Nitrate-induced NLP1 SUMOylation regulates nitrate signaling and root nodulation

Significance

Legume nodulation enables biological nitrogen fixation but is strongly repressed by nitrate. NIN-like proteins (NLPs) mediate this nitrate response, yet how their activity is regulated remains unclear. Here, we demonstrate that SUMOylation—a reversible posttranslational modification—is essential for the transcriptional activity and protein–protein interactions of MtNLP1 in Medicago truncatula, independently of its nitrate-induced nuclear localization. This modification is conserved in other NLPs, including Arabidopsis thaliana NLP7. Moreover, knockdown of SUMOylation-machinery components disrupts nodulation, suggesting that additional regulators in the symbiotic pathway also depend on SUMOylation. This work identifies SUMOylation as a conserved regulatory mechanism integrating nitrate signaling with root nodule symbiosis, with broad implications for improving plant nitrogen use efficiency.

Abstract

Nitrate serves both as an essential nutrient and a key signaling molecule that shapes plant growth. In legumes, high nitrate concentrations suppress symbiotic nitrogen fixation, a process mediated by MtNLP1 (NIN-like protein1). Although nitrate minimally affects NLP transcript levels, it strongly controls their nuclear localization. How posttranslational modifications regulate MtNLP1 function, however, has remained unclear. Here, we show that nitrate induces SUMOylation of MtNLP1 at lysine 589 and 795 and that this modification is essential for its biological activity. Loss of these SUMO sites compromises nitrate-mediated inhibition of nodulation and weakens MtNLP1 interactions with MtNIN and itself. Components of the SUMOylation machinery in Medicago truncatula physically interact and are essential for both nodulation and nitrate responsiveness, indicating broader roles for SUMOylation in symbiosis. A SUMO-deficient Arabidopsis thaliana AtNLP7_3KR mutant fails to complement the Atnlp7-1 phenotype, demonstrating that SUMOylation is a conserved regulatory mechanism among NLPs. Together, our findings reveal SUMOylation as a previously unrecognized layer of regulation that integrates nutrient signaling with root nodule symbiosis.

Nitrate is a vital macronutrient for plants, functioning not only as a primary nitrogen source but also as a key signaling molecule. In various plant species, NIN-like proteins (NLPs) act as central regulators of nitrate signaling (1–4). In Arabidopsis thaliana, AtNLP7 serves as a master transcription factor and nitrate sensor, coordinating transcriptional reprogramming in response to nitrate availability (5–8).

In legumes, nitrogen acquisition occurs through two complementary strategies: symbiotic nitrogen-fixation with rhizobia under nitrogen-limited conditions and direct nitrogen uptake from the soil when nitrogen is abundant. Because symbiotic nitrogen fixation is energetically costly, it is tightly regulated by external nitrogen levels—activated under low nitrate conditions and suppressed when nitrate is sufficient (9, 10). Recent studies in Medicago truncatula, Lotus japonicus, and Glycine max have identified NLP1 and NLP4 as key mediators of nitrate-dependent repression of root nodulation. Upon nitrate exposure, MtNLP1, LjNLP4, GmNLP1, and GmNLP4 translocate to the nucleus, where they interact with the nodulation-specific transcription factor Nodule Inception (NIN). This interaction inhibits NIN’s transcriptional activity, thereby suppressing nodule formation under high-nitrate conditions (11–14). In parallel, NLP1 directly activates NRT2.1, a nitrate transporter, enhancing nitrate uptake while further inhibiting nodulation (15, 16). Under high nitrate, NLP1 and NLP4 also induce CLE genes, which repress nodulation; conversely, under low nitrate, MtNLP1 facilitates CEP1 expression to stimulate systemic NRT2.1 induction and promote nodulation (16–18). These studies highlight the central role of NLP1 in integrating nitrate signaling with symbiotic development.

Although nitrate availability minimally affects NLP transcript levels, it profoundly alters the subcellular localization and activity of NLP proteins (11, 12). For example, the nuclear localization of AtNLP7 is regulated by phosphorylation in response to signals such as nitrate or cold-induced calcium fluxes, mediated by CDPK kinases (CDPK10/30/32 or CDPK28). Specific serine phosphorylation events promote AtNLP7 nuclear translocation and target gene activation (19, 20). In contrast, phosphorylation by SnRK1/KIN10 retains AtNLP7 in the cytoplasm, leading to its degradation and maintaining carbon-nitrogen homeostasis (21). These findings emphasize the importance of posttranslational modifications (PTMs) in fine-tuning NLP function.

Among PTMs, SUMOylation—a dynamic and reversible modification involving the conjugation of Small Ubiquitin-like Modifier (SUMO) proteins—plays a key role in modulating protein stability, localization, transcriptional activity, and interactions (22–26). SUMOylation operates a conserved enzymatic cascade: SUMO precursor maturation by ULP (Ubiquitin-like proteases), activation by the E1 enzyme SUMO-ACTIVATING ENZYME1 (SAE1) and SAE2, transfer to the E2 conjugates SUMO-CONJUGATING ENZYME1 (SCE1), and conjugation by E3 ligase such as SAP AND MIZ1 DOMAIN-CONTAINING LIGASE1 (SIZ1) (23–26). SIZ1-mediated SUMOylation is known to influence diverse plant processes, including flowering, seed germination, hormone signaling, stress responses, immunity, and nutrient assimilation (27–37).

In this study, we demonstrate that MtNLP1 undergoes SUMOylation in response to nitrate exposure. We show that components of SUMO pathway physically associate with MtNLP1, promoting its SUMOylation and modulating its transcriptional activity. SUMOylation also affects MtNLP1’s interactions with MtNIN and its ability to form homodimers. Furthermore, mutating SUMOylation sites (K845, K889, and K897) in A. thaliana AtNLP7 abolishes complementation of the Atnlp7-1 phenotype, suggesting that SUMOylation is conserved in other NLPs. Collectively, our findings reveal SUMOylation as a crucial regulatory layer in nitrate signaling, linking posttranslational modification of MtNLP1 to nitrate-mediated inhibition of nodulation.

Results

MtNLP1 SUMOylation Is Essential for Its Function and Transcriptional Activity.

NLP transcription factors mediate nitrate signaling and nitrate-dependent inhibition of nodulation in legumes (11–14). To investigate whether MtNLP1 is regulated by SUMOylation, we analyzed its amino acid sequence using the GPS-SUMO predication tool (http://sumosp.biocuckoo.org/online.php). Three putative SUMOylation motifs (ΨKxE/D; Ψ, large hydrophobic residue; K, acceptor lysine; x, any amino acid; E/D, glutamate or aspartate) were identified at lysine residues 589, 795, and 855 (Fig. 1A), with K795 showing the highest prediction score (Dataset S1). Additional analysis with JASSA (http://www.jassa.fr/index.php?m = jassa) further supported K589 and K795 as conserved SUMO-acceptor sites, suggesting these sites are functionally relevant in MtNLP1.

Fig. 1.

SUMOylation of MtNLP1 is essential for its function in nodulation and nitrate signaling. (A) Schematic of the MtNLP1 protein showing the RWP-RK and PB1 domains (blue) and predicted SUMOylation sites (K589, K795, and K855). (B and C) Nodule numbers on Mtnlp1-1 hairy roots expressing empty vector (EV), wild-type MtNLP1, or SUMO-deficient MtNLP1 variants at 2 wk postinoculation (wpi) with rhizobia under 10 mM KCl or KNO₃ treatment. Letters denote significant differences between samples (Two-way ANOVA, Tukey's multiple comparisons test, P < 0.05). (D and E) Relative expression levels of MtNRT2.1 and MtNIR1 in Mtnlp1-1 roots expressing EV, MtNLP1, and SUMO-site mutants following 4 h of 10 mM KNO₃ treatment. Plants were pregrown on FP medium for 7 d (n = 3, independent biological replicates, each with 6 to 10 plants). Error bars represent SD. (F and G) Dual-luciferase reporter assays examining transcriptional activation of MtNRT2.1 and MtNIR1 promoters by MtNLP1 or SUMO-deficient variants in Nicotiana benthamiana leaves, with nitrate treatment. EV serves as a negative control. LUC activity is normalized to REN activity. n = 6. (H and I) Shoot fresh weight of wild-type (Col-0), Atnlp7-1, and complemented lines expressing AtNLP7 or AtNLP7_3KR in Atnlp7-1 grown on MGRL medium with 5 mM KNO₃ for 14 d (H) or 21 d (I). Error bars indicate SD. In D-I, different letters denote significant differences (One-way ANOVA, Tukey's multiple comparisons test, P < 0.05). “n” represents the sample size.

To determine their biological significance, we generated single (K589R, K795R, or K855R), double (K589/795R, K795/855R, K589/K855R), and triple (K589/795/855R, 3KR) MtNLP1 SUMO-deficient variants. Each construct, driven by the native MtNLP1 promoter, was introduced into Mtnlp1-1 roots. Transgenic plants were inoculated with Sinorhizobium meliloti 2011, and nodulation phenotypes were evaluated two weeks postinoculation (2 wpi) under nitrate-free and nitrate-supplemented conditions.

Under nitrate-free conditions, all SUMOylation-deficient variants exhibited normal nodule formation and produced pink and white nodules comparable to wild-type MtNLP1 (Fig. 1 B and C, and SI Appendix, Fig. S1A). Only MtNLP1_K855R showed a slight reduction in pink nodules relative to the empty vector (EV) control (Fig. 1B). Under nitrate treatment, EV-transformed Mtnlp1-1 roots developed more pink nodules than wild-type R108, consistent with impaired nitrate suppression. This phenotype was fully rescued by wild-type MtNLP1, MtNLP1_K589R, MtNLP1_K855R, and MtNLP1_K589/855R. In contrast, MtNLP1_K795R and MtNLP1_K795/855R lines formed significantly more pink nodules than wild-type MtNLP1, indicating attenuated nitrate-mediated inhibition. In addition, MtNLP1_K589/795R and MtNLP1_3KR lines behaved similarly to the EV control under nitrate, whereas white nodule numbers remained unchanged (Fig. 1 B and C and SI Appendix, Fig. S1B). qRT-PCR confirmed comparable expression levels among MtNLP1 variants (SI Appendix, Fig. S1C).

To evaluate whether SUMOylation affects MtNLP1's transcriptional function, we quantified nitrate-induced expression of MtNRT2.1 and MtNIR1 four hours after nitrate treatment. A strong induction was observed in roots complemented with MtNLP1, MtNLP1_K589R, or MtNLP1_K855R, whereas lines expressing MtNLP1_K795R, MtNLP1_K589/795R, or MtNLP1_3KR displayed a reduced induction (Fig. 1 D and E). Dual-luciferase assays in Nicotiana benthamiana supported these findings, where MtNLP1_K795R, MtNLP1_K589/795R, and MtNLP1_3KR exhibited significantly reduced activation of the MtNRT2.1 and MtNIR1 promoters compared with wild-type MtNLP1 (Fig. 1 F and G and SI Appendix, Fig. S1D).

Together, these results demonstrate that SUMOylation is indispensable for MtNLP1 function, with K795—and cooperatively K589—being critical for nitrate-mediated suppression of nodulation and activation of downstream nitrate-responsive genes.

SUMOylation Is Conserved Among NLPs.

To assess the conservation of SUMOylation in the NLP family, we aligned NLP sequences from A. thaliana, M. truncatula, L. japonicus, and G. max. K589 was broadly conserved across most NLPs (with the exception of AtNLP8/9 and legume NLP2/3), whereas K795 and K855 were specific to NLP1 homologs (SI Appendix, Fig. S2). GPS-SUMO analysis predicted multiple SUMOylation sites in all surveyed NLPs, suggesting that SUMO modification is a widespread regulatory mechanism among plant NLPs (Dataset S1).

To examine its functional relevance in nonlegumes, we expressed either wild-type AtNLP7 or its SUMO-deficient triple mutant AtNLP7_3KR (K845, K889, K897) under its native promoter in Atnlp7-1 mutants. Under nitrate-replete conditions, Atnlp7-1 showed reduced shoot biomass compared to Col-0. Wild-type AtNLP7 fully rescued the mutant phenotype, whereas AtNLP7_3KR did not (Fig. 1 H and I), demonstrating that SUMOylation is essential for AtNLP7’s nitrate-responsive growth function. Collectively, these data support SUMOylation as a conserved and functionally critical modification across NLPs.

Direct Interactions Among Medicago SUMO Machinery Components.

To characterize the SUMOylation pathway in M. truncatula, we identified eight SUMO-related genes, two E2 conjugating enzymes (MtSCE1a, Medtr4g094872 and MtSCE1b, Medtr5g016630), and two putative E3 ligases (MtSIZ1, Medtr4g060510 and MtSIZ2, Medtr2g093730) (SI Appendix, Figs. S3A and S4A). MtSUMO1 (Medtr3g053520) and MtSUMO2 (Medtr3g053540) showed the highest similarity with AtSUMO1/2 and shared 98% sequence identity, including the conserved C-terminal QTGG conjugation motif (SI Appendix, Fig. S3B). GFP fusion of MtSUMO1 and MtSUMO2 localized to the cytoplasm and nucleus, whereas MtSIZ1 localized exclusively to the nucleus in M. truncatula roots (SI Appendix, Fig. S4B), consistent with their Arabidopsis homologs (38, 39). Protein expression was further confirmed by immunoblotting (SI Appendix, Fig. S4 C and D). RNA-seq (40, 41) and single-nucleus RNA-seq (42) data revealed constitutive coexpression of SUMO pathway components across tissues and root cell types (SI Appendix, Fig. S5).

Yeast two-hybrid assays (Y2H) demonstrated interaction among MtSUMO1/2, MtSCE1a/b, and MtSIZ1 (Fig. 2A). These interactions were validated using split-luciferase complementation assays (SLCA) in N. benthamiana (Fig. 2B) and bimolecular fluorescence complementation assays (BiFC) in both M. truncatula roots (Fig. 2C) and N. benthamiana (SI Appendix, Fig. S6). These results confirm the presence of all components of a functional SUMOylation cascade in M. truncatula.

Fig. 2.

Direct interactions among components of the Medicago SUMO machinery. (A) Y2H assays examining interactions among MtSUMO1/2, MtSCE1a/b (E2 enzyme), and MtSIZ1 (E3 ligase). Yeast was grown on SD/-LWH medium with 20 mM 3-amino-1,2,4-triazole (3-AT) for 5 d to assess interaction-dependent growth. (B) SLCA assays confirming interactions between MtSUMO1/2, MtSCE1a/b, and MtSIZ1 in N. benthamiana. Luciferase signal was imaged 2 d postinfiltration. (C) BiFC assays showing interactions among MtSUMO1/2, MtSCE1a/b, and MtSIZ1 in Medicago truncatula roots. Reconstituted Venus fluorescence (yellow) indicates protein–protein interaction. (Scale bar, 50 µm.)

MtNLP1 Interacts With MtSUMO1/2 and MtSCE1a/b.

We next examined whether MtNLP1 interacts with SUMO pathway components. BiFC assays in M. truncatula roots showed interactions between MtNLP1 and MtSUMO1/2 or MtSCE1a/b in the cytoplasm under KCl treatment, shifting to the nucleus after nitrate exposure (10 mM KNO₃, 30 min) (Fig. 3 A and B). This spatial shift mirrored MtNLP1's own nitrate-induced relocalization. SLCA and BiFC assays in N. benthamiana confirmed these interactions (Fig. 3 C and D and SI Appendix, Fig. S7 A and B). No direct interaction was detected between MtNLP1 and MtSIZ1 in either assay (Fig. 3 C and D and SI Appendix, Fig. S7 C and D). These results show that MtNLP1 physically associates with MtSUMO1/2 and MtSCE1a/b.

Fig. 3.

MtNLP1 interacts with MtSUMO1/2 and MtSCE1a/b. (A and B) BiFC assays showing interactions between MtNLP1 and MtSUMO1/2 or MtSCE1a/b in M. truncatula roots after 30 min of 10 mM KCl (A) or 10 mM KNO₃ (B) treatment. Yellow represents reconstituted Venus fluorescence. (Scale bar, 50 µm.) (C and D) SLCA assays showing MtNLP1 interactions with MtSUMO1/2 and MtSCE1a/b in N. benthamiana treatment with 10 mM KCl (C) or 10 mM KNO₃ (D). Constructs were coexpressed, and luciferase signals were recorded 2 d postinfiltration.

MtNLP1 Is Subject to SUMO Modification.

To test whether MtNLP1 is SUMOylated, we conducted in vitro SUMO conjugation assays using purified recombinant proteins. Reactions containing MBP-MtNLP1, Arabidopsis E1 (His-AtSAE1/2), E2 (His-AtSCE1), and E3 ligase (GST-AtSIZ1 or GST-MtSIZ1) were performed with either wild-type SUMO (His-AtSUMO1_GG) or a conjugation-deficient mutant (His-AtSUMO1_AA). High-molecular-weight SUMO-MtNLP1 conjugates were detected only when SUMO (His-AtSUMO1_GG) and an active E3 ligase (GST-AtSIZ1 or GST-MtSIZ1) were present (Fig. 4A). No conjugates were observed with the conjugation-deficient His-AtSUMO1_AA mutant, confirming E3-dependent SUMOylation of MtNLP1.

Fig. 4.

Nitrate induces MtNLP1 SUMOylation in a SIZ1-dependent manner. (A) In vitro SUMOylation assays showing that MtNLP1 is SUMOylated by AtSUMO1_GG in the presence of SIZ1 E3 ligase. SUMOylated MBP-MtNLP1 was detected by anti-AtSUMO1 antibodies. Red asterisks indicate SUMOylated MtNLP1. (B) In vivo SUMOylation assays in N. benthamiana. GFP-MtNLP1, MYC-MtSUMO1_GG, and cLUC-MtSIZ1 were coexpressed with or without 10 mM KNO₃ for 48 h. Proteins were immunoprecipitated using anti-GFP beads and detected by anti-MYC and anti-MtNLP1 antibodies. SUMOylated MtNLP1 was identified by anti-MYC and is indicated by red asterisks. (C) In vivo SUMOylation assays in M. truncatula roots. vYNE-MYC-MtSUMO1_GG or conjugation-deficient vYNE-MYC-MtSUMO1_AA was coexpressed with vYCE-HA-MtNLP1. Plants were grown on FP medium for 7 d and transferred to plates containing 10 mM KNO₃ for the indicated times (0 to 8 h, 20 to 30 plants per time point). SUMOylated MtNLP1 was detected by anti-MYC after anti-HA immunoprecipitation. (D) SUMOylation-site mutagenesis assays. GFP-tagged wild-type or lysine-mutated MtNLP1 (MtNLP1_K589/795R or MtNLP1_3KR) was coexpressed with MYC-MtSUMO1_GG or MYC-MtSUMO1_AA in N. benthamiana. Immunoblotting with anti-MYC showed reduced SUMOylation in mutant proteins. Red asterisks mark the SUMOylated form.

Nitrate Induces MtNLP1 SUMOylation In Vivo.

Coexpression of MtSUMO1_GG, MtSUMO2_GG, or AtSUMO1_GG with MtNLP1 and MtSIZ1 in N. benthamiana confirmed that all SUMO variants could conjugate to MtNLP1 (SI Appendix, Fig. S8), indicating functional conservation of these SUMO proteins and confirming that both MtSUMO1 and MtSUMO2 can mediate SUMOylation in planta. To determine nitrate responsiveness, coexpression assays with MtSUMO1_GG, MtSIZ1, and MtNLP1 revealed a nitrate-dependent SUMOylation, detectable only after nitrate treatment (10 mM KNO₃, 48 h) (Fig. 4B).

To substantiate these findings in a native context, we coexpressed MtNLP1 with either MtSUMO1_GG or MtSUMO1_AA in M. truncatula R108 roots and treated with nitrate (10 mM KNO₃) for varying durations. Immunoprecipitation using anti-HA beads followed by anti-MYC detection showed SUMO-conjugated MtNLP1 bands exclusively in samples treated with nitrate for 4 h and coexpressing MtSUMO1_GG (Fig. 4C). No such bands were observed with the SUMO-deficient MtSUMO1_AA mutant, further validating that MtNLP1 is a bona fide SUMOylation target and that this modification is triggered by nitrate signaling.

To pinpoint the specific SUMOylation sites on MtNLP1, we coexpressed MtSUMO1_GG or MtSUMO1_AA variant with wild-type MtNLP1 or its SUMOylation-deficient mutants (MtNLP1_K589/795R or MtNLP1_3KR), along with MtSIZ1 in N. benthamiana. Wild-type MtNLP1 formed distinct SUMO conjugates when coexpressed with MtSUMO1_GG, whereas SUMOylation was markedly reduced in the double (MtNLP1_K589/795R) and the triple (MtNLP1_3KR) mutants (Fig. 4D). No SUMO conjugates were detected when MtSUMO1_AA was used, confirming the requirement of a functional SUMO moiety. These results establish that K589 and K795 are the functional SUMOylation sites for MtNLP1 modification.

Downregulation of MtSUMO1/2 or MtSIZ1 Impairs Nodulation and Modulates Nitrate-Mediated Suppression.

To investigate the roles of MtSUMO1/2 and MtSIZ1 in nodulation, RNA interference (RNAi) constructs targeting these genes were introduced into M. truncatula roots. Knockdown efficiency was confirmed by qRT-PCR, with cross-suppression between MtSUMO1 and MtSUMO2, and partial off-target silencing between MtSIZ1 and MtSIZ2 (Fig. 5 A and B), likely due to their high sequence similarity.

Fig. 5.

Nodulation phenotype of MtSUMO1/2 and MtSIZ1 knockdown and knockout lines. (A and B) Relative transcript levels of MtSUMO1, MtSUMO2, MtSIZ1, and MtSIZ2 in RNAi (Ri) transgenic hairy roots, normalized to Ubiquitin. Error bars signify SD (n = 3). Asterisks denote significant differences compared to EV control (**P < 0.01, Student’s t test). (C) Nodule numbers in MtSUMO1/2-Ri and MtSIZ1-Ri lines at 2 wpi with S. meliloti 2011 under 10 mM KCl or KNO₃ treatment. Asterisks indicate significant differences (Two-way ANOVA, Dunnett's multiple comparisons test, P < 0.05). (D) Total nodule numbers and corresponding percentage reduction in MtSUMO1/2-Ri and MtSIZ1-Ri lines at 2 wpi. Letters indicate significant differences (One-way ANOVA, Tukey's test, P < 0.05). (E) Nodule numbers in CRISPR-Cas9-generated MtSIZ1 knockout lines under the same conditions. Asterisks indicate significant differences (Two-way ANOVA, Sidak's multiple comparisons test, P < 0.05). (F) Total nodule numbers and percentage reduction in CRISPR-SIZ1 at 2 wpi. Letters indicate significant differences (One-way ANOVA, Tukey's test, P < 0.05).

Under nitrate-free conditions, all RNAi lines produced significantly fewer pink nodules than EV control (Fig. 5C and SI Appendix, Fig. S9A), while white nodules remained unaffected. This suggests that both MtSUMO1/2 and MtSIZ1 are essential for efficient nodulation under permissive conditions. Under nitrate treatment, MtSUMO1/2 RNAi lines produced significantly more pink nodules than EV control, indicating reduced nitrate-mediated suppression (Fig. 5C and SI Appendix, Fig. S9B). MtSIZ1 RNAi lines showed no significant difference in pink nodules numbers but reduced suppression levels (52.6% and 56.1%) comparable with controls (72%) (Fig. 5 C and D). Furthermore, CRISPR/Cas9-mediated MtSIZ1 knockout phenocopied the RNAi lines, confirming defects in both nodulation and nitrate sensitivity, with a percentage reduction of 20.5% relative to 54% in EV control (Fig. 5 E and F and SI Appendix, Fig. S10 A and B). These results demonstrate that the MtSUMO1, MtSUMO2, and MtSIZ1 are required for both nodulation and nitrate-dependent suppression of nodulation.

SUMOylation of MtNLP1 Modulates Its Protein–Protein Interactions.

To determine whether SUMOylation affects MtNLP1 subcellular localization, GFP-tagged wild-type and SUMO-deficient MtNLP1 variants were expressed in M. truncatula roots. All variants were mainly cytoplasmic prior to nitrate exposure and translocated to the nucleus following nitrate treatment (Fig. 6 A and B), indicating that SUMOylation is dispensable for nitrate-induced nuclear import.

Fig. 6.

SUMOylation of MtNLP1 is required for its dimerization and interaction with MtNIN. (A and B) Confocal images showing subcellular localization of GFP, GFP-MtNLP1, and SUMOylation-site mutants (MtNLP1_K589/795R or MtNLP1_3KR) in M. truncatula roots under nitrogen starvation conditions (A) or after 20 min of 10 mM KNO₃ treatment (B). NLS-DsRED (magenta) marks the nucleus and confirms the transgene. (Scale bar, 100 µm.) (C and D) Quantification of BiFC mean gray values and fluorescence intensities showing interactions between MtNIN-C and MtNLP1-C, MtNLP1-C_K589/795R (C), or MtNLP1-C_3KR (D). (E and F) Quantification of BiFC mean gray value and fluorescence intensities showing MtNLP1 self-interaction and reduced homodimerization of MtNLP1_K589/795R (E) or MtNLP1_3KR (F). Asterisks indicate significant difference (*P < 0.05, **P < 0.01, Student’s t test). For each interaction, at least 10 Z-projection images were selected for quantitative analysis of fluorescence intensity (mean gray value) and number of fluorescent spots using ImageJ. ‘n’ represents the sample size. (G) Co-IP assays showing MtNLP1–MtNIN interaction and weakened binding of SUMOylation-deficient MtNLP1_K589/795R or MtNLP1_3KR. GFP-tagged MtNLP1 variants were coexpressed with Flag-MtNIN in N. benthamiana, followed by anti-Flag immunoprecipitation and anti-MtNLP1 immunoblotting. (H) Co-IP assays showing MtNLP1 homodimerization is SUMOylation-dependent. GFP-MtNLP1 or SUMO-deficient variants were coexpressed with Flag-MtNLP1 in N. benthamiana. Co-IP was performed with anti-Flag beads, and GFP-MtNLP1 was detected with anti-GFP. GFP alone served as a negative control.

Given the known interaction between MtNLP1 and MtNIN through their C-terminal PB1 domains, which represses symbiotic gene expression (11), we explored whether SUMOylation affects MtNLP1 interactions. BiFC assays in N. benthamiana revealed that SUMO-deficient MtNLP1 variants had weakened interactions with MtNIN and reduced homodimerization (Fig. 6 C–F and SI Appendix, Fig. S11). Coimmunoprecipitation (Co-IP) assays in N. benthamiana corroborated these findings: MtNLP1_K589/795R and MtNLP1_3KR failed to interact with MtNIN or form homodimers (Fig. 6 G and H). These results demonstrate that while SUMOylation is not required for MtNLP1 nuclear translocation, it is critical for its protein–protein interactions at least in a heterologous context.

Discussion

NLP transcription factors are central regulators of nitrate signaling and nitrate-induced regulation of root nodulation in legumes (11, 13, 14). While previous studies, particularly in Arabidopsis, have emphasized phosphorylation-dependent control of AtNLP7 nuclear localization (19–21), our work uncovers SUMOylation as a distinct and essential layer of NLP regulation.

We demonstrate that MtNLP1 undergoes SUMOylation at K589 and K795 and that this modification is indispensable for its biological function. Unlike phosphorylation, SUMOylation does not affect nitrate-induced nucleus translocation. Instead, it enhances MtNLP1 transcriptional activity likely through promoting its interaction with MtNIN and itself. Thus, SUMOylation primarily regulates MtNLP1's downstream signaling activation rather than its subcellular localization.

Although MtNLP1 SUMOylation required MtSUMO and E3 ligase MtSIZ1, we did not detect a stable MtNLP1-MtSIZ1 interaction. This could reflect a transient or indirect association, or a noncanonical conjugation pathway in which SUMO or E2 enzymes would directly recognize the consensus SUMOylation motifs in MtNLP1 and mediate conjugation independently or cooperatively with SIZ1, as described in the mammalian system (22). Importantly, SUMOylated MtNLP1 accumulated transiently and was detectable only within 4 h of nitrate treatment, consistent with the highly dynamic and reversible nature of SUMO modification, which generally affects only a small fraction of the total protein pool and is rapidly reversed by SUMO-specific proteases (43).

The conservation of NLP SUMOylation across species is supported by both predictive analyses and proteomic evidence from A. thaliana, where several NLPs have been identified as potential SUMOylation targets (44, 45). Notably, AtNLP1 interacts with SUMO conjugation enzymes and the SUMO protease ESD4 and was SUMOylated when expressed in Escherichia coli, suggesting that SUMOylation is a conserved regulatory feature across plant lineages (44). Our functional analyses further confirm that SUMOylation is essential for AtNLP7 and MtNLP1 function. Although K589 is conserved in most NLPs and is required for SUMOylation of both MtNLP1 and AtNLP7, detailed analyses will be necessary to determine which specific residues are functionally important in other NLP family members. Together with the conserved SUMO motifs across NLP homologs, these findings highlight SUMOylation as an evolutionarily conserved mechanism modulating NLP function.

Our study also expands the broader roles of the SUMO pathway in nodulation control. Unlike Mtnlp1 mutants, which form wild-type nodules under nitrate-free conditions but reduce nitrate sensitivity (11), silencing or knockout of MtSUMO1/2 or MtSIZ1 reduced nodulation even in the absence of nitrate and altered nitrate responsiveness. This suggests that SUMOylation regulates not only MtNLP1 but also additional factors required for nodulation. The relatively mild nitrate phenotype in MtSIZ1 knockdown or knockout roots may be explained by the reduced baseline nodulation, which may mask further inhibitory effects of nitrate.

In summary, our studies identify SUMOylation as a key posttranslational modification controlling NLP transcriptional activity and interaction dynamics in response to nitrate. Under nitrate exposure, MtNLP1 relocates to the nucleus and undergoes SUMOylation, which promotes its homodimerization and activation of nitrate-responsive genes. SUMOylation also enhances MtNLP1's interaction with MtNIN, reducing MtNIN-dependent activation of nodulation genes and ultimately suppressing the nodulation program. This conserved mechanism positions SUMOylation as a pivotal regulatory hub linking nutrient signaling with symbiotic development. Further studies exploring the interplay between SUMOylation, phosphorylation, and ubiquitination may reveal how plants integrate multiple environmental cues to balance growth and symbiotic nitrogen fixation.

Materials and Methods

Biological Materials and Strains.

In this study, the M. truncatula R108 ecotype was used as the wild type, and the Mtnlp1-1 mutant described by Lin et al. (11) was employed. M. truncatula seeds were scarified with sandpaper, sterilized in 10% NaClO for 3 min, and rinsed thoroughly with sterile water. Germination was conducted on 0.8% agar plates. Seedlings were cultivated on a perlite:vermiculite mixture (1:1) or on 1.2% nitrogen-free FP agar under a 16 h light (250 μmol m⁻² s⁻¹)/8 h dark photoperiod at 22 °C.

For nodulation assays, Sinorhizobium meliloti strain 2011 was used. Agrobacterium rhizogenes strain AR1193 facilitated hairy root transformation in M. truncatula, while A. tumefaciens strain EHA105 or GV3101 was used for transient expression in N. benthamiana. Plasmids were introduced into E. coli DH10B or DH5α for cloning, and into E. coli Rosetta (DE3) for protein expression. The yeast strain AH109 was used for yeast two-hybrid (Y2H) assays.

A. thaliana ecotype Columbia-0 (Col-0) served as the wild-type background for transformation. The mutant Atnlp7-1 was used, as previously described (21). To assess nitrate effects, seedlings were grown on MGRL medium (21).

In Vitro SUMOylation Assays.

E. coli cells transformed with the constructs His-AtSAE1/2, His-AtSCE1, His-AtSUMO1_GG/AA, MBP-MtNLP1, GST-AtSIZ1, or GST-MtSIZ1 were cultured overnight in LB medium at 37 °C. Overnight cultures were diluted 1:100 into fresh LB and grown for 3 h at 37 °C before induction with 0.5 mM IPTG. Protein expression was induced at 18 °C for 20 h. Recombinant proteins were purified using Ni-Sepharose beads (SMART, Cat.SA005005) for His-tagged proteins, Glutathione Beads 4FF (SMART, Cat.SA010010) for GST-tagged proteins, and Amylose Resin (SMART, Cat.SA026010) for MBP-tagged proteins. Purified proteins were incubated in a 30 µL reaction buffer (20 mM Tris-HCl pH 7.5, 150 mM NaCl, 5 mM MgCl₂, 5 mM ATP) at 30 °C for 3 h. SUMOylated MBP-NLP1 was detected by immunoblotting using anti-MBP (Abmart, T40007M) and anti-AtSUMO1 (Abcam, ab5316) antibodies.

In Vivo SUMOylation Assay.

In N. benthamiana, agroinfiltration was performed with MYC-MtSUMO1/2_GG, MYC-AtSUMO1_GG, and GFP-MtNLP1, GFP-MtNLP1_K589/795R, or GFP-MtNLP1_3KR, coinfiltrated with p19, a suppressor of gene silencing. Plants were cultivated for 48 h under standard conditions with or without nitrate treatment (10 mM KNO₃ for 48 h), then leaves were harvested and frozen in liquid nitrogen. Proteins were extracted using a SUMOylation buffer containing: 50 mM HEPES (pH 7.5), 100 mM NaCl, 5 mM EDTA, 10% (v/v) glycerol, 0.25% (v/v) Nonidet P-40, 0.5% (v/v) Triton X-100, 0.05% (w/v) SDS, 50 mM MG132 (Sigma), 20 mM N-ethylmaleimide (Sigma), and a protease inhibitor cocktail (Roche). Immunoprecipitation was carried out using anti-GFP beads. Immunoblots were performed using anti-MYC (Abmart, Cat.M2000M) and anti-MtNLP1 antibodies.

In M. truncatula roots, BiFC constructs pAtUb:vYCE-HA-MtNLP1 and pLjUb:vYNE-MYC-MtSUMO1_GG, or pLjUb:vYNE-MYC-MtSUMO1_AA were generated using Golden Gate cloning. These constructs were introduced into M. truncatula R108 roots. Plants were first grown on HRE plates for 10 d, then transferred to FP medium for 7 d before treatment with 10 mM KNO₃ for the indicated durations. Transgenic roots were scored for DsRED fluorescence using a Nikon SMZ1500 stereomicroscope. Fluorescent roots were harvested and flash-frozen. Protein extraction was performed as described above. Immunoprecipitation was carried out using anti-HA beads. Proteins were detected using anti-MYC (Abmart, Cat.M2000M) and anti-MtNLP1 antibodies.

All experiments were repeated twice with consistent results.

Hairy Root Transformation in M. truncatula.

For MtNLP1 complementation assays, the corresponding constructs carried by A. rhizogenes AR1193 were introduced into the roots of the Mtnlp1-1 mutant using the standard hairy root transformation protocol. Similarly, for RNAi or CRISPR-based knockdown/knockout of MtSUMO1/2 and MtSIZ1, constructs were transformed into the roots of M. truncatula R108. After 10 d, transformed roots were scored for GFP or RFP fluorescence under a Nikon SMZ1500 microscope. The transgenic plants were transplanted into pots containing a vermiculite and perlite (1:1) mixture. After acclimation for 7 d, plants were inoculated with S. meliloti 2011. The plants were watered twice weekly with either 10 mM KCl or KNO₃. Nodule numbers were recorded 2 wk postinoculation. All experiments were independently repeated at least three times.

Dual-Luciferase Reporter Assays.

The effector construct pMtNRT2.1:LUC was used as described previously (16). Reporter and effector constructs were expressed in GV3101 and subsequently infiltrated into N. benthamiana leaves via agroinfiltration. The dual-luciferase assay was conducted using the Dual-Luciferase® Reporter Assay System (Promega), following the manufacturer's instructions. Each treatment included at least three biological replicates, and the experiment was independently repeated three times.

Protein–Protein Interaction Assays.

Protein–protein interactions were examined by yeast two-hybrid (Y2H), split luciferase complementation assays (SLCA), bimolecular fluorescence complementation (BiFC), and coimmunoprecipitation (Co-IP). The detailed methods are included in the supporting information.

Statistical Analyses.

All statistical analyses and box plot visualization were performed using GraphPad Prism. In each box plot, the line within the box represents the median; the box limits indicate the upper and lower quartiles; and whiskers represent the minimum and maximum values. Statistical significance was determined using Student's t test, one-way ANOVA, or two-way ANOVA, as appropriate. Significance levels are denoted as *P < 0.05 and **P < 0.01. ANOVA analyses were followed by Tukey's or Dunnett's multiple comparisons test at a threshold of P < 0.05. Detailed statistical values are provided in the supporting information.

Supplementary Material

Appendix 01 (PDF)

Dataset S01 (XLSX)

Dataset S02 (XLSX)

Data, Materials, and Software Availability

All study data are included in the article and/or supporting information.