Wenqing Chen, Wei Wng, Yusong Lyu, Ywen Wu,Pingling Hung,Shiki Hu, Xingjin Wei, Guii Jio, Zhonghu Sheng, Shoqing Tng,Goneng Sho,*, Ju Luo,*

aState Key Laboratory of Rice Biology,China National Rice Research Institute,Hangzhou 310006,Zhejiang,China

bInstitute of Crop Science and Nuclear Technology Utilization,Zhejiang Academy of Agricultural Sciences,Hangzhou 310021,Zhejiang,China

Keywords:Rice Pre-harvest sprouting Seed dormancy OsVP1 Sdr4 ABA

ABSTRACT Pre-harvest sprouting (PHS) is a disadvantageous trait in cereal production worldwide,causing large economic losses each year. Its regulation mechanism is still unclear. We generated the Oryza sativa Viviparous1(OsVP1)mutant using gene editing technique,which shows increased PHS compared with that of the wild type Nipponbare. OsVP1 is localized mainly in the nucleus and expressed in various tissues and organs. Expression of Seed dormancy 4(Sdr4),a key gene controlling PHS,was sharply reduced in OsVP1 mutants.OsVP1 bound to the specific motif CACCTG in the promoter of Sdr4 and activated its expression in rice protoplasts. Overexpression of Sdr4 reduced the high seed germination rate of OsVP1 mutant cr-osvp1-1,showing that Sdr4 acts as a downstream target of OsVP1.Both OsVP1 and Sdr4 loss-of-function mutants were insensitive to exogenous ABA and employed the ABA signaling pathway in regulating seed dormancy.These findings shed light on the control of seed dormancy aimed at preventing PHS in rice.

1.Introduction

Pre-harvest sprouting (PHS) is an undesirable trait in agricultural production owing to its effect on seed viability, grain quality, and yield [1,2]. Especially in southeast Asia, PHS frequently occurs in rice owing to the long rainy periods during the harvest season. In view of the high annual economic losses caused by PHS [3,4], PHS resistance is considered one of the most important breeding goals for the improvement of rice and other cereal crops.

Seed dormancy,an indicator of PHS resistance,is defined as the inability of viable seeds to germinate within a certain period even under suitable environmental conditions, and is influenced by both genetic and environmental factors[1,5,6]. Abscisic acid (ABA) is a plant hormone that plays vital roles in the regulation of seed dormancy and germination in diverse plant species[7-9].Increased ABA biosynthesis promotes seed dormancy, whereas reducing ABA content or blocking the ABA signaling pathway leads to a decrease in seed dormancy, resulting in a PHS phenotype[10]. Recently, key regulators of ABA biosynthesis or signaling pathway have been characterized: LEC2, ABI3,and DOG1 in Arabidopsis [11-14]; TaPHS1 and TaMFT in wheat [2,15]; VP1 in maize [16]; and OsFbx352, OsNCED3,OsPDS, OsCRTISO, β-OsLCY, OsZDS, OsVP1, Sdr4, OsDSG1,OsABI3, OsABI5, PHS8, PHS9, and OsMFT2 in rice [1,7,17-30].All of these genes control seed dormancy, but their regulatory mechanisms remain elusive.

Gibberellin (GA), another phytohormone, plays a role opposite to that of ABA in the regulation of seed dormancy and germination [31]. GA promotes germination, which requires not only initiating the activity of the embryo, but also breaking the physical barrier of the seed coat surrounding the embryo [32]. In plants, the biologically active GA level is controlled by a balance between biosynthesis and degradation [33]. GA 20-oxidase (GA20ox) and GA 3-oxidase (GA3ox), identified in plant species including rice,Arabidopsis, barley, and wheat, are two key enzymes responsible for catalyzing the reactions of GA biosynthesis,whereas GA 2-oxidase (GA2ox) affects the biosynthesis of bioactive GAs [33-35]. In addition to GAs and ABA, other hormones including auxin, ethylene, brassinosteroids,jasmonic acid,salicylic acid, cytokinins, and strigolactones are involved in the control of seed dormancy and germination [6,36,37]. Recent studies [6,37,38]have revealed that crosstalk among those plant hormones occurs and that their dynamic balance modulates the seed transition between dormancy and germination.

Sdr4 strongly influences the seed dormancy difference between Nipponbare (japonica) and Kasalath (indica) [25].Haplotype analysis [25]showed that there are two genotypes of sdr4-n and sdr4-k, with sdr4-k conferring deep dormancy and reducing pre-harvest sprouting of seeds,indicating that Sdr4 acts as a determinant of seed dormancy in rice cultivars. OsVP1 was identified as a coactivator of TRAB1,which binds to ABA-responsive elements(ABREs)in the Sdr4 promoter region to regulate ABA-regulated transcription [39]. However, the biochemical and genetic relationship between OsVP1 and Sdr4 is still unclear. In this study, we found that OsVP1 directly binds to the motif CACCTG in the Sdr4 promoter region and activates its expression. Overexpression of Sdr4 partially rescued the PHS phenotype of osvp1. Both cr-osvp1 and cr-sdr4 mutants were insensitive to exogenous ABA in regulating seed germination. Thus, our study revealed the link between OsVP1 and Sdr4 at the molecular and genetic levels, and provided an important insight for seed dormancy to prevent pre-harvest sprouting in rice.

2.Materials and methods

2.1. Plant materials and growth conditions

The wild type Nipponbare (NIP; Oryza sativa ssp. japonica)and all the transgenic plants used in this study were grown in the experimental field and greenhouse of China National Rice Research Institute. Fertilizer and water management were those used for standard field production.

2.2. Vector construction and plant transformation

To generate cr-osvp1 and cr-sdr4 mutants, small guide RNAs(sgRNA)targeting OsVP1(Os01g0911700)and Sdr4(Os07g0585700)were ligated into the expression vector VK005-1 (Beijing Viewsolid Biotech. Co. Ltd., Catalog no. VK005-01). Hygromycin was used as a selection marker and the constructed vectors were transferred into NIP calli.

To create overexpression constructs,the coding sequences of OsVP1 and Sdr4 were amplified by polymerase chain reaction (PCR) and cloned into pCAMBI23A (overexpression vector), after which the constructs pCAMBI23A-OsVP1 and pCAMBI23A-Sdr4 were transferred into cr-sdr4 and cr-osvp1-1,respectively, using G418 as a selection marker.

The primers used in this study are listed in Table S1.Transgenic plants were generated by Agrobacteriummediated transformation of rice calli as described previously [40].

2.3. Seed germination assays

Panicles from the wild-type NIP and transgenic plants at the same heading stage were harvested at maturity. For the panicle germination assay,at least three panicles were soaked in the water for 4 h and wrapped in wet towel. For seed germination assay, 100 seeds from a harvested panicle were sown on filter paper in Petri dishes treated with exogenous ABA solution (0, 5, 10, or 50 μmol L?1). All seed germination assays were conducted at 30 °C in an illuminated incubator under 12 h/12 h light/dark regime for one week,and repeated at least three times for each assay.

2.4. RNA extraction and real-time PCR analysis

Total RNA of all tissues except developing seeds was extracted with TRIzol (Thermo Fisher Scientific, Catalog no. 5596018).Developing seed RNA was extracted using a modified SDSTRIzol method [1]. First-strand cDNA was synthesized with 2 μg of total RNA using the ReverTra Ace qPCR RT kit(Catalog no. FSQ-101) following the manufacturer's protocol. Quantitative real-time PCR (qRT-PCR) was performed using a Light Cycler 480 instrument(Roche)with the SYBR Green Real-time PCR Master Mix(Toyobo)in a 20 μL reaction volume.The qRTPCR conditions were as follows:95 °C for 30 s,40 cycles of 95 °C for 5 s,60 °C for 35 s,and 95 °C for 15 s.Assays included three biological replications. The actin gene (Os03g0718150) was used as an internal control, and relative expression levels were calculated by the 2?ΔΔCTmethod[41].

2.5. Subcellular localization

The coding sequences of OsVP1 and Sdr4 without a stop codon were amplified by PCR and cloned into the vector pAN580 to generate the constructs 35S:OsVP1-GFP and 35S:Sdr4-GFP. The D53 cDNA was cloned into the vector 163-mCherry to create the construct 35S:D53-mcherry that was able to express the fusion protein D53-mCherry as a nuclear localization reference [42]. 35S:OsVP1-GFP, 35S:Sdr4-GFP and the empty control vector pAN580-GFP were co-transformed with 35S:D53-mcherry into rice protoplasts following Zhang et al. [43]. Fluorescent signals were detected with a Zeiss LSM710 confocal laser scanning microscope (Karl Zeiss,Jena,Germany).

2.6. Yeast one- and two-hybrid assay

Yeast one-hybrid assays were performed. The OsVP1 coding sequence and Sdr4 promoter sequence(1611 bp forward from the start codon ATG) were inserted into pB42AD and pLacZ2u to generate the respective expression constructs pB42ADOsVP1 and ProSdr4:LacZ2u. pB42AD and ProSdr4:LacZ2u and pB42AD-OsVP1 and ProSdr4:LacZ2u were co-transformed into the yeast strain EGY48. The transformants were grown on SD/-Ura/-Trp plates and then transferred to SD/-Ura/-Trp plates containing 80 mg L?1X-Gal, 1% raffinose, 1 × BU salts and 2%galactose(Clontech).

The coding sequence of OsVP1 without a stop codon was inserted into the vector pGBKT7 (Catalog no. 630443). Yeast transformation was performed following the manufacturer's instructions (http://www.clontech.com).

2.7. Purification of tag-fused proteins and electrophoresis mobility shift assay (EMSA)

The full-length OsVP1 coding sequence was inserted into the vector pGEX-4T-1 to generate pGEX-4T-1-OsVP1, which was then transformed into E.coli DE3 cells.The soluble GST-OsVP1 fusion protein was expressed and purified using the GSTSefinose Kit(Catalog no.C600327-0001).

Probe 1 for EMSA analysis in a 40-nt length containing a motif CACCTG was synthesized and then labeled with an EMSA Probe Biotin Labeling Kit (Catalog no. GS008). DNA binding was performed in a 10 μL reaction volume containing EMSA/Gel-shift binding buffer, 2 nmol biotin-labeled probe 1,and 5 nmol purified recombinant protein. Unlabeled DNA oligos were used as competitors. EMSA assays were performed using the chemiluminescent EMSA kit (Thermos,Catalog No.20148).

2.8. Luciferase transient transcriptional activity assay

The coding sequence of OsVP1 was inserted into the control vector 35S:NONE as an effector and the 1611 bp length promoter sequence of Sdr4 was cloned into vector 190LUC as a reporter [44]. Protoplast preparation and transformation were performed following the method reported previously[45]. Firefly LUC and REN activities were detected using the Dual-Luciferase Reporter Assay System (Cat No. E1960).Relative luciferase activity was calculated as the ratio of rLUS1 to rLUS2.Three biological replications were performed.

3. Results

3.1. OsVP1 loss-of-function mutants showed pre-harvest sprouting

Oryza sativa Viviparous1 (OsVP1), a homolog of Zea mays Vp1 and Arabidopsis thaliana ABI3,encodes a B3 domain-containing transcriptional activator factor [12,16]. To investigate the underlying molecular function of OsVP1 in regulating seed dormancy, OsVP1 knockout transgenic plants were generated using CRISPR/Cas9 technology. A single guide RNA (sgRNA)targeting the OsVP1 coding region was designed(Fig.1-A).As a result, a total of 21 T0transgenic lines were generated, of which 19 plants carried the hygromycin B resistance gene as determined by PCR (Fig. S1-A). Sequencing the target regions of OsVP1 in the T1transgenic lines revealed two mutants, crosvp1-1 and cr-osvp1-2,one with a threonine(T)and the other with a guanine (G) insertion (Figs. 1-A; S1-C, E). Sequencing five putative off-target sites predicted by http://skl.scau.edu.cn/offtarget/ revealed no off-target effects in either cr-osvp1-1 or cr-osvp1-2 plants(Table S2).

Fig.1-Phenotypic characterization of OsVP1 loss-of-function mutants.(A)Generation of cr-osvp1-1 and cr-osvp1-2 by CRISPR/Cas9 technology;(B)PHS phenotypes of physiologically mature seeds of WT,cr-osvp1-1 and cr-osvp1-2 after seven consecutive rainy days.Bar,2 cm;(C)Comparison of fresh seed germination rates of WT,cr-osvp1-1 and cr-osvp1-2 calculated from(B).Values are means±SD of at least three replications. WT,NIP.

No PHS phenotype was observed in cr-osvp1-1 and cr-osvp1-2 in the absence of wetting after seed maturation (data not shown). However, in contrast to NIP, both cr-osvp1-1 and crosvp1-2 showed severe PHS phenotypes following seven consecutive rainy days after seed maturation (Fig. 1-B). The mean seed germination rates of cr-osvp1-1 and cr-osvp1-2 were respectively 94.0% and 96.7% on the seventh rainy day,whereas that of NIP was only 5.0%(Fig.1-C).

3.2.Subcellular localization and expression analysis of OsVP1

To investigate the expression profiles of OsVP1,we examined the subcellular localization of OsVP1 by transiently expressing OsVP1-GFP in rice protoplasts.Because OsVP1 is recognized as a nuclear localization transcriptional factor [46], the reported nuclear localization D53 protein fused with mCherry, a member of the mFruits family of monomeric red fluorescent proteins (mRFPs), was used as a marker. Subcellular localization assays showed that the fluorescence signaling of OsVP1-GFP was present in nuclei and completely co-localized with the red fluorescence signal of D53-mCherry (Fig. 2-A). Yeast transcriptional activation activity assays showed that OsVP1 had strong transactivation activity (Fig. 2-B). These results suggested that OsVP1 encodes a nucleus localization transcription factor with transcriptional activation activity.

Fig.2- Subcellular localization,transactivation,and tissue expression of OsVP1.(A)Subcellular localization of OsVP1.Free green fluorescent protein(GFP)and OsVP1 fusion protein(OsVP1-GFP) was transiently expressed in rice protoplasts. D53-mCherry serves as a nuclear localization reference.pAN580-GFP was used mainly as a control.Scale bar,10 μm.(B)Transactivation effects of OsVP1 in yeast cells.Yeast transformants were spotted onto control medium(SD/-Leu/-Trp)and selective medium(SD/-Leu/-Trp/-His/-Ade)with X-α-Gal.The interactions between pGADT7-T and pGBKT7-53(CK(+)),pGADT7 and pGBKT7(CK(?))were used as positive and negative controls,respectively.(C)Reverse transcription quantitative polymerase chain reaction analysis of the OsVP1 expression level in tissues and organs of NIP.Values are means±SD.DAF,days after fertilization.

To further reveal the expression patterns of OsVP1, the temporal and spatial expression profile of OsVP1 was investigated. The results showed that OsVP1 was constitutively expressed in all tissues and organs with highest expression levels in leaves, panicles, and seeds at different filling stages(Fig.2-C).

3.3. OsVP1 activated the expression of Sdr4

Previous studies showed that the expression level of Sdr4 was decreased in the OsVP1 mutants [25]. To further reveal the regulation relationship between OsVP1 and Sdr4, the expression levels of Sdr4 were determined in cr-osvp1-1 and cr-osvp1-2, and the results showed that the transcription of Sdr4 was markedly decreased in the two cr-osvp1 mutants in comparison with the wild type (Fig. 3-A), suggesting that Sdr4 acts downstream of OsVP1. The yeast one-hybrid experiment showed that OsVP1 activated the expression of LacZ driven by the Sdr4 promoter(Fig.3-B).The B3 domain of OsVP1 binds specifically to several motif sequences including CACCTG and CAACA [47]. The Sdr4 2-kb promoter sequence harbors a CACCTG motif starting 48 bp upstream of the start codon of OsVP1(Fig.3-C).In the EMSA,OsVP1 bound directly to probe 1(Fig. 3-C), and the shifted band signal was substantially weakened with the application of unlabeled probe 1. Thus,OsVP1 bound directly to the Sdr4 promoter.

To verify whether OsVP1 could activate the expression of Sdr4, luciferase (LUC) transient transcriptional activity assay was also performed in rice protoplasts.When 35S:OsVP1 or the control vector 35S:NONE was co-transformed with the reporter overexpressing firefly luciferase (LUC) driven by the Sdr4 promoter(ProSdr4:LUC)with the reference vector overexpressing renilla luciferase, OsVP1 activated Sdr4 expression in rice protoplasts (Fig. 3-D). Thus, OsVP1 bound directly to the Sdr4 promoter and activated its expression.

Fig.3-OsVP1 directly binds to the Sdr4 promoter and activates its expression.(A)Relative expression levels of Sdr4 in 15 DAF kernels of NIP,cr-osvp1-1 and cr-osvp1-2.Values are means±SD with at least three biological replicates.**,P<0.01 by Student's t-test.(B)Yeast one-hybrid assay showing that OsVP1 bound to the promoter of Sdr4.(C)EMSA assay showing that OsVP1 binds directly to probe 1 containing the motif CACCTG on the Sdr4 promoter.The 5-,10-,and 20-fold excess unlabeled probe 1 was used as a competitor in the assay.(D)Transient transcriptional activity assay showing that OsVP1 activates the expression of Sdr4 in rice protoplasts. Values are means±SD with three biological replicates. **,P< 0.01 by Student's t-test.

3.4. Sdr4 functions as a downstream gene of OsVP1 in the regulation of PHS

Previous studies indicated that Sdr4 was identified as a major QTL for seed dormancy [25]. To further understand the molecular function of Sdr4, CRISPR-Cas9 gene editing technique was used to generate Sdr4 loss-of-function mutant.As a result, one mutant cr-sdr4 was produced with a threonine (T)insertion in the target region (Figs. S1-B, D, E; S3-A). This mutant showed severe PHS phenotypes similar to those of the two OsVP1 mutants(Fig.S3-B),with a seed germination rate as high as 100% following seven consecutive rainy days after seed maturation (Fig. S3-C). Subcellular localization experiments showed that Sdr4-GFP was localized in the nuclei (Fig.S4-A).Also,temporal and spatial expression analysis demonstrated that Sdr4 was constitutively expressed in various tissues and organs like OsVP1(Fig.S4-B).

Thus, OsVP1 and Sdr4 loss-of-function mutants showed similar PHS phenotypes, and OsVP1 directly regulates Sdr4 expression, whereas Sdr4 had no effect on the expression of OsVP1 (Fig. S5), suggesting that Sdr4 acts downstream of OsVP1. When Sdr4 was overexpressed in the mutant cr-osvp1-1, its expression level of Sdr4 was significantly increased in the two transgenic plants OE-Sdr4-1/cr-osvp1-1 and OE-Sdr4-2/cr-osvp1-1(Fig.4-A).After seven days of imbibition,OE-Sdr4-1/cr-osvp1-1 and OE-Sdr4-2/cr-osvp1-1 showed decreased PHS phenotypes in comparison to the control cr-osvp1-1 (Fig. 4-B).Seed germination rates of OE-Sdr4-1/cr-osvp1-1 and OE-Sdr4-2/cr-osvp1-1 were significantly decreased compared with that of cr-osvp1-1 (Fig. 4-C). However, OE-OsVP1-1/cr-sdr4 and OEOsVP1-2/cr-sdr4, two OsVP1-overexpressing transgenic lines with the background of the mutant cr-sdr4, showed no reduction in seed germination rates compared with cr-sdr4(Fig.4-A,C).We inferred that Sdr4 acts as a downstream target of OsVP1 in regulating seed dormancy.

Fig.4- Overexpression of Sdr4 partially rescued the PHS phenotypes of cr-osvp1-1.(A)Relative expression levels of Sdr4 and OsVP1 in transgenic plants.Values are means± SD of at least three biological replicates. **,P <0.01 by Student's t-test.(B)Phenotypes of germinating seeds in panicles of wild-type and transgenic plants after seven days of imbibition.(C)Germination rates of fresh seeds harvested from wild-type and transgenic plants.Germination assays were conducted in a growth chamber under 12 h/12 h light/dark for one week.Values are means of at least three biological replicates.

3.5.OsVP1 and Sdr4 loss-of-function mutants show decreased sensitivity to ABA signaling

In previous studies, ABA synthesis or signal transduction pathways were involved in seed dormancy [48-50]. To investigate whether ABA could affect the molecular function of OsVP1 or Sdr4 in regulating seed dormancy, seeds of wild type NIP and two mutants cr-osvp1-1 and cr-sdr4 were harvested after maturation, stored for 3 months in indoors,and then assayed for ABA sensitivity. As shown in Fig. 5-A,increasing the ABA concentration from 5 to 50 μmol L?1significantly reduced the seed germination rate of NIP,but did not reduce the germination rates of the two mutants cr-osvp1-1 and cr-sdr4.At 50 μmol L?1ABA,the germination rates of crosvp1-1 and cr-sdr4 were 100% on the seventh day, while that of the NIP reached only 65% (Fig. 5-A). ABA treatment also markedly inhibited the growth of NIP seedlings but not those of the two mutants(Fig.5-B, C).

To data, several ABA-responsive related genes, such as OsLEA3, OsLIP9, OsRAB16A, OsEm1, and OsNAC19, have been identified[51-54].In this study,the endogenous transcripts of the above genes were examined in NIP,cr-osvp1-1 and cr-sdr4.The results showed that the expression levels of ABA-responsive genes OsLEA3 and OsRAB16A were reduced in the two mutants in comparison with NIP, whereas OsLIP9 transcripts were increased in the two mutants (Fig. 6-A, B,C).The finding that,Sdr4 functions downstream of OsVP1 and that the loss-of-function mutants cr-osvp1-1 and cr-sdr4 showed decreased sensitivity to ABA signaling reveals a direct connection between OsVP1 and Sdr4 in controlling seed dormancy in an ABA-dependent manner (Fig. 6-D).

4.Discussion

4.1.OsVP1 and Sdr4 loss-of-function mutants show increased PHS,but only slight effects on agronomic traits

OsVP1 knockdown transgenic plants were produced using RNA interference(RNAi),showing that OsVP1 regulated rice seed dormancy [55]. In the present study, two OsVP1 mutants, cr-osvp1-1 and cr-osvp1-2, and one Sdr4 mutant crsdr4 were generated using CRISPR/Cas9 gene editing.All the mutants showed increased PHS compared with the wild type following seven rainy days after seed physiological maturation (Figs. 1; S3), showing that both OsVP1 and Sdr4 are positive regulators of seed dormancy. Grain length,Grain width, Grain thickness, 1000-grain weight, plant height, and tiller number did not differ among NIP, crosvp1-1 and sdr4,except that the plant heights of cr-osvp1-1 and cr-osvp1-2 were less than that of NIP (Fig. S6). Thus,OsVP1 and Sdr4 play similar roles in controlling seed dormancy,but show only slight effects on agronomic traits,suggesting that PHS resistance could be improved by marker-assisted breeding without affecting agronomic traits.

4.2. The OsVP1-Sdr4 regulation pathway determines seed dormancy

In previous studies,TRAB1,a basic region leucine zipper(bZIP)factor responsible for ABA regulation, interacted with OsVP1.TRAB1 activated ABA-responsive elements(ABREs),but not by OsVP1, revealing a molecular mechanism for the OsVP1-dependent and ABA-inducible transcription that regulates seed dormancy [39]. Sdr4, a major quantitative trait locus for seed dormancy,encodes an unknown protein and contributes substantially to differences in seed dormancy between japonica and indica cultivars[25].TRAB1 bound to the putative five ABRE motifs in Sdr4 promoter regions and activated its expression, which was increased by the addition of OsVP1 proteins as a cofactor of TRAB1 in the reaction [25,39]. These findings revealed the biochemical and molecular relationship among OsVP1, TRAB1, and Sdr4 but did not show whether OsVP1 directly regulates the expression of Sdr4 or reveal their genetic relationships. In the present study, we showed that OsVP1 can bind the CACCTG motif in the promoter of Sdr4 and activate its expression (Fig. 3). Sdr4-overexpressing cr-osvp1-1 mutants showed greatly reduced seed germination rates,revealing OsVP1-Sdr4 as a novel regulation pathway in determining seed dormancy.

4.3.OsVP1 and Sdr4 are involved in ABA and GA biosynthesis and signaling pathways in regulating seed dormancy

In the ABA signaling pathway,ABA binds to its receptor PYL/PYR/RCARs, forming an ABA-PYL/PYR/RCAR complex that interacts with PP2C phosphatases and represses their phosphatase activity,consequently releasing activated SnRK2s to phosphorylate downstream targets to promote ABA response [56]. In the present study,seed germination rates and seedling growth of crosvp1-1 and cr-sdr4 were insensitive to exogenous ABA,suggesting that OsVP1 and Sdr4 function downstream of ABA signaling(Fig. 5). The expression levels of several ABA-responsive genes varied significantly.The expression levels of ABA receptor genes OsPYL1 and OsPYL2,showed significant differences from that of NIP.Thus,both OsVP1 and Sdr4 are involved in the ABA signaling pathway in controlling seed dormancy.

ODR1/AtSdr4L,a homolog of OsSdr4 in Arabidopsis,has been well characterized and plays an important role in the regulation of seed germination and dormancy [57,58]. Cao et al. [57]reported that Arabidopsis thaliana SEED DORMANCY 4-LIKE (AtSdr4L) is a novel specific regulator of dormancy and germination.AtSdr4L loss-of-function mutants are insensitive to gibberellin (GA), hypersensitive to the GA biosynthesisinhibitor paclobutrazol but not sensitive to ABA treatment[57].Several GA biosynthesis genes and GA-regulated cell wall remodeling genes were down-regulated in the mutant in comparison to that of the wild type, suggesting that the Atsdr4l mutation causes both decreased GA biosynthesis and decreased responses[57].These results agree with a previous finding [25]that OsSdr4 is a negative regulator of the expression of OsGA20ox-1. Liu et al. [58]found that AtABI3 suppressed the expression of AtODR1, which encodes a key factor that interacts with bHLH57 to inhibit bHLH57-mediated NCED6 and NCED9 expression to control ABA biosynthesis and seed dormancy.

Fig.5-cr-osvp1 and cr-sdr4 are insensitive to exogenous ABA.(A)Seed germination assay of NIP,cr-osvp1 and cr-sdr4 treated with three ABA concentrations.Values are means of at least three biological replicates. (B)Phenotypic characteristics of seed germination after seven days of ABA treatments.(C)Measurement of seedling length calculated from (B).

Fig.6-Expression levels of ABA-responsive genes and a proposed model of seed dormancy via activation of Sdr4 by OsVP1.(AC)The expression levels of ABA-responsive related genes OsLEA3(A),OsLIP9(B),and OsRAB16A(C).Values are means±SD of at least three biological replicates. **,P <0.01 and*,P <0.05 by Student's t-test.(D)Proposed model of seed dormancy via activation of Sdr4 by OsVP1.

Transcript levels of NCED3 and NCED4, which are required for ABA biosynthesis, were both up-regulated in sdr4 seeds compared with that of the wild type, whereas NCED5 was down-regulated in the Sdr4 loss-of-function mutant (Fig. S8).The expression of the GA biosynthesis genes GA20ox1 and GA3ox2 was up-regulated in both OsVP1 and Sdr4 loss-offunction mutants. The expression of the GA metabolic gene OsGA2ox3 was also up-regulated in both mutants. Two other GA metabolic genes, OsGA2ox6 and OsGA2ox9, were upregulated in cr-osvp1-1 and cr-sdr4, respectively (Fig. S9).Thus, the regulation networks AtABI3-AtODR1 in Arabidopsis and OsVP1-Sdr4 in rice might control seed dormancy in different ways involving ABA and GA biosynthesis and signaling pathways.

5.Conclusions

We generated OsVP1 and Sdr4 loss-of-function mutants, both of which showed increased PHS compared with the wild type NIP.OsVP1 bound directly to the specific motif CACCTG in the promoter of Sdr4,a key gene controlling PHS,and activated its expression in rice protoplasts.In agreement with this finding,the expression of Sdr4 was sharply reduced in OsVP1 mutants.Overexpression of Sdr4 significantly decreased the seed germination rates of cr-osvp1-1, revealing that Sdr4 acts as a downstream target of OsVP1. The OsVP1 and Sdr4 mutants were insensitive to exogenous ABA and employ the ABA signaling pathway to regulate seed dormancy.Collectively,we have established a new framework for the control of seed dormancy to prevent the PHS in rice.

CRediT authorship contribution statement

Wenqiang Chen and Wei Wang performed most of the research and drafted the manuscript. Yusong Lyu, Yawen Wu, and Pingliang Huang prepared rice protoplast cells and performed subcellular localization. Shikai Hu, Xiangjin Wei,Guiai Jiao,and Zhonghua Sheng gave experimental guidance.Shaoqing Tang revised the manuscript. Gaoneng Shao and Ju Luo designed the experiment, supervised the study, and revised the manuscript.

Declaration of competing interest

The authors declared that they have no conflicts of interest to this work.

Acknowledgments

This work is supported by grants from the National Major Science and Technology Program on New GMO Organism Variety Breeding (2016ZX08001-001) and Agricultural Science and Technology Innovation Program of Chinese Academy of Agricultural Sciences(CAAS).

Appendix A.Supplementary data

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2020.06.005.