMT1.3 was also expressed in the leaves, while the data for root expression is still to be determined . Our results show a high level of expression in root tissues that may indicate a role of ZmAMT1.3 in the root NH4+uptake.
The only known member of the AMT2 family in maize, ZmAMT2.1, was found expressed in all organs with some specificity to roots and tassels (Fig. 2d). Interestingly, Koegel et al.  indicated a similar expression pattern of the ZmAMT2.1 orthologue in sorghum. Indeed, the authors showed that in sorghum, SbAMT2.1 was expressed in all organs studied with a higher expression in roots and stamens. This analogous profile shows a conservation of expression patterns between species. Functional analysis is required to assess the conservation of function between ZmAMT2.1 and SbAMT2.1.
Transcripts of ZmAMT3.2 were detected in all organs but were higher in the older leaves (Fig. 2f). High expression could also be seen in the tassels although the data was variable and not conclusive. Its homologue, ZmAMT3.3, was also expressed in all organs except in the cobs (Fig. 2g). The broad expression patterns of ZmAMT3.2 and ZmAMT3.3 are similar to their close orthologues in sorghum . Transcripts of SbAMT3.2 and SbAMT3.3 were detected in roots, stems, shoots and pistils of field grown plants . SbAMT3.3 was also expressed in the stamens of sorghum. However, the function of the transporters and their involvement in the NH4+ transport has yet to be demonstrated.
Transcripts of the last member of the AMT3 family, ZmAMT3.1, were detected only in the OL at V7 and in the leaves and tassels at R1. This contrasts with the previous finding in sorghum that showed SbAMT3.1 to be expressed mainly in the roots . A detailed functional analysis of these two genes is required to resolve the dissimilarity in expression patterns between maize and sorghum. The only known member of the AMT4 family in maize, ZmAMT4, was the only AMT gene found not to be expressed in our samples (data not shown).
To provide a visual summary of indicative gene expression at the vegetative (V7) and reproductive stages (R1), gene expression relative to controls have been presented as colour indicative heat maps (Fig. 3). There is a clear definition in the expression of both nitrate and ammonium transport genes across the tissues and the two development phases of the plants. In the context of nitrogen transporter activity in reproductive tissues, ammonium transport (AMT1, AMT2 and AMF1) systems are clearly induced with expected roles in nitrogen redistribution in these important tissues. Activity of nitrate transport systems in flowering tissues (R1) are noticeably less than those of ammonium.
Response to nitrogen
Following starvation, we expect the expression of the HATS encoding genes to increase to compensate for the reduction of external nitrogen . A similar pattern has been previously described in A. thaliana [44, 45]. In hydroponically grown plants, AtNRT2.1 was demonstrated to be expressed rapidly and strongly after nitrogen starvation  peaking 24 h after the start of the experiment. The other NO3− HATS gene was found responsive in both roots and shoots. No expression of ZmNRT2.5 could be detected in our control conditions probably because of the younger age of the plants used in the starvation experiment (Fig. 4D and 5B). However, ZmNRT2.5 expression increased after starvation in both organs before decreasing after nitrogen resupply (Fig. 4D and 5B). In A. thaliana, the orthologous gene, AtNRT2.5, was also found to be induced after nitrogen starvation . The authors demonstrated that gene expression increased during the starvation period. These results highlight a conserved expression pattern of NRT2 genes between species. A deeper analysis of the NRT2 genes and their protein activities will be required to confirm if functional conservation exists between species.
ZmNRT3.1A expression followed the pattern of its putative partners, ZmNRT2.1 and ZmNRT2.2, as it increased during starvation in the roots (Fig. 4e, k). However, ZmNRT3.1A expression remained elevated after resupply. A longer period of resupply may be needed to detect a decrease in its expression. A comparable pattern of expression between NRT2 and NRT3 genes has previously been demonstrated in A. thaliana. Orsel et al. (2006) indicated that AtNAR2.1/AtNRT3.1 expression increased after 24 h of NO3− starvation in a similar fashion to AtNRT2.1 . The authors concluded that, since both proteins are required for functional HATS activity, their expression should be closely coordinated with the expression of both AtNRT2.1 and AtNAR2.1/AtNRT3.1 components. A corresponding protein association of ZmNRT2.1 and ZmNRT3.1A was recently demonstrated in maize roots . Our results are in accordance with these previous findings.
Although not significant, the expression of ZmNPF6.6 decreased after nitrogen starvation (~ 58%) but then returned to base level after nitrogen resupply (Fig. 4a, k). These results are in agreement with the recently published work by Wen et al.  where the authors described a decrease of ZmNPF6.6 expression after root starvation. This phenotype was reversible with the resupply of NO3−. Still, it is unknown whether the modulation of ZmNPF6.6 expression translate into a decrease of the transporter activity.
The only shoot specific putative NO3− transporter found responsive to changes in nitrogen supply was ZmNPF6.2. Transcripts increased (5.6-fold) after starvation and then reverted to initial levels after nitrogen resupply (Fig. 5a, h). Chiu et al. (2004) demonstrated the role of AtNPF6.2/AtNRT1.4 in NO3− storage of the petiole. The authors showed that, in the Atnrt1.4 mutant, the accumulation of NO3− in the petiole was reduced by half compared to the wild-type levels . The petiole NO3− content is commonly used as a rapid diagnostic test of the plant nitrogen status and an indicator of yield response in many crops like capsicum, cotton or potatoes [48,49,50,51,52]. Hence, AtNPF6.2/AtNRT1.4 might be an important marker of plant nitrogen status. In our experiments, ZmNPF6.2 expression responded to the availability of nitrogen to the plant which would support an involvement in the regulation of petiole NO3− content as seen in A. thaliana. The subcellular localisation of NPF6.2/NRT1.4 in both A. thaliana and maize needs to be validated.
ZmAMT1.1A transcripts were detected in most organs (Fig. 2a) however their individual responses were variable. In the roots, ZmAMT1.1A expression decreased by 46% after nitrogen starvation and slowly rose after resupply (Fig. 4f, k). On the other hand, the gene was found unresponsive to nitrogen in the shoots (Fig. 5c, h). This lack of response in the shoot might be due to a specificity of the gene to the roots. Previous studies already described the diminution by nearly half of ZmAMT1.1A transcripts following a nitrogen starvation treatment . Our data confirmed ZmAMT1.1A expression is nitrogen dependent in roots.
Although expressed in different organs, ZmAMT2.1 and ZmAMT3.2 presented a similar response to nitrogen in maize seedlings. In both roots and shoots, ZmAMT2.1 showed an upregulation of expression after nitrogen starvation (Fig. 4G, K and 5D, H). This expression returned to control levels after resupply in the shoots whereas in the roots, even after 24 h of nitrogen, the levels of ZmAMT2.1 were still similar to the starvation condition. A longer period of resupply might be necessary to see downregulation of ZmAMT2.1 in the roots. Only one gene of the AMT3 family was found to be responsive to nitrogen in both shoot and root. ZmAMT3.2 expression increased by nearly 5-fold in both roots and shoots after nitrogen starvation, although it was not significant in the shoots (Fig. 4H, K and 5E, H). After resupply, the transcript levels returned to base levels. The analogous nitrogen response of ZmAMT2.1 and ZmAMT3.2 was opposite to the pattern of ZmAMT1.1A where root expression decreased after nitrogen starvation (Fig. 4f, k). These results highlight different mechanisms in maize in response to an abiotic stress. However, a confirmation of ZmAMT2.1 and ZmAMT3.2 protein involvement in NH4+ uptake is required.
Although expressed similarly in all organs, ZmAMF1.1 and ZmAMF1.2 presented different responses to nitrogen. In both roots and shoots, ZmAMF1.1 expression increased by 1.5-fold after nitrogen starvation and decreased after resupply (Fig. 4I, K and 5F, H). ZmAMF1.2, on the other hand, did not present any significant response to nitrogen starvation but the transcript levels decreased after resupply.