Background

Rice is an important food crop that feeds more than half of the world population. With an increasing global population, food security and safety are becoming serious issues. Given the limited availability of arable land and the growing human population, a future increase in rice production will be a major challenge for rice breeders (Wang et al. 2008). Plant height and leaf angle are important agronomic traits directly affecting grain yield and crop architecture, which are major objectives of crop improvement (Ikeda et al. 2013).

Plant height, an important trait in crop improvement, is related to lodging resistance, grain yield, and biomass production. Reduced plant height and associated lodging resistance are predominant strategies in crop improvement (Ayano et al. 2014). Various factors cause reduced height in plants; gibberellin (GA) and brassinosteroids (BRs) are the most widely investigated factors affecting plant height in rice. Many rice GA- and BR-related mutants show dwarf or semi-dwarf phenotypes, such as sd1, d1 and d61 (Sasaki et al. 2002; Fujisawa et al. 1999; Zhang et al. 2016; Yamamuro et al. 2000). These dwarf mutants result from reduced cell numbers or cell length in stems, and analyzing more dwarf mutants may provide novel insights into the mechanisms controlling stem elongation.

Leaf angle, the degree of bending between the leaf blade and culm, is a key factor determining the plant architecture and grain yield (Zhao et al. 2010). Crops with erect leaves have increased photosynthetic effciency and nitrogen storage for grain flling and are suitable for dense planting (Sakamoto et al. 2006). Many genes or QTLs have been reported to control leaf angle, including D61/OsBRI1, ILI1, LC2, ILA1, RAV6, OsARF19, and SLG (Yamamuro et al. 2000; Zhang et al. 2009; Zhao et al. 2010; Ning et al. 2011; Zhang et al. 2015b; Zhang et al. 2015a; Feng et al. 2016). Most identified rice mutants with altered leaf inclination have abnormal cell division and or expansion and altered cell wall composition at the leaf:stem joint (Zhang et al. 2009; Zhao et al. 2010). Nevertheless, BR and auxin synergistically control leaf inclination in rice (Wada et al. 1981; Hirano et al. 2017).

The synergetic effects of auxin and BRs on various physiological events have strongly suggested the interdependency of their signaling, and molecular evidence for this has been reported. For example, the auxin-inducible genes, indole-5-acetic acid (IAA5), IAA19 and SAUR-AC1, are induced by brassinolide (BL) treatment, whereas the expression of IAA5 and IAA19 are down-regulated in the Arabidopsis BR-deficient mutant de-etiolated 2 (det2) (Nakamura et al. 2003). In addition, expression of BR biosynthetic gene DWARF4 and the BR receptor BRI1 are induced by auxin (Park et al. 2011; Sakamoto et al. 2012). BRASSINAZOLE RESISTANT 1 (BZR1) directly regulates many auxin-responsive genes (Sun et al. 2010). Nemhauser et al. (2004) found that 48 genes were co-regulated by auxin and BL (e.g. SAUR, Aux/IAA and GH3). They also found that a TGTCTC element in the auxin-responsive element (AuxRE) was enriched in genes up-regulated by both IAA and BL. These results strongly suggest that the two hormone pathways synergistically affect gene expression at the transcriptional level.

In this study, we identified a reduced BR-sensitivity mutant, ds1, showing dwarfness, smaller seed and decreased leaf angle compared to wild type. The small plant stature of the plants and decreased leaf angle resulted from reduced cell elongation. Our genetic mapping and molecular biology experiments revealed that the underlying DS1, a new allele of OsEMF1, encodes an EMF1-like protein involved in the development of plant height and leaf angle by interacting with OsARF11 to co-regulate OsBRI1 expression.

Results

Characterization of the ds1 mutant

To identify genetic factors regulating rice architecture, we screened our Nanjing 35 radiation mutagenesis population by isolating dwarf mutants. A mutant was identified with dwarfism and decreased leaf angles, ds1 (Fig. 1a-d).

Fig. 1
figure 1

Phenotypic comparison of wild type and ds1 mutant. a Gross morphology of the wild type and ds1 plants. Scale bar, 10 cm. b Comparison of internode and panicle lengths between WT and ds1. Scale bar, 2 cm. c Statistical data for internode lengths described in (b). Data are presented as means ± SD. d Young spikelet hulls of WT (left) and ds1 (right) plants. Scale bar, 1 mm

Full size image

Differences between wild type and ds1 mutant plants became apparent only after internode elongation when reduced height and darker leaf colour of the mutant were expressed. There was no difference between ds1 and WT in young seedlings. Leaves of the mutant remained erect until maturity (Fig. 1a). Spikelet number per panicle, numbers of primary and secondary panicle branches of ds1 were less than the WT (Additional file 1: Table S1). The panicles of ds1 at maturity were almost erect whereas the panicles of WT clearly bend down. All internodes of ds1 were shorter than those of WT (Fig. 1b-c). The dwarfism of ds1 belonged to the dn type based on the classification of Yamamuro et al. (2000). Flag and second leaf angles of ds1 were smaller than WT (Fig. 2a-b). These altered morphologies of ds1 suggested that DS1 participated in rice growth and development.

Fig. 2
figure 2

DS1 is a positive regulator of BR signaling in rice. a Flag and second leaf angles of WT (left) and ds1 (right) at maturity. b Statistical data for flag and second leaf angles shown in (a). Data are presented as means ± SD. c Lamina joint bending response to various amounts of 24-epiBL in WT (left) and ds1 mutant (right) seedlings. d Statistical data for lamina joint bending angle assay described in (c). Data are means ± SD.**, significant difference between WT and ds1 at P = 0.01

Full size image

The ds1 mutant is less sensitive to BRs

As ds1 had dwarfism, erect and dark green leaves, characteristic phenotypes of BR-related mutants, we hypothesized that DS1 was involved in the BR pathway. Leaf angle in rice was known to be sensitive to the concentration of BL or related compounds (Wada et al. 1981). We tested the sensitivity of ds1 to 24-epiBL in a lamina joint bending experiment. BL treatment caused a dose-dependent lamina joint inclination in WT, whereas ds1 plants were insensitive. Treatment with BL showed an increased level leaf angle in WT plants, but little change in ds1 plants (Fig. 2c-d).

DS1 regulates plant height and leaf angle by promoting cell elongation

To investigate the mutant phenotype of ds1 at a cellular level we performed scanning electron microscopy (SEM) on ds1 plants along with the WT (Fig. 3a-c). Dwarfism of ds1 plants was mainly due to reduced cell length as measured in 2nd internodes (Fig. 3a-b). Observations on adaxial cells, which influence leaf angle, showed that those cells were also shorted in the ds1 mutant (Fig. 3c). These results indicated that reduced cell length caused the dwarf phenotype of ds1.

Fig. 3
figure 3

Histological analysis of the 2nd internodes and lamina joints. a Transverse sections of WT and ds1 mutant 2nd internodes. Bar, 100 μm. b Longitudinal sections of WT and ds1 mutant 2nd internodes. Bar, 500 μm.** indicates significant difference at P < 0.01. c Cross section of lamina joints of flag leaves in WT (left) and ds1 (right). ** indicates significant difference at P < 0.01

Full size image

As the cell wall is the major factor restricting cell elongation, many genes involved in cell wall synthesis are reported to regulate cell elongation (Ning et al. 2011). We investigated the expression levels of cell wall synthesis-related genes in WT and the ds1 mutant. Expression levels of IRX10L, CESA6, and UGA4e were significantly down-regulated in ds1 compared to WT (Additional file 1: Figure S1). Consistent with the morphological observations, this indicated that the dwarfism in ds1 was caused by reduced cell length. To eliminate the possibility of decreased cell size in ds1, we analysed the expression levels of nine cell expansion regulators in WT and the ds1 mutant (Additional file 1: Figure S1). Seven genes were significantly down-regulated in ds1, indicating that cell expansion was reduced. These results further confirmed that reduced cell elongation was the cause of the phenotype of ds1.

Map-based cloning and functional analysis of DS1

To determine the gene responsible for the ds1 phenotype, we made a cross between ds1 and the wild type. F1 plants showed the WT phenotype and the F2 segregation ratio was consistent with that expected a single locus (χ23:1 = 0.98, P > 0.05). These results indicated that ds1 phenotype was due to a single recessive gene.

To map the DS1 locus, we next generated an F2 population from a cross between ds1 and indica cultivar 93–11. DS1 was roughly mapped on the short arm of chromosome 1 between SSR markers N1–12 and N1–017. We then used 568 recessive individuals from the same F2 population and narrowed down the DS1 locus to a 80 kb region between markers L-3 and L-2 within BAC clones P0452F10 and P0485D09. A total of 11 putative genes were localized in the 80 kb region according to the genome annotation of Nipponbare (http://www.gramene.org/). On comparison of the mapped genomic region between WT and ds1, we detected a single nucleotide deletion at the first exon and two nucleotide deletions in the second exon of LOC_Os01g12890 (Fig. 4a). We identified three domains in the DS1 polypeptide, including a nuclear localization signal (NLS), an ATP/GTP binding motif, and an LXXLL motif (Fig. 4b). The LXXLL motif had been demonstrated to mediate interactions that can activate or repress transcription (Heery et al. 1997; Torchia et al. 1997; Calonje et al. 2008).

Fig. 4
figure 4

Map-based cloning of DS1. a Fine mapping of the DS1 gene on chromosome 1. The DS1 locus was mapped to an 80 kb region that contained 11 predicted open reading frames (ORFs). b DS1 protein structure. Gray box, NLS; light gray box, GTP-ATP binding motif; black box, LXXLL motif. c Characterization of T1 transgenic plants when ds1 was transformed with the introduction of DS1 fragment. Bar, 10 cm

Full size image

The OsEMF1 gene, annotated as the LOC_Os01g12890 gene model, encodes an EMF1-like protein, involved in H3K27me3-mediated epigenetic silencing. In Arabidopsis, EMBRYONIC FLOWER 1 (EMF1), a plant-specific protein, participated in H3K27me3-mediated silencing of target genes by acting downstream of EMF2 and likely interacted with both PRC1 RING-finger proteins and PRC2 component MULTICOPY SUPRESSOR OFIRA1 (MSI1) (Aubert et al. 2001; Calonje et al. 2008). In rice, OsEMF1 plays an important role in palea development through maintaining H3K27me3-mediated epigenetic repression on OsMADS58 (Zheng et al. 2015; Yan et al. 2015). To confirm whether mutation of DS1 was responsible for the ds1 phenotype, a CDS fragment of 3174 bp containing the entire DS1 coding sequence and the DS1 promoter were cloned into the pCUbi1390 binary vector, and the complementation construct was then transformed into the ds1 mutant. All 8 independent DS1-complementation lines rescued normal leaf angle and plant height of ds1 plants (Fig. 4c). These genetic evidence and transformation results confirmed that DS1 was a newly identified OsEMF1 allele, which influenced leaf angle and plant height. Homology analysis found that amino acid sequence of DS1 had 37% similarity and 20% homology to Arabidopsis thaliana EMF1 (AtEMF1) (Additional file 1: Figure S2).

Expression pattern of DS1

To further explore the biological functions of DS1, we found that DS1 was highly expressed in the seeds,young leaves and panicles according to the Rice eFP Browser (http://bar.utoronto.ca/efprice/, Fig. 5a). Then we tested the DS1 expression pattern in various organs from WT and ds1 by quantitative real-time PCR. The results showed that DS1 was highly expressed in young leaves, panicles and seeds, with the lowest expression in the roots. In comparison with the WT, the expression level was not significantly different in all tested organs of the ds1 mutant (Fig. 5b). To determine whether the expression of DS1 was affected by BR, seeds were cultured on half-strength medium containing 1 μM 24-epiBL and expression of DS1 was determined by quantitative real-time PCR. As shown in Fig. 5c, the DS1 expression was induced by BL treatment in WT.

Fig. 5
figure 5

Spatiotemporal expression of DS1. a Expression pattern of DS1 based on the Rice eFP Browser. b Quantitative RT-PCR analysis of DS1 expression in various rice tissues. Roots and leaves were harvested from plants at the early booting stage. Internodes and stems were collected from 5-week-old seedlings. Panicles were collected when they had reached 2 cm length. Seeds were collected at grain filling. c Effect of 24-epiBL on DS1 expression. 5-day-old seedlings of the wild type (WT) were harvested after treatment with 1 μM 24-epiBL. CK represents the mock-treated control

Full size image

DS1 interacts with OsARF11

To gain a deeper insight into the molecular function of DS1, we performed a yeast two-hybrid (Y2H) screening using DS1 as bait. We obtained many candidate interacting proteins, including auxin response factor 11 (OsARF11). The interaction between DS1 and OsARF11 was further confirmed using the full-length cDNA in yeast (Fig. 6a-b). Auxin response factors (ARFs), including OsARF11 and OsARF19, were reported to regulating plant height and leaf angle. Loss-of-function osarf mutants showed decreased leaf angle and dwarfism, similar to the ds1 mutant. Also, OsARFs directly regulate expression of OsBRI1 by directly binding to the AuxRE element in the promoter of OsBRI1 (Shen et al. 2010; Sakamoto et al. 2013; Zhang et al. 2015a).To validate this interaction, we performed a bimolecular fluorescence complementation (BiFC) study. The N-terminal half of YFP was fused to DS1, and the C-terminal half of YFP was fused to OsARF11. Both constructs were then transiently expressed in tobacco (Nicotiana bethamiana). YFP fluorescence was observed in the nucleus, confirming that DS1 interacted with OsARF11 in planta (Fig. 6c). Taken together, these results indicated that DS1 might regulate plant height and leaf angle through interaction with OsARF11.

Fig. 6
figure 6

DS1 physically interacts with OsARF11. a Sketches showing the domain structures of DS1 and OsARF11 and various deletions. b Yeast two-hybrid assays showing the interactions between DS1, OsARF11, and their derivatives. Transformed yeast cells were grown on SD/−Trp/−Leu/-His/−Ade. c BiFC assay showing DS1-YFPN and OsARF11-YFPC interaction to form a functional YFP in the nucleus

Full size image

Expression pattern and subcellular localization of OsARF11

To verify whether the expression pattern of OsARF11 is similar to that of DS1, we found that OsARF11 was highly expressed in the shoot apical meristem and young panicles according to the Rice eFP Browser (http://bar.utoronto.ca/efprice/, Fig. 7a). We employed a quantitative RT-PCR (qRT-PCR) assay to examine OsARF11 expression in wild-type plants. Real time qPCR analysis showed that OsARF11 transcripts accumulated in all examined tissues, including sheaths, young panicles, roots, stems and leaves (Fig. 7b). These results indicated that OsARF11 showed a very similar expression pattern to DS1. We further, found that OsARF11 expression was induced by auxin treatment (Fig. 7c), suggesting that its function might be regulated by auxin. We next examined the subcellular localization of OsARF11 protein. A construct harboring a OsARF11-GFP fusion was introduced into tobacco epidermal cells by Agrobacterium. Fluorescent signals of OsARF11-GFP fusion protein were observed exclusively in the nucleus of tobacco cells (Fig. 7d), indicating that OsARF11 is a nuclear protein, consistent with the function of OsARF11.

Fig. 7
figure 7

Expression pattern and subcellular localization of OsARF11. a Expression pattern of OsARF11 based on the Rice eFP Browser. b Expression pattern of OsARF11 detected by qRT-PCR. The experiment was repeated three biological times with similar results. Error bars represent standard deviations (SD) (n = 3). R, root; S, stem; L, leaf; YP, young panicles; Sh, Leaf sheath. c OsARF11 expression under auxin treatments. One-week-old WT seedlings were grown in solution containing 10 μM IAA for 4 h. ** indicate significant difference at P < 0.01. Error bars represent the SD (n = 3). d Subcellular localization of the OsARF11-GFP fusion protein in tobacco epidermal cells

Full size image

DS1 and OsARF11 co-regulate D61/OsBRI1 expression

As shown in Fig. 2c, the ds1 mutant was insensitive to BL treatment. We investigated the effect of DS1 on the expression of BR-related genes and the results showed that disruption of DS1 led to significantly increased expression of D2 and DWARF4, two important brassinosteroid biosynthesis rate-limiting enzymes (Hong et al. 2003; Sakamoto et al. 2006). By contrast, DS1 had large effects on expression of BR signal genes, including D61/OsBRI1, OsBZR1, OsBU1 and XTR1 (Fig. 8). This indicated that DS1 could be involved in BR signaling.

Fig. 8
figure 8

Expression of BR biosynthesis and signal transduction genes. Real-time PCR analysis of BR biosynthesis and signal transduction genes in the wild type and ds1 seedling at 15 days after germination. Data are presented as means ± SD (n = 3)

Full size image

Previous studies showed that OsARF11 promotes plant growth and leaf angle by specific binding to an auxin response element (AuxRE) in the promoter of OsBRI1 (Sakamoto et al. 2013), and that OsARF19 binds to an AuxRE element in the promoters of OsBRI1 and OsGH3–5 (Zhang et al. 2015a). We analyzed the promoters of four BR signal genes (OsBRI1, OsBZR1, OsBU1 and XTR1), and found that only the promoter of OsBRI1 contained the auxin response element. To confirm the binding of DS1 or OsARF11 to the OsBRI1 promoter, we performed yeast one hybrid assays using in vitro-expressed DS1 and OsARF11. As shown in Fig. 9a, OsARF11 bound to the OsBRI1 promoter but DS1 failed to bind the promoter of OsBRI1.

Fig. 9
figure 9

DS1 interacts with OsARF11 to promote the expression of OsBRI1. a Yeast one-hybrid (Y1H) analysis of DS1, OsARF11 and OsBRI1 promoter. The bait vector containing the OsBRI1 promoter fragment-fused lacZ reporter gene, and the prey vector containing DS1/OsARF11-fused GAL1 activation domain were co-transformed into yeast cells (EGY48). b Transient expression assays of OsBRI1 transcriptional activity modulated by DS1/ds1 and OsARF11 in rice protoplasts. Various constructs used in transient expression assays are shown in the upper panel. pOsBRI1:LUC was cotransformed with either the effector or empty vector (control) into rice protoplasts. Relative LUC activity indicating the level of OsBRI1 expression activated by the effector is shown in the lower panel. Values are means ± SD of three biological replicates. c Model representing mechanism of action of DS1 in regulating OsBRI1

Full size image

To examine the regulation of DS1 and OsARF11 on the expression of its target gene OsBRI1, we performed transient expression assays using ~ 2 kb of the OsBRI1 promoter fused to the LUC gene as a reporter. Effector constructs for OsARF11 were expressed under control of the 35S promoter and transfected together with the reporter construct into rice protoplasts. Higher LUC activity was detected when OsARF11 protein was transfected with the reporter construct compared with the internal control (Fig. 9b). These findings strongly support the hypothesis that DS1 and OsARF11 coregulate OsBRI1 in controlling rice plant architecture.

Discussion

Plant architecture is a major factor underlying grain yield and determination of planting density and photosynthesis efficiency in rice. The molecular mechanisms of plant architecture have been extensively studied and many critical genes have been identified. Most of these genes are involved in the Auxin and BR pathways. Auxin and BR signaling share many transcriptional target genes in regulating rice plant architecture (Yin et al. 2002; Nemhauser et al. 2004), suggesting that interactions between TFs involved in auxin and BR signaling might act as the points of auxin–BR crosstalk, but only a few examples have been reported. One is the response of SAUR15 to auxin and BR. SAUR15 expression depends on the simultaneous interaction of BES1 and ARF5 within the promoter of the SAUR15, gene that produces a hormone with an Up at Dawn (HUD)-type E-box and AuxRE cis elements (Walcher and Nemhauser 2012). Another example is that BZR1 and ARF6 were shown to interact and regulate a large number of common target genes (Oh et al. 2014).

In this study, we characterized mutant ds1, which showed dwarfing and a reduced leaf angle compared to the WT. Through map-based cloning and genetic transformation, DS1 was identified as a new allele of OsEMF1. In rice, OsEMF1 functions in influencing the level of the H3K27me3 at the OsMADS58 locus, and repressing the expression of OsMADS58 to control palea development (Zheng et al. 2015; Yan et al. 2015). ds1 became less sensitive to BL treatment relative to the wild type in a lamina joint inclination test. We found that the expression level of OsBRI1 was down-regulated 5 times in the ds1 mutant. OsEMF1 inhibits the expression of target genes by mediating H3K27me3, which is not consistent the down-regulation of OsBRI1 expression in the ds1 mutant. DS1 may influence the expression of OsBRI1 through histone methylation or other unknown mechanisms. We demonstrated that DS1 physically interacted with OsARF11 to positively co-regulate expression of OsBRI1. Indeed, DS1 and OsARF11 showed similar expression profiles.

Following the initial discovery of the effect of auxin-BR crosstalk on lamina joint bending, several genes have now been reported to be involved (Luo et al. 2016). For instance, LPA1 negatively regulates lamina joint bending by inhibiting BR signaling and increasing the expression of auxin transporters (Liu et al. 2016). In a recent study, the SMOS1-SMOS2/DLT complex positively regulated expression of OsPHI-1, and treatment of smos mutants with auxin and BR still showed a synergistic effect on lamina joint bending (Hirano et al. 2017). In the current study, we showed that DS1-OsARF11 is invoved in auxin-BR crosstalk in rice. It is plausible that lamina joint bending is determined by a transcriptional network consisting of DS1-OsARF11, SMOS1-SMOS2/DLT, LPA1, and other factors regulated by auxin and BR signaling.

In high-density plantings, characteristic semi-dwarf phenotypes with erect leaves are ideal phenotypes for improving grain yield (Lu et al. 2013; Wu et al. 2008). Further, identification of other downstream genes of DS1/OsEMF1 will be needed for fully understanding the network regulating plant growth and development. In addition, the BR signaling pathway is necessary for auxin function in controlling leaf angle, indicating that the roles of auxin and BR are interdependent. Our study uncovers a previously unknown regulatory factor in plant architecture and the coordinating network of auxin and BR signaling. These findings might be useful for manipulating optimal rice architectures.

Conclusions

In this study, we found that DS1/OsEMF1 functions in BR signaling by interacting with OsARF11. The underlying mechanisms might be useful for engineering rice architecture to increase rice yield.

Methods

Plant material and growth conditions

The ds1 mutant obtained from an irradiated population of japonica variety Nanjing 35 was stably inherited. The mutant was self-pollinated for several generations until shown to be heritable. Crosses were made between the ds1 mutant and variety Nanjing 35 for genetic analysis. 93–11 (indica) was employed as a control in mapping of DS1. Pre-germinated seeds were sown in nursery beds in the field and the 1-month-old seedlings were transplanted into a paddy field in Nanjing.

Trait measurement

The grain length, width, and thickness were measured by an electronic digital display vernier caliper and fully filled grains were used for measuring the 1000-grain weight. The plant height, primary and secondary branch number, panicle length, and grain number per panicle were obtained from the measurement of the main stem. Numbers of tillers per plant were recorded at the time of harvest. Three biological repeats were measured for each sample.

Histological analysis

The second internodes and lamina joints of the flag leaves were harvested at heading date, respectively. All samples fixed in glutaric dialdehyde for 48 h. The fixed internode samples were soaked in 2% (w/v) OsO4 for 2 h, dehydrated in a graded ethanol series (70, 80, 90, 100, and 100%), cleared in a xylene series, and then sputter coated with platinum. Second internode sections were observed by SEM (HITACHI, S-3000 N). Fixed lamina joints were embedded in paraplast. After sectioning, 10 μm thick sections were dewaxed with xylene, rehydrated, stained with 1% toluidine blue, and observed with a Leica DM5000B microscope. Cell lengths of each organ were measured with IMAGEJ software.

Lamina joint bending assay

Brassinolide (BL) was dissolved in ethanol. For BL treatment, sterile water was added to reach final concentrations of 0.01, 0.1, and 1 μM. Germinated seedlings of WT and ds1 were grown in the light for 10 d at 30 °C. Lamina joint segments were floated on distilled water for 72 h containing various concentrations of brassinolide in the light. The lamina joint bending angles were then photographed and measured. Twenty plants were used for each treatment.

Genetic analysis and map-based cloning of DS1

Populations for genetic analysis and mapping were from crosses between ds1 and japonica cv. Nanjing 35 and indica cv. 93–11. For mapping of DS1, 7680 F2 plants were generated from a cross between ds1 mutant and indica cv. 93–11, and 568 individuals with the clear ds1 phenotypes were selected for mapping. Additional molecular markers were designed from comparative sequences of Nipponbare and 93–11. Molecular markers chosen for mapping are listed in Additional file 1: Table S2. The PCR procedure was: 95 °C for 5 min, followed by 33 cycles of 95 °C for 30 s, annealing for 30 s, 72 °C for 40 s and a final elongation step at 72 °C for 5 min. All the populations were grown in the field at Nanjing in Jiangsu province, and Lingshui County in Hainan province.

Complementation of the ds1 mutant

To test whether the mutation in LOC_Os01g12890 was responsible for ds1, an ~ 2 kb promoter and full length cDNA of DS1 were cloned into the pCUbi1390 binary vector (primer sequences are listed in Additional file 1: Table S3). The resulting binary plasmid was then introduced into Agrobacterium tumefaciens strain EHA105. Calli induced from homogeneous ds1 seeds were used for transformation. The transgenic plants were grown in a glasshouse to observe the phenotypes.

RNA extraction and qRT-PCR analysis

Total RNAs were isolated from various plant organs using Trizol reagent (Invitrogen) according to the manufacturer’s instructions. First strand cDNAs were synthesized from 2 μg total RNA using a PrimeScript 1st Strand cDNA Synthesis Kit (Takara, Dalian).

For qRT-PCR, SYBR Premix Ex TaqTM kit (Takara) was added to the reaction system and run on an ABI Prism 7500 real-time PCR system and the OsActin gene was used as an internal control. The relative expression level was calculated by 2-ΔΔCt. The primers used for BR-related genes and cell wall synthesis-related genes are listed in Additional file 1: Table S3. The PCR procedure was: 95 °C for 30 s, followed by 40 cycles of 95 °C for 5 s, 60 °C for 34 s.

Yeast two-hybrid assay

For yeast two-hybrid analysis, various fragments were cloned into pGBKT7 and pGADT7 to construct BD-DS1/N/C/ and AD-OsARF11/N1/N2/C/, respectively. The reported gene assay was performed following the manufacturer’s instructions (Clontech). Yeast transformation and screening procedures were performed according to the manufacturer’s instructions (Clontech). The yeast cells were transformed with the rice cDNA library constructed from young panicles of the japonica cultivar “Nipponbare”. Positive transformants were screened for growth on SD/−Trp/−Leu/-His/−Ade. The yeast clones were cultured and sequenced. The primers used for yeast two-hybrid analysis were listed in Additional file 1: Table S3.

BiFC assay

The full-length cDNA of DS1 was cloned into the p2YN vector to construct the DS1-YFPN fusion protein. OsARF11 was cloned into the p2YC vector to produce an OsARF11-YFPC fusion protein, (primer sequences are listed in Additional file 1: Table S3). BiFC analyses were performed in tobacco, as described previously (Waadt and Kudla 2008). The primers used for BiFC assay are listed in Additional file 1: Table S3.

Transient expression and yeast one-hybrid assays

To generate the pOsBRI1:LUC reporter construct, ~ 2 kb of the OsBRI1 promoter was cloned into pGreenII-0800-LUC. OsARF11 and DS1 full length cDNA were cloned into the modified pAN580 binary vector, respectively. The construct vectors were co-transformed into rice protoplasts. Rice protoplasts were prepared, transfected, and cultured as previously described (Wang et al. 2016). The luciferase activity assay was calculated following the manufacturer’s instructions (Promega) and the data presented are the averages of three biological replicates.

For yeast one-hybrid (Y1H) assays, the full length cDNA of DS1, OsARF11 and promoter of OsBRI1 were amplified using gene-specifc primers (Additional file 1: Table S3), and the amplifed products were cloned into the pB42AD vector and reporter plasmid pLacZi, respectively. Plasmids were co-transformed into yeast strain EGY48 according to the manufacturer’s manual (Clontech).