Ammonium transporter 1 increases rice resistance to sheath blight by promoting nitrogen assimilation and ethylene signalling

Summary Sheath blight (ShB) significantly threatens rice yield production. However, the underlying mechanism of ShB defence in rice remains largely unknown. Here, we identified a highly ShB‐susceptible mutant Ds‐m which contained a mutation at the ammonium transporter 1;1 (AMT1;1) D358N. AMT1;1 D358N interacts with AMT1;1, AMT1;2 and AMT1;3 to inhibit the ammonium transport activity. The AMT1 RNAi was more susceptible and similar to the AMT1;1 D358N mutant; however, plants with higher NH4 + uptake activity were less susceptible to ShB. Glutamine synthetase 1;1 (GS1;1) mutant gs1;1 and overexpressors (GS1;1 OXs) were more and less susceptible to ShB respectively. Furthermore, AMT1;1 overexpressor (AMT1;1 OX)/gs1;1 and gs1;1 exhibited a similar response to ShB, suggesting that ammonium assimilation rather than accumulation controls the ShB defence. Genetic and physiological assays further demonstrated that plants with higher amino acid or chlorophyll content promoted rice resistance to ShB. Interestingly, the expression of ethylene‐related genes was higher in AMT1;1 OX and lower in RNAi mutants than in wild‐type. Also, ethylene signalling positively regulated rice resistance to ShB and NH4 + uptake, suggesting that ethylene signalling acts downstream of AMT and also NH4 + uptake is under feedback control. Taken together, our data demonstrated that the AMT1 promotes rice resistance to ShB via the regulation of diverse metabolic and signalling pathways.


Introduction
Rice sheath blight disease (ShB), caused by Rhizoctonia solani K€ uhn (R. solani), significantly threatens worldwide rice cultivation (Molla et al., 2020). It is estimated that the yield reduction caused by ShB ranged from 8 to 50%, based on disease severity, crop stage of disease infection and environmental conditions (Savary et al., 2000). All known examples of ShB resistance are due to a quantitative trait that is controlled by multiple genes in rice, namely QTLs (quantitative trait loci). Many QTLs have been identified based on resistance to R. solani in different rice cultivars, some of which have been mapped and functionally characterized (Li et al., 1995;Richa et al., 2016Richa et al., , 2017. The underlying molecular mechanisms of rice resistance to ShB have been extensively investigated. PR (pathogenesis-related) genes are known to be significant contributors to plant defence. Specifically, the PR5 family gene OsOSM1 was confirmed to improve rice resistance against ShB (Xue et al., 2016). Similarly, overexpression of the ethylene (ET) biosynthetic gene OsACS2 results in enhanced ShB resistance (Helliwell et al., 2013). A recent genome-wide association study (GWAS) demonstrated that the F-box protein ZmFBL41 interacts and degrades ZmCAD (a lignin biosynthesis enzyme) to inhibit ShB resistance (Li et al., 2019). Previously, we identified that IDD14 and IDD13 activate PIN1a to promote rice resistance to ShB (Sun et al., 2019a(Sun et al., , 2020 and DEP1 interacts with IDD14 to negatively regulate rice defence to ShB ( Liu et al., 2021a). Our previous study demonstrated that brassinosteroids (BRs) are negative regulators of ShB resistance in rice, whereas ET can enhance the resistance. RAVL1, a key transcription factor of BR signalling, directly activates BR and ET signalling-related genes to modulate the rice immunity to ShB . Previous studies demonstrated that transcription factors such as OsWRKY4, 13, 30 and 80 enhance ShB resistance in rice (John Lilly and Subramanian, 2019;Peng et al., 2012Peng et al., , 2016Wang et al., 2015). Later, we proposed that OsWRKY53 functions as a negative regulator in rice resistance to ShB . In a more recent study, we identified that rice sugar transporters SWEET11 and SWEET14 negatively and positively regulate the rice resistance to ShB, respectively (Gao et al., 2018;Kim et al., 2021). Also, DOF11 promotes rice resistance to ShB by direct activation of SWEET14 (Kim et al., 2021).
Previous studies have revealed that high doses of nitrogen (N) fertilizer can cause a significant increase in the occurrence of ShB (Molla et al., 2020). However, limited N supply will restrict growth and yield in plants. Therefore, it is of great significance to identify genes with high nitrogen use efficiency (NUE), high resistance and high yield under low N conditions. Paddy-soil grown rice uses ammonium (NH 4 + ) as the primary nitrogen source (Britto et al., 2001). There are at least ten OsAMTs that mediate NH 4 + uptake in the rice genome. Three polarly localized members OsAMT1;1, OsAMT1;2 and OsAMT1;3 of the AMT1 subfamily are the primary ammonium transporters, specifically under low NH 4 + conditions. The three members are cooperatively responsible for NH 4 + uptake in rice (Konishi and Ma, 2021). Overexpression of OsAMT1;1, a key transporter of NH 4 + increases NUE, develops larger plants, increases yield under limited NH 4 + and is involved in the rice defence response against pathogens (Pastor et al., 2014;Ranathunge et al., 2014), suggesting that NH 4 + uptake plays important roles in the balance of rice growth and defence. However, the detailed underlying molecular mechanisms remain unknown.
Here, the role of AMT1-mediated NH 4 + uptake and subsequent assimilation in rice defence to ShB was investigated. The data suggest that N-metabolites rather than NH 4 + regulate rice defence. In addition, the genetic and physiological experiments demonstrated that chlorophyll and amino acids, but not c-Amino acid butyric acid (GABA) metabolism, positively regulate ShB resistance. N-dependent gene expression analysis in AMT1;1 OX and AMT1 RNAi identified that ethylene biosynthetic and signalling genes were under the control of AMT1. Interestingly, ethylene signalling controls the NH 4 + -dependent AMT1 induction via feedback regulation. Taken together, our analyses provide insight into the molecular mechanism of N transport and assimilation in ShB resistance in rice and identify the new signalling pathways by which rice modulates ShB defence under NH 4 + fertilizer.

Results
AMT1;1 D 358 N mutant accumulates less NH 4 + and is more susceptible to ShB Previously, we have isolated ShB-resistant and susceptible genes via Ds transposon tagging in rice mutants (Sun et al., 2019b). Among the lines, one more susceptible mutant (Ds-m (m)) was identified in this study (Figure 1a,b). However, Southern blot analysis indicated that Ds-m did not contain the Ds fragment ( Figure 1c). Since Ds-m leaves showed a pale green phenotype and accumulated less chlorophyll compared to wild-type (WT) (Figure 1d,e), AMTs, GS/GOGAT, and chlorophyll biosynthetic and catabolic genes were sequenced (data not shown). Interestingly, the sequencing results identified that G 1072 of AMT1;1was changed to A, which results in amino acid replacement from aspartic acid (D) 358 to asparagine (N) (Figure 1f). D 358 is located at the transmembrane helix (Figure 1g), and AMT1;1 D 358 N mutants accumulated less NH 4 + than did in WT plant roots ( Figure 1h).
To explore how N metabolism affects the interaction between rice and R. solani, the function of amino acids during R. solani growth was examined. The results indicated that some common amino acids, such as glutamate (Glu), glutamine (Gln), aspartic acid (Asp), asparagine (Asn), phenylalanine (Phe), proline (Pro) and alanine (Ala), promoted the growth of hyphae in low N concentrations, while the growth of mycelia was inhibited in high N concentrations (Figure 4f,g and Figure S3). Also, amino acid concentration measurements demonstrated that AMT1;1 OX mutant contained more, but AMT1 RNAi contains less Glu and Gln than WT plants ( Figure 4h). c-Amino Butyric Acid (GABA) is a non-protein amino acid that is also an important product of N metabolism. It has been reported to play key roles in a variety of physiological processes in plants (Deng et al., 2020). However, overexpressing glutamate decarboxylase (GDCi-OX) ( Figure S4a, b), an enzyme that converts glutamate to GABA, increased the susceptibility of rice to ShB ( Figure S4b,c). In addition, high concentrations of GABA (10 mM) promoted hyphae growth ( Figure S4d,e), suggesting that amino acids and not GABA may regulate rice resistance to ShB.

Chlorophyll accumulation promotes rice defence in response to ShB
Nitrogen metabolism is involved in the synthesis of many nitrogen-containing compounds (Baslam et al., 2020). A notable example is glutamate that forms 5-aminolevulinic acid (ALA) through glutamyl tRNA reductase (GluTR) and glutamate-1semialdehyde aminotransferase (GSA), functioning as the precursor to chlorophyll. Therefore, N is an important component of chlorophyll (Eckhardt et al., 2004). AMT1 RNAi and gs1;1 showed a pale green leaf phenotype and contain less chlorophyll, while AMT1;1 OX and GS1;1 OX accumulated more chlorophyll (Figure 5a,b,c). Also, our previous transcriptome data showed that R. solani inoculation altered N uptake, assimilation and chlorophyll synthesis (CHLD, CHLI, CHLM, PORB, DVR, CHLG and     (Figure 6a,b) and MN (Figure 6c,d) conditions. However, the positive effect of AMT1;1 to rice defence against R. solani was eliminated under the HN fertilization conditions with no significant resistance differences among AMT1;1 OX, AMT1 RNAi and WT plants (Figure 6e,f). Total N content was also measured in AMT1;1 OX, AMT1 RNAi (e) The interaction of AMT1;1 D 358 N and AMT1;1, AMT1;2 or AMT1;3 was tested via split-ubiquitin yeast two-hybrid assays. (f) Yeast growth assay was performed using Dmep123 strain to detect the NH 4 + transport activity of AMT1;1, AMT1;2 or AMT1;3 with co-expression of AMT1;1 D 358 N. Dmep123 was also used to generate the strain DL1 by integrating AMT1s into the Gap1 gene to generate DGap1::AMT1;1, DGap1::AMT1;2 and DGap1::AMT1;3. pDR-f1 vector was used to express AMT1;1 D 358 N. The yeast cells were grown on solid yeast nitrogen-based (YNB) medium at pH 5.2 containing 2% glucose, 2 mM ammonium chloride or 1 mM arginine as sole N source, at 28°C for 3 days.   (Figure 6g,h).
Feedback activation of NH 4 + uptake by ethylene signalling is important for rice resistance to ShB Our previous transcriptome study identified that ET biosynthesis and signalling genes were up-regulated after NH 4 + treatment . The qRT-PCR results verified the transcriptome data where ACO2, ACO3, EIN2, EIL1 and ERFs were significantly induced with NH 4 + treatment (Figure 7a). Previously, we identified that ET signalling positively regulates rice resistance to ShB   Figure S5). Interestingly, we found that NH 4 +mediated induction of AMT1;1 and AMT1;2 was inhibited in eil1 mutants   (Figure 7c) and that eil1 mutants accumulated less NH 4 + than WT plants (Figure 7d).

AMT1;1 OX increases yield and resistance under LN conditions
Previous studies demonstrated that overexpression of AMT1;1 enhances NH 4 + uptake and improves rice growth and yield at least under specialized N fertilization conditions (Ranathunge et al., 2014). Tillering is an important trait for grain yield in rice. Czapek-Dox medium with the addition of different concentrations of amino acids (glutamic acid and aspartic acid), and the colony diameter (with original cake diameter) was measured after 48 hours. The experiments were repeated at least ten times. Data represent means AE standard error (SE) (n > 10). Significant differences at P < 0.05 are indicated by different letters. (h) Glutamate and glutamine concentrations were measured from 10-day-old wildtype, AMT1;1 OX1 and AMT RNAi plant leaves grown in hydroponics for 4 weeks with a spectrophotometric method. Data represent the means AE standard error (SE) (n > 10).  Mature AMT1;1 RNAi plants developed significantly fewer tillers than in WT and AMT1;1 OX, while WT and AMT1;1 OX produced a similar number of tillers (Figure 8a,b). Furthermore, less filled grains per panicle and lower total grain yield per panicle were found in AMT1;1 RNAi than in WT and AMT1;1 OX. However, no differences were identified between WT and AMT1;1 OX (Figure 8c,d). The thousand-grain weight was similar between WT and AMT1;1 RNAi, while AMT1;1 OX was higher than WT ( Figure 8e).

Discussion
ShB is one of the most important diseases, which severely affects the quality and quantity of production in rice. However, the underlying rice defence mechanisms remain largely unknown. In this study, the AMT1;1 function in rice defence to ShB was explored by analysing the roles of NH 4 + and N-metabolites as well as ET signalling during the defence process. The data illustrated that AMT1;1-mediated NH 4 + transport accelerated N metabolism and regulated subsequent NH 4 + -dependent ethylene-related gene expression to promote rice resistance to ShB under limited N fertilizer conditions, suggesting that appropriate N uptake and assimilation are necessary for rice defence activation.
AMT1;1 D 358 N interacts with and inhibits AMT1;1, AMT1;2 and AMT1;3 to control rice defence A pale green mutant Ds-m was identified in the Ds-tagging mutant pool, which was more susceptible to ShB than WT. However, Southern blot results verified that the Ds-m phenotype was not caused by a Ds insertion. Since glutamate is the precursor of chlorophyll and forms ALA through GluTR and GSA (Eckhardt et al., 2004), AMT, GS/GOGAT, and chlorophyll biosynthetic and catabolic genes were sequenced in the Ds-m mutant.  (Yuan et al., 2013). Our analysis indicated that AMT1;1 D 358 N interacted with AMT1;1, AMT1;2 and AMT1;3 and inhibited their NH 4 + transport activity. AMT1;1, AMT1;2 and AMT1;3 are colocalized in the endodermis cell layer and are cooperatively responsible for the NH 4 + transport in rice (Konishi and Ma, 2021), suggesting that AMT1;1 D 358 N-mediated inhibition of AMT1;1, AMT1;2 and AMT1;3 activity can occur in planta.
N-metabolites, but not NH 4 + , promote ShB resistance in rice AMT1;1 D 358 N and AMT1 RNAi plants that contained less cellular NH 4 + were more susceptible. However, AMT1;1 OX and pAMT1;1-high-capacity Amtrac plants that accumulated more cellular NH 4 + were less susceptible to ShB, suggesting that cellular NH 4 + is positively correlated with rice resistance. Next, the key glutamine synthetase gene mutant gs1;1 was more susceptible while GS1;1 OX was less susceptible to ShB, despite the increased and reduced cellular NH 4 + in gs1;1 and GS1;1 OX respectively.
These results suggest that NH 4 + may not be the molecule responsible for the control of ShB resistance in rice. To definitively verify these results, AMT1;1 was overexpressed in the gs1;1 background. The results demonstrated that AMT1;1 OX/gs1;1 was similar to gs1;1 where increased cellular NH 4 + was accumulated and the plants were more susceptible to ShB. Therefore, AMT1;1-mediated rice resistance is required during the NH 4 + assimilation process. NH 4 + is incorporated into the glutamine amide group by GS (Mur et al., 2017). A recent study reported that amino acid metabolism is an important process in the nitrogen-mediated plant defence mechanism (Sun et al., 2020). Therefore, the direct role of amino acids on the growth of R. solani hyphae was investigated. The amino acids tested including Glu and Gln all inhibited R. solani growth at high concentrations. The AMT1;1 OX plants accumulated more Glu and Gln compared to WT. These data suggest that AMT1;1-mediated defence may partially act via accumulation of amino acids to inhibit R. solani growth. However, GABA is a non-protein amino acid that promoted R. solani growth even at high concentrations. Also, GABA biosynthetic gene overexpression plants GDCi OX were more susceptible to ShB, indicating that GABA negatively regulates rice resistance to ShB. Glutamate is the precursor of chlorophyll (Eckhardt et al., 2004), and AMT1;1 D 358 N, AMT1 RNAi and gs1;1 accumulated less chlorophyll and were more susceptible to ShB. Our transcriptome results suggested that R. solani infection significantly suppressed chlorophyll biosynthesis gene expression while inducing chlorophyll catabolic gene expression, suggesting a potential function of chlorophyll in rice defence. A genetic study by testing the chlorophyll biosynthetic gene DVR and YGL8 mutants (Kong et al., 2016;Nagata et al., 2005), as well as chlorophyll catabolic gene NYC3 mutant (Cao et al., 2021), revealed that chlorophyll content was positively correlated with rice resistance to ShB (Cao et al., 2021). These results suggest that AMT1-mediated NH 4 + transport and assimilation promote chlorophyll synthesis by which rice partially increased resistance to ShB.
Ethylene signalling activates NH 4 + uptake via feedback regulation to promote rice resistance to ShB Cellular NH 4 + is not only used to synthesize amino acids but also functions as a signal molecule to regulate global gene expression (Patterson et al., 2010). We further investigated whether other signalling pathways regulate AMT1;1-mediated rice resistance, aside from N metabolism. Plant hormone signalling is tightly associated with rice defence to ShB (Molla et al., 2020). We previously identified that NH 4 + treatment regulates the expression of auxin signalling genes  and demonstrated that auxin signalling activation via exogenous IAA application improves rice resistance to ShB (Sun et al., 2019a), implying that NH 4 + supply may modulate auxin signalling to regulate rice resistance to ShB. In addition, our previous studies identified that ET biosynthesis and signalling genes were induced by NH 4 + treatment  and that ET signalling promotes rice resistance to ShB , suggesting that NH 4 + signalling may activate ET signalling to promote rice resistance. Our analyses identified that EIL1 and EIL2 which activate ethylene signalling were positively regulated while ETR2 and ERS1, two negative regulators of ethylene signalling, were suppressed by AMT1;1 under the LN conditions. However, under the HN conditions, EIL1 and EIL2 expression levels were significantly lower while ETR2 and ERS1 levels were higher in AMT1;1 OX than in WT. These results suggest that ET signalling may be sensitive to the cellular N levels and may be associated with rice resistance to ShB. Furthermore, we identified that NH 4 + -mediated induction of AMT1 genes was inhibited in the key ET signalling gene eil1 mutant. The eil1 mutant accumulated less NH 4 + , suggesting that ethylene signalling controls NH 4 + transport via feedback regulation to fine-tune the cellular N transport and assimilation, which may be important for rice defence and growth.

AMT1;1 OX increases rice resistance and NUE under limited N fertilizer
Nitrogen fertilizers supplied to rice crops are partially lost via various mechanisms including ammonia volatilization, denitrification and leaching, causing environmental concerns by polluting the atmosphere, aquatic systems and groundwater (Choudhury and Kennedy, 2005 increase yield production at least under specialized N fertilization conditions (Ranathunge et al., 2014). Our data also confirmed that AMT1;1 OX produced a relatively higher yield, suggesting that AMT1;1 OX increased NUE and ShB resistance in rice. In this study, the data demonstrated that AMT1;1-mediated increase in rice resistance was via N-metabolite activation and ethylene signalling. This study demonstrates the precise use of nitrogen based on the underlying molecular mechanisms of N metabolism to improve yield production and immunity against ShB and other pathogens in rice.

Materials and methods
Plant growth and R. solani AG1-IA inoculation All of the rice plants treated with R. solani were cultured in the Shenyang Agriculture University greenhouse at 23-30°C, 80% relative humidity (RH) and 12-h light/12-h dark photoperiod. Nicotiana benthamiana plants were grown in environmental chambers at 22-24°C, 80% RH and 16-h light/8-h dark photoperiod for 4 weeks before use. The R. solani strain AG1-IA was cultured on solid PDA (Potato Dextrose Agar) medium at 28°C in an incubator. Rice was inoculated according to previously reported methods (Cao et al., 2021).

Molecular phylogenetic analysis using maximum likelihood
The amino acid sequences of AMT proteins in rice, maize, Arabidopsis, wheat and potato were used as bait for searching in the Uniprot database (http://www.uniprot.org/) using BLASTp. MSAs of these protein sequences were conducted using the Clustal Omega program (Larkin et al., 2007). Phylogenetic relationships were inferred using the maximum likelihood (ML) methods with 1,000 bootstrap iterations (Kumar et al., 2016).
Vector construction and transgenic plant generation AMT1;1 promoter was fused to an Amtrac high-capacity gene ORF in the pGA1611 binary vector. The primers used for plasmid construction are listed in Table S3. pAMT1;1:Amtrac high capacity was transformed into Japonica rice cultivar Dongjin (DJ) calli via Agrobacterium-mediated transformation method (Hiei et al., 1994). The gs1;1 mutant (PFG_3A-09512) was obtained from a rice T-DNA mutants collection (http://signal. salk.edu/cgi-bin/RiceGE/) (An et al., 2003). The overexpression of GS1;1 and GDCi was generated from the rice cultivar Zhonghua 11 (ZH11). The modified pCAMBIA1381-Ubi vector was used to construct GS1;1 and GDCi overexpression vectors at HindIII/KpnI and HindIII/HpaI respectively. The primers used for the GS1;1 and GDCi overexpression vector constructions are listed in Table S4.
Analysis of amino acid effects on R. solani growth R. solani AG1-IA was cultured on a Czapek-Dox medium with the addition of different concentrations of amino acids. Colonized PDA plugs (7 mm in diameter) were excised using a hole borer and transferred to the centre of the fresh media surface. These petri dishes were then cultured in a 37°C incubator for 42 hours, and the diameters of the colonies were measured. The assays were conducted repeatedly at least eight times.

Determination of NH 4 + and total N content
The NH 4 + content in roots and shoots of 7-day-old rice seedlings was measured using an F-kit (Roche) according to the manufacturer's instructions (Oliveira et al., 2002). The total N content in rice plants was determined by the Kjeldahl method using the Hanon k1160 Automatic Kjeldahl nitrogen determinator (Shandong, China).

RNA extraction and qRT-PCR analysis
Total RNA was extracted from the one-month-old leaves from tested rice plants using TRIzol reagent (Takara, Dalian, Liaoning, China). Elimination of genomic DNA and reverse transcription reactions were performed according to the manufacturer's instructions using the commercial kit (Takara, Dalian, Liaoning, China). qRT-PCR analysis was performed using the CFX96 real-time PCR system (Bio-Rad, Hercules, CA, USA) and ChamQ Universal SYBR qPCR Master Mix (Vazyme, Nanjing, Jiangsu, China). Gene expression values were normalized against Ubiquitin values in the same samples. Two technical and three biological replicates were used for each analysis. The primers used for qRT-PCR are listed in the supplemental Table S2.

Determination of chlorophyll content
The chlorophyll content in leaves of one-month-old plants was determined using the ultraviolet spectrophotometer following a previously reported method (Lichtenthaler, 1987).

Amino acid measurement
Gln and Glu content was measured using an L-Glu analysis kit (Yamasa, Tokyo, Japan) following the manufacturer's instructions (Hirano et al., 2008).
Split-ubiquitin yeast two-hybrid assay AMT1;1, AMT1;2 and AMT1;3 were fused to the N-terminus of Ubiquitin through Nub vector pXN25_GW and AMT1;1 D 358 N was fused to the C-terminus of Ubiquitin through Cub vector pMETYC_GW based on standard GATEWAY cloning protocol (Invitrogen, CA, USA). Yeast two-hybrid assays were performed according to a previously published method (Lalonde et al., 2010).
BiFC and southern blotting assays AMT1;1 D 358 N was cloned into a YFP N vector, while AMT1;1, AMT1;2 and AMT1;3 were cloned into CFP C plasmids. The constructs were co-transformed into tobacco leaves using Agrobacterium strain GV3101 (Kim et al., 2009). The YFP fluorescence signals were observed under a confocal microscope (Olympus FV1000, Japan) 36 to 48 hours after infiltration. Southern Blotting assay of Ds insertion was carried out with reference to the method described by a previous study (Xuan et al., 2016).

Statistical analyses
Statistical analyses were conducted using Prism 5.0 software (GraphPad, San Diego, CA, USA) with a one-way analysis of variance (ANOVA) for comparison of significant differences between multiple groups. Also, Student's t-test was used to compare the differences between the two groups. Differences between the groups were considered significant with at least P < 0.05.

Supporting information
Additional supporting information may be found online in the Supporting Information section at the end of the article. Table S1. Gateway primers used in this study. Table S2. qRT-PCR and RT-PCR primers used in this study. Table S3. Primers used in this study for the construction of plants expressing AtAMT1;3 T464D-A141E driven by AMT1;1 endogenous promoter.  Table S4. Primers used in this study for the construction of the overexpression vector. Figure S1. Sensitivity test of AMT1;1 RNAi and overexpression plants to methyl-ammonium (MeA). Figure S2. AtAMT1;3 T464D-A141E expression promotes rice resistance to ShB. Figure S3. Verification of the effects of amino acids on R. solani growth. Figure S4. Identification of the effect of GABA on rice resistance against ShB. Figure S5.