Clock component OsPRR59 delays heading date by repressing transcription of Ehd3 in rice

Yupeng Wng,Fuqing Wu,Shirong Zhou,Weiwei Chen,Chenyn Li,Erho Dun,Jihng Wng,Zhijun Cheng,Xin Zhng,Qiing Lin,Yulong Ren,Cilin Lei,Xiuping Guo,Ziming Wu,Shnshn Zhu,*,Jinmin Wn,,*

a National Key Facility for Crop Gene Resources and Genetic Improvement,Institute of Crop Sciences,Chinese Academy of Agricultural Sciences,Beijing 100081,China

b National Key Laboratory for Crop Genetics and Germplasm Enhancement,Jiangsu Plant Gene Engineering Research Center,Nanjing Agricultural University,Nanjing 210095,Jiangsu,China

c Key Laboratory of Crop Physiology,Ecology and Genetic Breeding,Ministry of Education,Jiangxi Agricultural University,Nanchang 330045,Jiangxi,China

Keywords:Circadian clock Ehd3 Heading date Oryza sativa Pseudo-response regulator

ABSTRACT Heading date(or flowering time),an important agronomic trait in crop species,is closely associated with regional adaptation and yield.Members of the Pseudo-Response Regulator(PRR)family play key roles in regulating flowering.However,their role and molecular mechanism controlling heading date in rice is not very clear.Here,we identified rice OsPRR protein,OsPRR59,which delayed heading under longday conditions.OsPRR59 positively regulates yield by affecting plant height,secondary branches number per panicle,grain number per panicle,seed setting rate,and grain weight per plant.OsPRR59 is expressed in most tissues and its protein is localized to the nucleus.We also found that OsPRR59 directly binds to the promoter of Ehd3 to inhibit its expression.Compared with the WT,osprr59 ehd3 showed a significantly delayed heading phenotype,as did the ehd3 mutant.This was opposite to the phenotype of the osprr59 mutant,confirming that Ehd3 acted downstream of OsPRR59 in regulating rice flowering.Our results identified a direct regulator of Ehd3,and revealed a novel molecular mechanism of clock component OsPRR proteins in regulating heading date and provide a new genetic resource for fine-tuning heading date in rice.

1.Introduction

Heading date(or flowering time)is an important agronomic trait for optimization of crop yield and is related to seasonal and regional adaptability[1].It is controlled by internal and external environmental factors,such as photoperiod,temperature,and hormonal effects[2].Photoperiod plays a leading role in regulating flowering in crop species[3].As a typical short-day(SD)plant,heading date of rice is accelerated under SD conditions,and delayed under long-day(LD)conditions.Hd3a(Heading date 3a)and RFT1(RICE FLOWERING LOCUS T 1),orthologs of Arabidopsis FT(FLOWERING LOCUS T),encode florigens that are expressed in leaf vasculatures and move to the shoot apical meristem to perform flowering function[4-6].The photoperiod flowering pathway in rice can be divided into two components:the Hd1(heading date 1)-Hd3a pathway similar to the CO(CONSTANS)-FT pathway in Arabidopsis thaliana;and the rice specific Ghd7(Grain number,plant height,and heading date 7)-Ehd1(Early heading date 1)-RFT1/Hd3a pathway[3].

Unlike CO in Arabidopsis which merely promotes flowering,Hd1 promotes flowering in rice by activating Hd3a under SD conditions,but delays flowering by inhibiting Hd3a under LD conditions.This functional conversion of Hd1 is regulated by phytochrome light signaling[7-9].Ehd1,encoding a rice-specific B-type response regulator,is another flowering signal integration factor in the rice flowering pathway[10].Ehd1 positively regulates the expression of RFT1 and Hd3a to promote rice heading under both LD and SD conditions,and its expression level is controlled by many regulatory factors,including OsRE1(regulator of Ehd1),SIP1(C2H2 zinc finger protein SDG723/OsTrx1/OsSET33 Interaction Protein 1),OsCOL10(OsCO-Like 10),Ehd4(Early heading date 4),Ghd7,DTH8(QTL for days to heading on chromosome 8),LFL1(OsLEC2 and FUSCA 3-like 1),and MADS50[11-18].Ghd7 encodes a CCTdomain protein and negatively regulates photoperiodiccontrolled expression of Ehd1 by interacting with Hd1[15,19].Recently,it was found that Ghd7,Ghd7.1 and Hd1 coordinate with the NUCLEAR FACTOR Y(NF-Y)B/C dimer through their CCT domains to specifically target the promoter of Hd3a[20].In addition,Ghd7 is regulated by many flowering regulators,including Ehd3(Early heading date 3),OsTrx1(OsTrithorax1),Hd16(Heading date 16),OsGI(OsGIGANTEA),PhyA(OsPhytochrome A),PhyB(OsPhytochrome B),and OsELF3-1(EARLY FLOWERING 3-1)[21-27].Among them,OsPhyA,OsPhyB and OsGI directly interact with Ghd7,and OsPhyA and OsPhyB inhibit interaction between OsGI and Ghd7,thereby stabilizing the Ghd7 protein[24].

Members of the Pseudo-Response Regulator(PRR)family,core circadian clock components in Arabidopsis,are widely involved in the regulation of flowering[28,29].Five PRR genes in Arabidopsis,PRR1/TOC1,PRR3,PRR5,PRR7,and PRR9,are expressed from dawn to dusk with different peak times in the order:PRR9,PRR7,PRR5,PRR3,and PRR1[28].PRR family members share two domains:PR(for Pseudo Receiver)and CCT(for CONSTANS,CONSTANS-like,TOC1);and two relatively conserved motifs:L(E/D)(L/I)S(L/I)(R/K)R and SXXSAF(S/T)(R/Q)Y(X is any amino acid)[29].PRR9,PRR7 and PRR5 promote flowering by repressing the mRNA levels transcriptional repressors of CO known as CYCLING DOF FACTORs(CDFs)[30,31].Concomitantly,they also bind the promoters of CCA1(CIRCADIAN CLOCK ASSOCIATED 1)and LHY(LATE ELONGATED HYPOCOTYL)and inhibit their expression,thus activating the expression of FLAVIN-BINDING,KELCH REPEAT F-BOX 1(FKF1)and GI,and promoting the degradation of CDFs[29,32,33].In addition,PRRs interact with CO to stabilize CO protein[34].Members of PRR family in crop species also play pivotal roles in controlling photoperiodic-induced flowering.Such as Ppd-H1 in barley,Ppd-D1 in wheat,and SbPRR37 in sorghum,and GmPRR37 in soybean have been well studied about the functions in photoperiod sensitivity[35-38].Recently,it was found that two homeologous PRRs Tof11(Time of Flowering 11)and Tof12 act via LHY homologs to promote expression of the legume-specific E1 gene and delay flowering under long day conditions[39].

There are also five homologous OsPRR family genes in rice,namely OsPRR1,OsPRR37,OsPRR59,OsPRR73,and OsPRR95[40,41].OsPRR37/DTH7 is a major effect gene that determines photoperiod sensitivity and regional adaptability of rice[42,43].It delays the heading date by inhibiting the expression of Ehd1 under LD conditions[42].Compared with WT,the heading date of OsPRR73-RNAi lines in the background of japonica rice variety Zhonghua 11 was delayed by about 12 days under LD conditions[44].It is quite clear that OsPRRs are important for regulation of heading date in rice,but the roles of individual OsPRRs and their regulation mechanisms remain to be elucidated.

In this study,we found that OsPRR59 delayed heading under LD conditions.OsPRR59 is expressed in most tissues and its protein is localized to the nucleus.Molecular and biochemical experiments revealed that OsPRR59 directly binds to the promoter of Ehd3 to inhibit its expression.Genetic evidence showed that OsPRR59 acts upstream of Ehd3 in the regulation of heading date.Our results identified a direct regulator of Ehd3,revealed the molecular mechanisms of OsPRRs regulating heading date,and provided a new gene resource for fine-tuning the heading date of cultivars.

2.Materials and methods

2.1.Plant materials and growth conditions

The background parent of all plant materials used in this study was Oryza sativa ssp.japonica cv.Nipponbare.The phenotypes of the materials were investigated after T2generation.Plants were grown in paddy fields under natural short-day(NSD)conditions(Sanya in Hainan,China,18°30′N)or natural long-day(NLD)conditions(Beijing,China,40°13′N)or in growth chambers under controlled long-day(CLD)conditions(14 h light at 30 °C/10 h darkness at 25 °C)or controlled short-day conditions(CSD)(10 h light at 30 °C/14 h darkness at 25 °C).To investigate floweringrelated genes regulated by OsPRR59,the WT and osprr59 plants were grown in growth chambers under CLD and CSD conditions.

2.2.Vector construction and rice transformation

An 18 bp gene-specific spacer sequence of OsPRR59 or Ehd3 was inserted into the sgRNA/Cas9 construct to obtain osprr59,ehd3 and osprr59 ehd3 mutants.The resulting plasmids were introduced into Agrobacterium tumefaciens strain EHA105 and then transformed into calli as described previously[45].Primers used in this assay are listed in Table S1.

2.3.Subcellular localization of OsPRR59

The coding sequence of OsPRR59 was amplified and cloned into the C-terminus of GFP in the pCAMBIA1390GFP vector,generating recombinant pUbi::OsPRR59-GFP.The recombinant vector was transformed into rice protoplasts according to protocols described previously[46].Fluorescence signals were visualized using a Zeiss LSM980 laser scanning confocal microscope(Zeiss,Oberkochen,Germany).Primers used in this assay are listed in Table S1.

2.4.Luciferase transient transcriptional repression assay

Transcriptional repression activity analysis was performed as described previously[47].Briefly,the full coding sequence of BDOsPRR59 was amplified and cloned into effector plasmid GAL4 by gateway reaction,generating a recombinant GAL4-BD-OsPRR59 construct.Effector plasmid vectors GAL4-BD-OsPRR59 and GAL4-BD were co-transformed into rice protoplasts with the two reporters,and luciferase(LUC)and Renilla reniformis(REN)activities were separately determined 14 h post-transformation using Dual-Glo Luciferase Assay System(E2920,Promega,Madison,WI,USA).Primers used in this assay are listed in Table S1.

2.5.RNA extraction and quantitative real-time PCR(qRT-PCR)analyses

To detect the expression of flowering-related genes regulated by OsPRR59,total RNA was extracted from the leaves of 45-dayold plants under CLD conditions or 21-day-old plants under CSD conditions using a ZR Plant RNA MiniPrep Kit(ZYMO Research,Orange County,CA,USA)and then reverse transcribed using Prime-Script II reverse transcriptase(TaKaRa,Dalian,Liaoning,China)with oligo(dT)18 primer.qRT-PCR analyses were performed in an ABI 7500 real-time PCR system or QuantStudio 7 Flex(Applied Biosystems,Waltham,MA,USA)with the SYBR Premix Ex Taq Kit(TaKaRa).The rice Ubiquitin(UBQ)gene was used as an internal control.Primers used in this assay are listed in Table S1.

2.6.Quantitative transactivation assay

A 2500 bp fragment from the promoter of Ehd3 for quantitative transactivation assays was cloned and fused into the vector pGreenII0800-LUC as the reporter(proEhd3::LUC).The OsPRR59-Flag was used as the effector and the empty vector served as the negative control.The plasmids were co-transformed into rice protoplasts and the Dual-Glo Luciferase Assay System(E2920,Promega)was used to measure LUC and REN activities.Primers used in this assay are listed in Table S1.

2.7.Chromatin immunoprecipitation(ChIP)-qPCR

ChIP assay in rice protoplasts was performed as previously described[48].About 40μg of plasmid pUbi::OsPRR59-Flag or pUbi::Flag was transformed into rice protoplasts.After 16 h of culture,the protoplasts were cross-linked in 1% formaldehyde.Chromatin was then extracted from the samples,and the DNA was sheared into fragments of 200-500 bp.Flag antibody(Medical Biological Laboratories,PM020)protein A magnetic beads(Merck Millipore,16-661)were used for immunoprecipitation.The ChIPed DNA fragments were used for qPCR.Primers used in this assay are listed in Table S1.

2.8.Electrophoretic mobility shift assay(EMSA)

To perform the EMSA,the coding sequence of CCT domain of OsPRR59 was introduced into PEGX4T-1.GST-OsPRR59-CCT constructs and empty GST vectors were introduced into the E.coli strain DE3 to induce protein expression.The induced proteins were purified with glutathione Sepharose 4B(GE Healthcare)and then eluted with 20 mmol L-1glutathione.EMSA was performed using the LightShift Chemiluminescent EMSA Kit(Thermo Fisher Scientific,20148).Primers used in this assay are listed in Table S1.

2.9.Statistical analysis

Microsoft Excel 2016(Microsoft,Redmond,WA,USA)was used for data standardization and visualization.IBM SPSS Statistics 19(IBM Corporation,Armonk,NY,USA)was employed for data statistics and analysis.

3.Results

3.1.OsPRR59 negatively regulates flowering in rice under long-day conditions

By screening a rice mutant library constructed by the CRISPR/Cas9 system,we isolated a mutant of OsPRR59(LOC_Os11g05930)that significantly promoted heading date under LD conditions.To further confirm the function of OsPRR59 in regulation of heading date,the wild type(WT)and osprr59 mutant were grown in four environments:NSD conditions in Sanya(day length＜12 h),NLD conditions in Beijing(day length＞15 h),CSD conditions(10 h light)and CLD conditions(14 h light).Compared with WT,the osprr59 mutant was about 4 days earlier in heading under NLD and about 7 days earlier in heading under CLD conditions,but there was no difference under NSD and CSD conditions(Fig.1).To eliminate the possible effect of the CRISPR construct on phenotypic observation,we obtained Cas9-free mutant of OsPRR59 using the TKC(Transgene Killer CRISPR)construct[49]and named TKCosprr59.Consistent with the earlier results,TKC-osprr59 exhibited about a 3.5-day earlier in heading date under NLD conditions,but not in NSD conditions(Fig.2).These results suggested that OsPRR59 is a repressor of heading date under LD conditions but not under SD conditions.

Fig.1.The flowering phenotype of osprr59 mutant.(A)Sketch map of mutation sites in OsPRR59 knockout lines.Each dash(-)sign indicates a missing nucleotide.PAM,protospacer adjacent motif.(B)Plants morphology of WT and osprr59 mutant at heading.Scale bar,10 cm.(C,D)Comparison of days to flowering between WT and osprr59 mutant under NLD and NSD conditions(C),and under CLD and CSD conditions(D).n≥20,statistically significant difference is marked with asterisks(**,P＜0.01;Student’s t-test).

Fig.2.Flowering phenotype of TKC-osprr59 mutant.(A)Sketch map of mutation sites in OsPRR59 knockout lines.Each dash(-)sign indicates a missing nucleotide.PAM,protospacer adjacent motif.(B)Plants morphology of WT and TKC-osprr59 mutants at heading.Scale bar,10 cm.(C,D)Comparison of days to flowering between WT and TKCosprr59 mutants under NLD(C)and NSD(D)conditions.n≥20,statistically significant difference is marked with asterisks(**,P＜0.01;Student’s t-test).

We next compared yield-related agronomic traits of WT and TKC-osprr59 plants grown in Beijing.There were no significant differences in tiller number,panicle length and primary branches number per panicle,but plant height,secondary branches number per panicle,grain number per panicle,seed setting rate,and grain weight per plant were lower in the TKC-osprr59 mutant(Fig.3).These data indicated that OsPRR59 is a positive regulator of yield under NLD conditions.

3.2.OsPRR59 was localized to the nucleus as a transcriptional repressor

Sequence analysis showed that OsPRR59 encodes a typical PRR protein with a pseudo receiver(PR)domain in the N-terminal and a CCT domain in the C-terminal(Fig.S1).BLAST searches of databases revealed the widespread presence of homologs of OsPRR59 in land plants;the protein sequences are highly conserved in the PR domain,CCT domain and SXXSAF(S/T)(R/Q)Y motif(Fig.S2).Phylogenetic analysis indicated that the OsPRR59 homologs were differentiated between monocotyledons and dicotyledons,indicating that their function may have changed during evolution(Fig.S3).To examine the temporal and spatial expression pattern,we used qRT-PCR and found that OsPRR59 was expressed in various tissues,with the highest level in leaf(Fig.4A).OsPRR59 showed an obvious diurnal rhythmic expression pattern under both LD and SD conditions,peaking during the morning and evening(Fig.4B,C).We also compared the expression of circadian clock component genes OsCCA1,OsPRR1,OsPRR73,and OsPhyC in WT and osprr59 mutant and found that expression of the four genes was decreased to different degrees in the osprr59 mutant(Fig.S4).These results indicated that OsPRR59 was indeed involved in regulation of the circadian clock system in rice.

To investigate the subcellular localization of OsPRR59 protein,we fused the full-length CDS of OsPRR59 with GFP and a transient expression assay in rice protoplasts showed that the OsPRR59-GFP fusion protein was localized to the nucleus,overlapping with the nuclear marker D53-mCherry(Fig.4D)[50].A luciferase transient transcriptional assay revealed that OsPRR59 had significant transcriptional repression activity compared with the control BD in rice protoplasts(Fig.4E,F),indicating that OsPRR59 acts as a transcriptional repressor in the nucleus.

3.3.OsPRR59 may regulate flowering through the Ghd7-Ehd1-Hd3a/RFT1 pathway

Fig.3.Agronomic traits of WT,TKC-osprr59-1 and TKC-osprr59-2 grown under NLD conditions.Comparisons of plant height(A),tiller number(B),panicle length(C),primary branches number per panicle(D),secondary branches number per panicle(E),grain number per panicle(F),seed setting rate(G),and grain weight per plant(H)between WT and TKC-osprr59 mutant plants.n=20,statistically significant difference is marked with asterisks(**,P＜0.01;Student’s t-test).

Fig.4.Expression patterns of OsPRR59 and characteristic of OsPRR59 protein.(A)qRT-PCR analysis of transcriptional levels of OsPRR59 in various tissues.(B,C)Diurnal expression pattern of OsPRR59 under SD(B)and LD(C)conditions.(D)Subcellular location of OsPRR59-GFP fusion protein in rice protoplasts.D53-mCherry was used as nuclear marker.Scale bar,5μm.(E,F)Transcriptional repression activity analysis of OsPRR59 protein in rice protoplasts.Schematic of effector and reporter vectors is shown in(E);relative LUC/REN activity is shown in(F).Data are means±SE(n=3).The experiment was performed with three biological replicates,and similar results were obtained.Statistically significant difference is marked with asterisks(**,P＜0.01;Student’s t-test).

To determine the regulatory relationship between OsPRR59 and various flowering-time regulators,the expression of major flowering genes was detected in the WT and osprr59 mutant under LD conditions.The mRNA levels of two florigen genes Hd3a and RFT1,and their downstream genes OsMADS14 and OsMADS15,were significantly upregulated in the osprr59 mutant compared to the WT(Fig.5).Hd1 and Ehd1 are major floral integrators in transduction of floral signals in rice.To investigate whether the regulation of Hd3a and RFT1 via OsPRR59 is mediated through Hd1 or Ehd1,or both,we compared their mRNA levels in osprr59 mutant and WT plants.The expression of Ehd1,but not Hd1,was significantly upregulated in the osprr59 mutant,suggesting that OsPRR59 regulates Hd3a and RFT1 through Ehd1 but not Hd1(Fig.5).We then investigated the expression of other known regulators of Ehd1.The expression levels of negative regulators Ghd7 and OsCOL10 were significantly downregulated,and positive regulator Ehd3 was up significantly upregulated.Negative regulators OsPRR37(DTH7)and OsLFL1,and positive regulators OsTrx1,OsMADS50,OsGI,and OsELF3 were not significantly affected in the osprr59 mutant(Fig.5).The expression of DTH2(for Days to heading on chromosome 2),which promotes heading by inducing the florigen genes Hd3a and RFT1 and acts independently of the known floral integrators Hd1 and Ehd1[51],was also not significantly different between WT and the osprr59 mutant(Fig.5),indicating that OsPRR59 does not regulate flowering through DTH2-mediated pathway.It was reported that Ehd3 acts upstream of Ghd7 to inhibit its expression in LD conditions[21].Therefore,our results suggested that OsPRR59 may regulate flowering through Ehd3-Ghd7-Ehd1-Hd3a/RFT1 pathway.

Fig.5.Relative expression of flowering-related genes in WT and osprr59 plants.Relative expression of OsMADS14,OsMADS15,Hd3a,RFT1,Ehd1,Hd1,Ghd7,Ehd3,OsTrx1,OsCOL10,OsPRR37(DTH7),OsLFL1,MADS50,OsGI,OsELF3,and DTH2 in WT and osprr59 plants under LD conditions.RNA isolated from leaves of 45-day-old plants grown under CLD conditions was used for RT-PCR.Closed bars at the bottom indicate the dark period;open bars,indicate the light period.The y-axes show transcript levels relative to rice Ubiquitin.

3.4.OsPRR59 physically associates with the Ehd3 promoter

The mRNA level of Ehd3 was significantly up-regulated in the osprr59 mutant and Ehd3 acted upstream of Ghd7-Ehd1-Hd3a/RFT1 pathway.Previous studies reported that the G-box-like motif was necessary for transcriptional regulation by PRRs and that there was direct binding of those proteins and G-box-like elements[39,52-54].We identified three G-box-like motifs in the promoter of Ehd3(Fig.6A)and tested whether OsPRR59 directly regulate transcription of Ehd3.DNA electrophoretic mobility shift assays(EMSA)with the purified GST-tagged CCT domain(the DNAbinding domain of PRRs[55])of OsPRR59(Fig.S5),showed that GST-OsPRR59-CCT,but not the free GST protein,was be able to bind all three DNA probes containing the G-box-like element(Fig.6B).Moreover,the competing probe without biotin labeling effectively competed for GST-OsPRR59-CCT binding with the G-box-like motif-containing probe(Fig.6B).ChIP-qPCR assays showed that OsPRR59 was significantly enriched at the promoter regions containing the G-box-like motifs(Fig.6C).Transient expression in rice protoplasts showed that OsPRR59 significantly inhibited activity of the Ehd3 promoter(Fig.6D,E).Those results verified that OsPRR59 directly binds to the Ehd3 promoter to inhibit its expression.

To further investigate the genetic relationship between OsPRR59 and Ehd3,ehd3 and osprr59 ehd3 double mutants was constructed by CRISPR/Cas9 technology(Fig.S6).Compared with the WT,both the ehd3 and osprr59 ehd3 mutants showed delayed flowering under NLD conditions.This effect was opposite to that of the osprr59 mutant(Fig.6F,G)thus supporting the notion that OsPRR59 acts upstream of Ehd3 in the photoperiodic-controlled flowering pathway.

4.Discussion

PRR family members are core components of the plant circadian clock[28,40].They play important roles not only in maintaining the stability of the circadian clock system,but also in regulating biological processes,such as flowering,ABA signaling and salt tolerance,by direct interaction with key genes at the transcriptional or post transcriptional level[29,34,54,56,57].PRRs generally function as transcriptional repressors,and both L(E/D)(L/I)S(L/I)(R/K)R and SXXSAF(S/T)(R/Q)(Y/F)motifs are critical for their repression activity[29].Previous study found that PRR5,PRR7 and PRR9 can interact with transcriptional corepressor TOPLESS complexes through the L(E/D)(L/I)S(L/I)(R/K)R motif to inhibit gene expression[58].Our results showed that the expression of circadian clock marker genes in the osprr59 mutant was significantly altered(Fig.S4),indicating that OsPRR59 is indeed a core component of the circadian clock in rice.Through sequence alignment,we found that OsPRR59 had strong transcriptional repression activity although missing the L(E/D)(L/I)S(L/I)(R/K)R motif(Fig.S2).This suggested that OsPRR59 might inhibit gene transcription in other pathways rather than through direct interaction with transcriptional corepressor TOPLESS complexes.Compared with WT,the heading date of the osprr59 mutant was only slightly earlier(Figs.1,2),but it displayed highly significant decreases in yield-related agronomic traits(Fig.3E-H).This indicates that OsPRR59 may regulate those traits through alternate pathways that do not affect heading date.

Fig.6.OsPRR59 binds to the Ehd3 promoter.(A)G-box element analysis of the Ehd3 promoter.P1,P2,P3,P4,and P5 are the sites for the primers used in ChIP-qPCR assays.(B)EMSA assays show that OsPRR59 binds to the region of the G-box in the Ehd3 promoter.Plus(+)and minus(-)denote presence or absence of the probe or protein in each sample.(C)ChIP-qPCR assays show that OsPRR59 binds to theEhd3 promoter in vivo.The ChIP assay in rice protoplasts was performed three times with similar results.(D,E)Effect of OsPRR59 on promoter activity of Ehd3 in rice protoplasts.Schematic of effector and reporter vectors is shown in(D);relative LUC/REN activity is shown in(E).Data are means±SE(n=3).The experiment was performed with three biological replicates,and similar results were obtained.(F)Plants morphology of WT,and osprr59,ehd3 and osprr59 ehd3 mutants at heading.Scale bar,10 cm.(G)Days to flowering in WT and osprr59,ehd3 and osprr59 ehd3 mutants under NLD conditions.n≥20,statistically significant difference is marked with asterisks(**,P＜0.01;Student’s t-test).

The circadian clock can control flowering time in plants by regulating expression of downstream flowering genes in response to photoperiod[3].LHY and PRRs,the core components of the circadian clock in Arabidopsis,regulate flowering time by controlling expression of CDFs,the CO repressors[29,30,59];It has been found recently that PRRs control expression of LHY and CCA1 to regulate flowering in soybean.LHY/CCA1 regulates flowering time by directly inhibiting the expression of soybean flowering specific factor E1[39].It is not clear how the circadian clock in other crops relates to the downstream flowering pathway.In this study,we found that clock component OsPRR59 can directly inhibit expression of Ehd3,thus regulating the rice-specific flowering pathway Ghd7-Ehd1-Hd3a/RFT1,and thereby delaying flowering under LD conditions(Fig.6).This mechanism is different from that shown in previous studies in Arabidopsis and soybean,indicating that the rice circadian clock evolved a different flowering regulation mode.

Ehd3,encoding a plant homeodomain finger-containing protein,is a critical promoter of flowering in rice[21];The ehd3 mutant flowers much later under LD conditions than under SD conditions[21],indicating that the action of Ehd3 depends on light signals.OsPRR59 negatively regulates flowering in rice only under LD conditions(Figs.1,2),suggesting that its function during photoperiodinduced flowering is precisely regulated by light signals.Our results that OsPRR59 directly represses transcription of Ehd3 therefore at least partly explain the difference in heading date of the ehd3 mutant between LD and SD conditions.In addition,the transcription level of Ehd3 was almost steady under both SD and LD conditions during developmental process.This suggests that Ehd3 is essential for transcription of downstream genes during plant development,but not for control of their transcription patterns[21].We speculate that as a clock gene,OsPRR59 strictly controls the expression of Ehd3 to efficiently regulate expression levels of downstream components.

Heading date is an extremely complicated biological process and regulated by many genes.Some genes controlling heading date are affected by photoperiod,and their functions can be altered under different photoperiod conditions,which leads to the difference of regulation network in heading date between LD and SD conditions[3,60].For instance,Hd1 promotes flowering by activating Hd3a under SD conditions,but delays flowering by repressing Hd3a under LD conditions[7].DTH8 associates with the Hd3a promoter to modulate the level of H3K27 trimethylation(H3K27me3)at the Hd3a locus under LD conditions,while there is no DTH8-mediated epigenetic modification under SD conditions[61].Ghd7 delays flowering by repressing Ehd1 expression under LD,while this repression was very weak under SD because of the red-lightdependent peak of Ghd7 inducibility shifts to night,resulting in reduced Ghd7 expression in the morning[25].Here,we found that OsPRR59 regulates heading date and the expression of Ehd3 under LD but not SD conditions(Figs.1,5,and Fig.S7).The reasons why there is no difference in flowering time between WT and osprr59 under SD conditions may be as follows.Firstly,OsPRR59 may need some essential cofactors to regulate the expression of Ehd3,while the expression pattern of those cofactors is inconsistent with OsPRR59 or the expression level of them is very low under SD conditions.Secondly,the function of OsPRR59 may be regulated by an unknown mechanism,which switches the inhibition of OsPRR59 on Ehd3 under SD conditions.Thirdly,there may be functional redundant proteins with OsPRR59 in regulation of Ehd3 under SD conditions,and the function of these proteins is turned off under LD conditions.The definite reason why there is different molecular mechanism of OsPRR59 in regulating heading date under LD and SD needs more studies to clarify in the future.

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.

CRediT authorship contribution statement

Yupeng Wang:Conceptualization,Formal analysis,Funding acquisition,Investigation,Methodology,Writing-original draft,Writing-review&editing.Fuqing Wu:Conceptualization,Formal analysis,Funding acquisition,Investigation,Methodology,Writing-original draft,Writing-review & editing.Shirong Zhou:Data curation,Methodology.Weiwei Chen:Investigation,Software,Data curation.Chenyan Li:Investigation,Software,Data curation.Erchao Duan:Investigation,Software,Data curation.Jiachang Wang:Investigation,Software,Data curation.Zhijun Cheng:Methodology,Validation.Xin Zhang:Methodology,Validation.Qibing Lin:Methodology,Validation.Yulong Ren:Methodology,Validation.Cailin Lei:Methodology,Validation.Xiuping Guo:Methodology,Validation.Ziming Wu:Conceptualization,Investigation.Shanshan Zhu:Conceptualization,Formal analysis,Funding acquisition,Investigation,Methodology,Writing-original draft,Writing-review&editing.Jianmin Wan:Resources,Project administration,Supervision.

Acknowledgments

This work was supported by the National Natural Science Foundation of China(31771886 and 31771764),China Postdoctoral Science Foundation(2019T120164),and Central Public-interest Scientific Institution Basal Research Fund(Y2020YJ10).

Appendix A.Supplementary data

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2022.04.007.