Rice DWARF14 acts as an unconventional hormone receptor for strigolactone

This work reveals the conserved function of D14 from dicots and monocots, and defines rice D14 as an unconventional hormone receptor that can restore strigolactone signaling in the Arabidopsis d14 mutant

Abstract

Strigolactones (SLs) act as an important class of phytohormones to regulate plant shoot branching, and also serve as rhizosphere signals to mediate interactions of host plants with soil microbes and parasitic weeds. SL receptors in dicots, such as DWARF14 in Arabidopsis (AtD14), RMS3 in pea, and ShHTL7 in Striga, serve as unconventional receptors that hydrolyze SLs into a D-ring-derived intermediate CLIM and irreversibly bind CLIM to trigger SL signal transduction. Here, we show that D14 from the monocot rice can complement Arabidopsis d14 mutant and interact with the SL signaling components in Arabidopsis. Our results further reveal that rice D14, similar to SL receptors in dicots, also serves as an unconventional hormone receptor that generates and irreversibly binds the active form of SLs. These findings uncover the conserved functions of D14 proteins in monocots and dicots.

Introduction

Strigolactones (SLs) are a family of carotenoid-derived phytohormones (Gomez-Roldan et al., 2008; Umehara et al., 2008) that play a vital role in the control of plant shoot branching, a key agricultural trait that determines plant architecture and affects crop yield (Song et al., 2017; Zhang et al., 2017). SLs also act as rhizosphere signals to establish symbiotic interactions between plants and soil microbes (Akiyama et al., 2005; Kretzschmar et al., 2012; Gutjahr et al., 2015), and regulate parasitic interactions by stimulating germination and growth of parasitic weeds such as Striga (Cook et al., 1966; Cardoso et al., 2014; Conn et al., 2015; Toh et al., 2015; Tsuchiya et al., 2015; Gobena et al., 2017; Lumba et al., 2017a). Moreover, SLs also regulate hypocotyl elongation, root growth, leaf development and senescence (Snowden et al., 2005; Kapulnik et al., 2011; Ruyter-Spira et al., 2011; Waters et al., 2012b; Waters and Smith, 2013; Yamada et al., 2014; Soundappan et al., 2015; Ueda and Kusaba, 2015; Wang et al., 2015; Li et al., 2016). Recently, SLs were revealed to regulate various plant stress responses including drought tolerance and disease resistance (Umehara et al., 2010; Dor et al., 2011; Kohlen et al., 2011; Bu et al., 2014; Ha et al., 2014; Decker et al., 2017).

Genetic and molecular characterization of highly branched mutants in various plant species, such as rice dwarf14 (d14), d3, d10, d17, d27, and d53 (Ishikawa et al., 2005; Arite et al., 2007, 2009; Gao et al., 2009; Lin et al., 2009; Liu et al., 2009; Jiang et al., 2013; Zhou et al., 2013), Arabidopsis more axillary growth1 (max1), max2, max3, max4, and d14 (Woo et al., 2001; Stirnberg et al., 2002; Sorefan et al., 2003; Booker et al., 2004, 2005; Shen et al., 2007; Nelson et al., 2011; Waters et al., 2012a, b; Abe et al., 2014; Chevalier et al., 2014; Yao et al., 2016), pea ramosus1 (rms1), rms2, rms3, rms4, and rms5 (Beveridge et al., 1996; Morris et al., 2001; Foo et al., 2005; Johnson et al., 2006; de Saint Germain et al., 2016), and petunia decreased apical dominance 1 (dad1), dad2, and dad3 (Snowden et al., 2005; Simons et al., 2007; Hamiaux et al., 2012), suggests that SL biosynthesis and signaling pathway are largely conserved in diverse plant species.

Rice D3 or its ortholog in Arabidopsis (MAX2), petunia (PhMAX2A), or pea (RMS4) encodes an F-box protein (Johnson et al., 2006; Shen et al., 2007; Stirnberg et al., 2007; Drummond et al., 2011; Hamiaux et al., 2012), a subunit of SCF (Skp1–Cullin1–F-box protein) ubiquitin ligase complex that functions in substrate recognition for proteasome-mediated proteolysis. Rice D14 or its orthologs, such as Arabidopsis D14, petunia DAD2, or pea RMS3, encodes an α/β hydrolase that hydrolyzes SLs (Hamiaux et al., 2012; Nakamura et al., 2013; Zhao et al., 2013; de Saint Germain et al., 2016; Yao et al., 2016) and interacts with its respective F-box protein in an SL-dependent manner to recruit various repressors, such as D53 in rice and SMXL6/7/8 (SUPPRESSOR OF MAX2 1-LIKE6/7/8) proteins in Arabidopsis, for ubiquitination and degradation (Hamiaux et al., 2012; Jiang et al., 2013; Zhou et al., 2013; Soundappan et al., 2015; Wang et al., 2015; Liang et al., 2016; Yao et al., 2016). Degradation of these repressors will subsequently de-repress their directly targeted transcription factors, such as IPA1 (Ideal Plant Architecture 1) in rice (Song et al., 2017), to activate downstream genes essential for various SL-regulated plant responses (Lumba et al., 2017b; Waters et al., 2017). It is intriguing that the receptors AtD14 and D14 are also degraded in an SL-induced and MAX2/D3-dependent manner (Chevalier et al., 2014; Hu et al., 2017).

Recently, AtD14 was defined as a non-canonical hormone receptor that possesses dual functions: AtD14 acts as an enzyme to hydrolyze SLs and generate the active form of the hormone molecule CLIM (the covalently linked intermediate molecule), and also serves as a receptor of SL to bind CLIM irreversibly and undergo significant conformational change for interacting with downstream components and triggering SL signal transduction (Snowden and Janssen, 2016; Yao et al., 2016; Fang and Chen, 2017; Yao et al., 2017). Similarly to AtD14, RMS3 in pea and ShHTL7 in Striga hermonthica also hydrolyze strigolactone into the D-ring-derived intermediate CLIM and covalently bind CLIM in an irreversible manner to trigger SL signaling (de Saint Germain et al., 2016; Yao et al., 2017).

Rice is an important crop and also serves as a model plant for the study of monocots. It is known that rice, but not Arabidopsis, is the host plant for both symbiotic arbuscular mycorrhizal fungi (AMF) and parasitic Striga (Yoshida and Shirasu, 2009; Waters et al., 2017). In this study, we employ bioinformatics, genetic, and biochemical approaches to investigate functional conservations between rice D14 and Arabidopsis AtD14, and examine whether the rice D14 is similar to AtD14 in generating and perceiving the active form of SLs.

Materials and methods

Generation of transgenic plants

The modified binary vector pCAMBIA1300-cFlag (Yao et al., 2016) carrying the full coding sequence of Arabidopsis thaliana D14 (AtD14), Oryza sativa D14, or N-terminus (amino acids 1–51) truncated O. sativa D14 (D14ΔN) under the control of the 35S promoter was introduced into the Atd14-5 mutant (Yao et al., 2016) by using the Agrobacterium-mediated floral dip method. Similarly, the pCAMBIA1300-cFlag vector carrying D14ΔN under the control of the AtD14 promoter was introduced into the Atd14-1 mutant (Waters et al., 2012b) to generate transgenic plants, AtD14pro:D14ΔN. The primary rosette branching numbers were counted for 7-week-old plants which were germinated on plates and grown in soil under a light/dark photoperiod of 16 h/8 h at 22 °C.

Real-time PCR (RT-PCR) analysis

Rosette leaves of plants were collected for RNA extraction. Total RNA was prepared with a TransZol Kit (TransGen) and used in the reverse transcription reaction with the reagent TransScript^® RT/RI Enzyme Mix (TransGen). The first-strand cDNA was used as the template for the subsequent RT-PCR, which was performed to amplify AtD14 (primers 5'-ATGAGTCAACACAACATCTTAG-3' and 5'-GATGATTCCGATCATAGCG-3'), and D14 (primers 5'-TGACCTCTTCGCCAAGCTTG-3' and 5'-TCTTGAAGACG GTCTGGCAGAC-3') in the plants with the indicated genotypes. The A. thaliana ACTIN1 was employed as a control (primers 5'-TGTTGAGAAGAACTACGAGC-3' and 5'- AAGCACTTCCTGTGAACAAT-3').

Hypocotyl measurements

The Arabidopsis seeds were sterilized and germinated on Murashige and Skoog (MS) medium with or without 3 μM rac-GR24 (an SL analog, Chiralix) under continuous low light at 22 °C for 7 d. Hypocotyl length was measured by Digimizer software.

Leaf morphology analysis

Plants were grown in soil under a light/dark photoperiod of 16 h/8 h at 22 °C for an additional 3 weeks after germination and growth on MS medium for 1 week. For each genotype, 20 plants were used for observation of leaf morphology. Whole plants and their sixth leaves were photographed and harvested for further measurement. The leaf length (the distance between the leaf tip and the base of petiole) and leaf width (the greatest distance across the leaf lamina perpendicular to the proximal/distal axis of the leaf) were measured manually using a ruler.

Protein preparation

N-terminus-truncated O. sativa D14ΔN or full-length A. thaliana D14 (AtD14) was expressed in Escherichia coli strain BL21 (DE3) (Novagen) as an N-terminal glutathione S-transferase (GST) tag-fusion protein. After being purified by glutathione Sepharose 4B (GE Healthcare) affinity chromatography, GST–D14ΔN or GST–AtD14 protein was released by 10 mM glutathione (GSH) elution or on-column cleavage to remove the GST tag, then further purified by HiTrap Q (GE Healthcare) followed by Superdex 200 10/300 (GE Healthcare) in a buffer containing 10 mM Tris–HCl, pH 8.0, 150 mM NaCl, and 5 mM DTT.

Full-length A. thaliana SMXL6 was expressed in sf9 insect cells with an N-terminal Flag tag and purified by anti-Flag beads (Sigma, A2220) according to the manufacturer’s manual.

The full-length O. sativa D3 or A. thaliana MAX2 was fused with His₆ and co-expressed with ASK1, which stabilizes F-box proteins (Yan et al., 2013; Yao et al., 2016; Li et al., 2017), in sf9 insect cells. After purification by Ni-NTA (Novagen) affinity chromatography, the His-D3–ASK1 or His-MAX2–ASK1 complex was eluted and further purified by HiTrap Q followed by Superdex 200 10/300 in a buffer containing 20 mM MES, pH 6.5, 150 mM NaCl, and 5 mM DTT.

Pull-down assay

For the interaction between D14 and MAX2, ~20 μg of His-MAX2–ASK1 was used as the bait and ~12 μg of GST–D14 was used as the prey in the presence of 20 μM rac-GR24 or its solvent DMSO as the control. The reaction mixtures were incubated with Ni-NTA beads (Qiagen) at 4 °C for 1 h in the reaction buffer [50 mM Tris–HCl, pH 6.8, 100 mM NaCl, 25 mM imidazole, 10% (v/v) glycerol, 0.1% Tween-20, 20 mM 2-mercaptoethanol]. After washing six times with the reaction buffer, the protein complexes on the beads were released and then subjected to western blot analysis. The pull-down assay of AtD14 with MAX2 was performed in a similar way to serve as a positive control.

For the interaction between D14 and SMXL6, ~20 μg of Flag-SMXL6 protein was used as the bait and ~12 μg of GST–D14 as the prey in the presence of 20 μM rac-GR24 or its solvent DMSO as the control. The reaction mixtures were incubated with anti-flag beads at 4 °C for 1 h in the reaction buffer (50 mM Tris–HCl, pH 7.0, 150 mM NaCl, 0.5% Tween-20). After washing six times with the reaction buffer, the protein complexes on the beads were released and then subjected to western blot analysis. The pull-down assay of AtD14 with SMXL6 was performed in a similar way to serve as a positive control.

GST-fused proteins were detected by a monoclonal anti-GST antibody (Abmart). The polyvinylidene difluoride (PVDF) membrane was stained with Memstain (Applygen) to show equal loading.

Co-immunoprecipitation (Co-IP) assay in protoplasts

Protoplasts prepared from the smxl6 smxl7 smxl8 triple mutant or the wild type were transformed with transient expression plasmids as described (Wang et al., 2015). After transformation with the hemagglutinin (HA)-AtD14 and green fluorescent protein (GFP)–MAX2 plasmids and incubation at 21 °C for 11 h, protoplasts were pre-treated with 100 µM rac-GR24 for 1 h in W5 solution. Cells were then collected and broken in the protein extraction buffer [50 mM Tris–HCl, 150 mM NaCl, 10% (v/v) glycerol, 0.1% Nonidet P-40, and 1×complete protease inhibitor cocktail], and immunoprecipitation (IP) with agarose-conjugated anti-GFP monoclonal antibody (MBL) was subsequently carried out in the presence or absence of 100 µM rac-GR24 at 4 °C. The HA-AtD14 recombinant proteins were detected with the anti-HA monoclonal antibody (Millipore), and the GFP–MAX2 fusion proteins and GFP were detected with the anti-GFP monoclonal antibody (Sigma). The total proteins extracted from protoplasts before IP were used as inputs.

Size exclusion chromatography (SEC) assay

Purified D14 (~10 μM) and D3–ASK1 proteins (~5 μM) were incubated with 200 μM 5-deoxystrigol (5DS; OlChemIm Ltd) or an equal amount of DMSO as the solvent control at 25 °C for 1 h in buffer containing 20 mM MES, pH 6.5, 150 mM NaCl, 5 mM DTT. The reaction mixture was then injected onto a Superdex 200 10/300 column for analysis at a flow rate of 0.3 ml min^–1. The fractions (0.5 ml per fraction) were analyzed by SDS–PAGE and visualized by Coomassie Brilliant Blue staining.

Mass spectrometric analysis of covalent modification

The gel bands of D14 from the SEC-separated D14–D3–ASK1 complex induced by 5DS were excised for mass spectrometric analysis as previously described (Yao et al., 2016). Briefly, tandem mass spectrometry (MS/MS) spectra from each LC-MS/MS run were searched against the D14 protein database using the Proteome Discoverer (Version 1.4) searching algorithm. The search criteria were as follows: full enzymatic specificity for trypsin was required, two missed cleavages were allowed, carbamidomethylation was set as a fixed modification (on the cysteine residue), oxidation (on the methionine residue) was set as a variable modification, precursor ion mass tolerance was 10 ppm for all mass spectra acquired in the Orbitrap mass analyzer, and fragment ion mass tolerance was 0.02 Da for all MS/MS spectra acquired in the ion trap. A high confidence score filter [false discovery rate (FDR) <1%) was used to select the ‘hit’ peptides, and their corresponding MS/MS spectra were manually inspected.

Results

Phylogenetic analysis and sequence alignment of D14 orthologs from monocots and dicots

To investigate the evolutionary relationships among D14 orthologs in monocots and dicots, we searched public sequence databases using BLAST with the rice (O. sativa) D14 protein sequence as a query to obtain the predicted sequences of D14 orthologs from important monocots and dicots (Waters et al., 2012b; Conn et al., 2015; Bythell-Douglas et al., 2017) (Fig. 1A). The phylogenetic analysis showed that D14 proteins from the monocots such as O. sativa, Brachypodium distachyon, Triticum asetivum, Hordeum vulgare, Setaria italica, Zea mays, Sorghum bicolor, and Saccharum hybrid have closer phylogenic relationships, while D14 orthologs from all the tested dicots exhibit closer relationships (Fig. 1A).

Fig. 1.

Phylogenetic analysis and sequence alignment of D14 orthologs from monocots and dicots. (A) Phylogenetic analysis of D14 orthologs from monocots and dicots. The phylogenetic tree was generated with 23 full-length amino acid sequences of D14 orthologs using the Maximum Likelihood method based on the WAG model (100 replicates) in MEGA7 (Whelan and Goldman, 2001; Kumar et al., 2016). The percentage of trees in which the associated taxa clustered together is shown next to the branches. The tree is drawn to scale, with branch lengths measured in the number of substitutions per site. The GenBank accession numbers of D14 orthologs in the presented species are, from top to bottom: Arabidopsis thaliana D14 (NP_566220), Brassica rapa D14 (XP_009130408), Brassica oleracea D14 (XP_013638430), Brassica napus D14 (CDY42894), Punica granatum D14 (OWM70752), Nelumbo nucifera D14 (XP_010248100), Gossypium raimondii D14 (XP_012451974), Hevea brasiliensis D14 (XP_021646820), Populus trichocarpa D14 (XP_002302409), Glycine max D14 (XP_003557012), Medicago truncatula D14 (XP_003589086), Pisum sativum RMS3 (AMB61024), Nicotiana attenuata D14 (XP_019258478), Petunia hybrida DAD2 (AFR68698), Solanum lycopersicum D14 (XP_004238093), Saccharum hybrid D14 (AJY78078), Sorghum bicolor D14 (XP_002468316), Zea mays D14 (NP_001150635), Setaria italica D14 (XP_004985292), Hordeum vulgare D14 (AJP07999), Triticum asetivum D14 (AK332360), Brachypodium distachyon D14 (XP_003558555), and Oryza sativa D14 (XP_015631400). (B) Sequence alignment and structural annotation of D14 orthologs. ESPript was used to analyze the multiple sequence alignments generated by Clustal Omega (Sievers et al., 2011; Robert and Gouet, 2014) with the D14 orthologs listed in (A). Secondary structure elements of the rice D14 crystal structure (PDB code: 4IH9) are displayed on top of the alignments. Identical and conserved residues are highlighted by red and yellow backgrounds, respectively. The three catalytic residues, Ser, Asp, and His, are indicated by green stars.

Further sequence alignment and structural annotation showed that the examined D14 orthologs from various plant species all exhibit considerable identities at the amino acid level and have the same catalytic triad Ser–His–Asp and α/β hydrolase fold (Fig. 1B; Supplementary Table S1 at JXB online), suggesting the conserved physiological functions for different D14 proteins. Consistent with phylogenetic analysis (Fig. 1A), protein sequences of D14 orthologs from monocots or dicots, respectively, are more conserved. For example, the rice D14 exhibits no less than 80% identity with its orthologs from all the tested monocots, but displays only ~50% identity with those from the examined dicots (Supplementary Table S1). Interestingly, all the examined monocot D14 proteins (including rice D14) have an additional glycine- and serine-rich N-terminus (at least 7 glycine and 10 serine residues among the N-terminal 55 residues), which is absent in all the examined dicot D14 proteins (Fig. 1B). It is unclear whether full-length D14 proteins in monocots and dicots might have some divergent physiological functions due to their discrepant N-termini.

Rice D14 rescues phenotypes of the Arabidopsis d14 mutant and interacts with Arabidopsis MAX2 and SMXL6 proteins

We further investigated whether the physiological function of D14 proteins is conserved in the monocot rice and the dicot Arabidopsis. We generated the transgenic Arabidopsis 35Spro:D14 and 35Spro:D14ΔN by introducing full-length rice D14 or N-terminus (residues 1–51) truncated rice D14 (D14ΔN) under the control of the Cauliflower mosaic virus (CaMV) 35S promoter into the Arabidopsis Atd14-5 mutant, a weak allele of the Atd14 mutant (Yao et al., 2016). We also introduced the 35S promoter-driven AtD14 into Atd14-5 to generate the 35Spro:AtD14 plants for comparison. As shown in Supplementary Table S2, 66.1% of 35Spro:AtD14 plants (39 out of 59 transgenic lines) and 33.9% of 35Spro:D14ΔN plants (21 out of 62 lines) display a similar branching phenotype to the wild-type Col-0 (with ≤3 branches) while only 3 out of 52 (5.8%) 35Spro:D14 transgenic lines rescues the branching phenotype of Atd14-5 well. These genetic complementation results demonstrate that the highly branched phenotype of the Arabidopsis Atd14-5 mutant can be rescued by both the full-length rice D14 and the N-terminus-truncated rice D14 (D14ΔN), but the complementation ratio is very low for the case of full-length rice D14.

We further introduced D14ΔN under the control of the native AtD14 promoter into the T-DNA insertion knockout mutant Atd14-1 (Waters et al., 2012b) to generate the AtD14pro:D14ΔN plants. The results showed that 50% of the AtD14pro:D14ΔN plants exhibit a similar branching phenotype to the wild-type Col-0 (Fig. 2A, B; Supplementary Table S2), demonstrating that the highly branched phenotype of the Arabidopsis Atd14-1 mutant can be well rescued by rice D14ΔN (Fig. 2A–C).

Fig. 2.

Rice D14 rescues the Arabidopsis d14 mutant. (A and B) Rice D14 rescued the branching phenotype of Atd14-1. (A) Representative branching phenotypes of 7-week-old Col-0, Atd14-1, and four independent T₃ transgenic lines AtD14pro:D14ΔN L2 (line 2), L3, L5, and L6 in the Atd14-1 background; scale bars=1 cm. (B) Quantitative analysis of primary rosette branches of the indicated plants; data are means ±SD (n=20). Error bars indicate the SD; bars with the same letter are not significantly different from one another (ANOVA+Tukey HSD, P<0.01). (C) RT–PCR analysis of the AtD14 or D14 transcript levels in the indicated plants described in (A). The Arabidopsis ACTIN1 was used as an internal control. (D and E) Rice D14 rescued the hypocotyl phenotype of Atd14-1. (D) Representative hypocotyl phenotypes of 7-day-old Col-0, Atd14-1, and AtD14pro:D14ΔN L3 (T₃) seedlings; scale bars=5 mm. (E) Relative hypocotyl lengths of the indicated seedlings; data are means ±SD (n=30). Error bars indicate the SD; bars with the same letter are not significantly different from one another (ANOVA+Tukey HSD, P<0.01). (F and G) Rice D14 rescued the leaf phenotype of Atd14-1. (F) Representative leaf phenotypes of 4-week-old Col-0, Atd14-1, and AtD14pro:D14ΔN L3 (T₃); scale bars=1 cm. (G) Quantitative analysis on the leaf length/leaf width ratio for the sixth leaves of the indicated plants; data are means ±SD (n=20). Error bars indicate the SD; bars with the same letter are not significantly different from one another (ANOVA+Tukey HSD, P<0.01).

Moreover, we explored whether rice D14ΔN is able to complement Atd14-1 in other SL-regulated physiological phenotypes including hypocotyl elongation (Scaffidi et al., 2014; Umehara et al., 2015) and leaf morphology (Waters et al., 2012b). As shown in Fig. 2D–G, the Atd14-1 mutant shows impaired sensitivity to rac-GR24 treatment on hypocotyl inhibition and has rounder and broader leaves than the wild-type Col-0, which is consistent with previous observations (Waters et al., 2012b; Scaffidi et al., 2014; Umehara et al., 2015). However, the hypocotyl elongation of AtD14pro:D14ΔN plants (L3) was obviously inhibited when treated with rac-GR24, and the leaf phenotypes of AtD14pro:D14ΔN plants (L3) were also similar to those of Col-0 (Fig. 2D–G). These results demonstrate that rice D14 can rescue many SL-regulated physiological phenotypes of the Arabidopsis d14 mutant well.

Consistent with the functional complementation of Atd14 by rice D14 and D14ΔN, biochemical pull-down assays and structural analysis showed that, similar to AtD14, both rice D14 and D14ΔN are able to interact efficiently with Arabidopsis MAX2 and SMXL6 proteins in an SL-dependent manner (Fig. 3; Supplementary Fig. S1). These results reveal the molecular basis of the functional complementation by showing the conserved functions of D14 proteins at the protein–protein interaction level.

Fig. 3.

Rice D14 physically interacts with the Arabidopsis SL signaling components. (A) Rice D14 efficiently bound Arabidopsis MAX2 in the presence of rac-GR24. Pull-down assay using recombinant His₆-MAX2 and GST–D14 or GST–AtD14 in the absence or presence of rac-GR24. GST-fused proteins were detected by anti-GST antibody and the PVDF membrane was stained with MemStain to show equal loading. (B) Rice D14 efficiently bound Arabidopsis SMXL6 in the presence of rac-GR24. Pull-down assay using recombinant Flag-SMXL6 and GST–D14 or GST–AtD14 in the absence or presence of rac-GR24. GST-fused proteins were detected by anti-GST antibody and the PVDF membrane was stained with MemStain to show equal loading.

Taken together, our results demonstrate that rice D14 can rescue the phenotype of the Arabidopsis d14 mutant well, which is probably attributed to the conserved function of rice D14 to interact with Arabidopsis MAX2 and SMXL6 proteins in the presence of SLs.

The in vivo interaction of the SL receptor with its F-box protein does not require repressors

The interaction dynamics or sequential binding among SL signaling components remains an open question (Wang and Smith, 2016). It is known that the SL-induced interaction of receptor with repressor does not depend on the F-box protein MAX2 in vitro or in vivo (Wang et al., 2015). However, it is unclear whether or not the SL-induced in vivo interaction of the SL receptor with its F-box protein is independent of repressors. To answer this question, we employed our well-established Arabidopsis protoplasts transformation and Co-IP system to investigate the interaction between AtD14 and MAX2 in the wild-type Col-0 and the smxl6 smxl7 smxl8 triple mutant. We found that AtD14 was able to interact weakly with MAX2 in both the smxl6 smxl7 smxl8 triple mutant (Fig. 4A) and the wild-type Col-0 (Fig. 4B), and that such interactions in both the triple mutant and Col-0 were obviously enhanced by the addition of exogenous rac-GR24 (Fig. 4). These data demonstrate that the SL-induced in vivo interaction of AtD14 with MAX2 is independent of SMXLs. Together with previous studies (Wang et al., 2015), these results imply that the in vivo interactions between any two components among the receptor (AtD14 or D14), the F-box protein (MAX2 or D3), and the repressor (SMXLs or D53) do not require the presence of the third one.

Fig. 4.

AtD14 interacts with MAX2 independently of SMXL6/7/8. The in vivo interactions between HA-AtD14 and GFP–MAX2 revealed by co-immunoprecipitation (Co-IP) assay in protoplasts prepared from the smxl6 smxl7 smxl8 triple mutant (A) and the wild-type Col-0 (B). After transformation and incubation for 11 h, protoplasts were pre-treated with rac-GR24 for 1 h, cells were broken, and then immunoprecipitation (IP) using agarose-conjugated anti-GFP monoclonal antibody was carried out in the presence or absence of 100 µM rac-GR24. The HA-AtD14 recombinant protein was detected by anti-HA monoclonal antibody, and the GFP–MAX2 fusion protein and GFP were detected by anti-GFP monoclonal antibody. Input means extracted crude proteins without immunoprecipitation.

Rice D14 is an unconventional hormone receptor for SLs

To investigate further whether rice D14, similar to AtD14, generates and covalently binds the active SL molecule CLIM, we employed the SEC approach to prepare the 5DS-induced D14–D3–ASK1 complex for MS/MS analysis. The D14–D3–ASK1 complex was eluted earlier (fraction peak ~13.4 ml) (Fig. 5A, upper panel), and then subjected to SDS–PAGE to separate D14 protein (Fig. 5A, lower panel) for further trypsin digestion followed by MS/MS analysis. Peptide matching from MS/MS spectra identified a chemically modified peptide (287-TTVEFLQTEGHLPHLSAPSLLAQVLR-312) of D14 with a molecular weight shift of 96.0211 Da on the catalytic residue H297 (Fig. 5B), which is identical to the accessional molecular weight on the corresponding histidine residue (H247) of AtD14 (Yao et al., 2016). As the control, no modified peptide was identified when D14 without 5DS treatment was subjected to MS/MS analysis (Supplementary Fig. S2).

Fig. 5.

Rice D14 generates and covalently binds the active form of SLs. (A) 5DS induced the interaction of rice D14 and D3 in the SEC assay. Upper panel: SEC analysis of the interaction between D14 and D3–ASK1 in the presence of 5DS; the elution volumes of the molecular weight markers are indicated above the peaks. Lower panel: SDS–PAGE analysis of peak fractions from the upper panel; M, molecular weight ruler (kDa). (B) Rice D14 hydrolyzed 5DS and generated the C₅H₅O₂ modification on the catalytic residue H297. A quadruply charged peptide (287-TTVEFLQTEGHLPHLSAPSLLAQVLR-312) of D14 with the 5DS-derived C₅H₅O₂ modification on H297 was identified by MS/MS (m/z=739.40202). The modified peptide was isolated from the trypsin digestion products of D14 in the 5DS-induced D14–D3–ASK1 complex collected in SEC (A). Labeled peaks correspond to masses of y and b ions of the peptide displayed on the top, respectively. The asterisked ‘H’ indicates the modified H297.

Taking advantage of the comprehensive analyses on D14-mediated SL perception in our recent work (Yao et al., 2016), we are able to deduce that this molecular weight shift of 96.0211 Da (Fig. 5B) corresponds to the chemical formula C₅H₄O₂, which indicates covalent C₅H₅O₂ modification on H297 of D14 (compound 5, Fig. 6A) and denotes the presence of D-ring-derived intermediate compound 4 (CLIM; Fig. 6A) as the active form of SLs in the SL-induced D14–D3 complex. Moreover, the same C₅H₅O₂ modification on rice D14 was also detected in planta when 35Spro:D14ΔN plants were treated with 5DS (Supplementary Fig. S3).

Fig. 6.

Proposed mechanism of SL perception by rice D14. (A) Schematic diagram of a proposed rice D14-mediated hydrolysis process of 5DS. The hydrolysis of 5DS (1) is proposed to involve a nucleophilic attack by S147, which produces ABC-OH (2) and compound 3, and the generation of D-OH (6). The N^ε2 atom of H297 attacks the aldehyde carbon atom of the S147-linked compound 3 to form the H297- and S147-linked linear compound 4, referred to as the covalently linked intermediate molecule (CLIM). Compound 4 initiates an intramolecular nucleophilic attack to generate the H297-linked circular compound 5, which appears as a C₅H₅O₂ modification on H297 detected by MS/MS and denotes the existence of CLIM inside the D3-bound D14. Compound 5 would be further hydrolyzed from H297 to produce D-OH (6). A similar deduction of the AtD14-mediated (+)-GR24 hydrolysis process can be found in detail in our recent study (Yao et al., 2016). (B) A simplified model of SL perception. D14/AtD14 docks SL in the catalytic cavity, hydrolyzes SL into a D-ring-derived intermediate (CLIM), which is covalently sealed inside the catalytic center of D14/AtD14 to promote the interaction with the D3/MAX2-based SCF complex and the repressor D53/SMXLs for triggering SL-regulated plant branching (Yao et al., 2016).

Together with previous studies about D14 function in rice (Arite et al., 2009; Gao et al., 2009; Liu et al., 2009; Jiang et al., 2013; Zhou et al., 2013), our results collectively uncover the conserved function of D14 proteins in the monocot rice and the dicot Arabidopsis, and suggest that rice D14 acts as an unconventional hormone receptor to generate and perceive the active form of SLs.

Discussion

Understanding of hormone perception is central to comprehending hormone action. Biologists over the past century have established a general perception mechanism for phytohormones: receptors specifically and reversibly bind their ligands with high affinity to initiate hormone signaling, and eventually release the unchanged ligands for the next round of perception. However, recent works on SL perception in dicots (de Saint Germain et al., 2016; Yao et al., 2016, 2017) have defined Arabidopsis D14, pea RMS3, and Striga HTL7 as unconventional receptors that hydrolyze SLs into the active form of hormone (CLIM), covalently bind CLIM to trigger SL signaling, and ultimately release an inactive hydrolysis product D-OH. Here, our data suggest that D14 in the monocot rice possesses the same physiological functions as AtD14, and also acts as an unconventional hormone receptor to generate and perceive CLIM, and expectedly undergo conformational changes (Hamiaux et al., 2012; Zhao et al., 2015; Yao et al., 2016) for recruitment of signaling components (such as D3 and D53) (Jiang et al., 2013; Zhou et al., 2013), thereby triggering SL signal transduction (Fig. 6B).

It is intriguing that a glycine- and serine-rich N-terminus is present in all the examined monocot D14s but absent in those of all the tested dicots (Fig. 1), and the N-terminus-truncated rice D14 showed a much higher complementation ratio than the full-length rice D14 when expressed in the Arabidopsis mutant Atd14-5 (Supplementary Table S2). However, the underlying molecular mechanism remains to be investigated in the future. Such a glycine- and serine-rich N-terminus does not affect the D14 interaction with MAX2 or SMXL6 (Supplementary Fig. S1), and is possibly structurally flexible (Kagiyama et al., 2013). Given that D14 can be transported via phloem in rice (Kameoka et al., 2016), it would be interesting to investigate whether the additional glycine- and serine-rich N-terminal sequence of rice D14 functions as a signal peptide for D14 localization and/or transport in vivo.

Divergent features of the SL signaling pathway in monocotyledonous and dicotyledonous species are also found in the downstream signal transduction process. In the dicots Arabidopsis and pea, the gene BRANCHED1 (BRC1), which encodes a TCP transcription factor, has been demonstrated to be a key SL-responsive gene in the downstream SL signaling pathway (Aguilar-Martínez et al., 2007; Póza-Carrion et al., 2007; Mashiguchi et al., 2009; González-Grandío et al., 2013; Wang et al., 2015). However, in the monocots rice and maize, the BRC1 ortholog TEOSINTE BRANCHED1 (TB1) was not up-regulated by GR24 treatment (Minakuchi et al., 2010; Guan et al., 2012). Moreover, a recent study identified rice IPA1 (OsSPL14), a member of the SQUAMOSA PROMOTER BINDING PROTEIN-LIKE (SPL) transcription factor family (Jiao et al., 2010; Miura et al., 2010), as a direct target of the repressor D53 to participate in the SL-mediated regulation of rice tillering (Song et al., 2017). The loss-of-function mutant of IPA1 is insensitive to GR24 treatment and shows more tillers than wild-type rice plants (Song et al., 2017). Similar regulation of Triticum aestivum (Ta)D53 on TaSPL13/17 from bread wheat was also observed recently (Liu et al., 2017). However, the Arabidopsis mutant containing loss-of-function mutations in both SPL9 and SPL15 (the orthologs of the IPA1/OsSPL14 gene) still responds to the SL analog GR24 and shows reduced branching to a level similar to that of the wild type (Bennett et al., 2016). These phenomena suggest that the downstream SL signaling pathway seems not to be fully conserved between monocots and dicots. Further study is needed to better understand the elusive downstream SL signaling pathway in various plant species.

Supplementary data

Supplementary data are available at JXB online.

Fig. S1. The N-terminus of rice D14 does not affect the interaction with Arabidopsis components MAX2 or SMXL6.

Fig. S2. No modified peptide was identified when D14 without 5DS treatment was subjected to MS/MS analysis.

Fig. S3. SL can generate the C₅H₅O₂ modification of rice D14 in planta.

Table S1. Protein identities between D14 orthologs and D14 or AtD14.

Table S2. Branch numbers of Atd14 transgenic plants with full-length or truncated rice D14.

Glossary

Abbreviations:

BRC

BRANCHED

CLIM

covalently linked intermediate molecule

D

DWARF

DAD

DECREASED APICAL DOMINANCE

GFP

green fluorescent protein

IP

immunoprecipitation

IPA

Ideal Plant Architecture

MAX

MORE AXILLARY GROWTH

MS/MS

tandem mass spectrometry

RMS

RAMOSUS

SCF

Skp–Cullin–F-box

SEC

size exclusion chromatography

SL

strigolactone

SMXL

SUPPRESSOR OF MAX2 1-LIKE

SPL

SQUAMOSA PROMOTER BINDING PROTEIN-LIKE

TB

TEOSINTE BRANCHED.