AMT2 sequences were searched in a sugarcane BAC library from the commercial cultivar ‘R570’ (Tomkins et al., 1999). Analyses of ScAMT2;1 expression in sugarcane organs upon various N sources and levels were performed using the commercial cultivar SP80-3280. The S. cerevisiae mutant 31019b (triple mepΔ: mep1, mep2::LEU2, mep3::KanMX2, ura3) (Marini et al., 1997) defective for ammonium uptake was complemented with AtAMT1;1 or ScAMT2;1. The Arabidopsis genotype Columbia-0 (Col-0), the quadruple AMT-knockout mutant qko (amt1;1, amt1;2, amt1;3, and amt2;1) (Yuan et al., 2007), and the respective complemented lines were used in complementation assays.

Sugarcane AMTs were sought in a BAC library that consists of 269 plates with 384 clones each in a total of 103,296 clones representing a 4.5X coverage of the sugarcane genome (Tomkins et al., 1999). The search was performed by real-time PCR amplification of the three-dimensional pool of clones (de Setta et al., 2014). For that, A. thaliana and Oryza sativa AMT2;1 were used to find orthologue sequences in the sugarcane expressed sequence tag (SUCEST) database (https://sucest-fun.org/) to design the primers ( Supplementary Table S1 ). First, superpools were screened for positive blocks, and positive blocks were further screened for the specific coordinates of positive clones, which were then isolated for confirmation and sequenced using the 454/Roche sequencing platform, assembled, and automated annotated as previously described (de Setta et al., 2014).

AMT gene automated annotation was curated using Artemis Genome Browser and Annotation Tool (v. 16.0.11) (Rutherford et al., 2000), and sorghum AMT2 was used as a reference. ScAMTs were aligned with AMTs from maize, rice, sorghum, and S. spontaneum by ClustalW (Thompson et al., 2003), including a sugarcane (‘SP80-3280’) AMT2;1 root-expressed sequence, identified here as ‘comp105883’ (NCBI id# OM966894). The evolutionary history was inferred by using the Maximum Likelihood method and JTT matrix-based model using MEGA11 (Tamura et al., 2021). A discrete Gamma distribution was used to model evolutionary rate differences among sites [5 categories (+G, parameter = 1,1177)]. This analysis involved a total of 537 positions in the final amino acid alignment. A physical map of genomic sequences (~100 kb) containing AMT2;1 from sugarcane (Saccharum spp. ‘R570’) BAC clones (032_A12, 038_G02, 118_C18, 216_D16, and 235_F05) and S. bicolor (chromosome 9; NC_012878) was manually generated.

ScAMT2;1 identified in the BAC clones were analyzed to select the sequence to be functionally characterized. Regulatory upstream (~ 3 kb from the start codon) and coding regions were aligned and compared by ClustalW using BioEdit (Hall, 1999). Conceptually translated amino acid sequences were analyzed for specific elements/domains of the MEP/AMT/Rh transporter superfamily using Prosite (Hulo et al., 2006), TMHMM (Krogh et al., 2001), and WebLogo (Crooks et al., 2004). The presence of transposable elements (TEs) in the ScAMT2;1 regulatory region was predicted by Censor (Kohany et al., 2006) using the Viridiplantae database, and the co-occurrence of transcription factor-binding sites (TFBSs) was analyzed by PlantPAN 2.0 (Chow et al., 2016).

‘SP80-3280’ plantlets derived from in vitro meristem culture were grown hydroponically in 5 L plastic pots with full-strength nutrient solution (Hoagland and Arnon, 1950) containing 1 mM NH₄NO₃ (pH adjusted to 5.8) under greenhouse conditions for three months. The nutrient solution was aerated and renewed weekly. Prior to treatment, plants received a nutrient solution containing 2 mM NH₄NO₃ for 2 d. Subsequently, the plants were subjected to either an N-free nutrient solution (-N), 2 mM NH₄NO₃ (+N), 4 mM KNO₃ ( NO 3 - ), 4 mM NH₄Cl ( NH 4 + ), or 5 mM NH₄NO₃ (high N) for 14 d. Roots, culms, and young (+1) and mature (+3) leaves were collected, frozen in liquid N and stored at -80°C. Three plants per treatment were used for ScAMT2;1 tissue-specific expression.

Arabidopsis seeds were surface sterilized and grown for 30 d in substrate and vermiculite (1:1) in a growth chamber at 22°C, 80% humidity, and a 16/8 h light/dark phase at 200 µmol m^-2 s^-1. For the selection of transgenic events and experiments in agar plates, seeds were sown onto modified half-strength MS with 1 mM NH₄NO₃ as the sole N source, with the pH adjusted to 5.8. After a 4 d vernalization at 4°C in the dark, plates were placed in a growth cabinet at 24°C, 16/8 h light/dark phases, and 100 μmol m^-2 s^-1. For experiments in agar plates, Arabidopsis seeds were kept for 3 d in half-strength MS medium with 5 mM KNO₃, with plates positioned vertically. Seedlings were then transferred onto media supplemented with various N sources at the indicated concentrations under the same environmental conditions. Treatments included either 0.5 mM KNO₃ or 2 mM NH₄Cl for experiments with plants bearing p35S::ScAMT2;1 and 2 mM KNO₃ or 0.2, 2, and 4 mM NH₄Cl for experiments with qko+p2ScAMT2;1::ScAMT2;1 plants (sugarcane endogenous promoter). After 14 d of treatment, seedlings were harvested, and the dry or fresh weight was measured.

Total RNA was isolated from sugarcane leaves as described (Leal et al., 2007) or from Arabidopsis using TRIzol (Thermo Fisher Scientific; Waltham, MS, USA). cDNA was synthesized using SuperScript III Reverse Transcriptase (Thermo Fisher Scientific). Primers were designed based on the ScAMT2;1 sequence from clone BAC 118_C18 ( Supplementary Table S1 ). RT–qPCR was performed with 5 µL of KAPA SYBR FAST (Kapa Biosystems, Wilmington, MA, USA), 0.2 µM of each primer ( Supplementary Table S1 ), and 1 µL of diluted cDNA (1:10) in a final volume of 10 µL. Reactions were run in a RotorGene-6000 (Qiagen) with the following settings: 50°C for 10 min and 95°C for 2 min, followed by 40 cycles of 95°C for 20 s, 62°C for 25 s, and 72°C for 25 s. UBIQUITIN2 was used as a reference gene in sugarcane (ScUBQ2) and Arabidopsis (AtUBQ2) ( Supplementary Table S1 ). All reactions were performed in triplicate with three biological replicates. Relative expression levels were calculated as described (Livak and Schmittgen, 2001). Normalization is indicated for each experiment.

The full-length ScAMT2;1 coding sequence from clone BAC 118_C18 was synthesized (Biomatik; Cambridge, Ontario, Canada); AtAMT1;1 was used as a positive control because of its well-established function, and the empty vector was used as a negative control (final constructs in Supplementary Table S2 ). AtAMT1;1 and ScAMT2;1 sequences were amplified from Arabidopsis cDNA or synthetic vector, respectively, with primers containing restriction enzyme sites ( Supplementary Table S1 ), cloned into pGEM-T Easy (Promega, Madison, WI, USA), sequenced, and subcloned into the expression vector pDR196 (Rentsch et al., 1995). Triple mepΔ (strain 31019b) yeast cells were transformed by the lithium acetate method (Gietz and Schiestl, 2007). Confirmed positive clones were inoculated in liquid YNB-AA/AS (0.17% yeast nitrogen base without amino acids or ammonium sulfate) containing 1 mM arginine (positive control) and 50 mg L^-1 ampicillin for 36 h at 30°C at 200 rpm. A growth test was performed with a serial dilution (DO_600nm= 1, and subsequent dilution of 10^-1, 10^-2, and 10^-3) plated onto YNB-AA/AS supplemented with 3% glucose and one source of N [0.5, 2, 3, or 5 mM NH₄Cl (NH₄ ⁺), 100 mM methylammonium (MeA), or 1 mM arginine (Arg)]. MES-Tris was added at 20 mM to maintain the pH at 5.0, 6.0, or 7.5. The plates were incubated at 30°C for 6 d.

The ScAMT2;1 promoter region fragment from clones BAC 118_C18 (2,936 bp; p1ScAMT2;1) and BAC 235_F05 (2,962 bp; p2ScAMT2;1), hereafter called endogenous promoters, as well as the coding region from the synthetic ScAMT2;1 gene, were amplified (see above). The amplified products were cloned into pDONR or pCR8 (Thermo Fisher Scientific) and then subcloned into the final pMDC vectors (primers in Supplementary Table S1 ; vectors and final constructs in Supplementary Table S2 ) using the Gateway system (Thermo Fisher Scientific). Arabidopsis plants (Col-0 or qko) were transformed by floral dipping (Narusaka et al., 2010) using Agrobacterium tumefaciens GV3101 bearing the constructs indicated in Supplementary Table S2 . Transformed lines were selected for hygromycin resistance. Homozygous lines were confirmed by PCR and RT–qPCR.

Seedlings of Arabidopsis transgenic lines containing the GUS reporter gene (uidA, pMDC164) driven by ScAMT2;1 endogenous promoter were cultivated on half-strength MS media without N (-N) or supplied with 2 mM NH₄Cl or 1 mM NH₄NO₃ as the sole N source for up to 10 d. For GUS staining, plants were transferred to buffer containing 5‐bromo‐4‐chloro‐3‐indolyl glucuronide (X-Gluc; Jersey Lab and Glove Supply, Livingston, NJ, USA) at 37°C for 4 h 30 min and then washed in 70% ethanol (Jefferson et al., 1987). Plant tissues were analyzed and photographed under a Nikon SMZ18 stereo microscope.

Wild-type Arabidopsis and homozygous (T₃) ScAMT2;1-complemented qko plants were hydroponically grown in N-sufficient conditions (2 mM NH₄NO₃) with pH adjusted to 5.8 with 2-(N-morpholino)ethanesulfonic acid (MES) for 40 d. For qko+p35S::ScAMT2;1 lines, plants were subjected to N deficiency (-N, no N) or N sufficiency (+N, 1 mM NH₄NO₃), whereas qko+p2ScAMT2;1::ScAMT2;1 plants were transferred to -N, 2 mM KNO₃ or NH₄Cl as sole N sources. After 3 d under treatment, plants were exposed to a short-term ¹⁵N-ammonium influx assay with a 10 min incubation in a full-strength nutrient solution containing 0.2 mM (¹⁵NH₄)₂SO₄ (60% of ¹⁵N-ammonium). To assess ¹⁵N accumulation in roots and shoots, plants were subjected to -N for 3 d and then transferred to a ¹⁵N-labelled nutrient solution with 2 mM (¹⁵NH₄)₂SO₄ (60% of ¹⁵N-ammonium) for 1 h. For concentration-dependent influx of NH₄ ⁺ into roots of qko and qko+p35S::ScAMT2;1 lines, 40-d-old plants grown hydroponically under the same conditions mentioned above were transferred to -N for 3 d. Roots were then incubated for 10 min in full nutrient solution containing increasing concentrations 0, 25, 50 100, 150, 200, 300, and 500 mM of (¹⁵NH₄)₂SO₄ (60% of ¹⁵N-ammonium). To further assess the contribution of AMT2;1 to ¹⁵N root-to-shoot translocation, 40-d-old hydroponically grown qko and qko+ p2ScAMT2::ScAMT2;1 plants were subjected to -N for 3 d followed by one h-root exposure to a nutrient solution containing 0.2 mM or 4 mM (¹⁵NH₄)₂SO₄ (60% of ¹⁵N-ammonium). In all ¹⁵N experiments, roots were first rinsed with 1 mM CaSO₄ for 1 min before exposure to ¹⁵N, followed by washing with 1 mM CaSO₄ prior to sample collection. Roots and shoots were collected separately, dried, ground, and analyzed for total ¹⁵N content using continuous-flow isotope ratio mass spectrometry (ANCA SL, Sercon, Cheshire, UK).

A completely randomized design was used in all experiments. The number of biological replicates is indicated for each experiment. Analysis of variance (ANOVA) was performed, and means were compared using Tukey’s test at 5% significance or t test (p ≤ 0.10 and p ≤ 0.05), as indicated for each experiment, using SAS software (SAS Institute Inc., Cary, NC, USA).

Through real-time amplification of the three-dimensional pool of BAC clones using AMT2;1-specific primers followed by screening for the specific coordinates of positive clones, we identified five clones containing sequences closely related to ScAMT2;1 transcript, namely, BAC 032_A12, BAC 038_G02, BAC 118_C18, BAC 216_D16, and BAC 235_F05 ( Supplementary Figure S1 ). Each BAC clone contained a unique locus, except for BAC 216_D16 and BAC 235_F05, which shared the same protein sequence. The BAC 032_A12, BAC 038_G02, and BAC 118_C18 AMT2.1 loci were phylogenetically closer to the root-expressed assembled transcript (comp105883). On the other hand, BAC 216_D16 and BAC 235_F05 did not group closely with the transcript sequence ( Supplementary Figure S1 ).

As sugarcane cultivars are polyploids derived from interspecific crosses between S. officinarum and S. spontaneum (Nogueira et al., 2005; Graff et al., 2011), the identified BAC clones containing distinct ScAMT2;1 loci are expected to differ for the surrounding topology ( Figure 1 ). Of the five clones, BAC 216_D16 and BAC 235_F05 were highly similar, with 62.0% identity.

The alignment of the deduced amino acid of AMT2;1 sequences from the BAC clones, comp105883, and one S. spontaneum AMT2;1 (Wu et al., 2021) indicated that BAC 118_C18 had an identical protein sequence as the transcribed assembled sequence (comp105883) ( Supplementary Figure S3A ). The complete ScAMT2;1 gene from BAC 118_C18 (NCBI# OM471796) is 1,473 bp long with three exons encoding 490 amino acids, conceptually estimated to be a protein of 52 kDa ( Supplementary Figure S2B ). Similar gene structure and protein features were observed for S. spontaneum SsAMT2;1 from the chromosome 7 (Sspon.03G0003380-4D; Wu et al., 2021) ( Supplementary Figure S2B ; Supplementary Figure S3A ). In addition, the presumed ScAMT2;1 amino acid sequence from BAC 118_C18 was identical to the SsAMT2;1 protein Sspon.03G0003380-4D ( Supplementary Figure S3A ) and contained the expected 11 transmembrane domains predicted by TMHMM ( Supplementary Figure S3B ), along with the conserved signature motif for the MEP/AMT/Rh superfamily ( Supplementary Figure S3C ).

Various plant AMT1 and AMT2 sequences, including the identified sugarcane members, were compared to verify the conservation of C- and N-terminal regions concerning amino acids essential for transport function ( Supplementary Figure S4 ). In the N-terminus of the tomato protein LeAMT1;1, two cysteines (C3 and C27) have been proven to be fundamental for AMT1 oligomer stability (Graff et al., 2011). While AMT1 proteins except for SlAMT1;3 contained these two conserved Cys residues, these residues were absent in all AMT2 homologues ( Supplementary Figures S4A, C ). In the C-terminus, some residues have been associated with transport regulation, including glycine-456 (G456; SlAMT1;1) and threonine-460 (T460) (Ludewig et al., 2003). G456 was found in all AMTs evaluated to date, whereas T460 was absent in all AMT2 subfamily members, including ScATM2;1 ( Supplementary Figures S4B, D ).

Multiple alignment of the various ScAMT2;1 promoter sequences identified in the BAC clones (approximately 3 kb upstream of the predicted translation start codon) allowed the arbitrary separation of the clones into two groups, in which sequences from BAC 032_A12, BAC 038_G02, and BAC 118_C18 were more similar between each other, differing from BAC 216_D16 and BAC 235_F05 ( Supplementary Figure S5A ). We analyzed whether this separation could be due to transposable element (TE) insertions, which are commonly found in promoter regions of sugarcane sequences (de Setta et al., 2014). TE insertion was assessed by Censor, which identified repetitive elements by comparison with known repeats and assigned a score for probability. The results reinforced the similarity of BAC 216_D16 and BAC 235_F05, showing a similar TE insertion profile ( Supplementary Figure S5C ). To further investigate the presence of regulatory elements and presumed synteny of regulatory motifs, we chose clones from each group, BAC 118_C18 and BAC 235_F05. Only a few conserved regions exist between the selected regulatory regions, indicating significant variation between the two ScAMT2;1 promoters ( Supplementary Figure S5C ). As loci from BAC 118_C18 were not functional in driving the expression of uidA in the GUS assay (see below), the sequence from the BAC 235_F05 clone was chosen to be further analyzed as a functional ScAMT2;1 endogenous promoter. Concerning the gene sequence, ScAMT2;1 from BAC 118_C18 was selected for functional validation due to greater similarity with the root-expressed ScAMT2;1 sequence comp105883 and S. spontaneum Sspon.03G0003380-4D ( Supplementary Figure S2B ).

The transcriptional profile of ScAMT2;1 was examined in the organs of sugarcane plants grown under various N conditions. At the stage of generative growth (90-d-old plants) under N-sufficient conditions, ScAMT2;1 was expressed in all organs analyzed, with more transcript accumulation in roots, but it was also largely expressed in mature leaves, followed by young leaves, and less abundant in culms ( Figure 2A ). To assess how ScAMT2;1 expression is regulated by N supply, transcript levels were determined in various organs in plants grown in nutrient solution containing distinct N sources or without N for 14 d and compared with the +N treatment ( Figure 2B ). In the presence of nitrate as the sole N source, ScAMT2;1 transcripts accumulated approximately 2- to 3-fold more in roots, mature leaves, and culms but not in young leaves compared with plants grown in ammonium nitrate (+N). Thus, exposing the plants at the same high N level (4 mM) but changing the source from 2 mM NH₄NO₃ to 4 mM KNO₃ (4 mM N with no ammonium) was sufficient to induce ScAMT2;1 expression in roots and shoots more than the change from 2 mM NH₄NO₃ to 4 mM NH₄Cl.

To further investigate whether ScAMT2;1 is transcriptionally modulated by N availability, the expression profile was evaluated in sugarcane plants grown under 5 mM NH₄NO₃ (high N) or no N (-N) for 10 d ( Supplementary Figure S6 ). Transient and temporal transcript accumulation was detected in N-deficient mature leaves relative to high N supply. In culms, ScAMT2;1 transcripts showed some accumulation in both treatments; however, this transcriptional response was not observed in roots or young leaves. Altogether, these observations indicate that the N source and the plant N status modulate the expression of ScAMT2;1 in sugarcane.

To investigate whether the selected ScAMT2;1 gene (BAC118_C18) encodes a functional ammonium transporter, we complemented the S. cerevisiae triple mep mutant (31019b) (Marini et al., 1997). The positive control (triple mep complemented with AtAMT1;1) completely restored growth under all N conditions tested ( Figure 3 ). By increasing the external NH₄ ⁺ concentration, triple mep cells complemented with ScAMT2;1 showed slightly more growth than the negative control suggesting that ScAMT2;1 is a functional protein that mediates ammonium transport. At 5 mM ammonium, the growth of the triple mep complemented with ScAMT2;1 was strongly pH dependent. The ScAMT2;1-expressing triple mep grew slightly better than the negative control (empty pDR196) at a pH of 5.0 and 6.0. Raising the pH further to 7.5 may have increased the concentration of ammonia (NH₃), resulting in similar growth between triple mep expressing ScAMT2;1 and the negative control.

In contrast to type 1 AMT proteins, AMT2 has been proposed to be impermeable to the transport of the ammonium toxic analogue methylammonium (MeA) (Sohlenkamp et al., 2000; Sohlenkamp et al., 2002). The growth of triple mep complemented with ScAMT2;1 was evaluated on media supplemented with 100 mM MeA ( Figure 3 ). The toxic effect of MeA drastically reduced the growth of triple mep cells expressing AtAMT1;1, whereas those complemented with ScAMT2;1 or the empty vector displayed no visible sensitivity towards MeA.

ScAMT2;1 driven by the CaMV35S promoter (p35S) was expressed in the Arabidopsis quadruple AMT mutant line (qko) (Yuan et al., 2007). Three independent T₃ homozygous lines were characterized for ScAMT2;1 expression in relation to Col-0 plants ( Supplementary Figure S7 ) and then used for phenotypic evaluation. ScAMT2;1-complemented events grown in the presence of ammonium as the only N source accumulated significantly more total dry biomass than qko, with values approximately 65% (event #1) and 51% (event #2) higher under 2 mM NH₄ ⁺ ( Figures 4A, B ). In contrast, no significant difference between qko and the complemented lines was observed when only nitrate was supplied ( Figures 4A, B ), suggesting that the ectopic expression of ScAMT2;1 restored the qko mutant growth phenotype only under ammonium nutrition, likely by mediating ammonium uptake into roots.

We then evaluated the short-term influx of ¹⁵N-NH₄ ⁺ in the qko ScAMT2;1-complemented lines. Under -N, the root ammonium uptake capacity of qko+p35S::ScAMT2;1 increased by 87% compared with qko ( Figure 4C ), corroborating the function of ScAMT2;1 in NH₄ ⁺ uptake in roots. To estimate the substrate affinity of ScAMT2;1, six-week-old ScAMT2;1-overexpressing (p35S) qko plants were grown under -N for 3 d, followed by concentration-dependent ¹⁵N-NH₄ ⁺ influx analyses. In this experiment, ScAMT2;1 function was saturated above 90 µM ( Figure 4D ). The estimated net ammonium influx fitted the Michaelis–Menten equation well, resulting in a K_m = 90.17 µM and a V_max of 338.99 µmoles h^-1 g^-1 root DW, determined by subtracting the values of qko ( Figure 4D ). These results demonstrate that the ScAMT2;1 protein can contribute to high-affinity ammonium transport in planta.

To help determining the ScAMT2;1 function, we conducted localization experiments in Arabidopsis by expressing the GUS reporter gene driven by the ScAMT2;1 regulatory region from BAC 118_C18 (p1ScAMT2;1) and BAC 235_F05 (p2ScAMT2;1). No reporter expression was detected with the promoter p1ScAMT2;1, which was apparently nonfunctional ( Supplementary Figure S8 ). Arabidopsis lines expressing p2ScAMT2;1::GUS allowed tracing promoter activity in vascular bundles and outermost cells in leaves, either under ammonium or ammonium nitrate ( Figure 5A ). In contrast, leaves from N-deficient plants displayed no p2ScAMT2;1 activity in outer cells, and activity appeared to predominate at vascular bundles ( Figure 5A ). In roots, GUS was mainly detected in the innermost tissues ( Figure 5B ). Altogether, these results suggest that the ScAMT2;1 regulatory region is associated with root and leaf vascular tissues, but tissue-specific expression depends particularly on the N status in leaves rather than roots.

We then assessed the contribution of ScAMT2;1 to ammonium uptake by generating qko lines complemented with ScAMT2;1 driven by the endogenous regulatory region p2ScAMT2;1. While all p2ScAMT2;1::ScAMT2;1-complemented lines and qko grew similarly on agar medium supplemented with either 2 mM nitrate or 0.2 mM ammonium ( Figure 6A ), the total biomass of qko+p2ScAMT2;1::ScAMT2;1 plants was clearly superior to that of qko under higher external NH₄ ⁺ concentrations. At 2 mM NH₄ ⁺, qko+p2ScAMT2;1::ScAMT2;1 accumulated approximately 83% (#1), 103% (#2), or 28% (#3) more shoot biomass than qko ( Figure 6A ). The biomass accumulation for plants grown at 4 mM NH₄ ⁺ was 102% (event #1), 75% (event #2), or 107% (event #3) higher than the control plants ( Figure 6A ). These results suggest that ScAMT2;1 under the control of the sugarcane endogenous promoter significantly increased biomass at elevated external ammonium levels and confirm the functionality of ScAMT2;1 in facilitating NH₄ ⁺ uptake.

To evaluate the regulatory level of the response of the ScAMT2;1 promoter to high external N supply, short-term ¹⁵N-ammonium influx analysis was performed with qko+p2ScAMT2;1::ScAMT2;1 plants in the presence of 2 mM of either ammonium or nitrate or no N (-N). The influx of ¹⁵N-NH₄ ⁺ into the roots of qko and complemented lines subjected to -N was similar and nonsignificant ( Figure 6B ). Complementation with ScAMT2;1 significantly increased uptake levels by 6% (#1) and 43% (#2) compared with qko when subjected to 2 mM nitrate and by 61% (#1) and 78% (#2) in complemented plants subjected to 2 mM ammonium ( Figure 6B ). Altogether, these results indicate that the regulation of ScAMT2;1 in ammonium uptake depends strictly on the preconditioning of plants to an externally high N form but not to -N.

Our experiments indicated that ScAMT2;1 contributes to root ammonium uptake mainly in NH₄ ⁺-supplied plants ( Figure 6 ) and that the ScAMT2;1 promoter (p2) drives gene expression in the inner vascular root cells ( Figure 5 ). These results prompted us to evaluate whether ScAMT2;1 mediates root-to-shoot ammonium transport under ammonium supply. To this end, we evaluated ScAMT2;1-specific functions by estimating ¹⁵N accumulation in roots and shoots of qko and qko+p2ScAMT2;1::ScAMT2;1 lines subjected to either 0.2 mM or 4 mM ¹⁵N-NH₄ ⁺ for 1 h to allow time for root-to-shoot translocation ( Figure 7A ). At 0.2 mM ¹⁵N-NH₄ ⁺ supply, no significant ¹⁵N was accumulated in shoots compared with qko, whereas some ¹⁵N accumulation in roots occurred for one transgenic line (event #2). When plants were grown in the presence of 4 mM ¹⁵N-NH₄ ⁺, roots of qko and complemented lines accumulated ¹⁵N in a similar pattern. In contrast, significantly more ¹⁵N accumulated in the shoots of both qko+p2ScAMT2;1::ScAMT2;1 lines, approximately 35% and 25% more than in qko shoots ( Figure 7A ). The rate of ¹⁵N accumulated in the shoot in relation to the whole plant was 7 and 6.3% for the qko+p2ScAMT2;1::ScAMT2;1 lines, significantly superior to the 4.8% observed in qko when plants were subjected to 4 mM ¹⁵N-NH₄ ⁺ ( Figure 7B ). These results suggest that ScAMT2;1 activity in roots might contribute to ammonium translocation to shoots under high external ammonium conditions.