Rice transcription factors OsLBD37/38/39 regulate nitrate uptake by repressing OsNRT2.1/2.2/2.3 under high-nitrogen conditions

Xinxin Zhu,Dujun Wang,Lijuan Xie,Tao Zhou,Jingyi Zhao,Qian Zhang,Meng Yang,Wenjuan Wu,Xingming Lian

National Key Laboratory of Crop Genetic Improvement and National Center of Plant Gene Research(Wuhan),Huazhong Agricultural University,Wuhan 430070,Hubei,China

Keywords:Oslbd37/Oslbd38/Oslbd39 Transcription factor Nitrate uptake Rice

ABSTRACT Nitrate()uptake involves a finely regulated and complex multilevel response system.Elucidating the molecular mechanism of nitrate uptake may lead to improving the growth and productivity of plants in the presence of dynamic variation in nitrate concentration.In this study,we identified three lateral organ boundaries domain(LBD)transcription factors,OsLBD37,OsLBD38,and OsLBD39,as regulators of nitrate uptake in response to nitrogen(N)availability.OsLBD37,OsLBD38,and OsLBD39 were induced by ammonium and glutamine in rice roots.Individual or collective knockout of OsLBD37,OsLBD38,and OsLBD39 led to increased concentrations of nitrate and increased expression of OsNRT2.1,OsNRT2.2,and OsNRT2.3 respectively under high-N conditions,whereas overexpression of each of these three LBD genes produced opposite effects where N accumulation was reduced.Dual-luciferase reporter assay further confirmed that OsLBD37,OsLBD38,and OsLBD39 possessed transcription inhibitory activities in rice protoplast cells,downregulating the expression of OsNRT2.1/OsNRT2.2/OsNRT2.3.Yeast two-hybrid and bimolecular fluorescence complementation assays showed that OsLBD37 interacted with OsLBD37,OsLBD38,and OsLBD39 in the nucleus.Together,these results show that OsLBD37,OsLBD38,and OsLBD39 collaborate to inhibit the expression of OsNRT2.1/OsNRT2.2/OsNRT2.3 transporters under N-sufficient conditions,thereby helping rice plants avoid excessive nitrate accumulation that may affect their growth.

1.Introduction

Nitrogen(N)is required for plant metabolism,growth,and development and is a major component of chlorophyll,nucleic acids,proteins,and secondary metabolites.Nitrate()and ammonium()are the major inorganic N sources for rice,and as much as 40% of total N uptake in irrigated rice is absorbed as nitrate after nitrification in the rhizosphere[1,2].Apart from its nutritional value,nitrate performs signaling functions in multiple biological processes[3-5].To cope with the heterogeneity and dynamic variation of nitrate concentrations,which range from＜100μmol L-1to higher than 10 mmol L-1in the soil solution[6],plant roots have evolved two uptake systems with different affinities:a low-affinity nitrate transport system(LATS)and a highaffinity nitrate transport system(HATS)[6,7].

Although it operates mainly under low-nitrate conditions,the HATS system is required to be instantly activated under nitratelimited conditions to meet the demands of N for plant growth,and to almost shut down under nitrate-sufficient conditions to avoid excessive accumulation and unnecessary energy loss[25,26].The expression in rice of NRT2 members,which are major components of the HATS system,can be finely controlled.Several transcription factors have been found to regulate the expression of NRT2 genes in rice.OsMADS57,a controller of rice tillering[27],was found to regulate nitrate translocation from roots to shoots by directly binding to the OsNRT2.3a promoter[28].OsBT is likely a repressor of OsNRT2.1,OsNRT2.2,and OsNRT2.3,and prevents the inducible expression of these NRT2 genes under low-nitrate conditions to coordinate N availability and plant growth[29].OsMYB305 was possibly involved in regulating the NRT2 genes,as overexpression of OsMYB305 increased the expression of OsNRT2.1,OsNRT2.2,OsNAR2.1,and OsNiR2,as well as the uptake of nitrate,but repressed genes involved in lignocellulose biosynthesis[30].The regulation of the NRT2 genes by the identified transcription factors was more likely to coordinate the relationship between nitrate absorption and plant growth than to directly control the expression of NRT2 under low-or highnitrate conditions.The regulatory mechanisms by which NRT2 genes are activated under low-nitrate conditions and how they are restricted under high-nitrate conditions are still unclear.

In this study,we identified three LBD transcription factors,OsLBD37,OsLBD38,and OsLBD39,as transcription suppressors of OsNRT2 transporters under nitrogen-sufficient conditions that assist rice plants to avoid excessive nitrate accumulation.

2.Materials and methods

2.1.Plant materials and growth conditions

The wild-type rice used in this study was Zhonghua 11(O.sativa L.japonica,ZH11).To generate overexpression constructs,the fulllength coding sequence of OsLBD37/OsLBD38/OsLBD39 was amplified from total cDNA of wild-type rice ZH11.Single lbd37,lbd38,lbd39,and triple lbd37/38/39 mutants were generated using the clustered regularly interspaced short palindromic repeat(CRISPR)/CRISPR-Cas9-mediated editing method[31]in ZH11.The constructs were introduced into Agrobacterium tumefaciens strain EHA105 and then transferred into ZH11,as described previously[32].

A hydroponic experiment was performed using standard rice nutrient solution containing 1.44 mmol L-1NH4NO3;0.5 mmol L-1K2SO4;1.0 mmol L-1CaCl2;1.6 mmol L-1MgSO4;0.17 mmol L-1Na2SiO3;0.3 mmol L-1NaH2PO4;50μmol L-1Fe-EDTA;0.06μmol L-1(NH4)6Mo7O24;15μmol L-1H3BO3;8μmol L-1MnCl2;0.12μmol L-1CuSO4;0.12μmol L-1ZnSO4;29μmol L-1FeCl3;and 40.5μmol L-1citric acid(pH 5.5)[33].All nutrient solutions were refreshed every three days.NH4NO3was supplied as an N source at 0.288 mmol L-1(LN;low nitrogen)and 1.44 mmol L-1(HN;high N).All experiments were performed in triplicate.

2.2.RNA extraction and qRT-PCR analysis

Total RNA was extracted from roots and leaves of thr ee-weekold rice plants under LN or HN conditions using the TransZol RNA Extraction Kit(TransGen,Beijing,China).Approximately 3μg of total RNA treated with DNase I was used to synthesize first-strand cDNA using oligo(dT)18 as primer.The product of first-strand cDNA was used as a template for PCR.For RT-qPCR,SYBR Premix Ex Taq(TaKaRa Bio Inc.,Shiga,Japan)was added to the reaction mixture in an optical 384-well plate,and the mixture was amplified on a QuantStudio 6 Flex PCR system(Applied Biosystems,Foster City,CA,USA)according to the manufacturer’s instructions.Three replicates were performed for each gene.The rice UBI gene was used as an internal control.The primer sequences are listed in Table S1.

2.3.RNA sequencing(RNA-seq)analysis

Rice seedlings were grown in culture solutions for 2 weeks,transferred to N-deficiency nutrient solution for 7 days,and then resupplied with N sources(1.44 mmol L-1NH4NO3).Root samples were collected 0 and 2 h after N reintroduction.Total RNA was extracted using a TransZol RNA Extraction Kit.Sequencing was performed using Illumina HiSeq Xten in the PE150 mode.Differentially expressed genes were identified by DEseq(https://www.huber.embl.de/users/anders/DESeq/)with a false discovery rate(FDR)＜0.01 and an absolute value of fold change value≥2.

2.4.Total root length assay

Roots of ZH11 and OsLBD37/38/39 transgenic rice seedlings were grown in N-sufficient(1.44 mmol L-1NH4NO3)medium for 10 days.The roots were then floated in water in a transparent plastic tray(20×15 cm)and scanned with a scanner(Epson V700,Beijing,China).Total root length and root diameter(0-0.2,0.2-0.5,0.5-1 and＞1 mm)in the images were estimated with WinRHIZO 2003b(Regent Instruments,Quebec,Canada).

2.5.Total N,nitrate,and ammonium concentration assays

Rice shoots and roots were harvested,dried at 80 °C for 3 days,and ground into powder.Powder samples(0.2 g)were digested with 5 mL of 98% H2SO4and 5 mL of 30% hydrogen peroxide.The samples were then held at 376 °C for 90 min.After cooling,each digested sample was diluted to 100 mL with distilled water.The total N concentration in the solution was determined colorimetrically at 660 nm using a modified Berthelot reaction with salicylate,dichloroisocyanurate,and complex cyanides in an automated discrete analyzer(SmartChem 200,Alliance,Paris,France).

Fresh rice roots and shoots were freeze-dried and ground into powder.For nitrate and ammonium analyses,20 mg of powder was extracted with 1 mL of H2O at 80 °C for 20 min.After centrifugation(13,000×g,10 min),the nitrate and ammonium concentrations in the supernatant were measured as previously described[34].

2.6.Collection and analysis of xylem sap

Wild type ZH11,single lbd37,lbd38,lbd39,and triple lbd37/38/39 mutants were cultured hydroponically.For measuring nitrate and ammonium concentrations,plants were grown in 0.288 mmol L-1NH4NO3or 1.44 mmol L-1NH4NO3for 6 weeks.The shoots(5 cm above the roots)were excised with a razor,and xylem sap was collected with a micropipette for 2 h after excision.Nitrate concentrations in the xylem sap were determined using inductively coupled plasma-mass spectrometry(ICP-MS)Agilent 7700 series(Agilent Technologies,Santa Clara,CA,USA).

2.7.15N uptake measurement

Nitrogen influx rate was assayed with15N as previously described[35].Wild type ZH11 and transgenic rice seedlings were grown in 1.44 mmol L-1NH4NO3nutrient solution for 10 d and then transferred to nitrogen-deficiency medium for 3 days.The seedlings were then transferred to 0.1 mmol L-1CaSO4for 1 min to remove any compounds adsorbed to the root surface,followed by 30 min exposure to 1.44 mmol L-115N-labeled NH4NO3(atom%15N:98%).At the end of the incubation period,roots were immediately rinsed with 0.1 mmol L-1CaSO4for 1 min.Shoots and roots were harvested,dried at 80°C for 3 days,and ground into powder.15N content was measured in an isotope mass spectrometer(Isoprime 100,Elementar Analysensysteme GmbH,Langenselbold,Germany).For this study,four plants were pooled as one biological replicate,and each treatment had three independent biological replicates.

2.8.Subcellular-localization assay

To investigate the subcellular localization of OsLBD37,OsLBD38,and OsLBD39,full-length cDNAs of OsLBD37/OsLBD38/OsLBD39 coding sequences were cloned into the pH7WGF2 vector to produce OsLBD37-GFP,OsLBD38-GFP,and OsLBD39-GFP fusion constructs driven by the cauliflower mosaic virus(CaMV)35S promoter(35S:GFP:OsLBD37/OsLBD38/OsLBD39).OsbZIP46 was cloned into the pH7WGR2 vector to produce the OsbZIP46-RFP fusion construct as a nuclear marker[36].The resulting vectors were transformed into rice protoplasts.The fluorescence signal was observed with a confocal microscope(FV1000;Olympus,Tokyo,Japan)after transformation for 16 h.The primer sequences are listed in Table S1.

2.9.Dual-luciferase reporter assay

An effector plasmid was constructed by fusing OsLBD37,OsLBD38,and OsLBD39 with a GAL4 DNA-binding domain(GAL4 DB)under the control of a CaMV 35S promoter(GAL4 DBOsLBD37/OsLBD38/OsLBD39).An empty vector(none)was used as control.The reporter plasmid contained the firefly luciferase gene driven by five copies of the GAL4 binding site(5×GAL4)or the CaMV 35S promoter+5×GAL4[37].The Renilla luciferase gene(R-LUC)driven by the 35S promoter was simultaneously co-transformed as an internal control.The luciferase assay(E1910;Promega,Madison,WI,USA)was performed according to the manufacturer’s instructions.Fluorescence was detected with a microplate spectrophotometer(SPARK;Tecan GmbH,Gr?dig,Austria).

2.10.Yeast two-hybrid assay

The coding region of OsLBD37 was cloned into the pDEST32 vector using Gateway technology(Invitrogen).OsLBD37/OsLBD38/OsLBD39 coding sequences were cloned into the pDEST22 vector.The combined plasmids were transformed into yeast strain MaV203,which was then plated on SD-Leu-Trp-His medium containing either 0(control)or 15,25,35,or 40 mmol L-13-amino-1,2,4-triazole(3-AT).

2.11.Bimolecular fluorescence complementation(BiFC)assay

Full-length cDNAs corresponding to OsLBD37,OsLBD38,and OsLBD39 genes were amplified from ZH11.The resulting amplicons were inserted into the cYFP(C-terminal fragment of YFP vector)or nYFP(N-terminal fragment of YFP vector)vectors to generate fusion constructs.Co-transformation of constructs(for example,those encoding LBD37-cYFP and LBD37-nYFP)into rice protoplast cells by Agrobacterium-mediated infiltration enabled testing of protein-protein interactions.After 24 h of incubation in the dark,the YFP signal was observed and photographed using a confocal microscope(FV1000;Olympus,Tokyo,Japan).Each BiFC assay was repeated at least three times.The relevant primer sequences are listed in Table S1.

3.Results

3.1.Expressions of transcription factors OsLBD37,OsLBD38,and OsLBD39 were highly induced by ammonium and glutamine

To investigate the molecular mechanism of the response of rice plants to external nitrogen,we first measured the changes in genome-wide transcription factors in rice roots with NH4NO3resupply after nitrogen deficiency using RNA-seq analysis.Removal of genes with absolute values of log2fold change(N sufficiency/N deficiency)and-log10FDR＞2 left 30 upregulated and 39 downregulated transcription factors(Fig.S1A).Among the 30 upregulated transcription factors,OsLBD37,OsLBD38 and OsLBD39 were from the same gene family and displayed highly significant NH4NO3-induced expression(Fig.S1B;Table S2).qRT-PCR further confirmed that the expression levels of OsLBD37,OsLBD38 and OsLBD39 were upregulated by respectively 7.1,9.6 and 3.2 times when resupplied with 1.44 mmol L-1NH4NO3after N deficiency compared to the absence of NH4NO3(Fig.S1C).To further investigate the specificity of OsLBD37/38/39 response to different N sources,we monitored the time-course expression of these two genes when plants were resupplied with NH4NO3,NH4Cl,KNO3,or glutamine(Gln)as an N source or with KCl as the control after N deficiency.OsLBD37/38/39 were induced by NH4NO3,NH4Cl,and Gln but showed nearly no response to KNO3,compared to the control KCl(Fig.1A-C).These results suggested that OsLBD37/38/39 were induced specifically byor Gln instead of.Considering that Gln was the direct assimilation product of,we further introduced methionine sulfoximine(MSX),a blocker of Gln synthesis from,into the expression experiment to distinguish the signals of Gln and.MSX partially blocked the induction of OsLBD37/38/39 when triggered bybut was not affected when triggered by Gln(Fig.1D-F).These results indicated that the presence of Gln was the direct cause of OsLBD37/38/39 induction by N supply.

3.2.Overexpression of OsLBD37,OsLBD38,and OsLBD39 led to reduced N accumulation

To further investigate the functions of OsLBD37,OsLBD38,and OsLBD39,overexpression lines driven by a CaMV 35S promoter were generated separately.For each of the three genes,two independent transgenic lines,which had been confirmed by qRT-PCR to be substantially overexpressed compared to wildtype plants,were selected for further study(Fig.2A-D).

As OsLBD37,OsLBD38,and OsLBD39 were all induced by N supply(Figs.1,S1),we investigated the growth of their overexpression lines(LBD37/LBD38/LBD39-OE)and the wild-type,hydroponically cultured at two N levels:low and high.Overexpression lines of each of the three genes showed no consistent differences in plant height,root length,and dry weight compared with the wild type under both LN and HN conditions,although the dry weights of shoots and roots were slightly lower in some overexpression lines than in the wild type(Fig.2E-H).In contrast,total N concentrations in both shoots and roots of most overexpression lines of the three genes were significantly lower than that of the wild type under both LN and HN conditions(Fig.2I,J).

Fig.1.OsLBD37,OsLBD38 and OsLBD39 are transcription factors involved in regulating N response that are strongly induced by ammonium or glutamine.(A-C)Induction of OsLBD37/38/39 by two N sources in roots of rice seedlings,validated by qRT-PCR.KCl was used as negative control.(D-F)Expression of OsLBD37/38/39 treated with methionine sulfoximine(MSX)used to inhibit the activity of glutamine synthetase(GS).Values are means±SD.n=3 biologically independent samples.Statistical comparison was performed by Tukey’s test:*,P＜0.05;**,P＜0.01.

Fig.2.Overexpression of OsLBD37,OsLBD38,and OsLBD39 reduced N accumulation under different levels of N supply.(A)The growth morphology of ZH11 and LBD37/LBD38/LBD39-OE transgenic lines grown in hydroponic culture.LN,0.288 mmol L-1 NH4NO3;HN,1.44 mmol L-1 NH4NO3.(B-D)Expression of OsLBD37/OsLBD38/OsLBD39 in the roots of ZH11 and LBD37/LBD38/LBD39-OE lines by qRT-PCR.(E-H)Plant height,root length,dry weight of shoots and roots of ZH11,and LBD37/LBD38/LBD39-OE transgenic lines.(I,J)Total N concentrations in shoots and roots.Con,concentration;DW,dry weight.Values are means±SD.n=6 biologically independent samples.Asterisks indicate significant differences between the ZH11 control and LBD37/LBD38/LBD39-OE transgenic lines by two-tailed Student’s t-test:*,P＜0.05;**,P＜0.01.

3.3.OsLBD37,OsLBD38,and OsLBD39 knockout resulted in weaker growth and increased N accumulation

To further confirm the roles of OsLBD37,OsLBD38,and OsLBD39 in rice growth and development,we generated single and triple mutant lines of OsLBD37,OsLBD38,and OsLBD39 via CRISPR/Cas9 technology(Fig.S2).Homozygous mutant lines were obtained by insertion or deletion at the designated target sites.T2seedlings were used to assess plant phenotypes in hydroponic cultures under both LN and HN conditions.Compared with the WT,the single lbd37,lbd38,lbd39,and triple lbd37/38/39 mutants showed no visible differences in plant height and root length,but the tiller numbers and the dry weights of shoots and roots in most mutant lines were significantly reduced(Fig.3A-E).Moreover,the triple mutants of these three genes seemed to show weaker growth than the single mutants,as revealed by tiller numbers and dry weights under both N-level conditions(Fig.3C-E).

Under LN conditions,the shoots of mutant lines(except for one line of LBD37)accumulated significantly higher(8.3%-26.1%)total N concentrations than those of the wild type(Fig.3F).When grown under HN conditions,the total N concentrations of shoots in only triple mutant lines were significantly higher(3.7%-4.6%)than those in the wild type(Fig.3F).In contrast,except for the two LBD38 mutant lines grown under LN conditions,the N concentrations in the roots were generally similar between the mutant and wild-type lines(Fig.3G).Root phenotype influences efficient N use by plants.In Arabidopsis,auxin-inducible LBD16 and LBD29 were identified as direct targets of ARF7 and ARF19 transcription activators that regulate lateral root formation[38].Generally,root diameter of rice is＜2 mm,and roots with 0.3-2 mm diameter are considered adventitious while smaller roots are considered laterals[39].In our study,the roots were divided into four groups according to their diameters(0-0.2,0.2-0.5,0.5-1 and＞1 mm).Compared with the ZH11,the total root length of the OsLBD37/38/39 mutants was significantly decreased in all groups(Fig.S3A).In contrast,the total root length of the overexpression seedlings showed no significant difference,especially in the smaller root diameter groups(0-0.2 and 0.2-0.5 mm)(Fig.S3A).These results suggested that knocking out the OsLBD37,OsLBD38,and OsLBD39 strongly affected lateral-root development.N concentrations were also measured in plants grown under field conditions.The straw of almost all single and triple mutant lines accumulated significantly higher N(12.0%-27.8%)than that of wild-type plants(Fig.S3B,C).

3.4.OsLBD37,OsLBD38,and OsLBD39 knockout increased nitrate accumulation under high-N conditions

To further investigate the mechanism by which OsLBD37,OsLBD38,and OsLBD39 affect N accumulation,we determined the contents of the two major absorbed N forms,and,in mutant and wild-type plants.At high N supply,the shootconcentrations of lbd37,lbd38,lbd39,and triple mutant lines were respectively 9.0%-13.7%,15.9%-27.1%,22.8%-28.3%,and 31.6%-39.2% higher than in the wild type,and those in roots were 15.5%-18.3%,40.8%-41.3%,45.9%-46.8%,and 38.4%-44.4% higher than in the wild type(Fig.4A,B).In contrast,at low N supply,shootconcentrations showed no difference between the mutant and the wild-type lines,and rootconcentrations in the lbd38 and lbd39 single mutants were slightly higher than those in the wild type(Fig.4A,B).Under HN conditions,theconcentrations in the xylem sap of almost all mutant lines were much higher than that of the wild type,whereas under the LN condition,triple mutant lines instead of the single lines accumulated higherconcentrations in xylem sap than the wild type(Fig.4C).In contrast to,the concentrations ofshowed almost no significant difference between the mutant lines and wild type plants under both LN and HN conditions(Fig.S3D,E),and the concentrations ofwithin the shoots of lbd37/38/39 triple mutant lines grown under HN conditions were slightly higher than those of the wild type(Fig.S3E).These results suggest that knocking out the OsLBD37,OsLBD38,and OsLBD39 promoted the accumulation of nitrate,but not of ammonium,under high N-supply conditions.

As nitrate uptake by roots is the main source of nitrate accumulation in plants,we further tested whether nitrate uptake was elevated in OsLBD37/38/39 mutants,using a15N trace experiment.Under 1.44 mmol L-1NO3,nitrate influx rates increased by 14.9%-29.8% in the single lbd37,lbd38,lbd39,and triple lbd37/38/39 mutants(Fig.S4).

3.5.OsLBD37/OsLBD38/OsLBD39 overexpression and knockout altering expression levels of OsNRT2 genes

Members of the NRT2 family are the major transporters for high-affinity absorption of nitrate in rice[8].To investigate the molecular mechanisms underlying the control by OsLBD37/OsLBD38/OsLBD39 ofaccumulation of,we measured the expression levels of key genes(OsNRT1,OsNRT2,OsNR2 and OsNiR1)involved in N uptake and assimilation.Compared with the wild type,the transcription levels of OsNRT2.1,OsNRT2.2,and OsNRT2.3 were significantly decreased in the LBD37/LBD38/LBD39-OE lines under both LN and HN conditions(Fig.5A-C).Correspondingly,the transcription levels of OsNRT2.1,OsNRT2.2,and OsNRT2.3 were dramatically upregulated in the single lbd37,lbd38,lbd39,and triple lbd37/38/39 mutants under HN conditions(Fig.5D-F).However,when plants were grown under LN conditions,the expression of OsNRT2.1,OsNRT2.2,and OsNRT2.3 did not differ significantly between the mutant and wild type lines(Fig.5D-F).The transcription of OsNR2 was significantly upregulated in the roots of single lbd37,lbd38,lbd39,and triple lbd37/38/39 mutants under HN conditions,and OsNiR1 was also upregulated in most mutant lines(Fig.S5A,B).However,OsNRT1.1 and OsNRT1.2 showed no marked differences in the single lbd37,lbd38,lbd39,and triple lbd37/38/39 mutants under HN conditions(Fig.S5C,D).These results further confirmed that OsLBD37/OsLBD38/OsLBD39 regulate nitrate uptake and assimilation under high-N conditions.

3.6.OsLBD37/OsLBD38/OsLBD39 subcellular localization and transcriptional activity

To determine the subcellular localization of OsLBD37,OsLBD38,and OsLBD39,the OsLBD37-GFP,OsLBD38-GFP,and OsLBD39-GFP fusion proteins were separately constructed under the control of the cCaMV 35S promoter(Pro35S)and were then co-transformed into rice protoplast cells along with a nuclear localization marker,OsbZIP46-RFP.The GFP fluorescence signal revealed that OsLBD37-GFP and OsLBD39-GFP fusion proteins were localized predominantly in the nucleus,whereas the OsLBD38-GFP fusion protein was localized in both the cytoplasm and nucleus(Fig.6A).Thus,OsLBD37/OsLBD38/OsLBD39 reside primarily in the nucleus.

A dual-luciferase reporter assay was used to assess the transcriptional activity of OsLBD37/OsLBD38/OsLBD39.Effector and reporter plasmids were co-transformed into rice protoplast cells using REN luciferase as an internal reference(Fig.6B).LUC luciferase activity was significantly repressed in the experimental target(GAL4 BD-OsLBD37,GAL4 BD-OsLBD38,GAL4 BD-OsLBD39)compared to the control(GAL4 BD)(Fig.6C).These results revealed that OsLBD37/OsLBD38/OsLBD39 acts as a transcriptional suppressor to directly downregulate gene expression.

Fig.3.OsLBD37,OsLBD38,and OsLBD39 knockout increased N accumulation in hydroponic culture.(A-C)Plant height,root length,and tiller number per plant of ZH11,single lbd37,lbd38,lbd39 and triple lbd37/38/39 mutants under both LN and HN conditions.(D,E)Dry weight of shoots and roots.(F,G)Total N concentrations in shoots and roots.Con,concentration;DW,dry weight.Values are means±SD.n=6 biologically independent samples.Asterisks indicate significant differences between the ZH11 control and single lbd37,lbd38,lbd39 and triple lbd37/38/39 mutants by two-tailed Student’s t-test:*,P＜0.05;**,P＜0.01.

Fig.4. concentrations in shoots,roots,and xylem sap.Con,concentration;DW,dry weight.Values are means±SD.n=6 biologically independent samples.Asterisks indicate significant differences between the ZH11 control and single lbd37,lbd38,lbd39 and triple lbd37/38/39 mutants by two-tailed Student’s t-test:*,P＜0.05;**,P＜0.01.

Fig.5.OsNRT2.1/OsNRT2.2/OsNRT2.3 are regulated by OsLBD37/OsLBD38/OsLBD39.(A-C)Expression analysis of the high-affinity nitrate transport genes OsNRT2.1,OsNRT2.2,and OsNRT2.3 in roots of the ZH11 and LBD37/38/39-OE lines by qRT-PCR.Relative expression in ZH11 is defined as 1 under the LN condition.(D-F)Expression of the OsNRT2.1,OsNRT2.2,and OsNRT2.3 genes in roots of ZH11 and mutants.Relative expression in ZH11 is defined as 1 under HN condition.Values are means±SD.n=3 biologically independent samples.Statistical comparison was performed by Tukey’s test:*,P＜0.05;**,P＜0.01.

To investigate whether OsLBD37/OsLBD38/OsLBD39 was involved in the regulation of OsNRT2.1/2.2/2.3,we performed a transient assay using rice protoplast cells.The promoter of each OsNRT2 gene was fused with the LUC reporter gene(Fig.6D).The expression levels of OsNRT2.1/2.2/2.3 were found to be significantly reduced in the presence of OsLBD37,OsLBD38,or OsLBD39(Fig.6E).These results further demonstrated that the transcription of OsNRT2.1/2.2/2.3 was inhibited by OsLBD37/OsLBD38/OsLBD39 in rice cells.

Fig.6.Subcellular localization and transcription repression analysis of OsLBD37/OsLBD38/OsLBD39.(A)Subcellular localization of OsLBD37/OsLBD38/OsLBD39.OsLBD37/OsLBD38/OsLBD39 was fused with GFP at the N terminal.The fused protein OsbZIP46-RFP was used as the nucleus marker.(B)Schematic representation of recombinant reporter,reference,and effector plasmids for dual-luciferase reporter analysis.(C)Transcriptional repression assay of OsLBD37/OsLBD38/OsLBD39 in rice protoplast cells.The effector and reporter vectors were co-transformed into rice protoplast cells.After 16 h,the fluorescence of firefly luciferase and Renilla luciferase(internal control)were detected.(D)Schematic representation of the effectors and reporters.The effector plasmids express full length OsLBD37,OsLBD38,and OsLBD39 under control of the 35S promoter with a viral translation enhancer.The reported plasmids consisted of the 2000-bp promoter regions of OsNRT2.1/OsNRT2.2/OsNRT2.3 fused with the LUC reporter gene.(E)Transient assay showed that the OsLBD37/OsLBD38/OsLBD39 repressed the OsNRT2.1/OsNRT2.2/OsNRT2.3 promoters.Values are means±SD.n=3 biologically independent samples.Asterisks indicate significant differences between the GAL4 DB control and GAL4 DB-OsLBD37/OsLBD38/OsLBD39 as evaluated by two-tailed Student’s t-test:**,P＜0.01.

3.7.OsLBD37 interaction with OsLBD37,OsLBD38,and OsLBD39

Given that OsLBD37,OsLBD38,and OsLBD39 showed high sequence similarity and response to N,the interactions among these three proteins were investigated.Considering that OsLBD37 showed a stronger response to N than the other two genes,we focused on their interaction with OsLBD37(Fig.S1).In the yeast two-hybrid experiment,we first tested the self-activation activity of OsLBD37 by fusing its full-length protein with the GAL4 DNAbinding domain in the prey vector,followed by cotransformation with an empty bait vector.The results showed that the full OsLBD37 protein harbored slight self-activation activity as revealed by yeast growth under different concentrations of 3-AT,a competitive inhibitor of the reporter gene product(Fig.7A).As OsLBD37 protein contains a conserved LOB domain at the N terminus and a disordered region at the C terminus,we roughly divided OsLBD37 into three fragments:A section(1-110 aa that contains the full LOB domain),B section(111-152 aa),and C section(153-204 aa that contains the disordered region),and then tested their self-activation activity(Fig.7A).Two other OsLBD37 variants with a deleted A or C section were also tested.The results showed that only the C section(as well as other fragments harboring the C section)rather than the A and B sections contributed to the slight self-activation activity(Fig.7A).We then tested the interaction of full length and various fragments of OsLBD37 with the fulllength OsLBD37,OsLBD38,and OsLBD39 based on this yeast twohybrid system.When the bait vector expressed OsLBD37,OsLBD38,or OsLBD39,only prey vectors expressing OsLBD37 fragments containing the A section strongly promoted yeast growth under high concentrations of 3-AT,whereas yeast cells carrying other bait vectors showed similar growth to the self-activation test(Fig.7B).These results suggest that OsLBD37 interacts with OsLBD37,OsLBD38,and OsLBD39 through its LOB domain(A section).BiFC in rice protoplast cells further confirmed that OsLBD37 can interact with OsLBD37,OsLBD38,and OsLBD39 in the nucleus(Fig.7C).These results indicate that OsLBD37 can directly interact with OsLBD37,OsLBD38,and OsLBD39 in the nucleus and can form homodimers and/or heterodimers.

4.Discussion

Fig.7.Functional interaction between OsLBD37 and OsLBD37/OsLBD38/OsLBD39.(A)Schematic representation of truncation fragments for self-activation activity analysis.(B)Yeast two-hybrid assay of OsLBD37 interaction with OsLBD37/OsLBD38/OsLBD39.MaV203 cells co-transformed with pDEST22(AD)and pDEST32(BD)plasmids were grown in SD-Trp-Leu-His medium containing either 0(control)or 15,25,35,or 40 mmol L-1 3-AT.(C)BiFC assay of OsLBD37 interaction with OsLBD37/OsLBD38/OsLBD39.OsLBD37,OsLBD38,and OsLBD39 were fused with the N terminal(nYFP)and the C terminal(cYFP),respectively.The empty vectors of nYFP and cYFP were used as control.Scale bars,10μmol L-1.Red triangles indicate positive interactions.

Rice plants in paddy fields preferentially use ammonium rather than nitrate as a major N source,becauseis the predominant species of mineral N in bulk soil of paddy rice fields and because the assimilation ofrequires lower energy than that of[1,2].In this study,OsLBD37,OsLBD38,and OsLBD39 in rice were induced mainly byor Gln and not by(Fig.1),indicating that these three LBD genes suppress OsNRT2 genes mainly under ammonium-sufficient conditions instead of nitrate-sufficient conditions.This strategy could help rice plants preferentially useby limiting the uptake of nitrate under ammoniumsufficient conditions.In Arabidopsis,AtLBD37,AtLBD38,and AtLBD39 were also suggested[41]to function as transcriptional repressors involved inuptake and assimilation;however,AtLBD37,AtLBD38,and AtLBD39 were more strongly induced bythan byand Gln.Such divergence in these LBD genes between rice and Arabidopsis in response to N sources may be attributed to the different preferences of rice and Arabidopsis for N sources,as Arabidopsis generally grows in aerated soil where nitrate is the major inorganic N source.These results suggest that LBD37/LDB38/LBD39 plays a conserved role in suppressinguptake and assimilation under N-sufficient conditions across monocots and dicots,as shown by Arabidopsis and rice(Fig.5)[41].These LBD genes in different species have undergone adaptive evolution by modifying their expression responses to environmental N availability to fit their actual requirements.

LBD proteins have a conserved LOB domain at the N terminus.Several motifs that can be bound by LBD proteins have been identified in Arabidopsis.For example,selection and amplification binding assays have shown[42]that the LBD family of proteins specifically recognizes the hexamer GCGGCG as a core sequence,termed the LBD motif.LBD29 preferentially binds to the G-box(CACGTG),TGGGC[C/T],TGTCTC,and GAGACA motifs,as revealed by ChIP-seq analysis[43].Pandey et al.[44]showed that the nucleotide sequence CCGGXTTTXXXG,where X is any nucleotide,is critical for LBD18 binding.LBD transcription factors may bind in diverse manners to the promoters of their target genes.However,within the promoter sequences of rice OsNRT2 genes,we did not find any potential target motifs of LBD genes.We tested for direct interaction between the promoter sequences of OsNRT2 genes and the proteins of OsLBD37,OsLBD38,and OsLBD39 using yeast one-hybrid and electrophoretic mobility shift assays,but neither experiment showed a positive signal of direct interaction between them(data not shown).The specific relationship between OsNRT2 genes and LBD37/LDB38/LBD39 in Arabidopsis is still unclear[41],although the latter have been clearly confirmed to be involved in regulating the transcription of the former.There may thus be some additional unknown transcription factors involved in the regulation of LBD37/LDB38/LBD39 and NRT2 genes.

Several LBD proteins have been reported to form homodimers and/or heterodimers with other LBD family members,and these contribute to transcriptional regulation for initiating a sequence of regulatory events leading to a specific cellular and biological response[45,46].Two major classes of LBD genes are characterized by the presence(class I)or absence(class II)of functional leucinezipper-like domains[47].Many of the class I LBD proteins are predicted to form a coiled-coil motif that may function in protein-protein interactions,whereas class II LBD proteins have an incomplete leucine zipper,which cannot form a coiled-coil structure[47,48].However,some class II LBD proteins that were predicted not to homodimerize can interact with other LBD family members[45,49].In our study,the yeast two-hybrid and BiFC assays showed that OsLBD37 could interact with OsLBD37,OsLBD38,and OsLBD39 in the nucleus(Fig.7),although all three LBD proteins belong to class II.These findings suggest that OsLBD37 forms homodimers and/or heterodimers with OsLBD38 and OsLBD39 to exercise its biological functions duringuptake and assimilation.In agreement with this scenario,OsLBD37,OsLBD38,and OsLBD39 presented similar expression patterns in response to N and showed consistent phenotypes when knocked out in rice plants.

CRediT authorship contribution statement

Xinxin Zhu:Writing-original draft,Data curation,Investigation,Formal analysis.Dujun Wang:Data curation,Investigation,Resources.Lijuan Xie:Data curation.Tao Zhou:Data curation.Jingyi Zhao:Data curation.Qian Zhang:Data curation.Meng Yang:Data curation,Writing-review & editing.Wenjuan Wu:Data curation.Xingming Lian:Project administration,Funding acquisition,Writing-review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This work was supported by the National Natural Science Foundation of China(32171943 and 31821005).

Appendix A.Supplementary data

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2022.03.011.