Genome-wide analyses of NPF and NRT2 reveal a potential role for a key gene, CqNRT2.2, in the quinoa response to nitrate Authors: Xiangxiang Wang, Xiaoyuan Li, Wenhua Dongcheng, Liangqin Zeng, He Wu, Xiaoshan Shao, Yingfa Zhang, Yan Yang, Xuexiang Nong, Shuai Zhang, Yongjun Xiong, Yiling Liu, Zhengjie Liu, Chun Lin Published: 2026 ABSTRACT BackgroundNO₃⁻ serves as the primary nitrogen source for plants and functions as a key signaling molecule in regulating growth and development. Despite its recognized importance, research on NO₃⁻ uptake and signaling in quinoa remains limited.ResultsIn this study, a genome-wide analysis identified 114 CqNPF and 15 CqNRT2 genes in quinoa. Heterologous complementation assays in the NO₃⁻-transporter-deficient yeast mutant Δynt confirmed that 11 CqNPF and 3 CqNRT2 transporters possess NO₃⁻ uptake capacity. Under NO₃⁻-deficient conditions, CqNRT2.2 was highly expressed in the roots of the quinoa cultivar ‘W32’. Using WGCNA and promoter binding site prediction, CqNLP2.1 is predicted to regulate CqNRT2.2 Further experiments showed that the transcription factor CqNLP2.1 binds to the CqNRT2.2 promoter. Overexpression of either CqNRT2.2 or CqNLP2.1 in Arabidopsis significantly promoted primary root elongation. Molecular docking predicted that both CqNLP2.1 and CqNLP2.2 may bind directly to NO₃⁻ ions, suggesting a potential but unconfirmed role in nitrate sensing. However, these are computational predictions only and do not demonstrate biological nitrate-sensing function, which requires experimental validation. Collectively, these findings indicate that CqNLP2.1 acts as a transcription factor in the quinoa NO₃⁻ signaling pathway, potentially regulating growth and development by modulating expression of the NO₃⁻-responsive gene CqNRT2.2.ConclusionsThis study offers novel insights into the molecular mechanisms of NO₃⁻ signal transduction in quinoa and provides a theoretical framework for enhancing nitrogen use efficiency in quinoa cultivation.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s12870-026-09242-x. Introduction Nitrogen (N) is a critical nutrient for plant growth and development. Severe nitrogen deficiency can severely inhibit plant growth, resulting in reduced biomass in both roots and shoots [1], impaired leaf development [2], decreased seed germination rates [3], fewer tillers, and lower grain yield [4]. In contrast, proper nitrogen fertilization can significantly enhance crop productivity. Consequently, understanding the molecular mechanisms that regulate nitrogen use efficiency (NUE) in plants could provide new strategies for improving yield, enhancing crop quality, and reducing nitrogen fertilizer consumption [5–7]. In plants, NO₃⁻ transport is mediated by four protein families: the NPF family, the NRT2 family, the CLC family, and the SLC/SLAH family [8, 9]. Among these, only specific members of the NRT1/NPF and NRT2 families have been confirmed to take up NO₃⁻ from the soil or external media [10, 11]. Based on their kinetic properties, these transporters generally operate as high-affinity transport systems (HATS) when external NO₃⁻ concentrations are below 0.5 mmol, or as low-affinity transport systems (LATS) when concentrations exceed 0.5 mmol [11–13]. Over the past three decades, numerous NPF and NRT2 genes have been identified, and their functions have been extensively studied in model plants [14, 15]. In Arabidopsis thaliana, at least 53 NPF genes and 7 NRT2 genes have been characterized. Most NRT1-type transporters function as low-affinity NO₃⁻ permeases, with one notable exception: AtNRT1.1 (AtNPF6.3/CHL1), which acts as a biphase-affinity NRT. In the LATS, AtNRT1.2 (AtNPF4.6) is critical for NO₃⁻ uptake, while the HATS is primarily mediated by AtNRT2.1, AtNRT2.2, AtNRT2.4, and AtNRT2.5 [16–19]. Interestingly, AtNRT1.1 also functions as a NO₃⁻ sensor, detecting external NO₃⁻ fluctuations to coordinate metabolic adaptation and root architecture remodeling. This sensing function is linked to its auxin transport activity, which is post-translationally regulated under low NO₃⁻ conditions, classifying AtNRT1.1 as a “transceptor” [20]. In contrast to Arabidopsis, rice, a key staple crop for global calorie supply, possesses a larger number of NRT/PTR family members, reflecting a more complex and precise regulation of NO₃⁻ uptake and transport. In rice, four NRT2 transporters are key components of the HATS. OsNRT2.1 is the major NO₃⁻ uptake transporter and plays a role in regulating root elongation by modulating auxin-related transporter genes [21, 22]. OsNRT2.2 participates in NO₃⁻ uptake in both roots and seeds [23]. The transport activity of OsNRT2.3b is sensitive to cellular pH changes [24]. Unlike AtNRT2.4 in Arabidopsis, OsNRT2.4 exhibits dual-affinity NO₃⁻ transport activity [25]. NO₃⁻ also plays a central role in activating the NO₃⁻ signaling pathway [9], which regulates several developmental processes, including primary and lateral root formation [26], seed dormancy [27], flowering time [28], and the coordinated uptake of other nutrients [1, 29]. The core of NO₃⁻ signal transduction involves transcription factors (TFs) from the NLP subfamily, which are essential for the expression of NO₃⁻-responsive genes [30]. NLP transcription factors are activated by nitrogen limitation and play a pivotal role in the nitrogen signaling network of plants [31]. In A. thaliana, the signaling pathways involving calcium-sensing protein kinases and NLPs have been well characterized, further highlighting the critical role of NLPs in NO₃⁻ signal transduction [32]. Recent studies provide additional evidence that NLP7 functions as a NO₃⁻ sensor, the conserved residues within the HLY motif of the NLP protein are essential for NO₃⁻ ion binding [29]. In essence, NLPs regulate the transcriptional activation of NO₃⁻-responsive genes [33]. Specifically, NLPs directly bind to the promoters of NRT genes to activate their expression, thereby linking NO₃⁻ signal perception to downstream transport activity [34]. Quinoa grows in drylands where NO₃⁻ is the dominant nitrogen form. Considering the key regulatory roles of NLP transcription factors on NRT genes in model plants [33–36], quinoa may likewise rely on the NLP-NRT regulatory module to achieve efficient NO₃⁻ utilization. In fact, the stage of greatest nutritional demand throughout the entire growth and development of quinoa occurs when the panicle emerges at the top (with 11–16 whorled leaves) [37]. Therefore, applying NO₃⁻-dominated fertilizers to the soil is a better agronomic practice to meet the nitrogen demand of quinoa. Previous studies have shown that when supplied with equivalent concentrations of nitrogen, NO₃⁻ can support similar or even better growth and yield of quinoa compared with ammonium [38, 39], indicating that quinoa possesses a robust NO₃⁻ absorption and assimilation system. NO₃⁻ uptake in plants is controlled by NRTs and their regulatory factors, NLPs [34, 40]. This strongly suggests that in quinoa, the core components responsible for NO₃⁻ uptake and signal transduction (such as the NRT and NLP families) may possess unique molecular features and regulatory mechanisms that enable efficient utilization of NO₃⁻. Although NRTs and their expression have been reported in many plants, and large omics datasets concerning quinoa are currently available [41], the NRT gene family, their functions, and regulatory mechanisms in quinoa remain unclear. This study investigated the quinoa CqNRT family, the functions of its gene members, and their regulatory networks based on a multi-omics strategy and molecular methods, aiming to better understand NO₃⁻ absorption, assimilation, and redistribution in quinoa, thereby further improving nitrogen use efficiency and optimizing cultivation practices for both yield and quality in quinoa production. Materials and methods Plant growth and experimental design The Bolivian quinoa cultivar W32 was introduced to Yunnan Agricultural University, where it underwent 10 successive generations of selfing. The W32 cultivar was grown under hydroponic conditions to enhance its performance in hydroponic cultivation. Using a Hoagland nutrient solution, three different concentrations of NO3– (0, 8, and 21 mM) were employed as the sole nitrogen source for the cultivation of W32 under hydroponic conditions, enabling easy assessment of root phenotypes. The nutrient solutions were prepared as follows: 0 mM NO3–: 1 mM KH2PO4, 2 mM MgSO4·7H2O, 4 mM CaCl2, 5 mM KCl; 8 mM NO3–: 1 mM KH2PO4, 2 mM MgSO4·7H2O, 4 mM Ca(NO3)2·4H2O, 5 mM KCl; 21 mM NO3–: 1 mM KH2PO4, 2 mM MgSO4·7H2O, 4 mM Ca(NO3)2·4H2O, 5 mM KCl, 13 mM NaNO3. The specific cultivation method followed the protocol described in [42]. Young seedlings exhibiting similar growth statuses were selected, grown in a nitrogen-free environment for one week, and then transferred to solutions containing different NO₃⁻ concentrations (0, 8, and 21 mM). The plants were cultivated for 21 and 27 days, respectively. Illumina RNAseq and analysis RNA sequencing (RNA-seq) data (Table S1) were evaluated using FastQC (https://github.com/topics/fastqc) and filtered with Fastp v0.23.4 (https://github.com/OpenGene/fastp) using default parameters to obtain clean data. RNA-seq data analysis was performed as previously described [42]. Briefly, the experiment included three biological replicates per condition. Raw reads were mapped to the quinoa reference genome, and gene expression levels were normalized using TPM (Transcripts Per Million). Differential expression analysis was conducted using DESeq2, with genes considered significantly differentially expressed if they met the criteria of |log2FC|≥ 1 and adjusted P-value (FDR) < 0.05. To gain a comprehensive understanding of the functions of the CqNPF and CqNRT2 gene families, this study utilized transcriptome data from the quinoa variety TY-20 grown at four different seed development stages, which has been published in previous research [43]. All RNA-seq data have been uploaded to CNCB with the IDs CRA017546 (https://ngdc.cncb.ac.cn/gsa/s/154IMIH0), CRA014457, and CRA016779 (https://ngdc.cncb.ac.cn/?lang=zh). Identification and characterization of CqNPF and CqNRT2 gene family in quinoa To identify NPF and NRT2 proteins in quinoa, 53 AtNPF protein sequences and 7 AtNRT2 protein sequences were retrieved from the TAIR website (https://www.arabidopsis.org/). These sequences served as reference sequences for genomic mining and comparison of NPF and NRT2 genes in quinoa. The quinoa genome assembly version Quinoa_v2.0 was obtained from the Phytozome database [41] (https://phytozome-next.jgi.doe.gov/info/Cquinoa_v1_0). To ensure data quality, redundant transcripts were removed by retaining only the longest transcript per gene using in-house scripts. Furthermore, truncated sequences with a protein length less than 50% of the typical NPF/NRT2 domain length were discarded before further analysis. The analysis was performed using BLAST v2.11.0 +, with filtering criteria of greater than 40% similarity, greater than 80% coverage, and an e-value of less than 1e–6. Conserved domains for NPF (PF00854) and NRT2 (PF07690) were then searched for in the quinoa genome database using the Pfam database (https://www.ebi.ac.uk/interpro/), Retain the protein sequences with an E-value less than 0.01. Genes identified via BLASTP and Hmmsearch, with intersections across methods, were considered potential members of the CqNPF and CqNRT2 gene families. The NPFs from A. thaliana, along with the NRT2 protein sequences from A. thaliana, Oryza sativa, and Zea mays, were obtained from a previous study. Additionally, the CqNPFs and CqNRT2s identified in this study were included in the phylogenetic analysis. The corresponding IDs are listed in Tables S2 and S3. The multiple sequence alignment at the amino acid level was performed by MUSCLE [44]; the phylogenetic tree was constructed based on referring to the maximum likeli-hood method of RAxML [45]. There were 1000 bootstrap replicates and it was beautified with the iTOL online website (https://itol.embl.de/). Chromosomal localization and physicochemical property analysis CqNPFs and CqNRT2s were mapped to the chromosomes of quinoa using the Gene Location Visualizer from the GTF/GFF function of TBtools (https://github.com/CJ-Chen/TBtools). The ExPASy online tool (http://www.expasy.ch/tools/pi_tool.html) was employed to obtain the amino acid lengths, molecular weights (MW), and isoelectric points (pI) of NPF and NRT2 proteins. Characterization of CqNPF and CqNRT2 gene family Gene sequence data for CqNPFs and CqNRT2s were downloaded from the quinoa-Genome-Hub. Promoter regions (2.0 kb upstream of the start codon ATG) of all CqNPFs and CqNRT2s sequences were extracted using TBtools, followed by cis-acting element prediction via PlantCARE [46]. Motif analysis was performed using the MEME online software. Finally, gene structure, conserved motifs, and domains of CqNPF and CqNRT2 genes were integrated into a map using TBtools [47]. Chromosome locations were determined using TBtools, and genome collinearity analysis was conducted with MCScanX software, using default settings [48]. Subcellular localization prediction of CqNPFs and CqNRT2s The subcellular localization of CqNPFs and CqNRT2s were predicted using the Cell-PLoc online tool (http://www.csbio.sjtu.edu.cn/bioinf/plant-multi/). Quantitative Real-Time PCR (qRT-PCR) analysis Total RNA was extracted from the samples described above, using the Hipure HP Plant RNA Mini Kit (Magen, Shanghai, China) according to the manufacturer’s protocols. Of the extracted total RNA, ~ 0.9 μg was used to synthesize the first-strand cDNA using FastKing gDNA Dispelling RT Super Mix (TIANGEN, Beijing China). qRT-PCR was performed with a SYBR qPCR Master Mix (BestEnzymes Biotech Co., Ltd.Jiangsu China) in an ABI7500 StepOnePlus Real-Time PCR System (ABI, MA, USA). The PCR mixtures were first incubated in a 96-well optical plate at 95 °C for 60 s, followed by 40 cycles at 95 °C for 10 s and 60 °C for 30 s. A melt curve was drawn by the instrument default acquisition program. The expression levels were calculated using the 2 − ∆∆CT method [49] with the primers listed in Table S13. CqNPFs and CqNRT2s cloning and yeast functional complementation The putative open reading frames (ORFs) of CqNPF5.20, CqNRT2.2, CqNPF3.1, CqNRT2.3, CqNRT2.6, CqNPF1.5, CqNPF6.12, CqNPF2.13, CqNPF4.3, CqNPF7.5, CqNPF5.19, CqNPF6.13, CqNPF8.7, and CqNPF6.1 were successfully amplified by PCR using specific primers containing the SalI and SpeI restriction sites (listed in Table S4). The purified PCR products were then cloned into the yeast expression vector pYNR via homologous recombination. The pYNR-CqNPFs/CqNRT2s plasmids were introduced into the Hansenula polymorpha YNT1 mutant strain △ynt (△ynt, △leu), which is unable to grow on medium containing NO3– as the sole nitrogen source at concentrations below 0.5 mM. As a positive and negative control, the pYNR plasmid was introduced into the yeast strains NCYC495 (△leu) and △ynt (△ynt, △leu), respectively [50–52]. All transformants were initially selected on YG agar medium (2% galactose, 2% agar, 0.17% yeast nitrogen base without amino acids, and ammonium sulfate from Difco, Detroit, MI, USA) supplemented with 5 mM NaNO3 as the nitrogen source. Monoclonal bacterial cultures confirmed by specific primer PCR were streaked onto YPD plates until single colonies formed. A single colony was picked, suspended in 90 µL of water, serially diluted, and spotted (3 µL) onto YG agar medium containing 0.5 mM NO3– as the sole nitrogen source. The medium pH was adjusted using 1 M HCl or KOH. Enzymatic assays and phenotypic measurements of quinoa roots Key enzymes involved in nitrogen metabolism, including GS, GOGAT, NR, and NiR, were quantified using a microassay kit (Jiangsu Grise Biotechnology Co., Ltd., Suzhou, China) following the manufacturer’s specifications. Root systems were scanned and analyzed using the WinRHIZO 2019b root scanner (Regent Instruments Inc., Quebec City, Canada). Predicting the upstream regulatory factors of CqNPFs and CqNRT2s In this study, weighted gene co-expression network analysis (WGCNA, v1.47) was performed in the R environment to conduct co-expression module analysis. To ensure the network approximates a scale-free topology, the soft-thresholding power was set to 8 (scale-free \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${R}^{2}=0.9$$\end{document}). A minimum module size of 30 was used to avoid generating too many small, potentially spurious modules. Modules with a similarity above 0.85 (i.e., merge cut height = 0.15) were merged to reduce redundancy. The specific parameters used were: power = 8, minimum module size = 30, and merge cut height = 0.15. Gene modules (GMs) containing CqNPFs and CqNRT2s were further utilized to construct transcriptional regulatory networks related to nitrogen transport and assimilation. Transcription factor binding sites were predicted in the 2000 bp promoter region upstream of the initiation codon (ATG) of target genes. Pearson correlation analysis was then used to identify associations between transcription factors and target gene expression, with the selection criteria set as absolute R-value > 0.8 and p-value < 0.05. Additionally, correlations between the expression of putative transcription factors and at least three phenotypic traits of quinoa (fresh weight, dry weight, chlorophyll a, chlorophyll b, carotenoids, average root diameter, number of root tips, root length, root surface area, GS, GOGAT, NiR, and NR) as well as at least three seed quality traits (water content, seed size, total sugars, soluble sugars, starch, total proteins, soluble proteins, amino acids, lipids, flavonoids, and saponins) were analyzed. Data on seed quality traits were obtained from a previously published article by the research group [43]. Finally, the constructed correlation networks were visualized using Cytoscape (v3.7.1). Construction and genetic transformation of the CqNRT2.2 and CqNLP2.1 Arabidopsis overexpression vector For Arabidopsis complementation analysis, the genomic DNA fragments of CqNRT2.2 and CqNLP2.1 were cloned into the pCAMBIA1300 vector. After verification through extensive restriction digestion and DNA sequencing, the construct was transformed into A. thaliana Col. All primers used for vector construction and qRT-PCR are listed in Table S5. GUS activity assay of the CqNRT2.2 promoter The plant expression vector pCAMBIA3301 was linearized by digestion with the restriction enzymes HindIII and NcoI. The CqNRT2.2 promoter fragments (PF1, PF2, PF3 and PF4, excluding the 5’ end, with PF representing the full 2 kb promoter sequence) were then ligated to the linearized vector using homologous recombination to replace the 35S promoter, generating a recombinant plasmid. The recombinant plasmid and the empty vector pCAMBIA3301 were introduced into Agrobacterium GV3101, which was cultured in YEP liquid medium containing 50 mg/L kanamycin. The culture was incubated until the OD600 value reached 0.8–1.5. The bacterial suspension was centrifuged, and the OD600 of the suspension was adjusted to 1.0 using suspension solution (10 mM MgCl2, 10 mM MES, and 200 μM As). This bacterial suspension was then infiltrated into four-week-old tobacco leaves. After one day of dark incubation and two days of normal light incubation, GUS staining was performed using the GUS staining kit (Beijing Solibo Technology Co., Ltd.) for qualitative analysis of the GUS gene. All primers used for vector construction are listed in Table S6. Yeast one-hybrid assays For the yeast one-hybrid (Y1H) assay, the NLP-binding core sequences from the promoters of CqNRT2.2 (designated as CqNRT2.2-PF) and their mutated versions (CqNRT2.2-MPF) were cloned into the pAbAi vector (Clontech, Palo Alto, CA, USA). Concurrently, the coding sequences of CqNLP2.1 and CqNLP2.2 were inserted into the pGADT7 vector as prey constructs (Clontech). The resulting pAbAi-CqNRT2.2-PF/MPF, pGADT7-CqNLP2.1, and pGADT7-CqNLP2.2 vectors were transformed into Saccharomyces cerevisiae Y1HGold strain for validation via the Y1H assay. The Y1H assays were performed as described by [53]. Table S7 lists the primers used to generate the clones for these analyses. Dual luciferase transient expression assays The full-length coding sequence of CqNLP2.1 was cloned as the effector gene into the vector pGreenII 62-SK, and the promoter of CqNRT2.2 was cloned into the dual-reporter vector pGreenII 0800-LUC. The empty vector pGreenII 62-SK was used as the negative control. The plasmid was transfected into GV3101 and co-injected into tobacco leaves. After two days of cultivation, the absolute LUC and REN activities were measured using a multi-well plate reader (Spark) and a dual luciferase assay kit (TransGen Biotech). The transcriptional activation status of the target promoter was calculated by the ratio of LUC to REN. All primers used for vector constructions are listed in Table S8. Prediction of the quinoa CqNLP2.1/2.2 protein structure and molecular docking of NO₃⁻ ions Molecular docking simulations were performed using AutoDock Vina v1.2.5 [54] to dock NO₃⁻ ions into the predicted structure of the CqRWP-RK protein. The three-dimensional structure of the NO₃⁻ ion (CID_94310) was retrieved from the PubChem database (https://pubchem.ncbi.nlm.nih.gov/) [55]. Before docking, the predicted protein model was prepared with PyMOL v3.1.0 [56], which involved removing water molecules and any bound ligands, followed by the addition of hydrogen atoms. NO₃⁻ ions underwent similar preparation steps. Flexible docking was then conducted to explore potential binding modes between the protein and NO₃⁻. The resulting docking poses were filtered based on two criteria: binding energy lower than zero and the presence of at least one well-defined binding site. Models meeting these criteria were selected for further analysis. Statistical analyses Statistical analysis of the data was performed using RStudio. Normality and homogeneity of variances were first checked using the Shapiro–Wilk test and Levene’s test, respectively, and all data met these assumptions unless otherwise stated. Differences between samples were evaluated using one-way analysis of variance or t-tests. For within-group comparisons, t-tests were employed, with significance levels indicated by asterisks. For inter-group comparisons, one-way ANOVA was conducted, with different letters denoting significant differences. Results The phenotypic changes of quinoa seedlings under different NO₃⁻ concentrations To elucidate the mechanisms underlying efficient nitrogen utilization and regulation in quinoa, the hydroponically adapted cultivar W32 was subjected to NO3– treatments at concentrations of 0, 8, and 21 mM. A comprehensive analysis of growth parameters and physiological indicators was then performed (Fig. 1). As shown in Fig. 1A and C, nitrogen deficiency (0 mM NO₃⁻) significantly inhibited plant growth, with leaves exhibiting typical nitrogen-deficiency symptoms, including pronounced chlorosis in the lower leaves. With increasing nitrogen availability, plant growth performance gradually improved, culminating in complete phenotypic recovery at optimal nitrogen levels, as evidenced by vigorous growth, restored leaf greening, and increased biomass. Meanwhile, compared with normal or high nitrogen levels, quinoa plants exhibited more developed root systems and greater biomass under nitrogen-deficient conditions (Fig. 1B, C). As NO₃⁻ concentrations increased, root growth was progressively inhibited, with consistent decreasing trends observed in both root length and surface area. In contrast, average root diameter increased. No significant difference in root diameter was observed among the 0—8 mM NO₃⁻ treatments, while the highest nitrogen concentration (21 mM) resulted in a significantly larger root diameter compared to lower nitrogen treatments. Notably, the number of root tips peaked at 8 mM NO₃⁻-N, showing significant differences from the other treatment groups (Fig. 1D). Additionally, increasing NO₃⁻ concentrations progressively enhanced photosynthetic capacity, as evidenced by elevated net photosynthetic rate, respiration rate, CO2 concentration, and stomatal aperture (Fig. S1A). Alongside changes in root architecture and photosynthetic parameters, the activities of key nitrogen-assimilating enzymes in quinoa leaves and roots—including glutamate synthase (GOGAT), glutamine synthetase (GS), nitrite reductase (NiR), and nitrate reductase (NR)—demonstrated an overall increasing trend with elevated NO₃⁻ concentrations. Interestingly, GOGAT and NiR activities exhibited a biphasic response: enhanced under both nitrogen-deficient (0 mM NO3–) and high-nitrogen (21 mM NO3–) conditions, but suppressed at intermediate NO₃⁻ supply (8 mM NO3–). This pattern likely reflects a homeostatic mechanism in internal nitrogen metabolism (Fig. S1B, C).Fig. 1Phenotypic and physiological responses of quinoa to NO₃⁻ supply. A Growth phenotype of quinoa plants under different NO₃⁻ concentrations. B Root architecture of quinoa under different NO₃⁻ regimes. C Shoots and roots dry weights of quinoa under different concentrations of NO₃⁻. D Physiological response of quinoa roots to different concentrations of NO₃⁻. Data are presented as mean ± SD (n = 14 biological replicates for biomass parameters; n = 11 for root parameters). Statistical significance was determined by one-way ANOVA followed by Tukey’s HSD post-hoc test. Different letters indicate significant differences at p < 0.05 (95% confidence interval). The same or shared letters indicate no significant difference. Scale bar = 2.5 cm Phenotypic and physiological responses of quinoa to NO₃⁻ supply. A Growth phenotype of quinoa plants under different NO₃⁻ concentrations. B Root architecture of quinoa under different NO₃⁻ regimes. C Shoots and roots dry weights of quinoa under different concentrations of NO₃⁻. D Physiological response of quinoa roots to different concentrations of NO₃⁻. Data are presented as mean ± SD (n = 14 biological replicates for biomass parameters; n = 11 for root parameters). Statistical significance was determined by one-way ANOVA followed by Tukey’s HSD post-hoc test. Different letters indicate significant differences at p < 0.05 (95% confidence interval). The same or shared letters indicate no significant difference. Scale bar = 2.5 cm Identification and structure of quinoa CqNPF and CqNRT2 genes To identify the NPF and NRT2 family genes in quinoa, a BLASTP search was employed to identify genes containing the core domains of the NPF/PTR2 (PF00854) and NRT2 (PF07690) subfamilies. A total of 114 NPF and 15 NRT2 genes were identified in the quinoa genome (Tables S2, S3). These 114 NPF and 15 NRT2 genes were unevenly distributed across the 19 and 7 quinoa chromosomes, respectively (Figs. S2, S3). Additionally, the characteristics of the CqNPF and NRT2 genes were systematically evaluated (Tables S2, S3). The CDS length of the CqNPF genes ranged from 381 base pairs (CqNPF1.11) to 3099 base pairs (CqNPF5.23), with corresponding protein lengths ranging from 126 amino acids to 1032 amino acids (Table S2). Meanwhile, the amino acid count within the NRT2 gene family ranged from 466 (CqNRT2.11) to 942 (CqNRT2.14), and the corresponding MW varied from 50.99 kDa to 101.59 kDa (Table S3). To explore the structural diversity and evolutionary relationships within the CqNPF and CqNRT2 gene families, their gene structures and conserved protein patterns were analyzed (Figs. S4, S5). Motif analysis of the CqNRTs proteins identified a total of 12 conserved motifs, numbered from 1 to 12 (Figs. S4B, S5B). The composition of these motifs corresponds to the phylogenetic relationships, with proteins within the same subgroup typically exhibiting similar motifs. Further functional analysis of these conserved motifs was conducted using the Pfam database. The results revealed that the amio acid residue lengths of the NPF conserved motifs ranged from 19 to 41, while those of the NRT2 conserved motifs ranged from 15 to 50. All motifs were annotated with NRT functions. However, significant differences in motif composition were observed across different groups (Tables S9, S10). Notably, Motif 2 and Motif 3, which possess the NRT1/PTR structure, were found in all CqNPF proteins, suggesting their critical role in the functions of these proteins. To investigate structural variations in the CqNRT family genes during evolution, the structures of the CqNPF and CqNRT2 genes were analyzed (Figs. S4C, S5C). The analysis revealed relatively high structural diversity among the CqNPF genes. The number of exons in the 114 CqNPF genes ranged from 1 to 7, with CqNPF5.39 having the most exons (n = 7). Intron distribution varied across CqNPF subfamilies, but most CqNPF genes contained at least three introns, indicating that intron number has fluctuated throughout the evolution of the CqNPF gene family, possibly contributing to the functional diversity of CqNPF genes. Similarly, the exon–intron structure of CqNRT2s was analyzed to better understand the structural characteristics of these genes (Fig. S5C). The results showed that most CqNRT2 genes shared similar structures, with 2–3 exons per gene. Of the 15 CqNRT2 genes, five genes (CqNRT2.13, CqNRT2.9, CqNRT2.10, CqNRT2.11, and CqNRT2.12) contained one intron, while seven genes (CqNRT2.5, CqNRT2.6, CqNRT2.4, CqNRT2.3, CqNRT2.1, CqNRT2.2, and CqNRT2.15) had two introns. The remaining three NRT2 genes had more than two introns. These results suggest that the structural similarities observed in the NPF and NRT2 genes of quinoa may be the result of repeated evolutionary events within the species. Cis-regulatory elements play a pivotal role in transcriptional regulation, influencing processes such as plant hormone responses, defense mechanisms, and developmental pathways. To further investigate the potential regulatory mechanisms controlling the CqNPF and CqNRT2 genes, the cis-acting elements in the promoters of these genes were analyzed using the PlantCARE database. A total of 14 and 15 cis-acting regulatory elements (CAREs) were identified in the 114 CqNPF and 15 CqNRT2 genes, respectively. These elements included hormone-responsive elements, defense and stress-responsive elements, light-responsive elements, and growth-related elements (Figs. S6, S7). These results indicate that the members of the CqNPF and CqNRT2 families may play significant roles in various aspects of plant hormone regulation, stress responses, and light signaling. Phylogenetic analysis of CqNPFs and CqNRT2s To explore the phylogenetic relationships between CqNPFs and CqNRT2s in quinoa and other species, a protein multiple sequence alignment was performed, followed by the construction of a neighbor-joining (NJ) phylogenetic tree for the NPF family, which included 53 AtNPFs. Additionally, an NJ phylogenetic tree was generated for the NRT2 family, consisting of 7 AtNRT2s, 4 OsNRT2s, and 4 ZmNRT2s (Fig. 2A, C). The total number of NPF and NRT2 genes in quinoa is significantly larger than in rice, maize, and Arabidopsis. This expansion can partly be attributed to quinoa’s tetraploid nature. Notably, the number of both NPF and NRT2 genes in quinoa was approximately twice that of their orthologs in Arabidopsis. These results suggest that the CqNPFs and CqNRT2s in quinoa may have originated from the A and B genomes of diploid ancestors [57]. To further investigate this, homologous gene searches were conducted in the diploid species Chenopodium watsonii (genotype A, CW) and Chenopodium suecicum (genotype B, Csu), which resulted in the identification of 55 NPF homologs in CW and 29 in Csu. Additionally, 8 and 6 NRT2 homologs were found in CW and Csu, respectively (Tables S11, S12). Collinearity analysis revealed that homologous CqNPF and CqNRT2 genes maintain 1-to-1 or 1-to-many homology relationships between quinoa and both CW and Csu. Interestingly, CqNRT2.11 is present only in the co-linear region between quinoa and C. suecicum, where it is homologous to Cw010535. A similar pattern was observed in the NPF gene family, where genes such as CqNPF1.9, CqNPF1.11, CqNPF1.12, CqNPF5.8, CqNPF5.10, CqNPF5.15, and CqNPF5.29 have no homologs in C. suecicum or C. watsonii, yet they are located in the co-linear regions between quinoa and the diploid species of the A and B genomes (Fig. 1B, D; Tables S11, S12). This phenomenon is not primarily due to gene loss during polyploid formation but rather results from the retention of genes from the dominant sub-genome and the loss of genes from the inferior sub-genome (fractionation) during the long evolutionary process following polyploidization. This characteristic is a defining feature of polyploid genome evolution and represents a key evolutionary force driving plants to adapt to new environments and develop novel regulatory networks. A previous study classified the NPF family of 33 plant species into eight subfamilies [14, 58]. In our analysis, the NPF5 and NPF6 subfamilies, which were originally grouped together, were further subdivided into two distinct subfamilies. The distribution of CqNPFs across the different subfamilies is as follows: NPF1 (16), NPF2 (16), NPF3 (5), NPF4 (7), NPF5 (33), NPF5.1 (6), NPF6 (12), NPF6.1 (2), NPF7 (10), and NPF8 (7). The variation in the number of CqNPF genes across these subfamilies indicates a significant expansion trend (Fig. 2A). In contrast, the CqNRT2 gene family is divided into three subgroups. Subgroup 2 contains both monocot and dicot plants, while Subgroup 1 includes only dicotyledonous plants. Subgroup 3, the largest subgroup, contains 9 CqNRT2 genes along with 1 AtNRT2 gene. The three highly homologous CqNRT2 genes are located on chromosomes 5, 7, 12, and 18 (Fig. S3), and the 9 CqNRT2 genes in Subgroup 3 are homologous to AtNRT2.5. These results suggest that gene duplication and amplification processes have contributed significantly to the expansion of the CqNRT2 gene family in quinoa (Fig. 2C).Fig. 2Phylogenetic and syntenic analysis of NRT genes in Chenopodium species. A Phylogenetic tree of NPF proteins from Arabidopsis thaliana and Chenopodium quinoa. Protein clades are color-coded, with CqNPF proteins highlighted in green. B Retention and loss of CqNPF genes across C. watsonii (Cw), C. suecicum (Cs), and C. quinoa (Cq). C Phylogenetic tree of CqNRT2 proteins from A. thaliana, Oryza sativa, Zea mays, and C. quinoa. Clades are distinguished by color, and CqNRT2 proteins are marked in green. D Retention and loss of CqNRT2 genes among C. watsonii (Cw), C. suecicum (Cs), and C. quinoa (Cq) Phylogenetic and syntenic analysis of NRT genes in Chenopodium species. A Phylogenetic tree of NPF proteins from Arabidopsis thaliana and Chenopodium quinoa. Protein clades are color-coded, with CqNPF proteins highlighted in green. B Retention and loss of CqNPF genes across C. watsonii (Cw), C. suecicum (Cs), and C. quinoa (Cq). C Phylogenetic tree of CqNRT2 proteins from A. thaliana, Oryza sativa, Zea mays, and C. quinoa. Clades are distinguished by color, and CqNRT2 proteins are marked in green. D Retention and loss of CqNRT2 genes among C. watsonii (Cw), C. suecicum (Cs), and C. quinoa (Cq) Expression profiles of CqNPFs and CqNRT2s in various quinoa cultivars at different developmental stages RNA-seq data from various tissues of W32 grown in modified Hoagland solution containing 0, 8, and 21 mM NO₃⁻, as well as RNA-seq data from W25 and TY-20 (representing high-altitude ecotype varieties) grown in the field, were analyzed (Figs. 3, S8). Volcano plots showing differentially expressed genes in quinoa roots and leaves under different NO₃⁻ concentrations are presented in Fig. S9A-L. The expression patterns of CqNPFs and CqNRT2s were examined in several organs: root (Rs) and leaf (Ls) for W32 seedlings (representing the vegetative growth stage), leaf (Ls) and spike (Ps) for W25 during flowering (F) and seed filling (S) stages (representing the reproductive growth stage), and TY20 seeds harvested 7 days (stage 1), 17 days (stage 2), 27 days (stage 3), and 42 days (stage 4) after flowering (stage 4). The RNA-seq data revealed that 64.1% of CqNPF genes were highly expressed in leaves (0/8/21 mM NO3–) and 71.9% were highly expressed in roots, with some CqNPF genes showing elevated expression under nitrogen-deficient conditions. Meanwhile, 27.1% of CqNPF genes were highly expressed in reproductive organs (FL, SL, FP, SP, and various seed stages after flowering). The heatmap indicated that most CqNRT2 genes were predominantly expressed in roots, Notably, CqNRT2.2 exhibited the highest expression level under 0 mM NO₃⁻ conditions, suggesting that it may be a key gene promoting nitrogen uptake in quinoa roots under low-nitrogen conditions. However, seven genes (CqNRT2.7, CqNRT2.8, CqNRT2.14, CqNRT2.15, CqNRT2.2, CqNRT2.1, and CqNRT2.13) exhibited high expression levels in TY-20 grains, suggesting their involvement in nitrogen accumulation in seeds. Furthermore, CqNRT2.7, CqNRT2.8, CqNRT2.14, and CqNRT2.15 were highly expressed in the leaves, indicating a potential role in NO₃⁻ transport and assimilation. Based on tissue expression patterns, the NRT2 family in quinoa was categorized into two groups: the first group exhibited low or no expression in roots and other tissues, while the second group showed high specificity in roots. These results highlight the critical role of these genes in NO3– absorption and transportation under nitrogen-limiting conditions. Expression levels of 11 CqNPF genes and 15 CqNRT2 genes were further analyzed using qRT-PCR. The qPCR data largely corroborated the RNA-seq expression profiles (Figs. S10, S11). Overall, the expression patterns of CqNPFs and CqNRT2s highlight the diverse roles that the CqNRT family plays in quinoa growth and development through the absorption, transfer, and redistribution of NO3–.Fig. 3Expression profiles of CqNRT2 genes. Expression levels in leaves and roots of cultivar W32 under 0, 8, and 21 mM NO₃⁻ at 21 and 27 days after treatment; Expression patterns in different reproductive tissues of field-grown cultivars W25 and TY-20: W25-FL (leaves at flowering stage), W25-SL (leaves at seed-filling stage), W25-FP (panicles at flowering stage), and W25-SP (panicles at seed formation stage). Expression dynamics in developing seeds at four stages: Stage1 (7 days after flowering), Stage2 (17 DAF), Stage3 (27 DAF), and Stage4 (42 DAF) Expression profiles of CqNRT2 genes. Expression levels in leaves and roots of cultivar W32 under 0, 8, and 21 mM NO₃⁻ at 21 and 27 days after treatment; Expression patterns in different reproductive tissues of field-grown cultivars W25 and TY-20: W25-FL (leaves at flowering stage), W25-SL (leaves at seed-filling stage), W25-FP (panicles at flowering stage), and W25-SP (panicles at seed formation stage). Expression dynamics in developing seeds at four stages: Stage1 (7 days after flowering), Stage2 (17 DAF), Stage3 (27 DAF), and Stage4 (42 DAF) Molecular cloning of CqNPFs and CqNRT2s and heterologous complementation analysis of yeast mutants △ynt (△ynt, △leu) To evaluate the functionality of CqNPFs and CqNRT2s in NO₃⁻ transport, a heterologous functional complementation assay was conducted in Hansenula polymorpha. The ORFs of 11 CqNPFs (CqNPF8.7, CqNPF3.1, CqNPF6.12, CqNPF1.5, CqNPF6.13, CqNPF4.3, CqNPF5.20, CqNPF7.5, CqNPF2.13, CqNPF6.1, and CqNPF5.19) and 3 CqNRT2s (CqNRT2.2, CqNRT2.6, and CqNRT2.3) were cloned into the pYNR vector and introduced into the yeast mutant △ynt (△ynt, △leu), which is unable to grow when 0.5 mM NO3– is the sole nitrogen source. The results demonstrated that, compared to transformants carrying the empty vector pYNR, yeast cells expressing CqNPFs and CqNRT2s resumed growth at 0.5 mM NO3–, with growth conditions comparable to the WT strain NCYC495 (△leu) transformed with pYNR as a positive control (Fig. 4).Fig. 4Functional complementation of a yeast NO₃⁻-uptake mutant by heterologous expression of CqNPF and CqNRT2 genes. The NO₃⁻ transporter-deficient yeast mutant △ynt1 (△ynt1, △leu) and its wild-type strain NCYC495 (△leu) were transformed with the empty yeast expression vector pYNR or with pYNR harboring CqNPF or CqNRT2 coding sequences. Transformants were first selected on NAAG medium containing 5 mM NO3– as the nitrogen source. A single colony was resuspended in sterile water, serially diluted, and spotted (3 µL) onto NAAG medium with 0.5 mM NO₃⁻ as the sole nitrogen source. The △ynt1 mutant expressing CqNPF or CqNRT2 genes showed restored growth on NO₃⁻, comparable to the wild-type positive control, whereas △ynt1 carrying the empty vector failed to grow (negative control) Functional complementation of a yeast NO₃⁻-uptake mutant by heterologous expression of CqNPF and CqNRT2 genes. The NO₃⁻ transporter-deficient yeast mutant △ynt1 (△ynt1, △leu) and its wild-type strain NCYC495 (△leu) were transformed with the empty yeast expression vector pYNR or with pYNR harboring CqNPF or CqNRT2 coding sequences. Transformants were first selected on NAAG medium containing 5 mM NO3– as the nitrogen source. A single colony was resuspended in sterile water, serially diluted, and spotted (3 µL) onto NAAG medium with 0.5 mM NO₃⁻ as the sole nitrogen source. The △ynt1 mutant expressing CqNPF or CqNRT2 genes showed restored growth on NO₃⁻, comparable to the wild-type positive control, whereas △ynt1 carrying the empty vector failed to grow (negative control) Predictive analysis of upstream regulatory factors of CqNPFs and CqNRT2s N is an essential element for the growth and productivity of plants, animals, and microorganisms. Its critical roles in various biological processes highlight its importance in sustaining life and driving agricultural productivity. In plants, nitrogen is fundamental to metabolism, influencing not only growth and development but also crop health and yield. To investigate the gene regulatory network, particularly putative key regulator factors involved in NO3– transport and assimilation, WGCNA was performed, and a gene co-expression module was constructed based on RNA-seq data (in TPM format). Fifteen Gene modules (denoted by different colors) were identified, reflecting various gene expression patterns (Fig. S12A-C), with the distribution of genes in each module provided in Tables S14-S27. Previous studies have shown that NIN-like transcription factors regulate most NO₃⁻ signaling genes, enhancing tiller number, grain yield, and NUE [31, 35, 36]. From the Gene modules, three out of thirteen CqNLPs were grouped into four Gene modules: CqNLP2.1, CqNLP2.2, and CqNLP1.3 were clustered in the greenyellow GM (GGM); CqNLP1.3, CqNLP1.4, CqNLP1.5, CqNLP1.6, CqNLP5.1, and CqNLP5.2 were grouped into the blue GM (BGM); CqNLP1.1 and CqNLP1.2 were placed in the red GM (RGM); CqNLP3.1 and CqNLP3.2 were grouped into the salmon GM (SGM) (Tables S24-S27). Among these modules, the Greenyellow module (GGM) was selected for further analysis because it showed the strongest positive correlation between NLP transcription factors and NRT target genes (average Pearson R > 0.85), which was higher than that observed in other modules. A total of 1,359 genes were included in the GGM gene set (Table S27), of which 87 were transcription factor-encoding genes (Table S28). Among these, four amino acid transport-related genes, two CqNPF-coding genes, and three CqNRT2-coding genes were identified as target genes (Table S29). Correlated expression was observed between CqNRTs, other amino acid transport-related genes, and transcription factors within the same module (absolute R > 0.7 and p-value < 0.05) (Table S30-31). To narrow the scope of transcription factor screening, particularly concerning root growth, the researchers analyzed phenotypic traits and the activities of key enzymes involved in nitrogen metabolism. The results revealed significant changes in root phenotypes and physiological characteristics of quinoa under varying NO₃⁻ ion concentrations (Figs. 1, S1). These alterations in traits and physiological indicators are likely due to changes in the expression of key nitrogen metabolism genes triggered by low nitrogen stress, which in turn affects the physiological and phenotypic characteristics of quinoa roots. The correlations between the expression of transcription factors and target genes and various root phenotypes (such as fresh weight, dry weight, root length ratio, surface area, average diameter, and number of root tips) were analyzed, as well as the activities of key nitrogen metabolism enzymes (including GOGAT, GS, NiR, and NR). Additionally, the correlations between transcription factors/target genes and compositional changes during different seed developmental stages (including water content, seed size, total sugars, soluble sugars, starch, total proteins, soluble proteins, amino acids, lipids, flavonoids, and saponins) were also examined (Fig. S12). After filtering the results based on the criteria of an absolute R value > 0.7 and p-value < 0.05, both positive and negative correlations between at least three phenotypic parameters and three enzyme activities were identified in quinoa roots and seeds. As a result, 21 core regulatory transcription factors and 9 target genes were identified within the GGM module (Fig. S13). The final 21 transcription factors in the GGM were classified into 9 transcription factor families: bHLH, WRKY, MYB, MIKC_MADS, GRAS, NAC, HSF, bZIP, and NIN-like (Fig. 5A). Notably, the expression patterns of three NIN-like transcription factors—CqNLP2.2, CqNLP2.1, and CqNLP1.3 were highly consistent with those of the four NRT target genes (CqNPF2.10, CqNRT2.2, CqNRT2.1, and CqNRT2.3) identified in the GGM module. This consistency was most pronounced under the 0 mM NO₃⁻ nitrogen condition (Fig. 5A). Building on the above findings, this phenomenon aligns with previous studies reporting that NIN-like transcription factors can regulate the majority of NO₃⁻ signaling genes, thereby improving NUE in crops and consequently increasing tiller number and grain yield [33, 35, 36, 59]. To investigate further, the binding sites (TFBS) for the promoters of genes predicted to be involved in nitrogen metabolism within the GGM were analyzed. Notably, 14 types of cis-elements were identified, including 50 ABRE, 104 MYC, 67 STRE, 32 CGTCA motifs, 32 as-1, 32 TGACG motifs, and 62 ARE elements. Additional binding motifs for ERE, WUN, and GARE were also predicted within the promoter regions of these genes (Fig. S14). Based on these findings, the gene regulatory network (including transcription factors) in the GGM is hypothesized to be involved in nitrogen metabolism and amino acid transport (Fig. 5B).Fig. 5Construction of a co-expression network and identification of associated transcription factors (TFs). A Expression patterns of selected transcription factors and nitrogen metabolism-related genes across different tissue types and nitrogen treatment conditions in the Greenyellow gene module (GGM). B Co-expression network of the 21 transcription factors showing the highest connectivity within the GGM Construction of a co-expression network and identification of associated transcription factors (TFs). A Expression patterns of selected transcription factors and nitrogen metabolism-related genes across different tissue types and nitrogen treatment conditions in the Greenyellow gene module (GGM). B Co-expression network of the 21 transcription factors showing the highest connectivity within the GGM CqNLP2.1 is a key transcription factor in the NO₃⁻ signaling pathway To further investigate the activity of the quinoa CqNRT2.2 gene promoter, transient expression assays using 5’-deletion fragments of varying lengths were conducted in tobacco. GUS staining results revealed blue spots in all three biological replicates of tobacco leaves infiltrated with pCAMBIA3301-CqNRT2.2-PF/PF1/PF2/PF3/PF4/. The most intense blue staining was observed in leaves containing pCAMBIA3301-CqNRT2.2-PF (−2000 to 0 bp), pCAMBIA3301-CqNRT2.2-PF1 (−1800 to 0 bp), PF2 (−1369 to 0 bp), and PF3 (−809 to 0 bp), while pCAMBIA3301-CqNRT2.2-PF4 (−310 to 0 bp) exhibited only faint blue spots along the leaf margins. These results indicate that the PF1-PF3 fragments exhibit the strongest promoter activity, driving high expression of the GUS gene, suggesting that these regions may harbor key elements enhancing promoter activity. Binding site prediction analysis of the CqNRT2.2 promoter via the PlantTFDB database (https://plantregmap.gao-lab.org/binding_site_prediction.php) revealed that all PF1-PF3 fragments contain NLP transcription factor binding sites (Fig. 6B). Concurrent cis-acting element analysis identified multiple light-responsive core elements and environmental stress-responsive modules within these regions (Fig. 6A), highlighting the importance of NLP binding motifs, light-responsive elements, and environmental stress-responsive elements in the CqNRT2.2 promoter.Fig. 6The transcription factor CqNLP2.1/2.2 binds to the promoter of CqNRT2.2 and activates its expression. A Schematic of the CqNRT2.2 promoter deletion series: PF–PF4 represent constructs truncated at 0, 200, 631, 1181 and 1690 bp upstream of the transcription start site, respectively. B GUS staining of N. benthamiana leaves transiently expressing the GUS reporter gene driven by the CqNRT2.2 promoter. C Schematic of constructs used for dual-luciferase transient expression assays. The CqNRT2.2 promoter was inserted into the reporter vector, and CqNLP2.1 was cloned into the effector vector pGreenII 62-SK. Results of dual-luciferase assays in N. benthamiana leaves co-infiltrated with the effector and reporter constructs. Luciferase activity was measured 48 h after infiltration. D Sequence of the CqNRT2.2 promoter region used in yeast one-hybrid assays. The predicted CqNLP-binding motif is highlighted in red, with mutated bases indicated below, We used the NLP binding site motif between 1000 and 1200 bp for alignment. E Yeast one-hybrid assay demonstrating binding of CqNLP2.1/2.2 to the CqNRT2.2 promoter. Interactions were assessed by yeast growth on selective medium (-Leu) supplemented with 200 ng/mL aureobasidin A (AbA). Serial dilutions (10⁰, 10⁻1, 10⁻2, 10⁻3) of yeast cultures are shown The transcription factor CqNLP2.1/2.2 binds to the promoter of CqNRT2.2 and activates its expression. A Schematic of the CqNRT2.2 promoter deletion series: PF–PF4 represent constructs truncated at 0, 200, 631, 1181 and 1690 bp upstream of the transcription start site, respectively. B GUS staining of N. benthamiana leaves transiently expressing the GUS reporter gene driven by the CqNRT2.2 promoter. C Schematic of constructs used for dual-luciferase transient expression assays. The CqNRT2.2 promoter was inserted into the reporter vector, and CqNLP2.1 was cloned into the effector vector pGreenII 62-SK. Results of dual-luciferase assays in N. benthamiana leaves co-infiltrated with the effector and reporter constructs. Luciferase activity was measured 48 h after infiltration. D Sequence of the CqNRT2.2 promoter region used in yeast one-hybrid assays. The predicted CqNLP-binding motif is highlighted in red, with mutated bases indicated below, We used the NLP binding site motif between 1000 and 1200 bp for alignment. E Yeast one-hybrid assay demonstrating binding of CqNLP2.1/2.2 to the CqNRT2.2 promoter. Interactions were assessed by yeast growth on selective medium (-Leu) supplemented with 200 ng/mL aureobasidin A (AbA). Serial dilutions (10⁰, 10⁻1, 10⁻2, 10⁻3) of yeast cultures are shown The NLP transcription factor family plays a central role as putative key regulators, integrating plant nitrogen nutrient signaling with growth and developmental processes. CqNLP2.1/2.2 and CqNRT2.2 exhibit a coordinated expression pattern and were concurrently identified within the Greenyellow module. The co-expression network analysis revealed that CqNRT2.2 is a target gene of the CqNLP2.1/2.2 transcription factors and is positively regulated by them. This suggests that CqNLP2.1/2.2 are directly involved in regulating the expression of the CqNRT2.2 gene (Fig. 5). To further investigate, A dual-luciferase reporter system was established using the Renilla luciferase gene as the reporter (driven by the CqNRT2.2 promoter) and CqNLP2.1 as the effector. In tobacco leaves injected with the pGreenII 0800-LUC + pGreenII 62-SK-CqNLP2.1 combination, the luciferase activity ratio did not change significantly compared to the pGreenII 0800-LUC + pGreenII 62-SK control group. In contrast, tobacco leaves injected with the pGreenII 0800-LUC-CqNRT2.2 + pGreenII 62-SK-CqNLP2.1 combination exhibited a significantly increased luciferase activity ratio compared to the pGreenII 0800-LUC-CqNRT2.2 + pGreenII 62-SK control group. The fluorescence imaging results were consistent with the quantitative data (Fig. 6C), indicating that CqNLP2.1 activates the CqNRT2.2 promoter. To confirm direct binding, yeast one-hybrid (Y1H) assays were performed. Promoter fragments of CqNRT2.2 containing the NLP binding sites were cloned into the pBait-AbAi plasmid and integrated into the genome of the Y1H Gold yeast strain using homologous recombination. The results showed that CqNLP2.1/2.2 could directly bind to the core motifs within the promoter of CqNRT2.2 (fragments marked in red in Fig. 6D). As a negative control, site-directed mutagenesis of the predicted core region within the promoter completely abolished the binding of CqNLP2.1/2.2 proteins to the promoter (Fig. 6E). Overexpression of CqNRT2.2 and CqNLP2.1 in Arabidopsis enhances root growth under low NO₃⁻ After harvesting T0 seeds from transformed A. thaliana, three generations of antibiotic selection combined with PCR screening were performed to obtain non-segregating T3 homozygous seeds. Total RNA was extracted from Arabidopsis, reverse-transcribed, and analyzed using quantitative real-time PCR (qRT-PCR). qRT-PCR confirmation of transgene expression in the overexpression lines is shown in Fig. 7C, F. To elucidate the physiological functions of CqNRT2.2 and CqNLP2.1 in NO₃⁻ metabolism, growth phenotypes of wild-type (Col), CqNRT2.2 overexpression lines, and CqNLP2.1 overexpression lines were evaluated after 12 days of treatment on media containing different NO₃⁻ concentrations (0/4/10 mM NO3–) (Fig. 7A, B, D, E). Compared to wild-type lines, the overexpression lines of CqNRT2.2 and CqNLP2.1 displayed longer roots, more lateral roots, and overall improved growth, with statistically significant differences observed between the wild-type and overexpression lines. Notably, overexpression lines grown under NO₃⁻-supplemented conditions (4/10 mM NO3–) exhibited better growth than those grown under 0 mM NO3– (Fig. 7A, D). Subsequent qRT-PCR analysis of carbon and nitrogen metabolism-related genes in Arabidopsis overexpressing CqNRT2.2 and CqNLP2.1 revealed up-regulation of all tested genes compared to wild-type lines. Among these, Rubisco and GLT3 showed the most significant up-regulation (Fig. 7C, F).Fig. 7Functional validation of CqNRT2.2 and CqNLP2.1 in NO₃⁻ response. A Phenotypic comparison of wild-type (Col-0) and three independent CqNRT2.2-overexpression lines (OE1, OE3, OE5) under NO₃⁻ treatment. B Primary root length of wild-type and CqNRT2.2-OE seedlings grown on media containing 0, 4, or 10 mM NO₃⁻. C Heatmap showing expression profiles of carbon- and nitrogen-metabolism genes in wild-type and CqNRT2.2-OE plants under 4 mM NO₃⁻. D Phenotype of wild-type (Col-0) and CqNLP2.1-overexpression lines (OE3, OE6, OE9) under NO₃⁻ treatment. E Primary root length of wild-type and CqNLP2.1-OE seedlings under 0, 4, or 10 mM NO₃⁻. F Heatmap of carbon- and nitrogen-metabolism gene expression in wild-type and CqNLP2.1-OE plants under 4 mM NO₃⁻. Data are presented as mean ± SD (n = 5 biological replicates). Statistical significance was determined by one-way ANOVA followed by Tukey’s HSD post-hoc test. Different letters indicate significant differences at p < 0.05. Scale bar = 1 cm Functional validation of CqNRT2.2 and CqNLP2.1 in NO₃⁻ response. A Phenotypic comparison of wild-type (Col-0) and three independent CqNRT2.2-overexpression lines (OE1, OE3, OE5) under NO₃⁻ treatment. B Primary root length of wild-type and CqNRT2.2-OE seedlings grown on media containing 0, 4, or 10 mM NO₃⁻. C Heatmap showing expression profiles of carbon- and nitrogen-metabolism genes in wild-type and CqNRT2.2-OE plants under 4 mM NO₃⁻. D Phenotype of wild-type (Col-0) and CqNLP2.1-overexpression lines (OE3, OE6, OE9) under NO₃⁻ treatment. E Primary root length of wild-type and CqNLP2.1-OE seedlings under 0, 4, or 10 mM NO₃⁻. F Heatmap of carbon- and nitrogen-metabolism gene expression in wild-type and CqNLP2.1-OE plants under 4 mM NO₃⁻. Data are presented as mean ± SD (n = 5 biological replicates). Statistical significance was determined by one-way ANOVA followed by Tukey’s HSD post-hoc test. Different letters indicate significant differences at p < 0.05. Scale bar = 1 cm Molecular docking predicts NO₃⁻ binding by CqNLP2.1 and CqNLP2.2 To investigate the evolutionary conservation of the RWP-RK gene family in quinoa, a phylogenetic tree was constructed using the NJ method with 1000 bootstrap replicates, based on 125 RWP-RK-related proteins from 7 species (Table S32). The genetic evolutionary relationships revealed that the proteins could be classified into two main categories: NLPs and RKDs (Fig. 8A). Notably, key functional residues such as H404, L406, and Y436 were identified as essential for AtNLP7 to bind to NO₃⁻ in A. thaliana [29]. To explore the role of CqNLPs in the NO₃⁻ signaling pathway in quinoa, the NO₃⁻-binding sites of CqNLP2.1 and CqNLP2.2 were compared to those of AtNLP7 (NreA). The results showed that CqNLP2.1/2.2 exhibit high conservation at the amino acid residue sites of H404, L406, and Y436 within the NO₃⁻-binding pocket of the NreA NO₃⁻ sensor (Fig. 8B, C), suggesting that CqNLP2.1/2.2 may have a potential role in NO₃⁻ sensing, based on computational predictions. Models generated by AlphaFold 3 indicated that the NO₃⁻-binding domain of CqNLP2.1/2.2 closely resembles that in AtNLP7 (Fig. 8D). The comparative analysis of homologous proteins across multiple species further confirmed that the NO₃⁻-sensing regions of CqNLP2.1/2.2 are highly conserved (Fig. 8E), providing additional evidence for the function of CqNLP2.1/2.2 as NO₃⁻ receptors. Within the NO₃⁻-binding domain of CqNLP2.1/2.2, strong interactions were observed between NO₃⁻ ions and specific amino acid residues, notably PHE-429 (2.7 Å), SER-441 (2.7 Å), ARG-445 (2.1 Å), PHE-403 (2.2 Å), THR-401 (2.7 Å), SER-411 (2.1 Å), and ARG-415 (2.1 Å) (Table S33). These residues form stable hydrogen bonds with the NO₃⁻ ion (Fig. 8F, G). Notably, these residues are located within the NO₃⁻-binding pocket of quinoa CqNLP2.1/2.2 (Fig. 8C). According to the molecular docking analysis, it is possible that CqNLP2.1/2.2 are involved in nitrate signal perception via potential hydrogen bond interactions between amino acid residues within their predicted nitrate-binding pocket and the nitrate ion.Fig. 8Prediction of the response of NLP subfamilies to NO₃⁻. A Phylogenetic tree of RWP–RK proteins constructed using the neighbor‑joining method. B The NO₃⁻‑binding domains of CqNLP2.1/2.2 in quinoa are homologous to those of AtNLP7 in Arabidopsis thaliana. Homologous regions are highlighted in red, with numerical values indicating specific amino acid positions. C Sequence alignment of the NO₃⁻‑binding domains of AtNLP7 and CqNLP2.1/2.2. Boxes of the same color represent conserved residues, while boxes of different colors denote semi‑conserved residues. Three critical NO₃⁻‑binding residues are outlined with red boxes. D Comparison of the three‑dimensional NO₃⁻‑binding structures of CqNLP2.1/2.2 proteins with the structural domain of AtNLP7. E The NO₃⁻‑binding region is highly conserved within the NLP‑1 subfamily, with red boxes highlighting the key NO₃⁻‑binding residues. F‑G Molecular docking simulations of CqNLP2.1/2.2 with NO₃⁻ ions were performed using Autodock software. The spherical models represent NO₃⁻ ions, yellow dashed lines indicate hydrogen‑bond interactions, and the purple stick structures represent the key residues forming hydrogen bonds with CqNLP2.1/2.2 Prediction of the response of NLP subfamilies to NO₃⁻. A Phylogenetic tree of RWP–RK proteins constructed using the neighbor‑joining method. B The NO₃⁻‑binding domains of CqNLP2.1/2.2 in quinoa are homologous to those of AtNLP7 in Arabidopsis thaliana. Homologous regions are highlighted in red, with numerical values indicating specific amino acid positions. C Sequence alignment of the NO₃⁻‑binding domains of AtNLP7 and CqNLP2.1/2.2. Boxes of the same color represent conserved residues, while boxes of different colors denote semi‑conserved residues. Three critical NO₃⁻‑binding residues are outlined with red boxes. D Comparison of the three‑dimensional NO₃⁻‑binding structures of CqNLP2.1/2.2 proteins with the structural domain of AtNLP7. E The NO₃⁻‑binding region is highly conserved within the NLP‑1 subfamily, with red boxes highlighting the key NO₃⁻‑binding residues. F‑G Molecular docking simulations of CqNLP2.1/2.2 with NO₃⁻ ions were performed using Autodock software. The spherical models represent NO₃⁻ ions, yellow dashed lines indicate hydrogen‑bond interactions, and the purple stick structures represent the key residues forming hydrogen bonds with CqNLP2.1/2.2 Discussion Unveiling NRTs in quinoa: phylogenetic insights and expression profiles Quinoa (Chenopodium quinoa Willd.), an annual crop in the Chenopodium genus of the Amaranthaceae family, is highly regarded for its rich nutritional composition and adaptability. The cultivation yield and nutritional quality of quinoa are strongly influenced by nitrogen availability. NO3– is a primary nitrogen source absorbed by plants, and as such, plants rely on a suite of genes responsible for the transport, assimilation, and regulation of NO₃⁻, which are essential for maintaining nitrogen homeostasis and ensuring efficient nitrogen utilization for growth and development [60]. Building upon the characterization of NRTs in various plant species, including Arabidopsis and rice, 114 CqNPFs and 15 CqNRT2s were identified in quinoa [18, 19, 25, 61]. Comparative analysis reveals significant divergence in NPF and NRT2 family members across plant species, with A. thaliana containing 53 NPFs and 7 NRT2 genes [62, 63]. All CqNPF and CqNRT2 proteins were predicted to contain conserved motifs for NO3– transport, though variations in transmembrane domains and gene structures were observed across different CqNRT members. The classification of the NPF subfamily has been a matter of ongoing discussion, with difering approaches proposed by various researchers. Wittgenstein [64] divided the NPF family into 10 supergroups based on the phylog-eny of NPF from 20 plant species. In this study, we follow the classification proposed by Léran et al. [14], which split the NPF family into 8 subfamilies across various specie. However, our analysis revealed some lineage-specific expansions and rearrangements within the NPF family that do not fully align with this classifications (Fig. 2A). Moreover, all NRT2 genes contain a conserved and several transmembrane domains (Table S10). CqNRT2.2 clustered into the same clade of Group II with AtNRT2.1, AtNRT2.2, AtNRT2.3, AtNRT2.4, and AtNRT2.6, and showed relatively high homology with OsNRT2.1, OsNRT2.2, and OsNRT2.3 in rice. This phylogenetic relationship suggests that CqNRT2.2 may also function as a component of the HATS, similar to its Arabidopsis and rice counterparts. In Arabidopsis, AtNRT2.1, AtNRT2.2, AtNRT2.4, and AtNRT2.5 govern the HATS [16–19]. Meanwhile, four NRT2 transporters in rice have been identified as critical components of the HATS for nitrate uptake [21, 23, 24, 65–67]. Among these, OsNRT2.1 is the primary transporter responsible for nitrate uptake, impacting root elongation by regulating auxin transporter genes [21, 67]. OsNRT2.2 is involved in nitrate uptake in both roots and seeds [23]. The nitrate transport function of OsNRT2.3b is responsive to variations in cellular pH levels [24]. In contrast to AtNRT2.4, OsNRT2.4 exhibits a dual affinity for nitrate [66]. Spatiotemporal expression analysis of CqNRT genes revealed that CqNRT2.2 was expressed at significantly higher levels than other family members, with the highest expression observed in roots under nitrogen-deficient conditions, highlighting its potential key role in NO₃⁻ uptake (Fig. 3). Given its strong root expression and upregulation under N deficiency, CqNRT2.2 may also play a role in NO₃⁻ signaling, a possibility explored further in the discussion. Additionally, several CqNPF and CqNRT2 genes showed distinct expression changes under varying nitrogen conditions, suggesting their involvement in the regulation of nitrogen acquisition and allocation. These findings support a model in which root NO₃⁻ uptake is modulated through integrated root–shoot–root signaling [68]. Functional identification of CqNRTs in a yeast heterologous expression system Among the 114 CqNPF and 15 CqNRT2 genes screened, 11 CqNPF and 3 CqNRT2 genes were identified as functional in nitrate uptake based on complementation assays in the triple mutant yeast strain △ynt1 (△ynt1, △leu) (Fig. 4). Although plants and yeast possess distinct nitrate transport and assimilation systems, prior studies have shown that the GS/GOGAT pathway contributes to glutamate synthesis and nitrogen assimilation. This is particularly evident in Saccharomyces cerevisiae, where mutant strains lacking glutamate dehydrogenase (GDH) genes (gdh1Δ2Δ3Δ) have been used to elucidate the functional overlap [69]. The redundant nitrate assimilation pathways—including GDH, the GS/GOGAT cycle, and proline utilization (PUT)—collectively ensure efficient conversion of nitrate to glutamate in yeast [70]. While these genes complemented the yeast mutant to some extent (allowing growth on nitrate), the extent of growth rescue did not show clear discrimination between HATS and LATS candidates under our assay conditions. This could be due to functional redundancy from endogenous yeast assimilation pathways (GDH, GS/GOGAT, PUT), or alternatively, these transporters may require additional post-translational modifications or interacting partners not present in yeast. CqNLP2.1 directly regulating the expression of CqNRT2.2 Through WGCNA analysis and promoter binding site prediction, CqNLP2.1 and CqNLP2.2 are predicted to regulate the CqNRT2.2 gene in quinoa, suggesting their potential involvement in the nitrate signaling pathway (Fig. 5). Previous studies have reported that AtNLP7 in Arabidopsis affects the expression of target genes by binding to the promoters of AtNRT1.1 and AtNRT2.1 [33]. Furthermore, it has been reported that in rice, OsNLP3 regulates the expression of OsNRT1.1B, OsNRT1.1C, OsNRT2.1, OsNRT2.3A, and OsNRT2.4, thereby increasing tiller number, grain yield, and nitrogen use efficiency [36]. In our study, yeast one-hybrid assays confirmed that both CqNLP2.1 and CqNLP2.2 directly bind to the core NLP-binding motif within the CqNRT2.2 promoter. Consistently, GUS staining revealed that fragments containing the NLP-binding motif exhibited higher promoter activity. Furthermore, transient expression assays demonstrated that CqNLP2.1 activates the CqNRT2.2 promoter and enhances luciferase reporter activity (Fig. 6). These results suggest that CqNLP2.1 may act as a key transcription factor in the quinoa nitrate signaling network, regulating growth and development through transcriptional control of downstream PNR genes. Overexpression of CqNRT2.2 and CqNLP2.1 in Arabidopsis improved growth phenotypes relative to wild-type plants, with particularly notable changes in root architecture (Fig. 7). While these results demonstrate the functional capacity of both genes to promote root growth in a heterologous system, they do not provide definitive proof of their endogenous roles in quinoa. The observed phenotypes may instead reflect a general enhancement of nitrogen assimilation rather than quinoa-specific functions. Future studies using quinoa mutants or gene silencing are required to confirm their native functions. As such, their expression levels could serve as reliable molecular indicators for nitrogen fertilizer demand in quinoa cultivation, presenting significant potential for enhancing NUE, reducing production costs, and minimizing environmental pollution risks. Molecular docking and structural modeling predict that CqNLP2.1 and CqNLP2.2 may bind NO₃⁻ ions via conserved residues similar to AtNLP7 [29]. However, these are according to the molecular only. Experimental validation (e.g., mutagenesis and binding assays) is required to confirm their NO₃⁻-sensing function. Putative signaling events for NO₃⁻ and low-nitrogen perception NO₃⁻ serves as a key signaling molecule in plants, orchestrating regulatory processes that range from immediate physiological responses to long-term developmental adaptations [9]. A key component of the NO₃⁻ signaling pathway is the NO₃⁻ transceptor CHL1/NRT1.1/NPF6.3. This dual-affinity NRT also facilitates auxin transport, thereby linking NO₃⁻ perception to the regulation of root growth through auxin signaling [20]. CHL1 has been shown to regulate genes involved in auxin biosynthesis, such as TAR2, and genes like LAX3, a root-specific auxin transporter [71]. In this study, several quinoa NRTs were identified, including CqNPF2.13, CqNPF6.1, CqNRT2.2, CqNRT2.3, CqNRT2.1, and CqNRT2.6, all of which are located within the Greenyellow module (Fig. 5). These genes are specifically expressed in roots and are significantly upregulated under nitrogen-deficient conditions, suggesting their critical roles in NO₃⁻ uptake and signal transduction, consistent with previous research (Figs. 3, S8). NO₃⁻ signaling operates through a multilayered regulatory mechanism. On one hand, the calcium channel CNGC15 modulates the activity of NRT1.1 through dynamic protein interactions in response to fluctuating NO₃⁻ concentrations [72]. On the other hand, NO₃⁻-induced nuclear translocation of NLP7 controls the expression of most NO₃⁻-responsive genes and works in conjunction with NLP6 and TCP20 to activate key genes like NR [73]. The calcium signaling components CIPK23 and CBL1/9 regulate the NO₃⁻ affinity of the dual-affinity receptor NRT1.1 via phosphorylation [20, 74]. Upon NO₃⁻ perception, NRT1.1 activates downstream calcium signaling, which is sensed and transduced by CPKs such as CPK10, CPK30, and CPK32. This leads to the phosphorylation of the transcription factor NLP7, which remains in the nucleus for extended periods, activating downstream NO₃⁻-responsive target genes. Furthermore, NLP7 interacts with transcription factors like NRG2 to synergistically upregulate genes involved in nitrogen metabolism [32]. Our genomic analysis of quinoa identified several calcium-dependent protein kinases (CqCPKs) with root-specific expression. Additionally, NLP family transcription factors CqNLP2.1, CqNLP2.2, and CqNLP1.3 were co-identified in the Greenyellow module (Fig. 6). Their expression patterns closely mirrored those of the calcium-dependent protein kinases. Promoter analysis of CqNLP genes revealed an enrichment of multiple cis-acting elements, such as W boxes, ABREs, and MYB binding sites, suggesting that other transcription factors may cooperate in NO₃⁻ signal transduction in quinoa (Fig. S14). Concurrently, quinoa CqCNGCs (cyclic nucleotide-gated ion channels) exhibited high tissue specificity and environmental responsiveness, being expressed in roots, floral organs, and leaves. These CNGCs likely function in perceiving extracellular signals and mediating the flux of secondary messengers like Ca2+, playing a pivotal role in plant growth as well as NO₃⁻ uptake and transport. The recently identified NIGT1/HRS1/HHO type transcription factors have been recognized as core regulators in the plant nitrogen starvation response (NSR), integrating into nitrogen signaling networks to modulate nitrogen uptake and assimilation [75]. In this study, several CqHRS1/HHO1 genes showed root-specific expression, with particularly high levels observed under nitrogen stress (e.g., AUR62017788). This expression pattern aligns with observations in other crops, suggesting that CqHRS1/HHO1 may function as a core NSR transcription factor, working in concert with other regulatory factors. Research on NO₃⁻-mediated transcriptional networks has advanced with the identification of key transcription factors. TGA1 and TGA4, members of the plant-specific TGA subfamily of bZIP transcription factors, were among the first core components discovered [76]. Subsequently, factors such as NAC4 [77] and the Arabidopsis NO₃⁻-regulated protein ANR1 [78] were integrated into the regulatory framework. Notably, within the Greenyellow module, CqbZIP (AUR62023129), two CqNACs (AUR62027390, AUR62031305), and two CqANR1s (AUR62023162, AUR62014383) displayed expression patterns highly consistent with the identified NRTs, showing elevated expression in roots under nitrogen starvation. Additionally, the expression patterns of CqSPL, CqLBD, and CqANR1 in quinoa were similar to those of CqNPF3.1, CqNPF6.13, CqNPF6.12, and CqNPF1.5 from the Greenyellow module, with peak expression in spikes during flowering. Other critical transcription factors in the NO₃⁻ regulatory network include LOB domain-containing proteins 37 and 38 (LBD37/38) [79], and NO₃⁻-regulated gene 2 (NRG2) [80]. In this study, the expression patterns of CqLBD and CqNRG2 also mirrored those observed for the NRTs in the Greenyellow module. Based on these findings, a hypothetical signaling model for NO₃⁻ and low-nitrogen perception in quinoa is proposed. By analogy with the Arabidopsis NRT1.1-CNGC15 module, NO₃⁻ signals might be transduced through a potential interaction between CqNRT2.2 and a quinoa CNGC15 homolog. However, direct evidence for such an interaction is still lacking. Notably, the NO₃⁻ transporter CqNRT2.2 may also contribute to signaling, but whether this is independent of its transport activity remains to be determined. CqNLP then cooperates with a variety of transcription factors, including CqNRG2, CqANR1, CqLBD, CqTCP, CqbHLH, CqbZIP, CqSPL, CqTGA, CqNAC, and CqHRS1/HHO1, to collectively upregulate genes involved in nitrogen metabolism. The initial events triggering the perception of low nitrogen conditions remain unclear, but transcription factors identified in this study, such as CqTCP, CqLBD, and CqHRS1/HHO1, are likely to play central roles in the NSR. It should be emphasized that the proposed model is primarily based on expression patterns, phylogenetic relationships, and promoter analyses; experimental validation (e.g., transport assays, protein–protein interaction, and mutant analysis) is required to confirm the assigned roles. Unlike Arabidopsis, which employs the dual-affinity NRT1.1 as a primary NO₃⁻ sensor and signaling hub, quinoa appears to rely more heavily on NRT2-type transporters, particularly CqNRT2.2, for both nitrate uptake and potentially signaling. Whether quinoa has lost a functional NRT1.1 ortholog, or whether CqNRT2.2 has evolved to acquire signaling capacity independent of its transport activity, remains an open question. This putative reliance on NRT2.2 could represent an adaptive feature for quinoa’s growth in dryland environments, where NO₃⁻ availability is often low and fluctuating, potentially favoring high-affinity uptake systems over dual-affinity sensors. Therefore, although our model proposes a potential signaling role for CqNRT2.2, it remains equally possible that its primary function is nitrate uptake and that any signaling effects are indirect. Future studies using transport-dead mutants of CqNRT2.2 are needed to dissect its transport and signaling functions (Fig. 9).Fig. 9Putative signaling events for NO₃⁻ and low-nitrogen perception and downstream gene expression. The figure is adapted and modified from [8, 81–83] Putative signaling events for NO₃⁻ and low-nitrogen perception and downstream gene expression. The figure is adapted and modified from [8, 81–83] Conclusions This study systematically identified and characterized the CqNPF and CqNRT2 gene families in quinoa. We confirmed that 11 CqNPFs and 3 CqNRT2s exhibit NO₃⁻ transporter activity. In addition, the CqNLP subfamily is involved in the NO₃⁻ response, with CqNLP2.1/2.2 showing potential for direct NO₃⁻ binding. CqNLP2.1 in particular plays a potential role in regulating NO₃⁻ signaling and modulates quinoa growth and development by controlling the NO₃⁻-responsive gene CqNRT2.2. We propose a signaling model for NO₃⁻ and low-nitrogen perception involving CqNRT2.2 and downstream gene regulation in quinoa. This model provides a theoretical foundation for reducing nitrogen fertilizer use and improving nitrogen use efficiency in quinoa cultivation. Furthermore, this study offers valuable genetic resources for enhancing quinoa’s nitrogen utilization efficiency. Supplementary Information Supplementary Material 1. Supplementary Material 2. Supplementary Material 1. Supplementary Material 2.