Plant Mol Biol DOI 10.1007/s11103-015-0410-2

Quantitative phosphoproteomic analysis of early seed development in rice (Oryza sativa L.) Jiehua Qiu1 • Yuxuan Hou1 • Xiaohong Tong1 • Yifeng Wang1 • Haiyan Lin1 Qing Liu1 • Wen Zhang1 • Zhiyong Li1 • Babi R. Nallamilli2 • Jian Zhang1

•

Received: 13 April 2015 / Accepted: 23 November 2015 Ó Springer Science+Business Media Dordrecht 2015

Abstract Rice (Oryza sativa L.) seed serves as a major food source for over half of the global population. Though it has been long recognized that phosphorylation plays an essential role in rice seed development, the phosphorylation events and dynamics in this process remain largely unknown so far. Here, we report the first large scale identification of rice seed phosphoproteins and phosphosites by using a quantitative phosphoproteomic approach. Thorough proteomic studies in pistils and seeds at 3, 7 days after pollination resulted in the successful identification of 3885, 4313 and 4135 phosphopeptides respectively. A total of 2487 proteins were differentially phosphorylated among the three stages, including Kip related protein 1, Rice basic leucine zipper factor 1, Rice prolamin box binding factor and numerous other master regulators of rice seed development. Moreover, differentially phosphorylated proteins may be extensively involved in the biosynthesis and signaling pathways of phytohormones such as auxin, gibberellin, abscisic acid and brassinosteroid. Our results strongly indicated that protein phosphorylation is a key mechanism regulating cell proliferation and enlargement, phytohormone biosynthesis and signaling, grain filling and

grain quality during rice seed development. Overall, the current study enhanced our understanding of the rice phosphoproteome and shed novel insight into the regulatory mechanism of rice seed development. Keywords Rice (Oryza sativa L.)
Phosphorylation
Proteome
Seed development Abbreviations DAP Days after pollination DP Differentially phosphorylated FDR False discovery rate MASS Mass spectrum MAPK Mitogen-activated protein kinase MLPK M-locus protein kinase PTM Post-translational modification SI Self-incompatibility SRK S-locus receptor kinase BR Brassinosteroid IAA Indole-3-acetic acid TCA Trichloroacetic acid

Jiehua Qiu and Yuxuan Hou have contributed equally to this paper.

Electronic supplementary material The online version of this article (doi:10.1007/s11103-015-0410-2) contains supplementary material, which is available to authorized users. & Jian Zhang [email protected] 1

State Key Lab of Rice Biology, China National Rice Research Institute, Hangzhou 311400, People’s Republic of China

2

Department of Human Genetics, Emory University School of Medicine, Atlanta, GA 30322, USA

Introduction Rice (Oryza sativa L.) is one of the most important cereal crops in the world as it serves as a major food source for more than half of the global population. Besides its economic importance, rice is also an ideal model plant in biological research due to its mature genetic transformation system, relatively small genome size, released genome sequences, available genetic resources and genome colinearity to other grass species (Zhang et al. 2007).

123

Plant Mol Biol

Rice seed is comprised of a maternal caryopsis coat, a diploid embryo and a triploid endosperm, in which most of the nutrients are stored in the form of starch, protein, lipids and other trace substances. Like in many other angiosperms, double fertilization is a very critical process which initiates seed development in rice (Bleckmann et al. 2014). After adhering to the stigmas, the pollen grain releases the two sperms into the embryo sac. In the embryo sac, one sperm fertilizes the egg cell to form the diploid embryo, while the other sperm fuses with the two haploid polar nuclei of the large central cell to develop into the triploid endosperm. Embryo development starts at about 10 h after the double fertilization, and grows from one cell to several hundred at around 3 days after pollination (DAP) as a consequence of rapid cell proliferation. From 3 DAP to 7 DAP; the embryo undergoes the process of apical meristem and leaf primordium differentiations. After 10 DAP, the rice embryo enters the late embryogenesis stage in which the embryo becomes mature and dormant. In contrast to the embryo development, the endosperm develops in a much faster manner as the primary endosperm nuclei division starts immediately after double fertilization. At 3 DAP, the endosperm nuclei stop division and become cellularized with their cell walls. After this, the new endosperm cells continue dividing to fill the embryo sac cavity. Accompanied with the division, DNA endoreduplication elevates the nuclear DNA content to 12–24C. At around 7 DAP, the outer most endosperm cells gradually differentiate into aleuronic cells, while the inner cells become starch storage cells. Afterwards, endosperm cells continue to grow by cell enlargement with nutrient deposition, and finally become mature and desiccated after 20 DAP. Morphologically, rice seed reaches its full length and full size (width and thickness) at approximately 7 DAP and 15 DAP, respectively. Given the importance of rice as major food resource, rice seed development has been one of the focuses of the plant science community. In the past decades, several agronomically important genes were reported to be related with rice embryo development (Nagasaki et al. 2007), endosperm quality, nutrient accumulation (Nallamilli et al. 2013; Zhang et al. 2010) and seed size (Rao et al. 2014). Novel OMIC methods like transcriptomic, epigenomic and proteomic techniques were also applied to systematical investigations related to the regulatory network of rice seed development in various hierarchies (Deng et al. 2013; Lan et al. 2012; Malone et al. 2011; Xue et al. 2012; Zemach et al. 2010). However, these reports majorly focused on the profiling or quantity change of RNAs or proteins. The posttranslational modifications (PTMs) of the pre-existing proteins, which is another important mechanism regulating protein activity has been barely studied. Previous studies have revealed over three hundred types of PTM including phosphorylation, acetylation, nitrosylation, glycosylation

123

etc., among which phosphorylation is the most well documented due to its rich abundance, relatively high stability and functional importance in plant development. Phosphorylation is a reversible, covalent modification majorly occurring on the hydroxyl group of hydroxyl amino acids like serine, threonine and tyrosine (Reinders and Sickmann 2005). The addition or removal of the phosphate group on specific sites of a protein could alter its enzyme activity, substrate specificity, structure stability or intracellular localization, thus to confer the modified proteins with new functions. To date, most of the reported signaling pathways are mediated by phosphorylation/dephosphorylation of the proteins in the cascade, which could play essential roles in plant pollination and seed development (de la Fuente van Bentem and Hirt 2007; Schmelzle and White 2006). In the early 1990s, scientists found that the application of protein phosphatase inhibitors could destroy the stigma receptivity to pollination (Cabrillac et al. 2001; Kandasamy et al. 1993; Rundle et al. 1993). The self-incompatibility (SI) response of pollen-pistil interaction in Brassicaceae is suggested to be based on phosphorylation cascade signaling and ubiquitin-mediated degradation pathway (Tantikanjana et al. 2010). It was clearly indicated that the activation of the SRK (S-locus receptor kinase) by autophosphorylation is the first step that triggers the SI signaling (Kachroo et al. 2001; Takayama et al. 2001). The phosphorylated SRK combines with another kinase MLPK (M-locus protein kinase), directly phosphorylates, and turns on the Armadillo Repeat-Containing protein 1 ubiquitin ligase, which mediates the degradation of EXO70A1 and other ‘compatibility’ factors, and eventually leads to self-pollen rejection (Gu et al. 1998; Kakita et al. 2007; Murase et al. 2004; Stone et al. 1999, 2003). Similar cases were also reported in Antirrhinum hispanicum and Papaver rhoeas (Lai et al. 2002; Wheeler et al. 2009). Several Mitogen-activated protein kinase (MAPK) cascade members were reported to be involved in seed development as well (Xu and Zhang 2015). In Arabidopsis, MPK6 T-DNA mutants produced abnormal anther and seeds (Bush and Krysan 2007). Knock-out of OsMKK4/SMG1, a MAPKK in rice, significantly influenced the grain size possibly by regulating brassinosteroid (BR) responses (Duan et al. 2014). In addition, some other kinase genes were found to be required for endoreduplication and aleurone layer formation in rice (Barroco et al. 2006; Pu et al. 2012). Despite the realization of the importance of phosphorylation in rice seed development, our current knowledge in this field is very limited due to the fact that most of the studies were carried out on a single phosphoprotein using traditional genetic and biochemical analysis or a few phosphoproteins by using low throughput proteomic methods like 2D gel resolving associated with specific dye staining (Chen et al. 2011; Zhang et al. 2014b). Recently, novel high throughput

Plant Mol Biol

phosphoproteomic technologies provide great opportunities to identify phosphoproteins and phosphosites in a large scale. A study by Nakagami et al. (2010) identified 6919 phosphopeptides in rice whole-cell lysates. Most of these phosphosites are conserved among plant species. A quantitative phosphoproteomic analysis of rice seed early germination revealed that 149 phosphoproteins including three core BR signal transduction components were differentially phosphorylated, suggesting that BR triggers rice seed germination (Han et al. 2014). Nevertheless, no reports regarding the phosphorylation regulatory network in rice seed development are available thus far. Until very recently, a phosphoproteomic analysis identified over 2300 phosphosites corresponding to 1588 phosphoproteins in the rice pistil (Wang et al. 2014a). However, this work only qualitatively profiled the phosphosites in the pistil, key information to understand the phosphorylation regulatory roles in rice development, such as the dynamics of phosphoproteins, phosphosites and phosphorylation intensities during seed development remain a mystery. Here, we report the first large scale, comprehensive analysis of the phosphorylation events during early rice seed development through a non-gel, quantitative phosphoproteomic approach. 2487 proteins were found to be differentially phosphorylated including kinases, transcription factors, epigenetic controlling factors involved in hormone regulation, cell division, cell size determination, grain filling and quality control. The results obtained from this work not only provided a large number of phosphorylation proteins and site information for rice, but also shed light on the regulatory roles of protein phosphorylation in rice seed development.

complex, and 1 mM phenylmethylsulfonyl fluoride) for 30 min at 4 °C, and sheared by sonication. After centrifugation, the supernatant was transferred to a new tube and proteins were precipitated in 100 % acetone. The protein pellets were washed in 75 % ethanol and dissolved in the lysis buffer. Qubit 2.0 fluorometer (Invitrogen) was used for the quantification of the extracted total proteins. Western blot analysis In brief, 20 lg of the extracted total proteins were resolved on 10 % SDS–polyacrylamide gels, and subsequently transferred onto a polyvinylidene fluoride fluoropolymer (PVDF) membrane using an electrophoretic blotting system (BioRad). Then, primary antibodies (1:1000 dilution) were used to detect corresponding proteins, a HRP conjugated goat anti rabbit IgG was used as secondary antibody (1:10,000 dilution). Lastly, the enhanced chemiluminescence (Pierce) method was used for signal visualization. The antibodies against D1, KRP1, OsABF1, PPKL2, RISBZ1 and SMG1 were commercially synthesized (Genescript), and anti-actin (cat No. M20009) was purchased from Abmart Company. Quantification of the band intensities on the immunoblots was performed using the ImageJ software according to the instructions (http://rsb.info.nih.gov/ij/docs/menus/analyze. html#gels). All the sample intensities were first normalized to the control actin, and then calculated based on the ratio to set the expression level of S0 into 1. Three technical repeats were conducted. Protein digestion

Materials and methods Plant growth conditions and sample collection Rice japonica subspecies variety Nipponbare plants were grown in the field of China National Rice Research Institute (CNRRI) in the summer of 2014. To prevent contamination from pollinated pistils, the mature pistils were manually dissected and collected before the panicles fully headed out. For the developing seeds, each panicle was labeled on the day of anthesis, and seeds were manually collected with glumes removed at 3 DAP (around 6 mm in length) and 7 DAP (around 10 mm in length) respectively. Three biological replicates of pistil, 3 DAP seeds and 7 DAP seeds were harvested and immediately stored in liquid nitrogen until use. Total protein extraction Briefly, 1 gram of each sample was ground into fine powders in liquid nitrogen, shook in 5 ml lysis buffer (150 mM Tris– HCl pH 8.0, 8 M urea, 19 phosphoprotein protease inhibitor

Protein was reduced with 5 mM DTT at 56 °C for 30 min, cooled to room temperature, alkylated with 20 mM IAA in dark for 30 min, and then 5 mM DTT was added in the dark for 15 min. The reduced and alkylated proteins were digested on the 30 kDa filter unit (millipore) over night with trypsin at pH 8.0 (enzyme:protein = 1:50). Peptides obtained by filter-aided sample preparation (FASP) were desalted using C18 Sep-Pak (Waters). Phosphopeptide enrichment The digested peptides were resolved with binding buffer (80 % ACN, 5 % TFA, 1 M lac acid), then incubated with TiO2 beads (GL sciences, peptide:TiO2 = 1:4) three times, each time for 30 min and then washed with binding buffer twice. Then, the TiO2 beads were transferred into a 200 mL homemade StageTip with two pieces of C18 solid phase extraction disk (3 M). The enriched phosphopeptides were washed with elution buffer (40 % ACN, 15 % NH3H2O) 4 times. Eluates were subsequently dried to *5 lL

123

Plant Mol Biol

in a SpeedVac and reconstituted with 5 % MeOH in 1 % TFA solution for LC–MS/MS analysis.

protein signals were observed under a confocal microscope (Leica).

LC–MS/MS and data analysis

Results Firstly, peptides were loaded onto a homemade reversedphase column (75 lmID 9 15 CM), separated over 1 h using a linear 5–30 % acetonitrile gradient at a flow rate of 300 nL/min on an Easy-nLC1000 liquid chromatography system (Thermo). Full-scan mass spectra were acquired by using a Q Exactive Plus spectrometer (Thermo) over a mass range of 300–1400 m/z with a resolution of 70,000. Subsequently, raw spectral data were processed for phosphopeptide identification and phosphosite quantification with Mascot search engine against the rice genome annotation project database: (ftp://plantbiology.msu.edu/pub/ data/Eukaryotic_Projects/o_sativa/annotation_dbs/pseudo molecules/version_7.0/all.dir/). In the Mascot searches for phosphopeptides, Oxidation (M), Acetyl (Protein-N term) and Phospho(S/T/Y) was set as variable modifications, and one missed cleavage on trypsin was allowed. Peptide mass tolerance was set at 20 ppm for precursor and 50 mmu for the tolerance of product ions. Mascot results were filtered with the Percolator tool of the Protein Discoverer package with false discovery rates (FDR) set to \1 % for peptide identification of all searches. For reliable phosphorylation site analysis, all phosphopeptide hits were automatically re-analyzed by the phosphoRS software within the Protein Discoverer software suite. PhosphoRS probability higher than 90 % was required for a phosphorylation site to be considered as localized. Only those peptides which were phosphorylated in at least two of the three biological replicates were considered as truly phosphorylated. The differentially phosphorylated protein was defined to have an over twofold change in the normalized average intensity with the credible student’s t test (P \ 0.05).

Identification of phosphorylation sites, peptides and proteins Protein phosphorylation, an important posttranslational modification, plays vital roles in complex seed development and maturation. To dissect the phosphoproteins involved in the rice seed developmental regulatory mechanisms, a quantitative, non-gel, label-free phosphoproteomic study was performed for the unpollinated pistil (S0), seed at 3 DAP (Day After Pollination) (S3) and 7 DAP (S7) of Japonica rice variety Nipponbare, which represented the pre-pollination, post-pollination, fast proliferation and differentiation status of rice seed, respectively (Fig. 1). We did not include the developing seeds in later stages in this study because no ideal techniques were available so far to remove the rich amount of polysaccharide, lipids and storage proteins, which could interfere with the outcome of MS identification of the regulatory proteins. Phosphopeptides were enriched from total seed proteins by TiO2-MOAC method and analyzed by LC–MS/MS assay. In this study, a total of 3885, 4313 and 4135 phosphopeptides were identified covering 4120, 4611 and 4135 phosphosites in S0, S3 and S7 samples respectively. For all the three stages, the majority of the phosphopeptides carried only one phosphorylation modification, while around

Subcellular localization analysis To generate the rice protoplast, around 5 grams rice leaf strips in 0.5 mm size were digested in 10 mL enzyme solution (1.5 % cellulose R10, 0.75 % macerozyme R10, 0.6 M mannitol, 10 mM MES pH = 7.5) for 6 h in dark with gentle shaking (40 rpm) at 28 °C. The protoplasts were filtered and harvested by centrifugation, then washed with 10 mL ice cold W5 solution (154 mM NaCl, 125 mM CaCl2, 2 mM KH2PO4, 2 mM MES, 5 mM glucose, pH = 5.7) two times, and finally suspended in 500lL MMG solution (0.4 M mannitol, 15 mM MgCl2, 4 mM MES, pH = 5.8). The 35S:KRP1-GFP and 35S:D53mKate plasmid DNA were transformed into protoplast by incubating in PEG (0.6 M mannitol, 100 mM CaCl2, 40 % PEG4000) for 30 min at room temperature. The fluorescent

123

Fig. 1 Seed morphology of the samples tested in this study. Scale bar 2 mm. The sample order from left to right: S0, S3 and S7

Plant Mol Biol

6 % of the phosphopeptides carried two phosphorylation groups, three or more phosphorylation modifications were scarcely detected on one phosphopeptide in this study (Fig. 2a). In S0, there were 3743 phosphoserine, 369 phosphothreonine and 8 phosphotyrosine, representing a percentage of 90.8, 9.0 and 0.2 of the total phosphosites respectively. The phosphosite types of S3 and S7 also shared a similar distribution pattern with S0 (Fig. 2b). Interestingly, even though different tissues and plant species were studied, the ratios of phosphorylation which occurred on serine, threonine and tyrosine observed in this study are very clsoe with the data of previous reports (Han et al. 2014; Lv et al. 2014; Wang et al. 2014a; Zhang et al. 2014a), indicating a conserved phosphosite distribution pattern among plants. A database search finally corresponded the S0, S3 and S7 phosphopeptides to 1999, 2143 and 2099 phosphoproteins respectively. By using a combination of TCA-acetone and phenol extraction methods, Wang et al. (2014a) identified 3143 phosphopeptides corresponding to 1588 phosphoproteins from rice pistil, which is essentially the S0 samples used in this study. A comparable number of 3885 phosphopeptides and 1999 phosphoproteins identified from our S0 samples strongly indicated the successful performance of our experiments. On the other hand, we noticed that only 888 (44.4 %) of S0 phosphoproteins matched with Wang et al’s pistil results, possibly due to the difference of protein extraction methods

and data analysis procedures (Supplementary Table 1). Therefore, the urea method used in this study would be an effective complementation of the TCA and phenol methods to achieve wider proteome coverage. Subcellular localization of phosphoproteins Eukaryotic cells are comprised of several membrane-bound subcellular compartments such as nucleus, cytoplasm, mitochondria etc. After translation, proteins are sorted and transported to different subcellular compartments where functional proteins play distinct roles. To analyze the putative subcellular localization, the sequences of S0, S3 and S7 phosphoproteins were used to search against the ‘‘Eukaryotes’’ database of CELLO using the default setting (http://cello.life.nctu.edu.tw/) (Yu et al. 2004, 2006). As shown in Fig. 2c–e, phosphoproteins from the three stages showed similar cellular compartment distributions. The nuclear phosphoproteins represented more than 60 % of the input phosphoproteins. Moreover, there were around 14, 11 and 7 % of the phosphoproteins assigned to cytoplasm, plasma membrane and chloroplast, respectively. However, other compartments, such as mitochondrial, golgi and ER, represented less than 10 % of the phosphoproteins in total. To verify the subcellular localization prediction result, we fused KRP1 (LOC_Os02g52480) with GFP protein, and co-transformed the construct into rice protoplast with a

Fig. 2 a The counts of phosphopeptides carrying single, double and triple phosphorylation modifications in S0, S3 and S7. b The counts of phosphosites in serine, threonine and tyrosine in S0, S3 and S7. c–e Distribution of the S0, S3 and S7 phosphoprotein subcellular localization

123

Plant Mol Biol

Fig. 3 Motif-X analysis of the over-represented motifs around the phosphosites of the identified rice seed phosphoproteins. a [sP], b [LxRxxs], c [Rxxs], d [Lxxxxs], e [sF], f [RsxS], g [sS], h [sDxE], i [tP], j [Pt]

nuclear marker 35S:D53-mKate (Jiang et al. 2013; Zhou et al. 2013). As shown in Supplementary Fig. 1, KRP1 colocalized with D53 in the nucleus, indicating KRP1 is a nuclear protein which is consistent with the CELLO prediction result. In addition, several reported cases in rice also strongly supported the prediction result of CELLO, suggesting the subcellular localization result in this study is highly reliable (Yu et al. 2015). Interestingly, it seems that the phosphoprotein compartment distribution patterns varied from species, tissues, protein extraction methods and prediction tools as divergent results were reported by different plant phosphoproteomic studies (Fan et al. 2014; Han et al. 2014; Wang et al. 2014a). Conserved phosphorylation motif analysis of the phosphosites The specific amino acid (AA) sequence features or motifs around the phosphosites usually determine the kinase-substrate specificity. In this study, Motif-X tool (http://motif-x.

123

med.harvard.edu/motif-x.html) (Chou and Schwartz 2011) was used to detect the over-represented motifs around the identified phosphosites. After combining all the phosphosites from the three stages and removing the redundant sites, we obtained totally 5412 unique phosphosites for the motifX analysis. With the aid of bioinformatics tools, a phosphosite centered, 13 AA in length sequence was extracted from each phosphosite, and totally 4885 unique amino acid sequence extractions, including 4409 centered by phosphoserine, 466 centered by phosphothreonine and 10 tyrosine-centered phosphopeptides, were applied for overrepresented motif search. As a result, at least eight overrepresented motifs were detected from the phosphoserine sites (Fig. 3). [sP] and [sS] were the two most enriched motifs with 2365 and 1469 matches respectively (Fig. 3a, b). Motif [Rxxs], [Lxxxxs] and [sF] also appeared very frequently (Fig. 3c–e). In addition, we identified over one hundred [RsxS], [LxRxxs] and [sDxE] motifs (Fig. 3f–h). [tP] and [Pt] were the only two conserved motifs found around the phosphothreonine sites (Fig. 3i, j).

Plant Mol Biol

Unfortunately, we failed to find any over-represented motifs from the phosphotyrosine sites due to the fact that only ten sequences were extracted in this study. Previous metaanalysis of phosphoproteomic data has associated several enriched motifs with certain kinase substrates (van Wijk et al. 2014). For example, [sP], an extremely common proline-directed motif, was over-represented in various plant species including Arabidopsis, rice and wheat (van Wijk et al. 2014; Wang et al. 2014a; Zhang et al. 2014a). Most of the [sP] motif-containing proteins were found to be located in the nucleus and cytosol, and could be potential substrates for MPK, SnRK2, RLK, AGC, CDK, CDPK and SLK kinases (van Wijk et al. 2014). [Rxxs] motif is another common plant protein motif which could be recognized by MAPKK, CaMK-II and protein kinase A or C (van Wijk et al. 2014; Zhang et al. 2014a). In agreement with the previous reports, this study found that several important rice regulators were phosphorylated at [sP] or [Rxxs] motif, including D1, DGL1, FLO2, RPBF, TRAB1 etc. (Table 1), suggesting these proteins are potential substrates of the corresponding kinases. And indeed, TRAB1/bZIP66 has been proven to be phosphorylated by a rice SnRK2 kinase SAPK10 (Kobayashi et al. 2005). In addition to TRAB1, we also found that 7 other bZIP TFs were phosphorylated at [sP] motif (Supplementary Table 1). Considering that bZIP is a major group of ARFs (ABA responsive factors), these bZIP TFs are very likely to be the substrates of rice SnRK2 s whose function is to phosphorylate ARFs in the ABA signaling (Cutler et al. 2010). Acidic SD type motif [sDxE] is much more conserved in SDPK substrate proteins. By far, [tP] is the most common phosphothreonine motif found in plants (van Wijk et al. 2014). Though the [sS], [Rxxs], [Lxxxxs], [sF] and [Pt] were predominantly over-represented in our case, another study annotated them as low-frequency motifs without pointing to any functions (van Wijk et al. 2014). Taken together, the conserved motifs identified in this study are not only useful in phosphosite prediction for unknown phosphoproteins, but also provide hints on how the phosphorylated proteins are associated with their kinases. Differentially phosphorylated (DP) peptides and proteins during early seed development After the quantitative normalization of the phosphorylation intensity of three biological replicates, 4803 peptides were found to be differentially phosphorylated with twofold change or more among the S0, S3 and S7 (P \ 0.05). 1802 peptides were phosphorylated in all three samples but with significantly different intensities. 477, 523 and 673 peptides were specifically phosphorylated in S0, S3 and S7, respectively. Meanwhile, 588, 206 and 534 were specifically not phosphorylated in S0, S3 and S7, respectively

(Fig. 4a). Due to the condition that multiple phosphopeptides could correspond to one phosphoprotein, we totally identified 2487 DP proteins from the three stages, including 286, 329 and 581 phosphoproteins that were specifically phosphorylated, and 341, 109 and 287 that were specifically not phosphorylated in S0, S3 and S7 respectively. 554 proteins showed a constitutive phosphorylation pattern but with various intensity among the three stages (Fig. 4b). These DP proteins covered 166 transcription factors, 89 epigenetic controlling factors and 209 kinases, indicating the complexity of the phosphorylation regulation in seed (Supplementary Table 1). Based on the protein phosphorylation dynamic tendency, the 2487 DP proteins could be divided into 5 clusters, implying their functions in different stages of rice seed development (Fig. 4c). KEGG pathway analysis in DAVID (http://david.ncifcrf.gov/) (Dennis et al. 2003) revealed that DP proteins are majorly over-represented in the pathways of spliceosome and basal transcription, indicating protein phosphorylation plays key roles in the mRNA transcription and processing during seed development (Supplementary Fig. 2). Gene ontology of DP proteins A gene ontology analysis of DP proteins was conducted in the vocabulary of ‘‘cellular component,’’ ‘‘biological process’’ and ‘‘molecular function’’ (Fig. 5a). From the perspective of ‘‘biological process,’’ DP proteins were preferentially cataloged into ‘‘cellular process,’’ ‘‘metabolic process’’ and ‘‘response to stimulus,’’ whereas ‘‘death’’ and ‘‘multi-organism process’’ were under-represented. In terms of ‘‘molecular function,’’ phosphoproteins related to ‘‘catalytic,’’ ‘‘binding’’ and ‘‘transcription regulator’’ accounted for over 30 % of all the phosphoproteins identified, but the ‘‘enzyme regulators’’ and ‘‘translation regulator’’ only took less than 2 %. Considering that ‘‘binding’’ and ‘‘transcription regulator’’ are closely related to mRNA transcription, the GO analysis results also indicated that protein phosphorylation may be involved in the gene transcription, which is consistent with our KEGG pathway results. From the ‘‘cellular component’’ perspective, ‘‘cell,’’ ‘‘intracellular’’ and ‘‘organelle’’ were over-represented in our DP proteins when the whole-genome encoding proteins were used as a control, only less than 1 % of the phosphoproteins were related to ‘‘envelope’’ and ‘‘endomembrane system.’’ The phosphorylation of DP protein is uncorrelated to protein abundance To evaluate the relationship between protein quantity and protein phosphorylation intensity of the DP proteins, we semi-quantified the protein abundance of six DP proteins which are master regulators of rice seed development in S0,

123

ID

LOC_Os01g44220

LOC_Os01g49000

LOC_Os02g15350

LOC_Os02g34850

LOC_Os02g52480

LOC_Os02g58480

LOC_Os04g55230

LOC_Os05g05240

LOC_Os05g06280

LOC_Os05g06480

LOC_Os05g11414

LOC_Os05g26890

LOC_Os05g33570

LOC_Os06g49840

LOC_Os07g08420

LOC_Os07g39480

Sequence

CVFTSDADRDTPHLR

123

SSSPPIR

ASGGALSPVEEKPTVVK

TPVHVSR

CSSTASSVDAAAQDR

RSDSIADMMPEALR

KPSGPDESGNAQSS NNINKSPSGR

AMDATWMQELNIQRPPTPTR

QLQQQQRPTSASASQNSSR

VERESDGGDGDGDGVD DEEDGVDYR

EQEAEPSTGLMMPEPAP VASPGSGGSGGSGSVGAEK

YVISPDNQEIGEK

AGLDYVSCSPFR

VVPSQPNLHGMAYGGNHDLR

CPSGWNLER

SANPDILPSPR

Superfamily of TFs having WRKY and zinc finger domains

bZIP transcription factor

MADS-box family gene with MIKCc type-box

Pyruvate, phosphate dikinase

G-protein alpha subunit

MADS-box family gene with MIKCc type-box

Inorganic H? pyrophosphatase

ATKINESIN-13A/ KINESIN-13A

Serine/threonine protein phosphatase

Tetratricopeptide repeat domain containing protein

Sucrose synthase

Cyclin-dependent kinase inhibitor

Histone-lysine Nmethyltransferase ASHH2

Dof zinc finger domain containing protein

Katanin p60 ATPasecontaining subunit

Glucose-1-phosphate adenylyltransferase large subunit

Annotation

Table 1 Some selected examples of the DP proteins in this study

Regulates grain filling

Regulates internodes and seed development

OsWRKY78

Regulates flower development

OsMADS16/ SPW1 RISBZ1/ OsbZIP58

Regulates starch and fatty acid biosynthesis and accumulation

Regulates grain shape and size

Regulates flower development

Regulates the chalkness content

Controls seed shape and size

Positively regulates the grain size

Regulates starch content and grain size

Involved in starch synthesis

Regulates grain filling rate

Epigenetically regulates BR synthesis and signaling

Controls the storage protein, stach and fatty acid content

FLO4/ OsPPDKB

D1/RGA1

OSMADS58

Chalk5

SRS-3

OsPPKL2

FLO-2

SUS6

KRP1

SDG725

RPBF/ OsDof3

Regulates grain shape and size

Involved in starch synthesis

OsAPL2/ osagpl2-3 DGL1

Reported function

Abbreviation

S9

S3

7.714E7

0.000E0

2.917E7

0.000E0

S9

S4

0.000E0

2.568E7

0.000E0

1.570E7

0.000E0

0.000E0

2.760E7

0.000E0

0.000E0

3.780E6

0.000E0

0.000E0

S0 intensity

S4

S20

S5

S10

T17

S20

S4

S3

T1

S7

S3

T11

Phosphosite

1.577E8

0.000E0

0.000E0

0.000E0

1.276E7

0.000E0

0.000E0

1.746E7

2.052E7

0.000E0

4.436E6

1.898E7

3.111E6

2.653E7

3.046E7

8.556E7

S3 intensity

1.692E8

2.898E7

0.000E0

4.341E7

3.981E7

0.000E0

1.402E8

0.000E0

1.001E8

8.824E7

0.000E0

2.411E7

0.000E0

5.357E8

0.000E0

9.083E8

S7 intensity

Zhang et al. (2011)

Kawakatsu et al. (2009)

Nagasawa et al. (2003)

Kang et al. (2005)

Fujisawa et al. (1999)

Yamaguchi et al. (2006)

Li et al. (2014)

Kitagawa et al. (2010)

Zhang et al. (2012)

She et al. (2010)

Hirose et al. (2008)

Barroco et al. (2006)

Sui et al. (2012)

Kawakatsu et al. (2009)

Komorisono et al. (2005)

Akihiro et al. (2005)

Ref

Plant Mol Biol

Hobo et al. (1999) 6.243E7 6.185E7 4.657E7

LOC_Os08g36790

LOC_Os08g36790

LOC_Os08g36790

DFGSMNMDELLR

QGSLTLPR

GDGDLSSPMAPVP YPFEGVIR

Regulates grain filling TRAB1/ OsbZIP66 bZIP transcription factor

S7

Hobo et al. (1999) 1.298E8 2.762E8 1.499E8

LOC_Os08g36790 EASPGAAAADGGG GGGEQQQPR

Regulates grain filling TRAB1/ OsbZIP66 bZIP transcription factor

S3

Hobo et al. (1999) 0.000E0 4.669E6 0.000E0

LOC_Os08g25734 IFPSRSNVASEQQQSK

Regulates grain filling TRAB1/ OsbZIP66 bZIP transcription factor

S4

Hobo et al. (1999) 1.066E8 7.741E7 0.000E0

LOC_Os08g09230 EHINSDEETFDTYNR

Regulates grain filling TRAB1/ OsbZIP66 bZIP transcription factor

S3

Akihiro et al. (2005) 7.157E8 5.985E7 0.000E0 Involved in starch synthesis OsAPS2/ OsAGPS2b Glucose-1-phosphate adenylyltransferase large subunit

S6

Ryoo et al. (2007) 1.594E8 0.000E0 0.000E0 Involved in starch synthesis OsSSIIIa/ Flo5 Starch synthase III

S5

Reported function ID Sequence

Table 1 continued

Annotation

Abbreviation

Phosphosite

S0 intensity

S3 intensity

S7 intensity

Ref

Plant Mol Biol

S3 and S7, including D1/RGA1 (LOC_Os05g26890), OsABF1 (LOC_Os01g64730), KRP1 (LOC_Os02g52480), PPKL2 (LOC_Os05g05240), RISBZ1 (LOC_Os07g08420) and SMG1 (LOC_Os02g54600). The protein abundances were quantified based on the specific band intensities on the immune-blot by using ImageJ tool, and then normalized to the internal control actin (Fig. 5b). It is interesting that no significant correlation was observed between the protein quantity and phosphorylation intensity for all the six proteins (P [ 0.05), suggesting that the protein phosphorylation events during seed development are independent of the protein quantity. For example, D1/RGA1, PPKL2 and SMG1 showed increasing phosphorylation intensity with the development of rice seeds. However, the protein abundance of these proteins exhibited distinct inclinations. Though the amount of D1/RGA1 also had a very similar tendency with the protein phosphorylation intensity, the correlation is statistically not significant (r = 0.97, P [ 0.05). The quantity of SMG1 decreased to 0.37 in S3, but got back to 1.14 in S7, which is apparently different from the SMG1 phosphorylation intensity. More interestingly, PPKL2 abundance even showed an opposite tendency with the phosphorylation intensity, as the PPKL2 amount decreased with the development of rice seeds. From the results above, we suggest that the protein phosphorylation during seed development may work independently of the protein quantity changes. Nevertheless, due to the fact that only a few proteins were tested in this study, a correlation analysis of the phosphorylation intensity with the protein quantity at the proteomic level would be very necessary to draw a precise conclusion.

Discussion Phosphorylated floral meristem determinacy proteins In this study, we identified over four hundred cluster III DP proteins (Fig. 4c), which are highly phosphorylated in pistil, but less or not phosphorylated in S3 and S7, indicating their roles in late rice flower development. Among these proteins, several of them have been well documented for their functions in rice flower development. OsMADS16/SPW1 (LOC_Os06g49840) is a B class protein controlling the differentiation of lodicules and stamen. The homeotic mutation of spw1 converted the lodicules and stamens into pistils and glume-like organs (Lee et al. 2003; Xiao et al. 2003). Further evidence also emphasized that OsMADS16/SPW1 genetically interacts with OsMADS58, another master regulator of flower development (Yamaguchi et al. 2006), to play a key role in suppressing indeterminate growth of whorl-3 primordia (Yun et al. 2013). Interestingly, in our phosphoproteomic data, OsMADS16/SPW1and OsMADS58 were specifically

123

Plant Mol Biol Fig. 4 a Venn diagram showing the number of phosphopeptides identified in S0, S3 and S7. b Venn diagram showing the number of phosphoproteins identified in S0, S3 and S7. c Hierarchical clustering analysis of the DP proteins among S0, S3 and S7

phosphorylated in the pistil sample S0, but not phosphorylated in developing seeds S3 and S7. Such a temporal- and spatialspecific phosphorylation pattern of the flower development regulatory proteins also suggested that phosphorylation-dependent activation of the OsMADS16, OsMADS58 and many other flower development regulatory proteins might be required to keep the proteins functional, while dephosphorylation is an important mechanism inhibiting unnecessary protein functions in the case that they are not required, ultimately to avoid developmental disorder. Phosphorylation activates cell proliferation and enlargement in seed development Seed size determination is a consequence of both cell division and cell size expansion (Deng et al. 2013). For Eukaryotes, cell division rate is conservatively controlled

123

by heterodimeric complexes comprised of a cyclin and a cyclin-dependent-kinase (Dante et al. 2014). In the rice genome, there are at least 49 cyclin genes and 22 cyclindependent kinases. Interestingly, current phosphoproteomic data revealed that 5 cyclins and 9 cyclin-dependent kinases (CDKs) related proteins were differentially phosphorylated, ten out of which were not phosphorylated in S0, but highly phosphorylated in S3 and/or S7. This type of phosphorylation pattern is in coincidence with the robust cell proliferation event after double fertilization, suggesting that phosphorylation modification could activate or enhance the activity of these cell division-related proteins to suffice the rapid cell proliferation. It is noteworthy that the biological function of the DP proteins KRP1 (LOC_ Os02g52480), CDKB2 (LOC_Os08g40170), CDKA1 (LOC_Os03g02680) and CDKD1/R2 (LOC_Os05g32600) have been elucidated as core cell cycle regulators by

Plant Mol Biol Fig. 5 a Gene ontology analysis of the DP proteins. b Western blot analysis of the protein abundance of 6 DP proteins and their phosphorylation intensities. a– c indicate a statistically significant difference at P \ 0.05

previous studies (Barroco et al. 2006; Endo et al. 2012; Fabian-Marwedel et al. 2002; Umeda et al. 1999). In particular, KRP1encoding a CDK inhibitor was reported to play vital roles in rice seed development (Barroco et al. 2006). Ectopical expression of KRP1 dramatically reduced seed production, possibly via disturbing cell proliferation. The highest KRP1 abundance was detected in seeds at 8 DAP at which the seed cell proliferation is preliminarily finished, indicating an important function of KRP1 in the exit from the mitotic cell cycle during rice grain formation. We also found that the highest phosphorylation intensity was detected at 7 DAP (S7) (Table 1), suggesting that phosphorylation highly activated KRP1, which strongly

inhibited the cell division at the right time during development, thus to keep the rice seed at an appropriate size. Molecular manipulation of the KRP1 phosphorylation status via decorating the phosphorylation sites could be a potential method to genetically improve rice yield. In addition to cell division, cell enlargement is equally important in seed development control. In our study, some of the key cell enlargement regulators were differentially phosphorylated as well. For example, OsWRKY78 (LOC_Os07g39480) encoding an Ia type WRKY transcription factor is significantly up-phosphorylated in S3 and S7 (Table 1) (Zhang et al. 2011). Cytological analysis suggested that OsWRKY78 majorly regulates the cell

123

Plant Mol Biol

expansion instead of the cell division rate. Similarly, SRS3 (small and round seed 3), another differentially phosphorylated kinesin 13 subfamily protein, regulates the seed size through the same mechanism (Kitagawa et al. 2010). Role of phosphorylation in phytohormone regulation Seed development is a precisely coordinated, highly complex event, in which phytohormones play essential roles in cell proliferation, tissue specification and metabolism. Several literatures reviewed the hormonal control of seed development in various plant species (Hedden and Thomas 2007; Locascio et al. 2014; Rijavec and Dermastia 2010). In rice, Yang et al (2001) investigated the hormonal dynamics in the seeds during the grain filling stage, and revealed divergent accumulation patterns of different hormones (Yang et al. 2001). Auxin and cytokinin have been well-known for their functions in controlling several significant developmental processes in a either synergistic or antagonistic manner (Su et al. 2011). In rice, cytokinins and IAAs predominantly accumulated in the early seed development stages, but dropped very quickly after 12-15 DAP. Current phosphoproteomic data revealed at least 16 proteins related to auxin biosynthesis, signaling and polar transport that were differentially phosphorylated during seed development. However, we only found two DP proteins cytokinin-Oglucosyltransferase 1 (LOC_Os05g08480) and DEC (LOC_Os12g27994) which were cytokinin related. It has been clear that cytokinin-O-glucosyltransferase 1 could glucosylate cytokinin for hormone homeostasis (Hou et al. 2004), whereas DEC controls the phyllotactic pattern by affecting cytokinin signaling in rice (Itoh et al. 2012). The action of GAs and ABA on seed development is strictly correlated and antagonistic (Locascio et al. 2014). GAs were proposed to play a role in embryogenesis, while ABA seemed to be more relevant to seed maturation as ABA content reached a peak level at a much later developmental stage (Yang et al. 2001). Notably, the identified DP proteins covered at least 8 GA-related proteins and over 30 ABArelated proteins. DGL1 was cloned from a rice mutant with pleiotropic phenotype including dwarf, shorter roots and smaller seeds (Komorisono et al. 2005). DGL1 controls the cell elongation and division in rice by regulating gibberellin biosynthesis independently of gibberellin signaling. During seed development, DGL1 was specifically phosphorylated in S3, a phase with robust cell division and differentiation. However, in S0, when cell division remains un-triggered, and S7, when the GA content starts to decrease after peaking, DGL1 was not phosphorylated (Table 1). This type of phosphorylation pattern suggested that GA biosynthesis majorly occurs in the period of 0 DAP to 7 DAP, which is

123

also consistent with the fact that the maximal seed GA content was reached at 7-9 DAP. In rice, D1/RGA1 (LOC_Os05g26890) encoding an a subunit of the heterotrimeric G protein has been illustrated to affect the G proteindependent GA signaling (Ueguchi-Tanaka et al. 2000). d1 showed shorter internodes and smaller seed setting due to the impaired GA-mediated cell proliferation. Meanwhile, D1/ RGA1was not phosphorylated in S0, but was phosphorylated with high intensities in S3 and S7, respectively (Table 1). Moreover, it was pointed out that RGA1 also participates in signaling with other phytohormones such as BR (Wang et al. 2006). In addition to DGL1 and RGA1, many other DP proteins during seed development like PHYB (LOC_Os03g19590) and OsLOL2 (LOC_Os01g42710) are involved in GA biosynthesis (Hirose et al. 2012; Xu and He 2007), while GID2 (LOC_Os02g36974) participates in GA signaling (Gomi et al. 2004). In our result, the majority of DP ABA-related proteins are functionally related to ABA signaling. Up till now, a new ‘‘PYR/PYL/RCAR-PP2C-SnRK2’’cascade model for ABA signaling has been proposed and validated (Cutler et al. 2010; Umezawa et al. 2010). In rice, at least 78 PP2Cs and 10 SnRK2s have been identified (Kobayashi et al. 2004; Xue et al. 2008). Interestingly, 13 PP2Cs and 3 SnRK2s showed DP pattern among the three seed developmental stages. A couple of ABA responsive transcription factors, such as TRAB1 (LOC_Os08g36790) and other bZIP family members, were differentially phosphorylated as well. A previous study showed that TRAB1 is involved in embryo maturation and seed dormancy, and a SnRK2 type kinase SAPK10 could phosphorylate the 49th serine and 57th serine of TRAB1 in vitro (S94 and S102 respectively in the original paper which used a different reference sequence from this study) (Hobo et al. 1999; Kobayashi et al. 2005). Besides the known 49th serine, our MS analysis identified three new phosphosites, 11th serine, 73th serine and 215th serine, but did not detect the 57th serine (Supplementary Fig. 3). Except the 215th serine, the other three phosphosites were differentially phosphorylated. The 11th serine and 49th serine both had the maximal phosphorylation intensity at S3, but were significantly less or unphosphorylated in S7, suggesting that they are functional in early seed development. Nevertheless, 73th serine was unphosphorylated in S0, but the phosphorylation gradually augmented from S3 to S7, indicating a role in the later stage (Table 1). Phosphorylation on different phosphosites of the same target protein may confer the phosphoprotein different functions or activities (Nishi et al. 2015), the different phosphorylation patterns of TRAB1 phosphosites also indicated a versatile role of it in other processes beyond seed development. Besides the four phytohormones discussed above, BR and ethylene pathways seemed to be subject to

Plant Mol Biol Fig. 6 An overall view of the phosphorylation regulation in early rice seed development as revealed by this study

phosphorylation regulation as well. For example, SDG725 (LOC_Os02g34850) encodes a H3K36 methyltransferase involved in the epigenetic control of rice development. The knock-down of SDG725 caused pleiotropic phenotypes with dwarfism, erect leaves and small seeds, which mimicked the BR defective mutants. SDG725 protein regulates BR biosynthesis or signaling pathways by epigenetically modifying the lysine sites of histones associated with the key BR-related genes (Sui et al. 2012). Given the fact that SDG725 was specifically phosphorylated in S3 (Table 1), we deduced that there exists a phosphorylation regulating mechanism underlying the SDG725 epigenetic control of BR-related development. And this type of phosphorylation switch overriding the epigenetic regulation mechanism seems to be a very universal model in plants as a large amount of epigenetic factors were differentially phosphorylated in our case and other studies (Supplementary Table 1) (Hou et al. 2015; Wang et al. 2014a, b). Phosphorylation regulates grain filling and quality During rice grain filling, the regulatory proteins and enzymes involved in nutrient synthesis and alcoholic fermentation are highly translated and activated by PTM, in which phosphorylation has been attested to impose huge impacts. It was suggested that phosphorylation could enhance the activities of wheat starch synthesis enzymes such as SSI, SSII-a etc. and promote the substrate recognition (Tetlow et al. 2004, 2008). Similar conclusions were

obtained from Maize and Arabidopsis as well (Grimaud et al. 2008; Reiland et al. 2009). In an effort to study the molecular mechanism of poor grain filling of rice inferior spikelets, Zhang et al. (2014b) identified 42 phosphoproteins, most of which are functionally related to metabolism, starch synthesis and protein synthesis/destination. Consistent with previous reports, we identified a large number of differentially phosphorylated, grain filling-related proteins. Regarding starch synthesis, at least two 1,3-beta-glucan synthases, two sucrose synthases, eight monosaccharide transporters and starch synthase IIIa (LOC_Os08g09230) were identified with the majority of them only phosphorylated in developing seeds (S3 and/or S7). In particular, 2 starch synthesis rate-limiting enzymes ADP-glucose pyrophosphorylases OsAGPS2b (LOC_Os08g25734) and OsAGPL2 (LOC_Os01g44220) were phosphorylated at 13th serine and 68th threonine, respectively (Table 1). Other than the biosynthesis of starch, phosphorylation regulates grain filling as well. Rice grain filling has been revealed to be regulated by the synergism between RPBF (LOC_Os02g15350) and RISBZ1 (LOC_Os07g08420) transcription activators (Kawakatsu et al. 2009; Yamamoto et al. 2006). Simultaneous suppression of both RISBZ1 and RPBF significantly reduced the accumulation of the starch, storage proteins and lipids. RISBZ1 and RPBF worked physically and redundantly to ensure the coordination of various processes during seed development in rice. Intriguingly, RISBZ1 was specifically phosphorylated in S7, and RPBF was phosphorylated in all three stages but

123

Plant Mol Biol

showed a significant increasing tendency from S0 to S7 (Table 1) (Onodera et al. 2001). Phosphorylation may also regulate rice grain qualities such as chalkiness and starch structures. For example, Chalk5, a master regulator controlling rice chalkiness, was specifically phosphorylated in S7 (Table 1). It has been demonstrated that Chalk5 encodes a vacuolar H?-translocating pyrophosphatase (V-PPase) with inorganic pyrophosphate (PPi) hydrolysis and H?-translocation activity (Li et al. 2014). Elevated expression of Chalk5 could disturb the pH homeostasis of the endomembrane trafficking system in developing seeds, and greatly affect the storage substances compaction with small vesicle-like structures and air spaces, which finally result in chalky grain (Li et al. 2014). FLO2 (LOC_Os04g55230) and FLO4 (LOC_Os05g33570) are two more documented examples supporting our speculation (Table 1). flo2 and flo4 are phenotypically similar with floury-white endosperm and smaller grain size. However, different effects on the storage substance content were observed for them. flo2 had decreased protein and amylase content when compared with the wildtype, whereas knock-out of FLO4 could increase the seed protein and lipid content without affecting the starch content (Kang et al. 2005; She et al. 2010). Taken together, our phosphoproteomic analysis strongly indicated that phosphorylation-mediated regulation is a key mechanism controlling rice seed development, as numerous master regulators exhibited a differentially phosphorylated pattern (Table 1; Fig. 6). Considering that grain filling majorly takes place at 5–15 DAP, our latest stage S7 (7 DAP) in this study only caught a glimpse of the phosphorylation events in the early grain filling and quality control. Phosphoproteomic analysis of developing seeds in later time points like 10 or 15 DAP is very necessary for a deeper and more comprehensive insight into phosphorylation regulated grain filling. However, the accumulation of polysaccharides, lipids and storage proteins accompanied with the grain filling in the later stage would also greatly interfere with the extracted protein purity, phosphopeptides enrichment, which may eventually result in poor MS identification of the phosphorylated regulatory proteins. Thus, developing novel protein extraction methods, which could exclude the storage proteins without affecting the regulatory protein nature, is crucial for our future study. Acknowledgments We thank Dr. Zhiguo Er of China National Rice Research Institute for assistance in the bioinformatics analysis, Dr. Hana Mujahid of Mississippi State University, USA for critical review of the manuscript. This work was supported by Agricultural Sciences and Technologies Innovation Program of Chinese Academy of Agricultural Sciences (CAAS) to Rice Reproductive Developmental Biology Group, ‘‘Elite Youth’’ Program (CAAS) to Jian Zhang, and National Natural Science Foundation of China (Grant Number: 31401366).

123

Compliance with ethical standards Conflict of interest of interest.

The authors declare that they have no conflict

References Akihiro T, Mizuno K, Fujimura T (2005) Gene expression of ADPglucose pyrophosphorylase and starch contents in rice cultured cells are cooperatively regulated by sucrose and ABA. Plant Cell Physiol 46:937–946 Barroco RM, Peres A, Droual AM, De Veylder L, le Nguyen SL, De Wolf J, Mironov V, Peerbolte R, Beemster GT, Inze D, Broekaert WF, Frankard V (2006) The cyclin-dependent kinase inhibitor Orysa; KRP1 plays an important role in seed development of rice. Plant Physiol 142:1053–1064 Bleckmann A, Alter S, Dresselhaus T (2014) The beginning of a seed: regulatory mechanisms of double fertilization. Front Plant Sci 5:452 Bush SM, Krysan PJ (2007) Mutational evidence that the Arabidopsis MAP kinase MPK6 is involved in anther, inflorescence, and embryo development. J Exp Bot 58:2181–2191 Cabrillac D, Cock JM, Dumas C, Gaude T (2001) The S-locus receptor kinase is inhibited by thioredoxins and activated by pollen coat proteins. Nature 410:220–223 Chen X, Zhang W, Zhang B, Zhou J, Wang Y, Yang Q, Ke Y, He H (2011) Phosphoproteins regulated by heat stress in rice leaves. Proteome Sci 9:37 Chou MF, Schwartz D (2011) Biological sequence motif discovery using motif-x. Curr Protoc Bioinform Chapter 13: Unit 13:15–24 Cutler SR, Rodriguez PL, Finkelstein RR, Abrams SR (2010) Abscisic acid: emergence of a core signaling network. Annu Rev Plant Biol 61:651–679 Dante RA, Larkins BA, Sabelli PA (2014) Cell cycle control and seed development. Front Plant Sci 5:493 de la Fuente van Bentem S, Hirt H (2007) Using phosphoproteomics to reveal signalling dynamics in plants. Trends Plant Sci 12:404–411 Deng ZY, Gong CY, Wang T (2013) Use of proteomics to understand seed development in rice. Proteomics 13:1784–1800 Dennis G Jr, Sherman BT, Hosack DA, Yang J, Gao W, Lane HC, Lempicki RA (2003) DAVID: database for annotation, visualization, and integrated discovery. Genome Biol 4:P3 Duan P, Rao Y, Zeng D, Yang Y, Xu R, Zhang B, Dong G, Qian Q, Li Y (2014) SMALL GRAIN 1, which encodes a mitogen-activated protein kinase kinase 4, influences grain size in rice. Plant J 77:547–557 Endo M, Nakayama S, Umeda-Hara C, Ohtsuki N, Saika H, Umeda M, Toki S (2012) CDKB2 is involved in mitosis and DNA damage response in rice. Plant J 69:967–977 Fabian-Marwedel T, Umeda M, Sauter M (2002) The rice cyclindependent kinase-activating kinase R2 regulates S-phase progression. Plant Cell 14:197–210 Fan S, Meng Y, Song M, Pang C, Wei H, Liu J, Zhan X, Lan J, Feng C, Zhang S, Yu S (2014) Quantitative phosphoproteomics analysis of nitric oxide-responsive phosphoproteins in cotton leaf. PLoS ONE 9:e94261 Fujisawa Y, Kato T, Ohki S, Ishikawa A, Kitano H, Sasaki T, Asahi T, Iwasaki Y (1999) Suppression of the heterotrimeric G protein causes abnormal morphology, including dwarfism, in rice. Proc Natl Acad Sci USA 96:7575–7580 Gomi K, Sasaki A, Itoh H, Ueguchi-Tanaka M, Ashikari M, Kitano H, Matsuoka M (2004) GID2, an F-box subunit of the SCF E3

Plant Mol Biol complex, specifically interacts with phosphorylated SLR1 protein and regulates the gibberellin-dependent degradation of SLR1 in rice. Plant J 37:626–634 Grimaud F, Rogniaux H, James MG, Myers AM, Planchot V (2008) Proteome and phosphoproteome analysis of starch granuleassociated proteins from normal maize and mutants affected in starch biosynthesis. J Exp Bot 59:3395–3406 Gu T, Mazzurco M, Sulaman W, Matias DD, Goring DR (1998) Binding of an arm repeat protein to the kinase domain of the S-locus receptor kinase. Proc Natl Acad Sci USA 95:382–387 Han C, Yang P, Sakata K, Komatsu S (2014) Quantitative proteomics reveals the role of protein phosphorylation in rice embryos during early stages of germination. J Proteome Res 13:1766–1782 Hedden P, Thomas SG (2007) Plant hormone signaling. Blackwell, Oxford Hirose T, Scofield GN, Terao T (2008) An expression analysis profile for the entire sucrose synthase gene family in rice. Plant Sci 174:534–543 Hirose F, Inagaki N, Hanada A, Yamaguchi S, Kamiya Y, Miyao A, Hirochika H, Takano M (2012) Cryptochrome and phytochrome cooperatively but independently reduce active gibberellin content in rice seedlings under light irradiation. Plant Cell Physiol 53:1570–1582 Hobo T, Kowyama Y, Hattori T (1999) A bZIP factor, TRAB1, interacts with VP1 and mediates abscisic acid-induced transcription. Proc Natl Acad Sci USA 96:15348–15353 Hou B, Lim EK, Higgins GS, Bowles DJ (2004) N-glucosylation of cytokinins by glycosyltransferases of Arabidopsis thaliana. J Biol Chem 279:47822–47832 Hou Y, Qiu J, Tong X, Wei X, Nallamilli BR, Wu W, Huang S, Zhang J (2015) A comprehensive quantitative phosphoproteome analysis of rice in response to bacterial blight. BMC Plant Biol 15:163 Itoh J, Hibara K, Kojima M, Sakakibara H, Nagato Y (2012) Rice DECUSSATE controls phyllotaxy by affecting the cytokinin signaling pathway. Plant J 72:869–881 Jiang L, Liu X, Xiong G, Liu H, Chen F, Wang L, Meng X, Liu G, Yu H, Yuan Y, Yi W, Zhao L, Ma H, He Y, Wu Z, Melcher K, Qian Q, Xu HE, Wang Y, Li J (2013) DWARF 53 acts as a repressor of strigolactone signalling in rice. Nature 504:401–405 Kachroo A, Schopfer CR, Nasrallah ME, Nasrallah JB (2001) Allelespecific receptor–ligand interactions in Brassica self-incompatibility. Science 293:1824–1826 Kakita M, Murase K, Iwano M, Matsumoto T, Watanabe M, Shiba H, Isogai A, Takayama S (2007) Two distinct forms of M-locus protein kinase localize to the plasma membrane and interact directly with S-locus receptor kinase to transduce self-incompatibility signaling in Brassica rapa. Plant Cell 19:3961–3973 Kandasamy MK, Thorsness MK, Rundle SJ, Goldberg ML, Nasrallah JB, Nasrallah ME (1993) Ablation of papillar cell function in Brassica flowers results in the loss of stigma receptivity to pollination. Plant Cell 5:263–275 Kang HG, Park S, Matsuoka M, An G (2005) White-core endosperm floury endosperm-4 in rice is generated by knockout mutations in the C-type pyruvate orthophosphate dikinase gene (OsPPDKB). Plant J 42:901–911 Kawakatsu T, Yamamoto MP, Touno SM, Yasuda H, Takaiwa F (2009) Compensation and interaction between RISBZ1 and RPBF during grain filling in rice. Plant J 59:908–920 Kitagawa K, Kurinami S, Oki K, Abe Y, Ando T, Kono I, Yano M, Kitano H, Iwasaki Y (2010) A novel kinesin 13 protein regulating rice seed length. Plant Cell Physiol 51:1315–1329 Kobayashi Y, Yamamoto S, Minami H, Kagaya Y, Hattori T (2004) Differential activation of the rice sucrose nonfermenting1related protein kinase2 family by hyperosmotic stress and abscisic acid. Plant Cell 16:1163–1177

Kobayashi Y, Murata M, Minami H, Yamamoto S, Kagaya Y, Hobo T, Yamamoto A, Hattori T (2005) Abscisic acid-activated SNRK2 protein kinases function in the gene-regulation pathway of ABA signal transduction by phosphorylating ABA response element-binding factors. Plant J 44:939–949 Komorisono M, Ueguchi-Tanaka M, Aichi I, Hasegawa Y, Ashikari M, Kitano H, Matsuoka M, Sazuka T (2005) Analysis of the rice mutant dwarf and gladius leaf 1. Aberrant katanin-mediated microtubule organization causes up-regulation of gibberellin biosynthetic genes independently of gibberellin signaling. Plant Physiol 138:1982–1993 Lai Z, Ma W, Han B, Liang L, Zhang Y, Hong G, Xue Y (2002) An F-box gene linked to the self-incompatibility (S) locus of Antirrhinum is expressed specifically in pollen and tapetum. Plant Mol Biol 50:29–42 Lan Y, Su N, Shen Y, Zhang R, Wu F, Cheng Z, Wang J, Zhang X, Guo X, Lei C, Jiang L, Mao L, Wan J (2012) Identification of novel MiRNAs and MiRNA expression profiling during grain development in indica rice. BMC Genomics 13:264 Lee S, Jeon JS, An K, Moon YH, Chung YY, An G (2003) Alteration of floral organ identity in rice through ectopic expression of OsMADS16. Planta 217:904–911 Li Y, Fan C, Xing Y, Yun P, Luo L, Yan B, Peng B, Xie W, Wang G, Li X, Xiao J, Xu C, He Y (2014) Chalk5 encodes a vacuolar H(?)-translocating pyrophosphatase influencing grain chalkiness in rice. Nat Genet 46:398–404 Locascio A, Roig-Villanova I, Bernardi J, Varotto S (2014) Current perspectives on the hormonal control of seed development in Arabidopsis and maize: a focus on auxin. Front Plant Sci 5:412 Lv DW, Li X, Zhang M, Gu AQ, Zhen SM, Wang C, Li XH, Yan YM (2014) Large-scale phosphoproteome analysis in seedling leaves of Brachypodium distachyon L. BMC Genomics 15:375 Malone BM, Tan F, Bridges SM, Peng Z (2011) Comparison of four ChIP-Seq analytical algorithms using rice endosperm H3K27 trimethylation profiling data. PLoS ONE 6:e25260 Murase K, Shiba H, Iwano M, Che FS, Watanabe M, Isogai A, Takayama S (2004) A membrane-anchored protein kinase involved in Brassica self-incompatibility signaling. Science 303:1516–1519 Nagasaki H, Itoh J, Hayashi K, Hibara K, Satoh-Nagasawa N, Nosaka M, Mukouhata M, Ashikari M, Kitano H, Matsuoka M, Nagato Y, Sato Y (2007) The small interfering RNA production pathway is required for shoot meristem initiation in rice. Proc Natl Acad Sci USA 104:14867–14871 Nagasawa N, Miyoshi M, Sano Y, Satoh H, Hirano H, Sakai H, Nagato Y (2003) SUPERWOMAN1 and DROOPING LEAF genes control floral organ identity in rice. Development 130:705–718 Nakagami H, Sugiyama N, Mochida K, Daudi A, Yoshida Y, Toyoda T, Tomita M, Ishihama Y, Shirasu K (2010) Large-scale comparative phosphoproteomics identifies conserved phosphorylation sites in plants. Plant Physiol 153:1161–1174 Nallamilli BR, Zhang J, Mujahid H, Malone BM, Bridges SM, Peng Z (2013) Polycomb group gene OsFIE2 regulates rice (Oryza sativa) seed development and grain filling via a mechanism distinct from Arabidopsis. PLoS Genet 9:e1003322 Nishi H, Demir E, Panchenko AR (2015) Crosstalk between signaling pathways provided by single and multiple protein phosphorylation sites. J Mol Biol 427:511–520 Onodera Y, Suzuki A, Wu CY, Washida H, Takaiwa F (2001) A rice functional transcriptional activator, RISBZ1, responsible for endosperm-specific expression of storage protein genes through GCN4 motif. J Biol Chem 276:14139–14152 Pu CX, Ma Y, Wang J, Zhang YC, Jiao XW, Hu YH, Wang LL, Zhu ZG, Sun D, Sun Y (2012) Crinkly4 receptor-like kinase is required to maintain the interlocking of the palea and lemma,

123

Plant Mol Biol and fertility in rice, by promoting epidermal cell differentiation. Plant J 70:940–953 Rao Y, Li Y, Qian Q (2014) Recent progress on molecular breeding of rice in China. Plant Cell Rep 33:551–564 Reiland S, Messerli G, Baerenfaller K, Gerrits B, Endler A, Grossmann J, Gruissem W, Baginsky S (2009) Large-scale Arabidopsis phosphoproteome profiling reveals novel chloroplast kinase substrates and phosphorylation networks. Plant Physiol 150:889–903 Reinders J, Sickmann A (2005) State-of-the-art in phosphoproteomics. Proteomics 5:4052–4061 Rijavec T, Dermastia M (2010) Cytokinins and their function in developing seeds. Acta Chim Slov 57:617–629 Rundle SJ, Nasrallah ME, Nasrallah JB (1993) Effects of inhibitors of protein serine/threonine phosphatases on pollination in Brassica. Plant Physiol 103:1165–1171 Ryoo N, Yu C, Park CS, Baik MY, Park IM, Cho MH, Bhoo SH, An G, Hahn TR, Jeon JS (2007) Knockout of a starch synthase gene OsSSIIIa/Flo5 causes white-core floury endosperm in rice (Oryza sativa L.). Plant Cell Rep 26:1083–1095 Schmelzle K, White FM (2006) Phosphoproteomic approaches to elucidate cellular signaling networks. Curr Opin Biotechnol 17:406–414 She KC, Kusano H, Koizumi K, Yamakawa H, Hakata M, Imamura T, Fukuda M, Naito N, Tsurumaki Y, Yaeshima M, Tsuge T, Matsumoto K, Kudoh M, Itoh E, Kikuchi S, Kishimoto N, Yazaki J, Ando T, Yano M, Aoyama T, Sasaki T, Satoh H, Shimada H (2010) A novel factor FLOURY ENDOSPERM2 is involved in regulation of rice grain size and starch quality. Plant Cell 22:3280–3294 Stone SL, Arnoldo M, Goring DR (1999) A breakdown of Brassica self-incompatibility in ARC1 antisense transgenic plants. Science 286:1729–1731 Stone SL, Anderson EM, Mullen RT, Goring DR (2003) ARC1 is an E3 ubiquitin ligase and promotes the ubiquitination of proteins during the rejection of self-incompatible Brassica pollen. Plant Cell 15:885–898 Su YH, Liu YB, Zhang XS (2011) Auxin–cytokinin interaction regulates meristem development. Mol Plant 4:616–625 Sui P, Jin J, Ye S, Mu C, Gao J, Feng H, Shen WH, Yu Y, Dong A (2012) H3K36 methylation is critical for brassinosteroid-regulated plant growth and development in rice. Plant J 70:340–347 Takayama S, Shimosato H, Shiba H, Funato M, Che FS, Watanabe M, Iwano M, Isogai A (2001) Direct ligand–receptor complex interaction controls Brassica self-incompatibility. Nature 413:534–538 Tantikanjana T, Nasrallah ME, Nasrallah JB (2010) Complex networks of self-incompatibility signaling in the Brassicaceae. Curr Opin Plant Biol 13:520–526 Tetlow IJ, Wait R, Lu Z, Akkasaeng R, Bowsher CG, Esposito S, KosarHashemi B, Morell MK, Emes MJ (2004) Protein phosphorylation in amyloplasts regulates starch branching enzyme activity and protein–protein interactions. Plant Cell 16:694–708 Tetlow IJ, Beisel KG, Cameron S, Makhmoudova A, Liu F, Bresolin NS, Wait R, Morell MK, Emes MJ (2008) Analysis of protein complexes in wheat amyloplasts reveals functional interactions among starch biosynthetic enzymes. Plant Physiol 146:1878–1891 Ueguchi-Tanaka M, Fujisawa Y, Kobayashi M, Ashikari M, Iwasaki Y, Kitano H, Matsuoka M (2000) Rice dwarf mutant d1, which is defective in the alpha subunit of the heterotrimeric G protein, affects gibberellin signal transduction. Proc Natl Acad Sci USA 97:11638–11643 Umeda M, Umeda-Hara C, Yamaguchi M, Hashimoto J, Uchimiya H (1999) Differential expression of genes for cyclin-dependent protein kinases in rice plants. Plant Physiol 119:31–40 Umezawa T, Nakashima K, Miyakawa T, Kuromori T, Tanokura M, Shinozaki K, Yamaguchi-Shinozaki K (2010) Molecular basis of

123

the core regulatory network in ABA responses: sensing, signaling and transport. Plant Cell Physiol 51:1821–1839 van Wijk KJ, Friso G, Walther D, Schulze WX (2014) Meta-analysis of Arabidopsis thaliana phospho-proteomics data reveals compartmentalization of phosphorylation motifs. Plant Cell 26:2367–2389 Wang L, Xu YY, Ma QB, Li D, Xu ZH, Chong K (2006) Heterotrimeric G protein alpha subunit is involved in rice brassinosteroid response. Cell Res 16:916–922 Wang K, Zhao Y, Li M, Gao F, Yang MK, Wang X, Li S, Yang P (2014a) Analysis of phosphoproteome in rice pistil. Proteomics 14:2319–2334 Wang X, Zhou S, Chen L, Quatrano RS, He Y (2014b) Phosphoproteomic analysis of developmental reprogramming in the moss Physcomitrella patens. J Proteomics 108:284–294 Wheeler MJ, de Graaf BH, Hadjiosif N, Perry RM, Poulter NS, Osman K, Vatovec S, Harper A, Franklin FC, Franklin-Tong VE (2009) Identification of the pollen self-incompatibility determinant in Papaver rhoeas. Nature 459:992–995 Xiao H, Wang Y, Liu D, Wang W, Li X, Zhao X, Xu J, Zhai W, Zhu L (2003) Functional analysis of the rice AP3 homologue OsMADS16 by RNA interference. Plant Mol Biol 52:957–966 Xu C, He C (2007) The rice OsLOL2 gene encodes a zinc finger protein involved in rice growth and disease resistance. Mol Genet Genomics 278:85–94 Xu J, Zhang S (2015) Mitogen-activated protein kinase cascades in signaling plant growth and development. Trends Plant Sci 20:56–64 Xue T, Wang D, Zhang S, Ehlting J, Ni F, Jakab S, Zheng C, Zhong Y (2008) Genome-wide and expression analysis of protein phosphatase 2C in rice and Arabidopsis. BMC Genomics 9:550 Xue LJ, Zhang JJ, Xue HW (2012) Genome-wide analysis of the complex transcriptional networks of rice developing seeds. PLoS ONE 7:e31081 Yamaguchi T, Lee DY, Miyao A, Hirochika H, An G, Hirano HY (2006) Functional diversification of the two C-class MADS box genes OSMADS3 and OSMADS58 in Oryza sativa. Plant Cell 18:15–28 Yamamoto MP, Onodera Y, Touno SM, Takaiwa F (2006) Synergism between RPBF Dof and RISBZ1 bZIP activators in the regulation of rice seed expression genes. Plant Physiol 141:1694–1707 Yang J, Zhang J, Wang Z, Zhu Q, Wang W (2001) Hormonal changes in the grains of rice subjected to water stress during grain filling. Plant Physiol 127:315–323 Yu CS, Lin CJ, Hwang JK (2004) Predicting subcellular localization of proteins for Gram-negative bacteria by support vector machines based on n-peptide compositions. Protein Sci 13:1402–1406 Yu CS, Chen YC, Lu CH, Hwang JK (2006) Prediction of protein subcellular localization. Proteins 64:643–651 Yu H, Jiang W, Liu Q, Zhang H, Piao M, Chen Z, Bian M (2015) Expression pattern and subcellular localization of the ovate protein family in rice. PLoS ONE 10:e0118966 Yun D, Liang W, Dreni L, Yin C, Zhou Z, Kater MM, Zhang D (2013) OsMADS16 genetically interacts with OsMADS3 and OsMADS58 in specifying floral patterning in rice. Mol Plant 6:743–756 Zemach A, Kim MY, Silva P, Rodrigues JA, Dotson B, Brooks MD, Zilberman D (2010) Local DNA hypomethylation activates genes in rice endosperm. Proc Natl Acad Sci USA 107:18729–18734 Zhang J, Guo D, Chang Y, You C, Li X, Dai X, Weng Q, Chen G, Liu H, Han B, Zhang Q, Wu C (2007) Non-random distribution of T-DNA insertions at various levels of the genome hierarchy as revealed by analyzing 13 804 T-DNA flanking sequences from an enhancer-trap mutant library. Plant J 49:947–959 Zhang J, Nallamilli BR, Mujahid H, Peng Z (2010) OsMADS6 plays an essential role in endosperm nutrient accumulation and is

Plant Mol Biol subject to epigenetic regulation in rice (Oryza sativa). Plant J 64:604–617 Zhang CQ, Xu Y, Lu Y, Yu HX, Gu MH, Liu QQ (2011) The WRKY transcription factor OsWRKY78 regulates stem elongation and seed development in rice. Planta 234:541–554 Zhang X, Wang J, Huang J, Lan H, Wang C, Yin C, Wu Y, Tang H, Qian Q, Li J, Zhang H (2012) Rare allele of OsPPKL1 associated with grain length causes extra-large grain and a significant yield increase in rice. Proc Natl Acad Sci USA 109:21534–21539 Zhang M, Lv D, Ge P, Bian Y, Chen G, Zhu G, Li X, Yan Y (2014a) Phosphoproteome analysis reveals new drought response and

defense mechanisms of seedling leaves in bread wheat (Triticum aestivum L.). J Proteomics 109:290–308 Zhang Z, Zhao H, Tang J, Li Z, Chen D, Lin W (2014b) A proteomic study on molecular mechanism of poor grain-filling of rice (Oryza sativa L.) inferior spikelets. PLoS ONE 9:e89140 Zhou F, Lin Q, Zhu L, Ren Y, Zhou K, Shabek N, Wu F, Mao H, Dong W, Gan L, Ma W, Gao H, Chen J, Yang C, Wang D, Tan J, Zhang X, Guo X, Wang J, Jiang L, Liu X, Chen W, Chu J, Yan C, Ueno K, Ito S, Asami T, Cheng Z, Lei C, Zhai H, Wu C, Wang H, Zheng N, Wan J (2013) D14-SCF(D3)-dependent degradation of D53 regulates strigolactone signalling. Nature 504:406–410

123