The transcription factor OsMADS61 positively regulates root elongation and nitrogen use efficiency in rice

Abstract

Plant roots have evolved adaptive strategies mediated by transcriptional networks to cope with fluctuating nitrogen (N) forms and availability. However, the mechanisms linking root-foraging responses to N use efficiency (NUE) in crops remain poorly understood. Here, we show that rice exhibits enhanced root elongation under nitrate compared with ammonium, particularly under low N supply, suggesting a specific regulatory role for nitrate in root morphogenesis. We identify the transcription factor OsMADS61 as a key regulator of nitrate-dependent root morphological and physiological responses, as well as NUE, especially under N-limited conditions. OsMADS61 acts as a transcriptional activator of nitrate metabolism by directly binding to OsNRT2.1 and OsNR2 promoters. Nitric oxide produced via the nitrate reductase pathway, under the control of nitrate-responsive OsMADS61, precisely triggers cell proliferation in the root meristem. Moreover, single-nucleotide polymorphisms in the OsMADS61 promoter may be associated with differential root-foraging responses to nitrate availability. Therefore, enhancing N-adaptive root responses to optimize N uptake and assimilation represents a promising strategy for breeding crops with high NUE.

This study reports that rice exhibits enhanced root elongation under nitrate compared with ammonium, particularly under low nitrogen conditions. OsMADS61 functions as a key transcriptional activator that regulates nitrate metabolism (via OsNRT2.1 and OsNR2) and nitric oxide–mediated root meristem cell proliferation, thereby linking nitrate-dependent root foraging to nitrogen use efficiency (NUE) and providing a promising target for breeding high-NUE crops.

Introduction

Nitrogen (N) is one of the main mineral nutrients limiting crop production (Xu et al., 2012; Wang et al., 2018). The nitrogen use efficiency (NUE) of modern crops is typically low, and high yields currently depend on unsustainable and excessive levels of fertilization, potentially resulting in severe environmental pollution, climatic fluctuations, and biodiversity loss (Fang et al., 2006). Therefore, enhancing crop NUE, particularly under low N supply, represents an effective strategy to address these challenges. NUE is inherently complex; in crop plants, its two most important physiological components are N uptake efficiency and N utilization efficiency (Xu et al., 2012). Typically, NUE is higher under low N supply than under high N supply (Xu et al., 2012). Variations in NUE among crop plants grown under low N supply are primarily dictated by changes in N uptake efficiency, which largely depend on adaptive root system responses to fluctuating N availability (Le Gouis et al., 2000). However, little is known about the regulators that orchestrate N uptake and root-adaptive regulation.

Plant roots sense external environmental N signals and exhibit pronounced morphological and physiological plasticity; thus, improving root adaptive traits under different N regimes, especially under N limitation, has been proposed as a second green revolution in agricultural production (Lynch, 2007). Decreasing N supply consistently leads to an increased root-to-shoot ratio across plant species (Sun et al., 2020, 2023), thereby optimizing nutrient access and acquisition. Induced root elongation has been observed in several plant species in response to deficiencies of the two major inorganic N sources (Sun et al., 2014; 2016; 2021; 2023; Gao et al., 2015; Jia and Von Wirén, 2020; Jia et al., 2022; Xie et al., 2023). High-affinity N transporters and subsequent N assimilation pathways have evolved to modulate root uptake capacity and compensate for spatial and temporal variation in N availability (Xu et al., 2012; Wang et al., 2018). Underlying these N-regulated adaptive root-foraging responses are transcription factors (TFs), which play key roles in regulating nutrient uptake capacity by controlling metabolic genes and root morphology (Xuan et al., 2017; Gaudinier et al., 2018; Jia and Von Wirén, 2020; Liu et al., 2022). The NO₃₋-inducible Arabidopsis MADS-box transcription factor AtANR1 was the first identified regulator linking localized NO₃⁻ supply to root elongation (Zhang and Forde, 1998). Subsequently, numerous TFs involved in N-dependent regulatory networks and pathways—most of which regulate N uptake and assimilation—have been identified and extensively characterized in Arabidopsis and crop plants (Delgado et al., 2024; Zhang et al., 2024; Ruffel et al., 2025). Despite the importance of root development in crop responses to N availability, only a limited number of TFs, including OsMADS25/27/57 (Yu et al., 2015; Huang et al., 2019; Pachamuthu et al., 2022; Wu et al., 2023), OsSPL14/17 (Sun et al., 2021), and OsNLP1/3/4 (Alfatih et al., 2020; Wu et al., 2021; Zhang et al., 2022), have been shown to modulate root adaptation to soil N fluctuations. However, the transcriptional mechanisms linking root adaptive responses to NUE remain poorly understood. Further research is needed to identify key genes and favorable alleles for developing crop varieties with higher NUE to support sustainable agriculture.

Rice (Oryza sativa L.), one of the most important staple food crops worldwide, is usually cultivated under flooded conditions. Although ammonium (NH₄⁺) is the primary N source in paddy fields, rice roots are also exposed to nitrate (NO₃⁻) due to nitrification in the rhizosphere (Li et al., 2008). Increasing evidence indicates that approximately 40% of assimilated N in rice is in the form of NO₃⁻; moreover, NO₃⁻ plays a crucial role in enhancing rice yield (Li et al., 2008; Hu et al., 2015; Fan et al., 2016; Wang et al., 2018). To decipher the molecular machinery underlying rice adaptive responses to low N supply, we performed transcriptome analyses of rice roots subjected to two inorganic N forms (NH₄⁺ and NO_3⁻) at two concentrations (0.2 and 2.5 mM). This analysis led to the identification of a rice TF, OsMADS61, that orchestrates N metabolism and root adaptive regulation. OsMADS61, which is specifically induced by low NO₃⁻ supply, improves rice grain yield and NUE by promoting root elongation and NO₃⁻ uptake and assimilation. Under low NO₃⁻ supply, OsMADS61 enhances NO₃⁻ reductase activity in rice roots, leading to increased nitric oxide (NO) production that facilitates root elongation. Collectively, these results demonstrate that OsMADS61 is involved not only in remodeling root architecture under low NO₃⁻ supply but also in enhancing rice yield and NUE under both low and high N supply, making it a promising target for molecular breeding.

Results

OsMADS61 specifically regulates rice root elongation under low NO₃⁻ supply

To investigate the molecular mechanisms governing rice developmental adaptation to low N stress, rice plants (cv. Hwayoung) were treated with NH₄⁺ or NO₃⁻ (at concentrations of 0.2 or 2.5 mM) for 21 days. Seminal roots were longer in rice plants grown with NO₃⁻ than in those grown with NH₄⁺ at the same level of N supply (Figures 1A and 1B). Nitrogen-responsive root elongation (NRRE), defined as the ratio of seminal root length under low N supply to that under high N supply, was higher in NO₃⁻-treated plants than in NH₄⁺-treated plants (Figure 1C), implying that NO₃⁻ exerts a more positive effect on rice root architecture than NH₄⁺. Root elongation is determined by meristematic cell division and subsequent cell elongation. Therefore, rice root tips within 1 mm of the apex were collected for RNA sequencing (RNA-seq) under the four N treatments (Supplemental Figure 1A). The analysis revealed that 34 TFs were downregulated, whereas only one TF (Os04g0461300) was specifically upregulated under low NO_3⁻ supply relative to both NH₄⁺ supplies and 2.5 mM NO_3⁻ (Figure 1D; Supplemental Figure 1B). This low-NO_3⁻-induced TF belongs to the AGL17-like family, which features a MIKC domain, and was designated OsMADS61; however, inconsistent transcript annotations have been reported in the RAP (Os04g0461300), MSU (LOC_Os04g38770/LOC_Os04g38780), and NCBI (LOC4336057) databases. Alignment of OsMADS61 protein sequences with AGL17-like proteins from other plant species indicated that only LOC4336057 in the NCBI database (240 amino acids in the open reading frame) encodes a protein containing both MADS-box and K-box domains, leading to the conclusion that LOC4336057 represents the correct transcript (Supplemental Figure 2A). An RT–PCR assay was performed using a sense primer targeting the initiation codon of OsMADS61 and an antisense primer targeting the termination codon of LOC4336057, which corresponds to the longest transcript. The 723-bp cDNA amplified by RT–PCR was fully consistent with LOC4336057 based on sequence alignment (Figure 1E; Supplemental Figure 2B), suggesting that LOC4336057 in the NCBI database is the correct annotation for OsMADS61. Thereafter, a 720-bp cDNA was used to construct fusion proteins for subcellular localization analysis of OsMADS61. Both the OsMADS61–GFP and GFP–OsMADS61 fusion proteins were localized to the nucleus in tobacco leaves (Supplemental Figure 3A). RT–qPCR analysis further showed that OsMADS61 expression was induced by low NO_3⁻ supply in roots but not in shoots, with transcript levels approximately fivefold higher in roots than in those in aboveground tissues (Supplemental Figure 3B), supporting a role for OsMADS61 in mediating root responses under low NO_3⁻ supply.

Figure 1.

OsMADS61 is involved in low-nitrate regulation of rice root elongation.

(A–C) Root phenotype (A), seminal root length (B), and N-responsive root elongation (NRRE) (C). Seven-day-old Hwayoung seedlings (WT) were transferred to hydroponic medium with NH₄⁺ or NO_3⁻ supply for 21 days. Scale bar, 5 cm.

(D) RNA sequencing analysis showing differentially expressed transcription factors in root tips treated with NO_3⁻ or NH₄⁺ at 0.2 or 2.5 mM for 14 days. Vs. indicates the latter-to-former ratio (fold change); up, upregulated transcription factors with log₂ (fold change) >0.75; normal, common to both treatments (absolute value of log₂ [fold change] <0.75).

(E) OsMADS61 transcripts annotated in various public databases. Agarose gel electrophoresis shows the sizes of OsMADS61 CDSs amplified from various rice tissues by RT–PCR.

(F–K) Root phenotype (F and I), seminal root length (G and J), and NNRE (H and K) of WT and osmads61 (mads61-1 and mads61-2) seedlings under NO_3⁻ supply (F–H) or NH₄⁺ supply (I–K) for 21 days. Scale bar, 5 cm.

Data are presented as mean ± SD (n = 12 in B, G, and J; n = 3 in C, H, and K). Different letters denote significant differences (P < 0.05, Duncan’s multiple range test). P values were generated using two-tailed Student’s t-tests.

To clarify the biological function of OsMADS61, we generated two osmads61 mutants (mads61-1 and mads61-2) in the background of the wild-type (WT) cultivar Hwayoung using CRISPR technology (Supplemental Figure 4A). A shorter seminal root length was observed in both osmads61 mutant lines relative to the WT, exclusively under low NO_3⁻ supply (Figures 1F and 1G). Furthermore, NRRE was significantly reduced in mads61-1 (88%) and mads61-2 (68%) compared with the WT under NO_3⁻ supply (Figure 1H); similar trends were observed under NH₄⁺ supply (Figures 1I–1K). No differences were detected in adventitious root number or lateral root density between the WT and osmads61 mutants under any of the four N treatments tested (Supplemental Figures 5B and 5C). we also generated three OsMADS61 overexpression lines (OE1, OE2, and OE3; Supplemental Figure 4B) and found that both seminal root length and adventitious root number were increased in the OE lines compared with the WT (Supplemental Figures 5A and 5B). Together, these results demonstrate that OsMADS61 acts as a positive regulator of rice root elongation in response to low NO_3⁻ supply.

OsMADS61 enhances NUE and grain yield

To evaluate whether OsMADS61 could serve as a candidate gene for improving agronomic traits in rice, we cultivated OsMADS61 transgenic lines under flooded field conditions at two N rates (Figures 2A–2G, 75 kg N ha⁻¹, LN [low Nitrogen]; Figures 2H–2N, 250 kg N ha⁻¹, HN [high Nitrogen]). Grain yield was significantly higher in the OE lines than in the WT, with a greater increase under LN (40%, Figure 2D) than HN conditions (16%, Figure 2K). This increase was largely attributable to improved tiller number, 1000-grain weight, and seed-setting rate in the OE lines (Figures 2D–2G and 2K–2N; Supplemental Figures 6C, 6D, and 6G–6I). In contrast, grain yield was reduced in osmads61 mutants compared with the WT under both N supplies, mainly due to decreases in 1000-grain weight and seed-setting rate (Figures 2E, 2F, 2L, and 2M).

Figure 2.

Increased OsMADS61 abundance boosts grain yield and NUE in rice.

(A) Appearance of rice plants at maturity under 75 kg N ha⁻¹. WT, osmads61 mutants (mads61-1 and mads61-2), and OsMADS61 overexpression lines (OE1, OE2, and OE3) were grown under flooded field conditions. Scale bar, 30 cm.

(B and C) Absolute (B) and proportional (C) N distribution of rice plants shown in (A).

(D–G) Grain yield per plant (D), seed-setting rate (E), 1000-grain weight (F), and tiller number development (G) of rice plants shown in (A).

(H) Appearance of rice plants at maturity under 250 kg N ha⁻¹. Scale bar, 30 cm.

(I–J) Absolute (I) and proportional (J) N distribution of plants shown in (H).

(K–N) Grain yield per plant (K), seed-setting rate (I), 1000-grain weight (M), and tiller number (N) of rice plants shown in (H).

Data are presented as mean ± SD (n = 5 in B, C, I, and J; n = 20 in D–F and K–M; n = 36 in G and N). ∗ or ∗∗ indicate significant differences in N accumulation or distribution between WT and OsMADS61 transgenic plants (P < 0.05 or P < 0.01, two-tailed Student’s t-test). Different letters denote significant differences (P < 0.05, Duncan’s multiple range test).

Total N accumulation in aboveground tissues was significantly higher in OsMADS61 OE lines than in the WT under both N supplies, largely due to increased N accumulation in rice grains (Figures 2B and 2I). In parallel, enhanced root elongation was observed in the OE lines compared with the WT under both N supplies (Supplemental Figures 6A, 6B, 6E, and 6F). Furthermore, similar N distribution ratios in rice grains between WT and OsMADS61 OE plants under both N supplies implied that the positive modulation of rice grain yield by OsMADS61 might result from its coordinated regulation of root development and N uptake (Figures 2C and 2J). Conversely, the osmads61 mutants exhibited lower total N accumulation and grain N distribution, as well as reduced root elongation, compared with the WT, confirming that compromised root elongation substantially limits rice N uptake capacity.

OsMADS61 regulates root development by controlling meristem size

To elucidate the biological role of OsMADS61 in regulating rice root elongation, we generated transgenic rice plants expressing the GUS reporter driven by the OsMADS61 promoter (∼2.5 kb). As shown in Figure 3B, the strongest GUS activity was detected in rice roots supplied with 0.2 mM NO_3⁻ relative to those supplied with 2.5 mM NO_3⁻ or NH₄⁺, consistent with the RT–qPCR results (Supplemental Figure 3B). Importantly, GUS activity assays and RT–qPCR analyses indicated that OsMADS61 is highly expressed in the meristematic region (0–1 mm from the root tip) and the mature zone (≥ 3 mm from the root tip), whereas lower expression was observed in the elongation zone (1–3 mm from the root tip) (Figures 3A and 3C; Supplemental Figure 7). This spatial expression pattern of OsMADS61 in rice roots in response to N availability and form indicates a potential role in N-dependent modulation of rice development.

Figure 3.

OsMADS61 regulates root elongation by controlling meristem size.

(A) Relative expression of OsMADS61 in the 0–1 mm region from the root tips of Hwayoung seedlings.

(B and C)GUS staining of pOsMADS61:GUS transgenic seedlings. (B) Root tip. Scale bar, 1 mm. (C) Longitudinal root section under 0.2 mM NO_3⁻. Scale bar, 100 μm. Seven-day-old seedlings were transferred to hydroponic medium containing NH₄⁺ or NO_3⁻ for 14 days. Red and yellow arrowheads indicate stem cells and the beginning of the transition zone, respectively.

(D and E) Root phenotype (D) (scale bar, 5 cm) and confocal images of root tips (E) (scale bar, 100 μm). Seven-day-old seedlings were transferred to hydroponic medium containing 0.2 mM NO_3⁻ and grown for 21 days. Red and yellow arrowheads indicate stem cells and the beginning of the transition zone, respectively.

(F–H) Seminal root length, meristem length, and cell number in the root meristem shown in (E).

Data are presented as mean ± SD (n = 4 in A; n = 12 in F–H). Different letters denote significant differences (P < 0.05, Duncan’s multiple range test).

Root elongation is determined by cell division within the root meristem and subsequent cell elongation outside the meristem (Beemster et al., 2003). To clarify how OsMADS61 promotes root elongation under low NO_3⁻ supply relative to the WT (Figures 3D and 3F; Supplemental Figures 5 and 8), we examined cellular changes underlying root elongation by measuring cell proliferation in the meristematic zone and cortical cell length in the mature zone. Under low NO_3⁻ supply, mutation of OsMADS61 resulted in reduced meristem length and cell number, whereas cortical cell length in the mature zone was comparable to that in WT plants. In contrast, OsMADS61 OE lines exhibited increased meristem length and cell number (Figures 3E, 3G, and 3H; Supplemental Figure 9). These results suggest that OsMADS61 promotes rice root elongation under low NO_3⁻ supply, primarily by enhancing cell division in the root meristem rather than by affecting cell elongation.

OsMADS61 acts upstream of OsNRT2.1 to promote NO₃⁻ uptake

Given that OsMADS61 expression was linked to N accumulation in rice plants grown under field conditions, we next examined its role in modulating N metabolism. Rice plants were grown hydroponically under two NO_3⁻ concentrations for 21 days, and an approximately 20% reduction in dry weight was observed in WT plants under 0.2 mM NO_3⁻ compared with 2.5 mM NO_3⁻ (Supplemental Figure 10A). As expected, lower N content was detected in both the roots and shoots of osmads61 mutants only under 0.2 mM NO_3⁻ supply relative to the WT, whereas no difference was observed under 2.5 mM NO_3⁻. Conversely, higher N content was recorded in the roots and shoots of OsMADS61 OE lines relative to the WT under both NO_3⁻ concentrations (Figure 4A; Supplemental Figure 10B). We then performed a time-course uptake experiment using ¹⁵N-labeled NO_3⁻ to further elucidate the role of OsMADS61 under low NO₃⁻ supply (Figure 4B; Supplemental Figure 10C). Reduced root ¹⁵N accumulation was observed in osmads61 mutants, whereas increased accumulation was detected in OE lines at all time points examined (5, 30, and 60 min). A similar trend was observed in shoots at 30 and 60 min, suggesting that OsMADS61 enhances NO_3⁻ uptake by roots and promotes root-to-shoot translocation. To confirm the role of OsMADS61 in NO₃⁻ uptake, net NO_3⁻ fluxes were measured in seminal root tips using the high-resolution scanning ion-selective electrode technique (SIET; Figures 4C and 4D). Throughout the 6.5-min measurement period, an approximately 58% decrease in net NO₃⁻ influx was recorded in the mutants and an approximately 63% increase was observed in the OE lines compared with WT plants at three distances from the root tip, suggesting that changes in influx rates in transgenic lines supplied with 0.2 mM NO_3⁻ were consistent with the results of ¹⁵N-NO_3⁻ uptake.

Figure 4.

OsMADS61 regulates rice NO₃⁻ uptake by binding to the OsNRT2.1 promoter.

(A) N content in rice roots. Seven-day-old WT, osmads61 mutants (mads61-1 and mads61-2), and OsMADS61 overexpression lines (OE1, OE2, and OE3) were transferred to hydroponic medium containing 0.2 or 2.5 mM NO_3⁻ for 21 days.

(B)¹⁵N content in rice roots. Seven-day-old rice seedlings were treated with 0.2 mM NO_3⁻ for 14 days, deprived of N for 3 days, and then exposed to 0.2 mM [¹⁵N]NO_3⁻ for 5, 30, or 60 min.

(C and D) Net NO_3⁻ fluxes in the meristem of rice seminal roots supplied with 0.2 mM NO_3⁻ for 6.5 min (C) and mean NO_3⁻ flux rates at three distances from the root tip during the measurement period (D).

(E) Relative expression of OsNRT2.1 in seedling roots treated with 0.2 mM NO_3⁻ for 14 days.

(F) Yeast one-hybrid assays examining interactions between OsMADS61 and the OsNRT2.1 promoter.

(G) ProUbi:OsMADS61–FLAG-mediated ChIP–qPCR enrichment (relative to input) of CArG-containing promoter fragments from OsNRT2.1. ProUbi:FLAG transgenic plants were used as controls.

(H) EMSA analysis of OsMADS61 protein binding to the OsNRT2.1 promoter. The OsMADS61–His fusion protein was expressed in E. coli. Probes were labeled with biotin.

(I) Transactivation analysis in rice protoplasts.

(J)¹⁵N content in rice roots. Seven-day-old rice seedlings were treated with 0.2 mM NO_3⁻ for 14 days, deprived of N for 3 days, and then exposed to 0.2 mM [¹⁵N]NO_3⁻ for 60 min.

Data are presented as mean ± SD (n = 3). Different letters denote significant differences (P < 0.05, Duncan’s multiple range test). ∗∗ indicates a significant difference in relative enrichment between ProUbi:OsMADS61–FLAG lines and control plants (P < 0.01, two-tailed Student’s t-test).

To elucidate how the TF OsMADS61 modulates N metabolism via binding to target genes, we performed chromatin immunoprecipitation sequencing (ChIP-seq) experiments. OsNRT2.1 and OsNRT2.2, which encode high-affinity NO₃⁻ transporters, exhibited over twofold enrichment among genes associated with NO_3⁻ uptake (Supplemental Figure 11). Additionally, OsNRT2.1 displayed expression patterns similar to those of OsMADS61 in response to N supply; for example, both genes showed comparable expression under 0.2 or 2.5 mM NH₄⁺ but were strongly induced under 0.2 mM NO_3⁻ relative to 2.5 mM NO_3⁻ (Supplemental Figure 12). Importantly, compared with WT plants, OsNRT2.1 transcript levels were significantly decreased in osmads61 mutants and enhanced in OsMADS61 OE lines (Figure 4E). Moreover, GUS activity assays in transverse sections of rice roots indicated that OsMADS61 was expressed in most root cell types (Supplemental Figure 13), similar to the expression pattern reported for OsNRT2.1 (Feng et al., 2011).

MADS-box TFs recognize CArG motifs present in the promoters of target genes (Guo et al., 2013). A putative OsMADS61-binding CArG-box site (CTATAAATAG) was identified at −143 to −152 bp upstream of the OsNRT2.1 transcription start site (Supplemental Figure 14A). Yeast one-hybrid assays demonstrated that OsMADS61 binds to the OsNRT2.1 promoter (Figure 4F), and this interaction was further confirmed by ChIP assays (Figure 4G). Electrophoretic mobility shift assays (EMSAs) showed that OsMADS61 specifically binds to the CArG-box element within the OsNRT2.1 promoter (Figure 4H). Collectively, these results indicate a direct physical interaction between OsMADS61 and the OsNRT2.1 promoter. To assess the ability of OsMADS61 to activate OsNRT2.1 transcription, we conducted dual-luciferase (LUC) assays in planta. The OsNRT2.1 promoter was fused to a LUC reporter gene and co-transformed into rice protoplasts with constructs expressing OsMADS61. In the presence of OsMADS61, LUC activity driven by the OsNRT2.1 promoter was markedly enhanced relative to the control, suggesting that OsMADS61 transactivates OsNRT2.1 expression (Figure 4I).

To further verify OsNRT2.1 as a target gene of OsMADS61, we attempted to generate an nrt2.1/nrt2.2 double mutant, given the complete sequence identity of the coding sequences (CDSs) of OsNRT2.1 and OsNRT2.2. However, viable seeds of the nrt2.1/nrt2.2 mutants could not be obtained. Therefore, an OsNRT2.1 OE line was generated in both the WT and osmads61 mutant backgrounds. Because OsNRT2.1 has been shown to positively regulate NO_3⁻ uptake (Chen et al., 2016), we analyzed root ¹⁵N accumulation in the osmads61 mutant, the OsNRT2.1 OE line (NRT2.1OE), and OsNRT2.1-overexpressing mads61-1 plants (mads61-1/NRT2.1OE) under 0.2 mM NO_3⁻ (Figure 4J). OsNRT2.1 overexpression significantly improved NO₃⁻ uptake in mads61-1/NRT2.1OE plants compared with the osmads61 mutant. These results confirm that OsMADS61 acts upstream of OsNRT2.1 to regulate NO_3⁻ uptake.

OsMADS61 acts upstream of OsNR2 to modulate N assimilation and seminal root elongation

Nitrate is first reduced to NO₂⁻ by the cytoplasmic enzyme nitrate reductase (NR) within the cell, a step that appears to be rate-limiting in the N assimilation pathway (Campbell, 2001). ChIP-seq analysis showed that, in addition to OsNRT2.1/2.2, OsNR2 exhibited more than twofold enrichment among genes associated with NO_3⁻ metabolism (Supplemental Figure 11). Moreover, OsNR2 and OsMADS61 displayed similar expression patterns in response to N supply (Supplemental Figure 12). Reduced OsNR2 transcript abundance and NR activity were observed in osmads61 roots, whereas increased transcript levels and NR activity were observed in the roots of OsMADS61 OE lines relative to the WT (Figures 5A and 5B), indicating that OsMADS61 modulates NR activity at the transcriptional level. We thus examined the distribution of CArG elements in the OsNR2 promoter (Supplemental Figure 14B). Five putative OsMADS61-binding CArG-box sites (P1–P5) were identified, three of which contained an identical CArG motif (CTAAAATTTG) located at −1622 to −1613 bp (P1, NR2-1a), −1051 to −1042 bp (P3, NR2-1b), and −706 to −697 bp (P4, NR2-1c) upstream of the translation start site. Yeast one-hybrid assays confirmed that OsMADS61 directly binds only to the P3 site (Figure 5C; Supplemental Figure 14C). This interaction was further confirmed by ChIP assays, which verified specific binding of OsMADS61 to the CArG box at the P3 site (Figure 5D). EMSAs further validated the physical interaction between OsMADS61 and OsNR2 (Figure 5E). We also tested the transcriptional activation potential of OsMADS61 using dual-LUC assays in planta. In the presence of OsMADS61, LUC activity driven by the OsNR2 promoter was substantially enhanced compared with the control, suggesting that OsMADS61 transactivates OsNR2 expression (Figure 5F).

Figure 5.

OsMADS61 transactivates OsNR2.

(A and B) Relative expression of OsNR2(A) and relative nitrate reductase (NR) activity (B) in rice roots. Seven-day-old WT, osmads61 mutants (mads61-1 and mads61-2), and OsMADS61 overexpression lines (OE1, OE2, and OE3) were transferred to hydroponic medium containing 0.2 mM NO_3⁻ and grown for 14 days.

(C) Yeast one-hybrid assays testing the interaction between OsMADS61 and OsNR2.

(D) OsMADS61–FLAG–mediated ChIP–qPCR enrichment (relative to input) of CArG-containing promoter fragments from OsNR2. ProUbi:FLAG transgenic plants were used as controls.

(E) EMSA analysis of OsMADS61 binding to the OsNR2 promoter. The OsMADS61–His fusion protein was expressed in E. coli. Probes were labeled with biotin.

(F) Transactivation assays performed in rice protoplasts.

(G and H) Root phenotype (G) and N-responsive root elongation (NRRE) (H) of nr2 mutants after 21 days. Scale bar, 5 cm. Seven-day-old nr2 mutants in the WT or OsMADS61 overexpression line (OE1) background were transferred to hydroponic medium containing NH₄⁺ or NO_3⁻ for 21 days.

(I)¹⁵N content in rice roots. Seven-day-old rice seedlings were treated with 0.2 mM NO_3⁻ for 14 days, deprived of N for 3 days, and then exposed to 0.2 mM [¹⁵N]NO_3⁻ for 60 min.

(J) Seminal root length of rice seedlings supplied with 0.2 mM NO_3⁻ for 21 days.

Data are presented as mean ± SD (n = 3 in A, B, D, F, H, and I; n = 12 in J). ∗∗ indicates a significant difference in relative enrichment between ProUbi:OsMADS61–FLAG lines and control plants (P < 0.01, two-tailed Student’s t-test). Different letters denote significant differences (P < 0.05, Duncan’s multiple range tests). P values were generated using two-tailed Student’s t-tests.

Using CRISPR technology, we generated an nr2 mutant in the WT (Hwayoung) background and subsequently obtained an nr2 mutant in the OsMADS61 OE background through hybridization and molecular validation (Supplemental Figure 14D). Reduced root ¹⁵N accumulation was observed in nr2 mutant plants compared with their respective WTs (Figure 5I). Notably, seminal root length in the nr2 mutant did not respond to low NO_3⁻ supply relative to 2.5 mM NO_3⁻, resulting in an NRRE near zero under both NO_3⁻ supplies (Figures 5G and 5H; Supplemental Figure 15). In contrast, the NRRE of nr2 plants under both NH₄⁺ supplies was comparable to that of WT plants (Figures 1C and 5H). Furthermore, mutation of OsNR2 markedly suppressed seminal root elongation in the OsMADS61 OE background (Figure 5J). Consistent with previous reports (Gao et al., 2019), nr2 mutants exhibited lower grain yields compared with their respective WT plants (Supplemental Figure 16A). Together, these results indicate that OsMADS61 acts upstream of OsNR2 to regulate N metabolism and seminal root elongation.

NO accumulation is involved in OsMADS61-induced root elongation

Previous studies have shown that NR, a major enzyme responsible for NO production in plants, is involved in the regulation of root elongation in rice (Sanchez-Corrionero et al., 2023). Given the above finding that OsNR2 participates in OsMADS61-modulated seminal root elongation, we hypothesized that OsMADS61-induced seminal root elongation is mediated by NO. The extremely short half-life of NO currently hinders precise quantification of NO content in plants; thus, relative NO levels were assessed using a fluorescent probe, 4-amino-5-methylamino-2′,7′-difluorofluorescein diacetate (DAF-FM). We analyzed NO-associated fluorescence in the root tips of plants exposed to 0.2 mM NO_3⁻ for 0, 5, 30, and 60 min (Figures 6A and 6B). Compared with WT plants, DAF-FM fluorescence was significantly decreased in the root tips of osmads61 mutants and significantly increased in those of the OE lines within 5 min of treatment, consistent with changes in relative NR activity and OsNR2 transcript abundance in rice roots (Supplemental Figure 17).

Figure 6.

NO directly promotes root elongation.

(A and B) NO production in root tips shown as green fluorescence (A) and quantified as fluorescence intensity (B). Seven-day-old WT, osmads61 mutants (mads61-1 and mads61-2), and OsMADS61 overexpression lines (OE1, OE2, and OE3) were transferred to hydroponic medium containing 0.2 mM NO_3⁻ for 14 days, deprived of N for 3 days, and then resupplied with 0.2 mM NO_3⁻ for 5, 30, and 60 min. Scale bar, 1 mm.

(C and D) NO production in root tips shown as green fluorescence (C) and quantified as fluorescence intensity (D). Seven-day-old seedlings were grown in nutrient solution containing 0.2 mM NO_3⁻ without (mock) or with sodium nitroprusside (10 μM), carboxy-PTIO (cPTIO; 80 μM), or tungstate (Tu; 25 μM) for 14 days. Scale bar, 1 mm.

(E) Seminal root elongation of seedlings after 14 days of the treatments shown in (C).

(F) Linear regression analysis between seminal root elongation and DAF-FM fluorescence shown in (C).

Data are presented as mean ± SD (n = 10). Different letters denote significant differences (P < 0.05, Duncan’s multiple range tests).

Application of the NO donor sodium nitroprusside markedly increased DAF-FM fluorescence and seminal root elongation in both osmads61 and WT plants, leading to no detectable difference between the two genotypes under 0.2 mM NO_3⁻ (Figures 6C–6E). Conversely, treatment with the NO scavenger carboxy-PTIO or the NR inhibitor tungstate reduced DAF-FM fluorescence and seminal root elongation in WT and OE lines, ultimately resulting in similar levels between the WT plants and osmads61 mutants. Lower DAF-FM fluorescence was observed in the root tips of nr2 mutants relative to their respective WT plants (Supplemental Figure 18), consistent with decreased seminal root elongation (Figure 5J). Similarly, stronger DAF-FM fluorescence was detected in the root tips of OsNRT2.1 OE lines compared with their respective WT plants, paralleling increased seminal root elongation (Supplemental Figure 19).

NO has been reported to function as a signaling molecule with a dose–response relationship (Sanchez-Corrionero et al., 2023). Notably, plotting seminal root length formed during the experimental period against NO-associated fluorescence in WT and OsMADS61 transgenic plants revealed a strong correlation between seminal root elongation and NO content in root tips (Figure 6F), indicating that OsMADS61 regulates seminal root elongation in an NO-dependent, linear manner. These results suggest that OsMADS61 enhances NR-dependent NO production in rice root tips, thus promoting seminal root elongation.

Genetic variation and geographical distribution of OsMADS61 alleles

Given the prominent role of OsMADS61 in rice root development and NO_3⁻ metabolism, we examined its genetic variation using rice germplasm from the Rice3K database by analyzing nucleotide diversity and fixation index (F_ST) values for OsMADS61 and its flanking regions. Aus and its wild progenitor (O. nivara) exhibited higher nucleotide polymorphism than other Asian rice subgroups and their corresponding wild progenitors (Figure 7A). In addition, higher F_ST values for Aus vs. indica or Aus vs. japonica were detected in OsMADS61 and its flanking regions than in other comparisons, indicating that genetic variation in OsMADS61 likely arose after the divergence of Aus from O. nivara (Figure 7B). Two haplotypes (HapA and HapB) were identified in the CDS of OsMADS61, comprising four single-nucleotide polymorphisms (SNPs) located in the C-domain region (Figure 7D; Supplemental Figure 20A). However, only minor differences were detected between the two haplotypes at the C-terminal end of the predicted proteins (Figure 7D). Because the C domain of MIKC-type MADS proteins is generally considered less functionally critical, and because seminal root length and N content were similar between the two haplotypes under low NO_3⁻ supply (Supplemental Figures 20B and 20C), OsMADS61 was considered a conserved protein. Haplotype analysis of the OsMADS61 promoter identified three haplotypes (HapI, HapII, and HapIII) (Figure 7C). To determine whether genetic variation in the OsMADS61 promoter affects its transcriptional activity, the HapI and HapIII promoters were amplified and inserted upstream of the LUC gene in the pGreenII 0800-LUC vector. Dual-LUC assays showed that HapI exhibited higher transcriptional activity than HapIII (Supplemental Figure 21). Notably, HapIII was highly enriched in the Aus subgroup, with a frequency of 83% (Figure 7E). Thirty accessions representing HapI and HapIII were randomly selected and grown hydroponically under 0.2 mM NO_3⁻ for 14 days. Consistently, HapI accessions exhibited higher OsMADS61 expression, accompanied by increased seminal root length and N content, compared with HapIII accessions (Figures 7F–7H).

Figure 7.

Genetic variation in OsMADS61.

(A and B) Nucleotide diversity (A) and mean F_ST(B) of OsMADS61 and its flanking regions among rice subspecies and two wild accessions (Ruf, O. rufipogon; Niv, O. nivara). Gray dashed lines indicate the ATG and TGA positions of OsMADS61, respectively. Red dashed lines indicate significant differences among groups (B).

(C) Haplotypes of the OsMADS61 promoter and their positions relative to the ATG start codon (Rice Variation Map v2.0). Minor-allele frequency (MAF) ≥0.1 was used as the screening criterion for all variant sites; heterozygous and deletion sites were removed.

(D) Alignment of protein structures encoded by the two coding-region haplotypes of OsMADS61. Protein structures were predicted using AlphaFold3.

(E) Geographic distribution of OsMADS61 promoter haplotypes among rice varieties.

(F–H) Relative expression of OsMADS61 in roots (F), seminal root length (G), and N content (H) in HapI and HapIII accessions. Seven-day-old rice seedlings were transferred to hydroponic medium containing 0.2 mM NO_3⁻ for 21 days.

Data are presented as mean ± SD (n = 30 in F–H). P values were generated using two-tailed Student’s t-tests.

Discussion

The application of N fertilizers to crops has dramatically increased over the past half century; therefore, the development of crop varieties with improved NUE, particularly those with higher N uptake efficiency under low N supply, is critical for sustainable agriculture. As sessile organisms, plants continuously adapt to spatiotemporal N fluctuations through dynamic perception of available N forms, regulation of N transport and metabolic processes, and modification of root system architecture to sustain growth (Jia and Von Wirén, 2020). However, little is known about the regulators that orchestrate N uptake and root-adaptive regulation associated with NUE. In this study, we found that rice exhibits enhanced root elongation under NO₃⁻ relative to NH₄⁺; this differential response is particularly pronounced under low N conditions, underscoring the unique regulatory role of NO₃⁻ in root morphogenesis. OsMADS61 was identified as a specific regulator of root morphological and physiological adaptation to low NO₃⁻ supply, which ultimately affects rice grain yield and NUE, especially under limited N availability in paddy fields.

OsMADS61 belongs to the MIKC-type MADS-box superfamily, one of the most extensively studied TF families in plants. Functionally, MADS proteins play key roles in modulating diverse biotic and abiotic stress responses, ranging from pathogen resistance to environmental stresses such as drought, salinity, and cold, as well as in developmental processes including the maturation of roots, leaves, floral organs, and fruits (Zhang et al., 2024). Several studies have demonstrated regulatory roles for MADS family members in N-modulated root development. In Arabidopsis, AtAGL17, AtAGL21, and AtANR1—three of the four members of the AGL17-like clade—are preferentially expressed in roots (Zhang and Forde, 1998; Gan et al., 2005; Han et al., 2008; Yu et al., 2014). AtAGL21 plays a crucial role in sustaining lateral root growth under NO_3⁻ deficiency (Yu et al., 2014), whereas AtANR1 promotes lateral root elongation under localized NO_3⁻ supply (Zhang and Forde, 1998; Gan et al., 2012). In maize, ZmTMM1 (an ortholog of AtANR1) restores defective lateral root growth in the Arabidopsis anr1/agl21 double mutant (Liu et al., 2020). In rice, OsMADS23, OsMADS25, OsMADS27, and OsMADS57—four of the five members of the AGL17-like clade—are induced by NO_3⁻ supply in roots (Puig et al., 2013). OsMADS25 promotes primary and lateral root growth in the presence of NO_3⁻, suggesting a role for this AGL17-like gene in root nutrient foraging (Yu et al., 2015; Wu et al., 2023). Mutation of OsMADS57 inhibits seminal and adventitious root elongation, mainly under low NO_3⁻ conditions (Huang et al., 2019). Notably, the monocot-specific microRNA miR444 inhibits rice lateral root growth in a NO_3⁻-dependent manner, primarily by targeting OsMADS27 (Yan et al., 2014; Pachamuthu et al., 2022). To our knowledge, N-modulated root adaptive responses mediated by a MADS family protein that is directly associated with NUE have not been previously reported, highlighting the unique role of OsMADS61 in rice.

Plants have evolved adaptive strategies mediated by transcriptional networks to cope with and survive environmental challenges. Systemic signaling enables roots to respond to the overall N status of the plant and provides critical input for root-foraging responses (Guan et al., 2014). OsMADS61 functions as a transcriptional activator that regulates N form–dependent physical and morphological root responses by directly binding to the OsNRT2.1 and OsNR2 promoters. As primary N sources for plant metabolism, comparative analyses across monocot and dicot species have demonstrated preferential root elongation under NO_3⁻-enriched conditions compared with NH₄⁺-based environments (Kudoyarova et al., 1997; Cao et al., 2010; Manoli et al., 2013; Na et al., 2014; Sun et al., 2021; Ötvös et al., 2021). In the present study, substantial enhancement of root elongation was observed in rice plants supplied with 0.2 mM NO_3⁻ relative to 2.5 mM NO_3⁻, indicating that NO_3⁻ functions as a systemic signal that drives adaptive root responses under low NO_3⁻ supply and supports tolerance to abiotic stress. Importantly, OsMADS61 was identified as a causal regulator under NO_3⁻-limited conditions that integrates NO_3⁻ metabolism with adaptive root architecture. ChIP-seq analysis showed that OsNRT2.1/2.2 and OsNR2 exhibited more than twofold enrichment among genes associated with NO_3⁻ uptake and metabolism. Given that the expression pattern of OsNRT2.2 differed from that of OsMADS61, we performed additional experiments demonstrating that OsNRT2.1 and OsNR2 are downstream regulatory targets of OsMADS61. Although OsNRT2.1 is highly expressed throughout rice roots (Feng et al., 2011), transgenic plants harboring OsNRT2.1 under the control of a constitutive promoter did not show increased grain yield or NUE (Chen et al., 2016), consistent with our results (Supplemental Figure 16B). In this study, OE of OsMADS61 enhanced NUE mainly through coordinated regulation of NO_3⁻ uptake and downstream assimilation via modulation of OsNRT2.1 and OsNR2 expression, consistent with previous findings (Gao et al., 2019).

Several phytohormones have been reported to participate in rice root growth, among which auxin plays a predominant role (Sun et al., 2014; Chen et al., 2022). However, no significant enrichment of auxin-related genes was detected by ChIP-seq (data not shown). Increasing evidence indicates that NO is an important regulator of root elongation through its effects on cell proliferation and maintenance of the root apical meristem (Sun et al., 2016; Sanchez-Corrionero et al., 2023). Because the NR pathway is a major route for NO production (Gupta et al., 2022), multiple lines of evidence support the involvement of NO accumulation in OsMADS61-induced root elongation under low NO_3⁻ conditions. First, compared with WT plants, DAF-FM fluorescence was significantly reduced in the root tips of osmads61 mutants and significantly increased in OE lines within 5 min of treatment, consistent with changes in relative NR activity and OsNR2 transcript abundance in rice roots (Figure 6; Supplemental Figure 17). Second, application of the NO donor sodium nitroprusside markedly enhanced DAF-FM fluorescence and seminal root length in osmads61 and WT plants, eliminating differences between them under 0.2 mM NO_3⁻ (Figure 6). Conversely, application of an NO scavenger or an NR inhibitor decreased DAF-FM fluorescence and seminal root length in WT and transgenic lines, ultimately resulting in comparable levels between WT and osmads61 mutants. The relative expression of OsPIN1b in seminal root tips was consistent with NO-associated root elongation (Supplemental Figures 22A and 22B) and aligned with findings by Sun et al. (2018). Moreover, no enrichment exceeding twofold was detected in the OsPIN1b promoter based on ChIP-seq data (Supplemental Figure 22C), suggesting that OsMADS61 does not directly regulate OsPIN1b expression. Third, lower DAF-FM fluorescence was observed in the root tips of nr2 mutants (nr2 and MADS61OE/nr2) compared with their respective WT plants under 0.2 mM NO_3⁻, paralleling decreased seminal root elongation (Figure 5J; Supplemental Figure 18). In contrast, higher DAF-FM fluorescence was detected in the root tips of OsNRT2.1 OE lines (NRT2.1OE and mads61-1/NRT2.1 OE) relative to their respective WT plants, accompanied by increased seminal root elongation (Supplemental Figure 19). Collectively, these results indicate that OsMADS61 maintains appropriate NO levels to promote root elongation (Figure 6). Given that NO exerts dose-dependent effects on root development (Fernández-Marcos et al., 2011), only optimal regulation of NO levels effectively promotes root elongation (Correa-Aragunde et al., 2016); OsMADS61 appears to achieve this balance under low NO_3⁻ supply.

Plant-specific MIKC-type MADS-box proteins contain a MADS-box domain and three additional domains: a K (keratin) domain, a less conserved I (intervening) region, and a highly variable C-terminal region (Alvarez-Buylla et al., 2000). The protein structure of OsMADS61 is highly conserved in rice; although four SNPs are located in the C domain, none appear to affect key functional properties across more than 4700 rice accessions (Figure 7D; Supplemental Figure 20A). Haplotype analysis of the OsMADS61 promoter region revealed that HapIII, which is prevalent in the Aus subgroup, is associated with significantly shorter root length and reduced NUE. To develop high-yielding and high-quality hybrid super rice, distantly related Aus subgroups are widely used in crosses with indica and japonica rice (Chen et al., 2020). Notably, introduction of the HapIII genotype should be avoided in breeding programs aimed at improving NUE. Reducing N fertilizer application is a central objective of modern agricultural production systems, driven by the need for both economic viability and environmental sustainability. The identification of this N-responsive transcriptional network therefore provides a potential breeding strategy to reduce dependence on N fertilizers and to promote more efficient and sustainable agricultural development.

Methods

Plant materials

For gene disruption using the CRISPR–Cas9 system, gRNA constructs were generated to produce mads61-1, mads61-2, and nr2 mutants in the Hwayoung genetic background (WT). OsMADS61 OE plants (OE1, OE2, and OE3) were developed by introducing full-length OsMADS61 cDNA driven by the ubiquitin promoter into the WT background. To construct the fusion of the OsMADS61 promoter and the GUS CDS (pOsMADS61-GUS), a 2.5-kb region of the putative promoter (upstream of the ATG start codon of OsMADS61) was cloned into the GUS-pCAMBIA 1300GM vector. OsNRT2.1 OE lines (NRT2.1 OE) were obtained by introducing full-length OsNRT2.1 cDNA driven by its native promoter into the WT background. Transgenic rice plants were generated by Agrobacterium-mediated transformation, as previously described (Huang et al., 2019). All primers used for vector construction are listed in Supplemental Table 1.

Hydroponic culture conditions

Plants were grown in a greenhouse under natural light with day/night temperatures of 30°C/18°C. Rice seeds were surface-sterilized and germinated on half-strength Murashige and Skoog medium (PhytoTech Labs). Seven-day-old seedlings with uniform size and vigor were selected and transferred to tanks containing 8 l of IRRI nutrient solution (0.2 mM KH₂PO₄, 0.4 mM K₂SO₄, 1 mM CaCl₂·2H2O, 1 mM MgSO₄·7H₂O, 0.5 mM Na₂SiO₃·9H₂O, 20 μM Fe-EDTA, 9 μM MnCl₂·4H₂O, 0.39 μM Na₂MoO₄·2H₂O, 20 μM H₃BO₃, 0.77 μM ZnSO₄·7H₂O, and 0.32 μM CuSO₄·5H₂O [pH 5.5]). The nutrient solution was replaced with fresh solution every 2 days. NO_3⁻ and NH₄⁺ were supplied as Ca(NO₃)₂ and (NH₄)₂SO₄, respectively. To exclude potential effects of Ca²⁺ among treatments, solutions within the same experimental system were supplemented with CaCl₂ to match the Ca²⁺ concentration of the high-NO_3⁻ condition. To inhibit nitrification, 7 μM dicyandiamide was added to prevent NH₄⁺ oxidation. Pharmacological treatments were applied directly to the hydroponic medium using sodium nitroprusside (10 μM), carboxy-PTIO (80 μM), or tungstate (25 μM).

Field trial conditions

Three-week-old rice seedlings were transplanted to paddy fields in June and harvested in October. The field site was located in Baima, Nanjing (119°18′ E, 31°61′ N), under long-day conditions (13.5 h light) during summer. All plants, including transgenic lines (T2–T4), were cultivated at the same site with 20-cm spacing under two N application rates (75 and 250 kg ha⁻¹). N fertilizer (urea) was applied to the field 1 day before transplanting, at the tillering stage, and at the flowering stage, with 40%, 30%, and 30% of the total N applied at each respective stage.

Measurement of root system architecture and histological observation

Preliminary experiments showed that seminal and adventitious roots exhibited similar responses to N supply in both WT and transgenic rice plants; however, seminal roots were significantly longer than adventitious roots under the experimental conditions. Thus, seminal roots were selected as representative roots, and their lengths were measured using a ruler.

To analyze meristem length and cell number, longitudinal sections (50 μm) of root tips were prepared using a microtome (VT1200; Leica, Germany) after embedding the tissue in 5% agarose. Sections were stained with Calcofluor White (1:1000; Sigma, USA) for 5 min and observed using a laser scanning confocal microscope (TCS SP8; Leica, Germany). At the beginning of the transition zone, cell length shows a sudden increase compared with that of cells in the meristematic zone. Meristem length was defined as the distance from the stem cells to the transition zone. For cell counting, cells were manually labeled and counted using the Leica microscope with LAS X software.

RNA-seq analysis

Total RNA was extracted from root tip segments (∼1 mm) of 40 seedlings treated with NO_3⁻ or NH₄⁺ (0.2 or 2.5 mM) for 2 weeks, in accordance with the manufacturer’s instructions for TRIzol Reagent (Life Technologies, CA, USA). Three biological replicate RNA-seq libraries were prepared for each treatment. Sequencing was performed on the Illumina NovaSeq 6000 platform (Annoroad Gene Technology) using paired-end 150-bp reads, generating a total of 12 libraries. Low-quality reads, including adaptor-only reads, reads containing >5% unknown nucleotides, or reads with Q20 <20% (percentage of bases with sequencing error rates <1%), were removed using a Perl script. Clean reads were mapped to the rice reference genome IRGSP1.0 (https://rapdb.dna.affrc.go.jp/) using HISAT2. Aligned reads in BAM/SAM format were further processed to remove potential duplicate reads. After mapping, normalization was performed, and fragments per kilobase of exon per million mapped reads were calculated using StringTie. The PlantTFDB database was used to identify TFs, which are listed in Supplemental Table 3, based on comparisons of the fragments per kilobase of exon per million mapped reads values across three biological replicates. Venn diagrams were generated using custom R scripts. In this study, upregulated genes were defined as those with an absolute log₂ (fold change) >0.75 and adjusted P-value <0.05. Genes were considered not significantly different between treatments if the absolute log₂ (fold change) was below 0.75.

RNA extraction, cDNA synthesis, and RT–qPCR

Total RNA was extracted from various organs of rice plants grown under hydroponic conditions using TRIzol reagent (Invitrogen). First-strand cDNA was synthesized from total RNA using a cDNA synthesis kit (ACCURATE BIOTECHNOLOGY [HUNAN], Changsha, China; AG11728). RT–qPCR was performed on a QuantStudio 6 Flex Real-Time PCR System (Applied Biosystems, Thermo Fisher Scientific), in accordance with the manufacturer’s instructions (ACCURATE BIOTECHNOLOGY; AG11718). Relative transcript abundance was determined by normalization to the reference gene OsActin within the same sample. All primers used for RT–qPCR are listed in Supplemental Table 2.

Yeast one-hybrid assays

The Matchmaker Gold Yeast One-Hybrid Library Screening System and Yeastmaker Transformation System 2 kit (Takara Clontech) were used in accordance with the manufacturer’s instructions. A 30-bp fragment containing three tandem CArG boxes, or promoter fragments containing CArG boxes from the OsNRT2.1/2.2 or OsNR2 promoters, as well as mutated CArG-box sequences, were inserted into the pAbAi vector. Full-length OsMADS61 cDNA was cloned into the pGADT7AD vector to generate pGADT7AD–OsMADS61. After transformation, yeast cells were cultured in liquid medium to an OD₆₀₀ of 1 and then serially diluted tenfold (10⁻¹ to 10⁻³). From each dilution, a 4-μl aliquot was spotted onto solid synthetic dropout medium (−Leu) to select for plasmids and protein–DNA interactions. Media were supplemented with various concentrations (0–1000 ng/ml) of aureobasidin A to suppress background growth and assess interaction strength.

ChIP–qPCR and ChIP-seq analysis

ChIP–qPCR assays were performed using the EpiQuik Plant ChIP Kit (Epigentek, USA), in accordance with the manufacturer’s instructions. Briefly, 1.0 g of fresh rice seedlings carrying ProUbi:FLAG (control) or ProUbi:MADS61–FLAG was harvested for ChIP analysis. Samples were fixed with 1% (v/v) formaldehyde, after which chromatin was isolated and sheared by sonication (Bioruptor Pico, Diagenode) to generate DNA fragments with an average size of approximately 500 bp. Anti-FLAG monoclonal antibodies (Cell Signaling Technology, USA) immobilized on protein A/G–coated resin in strip wells were used to immunoprecipitate genomic DNA fragments. qPCR was performed using immunoprecipitated genomic DNA, and enrichment was calculated as the ratio of immunoprecipitated DNA to input DNA.

For ChIP-seq analysis, purified DNA from both immunoprecipitated and input samples was used for library construction. Sequencing was performed on the Illumina NovaSeq 6000 platform (Annoroad Gene Technology) using paired-end 150-bp reads for both immunoprecipitated and input libraries. After low-quality reads were filtered using fastp, clean reads were aligned to the rice reference genome IRGSP1.0 using Bowtie2. Peak calling was conducted using MACS3 with default parameters. Peaks were annotated using the R package ChIPseeker, and genes with peaks located within 3 kb upstream of the transcription start site were subjected to Gene Ontology and Kyoto Encyclopedia of Genes and Genomes enrichment analyses. Peaks were visualized using the Integrated Genomics Viewer (Broad Institute, MA, USA).

EMSA

The full-length CDS of OsMADS61 was cloned into the pET-28a vector and transformed into the Escherichia coli (E. coli) BL21 (DE3) pLysS strain to express the OsMADS61–His fusion protein. After induction with 0.5 mM isopropyl β-D-1-thiogalactopyranoside and incubation at 37°C for 5 h, the recombinant protein was purified using Ni–NTA agarose (Beyotime Biotech, P2210). EMSAs were performed with the Light Shift Chemiluminescent EMSA Kit (Beyotime Biotech, GS009) using 5′ biotin-labeled DNA probes synthesized by Sangon Biotech.

Transient expression assays in rice protoplasts

Approximately 2-kb promoter fragments of OsNRT2.1 and OsNR2 were amplified and inserted upstream of the firefly LUC gene in the pGreenII 0800-LUC vector. The effector construct contained the full-length OsMADS61 CDS driven by the CaMV35S promoter. Transient transactivation assays were performed using rice protoplasts, and luciferase activities were measured with the dual-LUC Reporter Assay System (TRANSGEN, FR201) to detect LUC and Renilla LUC (REN) signals.

Determination of total N content and ¹⁵N accumulation

Rice seedlings were grown in IRRI nutrient solution containing 0.2 or 2.5 mM NO_3⁻ for 14 or 21 days. Roots and shoots were harvested separately, rinsed with 0.1 mM CaSO₄ for 1 min, heated at 105°C for 30 min to inactivate enzymes, and dried at 70°C to constant weight. Total N content was determined using the Kjeldahl method.

For ^{15N accumulation analysis, r}ice seedlings were grown in IRRI nutrient solution containing 0.2 mM NO_3⁻ for 14 days, then deprived of N for 3 days. Plants were sequentially transferred to 0.1 mM CaSO₄ for 1 min, followed bycomplete nutrient solution containing 0.2 mM [¹⁵N]NO_3⁻ (atom % ¹⁵N: 99%, Shanghai Research Institute of Chemical Industry) for 5, 30, or 60 min, and then washed with 0.1 mM CaSO₄ for 1 min. Samples were ground and dried to constant weight at 70°C. Approximately 1 mg of dried powder from each sample was analyzed using an isotope ratio mass spectrometer (Thermo Fisher Scientific).

Net NO₃⁻ flux rate measurement using the SIET system

Rice seedlings were grown in IRRI nutrient solution containing 0.2 mM NO_3⁻ for 21 days and then deprived of N for 3 days. Net NO_3⁻ fluxes were measured at different root zones (300, 1200, and 2500 μm from the root tip) using the noninvasive SIET technique, as previously described (Huang et al., 2019). Seedling roots were equilibrated in the measuring solution for 30 min and then transferred to a measuring chamber containing solution with 0.2 mM NO_3⁻. Net NO_3⁻ flux was recorded for 6.5 min to minimize variability caused by transient fluctuations. Prior to flux measurements, ion-selective electrodes were calibrated using NO_3⁻ solutions of 0.05 and 0.5 mM. Each plant was measured once, and final flux values represent the means of five individual plants. The measuring solution contained 0.2 mM CaCl₂, 0.1 mM NaCl, 0.1 mM MgSO₄, and 0.3 mM MES (pH 6.0, adjusted with 1 M NaOH). Measurements were performed using the SIET system BIO-003A (Younger USA Science and Technology).

NO detection in root tips

NO was detected by DAF-FM staining and epifluorescence microscopy. Root tips were incubated with 10 μM DAF-FM in 20 mM HEPES–NaOH buffer (pH 7.5). After incubation in darkness for 30 min, root tips were washed three times with fresh buffer and immediately visualized using a stereomicroscope equipped with a color CCD camera; excitation was set at 488 nm and emission at 495–575 nm (Olympus MVX10). Green fluorescence intensity was quantified using ImageJ software. Data are presented as mean fluorescence intensities.

NR activity measurement

Rice roots were harvested and immediately frozen in liquid N₂. Frozen tissue was finely chopped and homogenized in 5 ml of extraction buffer containing 5 mM EDTA and 5 mM cysteine in 0.025 M phosphate buffer (pH 8.7); 2 mg ml⁻¹ NADH served as the electron donor. The mixture, together with a small amount of quartz sand, was transferred to a chilled mortar and thoroughly ground on ice until fully homogenized. The homogenate was transferred to a 10-ml centrifuge tube and centrifuged at 4000 rpm for 15 min at 4°C. The supernatant was used as the crude enzyme extract. For the assay, 1 ml of KNO₃, 0.6 ml of NADH, and 0.4 ml of the crude enzyme extract were combined in a separate centrifuge tube and incubated at 25°C for 30 min. The reaction was terminated by immediate addition of 0.5 ml sulfanilamide solution to neutralize excess NADH, followed by addition of 0.5 ml N-(1-naphthyl)ethylenediamine dihydrochloride. The mixture was centrifuged at 2000 rpm for 15 min, and the absorbance of the supernatant was measured at 540 nm using a microplate reader. A control reaction was performed by replacing NADH with 0.1 M phosphate buffer (pH 7.4). Nitrite (NO₂⁻⁾ production was calculated using a standard curve derived from a regression equation. NR activity was determined in the presence of 5 mM EDTA in both extraction and reaction mixtures. Relative NR activity was calculated by normalization to the WT control.

Genetic statistics and allelic variation analysis

SNP datasets of Asian rice (japonica, indica, Aus, and aromatic) and their wild progenitors (O. rufipogon and O. nivara), obtained from publicly available data from previous studies (Jing et al., 2023), were used to analyze genetic variation in OsMADS61 and its surrounding regions. Nucleotide diversity (π) and fixation index (F_ST) were calculated using a sliding window approach (window size: 400 bp; step size: 100 bp) for wild and cultivated rice subpopulations with VCFtools v0.1.16 (Danecek et al., 2011). Loess curve fitting was performed in R; π and F_ST values were plotted accordingly. Allele frequencies of SNPs in the OsMADS61 promoter region were calculated for 4726 varieties from six rice subgroups (japonica, indica I, indica II, indica III, aromatic, and Aus) and integrated into a geographic map based on subpopulation classification and origin information from the Rice Variation Map v2.0 (Zhao et al., 2021).

Funding

National Natural Science Foundation of China (grant nos. 32372815 and 31972501).

Acknowledgments

No conflict of interest declared.

Author contributions

Y.Z. conceived and supervised the project. D.W., Z.L., and Y.Z. designed the experiments. D.W., Z.L., and C.G. performed most of the experiments with assistance from Y.C., J.Z., D.W., and Y.Z. analyzed the data and wrote the paper.

Published: December 4, 2025