The small peptide CEP1 and the NIN-like protein NLP1 regulate NRT2.1 to mediate root nodule formation across nitrate concentrations

Abstract

Legumes acquire fixed nitrogen (N) from the soil and through endosymbiotic association with diazotrophic bacteria. However, establishing and maintaining N₂-fixing nodules are expensive for the host plant, relative to taking up N from the soil. Therefore, plants suppress symbiosis when N is plentiful and enhance symbiosis when N is sparse. Here, we show that the nitrate transporter MtNRT2.1 is required for optimal nodule establishment in Medicago truncatula under low-nitrate conditions and the repression of nodulation under high-nitrate conditions. The NIN-like protein (NLP) MtNLP1 is required for MtNRT2.1 expression and regulation of nitrate uptake/transport under low- and high-nitrate conditions. Under low nitrate, the gene encoding the C-terminally encoded peptide (CEP) MtCEP1 was more highly expressed, and the exogenous application of MtCEP1 systemically promoted MtNRT2.1 expression in a compact root architecture 2 (MtCRA2)-dependent manner. The enhancement of nodulation by MtCEP1 and nitrate uptake were both impaired in the Mtnrt2.1 mutant under low nitrate. Our study demonstrates that nitrate uptake by MtNRT2.1 differentially affects nodulation at low- and high-nitrate conditions through the actions of MtCEP1 and MtNLP1.

MtNLP1 promotes the expression of MtNRT2.1 to stimulate or suppress nodulation under low- and high-nitrate conditions, respectively, and CEP1 independently promotes nodulation by increasing the expression of NRT2.1 under low-nitrate conditions.

IN A NUTSHELL.

Background: Nitrogen (N) is an essential nutrient and signaling molecule for plant growth and development. Legume plants can not only use N from the soil as nitrate (NO₃⁻) but also obtain N through symbiotic nitrogen fixation with bacteria (rhizobia), which inhabit nodules formed on the plant roots. However, nodule formation and nitrogen fixation are energy-consuming processes for plants; therefore, plants control nodule formation depending on the concentration of available nitrate. High concentrations of nitrate inhibit nodule formation and low concentrations of nitrate promote nodule formation. However, how nitrate controls nodulation remains unclear.

Question: How do the legume nitrate transporters (NRTs) responsible for the uptake/transport of low/high nitrate control root nodule formation in Medicago truncatula?

Findings: Here we report that M. truncatula MtNRT2.1 has both high- and low-affinity nitrate uptake transport activity, and MtNRT2.1 is required for optimal nodule establishment under low nitrate and inhibiting nodulation when nitrate is sufficient. We showed that the transcription factor MtNLP1 directly activated MtNRT2.1 expression and regulation of nitrate uptake under low and high nitrate. Under low nitrate, the gene encoding the small peptide MtCEP1 was highly expressed, and the application of this peptide systemically promoted MtNRT2.1 expression in a MtCRA2-dependent manner. MtNRT2.1 is required for the peptide MtCEP1 to enhance nodulation and nitrate uptake. Our study demonstrates that nitrate uptake by MtNRT2.1 differentially affects nodulation under low- and high-nitrate conditions and this is controlled by MtCEP1 and MtNLP1.

Next steps: Soil N uptake and symbiotic N are important for the growth and yield of legume crops, including soybean (Glycine max). Evaluating the conservation of mechanisms identified in Medicago may provide insight on other legumes in the future.

Introduction

Nitrogen (N) is a macronutrient required for normal plant growth and can be taken up by plants in the form of nitrate (NO₃⁻), which acts as a local and systemic signal in various processes of plant metabolism and development. NO₃⁻ uptake from soil occurs through transporters belonging to two families: nitrate transporter 2 (NRT2) and nitrate transporter/peptide transporter (NPF; Gojon et al., 2011; Krapp et al., 2014; Noguero and Lacombe, 2016). Plant roots have two types of nitrate transport systems: a low-affinity transport system (LATS, >1 mM) and a high-affinity transport system (HATS, 1 μM to ∼1 mM), to adapt to fluctuating levels of nitrate in the soil (Crawford and Glass, 1998). In Arabidopsis thaliana, AtNPF6.3 (also known as CHL1/NRT1.1) was identified as a dual-affinity nitrate transporter and is thought to be the primary transporter involved in LATS (Liu et al., 1999), while AtNRT2s, which require interaction with nitrate assimilation related 2 (NAR2) family proteins for their activity, are the primary transporters involved in HATS (Li et al., 2007; Kotur et al., 2012).

In addition to using NO₃⁻, most legume plants can acquire N from the atmosphere through symbiotic N fixation, which involves a mutualism between legumes and N-fixing bacteria called rhizobia (Ferguson et al., 2019; Wang et al., 2022). This interaction results in the nodule formation on the plant's roots. Within the nodule, rhizobia use the plant photosynthesis products as energy to reduce (or fix) atmospheric N into ammonia for plants’ utilization.

N fixation is an energy-intensive process for plants; therefore, legume plants have evolved a systemic regulatory mechanism referred to as autoregulation of nodulation (AON) to coordinate growth and nodulation. The cost of N fixation is reflected by the inhibition of root nodule formation by legume plants when sufficient mineral N (>2 mM nitrate) is available (Streeter and Wong, 1988; Ferguson et al., 2019). Recent studies in Medicago truncatula and Lotus japonicus have identified a central role for nodule inception (NIN)-like protein (NLP) transcription factors in inhibiting nodulation under high nitrate (HN). NIN, the founding member of the NLP transcription factor family, is one of the most important nodulation genes. NIN is required for rhizobial infection, nodule organogenesis, N fixation, and AON (Liu and Bisseling, 2020; Feng et al., 2021; Jiang et al., 2021).

In contrast to NIN's positive role in infection and organogenesis, NLP1 and NLP4 are required to suppress nodulation at HN concentrations. M. truncatula nlp1 and L. japonicus nlp1 and nlp4 mutants have normal nodulation in the absence of nitrate, but are tolerant to nitrate suppression of nodulation (Lin et al., 2018; Luo et al., 2021; Nishida et al., 2018, 2021). Nitrate can trigger the redistribution of MtNLP1 and LjNLP4 from the cytosol to the nucleus, and MtNLP1 and LjNLP1/4 can interact with NIN through their C-terminal PB1 domains, thus reducing the activation of symbiotic gene expression by NIN (Lin et al., 2018; Luo et al., 2021; Nishida et al., 2018, 2021). By contrast, low concentrations of nitrate (<2 mM) stimulate nodulation in several legume species (Carroll et al., 1985; Barbulova et al., 2007; Fei and Vessey, 2009; Nanjareddy et al., 2014; Ferguson et al., 2019), but the underlying mechanism remains unknown.

A full understanding of nitrate regulation of nodulation in legumes requires identification of the specific nitrate transporters involved and their regulators. Several NPFs have been studied in legumes. Three AtNRT1.1/AtNP6.3 orthologs were identified in M. truncatula, and MtNPF6.5 and MtNPF6.7 were repressed and induced by nitrate, respectively, in an MtNLP1-dependent manner. Xiao et al. (2021) found that both MtNPF6.5 and MtNPF6.7 can mediate nitrate and chloride uptake, but nodulation was not studied in this work. Morère-Le Paven et al. (2011) showed that MtNRT1.3 encodes a dual-affinity nitrate transporter that is up-regulated in the absence of nitrate, suggesting that it may be involved in the response to N-limitation, and again nodulation was not investigated. The high-affinity nitrate transporter MtNPF1.7 [also named LATD/NIP, lateral root defective/numerous infection threads (ITs), polyphenolics] positively regulates root and nodule development (Yendrek et al., 2010; Bagchi et al., 2012). The high-affinity nitrate transporter MtNPF7.6 is induced by rhizobia and nitrate and is specifically expressed in the nodule vasculature within nodule transfer cells, where it mediates nitrate uptake and affects nodulation (Wang et al., 2020). In L. japonicus, a series of nitrate transporters that have positive roles in nodule N fixation have been characterized, including LjNPF3.1, LjNPF8.6, and LjNRT2.4 (Valkov et al., 2017, 2020; Vittozzi et al., 2021); by contrast, LjNPF2.9 has a negative role in regulating shoot biomass and nitrate content without affecting symbiotic performance (Sol et al., 2019). Most recently, LjNRT2.1 was shown to be activated by LjNLP1 and to be required for the translocation of LjNLP4 to the nucleus to mediate nitrate inhibition of nodulation (Misawa et al., 2022).

C-terminally encoded peptides (CEPs), a family of mobile peptide hormones, play an important role in plant nitrate uptake. In Arabidopsis, they act systemically, moving from roots to shoots where they activate the leucine-rich repeat receptor-like kinase (LRR-RLK) receptors AtCEPR1 and AtCEPR2, to promote expression of the shoot-to-root systemic effectors AtCEPD1 (CEP Downstream 1) and AtCEPD2 (Tabata et al., 2014; Ohkubo et al., 2017). AtCEPD1/2 are phloem-mobile signals that relay N-deprivation information to roots to up-regulate the expression of AtNRT2.1 (Ohkubo et al., 2017). In M. truncatula, MtCEP genes are induced under N deprivation conditions and promote susceptibility to rhizobial infection and nodulation (Imin et al., 2013; Gautrat et al., 2020). MtNLP1 directly binds to the half-nitrate response element (NRE) in the MtCEP1 promoter and represses its expression in the presence of nitrate (Luo et al., 2022). Synthetic MtCEP1 peptides enhance nitrate uptake in M. truncatula under low (100 and 500 μM) external nitrate concentrations, but had no effect at high (1 and 5 mM) concentrations (Roy et al., 2021). MtCEP1 binding to leaf vascular cells is dependent on the MtCRA2 (compact root architecture 2) receptor which is closely related to AtCEPR1/2 (Huault et al., 2014; Lee et al., 2021) and the cra2 mutant (tr185 allele) has reduced nitrate uptake rate and N content (Bourion et al. 2014).

Considering that AtNRT2s are the major HATS transporters and that MtNRT2.1 is expressed in root hairs under HN (Pellizzaro et al., 2015; Damiani et al., 2016), we investigated whether MtNRT2.1 may be involved in the regulation of nodulation under low-nitrate (LN) and HN conditions. We show that MtNRT2.1 is required both for enhancing nodulation at LN and inhibiting nodulation at HN. We further show that, similar to Arabidopsis, MtCEP1 peptides systemically induce MtNRT2.1 expression under LN, resulting in enhanced nodulation. Furthermore, we demonstrate that MtNLP1 activates MtNRT2.1 expression under LN and HN conditions to promote or repress nodulation, respectively.

Results

MtNRT2.1 encodes a high-affinity nitrate transporter

Three NRT2 homologs were identified in M. truncatula, and the predicted amino acid sequences of MtNRT2.1 (MtrunA17_Chr4g0028551.1) and MtNRT2.2 (MtrunA17_Chr4g0028541.1) differ by just a single residue, but MtNRT2.2 expression is very low (Pellizzaro et al., 2015). To establish their relationship with NRT2 family members in other plant species, a phylogenic tree is generated using NRT2s from M. truncatula, L. japonicus, and A. thaliana, which shows that MtNRT2.1/2.2 is closely related to LjNRT2.1 (Supplemental Figure 1). Analysis of MtNRT2.1 expression in roots using RT-qPCR reveals that MtNRT2.1 is expressed under both LN (0.5–2 mM KNO₃) and HN (5 mM KNO₃) conditions, and the expression level is higher in HN than in LN (Figure 1A). The spatial and temporal expression patterns of MtNRT2.1 during root nodulation are determined using pMtNRT2.1:GUS transformed M. truncatula hairy roots, which shows that it is expressed in root vascular tissues and the nodule meristem (Supplemental Figure 2).

Figure 1.

MtNRT2.1 was induced by nitrate and has nitrate transport activity in X. laevis oocytes. A, Relative expression of MtNRT2.1 in roots of M. truncatula seedlings grown on 0, 0.5, 1, 2, and 5 mM KNO₃. Plants were grown on a nitrogen-free FP medium for 5 days and then transferred to new plates containing different concentrations of KNO₃ for an additional 3 days (n = 3, each made up of 8–12 plants). Different letters represent statistically significant differences (P < 0.05, one-way ANOVA with Tukey's multiple comparisons). B and C, Uptake of 0.5 Mm (B) and 5 mM (C). K¹⁵NO₃ into oocytes injected with water (control), single MtNRT2.1 cRNA or MtNRT2.1 and MtNAR2.1 mixtures. Oocytes were incubated with the indicated concentrations of K¹⁵NO₃ at pH 5.5 for 3 h, and ¹⁵N contents were determined. Values are means ± SD (n = 10–18 oocytes). Asterisks indicate significant differences compared with the H₂O-injected oocytes (two-tailed t test, * P < 0.05; ns, not significant).

In order to determine the function of MtNRT2.1 in root nodule formation, we isolated a Tnt1 insertion mutant line, NF2226, which has a Tnt1 insertion in the first exon of MtNRT2.1 (Supplemental Figure 3A). A primer set consisting of one primer common to both NRT2.1 and NRT2.2 and a primer specific to the 3′ UTR of MtNRT2.1 is used for genotyping (Supplemental Figure 3B). Further RT-qPCR analysis confirms that MtNRT2.1 expression is dramatically reduced in Mtnrt2.1 roots (Supplemental Figure 3C). RT-PCR using NRT2.2-specific primers show that the NRT2.2 cDNA is barely detectable in the roots of wild type and the nrt2.1 mutant (Supplemental Figure 3D), consistent with a previous study (Pellizzaro et al., 2015). Thus, we focused our study on MtNRT2.1.

MtNRT2.1 homologs in A. thaliana and rice (Oryza sativa) were shown to be HATS components (Filleur et al., 2001; Orsel et al., 2004; Feng et al., 2011), and NAR2 was required for their activity (Okamoto et al., 2006; Orsel et al., 2006). The nitrate transport activity of MtNRT2.1 was assessed by measuring ¹⁵NO₃⁻ uptake activity in MtNRT2.1/MtNAR2-cRNA injected Xenopus laevis oocytes. The injected oocytes were incubated with 0.5 or 5 mM ¹⁵N nitrate at pH 5.5, then their ¹⁵NO₃⁻ content was determined. The measurement of ¹⁵N enrichment in oocytes shows that only the combination of injecting with MtNRT2.1 and MtNAR2 results in a significant uptake when compared with water and MtNRT2.1-injected oocytes, at both 0.5- and 5-mM nitrate (Figure 1, B and C). This indicates that MtNRT2.1 has both high- and low-affinity nitrate uptake activity which requires NAR2.

MtNRT2.1 mediates nitrate uptake

To determine the role of MtNRT2.1 in nitrate uptake in M. truncatula, we assessed the nitrate content of wild-type (WT) and Mtnrt2.1 roots using high-performance liquid chromatography (HPLC) 6- or 12-h after treatment with LN (0.5 mM KNO₃) or HN (5 mM KNO₃). The nitrate content of WT is higher at 12 h after treatment with LN or HN than at 6 h, but remains low in Mtnrt2.1 roots (Figure 2, A and D). The nitrate uptake ability of WT, Mtnrt2.1, and Mtnlp1 roots under LN and HN was further examined using ¹⁵NO₃⁻. In this assay, the ¹⁵NO₃⁻ content was assessed after 30 min of incubation. The results show that the root and shoot ¹⁵NO₃⁻ contents of both Mtnrt2.1 and Mtnlp1 are significantly lower than WT under LN and HN conditions (Figure 2, B, C, E, and F).

Figure 2.

MtNRT2.1 mediates nitrate uptake in M. truncatula. A and D, Nitrate content in WT and Mtnrt2.1 mutants after treatment with 0.5 mM KNO₃ (A) and 5 mM KNO₃ (D) as determined by HPLC. n = 5–6 with each data point representing a pool of roots from 3 to 5 plants. Plants were grown on a nitrogen-free FP medium for 12 days and then supplied with 0.5 or 5 mM KNO₃ for 6 h or 12 h. Different letters represent statistically significant differences (P < 0.05, two-way ANOVA with Tukey's test for multiple comparisons). B and C, Nitrate uptake by roots (B) and shoots (C) of WT and Mtnrt2.1 in 0.5 mM KNO₃. E and F, Nitrate uptake in roots (E) and shoots (F) of WT and Mtnrt2.1 in 5 mM KNO₃. Plants were grown on a nitrogen-free FP medium for 7 days, pretreated with 0.5 mM or 5 mM KNO₃ for 4 h, then transferred to FP hydroponics medium with 0.5 or 5 mM K¹⁵NO₃ for 30 min. ¹⁵N content of shoots and roots was analyzed. * P < 0.05 by a two-sided Student's t test. G–I, Relative gene expression of MtCLE35 (G), MtNIA1 (H), and MtNIR1 (I) in WT and Mtnrt2.1 roots after treatment with 0, 0.5, 1, 2, and 5 mM KNO₃. Plants were grown on FP medium for 5 days and then transferred to new plates containing 0, 0.5, 1, 2, and 5 mM KNO₃ for additional 3 days (n = 3, each made up of 9–12 plants). * P < 0.05 indicate significant differences compared with WT with ZN by a two-sided Student's t test. Error bars show SD.

To verify the relationship between nitrate uptake/transport and gene expression, the expression of three nitrate-inducible genes, MtCLE35, Nitrate Reductase 1 (NIA1), and Nitrite Reductase 1 (NIR1), was analyzed in noninoculated roots by RT-qPCR. These genes show no induction at ≤ 0.5 mM nitrate, are weakly induced at 1.0 or 2.0 mM nitrate, and are highly induced at 5 mM nitrate; moreover, the induction level is significantly attenuated in Mtnrt2.1 mutants (Figure 2, G–I). These data indicate that MtNRT2.1 is responsible for the majority of nitrate taken up by M. truncatula in the 0.5–5 mM range and that it is required for nitrate-induced gene expression.

MtNRT2.1 promotes nodulation at LN

A low concentration of nitrate can stimulate nodulation in several legume species (Nanjareddy et al., 2014; Ferguson et al., 2019). Given that MtNRT2.1 has high-affinity nitrate transport activity, we speculated that it may function in the stimulation of nodulation under LN. First, we analyzed the temporal expression pattern of MtNRT2.1 after LN treatment. RT-qPCR analysis shows that it is induced after one day of LN treatment and the induction level increases over time (Supplemental Figure 4A). The expression of three nitrate-induced genes, MtCLE35, MtNIA1, and MtNIR1, was analyzed at 1, 3, and 5 days after transfer to LN in WT and Mtnrt2.1. This reveals that after 5 days of LN treatment, the expression of these genes is weakly induced by LN and that this induction is abolished in Mtnrt2.1 (Supplemental Figure 4, B–D).

We then analyzed the nodulation phenotype of the Mtnrt2.1 mutants in the absence of nitrate (zero nitrate, ZN) and under LN conditions. Consistent with previous reports (Fei and Vessey, 2009), we found that WT plants develop ∼60% more nodules at LN than in ZN conditions at 1, 2, and 3 weeks post inoculation (wpi; Figure 3, A and B). By contrast, the nodule number of Mtnrt2.1 does not increase compared with ZN and LN at all time points examined (Figure 3B). The number of rhizobial ITs at 1 wpi increases in WT at LN, but does not change in Mtnrt2.1, in line with the nodule number (Figure 3C). Acetylene reduction activity (ARA) was used to estimate nodule nitrogenase activity at 2 wpi. This shows that ARA is higher in WT at LN compared to ZN and that this effect is lost in Mtnrt2.1 (Figure 3D). Together, these analyses suggest that the enhancement of symbiosis at LN relative to ZN is dependent on MtNRT2.1. This is consistent with the lower nitrate accumulation in Mtnrt2.1 roots under LN (Figure 3E).

Figure 3.

Mtnrt2.1 nodulation and nitrate accumulation phenotypes under LN. A–E, Nodule phenotype (A), nodule numbers (B), IT numbers (C), nitrogenase activity (D), and nitrate content (E) of WT and Mtnrt2.1 after continuous treatment with LN (0.5 mM KNO₃) 1∼3 (B), 1 (C), or 2 (A, D, and E) wpi with S. meliloti (n = 10–15 plants). n = 3 in (D), n = 5 (E), each data point represents a pool of roots from 2 to 3 plants. F–H, Relative gene expression of MtNIN (F), MtCLE13 (G), and MtNF-YA1 (H) in WT and Mtnrt2.1 roots after treatment with 0.5 mM KNO₃ at 7 dpi (n = 3, each made up of 9–12 plants). Expression levels were normalized relative to the roots without inoculation. Different letters represent statistically significant differences (P < 0.05, two-way ANOVA with Tukey's multiple comparisons test). Asterisk (change figure to one asterisk) indicates significant differences (* P < 0.05, two-tailed t test). Error bars show SD.

To further validate the role of MtNRT2.1 in nodulation under LN, the expression of three symbiotic genes, MtNIN, MtCLE13, and MtNF-YA1 (Nuclear Factor YA1), was examined in WT and Mtnrt2.1 at 1 wpi under ZN and LN. The results show that these symbiotic genes are relatively more highly expressed in WT at LN, but that this increase is abolished or reduced in Mtnrt2.1 (Figure 3, F–H). The shoot and root weight of WT is higher at LN compared with ZN, and this increase is reduced in the Mtnrt2.1 mutant (Supplemental Figure 5). The nodulation and growth phenotypes and nitrate content of Mtnrt2.1 are complemented by the transgenic expression of MtNRT2.1, driven by its native promoter (Supplemental Figure 6). Together, these results demonstrate that MtNRT2.1 is required for LN stimulation of root nodulation and plant growth.

MtNRT2.1 is required for CEP1-induced nodulation

MtCEP1 is up-regulated by LN and the overexpression of MtCEP1 or application of synthetic MtCEP1 peptide to roots can enhance nodule number in M. truncatula (Imin et al., 2013). We used RT-qPCR to test whether MtNRT2.1 is required for MtCEP1 expression at different nitrate concentrations. We found that in WT roots, CEP1 expression is higher at LN (0.5 mM KNO₃) and is lower at 1-, 2-, or 5-mM nitrate treatments, compared with ZN (Figure 4A). However, the effects of different nitrate concentrations on MtCEP1 expression are mostly lost in Mtnrt2.1 roots, with only a slight increase observed at LN (0.5 mM KNO₃) and no changes seen at the other concentrations tested, relative to ZN (Figure 4A). We then analyzed changes in MtCEP1 expression over time after LN treatment. Our results reveal that MtCEP1 expression shows MtNRT2.1- and NLP1-dependent increase at 3 and 5 days after LN treatment (Figure 4B).

Figure 4.

MtCEP1 enhances nodulation via MtNRT2.1. A, Relative expression of MtCEP1 in WT and the Mtnrt2.1 mutant with 0, 0.5, 1, 2, and 5 mM KNO₃. B, Relative expression of MtCEP1 in WT and Mtnrt2.1 mutant after treatment with 0.5 mM KNO₃ at 1, 3, and 5 days. Asterisks indicate significant differences compared with the plants with ZN (* P < 0.05, ns, not significant by a two-tailed t test). C, Relative expression of MtNRT2.1 after treatment with 1 μM MtCEP1. Asterisks indicate significant differences at * P < 0.05 compared with the plants’ treatment with water (Mock) by a two-tailed t test. Plants were grown on FP plates for 5 days and then transferred to a new plate containing indicated concentrations of KNO₃ (A, B) or 1 μM MtCEP1 peptide (C) for another 3 days. (n = 3, each comprising 8–10 plants). D–G, Nitrate uptake by roots (D, F) and shoots (E, G) of WT and Mtcra2 in 0.5 mM KNO₃ (D, E) and 5 mM KNO₃ (F, G). Plants were grown on a nitrogen-free FP medium for 7 days, pretreated with 0.5 mM or 5 mM KNO₃ for 4 h, then transferred to FP hydroponics medium with 0.5 mM or 5 mM K¹⁵NO₃ for 30 min. ¹⁵N content of shoots and roots was analyzed. H, Split-root system in which the roots of a plant were separated into left (L) and right (R) parts, and right (R) parts which were supplied with either mock (deionized water) or 1 μM MtCEP1 peptide for 3 days. I, Relative expression of MtNRT2.1 on each side of the split roots of WT (n = 3 independent pools of each part of roots from 4 to 6 plants). J, K Nitrate content of the WT (J) and the Mtnrt2.1 mutant (K) treated with 0.5 mM KNO₃ and supplied with 1 μM MtCEP1. The roots were harvested and then nitrate content was measured by HPLC (n = 4–5, with each biological consisting of 3–5 pooled roots). L, M Nodule numbers (L), nitrogenase activity (M) of WT and the Mtnrt2.1 mutant in the presence of 0.5 mM KNO₃ supplied with 1 μM MtCEP1 at 2 wpi with S. meliloti (n = 10–15 plants). n = 3 with each data point in (M) representing a pool of roots from 2 to 3 plants. Different letters represent statistically significant differences (P < 0.05, two-way ANOVA with Tukey's multiple comparisons test). Asterisk indicates significant differences (* P < 0.05, two-tailed t test). Error bars show SD. The % increase between mock and MtCEP1 treatments is indicated in (L).

In Arabidopsis, CEP1 peptide can systemically induce AtNRT2.1 to regulate N acquisition (Ohkubo et al., 2017). To examine whether similar regulation occurs in M. truncatula, we checked MtNRT2.1 expression levels in WT, Mtcra2, and Mtnlp1 roots after treatment with synthetic MtCEP1 peptide. This reveals that MtNRT2.1 is induced by MtCEP1 peptide in WT and Mtnlp1, but not in Mtcra2 roots (Figure 4C). This suggests that CEP1's effect on nodulation may depend on MtNRT2.1-mediated nitrate uptake. To address this possibility, we assessed the nitrate uptake ability of the Mtcra2 mutant using K¹⁵NO₃ at 0.5 mM (LN) or 5 mM (HN). This shows that the nitrate uptake ability of the mutant is reduced compared with WT both at LN and HN (Figure 4, D–G). These findings are consistent with observations made for the Arabidopsis cepr1 cepr2 mutant (Tabata et al., 2014) and the M. truncatula cra2 mutant (Bourion et al., 2014).

We then used a split-root system to investigate whether this effect was systemic. The split roots are grown under ZN conditions, with the media containing CEP1 peptide on just one side (R; Figure 4H). RT-qPCR analysis reveals that MtNRT2.1 expression is higher and essentially equal at both sides of the split root when one side is treated with MtCEP1 (Figure 4I). This suggests that, as in Arabidopsis (Ohkubo et al., 2017), MtCEP1 induces MtNRT2.1 expression in a systemic manner. We then tested nitrate uptake in WT and Mtnrt2.1 roots under LN conditions after MtCEP1 peptide treatment. We found that treatment with MtCEP1 enhances nitrate uptake in WT roots at LN (Figure 4J), consistent with an earlier report (Roy et al., 2021). However, MtCEP1's effect on nitrate uptake is totally abolished in Mtnrt2.1 roots (Figure 4K). This suggests that MtNRT2.1 is required for CEP1 promotion of nitrate uptake.

MtCEP1 peptide can increase nodule number in LN conditions (Imin et al., 2013). We examined nodulation in WT and Mtnrt2.1 after MtCEP1 treatment at LN. We found that MtCEP1 induces an increase in nodule number and ARA at LN, and that these effects are reduced in Mtnrt2.1 (Figure 4, L and M). In addition, MtCEP1 treatment results in higher nitrate accumulation, and increases shoot and root weight in WT at LN, and these increases are abolished in Mtnrt2.1 (Supplemental Figure 7). Taken together, our data suggest that the positive effects of MtCEP1 on plant growth and nodule number at LN partly require MtNRT2.1.

MtNRT2.1 is required for HN inhibition of nodulation

LjNRT2.1 was shown to have a central role in nitrate-mediated inhibition of nodulation (Misawa et al., 2022). As MtNRT2.1 is similarly induced by HN (Figure 1A), we further analyzed the Mtnrt2.1 nodulation phenotype in the presence of 5 or 10 mM KNO₃ 2 wpi with Sinorhizobium meliloti. As expected, the total number of nodules and the number of pink nodules in WT are suppressed at both 5 and 10 mM KNO₃, and the effect is stronger at higher concentrations (Figure 5, A–C). In Mtnrt2.1, the total nodule number is not suppressed at 5 mM, and is only partly suppressed at 10 mM KNO₃ (Figure 5B). In addition, Mtnrt2.1 produces more pink nodules and has higher ARA than WT at both KNO₃ concentrations, but still shows some nitrate suppression (Figure 5, A, C, and D). The nitrate content and shoot and root fresh weight of Mtnrt2.1 are lower than WT at HN (Figure 5E, Supplemental Figure 5).

Figure 5.

Mtnrt2.1 nodulation phenotype at HN. A–E, Nodule phenotype (A), total nodule number (B), pink nodule number (C), nitrogenase activity (D), and nitrate content (E) of WT and the Mtnrt2.1 mutant in the presence of HN at 2 wpi with S. meliloti (n = 19–26 plants). n = 3 in (D), n = 4–5 in (E), each data point representing a pool of roots from 2 to 3 plants. F–H, Relative expression of MtNIN (F), MtCLE13 (G), and MtNF-YA1 (H) in WT and Mtnrt2.1 roots after treatment with 5 mM KNO₃ at 7 dpi. (n = 3, each made up of 9–12 plants). Expression levels were normalized relative to the roots without inoculation. Different letters represent significant differences (P < 0.05, two-way ANOVA with Tukey's multiple comparisons). Scale bars = 1 mm in (A). Error bars show SD.

To further investigate MtNRT2.1's potential role in HN inhibition of nodulation, the expression of three symbiotic genes, MtNIN, MtCLE13, and MtNF-YA1, was analyzed. Their expression is reduced in WT at HN and this repression is lost in Mtnrt2.1 (Figure 5, F–H). The Mtnrt2.1 nitrate tolerant nodulation phenotype, nitrate content, and root growth defect at HN are rescued by complementation with MtNRT2.1 (Supplemental Figure 8). The data demonstrate that MtNRT2.1 plays an important role in the nitrate inhibition of nodulation.

MtNLP1 activates MtNRT2.1 in response to nitrate

MtNRT2.1 is expressed under LN, HN, and ZN conditions, and expression is lower in Mtnlp1 roots than in WT roots (Lin et al., 2018). We first confirmed that the loss of MtNLP1 results in lower expression of MtNRT2.1 and showed, using a time course, that LN does not promote MtNRT2.1 expression in nlp1 (Figure 6A). We then examined whether MtNLP1 is required for LN stimulation of nodulation and found that nodulation stimulation under LN is abolished in the Mtnlp1 mutant at 2 wpi with S. meliloti (Figure 6B). Overexpression of MtNLP1 (p35S-NLP1) results in increased induction of MtNRT2.1 in M. truncatula hairy roots, and the increase is higher at HN compared with LN (Figure 6C).

Figure 6.

MtNLP1 activates MtNRT2.1 expression. A, Relative expression of MtCEP1 in WT and Mtnlp1 roots after treatment with 0.5 mM KNO₃ for 1, 3, and 5 days. Plants were grown on a nitrogen-free FP medium for 5 days and then transferred to new plates containing 0 or 0.5 mM KNO₃ for an additional 1, 3, or 5 days (n = 3, each made up of 8–12 plants). B, Nodule number of WT and the Mtnlp1 in the presence of 0.5 mM KNO₃ at 2 wpi with S. meliloti (n = 19–26 plants). C, Relative expression of MtNRT2.1 in transgenic hairy roots of WT transformed with EV control and NLP1-overexpression (p35S-GFP-MtNLP1) constructs in the presence of 0.5 and 5 mM KNO₃ (n = 3, each made up of 7–11 plants). Asterisks indicate significant differences compared with the transformed roots with ZN (* P < 0.05, *** P < 0.001, one-way ANOVA with Dunnett's multiple comparisons test). D, Alignment of NRE motifs previously identified in MtCLE35, LjCLE-RS2, and MtNRT2.1 promoters, performed using the MEME algorithm (http://meme-suite.org/index.html). Red letters represent conserved nucleotides. E, Relative expression of MtNRT2.1 in pUb-MtNLP1-GR transgenic hairy roots supplemented with mock (DMSO), dexamethasone (DEX), or DEX + cycloheximide (CHX) for 4 or 24 h (n = 3, each consisting of pooled roots from 6 to 8 plants). Different letters represent statistically significant differences (P < 0.05, two-way ANOVA with Tukey's multiple comparisons test). F, Yeast one-hybrid assay of MtNLP1 binding to the NRE in the MtNRT2.1 promoter. Binding assays were performed with MtNLP1 against the synthetic NRE, or against the MtNRT2.1 with a deleted NRE region. Medium lacking leucine and tryptophan (-Leu -Trp) or leucine, tryptophane, and histidine (-Leu -Trp -His) was supplemented with 100 mg mL⁻¹ 3-AT. Yeast cells were diluted in a 10× dilution series. G, The luciferase activity of N. benthamiana leaves co-transformed with LUX driven by the MtNRT2.1 promoter and p35S-MtNLP1 in the presence of 0.5 and 5 mM KNO₃ (n = 6). Asterisks indicate significant differences compared with the EV control (* P < 0.05, one-way ANOVA with Dunnett's multiple comparisons test). Error bars show SD.

We further investigated the nature of this activation by analyzing the MtNRT2.1 promoter sequence. This reveals a putative NRE motif in the MtNRT2.1 promoter region at −248 bp upstream of the start codon (Figure 6D). Then we investigated whether MtNLP1 can directly bind to the MtNRT2.1 promoter to activate its expression using the pUb-MtNLP1-GR inducible system (Luo et al., 2021). The expression of MtNRT2.1 in M. truncatula pUb-MtNLP1-GR transformed hairy roots treated with dexamethasone (DEX) or DEX and cycloheximide (CHX) was investigated. RT-qPCR analysis revealed that the expression MtNRT2.1 was induced at 4 and 20 h by treatment with DEX. The addition of CHX does not repress the DEX-inducible expression and even enhances it at 4 h (Figure 6E), demonstrating that de novo protein synthesis is not required for the activation of MtNRT2.1 expression. The potential binding of MtNLP1 to the MtNRT2.1 promoter region was then examined using a yeast one-hybrid assay. The results show that MtNLP1 can bind to the MtNRT2.1 promoter region containing the NRE motif, however, this binding does not occur when this NRE motif is deleted (ΔNRE; Figure 6F). We then used a dual- luciferase reporter assay in which pMtNRT2.1:LUC or pMtNRT2.1ΔNRE:LUC was co-expressed with p35S-MtNLP1 or empty vector (EV) in N. benthamiana leaves to analyze whether MtNLP1 affects MtNRT2.1 transcriptional activity. This reveals that MtNLP1 can slightly enhance pMtNRT2.1:LUC expression at LN, and can highly activate pMtNRT2.1:LUC expression at HN, but that the MtNLP1 activation of MtNRT2.1 expression is absent when the NRE motif is removed (pMtNRT2.1ΔNRE:LUC; Figure 6G). These results suggest that MtNLP1 directly binds to the MtNRT2.1 promoter to activate its expression.

Nitrate can trigger MtNLP1 protein shuttling from the cytosol to the nucleus (Lin et al., 2018). We then tested whether this phenomenon also occurs at LN. A construct containing p35S:GFP-NLP1 and pLjUb-nls-DsRed was expressed in M. truncatula WT hairy roots, and localization of the GFP-fusion protein was analyzed 10–20 min after nitrate treatment. In agreement with our previous study (Lin et al., 2018), confocal microscopy observations showed that the MtNLP1 localized in the cytosol in the absence of nitrate and translocated to the nucleus in WT roots after HN treatment. After LN treatment, some MtNLP1 translocation to the nucleus is detectable, but is much lower compared to that observed at HN (Supplemental Figure 9). Taken together, our results demonstrate that MtNLP1 directly targets MtNRT2.1 to activate its expression, including a slight activation of MtNRT2.1 expression at LN, and that MtNLP1 protein translocation to the nucleus may be proportional to the level of nitrate present.

Overexpression of MtNRT2.1 can partially rescue the Mtnlp1 phenotype

To further investigate MtNRT2.1 function in nitrate regulation of root nodule formation, we overexpressed MtNRT2.1 from the Lotus Ubiquitin promoter (pUb-MtNRT2.1) in WT and Mtnlp1-1. The transgenic roots were watered with LN or HN, and the phenotype was characterized 2 wpi with S. meliloti. The WT plants transformed with pUb-MtNRT2.1 develop more total nodules and pink nodules at LN, but less of both types of nodules at HN compared with the EV control (Figure 7, A and B; Supplemental Figure 10). This result suggested that overexpressing MtNRT2.1 in WT can enhance nodulation at LN, but confers hypersensitivity to HN suppression of nodulation. However, the overexpression of MtNRT2.1 can enhance nitrate accumulation, as well as root weight at LN and HN conditions (Figure 7C; Supplemental Figure 10, A, C, E, and F). The overexpression of MtNRT2.1 in Mtnlp1 similarly enhances the total nodule number and the number of pink nodules at LN, but decreases nodule formation at HN compared with the EV control (Figure 7, A and B; Supplemental Figure 10B), while the nitrate uptake ability of Mtnlp1 is enhanced under both LN and HN conditions (Figure 7C). Moreover, the overexpression of MtNRT2.1 in Mtnlp1 enhances root weight at HN (Supplemental Figure 10, B, D, E, and G). These results suggested that overexpression of MtNRT2.1 can partially restore Mtnlp1's defects in nitrate uptake, growth, and nitrate inhibition of root nodule formation, and further demonstrates that MtNRT2.1 promotes nodulation at LN but inhibits nodulation at HN.

Figure 7.

The effects of overexpression MtNRT2.1 in WT or Mtnlp1 hairy roots on nodule number and nitrate content. A–C, Total nodule number (A), number of pink nodules (B), and nitrate content (C) in the hairy roots of WT and Mtnlp1 overexpressing MtNRT2.1. The phenotype was scored at 2 wpi with S. meliloti with continuous treatment with 0.5 or 5 mM KNO₃ (n = 8–16 plants). Each data point representing a pool of roots from 1 to 3 plants in (C). Asterisk indicates significant differences (* P < 0.05, two-tailed t test). Error bars show SD.

To further test whether NLP1 and NRT2.1 function in the same genetic pathway, a construct designed to target MtNLP1 using RNAi (Lin et al., 2018) was introduced into Mtnrt2.1 using hairy root transformation and the nodulation phenotype was analyzed in the presence of LN and HN. The results show that compared to EV control roots, knockdown of MtNLP1 in Mtnrt2.1 roots does not affect nodule number under either LN or HN conditions (Supplemental Figure 11). This suggests that MtNLP1 and MtNRT2.1 function in the same genetic pathway to regulate nodulation in response to nitrate.

Discussion

Legume N fixation is a highly energy-consuming process, and the inhibition of nodulation by HN availability ensures its efficient regulation. In recent years, several studies have demonstrated that NLPs play important roles in HN inhibition of nodulation (Lin et al., 2018; Nishida et al., 2018, 2021). However, compared to N-starvation, low concentrations of nitrate and ammonium, through unknown mechanisms, stimulate nodulation and N₂ fixation in several types of legumes (Barbulova et al., 2007; Fei and Vessey, 2009; Nanjareddy et al., 2014; Ferguson et al., 2019). Notably, the effective nitrate concentration stimulating nodulation differed across these studies. This discrepancy could result from many factors, including the species used, the size of the plant, the growth conditions used, and the type and frequency of N provided. Of the nitrate concentrations (0, 0.5, 1, 2, and 5 mM KNO₃) used in this study, only 0.5 mM was effective in promoting the nodulation of M. truncatula. This correlated closely with CEP1 expression, supporting the role of CEP1 in this phenomenon. We found that the promotion of nodulation under LN occurs through the nitrate transporter MtNRT2.1. Hence, MtNRT2.1 plays a dual role in nitrate regulation of nodulation, by promoting nodulation at LN, which requires both NLP1 and CEP1/CRA2, and inhibiting nodulation at HN, which requires MtNLP1.

Although legume plants can fix N, the uptake of inorganic N is still important, especially during early seedling growth when N₂ fixation has not yet been well developed. Harper (1974) reported that soybean plants that depend solely on N fixation show poor growth and low seed yield, suggesting that both soil N and symbiotic N are required for optimum soybean production. Several nitrate transporters have been identified in L. japonicus and M. truncatula, but none of them have been reported to be involved in LN stimulation of nodulation. Here, we found that MtNRT2.1 is expressed in root vascular tissues and the apex of nodules, implying that MtNRT2.1 might be involved in root nodulation. Indeed, the Mtnrt2.1 mutant develops less nodules and has lower ARA, and overexpression of MtNRT2.1 can enhance nodulation at LN, suggesting that MtNRT2.1 is essential for nitrate uptake and the stimulation of nodulation at LN. In A. thaliana, AtNRT2.1 interacts with AtNAR2.1 to form functional HATS (Orsel et al., 2006). Similarly, we found that MtNRT2.1's ability to uptake nitrate in X. laevis oocytes requires MtNAR2, suggesting that MtNRT2.1 is a typical HATS transporter. However, we also noticed that MtNRT2.1 functions at HN, and thus appears to also operate as a LATS transporter. Moreover, nitrate uptake ability was significantly impaired in Mtnrt2.1 and Mtnlp1 at both LN and HN, suggesting that they contribute to nitrate uptake across a broad range of nitrate concentrations. Mtnrt2.1 shoot and root biomass were smaller than the wild type at both LN and HN. This differs from Atnrt2.1 (Orsel et al., 2006), but is similar to Ljnrt2.1, which has reduced shoot biomass at HN (Misawa et al., 2022). These observations suggest that legume NRT2.1s may function differently than their orthologs in nonsymbiotic species.

MtNRT2.1 is up-regulated by LN and HN in a MtNLP1-dependent manner (Lin et al., 2018). MtNLP1 needs to translocate to the nucleus to function as a transcription factor, and this translocation was minimal after LN treatment. Nonetheless, MtNRT2.1 expression was very low in Mtnlp1 mutants even in the absence of external nitrate, and overexpression of MtNLP1 can slightly activate MtNRT2.1 expression under LN. This suggests that MtNLP1 activation of MtNRT2.1 is proportional to the external nitrate concentration, and that our inability to detect MtNLP1 in the nucleus in the absence of external nitrate may reflect the sensitivity of detection. A similar proportional effect of nitrate levels on gene expression was reported for the nodule-expressed NLP2 (Jiang et al., 2021).

The AtCEP1 peptide is recognized as a “hunger signal” by the plant, and systemically up-regulates the expression of AtNRT2.1 depending on the CEP1 receptors AtCEPR1 and AtCEPR2 (Ohkubo et al., 2017). A previous study found that MtCEP1 expression is higher in nitrate-starved plants compared to those fed with mixed N (Imin et al., 2013), while our data indicate that CEP1 expression is higher in plants under 0.5 mM KNO₃ than in nitrate-starved plants. This apparent discrepancy can be explained by the strong negative impact of ammonia on NRT2.1 expression (Muños et al., 2004). We thus conclude that MtCEP1 expression is higher under 0.5 mM nitrate than under nitrate starvation. This is consistent with a recent study that shows that synthetic MtCEP1 peptide can increase nitrate content at 0.5 mM KNO₃, but not at higher or lower concentrations (Roy et al., 2021). Similar to findings reported in Arabidopsis, we found that MtCEP1 can induce MtNRT2.1 expression in a systemic manner, and that MtNRT2.1 is required for MtCEP1 induction by LN. However, MtCEP1 expression is repressed by MtNLP1 at HN (Luo et al., 2022). Here, we found that LN promotes MtCEP1 expression in an MtNLP1 and MtNRT2.1-dependent manner. We speculate that under LN conditions, a limited amount of MtNLP1 can enter the nucleus, where it directly induces low-level MtNRT2.1 expression and activates MtCEP1 expression through an unknown mechanism. This suggests that plants dynamically adjust the expression of CEPs to respond to environmental fluctuations in nitrate availability to appropriately regulate growth. Moreover, the enhancement of nodulation by MtCEP1 peptide was impaired in Mtnrt2.1 at LN, suggesting that MtCEP1 peptide and MtNRT2.1 form a regulatory loop for plant growth and nodulation under LN. MtCEP1 belongs to a large gene family, some of which are similarly regulated and have activities in nodulation. MtCEP1, 2, and 12 were shown to redundantly regulate lateral root number and nodulation (Zhu et al., 2021). Recently we showed that MtCEP1 and MtCEP2 are highly expressed in Mtcra2 roots in the presence or absence of nitrate (Luo et al., 2021). Moreover, MtCEP7, which is transcriptionally activated by MtNIN, similarly promotes nodulation through CRA2 (Laffont et al., 2020). Thus, it appears likely that CEP1 acts together with other CEPs to promote nodulation under LN conditions. In the study of LjNRT2.1, 0.2 mM nitrate failed to stimulate nodulation in wild-type plants (Misawa et al., 2022). This might be due to the different nitrate concentrations used. Moreover, there are so far no reports on CEP function in legumes that form determinate nodules, so whether the mechanisms identified here in Medicago are present in other legumes remains to be determined.

The stimulation of nodulation and nitrate content by LN and MtCEP1, although attenuated, was intact in Mtnrt2.1, suggesting that other nitrate transporters, such as MtNRT2.2 may be involved. However, the expression of MtNRT2.2 is very low (Pellizzaro et al., 2015), implying that it may not play an important role in the promotion of nodulation under LN. Knockout of another high-affinity nitrate transporter, MtNPF7.6, reduced nitrate responsiveness both at LN and HN, but its involvement in LN promotion of nodulation was not studied (Wang et al., 2020).

Based on our findings and the findings of others (Roy et al., 2021; Luo et al., 2022; Misawa et al., 2022), we propose that LN induces MtCEP1 expression, which systemically up-regulates MtNRT2.1 expression in a MtCRA2-dependent manner. Meanwhile, under LN, limited nuclear localization of MtNLP1 activates low-level MtNRT2.1 expression, thus enhancing nitrate uptake to improve plant growth and nodulation (Figure 8, A and B). However, HN triggers a major migration of MtNLP1 to the nucleus, where it activates CLE35 expression, which negatively regulates nodulation through SUNN in the shoot. In addition, MtNLP1 activates MtNRT2.1 expression to promote nitrate uptake, and further inhibit nodulation (Figure 8, C and D), and suppresses MtCEP1 expression which would otherwise positively regulate nodulation through MtCRA2 in the shoot (Luo et al., 2021, 2022).

Figure 8.

Proposed a model for LN and HN regulation of nodulation in M. truncatula. A, In the presence of low NO₃⁻ (0.5 mM), MtCEP1 is expressed and the encoded peptides systemically induce MtNRT2.1 expression through MtCRA2 in the shoot. MtNRT2.1 mediated nitrate uptake further enhances nodulation and MtCEP1 expression. B, Under low NO₃⁻, limited nucleus localization of MtNLP1 activates low-level MtNRT2.1 expression. C and D, In the presence of high NO₃⁻ (5 mM), nitrate triggers MtNLP1 shuttling to the nucleus, then MtNLP1 activates MtCLE35and represses MtCEP1 expressions which further systemically regulate nodulation. Meanwhile, MtNLP1 activates MtNRT2.1 expression and promote nitrate uptake which further represses nodulation.

Materials and methods

Plant materials and growth conditions

In this study, the R108 ecotype of M. truncatula was used as the WT. The Mtnrt2.1 mutant was obtained from the Tnt1 retrotransposon-tagged mutants of M. truncatula as described below. The Mtcra2-2 mutant was previously described by Huault et al. (2014). The Mtnlp1-1 mutant was previously described by Lin et al. (2018). The Agrobacterium rhizogenes strain AR1193 was used for M. truncatula root transformation, and the Agrobacterium tumefaciens EHA105 or GV3101 strains were used for N. benthamiana transient expression. Sinorhizobium meliloti 2011 was used in nodulation assays. M. truncatula seeds were scarified using sandpaper, then immersed in 10% sodium hypochlorite for 3–5 min, and washed with sterilized water. The sterilized seeds were imbibed overnight in sterilized water, and then transferred to 0.8% agar plates for germination. The germinated seedlings were grown on a mixture of perlite and vermiculite (1:1) or 1.2% N-free FP agar plates under a 16 h/8 h light (250 µmol m⁻² s⁻¹)/dark photoperiod at 22°C.

Mtnrt2.1 mutant genotyping

The Mtnrt2.1 Tnt1 insertion mutant (line NF6636) was identified from the Tnt1-mutant collection at Oklahoma State University (https://medicago-mutant.dasnr.okstate.edu/mutant/index.php). The MtNRT2.1 specific reverse primer and Tnt-R1 primer were used for genotyping. The insertion site was identified by sequencing the PCR-amplification products using the 2×Hieff PCR Master Mix (Cat No.10102ES60; Yeasen, Shanghai, China). The relevant primers are shown in Supplemental Table 1.

MtNRT2.1 overexpression and Mtnrt2.1 mutant complementation

Full-length MtNRT2.1 coding region was obtained by PCR from M. truncatula A17 genomic DNA and cloned into pDONR207, then transferred into pUb-GW-GFP by LR recombination reaction to generate pUb-MtNRT2.1. To generate the MtNRT2.1pro-MtNRT2.1 construct, the Lotus Ubiquitin promoter was replaced by a 2-kb MtNRT2.1 promoter using MssI/XbaI restriction sites using the ClonExpress II One-Step Cloning kit (Vazyme Biotech co., Ltd). The relevant primers are listed in Supplemental Table 1. The constructs were introduced into WT or the Mtnrt2.1 mutant by A. rhizogenes mediated hairy root transformation. The GFP reporter gene was used to identify transgenic roots using a fluorescence lamp (LUYOR-3415RG, LUYOR). Three weeks after transformation, the transgenic plants were transferred to a vermiculite and perlite (1:1) mixture and acclimatized for 3–5 days prior to inoculation with Sm2011. The plants were watered with the indicated concentrations of KNO₃ twice a week. The nodule number was scored 2 weeks after inoculation with rhizobia. These experiments were repeated 3 times.

K¹⁵NO₃ uptake in X. laevis oocytes

Oocyte preparation, cRNA injection, and ¹⁵NO₃⁻ uptake measurements were conducted as previously described (He et al., 2017). Briefly, # the coding sequence of MtNRT2.1 or MtNAR2.1 was amplified by PCR from M. truncatula A17 cDNA and was cloned into the oocyte expression vector pOO2 using BamHI/SpeI restriction sites. cRNA was synthesized using the Ambion mMessage mMachine kit according to the manufacturer's protocol. Oocytes were injected with 50 ng indicated cRNA. Water-injected oocytes were used as the negative control. Oocytes were incubated in ND-96 Ringer solution for 2 days as described (Li et al., 2010). Then the oocytes were incubated with 0.5 and 5 mM K¹⁵NO₃ at pH 5.5 for 3 h. Each oocyte was dried for 2 days at 70°C and the ¹⁵N contents were determined by a continuous-flow isotope ratio mass spectrometer coupled to a carbon–nitrogen elemental analyzer (Vario EL III/Isoprime; Elementar).

Analysis of nitrate uptake using ¹⁵NO₃⁻

Nitrate uptake was assessed by the ¹⁵NO₃⁻ stable isotope as described (Zhang et al., 2014; Misawa et al., 2022). 7-day-old seedlings grown on N-free FP medium were then transferred to new FP liquid medium with 0.5 mM or 5 mM KNO₃ for 4 h. Then, they were washed with 0.1 mM CaSO₄ for 1 min, and placed in FP hydroponics medium with 0.5 mM or 5 mM K¹⁵NO₃ with 99% atom excess of ¹⁵N for 30 min. Afterward, plants were washed for 1 min in 0.1 mM CaSO4 and again washed at least three times by ultrapure water; then, roots were separated from shoots. Each sample was dried for 2 days at 65°C and analyzed for ¹⁵N contents by a continuous-flow isotope ratio mass spectrometer coupled to a carbon–nitrogen elemental analyzer (Vario EL III/Isoprime; Elementar).

Analysis of nitrate content using HPLC

Nitrate content was measured by HPLC (Agilent 1200 series) as described (Meng et al., 2016). Wild-type and Mtnrt2.1 mutant plants were grown in FP agar medium without nitrate for 12 days and then pretreated with 1 μM MtCEP1 peptide or ultrapure water control for 48 h as described by Roy et al. (2021). Next, 5 mL of the indicated concentration of nitrate solution in ultrapure water was added to each square plate (grown 8–10 seedlings). Roots were sampled and washed three times with ultrapure water. Three to five roots were pooled together to make a biological repeat: 5–6 biological repeats each time. This experiment was repeated three times.

Nitrate content measurement was performed using the salicylic acid method (Xu et al., 2016; Hachiya and Okamoto, 2017). Briefly, root samples (50 or 100 mg) were frozen in liquid N and crushed into powder using a SM-48R Freeze Grinder (Shanmi, China). Then, 1 mL of ultrapure water (preheated at 80°C) was added into a 2 mL tube followed by incubating at 100°C for 20 min. The tubes were centrifuged at 20,400g for 10 min. Next, 10 µL of the supernatant was transferred into a 1.5 mL tube and 40 µL of 0.5% (w/v) salicylic acid in sulfuric acid was added, and then, the samples were vortexed well. The reactions were incubated at room temperature for 20 min; then 1 mL of 8% NaOH solution was added gently and vortexed. The OD 410 nm absorbance values of the samples were determined by a Varioskan LUX Multimode Microplate Reader (Thermo Fisher Scientific). The nitrate content was calculated using the formula: [true nitrate concentration (mM)] × [extracted volume (mL)]/[fresh weight (g)]. A standard curve was made with KNO₃ dilution series (0, 0.125, 0.25, 0.5, 1, 2, 4, and 8 mM). This experiment was repeated 3 times.

ARA assays

ARA experiments were conducted according to the previous description (Luo et al., 2021). The plants were harvested at 2 wpi, roots were removed from the plant and placed in 10 mL sealed glass vials (three plants per sample), and then, 0.5 mL acetylene was injected into each vial and incubated for 30 min at 28°C. Acetylene and ethylene concentrations were measured using a gas chromatograph (GC9310, Shanghai Chromatographic Instruments Co., Ltd., Shanghai, China). These experiments were repeated 3 times: 3–5 biological repeats each time.

CEP1 peptides treatment and rhizobium inoculation

MtCEP1 domain 1 peptide with hydroxyprolines (HyP) 4- and HyP11-modification were synthesized by GL Biochem Pty Ltd (Shanghai, China; Mohd-Radzman et al., 2015). WT and Mtnrt2.1 mutant germinated seedlings were transferred to the vermiculite and perlite mixture pots grown for 7 days and pretreated with 0.5 mM KNO₃, then treated twice with 1 mM CEP1 peptide. One week later, the 14 days old plants were inoculated with Sm2011 (OD600 = 0.02). Then, the plants were continuously watered with the 0.5 mM KNO₃ supplement with 1 mM CEP1 twice a week. The nodule number, ARA, growth phenotype, and nitrate accumulation were scored at 2 wpi. This experiment was repeated 2 times: 10–15 plants for each biological repeat.

Split-root experiments

To generate split roots, we used A. rhizogenes AR1193-induced adventitious roots as described in the M. truncatula Handbook, and the roots were allowed to grow in N-free FP agar plates for additional 7 days. Then, the two-part roots of each plant were transferred into a 90 mm petri dish with two compartments (Leiboer, China) containing a FP medium on one side and a FP medium supplemented with 1 μM MtCEP1 peptide on the other side. Three days later, the two root parts were separately harvested to analyze the gene expression. This experiment was repeated 3 times.

Gene expression analysis

Total RNA was extracted from M. truncatula roots using the TRIzol Reagent (Invitrogen). RNA quantification was measured by using a Nanodrop 2000 (Thermo Fisher). First-strand cDNA was prepared from 1 μg total RNA using the TransScript One-Step gDNA Removal kit and the cDNA Synthesis SuperMix (Transgen, China). RT-qPCR was performed using a Step-one Plus PCR system (Thermo Fisher, USA) with a TB Green Premix Ex Taq (Takara, China). MtUbiquitin10 (MtUBQ10) was used as a reference gene for normalization. The relative levels of gene expression were calculated using the 2^−△△Ct method. The primers used in RT-qPCR are listed in Supplemental Table 1.

Promoter:GUS analysis

The upstream 1499 bp promoter region of MtNRT2.1 was amplified by PCR using M. truncatula A17 genomic DNA as a template. The PCR product was inserted into the pCAMBIA1381 vector via BamHI/NcoI restriction sites and confirmed by sequencing. The relevant primers are shown in Supplemental Table 1. The pMtNRT2.1:GUS construct was then introduced into M. truncatula by A. rhizogenes mediated transformation to generate transgenic roots. These transformed plants were then transferred into vermiculite/perlite (1:1) pots and inoculated with Sm2011. Transgenic roots were harvested at indicated time points and then incubated in a GUS staining solution at 37°C for 30–60 min.

MtNLP1 overexpression and dexamethasone/cycloheximide treatments

MtNLP1 overexpression and dexamethasone induction of gene expression were performed as described previously (Luo et al., 2021). A previously reported 35S-GFP-MtNLP1 construct (Lin et al., 2018) was used for overexpression. MtNLP1-overexpressing composite plants were supplemented with 0, 0.5, and 5 mM KNO₃. A previously created pUb-NLP1-GR plasmid (Luo et al., 2021) was used for the DEX induction of gene expression, and the transformed roots of plants were incubated with 10 µM dexamethasone (DEX) or with cycloheximide (CHX) + DEX combination (10 µM each) for 4 h or 24 h. The transgenic roots were then harvested for RNA extraction and analyzing gene expression.

Dual-luciferase (LUC) reporter assays

The upstream 2-kb promoter regions of MtNRT2.1 was amplified from M. truncatula A17 genomic DNA, and cloned into the pGreenII-0800 vector to generate MtNRT2.1pro:LUC reporter construct. MtNRT2.1pro-ΔNRE:LUC was amplified by PCR using the MtNRT2.1pro:LUC vector as a template and recombination by the ClonExpress II One-Step Cloning kit. Primers used are listed in Supplemental Table 1. A previously reported 35S-MtNLP1 was used as an effector construct (Luo et al., 2021). These constructs were then introduced into N. benthamiana leaves by A. tumefaciens GV3101-mediated transient transformation. The dual-luciferase assay was performed using the Dual-Luciferase Reporter Assay System (Promega), according to the manufacturer's protocol. This experiment was repeated 3 times.

Yeast one-hybrid assay

About 300 bp upstream of the initiation codon of MtNRT2.1 promoter was amplified using the MtNRT2.1pro:LUC vector as a template and the MtNRT2.1pro-ΔNRE was generated by PCR using the MtNRT2.1pro-ΔNRE:LUC vector as a template, then inserted into the pHIS2 vector through EcoRI and MluI restriction sites. Primers used are listed in Supplemental Table 1. The pGADT7-NLP1 construct was previously described (Luo et al., 2021). These different vectors were co-transformed into the Y187 yeast strain using the Matchmaker One-Hybrid Library Screening System (Clontech). Transformed yeasts were selected on SD/-Leu-Trp and SD/-Leu-Trp-His medium supplemented with the 100 mM 3-AT antibiotic, and yeast cells were diluted in a 10× dilution series.

NLP1 knockdown by RNAi

A previously reported pUb-MtNLP1-RNAi construct (Lin et al., 2018) was introduced into WT or the Mtnrt2.1 mutant by A. rhizogenes mediated hairy root transformation. The pUb-GWS-GFP EV was used as a control. The GFP reporter gene was used to identify transgenic roots. Three weeks after transformation, the transgenic plants were transferred to a vermiculite and perlite (1:1) mixture and acclimatized for 3–5 days prior to inoculation with Sm2011. The plants were watered with the indicated concentrations of KNO₃ twice a week. The nodule number was scored at 2 wpi.

NLP1 subcellular location

The pDONR207-MtNLP1 (Lin et al., 2018) was recombined into a modified pK7WGF2 (a pUb-nls-DsRed cassette was inserted using XbaI and BsPTI restriction sites) via LR reaction. This construct was expressed in WT hairy roots and the nls-DsRed reporter gene, which was used as a nuclear localization marker. The transgenic plants were grown on N-free FP medium for 10–14 days, and were then incubated with 0, 0.5, and 5 mM KNO₃ for 10–20 min before observation using a confocal microscope (TCS SP8 STED, Leica). This experiment was repeated 3 times, and 5–10 plants were used each time.

Phylogenetic analysis

The protein sequences were obtained from Phytozome (https://phytozome-next.jgi.doe.gov/), and were aligned using ClustalW. Phylogenetic analysis was performed in MEGA11 software using the neighbor-joining method with the bootstrapping value set at 1,000 replications. The percentage of replicate trees in which the associated taxa clustered is shown below the branches. The tree is drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree. The evolutionary distances were computed using the Poisson correction method. All ambiguous positions were removed for each sequence pair (pairwise deletion option). Sequence alignment and machine-readable tree files are provided as Supplemental Files 1 and 2, respectively.

Statistical analyses

Statistically significant differences were calculated using GraphPad Prism version 8 (GraphPad software). Statistical analyses were conducted using the two-tailed t test or ANOVA test. One-way or two-way ANOVAs were performed with Tukey's or Dunnett's multiple comparisons test at a significant level of P < 0.05. For each boxplot, the center line in the box shows the median; the box limits are the upper and lower quartiles; and the whiskers represent the maximum and minimum values. Statistical data are provided in Supplemental Data Set 1.

Accession numbers

MtNRT2.1, MtrunA17_Chr4g0028551; MtNRT2.2, MtrunA17_Chr4g0028541; MtCEP1, MtrunA17_Chr8g0374821; MtNIA1, MtrunA17_Chr5g0424491; MtNIR1, MtrunA17_Chr4g0044831; MtCLE13, MtrunA17_Chr4g0040951; MtCLE35, MtrunA17_Chr2g0325381; MNIN, MtrunA17_Chr5g0448621; MtNF-YA1, MtrunA17_Chr1g0177091; MtNLP1, MtrunA17_Chr3g0143921 from the M. truncatula genome v5.0 (https://medicago.toulouse.inra.fr/MtrunA17r5.0-ANR/); MtNRT2.1, MtrunR108HiC_017743; MtNRT2.2, MtrunR108HiC_017742 were retrieved from the M. truncatula R108 genome v1.0 (https://medicago.toulouse.inrae.fr/MtrunR108_HiC/).

Supplemental data

The following materials are available in the online version of this article.

Supplemental Figure 1. Phylogenetic tree of the NRT2 family from M. truncatula, L. japonicus, and A. thaliana. (Supports Figure 1.)

Supplemental Figure 2. MtNRT2.1 expression pattern in M. truncatula root nodule formation. (Supports Figure 1.)

Supplemental Figure 3. Identification of the Mtnrt2.1 mutant. (Supports Figure 1.)

Supplemental Figure 4. Time course of MtNRT2.1, MtCLE35, MtNIA1, and MtNIR1 gene expression after LN treatment. (Supports Figure 3.)

Supplemental Figure 5. The growth phenotypes of Mtnrt2.1 in the presence of low- and high- nitrate. (Supports Figure 3.)

Supplemental Figure 6. Genetic complementation of Mtnrt2.1 at low nitrate. (Supports Figure 3.)

Supplemental Figure 7. MtCEP1 peptide promotes nodulation at LN in a MtNRT2.1 dependent way. (Supports Figure 4.)

Supplemental Figure 9. MtNLP1 accumulates in the nuclei in response to low- and high-nitrate. (Supports Figure 6.)

Supplemental Figure 10. Phenotype of overexpression MtNRT2.1 in WT and Mtnlp1. (Supports Figure 7.)

Supplemental Figure 11. Nodulation phenotype in MtNLP1 RNAi in Mtnrt2.1. (Supports Figure 7.)

Supplemental Table 1. Primer sequences used in this study.

Supplemental Data Set 1. Statistical analysis data.

Supplemental File 1. The NRT2s protein sequence alignment for phylogenetic analysis.

Supplemental File 2. Machine-readable tree file for phylogenetic analysis.

Funding

This work was funded by grants from the CAS Project for Young Scientists in Basic Research (YSBR-011) and the Strategic Priority Research Program of the Chinese Academy of Sciences (XDB27040208).