Zhang Guangchen, Liu Zimeng, Liu Youhong, Kuya Noriyuki, Hua Yuchen, Shi Hongru, Zhao Weilin, Han Yuqing, Yamamoto Toshio, Chen Wenfu, Sun Jian

Research Paper

iTRAQ-Based Proteomics Investigation of Critical Response Proteins in Embryo and Coleoptile During Rice Anaerobic Germination

Zhang Guangchen1, #, Liu Zimeng1, #, Liu Youhong2, Kuya Noriyuki3, Hua Yuchen1, Shi Hongru4, Zhao Weilin1, Han Yuqing1, Yamamoto Toshio5, Chen Wenfu1, Sun Jian1

(Rice Research Institute, Shenyang Agricultural University, Shenyang 110161, China; Institute of Crop Cultivation and Tillage, Heilongjiang Academy of Agricultural Sciences / Heilongjiang Provincial Key Laboratory of Crop Molecular Design and Germplasm Innovation, Haerbin 150086, China; Institute of Crop Science, National Agriculture and Food Research Organization, Ibaraki 305-8518, Japan; Center of Seed Industry Development of Liaoning Province, Shenyang 110034, China; Institute of Plant Science and Resources, Okayama University, Okayama 710-0046, Japan;)

Direct-seeding of rice has become popular in recent years due to its low cost and convenience, however, hypoxic condition limits seedling establishment. In this study, weedy rice WR04-6 with high germination ability under anaerobic conditions was used as a gene donor, and we successfully improved the seedling establishment rate of rice cultivar Qishanzhan (QSZ) based on selection of a new rice line R42 from the recombinant inbred line population. R42 inherited high anaerobic germination (AG) ability, and was used for isobaric tags for relative and absolute quantitation (iTRAQ)-based comparative proteomic studies with QSZ to further explore the molecular mechanism of AG. A total of 719 differentially abundant proteins (DAPs) were shared by R42 and QSZ responded to AG, and thus defined as common response DAPs. A total of 300 DAPs that responded to AG were only identified from R42, which were defined as tolerance-specific DAPs. The common response and tolerance-specific DAPs had similar biochemical reaction processes and metabolic pathways in response to anoxic stress, however, they involved different proteins. The tolerance-specific DAPs were involved in amino acid metabolism, starch and sucrose metabolism, tricarboxylic acid cycle pathway, ethylene synthesis pathway, cell wall-associated proteins and activity of active oxygen scavenging enzyme. Theprotein-protein interactions for the top 60 DAPs indicated that tolerance-specific DAPs had relatively independent protein interaction networks in response to an anoxic environment compared with common response DAPs. The results of physiological indicators showed that α-amylase and superoxide dismutase activities of R42 were significantly increased under anoxic conditions compared with aerobic conditions. Multiple lines of evidence from western blot, physiological analysis and quantitative real-time PCR jointly supported the reliability of proteomics data. In summary, our findings deepened the understanding of the molecular mechanism for the rice response to AG.

iTRAQ-based proteomics; direct-seeding; anaerobic germination; weedy rice; differentially abundant protein

Abiotic stress seriously limits plant growth (Mittler, 2006). Drought and osmotic stresses are major constraints of rice production (Sheteiwy et al, 2018, 2019; Hamoud et al, 2019). Due to climate change, flooding has more recently become another major abiotic stress to crops that decreases the survival rate of crop plants (Setter and Waters, 2003). Rice (L.) is the only cereal crop that can germinate successfully under anaerobic conditions with oxygen deficiency and limited sources of energy (Pearce and Jackson, 1991; Magneschi and Perata, 2009; Yu et al, 2019). Nevertheless, the anaerobic germination (AG) capacity of rice varies greatly, and there are differences among different rice groups, as well as within subgroups based on previous studies (Kuya et al, 2019). For this reason, the strength of AG directly affects the seedling establishment of direct-seeding rice (Xiao et al, 2019).

Howell et al (2007) characterized the changes in water content and metabolic activity in embryos during rice seed germination under aerobic and anaerobic conditions. Narsai et al (2009) defined 10 metabolites and 1 136 transcripts as aerobic responders, and 13 metabolites and 730 transcripts as anaerobic responders in rice embryos and young seedlings. Narsai et al (2015) further clarified the gene expression patterns in oxygen-deprived rice coleoptiles and found that hypoxia- specific post-transcriptional regulation may involve photosynthetic components and an anoxia-specific expression pattern for starch metabolism functions. A comparative transcriptomics study between tolerant and sensitive varieties showed a more energy efficient system under hypoxic conditions in tolerant varieties along with better reactive oxygen species(ROS) handling and cellular pH maintenance (Vijayan et al, 2018). In addition, plant hormones have also been shown to be important factors in coping with and regulating water stress and germination (Hu et al, 2017; Sheteiwy et al, 2018).

Genetic studies of AG in rice have identified a series of QTLs and functional genes (Angaji et al, 2010; Manangkil et al, 2013;Septiningsih et al, 2013; Baltazar et al, 2014; Hsu and Tung, 2015; Zhang et al, 2017; Yang et al, 2019). Arice variety Khao Hlan On has strong AG ability, andis identifiedas a functional gene for AG that limits the accumulation of trehalose 6-phosphate and enhances carbohydrate flux from the source to supply seed germination and coleoptileextension (Kretzschmar et al, 2015). Calcineurin B-like protein10 (OsCBL10) promoter sequence modulates the expression of, especially in response to anaerobic stress. The expression ofshows that the initial calcium signal may localize upstream of CBL-interacting protein kinase 15 (OsCIPK15) and negatively regulate the downstream sugar-sensing pathway to generate energy through fermentative metabolism (Lu et al, 2007; Lee et al, 2009; Ye et al, 2018).

Weedy rice (f.) possesses robust ecological adaptability and high phenotypic plasticity, such as strong reproductive ability, invasivenessand resource competition (Sun et al, 2013, 2019). Weedy rice at Asian high latitudes can be considered as a natural direct-seeding rice because it can establish seedlings without limitation under flooding conditions. Based on this physiological characteristic, we used the weedy rice at Asian high latitudes, WR04-6, as a gene donor in breeding to improve the AG ability of direct- seeding rice. Proteomic studies including phosphor- proteomic analyses were performed to explore the complex processes underlying successful flooding signaltransduction in soybean (Yin et al, 2014; Komatsu et al, 2015; Yin and Komatsu, 2015), however, this research strategy is absent in rice flooding research. In this study, we used a tolerant breeding line and its sensitive parent for the isobaric tags for relative and absolute quantitation (iTRAQ)-based proteomic studies to further understand the molecular mechanism.

Results

AG ability of weedy rice and breeding application

We evaluated the ability of AG based on coleoptile lengths between weedy rice and different types of cultivated rice at 7 dafter sowing.WR04-6 had the highest AG ability, followed by,,and(Fig. 1-A). To improve the AG ability of direct-seeding rice, a RIL (recombinant inbred line) population was constructed by crossing WR04-6 with elite cultivated rice QSZ (Qishanzhan). The 168 RILs showed significant and continuous phenotypic variation in AG ability, which implies a quantitative trait locus (Fig. 1-B).

By considering the performance of AG and other comprehensive agronomic traits, we selected line R42 for both breeding applications and molecular mechanismresearch. The development of coleoptiles under anaerobic conditions showed that R42 inherited the AG ability from weedy rice parent WR04-6 (Fig. 1-C and Fig. S1).

Identification of differentially abundant proteins (DAPs) in response to anoxic conditions

In the first round of iTRAQ, a total of 39 860 peptides and 6 357 proteins were identified, while in the second round of iTRAQ, a total of 40 424 peptides and 6 374 proteins were identified. Additionally, the isoelectric points, molecular weights, protein sequence coverages and peptide counts were assessed. DAPs had both a fold-change of more than 1.50 or less than 0.67 with the-value of less than 0.05. The overall changes in protein abundance of R42 and QSZ under anoxic stress were illustrated(Fig. S2-A). For the comparison between QSZ under aerobic and anoxicconditions (QSZ-aero and QSZ-anox), 1 754DAPswere found with 663 up-regulated and 1 091 down- regulated. For the comparison between R42 under aerobic and anoxic conditions (R42-aero and R42-anox),1 019 DAPs were found with 497 up-regulated and 522 down-regulated.The 719 DAPs shared between the two groups can be considered as common response DAPs under anaerobic conditions. For the DAPs between R42-aero and R42-anox, after excluding the common response DAPs, the 300 remaining DAPs (with 167 up-regulated and 133 down-regulated) can be defined as tolerance-specific DAPs (Fig. S2-B; Tables S1 and S2). We focused on these two sets of DAPs in the following analyses.

Fig. 1. Phenotypic identification of anaerobic germination.

A, Distribution of coleoptile length in five subgroups under anoxic conditions on 7 d after sowing. WRAH, Weedy rice at Asian high latitudes; TEJ,; TRJ,; IN,; AUS,; QSZ, Qishanzhan.

B, Frequency distribution of coleoptile length of QSZ, R42, WR04-6 and recombinant inbred lines under anoxic conditions on 7 d after sowing.

C, Dynamic of coleoptile elongation of QSZ, R42 and WR04-6 under anoxic conditions.

D, Dynamic of coleoptile elongation of QSZ, R42 and WR04-6 under aerobic conditions.

Common response and tolerance-specific DAPs of embryos under anoxic conditions at rice germination stage

In order to distinguish the DAPs with similar and different expression patterns in 719 common response DAPs, we conducted hierarchical cluster analysis. The 649 common response DAPs (367 up-regulated and 282 down-regulated) shared concordant expression patterns, and just 70 common response DAPs (48 up- regulated and 22 down-regulated) exhibited different expression patterns (Fig. S3-A), suggesting that the two groups have similar expression changes of proteins for many common targets to respond to anoxic stress. Based on Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis, the proteins involved in ‘biosynthesis of secondary metabolites’ and ‘metabolic pathway’ functioned directly in response to anoxic stress. Specifically, key lignin biosynthetic pathways of cell walls ‘phenylpropanoid biosynthesis’ and key energy metabolism, such as ‘glycolysis/ gluconeogenesis’, ‘pentose and mannose metabolism’, ‘pentose phosphate pathway’ and ‘starch and sucrose metabolism’, were the biochemical basis of rice AG (Fig. S3-B). Moreover, the 70 common response DAPs that showed different expression patterns were determined to be involved in ‘pyruvate metabolism’ and ‘protein processing inendoplasmic reticulum’, the main metabolic and genetic information processing pathways based on KEGG pathway enrichment analysis (Fig. S3-C).

The KEGG and Gene Ontology (GO) analyses showed that in 300 tolerance-specific DAPs, the proportion of up-regulated DAPs that were annotated in GO terms and KEGG pathways were much higher than in common response DAPs (Fig. 2 and Fig. S4). These 300 tolerance-specific DAPs were annotated andassigned to 84 KEGG pathways, such as‘carbohydrate metabolism’, ‘amino acid metabolism’ and‘lipid metabolism’ (Fig. 2-A). Among the 300 tolerance- specific DAPs, 1 468 GO terms were annotated, of which 858 termswere involved in biological processes, 215 termsin cellular componentsand 395 termsin molecular function categories (Fig. S4-A). The enrichment results of GO and KEGG between commonresponse and tolerance-specific DAPs were similar, with just minor changes in the order of several pathways. In the GO enrichments of tolerance-specific DAPs, the terms ‘cellular process’ in biological process and ‘membrane’ in cellular components were ranked higher than those of common response DAPs (Fig. 2-B and Fig. S4-B). TheGO and KEGG enrichment results implied that regardless of tolerant or sensitive rice, the basic biochemical and metabolic pathways were consistent in response to anaerobic conditions. In these basic pathways, changes in the expression levels of some key proteins may cause differences in the AG capacity of rice.

Fig. 2. Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis of differentially abundant proteins during rice anaerobic germination.

A, KEGG analysis of tolerance-specific differentially abundant proteins (DAPs).

B, KEGG analysis of common response DAPs.

Comparative analysis of proteins abundance for rice AG in carbohydrate metabolisms

In the above analysis, we realized that carbohydrate metabolism is the key factor in the process of AG. We thus mapped the expression levels of some DAPs to the pathways, and metabolic pathways to reveal the protein basis of embryos and coleoptiles in response to anaerobic conditions from tolerant and sensitive rice.The abundances of cytosolic starch phosphorylase, cytosolic phosphoglucomutase and sucrose synthase were up-regulated in tolerance-specificDAPs (Fig. S5), which accelerated the conversion of starch to monosaccharide. Trehalose-6-phosphate phosphatase (TPP) is involved in starch mobilization during AG (Kretzschmar et al, 2015). We detected increased abundance of TPP in the tolerance-specific DAPs. After the monosaccharides enter the glycolysis pathway,the abundance of fructose-bisphosphate aldolase, 2,3- bisphosphoglycerate-independent phosphor-glycerate mutase, enolase, and pyruvate kinase were up-regulated in tolerance-specific DAPs.

Protein-protein interaction analysis for common response and tolerance-specific DAPs

To determine the interacting proteins that may respond to anoxic conditions during germination, 30 common response DAPs (4 down-regulated and 26 up-regulated)and 30 tolerance-specific DAPs (16 up-regulated and 14 down-regulated) with the highest fold change (FC) value were selected for co-expression network analyses with the Pearson correlation coefficient (Fig. S6; Tables S3 and S4). The correlated pairs were filtered using the correlation coefficient (> 0.95).protein-protein interaction (PPI) analysisindicated that common response and tolerance-specific DAPs exhibited relatively independent protein-interaction networks in response to anoxic stress, thus we illustrated them separately. The frequencies of protein- protein interactions in common response DAPs were significantly higher than those in tolerance-specific DAPs. The down-regulated DAP Os03t0277600-01 was the core interacting protein in PPI networks of the common response DAPs based on the FC value and number of interacting proteins (Fig. S6-A). According to the functional description,codes acid phosphatase. Down-regulated Os05t0192100-01 and up-regulated Os03t0785900-01 were the two core interacting proteins in PPI networks of the tolerance- specific DAPs (Fig. S6-B). Os05t0192100-01 is like Stem 28 kDa glycoprotein, which isrelated to nutrient reservoir; and Os03t0785900-01 is like Glutathione-S-transferase, which regulates glutathione transferase activity. As shown in Tables S3 and S4, we sorted 30 common response and 30 tolerance-specific DAPs according to the number of interacting proteins.

Validation of representative proteins and pathways of AG

To further validate the iTRAQresults,pyruvate kinase (Os04t0677500-02) and chitinase 9 (Os05t0399400-00) were selected for western blot analyses using specific antibodies (Fig. 3-A and -B). We found that pyruvate kinase abundance of the two rice samples significantly increased under anoxic conditions, and R42-anox showed the highest abundance. The chitinase 9 abundance of R42 obviously decreased under anoxic stress, whereas QSZ deficiency was quite the opposite.

Moreover, the proteomics results were also verified at the transcription expression level using quantitative real-time PCR (qRT-PCR) analysis. The up-regulated transcripts of six genes (,,,,and) were consistent with the up-regulated of proteins in R42 under anoxic stress, and fold changes of R42 were significantly higher than those of QSZ (Fig. 3-C).andhad weak correlations of fold changes between protein and transcription. Overall, the multiple lines of evidence from western blot, physiological analysis and qRT- PCR jointly support the reliability of proteomics data.

Fig. 3. Validation of isobaric tags for relative and absolute quantitation (iTRAQ) results.

A, Western blot analysis of pyruvate kinase (Os04t0677500-02) and chitinase 9 (Os05t0399400-00) for R42 under aerobic condition (R42-aero), Qishanzhan under aerobic condition (QSZ-aero), R42 under anoxic condition (R42-anox) and Qishanzhan under anoxic condition (QSZ-anox). Accumulated level of BIP-2 was used as a loading control.

B, Heat map showing the fold changes of selected differentially abundant proteins abundance in aerobic and anoxic conditions. The fold change was obtained relative to the aerobic conditions.

C, Quantitative real-time PCR analysis of the transcription levels of eight differentially abundant protein related genes. Data are Mean ± SD (= 3). *,≤ 0.05; **,≤ 0.01.

Effect comparison of alcohol dehydrogenase (ADH), α-amylase and superoxide dismutase (SOD) on AG

Based on the proteomic studies, many critical responseproteins of AG in embryos and coleoptiles were related to carbohydrate metabolism. ADH and α-amylase are the two key enzymes of carbohydrate metabolism. As shown in Fig. 4, ADH activity was significantly higher (≤ 0.01) under anoxic stress than that under aerobic conditions for both R42 and QSZ. However, the α-amylase activity of R42 under anoxic stress was highly significantly increased compared with R42 under aerobic conditions, whereas no significant difference was detected between the two conditions for QSZ. In addition, we determined that the SOD activity of R42 was significantly increased under anoxic conditions compared with aerobic conditions, whereas the SOD activity of QSZ significantly decreased under anoxic conditions. The SOD activity is also consistent withthehigh-level expression of SOD protein (Os05t0323900-01) based on the results of proteomic studies.

Discussion

Proteomics enhance the understanding of rice AG

AG is a key trait for improving direct-seeding of rice. Hsu and Tung (2015) observed significant variation in the AG ability in five subpopulations of; someaccessions exhibited high AG ability. In this study, some weedy rice accessions possessed higher AG ability thanand one of them was applied in the improvement of direct-seeding rice. Several previous studies have detectedexpression of key genes and pathways for coleoptile growth during AG based on gene expression patterns and comparative transcriptomics analysis. These studies found molecular mechanisms related to AG, including starch hydrolysis, mobilization of starch, glycolysis, anaerobic fermentation, hormone regulation and cell extension (Lasanthi-Kudahettige et al, 2007; Huang et al, 2009; Shingaki-Wells et al, 2011; Narsai et al, 2015; Miro et al, 2017; Vijayan et al, 2018). However, changes in protein abundance often occur after the expression of corresponding genes and are also affected by post- translational modifications and cell splicing events. Proteomics makes it easier to find the essence ofbiochemical reactions than transcriptomics analysis(Pandey and Mann, 2000). Proteomics analysis has been applied to the study of rice seed germination (He and Yang, 2013), however, proteomic studies of AG of rice have not been reported. Here, we detected both common response and anaerobic-inducible proteins in embryos using iTRAQ-based analysis, which provided a protein expression profile for further comprehensive understanding of rice biochemical responses to AG from DNA transcription to protein expression.

Key pathways affecting AG of weedy rice

The glycolysis pathway, starch and sucrose metabolism and citrate cycle pathway are the three main metabolic pathways in carbohydrate metabolism. O2-deficiency is the main limiting factor of rice growth in AG, which changes carbohydrate metabolism, enhances anaerobic respiration, and affects regulated fermentative pathways(Kawai and Uchimiya, 2000; Narsai et al, 2009).Tolerantcultivars can improve the activities of sucrose hydrolases, including α-amylase, aldolase and sucrose synthase (SuSy) (Miro and Ismail, 2013; Hsu and Tung, 2017). A recent study indicated that some genes associated with glycolysis were up-regulated and high level glycolysis maintained adenosine triphosphate (ATP) production in lowland rice YueFu under hypoxic conditions (Liu et al, 2020). In the present study, most tolerance-specific DAPs were up-regulated, and the abundance of most DAPs increased in R42 under anoxic stress, but there were no significant changes in QSZ after anoxic treatment. Based on KEGG enrichmentanalysis, ‘carbohydrate metabolism’ class was the mostsignificant in the tolerance-specific DAPs. This specificgroup of proteins in the carbohydrate metabolism pathway might be essential in responding to the biological processes of AG.

Fig. 4. Activity comparison of enzymes between R42 and Qishanzhan (QSZ) under aerobic and anoxic conditions.

A, Alcohol dehydrogenase (ADH)activity. B, α-amylase activity. C, Superoxide dismutase (SOD) activity.

Data are Mean ± SD (= 3). *,≤ 0.05; **,≤ 0.01.

Pyruvate metabolism plays the central role in carbohydrate metabolism when O2is limited, and gene expression involved in this metabolic pathway is very active (Lasanthi-Kudahettige et al, 2007; Narsai et al, 2009). One of the key factors is pyruvate kinase (Os04t0677500-02), a protein of the pyruvate metabolism pathway, which was identified as an up-regulated tolerance-specific DAP in the present study. Moreover, we found that seven proteins were up-regulated in tolerance-specific DAPs; and regulation of these proteins could promote the rate of glycolysis and influence pyruvate production for fermentative pathways (Fig. 2).

Glycolysis product phosphoenolpyruvate can be converted into three aromatic amino acids (phenylalanine,tryptophan and tyrosine), and phenylalanine and tyrosine are the core metabolites of three enriched pathways (phenylalanine biosynthesis, tyrosine biosynthesis and phenylalanine biosynthesis). Our results revealed that many critical response proteins of AG in embryo and coleoptile are related to carbohydrate metabolism. ADH and α-amylase are two key enzymes of carbohydrate metabolism, and both R42 and QSZ showed significantly higher ADH activity under anoxic stress, compared with aerobic conditions. However, R42 showed significantly higher α-amylase activity only under anoxic stress to maintain the energy supply and material accumulation required for germination (Fig. 4-B). In the present study, both physiological and proteomic evidences proved that R42 had high activity of carbohydrate metabolism in response to anoxic stress.

‘ATP binding’, ‘cell wall’, ‘peroxisome’ and ‘calcium ion binding’ also played important roles in AG since many DAPs in these categories were enriched based on GO enrichment analysis. The abundance of most DAPs in R42 under anoxic stress showed clear changes, but there was no significant change in QSZ (Fig. S7). The ATP citrate lyases (ALCs) are key kinases in the process of acetyl-CoA conversion to citrate that encode ATP citrate lyases. The abundance of ALCs (,and) were up-regulated in tolerance- specific DAPs (Fig. S7). After obtaining sufficient ATP, the coleoptile elongation ability is of paramount importance in AG. Cell elongation contributes significantly to coleoptile growth, and cell wall-related genes are uniquely expressed under AG (Cosgrove, 2005; Magneschi and Perata, 2009; Takahashi et al, 2011). Coleoptile elongation depends on cell proliferation and elongation, which is involved in the regulation of cell wall-related proteins. In this study, we identified 11 cell wall-related proteins with fold changes in both common and tolerance-specific DAPs with expected and reasonable up or down regulation.

Ca2+, as a secondary messenger, is involved in hypoxic signaling in plants because transientincreases in the cytosolic Ca2+concentration can bedetected soon after the floodingof maize roots or thehypoxic treatment of(Sedbrook et al, 1996). Ca2+signaling affects the germination ability of rice under submergence (Ho et al, 2017; Ye et al, 2018). We found that some of the calcium ion binding proteins were related to calcium signaling in tolerance-specific DAPs (Fig. S7). The elongation ability of the coleoptile enhances the success rate of germination and seedling establishment under anoxic conditions in rice (Nghi et al, 2019). Moreover, anoxic stress induces excessiveROS, which limits germination during coleoptile elongation (Bailey-Serres and Voesenek, 2008). SOD is the main scavenging enzyme of ROS (Gara et al, 1997), and we also detected that the high-level expression of SOD protein (Os05t0323900-01) was consistent with the high SOD activity in R42 under anoxic stress, which implied that peroxisome accumulation of R42 can ensure AG by reducing the concentration of ROS (Fig. 4 and Fig. S7).

AG is a complex process of seed development, which is involved in major changes in mobilization of reserves, signaling transduction and transcription activation. Here, we identified tolerance-specificproteins of rice AG, which provided important basic protein data for future gene function and molecular mechanism studies and had great potential for breeding applications. Weedy rice has been subjected to severe natural selection and then shaped as a natural direct-seeding rice. Therefore, gene resources need to be systematically researched and applied in weedy rice for AG, and also for other abiotic stress abilities.

Weedy rice is an important genetic resource for innovation of germplasm resource

In order to identify the protein expression profiles in response to anaerobic conditions, we selected a strong AG line derived from a weedy rice breeding population for iTRAQ-based proteomic studies. A total of 719 common and 300 tolerance-specific DAPs from two groups were identified, and they shared some biochemicalreaction processes and metabolic pathways in response to anoxic stress, however, they involved different proteins. In some well-known pathways, such as starch and sucrose metabolism and amino acid metabolism, we found many critical tolerance-specific proteins in embryos during AG. In summary, our proteomic study increased understanding of the molecular mechanism of rice AG, and also provided useful gene resources.

Methods

Plant materials, germination assays and sampling details

Fiverepresentative subpopulations (25 weedy rice,25, 11, 23and 9) were screened for AG.The weedy rice WR04-6, cultivated rice QSZ and 168 of their RILs including R42 were used in this study. After harvesting and drying, seeds were stored at 4 ºC. To break dormancy, seeds were incubated at 50 ºC for 7 d. All seeds were grown at a farm belonging to Shenyang Agricultural University, Liaoning, China.

Germination experiments of sterilized seeds were carried out. Anoxic water was prepared by adding 0.6% oxyrase enzyme (Oxyrase Inc., OH, USA) to sterilized water at 30 ºC overnight. For anaerobic conditions, 20 seeds were transferred to a 50-mL glass bottle (30 mm × 100 mm) filled with anoxic water (Kuya et al, 2019), and the bottle was covered with a lid. For aerobic conditions, 2 mL of sterile water was added to the seeds in the same bottle, which was then covered with a lid. The anoxic and aerobic seeds of R42 and QSZ were germinated for 3 d or 7 d in the dark at 30 ºC in an artificial climate chamber. The harvested embryos with coleoptiles of R42 in aerobic and anoxic germination of 3 d were respectively named R42-aero and R42-anox; the harvested embryo with coleoptiles of QSZ in aerobic and anoxic germination of 3 d were named QSZ-aero and QSZ-anox, respectively. The coleoptile lengths of seeds were investigated every day as an indicator of AG. Seeds were photographed and ImageJ (NIH, MD, USA) was used to measure coleoptile length.

Protein extraction and digestion

Briefly, coleoptiles harvested for each sample were ground separately with liquid nitrogen and extracted using the Plant Total Protein Extraction Kit (Bangfei Bioscience Co., Ltd, Beijing, China). The protein concentration was determined using a protein quantification kit (Dingguo Changsheng, Beijing, China) according to the manufacturer’s instructions. The Filter Aided Sample Preparation protocol was followed to digest all proteins (Wiśniewski et al, 2009).

iTRAQ labeling and liquid chromatography-electrospray ionization-mass spectrometer (MS)/MS analysis

To identify the proteins responsive to rice AG under anoxic conditions, iTRAQ-based proteomic technique was used in embryos (including coleoptiles and removing the endosperm) with or without submergence stress. The embryos of QSZ and R42 under aerobic and anoxic conditions for 3 d were sampled with three biological iTRAQs. We carried out two rounds of iTRAQ analyses for all 12 samples and one mixed sample; and the experimental design is shown in Fig. S7.

Protein peptides from all samples were labeled with two rounds of the 8-plex iTRAQ reagents multiplex kit (ABI, CA, USA). After the labeling reaction, each group was tested for labeling and extraction efficiency, and the sample was subjected to a matrix-assisted laser desorption ionization procedure after ziptip desalting. Each pool of mixed peptides was lyophilized and dissolved in solution A (2% acetonitrile and 20 mmol/L ammonium formate, pH 10). Then, the samples were loaded onto a reverse-phase column (Luna C18, 4.6 mm × 150 mm; CA, USA) and eluted using a step linear elution program at a flow rate of 0.8 mL/min. The samples were collected each min and centrifuged for 5–45 min. The collected fractions were divided into six pools and desalted on C18 cartridges (Sigma, MO, USA).

LC-MS analysis was executed on a Q-Exactive mass spectrometer (Thermo Fisher Scientific, MA, USA) coupled to a nano high-performance liquid chromatography system (UltiMate 3000 LC Dionex, Thermo Fisher Scientific, MA, USA). The samples were loaded onto a C18-reversed phase column (3 μm C18 resin, 75 μm × 15 cm) and separated on an analytical column (5 μm C18 resin, 150 μm × 2 cm; Ammerbuch, Germany) using the mobile phase at a flow rate of 300 nL/min, at a 150 min gradient. Spectra were acquired in data-dependent ‘top 10’ method. MS spectra were measured at a resolution of 15 000 at m/z 400. Dynamic precursor exclusion was allowed for 120 s after each MS spectrum measurement and was set to 17 500 at m/z 200. Normalized collision energy was 30 eV and the underfill ratio, which standardizes the minimum percentage of the target value likely to be reached at the maximum fill time, was defined as 0.1%. The instrument was run with peptide recognition mode enabled.

Proteomic data analysis

Proteome Discoverer 2.1 software was used for data analysis. Peptide identification was performed with the SEQUST search engine using shuidao_IRGSP-1.0_protein_170414. Decoys for the database search were generated with the revert function. The following options were used to identify the proteins: Peptide mass tolerance as ± 15 mg/L, MS/MS tolerance as 20 mmu, enzyme as trypsin, missed cleavage as 2; variable modification: oxidation (M) and iTRAQ8plex (Y), fixed modification: Carbamidomethyl (C), iTRAQ8plex (N-term), iTRAQ8plex (K), database pattern as decoy. The false discovery rate (FDR) for peptides and proteins was set to 0.01. Fold changes of > 1.50 or < 0.67 andvalue < 0.05 in all three replicates were used as thresholds for determining DAPs in the two groups.

Bioinformatics and data analyses

We performed a GO (http://www.geneontology.org) bioinformatics analysis on common and tolerance-specific DAPs to catalog the molecular functions, cellular components and biological processes.The biological pathways of these proteins were determined using the KEGG Pathway Database (https://www.genome.jp/ kegg/pathway.html) to better understand these DAPs in relation to this study.

Cytoscape software (version 3.6.1,Cytoscape, CA, USA) was used to map the protein-protein interaction network. The protein-protein interaction network map was made with> 0.95 as the threshold (Pujana et al, 2007).

Measurements of SOD, α-amylase and ADH

The activities of cellular SOD, α-amylase and ADH were determined according to the instructions of antioxidant enzyme assay kits (BC0175, BC0610, BC1085, Solarbio, China). Briefly, 0.1 g seeds with coleoptiles (0.1 g) of R42 and QSZ under aerobic germination and anoxic germination for 3 d were harvested and ground separately in liquid. The harvested sampleswere homogenized with extraction buffer and centrifuged at 8 000 ×for 10 min. The supernatants were used for the enzyme activity measurement and the colored products were determined using a spectrometer (BioTek, VT, USA).

Western blot analysis

The extracted total proteins (20 μg) of embryos in every sample and marker were separated on 12% SDS-PAGE gels. After separation, they were transferred onto a polyvinylidene difluoride (PVDF) membrane using an electrophoretic transfer system (Bio-Rad, CA, USA) in low temperature conditions (0 ºC).Skim milk (5%) was used to seal membranes for more than 1 h at room temperature. The rabbit polyclonal antibody to pyruvate kinase (Beijing Protein Innovation, Beijing, China) and rabbit polyclonal antibody to chitinase 9 (Beijing Protein Innovation, Beijing, China) were added to membranes at 4 ºC overnight. The next day, the membranes were incubated with goat antirabbit antibody (Beijing Protein Innovation, Beijing, China) for 1 h at room temperature. Immunological reaction of membranes was detected using an HRP-DAB Detection Kit (Tiangen, Beijing, China).

qRT-PCR analysis

Eight DAP-related genes were randomly selected to verify the proteomics results using qRT-PCR analysis. The eight genes are,,,,,,and. Total RNA was converted to cDNA by PrimeScriptTMRT Master Mix (TaKaRa, Tokyo, Japan) according to the instruction of manufacturer. The cDNA was used for qRT-PCR on an Applied Biosystems QuantStudio 3 (Thermo Fisher Scientific)with SYBR Premix Ex Taq II (TaKaRa, Tokyo, Japan). The transcript-specific primers were acquired at https://biodb.swu.edu.cn/qprimerdb/ and synthesized by Huada Biological Technology (Beijing, China) (Table S5). A relative standard curve for threshold values was used for relative quantification analysis. qRT-PCR data of each gene were standardized with the internal reference of.

Acknowledgements

This study was supported by the National Natural Science Foundation of China (Grant No. 31701503), the Provincial Matching Funds to the National Foundation of Applied Technology Research and Development Program in Heilongjiang Province (Grant No. GX18B002) and the National Key Research and Development Program of China (Grant No. 2016YFD0300501).

SUPPLEMENTAL DATA

The following materials are available in the online version of this article at http://www.sciencedirect.com/journal/rice-science; http://www.ricescience.org.

Fig. S1. Germination performance of QSZ, WR04-6 and R42 under anoxic and aerobic conditions at 7 d after sowing.

Fig. S2. Statistics of differentially abundant proteins in two comparison groups.

Fig. S3. Analysis of 719 common responsedifferentially abundant proteins.

Fig. S4. Gene ontology (GO) analysis of differentially abundant proteins during rice anaerobic germination.

Fig. S5. Changes of differentially abundant proteins (DAPs) in carbohydrate metabolism.

Fig. S6. Protein-protein interaction network.

Fig. S7. Heat map of tolerance-specific differentially abundant proteins (DAPs) related to ATP binding, cell wall, peroxisome and calcium ion binding.

Fig. S8. Experimental design of the present iTRAQ-based proteomic study.

Table S1. Common response differentially abundant proteins.

Table S2. Tolerance-specific differentially abundant proteins.

Table S3. Top 30 common response differentially abundant proteins.

Table S4. Top 30 tolerance-specific differentially abundant proteins.

Table S5. Primers used in this study.

Angaji S A, Septiningsih E M, Mackill D J, Ismail A M. 2010. QTLs associated with tolerance of flooding during germination in rice (L.)., 172(2): 159‒168.

Bailey-Serres J, Voesenek L A C J. 2008. Flooding stress: Acclimations and genetic diversity., 59: 313‒339.

Baltazar M D, Ignacio J C I, Thomson M J, Ismail A M, Mendioro M S, Septiningsih E M. 2014. QTL mapping for tolerance of anaerobic germination from IR64 and thelandrace Nanhi using SNP genotyping., 197(2): 251‒260.

Cosgrove D J. 2005. Growth of the plant cell wall., 6(11): 850‒861.

Gara L D, de Pinto M C, Arrigoni O. 1997. Ascorbate synthesis and ascorbate peroxidase activity during the early stage of wheat germination., 100(4): 894‒900.

Hamoud Y A, Wang Z C, Guo X P, Shaghaleh H, Sheteiwy M, Chen S, Qiu R J, Elbashier M M A. 2019. Effect of irrigation regimes and soil texture on the potassium utilization efficiency of rice., 9(2): 100.

He D L, Yang P F. 2013. Proteomics of rice seed germination., 4: 246.

Ho V T, Tran A N, Cardarelli F, Perata P, Pucciariello C. 2017. A calcineurin B-like protein participates in low oxygen signalling in rice., 44(9): 917‒928.

Howell K A, Cheng K, Murcha M W, Jenkin L E, Millar A H, Whelan J. 2007. Oxygen initiation of respiration and mitochondrial biogenesis in rice., 282(21): 15619‒15631.

Hsu S K, Tung C W. 2015. Genetic mapping of anaerobic germination- associated QTLs controlling coleoptile elongation in rice., 8: 38.

Hsu S K, Tung C W. 2017. RNA-seq analysis of diverse rice genotypesto identify the genes controlling coleoptile growth during submerged germination., 8: 762.

Hu Q J, Lin C, Guan Y J, Sheteiwy M S, Hu W M, Hu J. 2017. Inhibitory effect of eugenol on seed germination and pre-harvest sprouting of hybrid rice (L.)., 7(1): 5295.

Huang S B, Taylor N L, Narsai R, Eubel H, Whelan J, Millar A H. 2009. Experimental analysis of the rice mitochondrial proteome, its biogenesis, and heterogeneity., 149(2): 719‒734.

Kawai M, Uchimiya H. 2000. Coleoptile senescence in rice (L.)., 86(2): 405‒414.

Komatsu S, Tougou M, Nanjo Y. 2015. Proteomic techniques and management of flooding tolerance in soybean., 14(9): 3768‒3778.

Kretzschmar T, Pelayo M A F, Trijatmiko K R, Gabunada L F M, Alam R, Jimenez R, Mendioro M S, Slamet-Loedin I H, Sreenivasulu N, Bailey-Serres J, Ismail A M, Mackill D J, Septiningsih E M. 2015. A trehalose-6-phosphate phosphatase enhances anaerobic germination tolerance in rice., 1: 15124.

Kuya N, Sun J, Iijima K, Venuprasad R, Yamamoto T. 2019. Novel method for evaluation of anaerobic germination in rice and its application to diverse genetic collections., 69(4): 633‒639.

Lasanthi-Kudahettige R, Magneschi L, Loreti E, Gonzali S, Licausi F, Novi G, Beretta O, Vitulli F, Alpi A, Perata P. 2007. Transcript profiling of the anoxic rice coleoptile., 144(1): 218‒231.

Lee K W, Chen P W, Lu C A, Chen S, Ho T H D, Yu S M. 2009. Coordinated responses to oxygen and sugar deficiency allow rice seedlings to tolerate flooding., 2: ra61.

Liu J, Hasanuzzaman M, Sun H Z, Zhang J, Peng T, Sun H W, Xin Z Y, Zhao Q Z. 2020. Comparative morphological and transcriptomic responses of lowland and upland rice to root-zone hypoxia., 169: 103916.

Lu C A, Lin C C, Lee K W, Chen J L, Huang L F, Ho S L, Liu H J, Hsing Y I, Yu S M. 2007. The SnRK1A protein kinase plays a key role in sugar signaling during germination and seedling growth of rice., 19(8): 2484‒2499.

Magneschi L, Perata P. 2009. Rice germination and seedling growth in the absence of oxygen., 103(2): 181‒196.

Manangkil O E, Vu H T T, Mori N, Yoshida S, Nakamura C. 2013. Mapping of quantitative trait loci controlling seedling vigor in rice (L.) under submergence., 192(1): 63‒75.

Miro B, Ismail A M. 2013. Tolerance of anaerobic conditions caused by flooding during germination and early growth in rice (L.)., 4: 269.

Miro B, Longkumer T, Entila F D, Kohli A, Ismail A M. 2017. Rice seed germination underwater: Morpho-physiological responses and the bases of differential expression of alcoholic fermentation enzymes., 8: 1857.

Mittler R. 2006. Abiotic stress, the field environment and stress combination., 11(1): 15‒19.

Narsai R, Howell K A, Carroll A, Ivanova A, Millar A H, Whelan J. 2009. Defining core metabolic and transcriptomic responses to oxygen availability in rice embryos and young seedlings., 151(1): 306‒322.

Narsai R, Edwards J M, Roberts T H, Whelan J, Joss G H, Atwell B J. 2015. Mechanisms of growth and patterns of gene expression in oxygen-deprived rice coleoptiles., 82(1): 25‒40.

Nghi K N, Tondelli A, Valè G, Tagliani A, Marè C, Perata P, Pucciariello C. 2019. Dissection of coleoptile elongation inrice under submergence through integrated genome-wide association mapping and transcriptional analyses., 42(6): 1832‒1846.

Pandey A, Mann M. 2000. Proteomics to study genes and genomes., 405: 837‒846.

Pearce D M E, Jackson M B. 1991. Comparison of growth responses of barnyard grass () and rice () to submergence, ethylene, carbon dioxide and oxygen shortage., 68(3): 201‒209.

Pujana M A, Han J D J, Starita L M, Stevens K N, Tewari M, Ahn J S, Rennert G, Moreno V, Kirchhoff T, Gold B, Assmann V, Elshamy W M, Rual J F, Levine D, Rozek L S, Gelman R S, GunsalusK C, Greenbreg R A, Sobhian B, Bertin N, Venkatesan K, Ayivi- Guedehoussou N, Sole X, Hernandez P, Lazaro C, Nathanson K L, Weber B L, Cusick M E, Hill D E, Offit K, Livingston D M, Gruber S B, Parvin J D, Vidal M. 2007. Network modeling links breast cancer susceptibility and centrosome dysfunction.,39(11): 1338‒1149.

Sedbrook J C, Kronebusch P J, Borisy G G, Trewavas A J, Masson P H. 1996. Transgenicreveals organ-specific cytosolic Ca2+responses to anoxia inseedlings., 111(1): 243‒257.

Septiningsih E M, Ignacio J C I, Sendon P M D, Sanchez D L, Ismail A M, Mackill D J. 2013. QTL mapping and confirmation for tolerance of anaerobic conditions during germination derived from the rice landrace Ma-Zhan Red., 126(5): 1357‒1366.

Setter T L, Waters I. 2003. Review of prospects for germplasm improvement for waterlogging tolerance in wheat, barley and oats., 253(1): 1‒34.

Sheteiwy M S, Gong D T, Gao Y, Pan R H, Hu J, Guan Y J. 2018. Priming with methyl jasmonate alleviates polyethylene glycol- induced osmotic stress in rice seeds by regulating the seed metabolic profile., 153: 236‒248.

Sheteiwy M S, Shao H B, Qi W C, Hamoud Y A, Shaghaleh H, Khan N U, Yang R P, Tang B P. 2019. GABA-alleviated oxidative injury induced by salinity, osmotic stress and their combination by regulating cellular and molecular signals in rice., 20(22): 5709.

Shingaki-Wells R N, Huang S B, Taylor N L, Carroll A J, Zhou W X, Millar A H. 2011. Differential molecular responses of rice and wheat coleoptiles to anoxia reveal novel metabolic adaptations in amino acid metabolism for tissue tolerance., 156(4): 1706‒1724.

Sun J, Qian Q, Ma D R, Xu Z J, Liu D, Du H B, Chen W F. 2013. Introgression and selection shaping the genome and adaptive loci of weedy rice in northern China., 197(1): 290‒299.

Sun J, Ma D R, Tang L, Zhao M H, Zhang G C, Wang W J, Song J Y, Li X, Liu Z M, Zhang W X, Xu Q, Zhou Y C, Wu J Z, Yamamoto T, Dai F, Lei Y, Li S, Zhou G, Zheng H K, Xu Z J, Chen W F. 2019. Population genomic analysis andassembly reveal the origin of weedy rice as an evolutionary game., 12(5): 632‒647.

Takahashi H, Saika H, Matsumura H, Nagamura Y, Tsutsumi N, Nishizawa N K, Nakazono M. 2011. Cell division and cell elongationin the coleoptile of rice alcohol dehydrogenase 1-deficient mutantare reduced under complete submergence., 108(2): 253‒261.

Vijayan J, Senapati S, Ray S, Chakraborty K, Molla K A, Basak N, Pradhan B, Yeasmin L, Chattopadhyay K, Sarkar R K. 2018. Transcriptomic and physiological studies identify cues for germination stage oxygen deficiency tolerance in rice., 147: 234‒248.

Wiśniewski J R, Zougman A, Nagaraj N, Mann M. 2009. Universal sample preparation method for proteome analysis., 6(5): 359‒362.

Xiao G Y, Xiao Y L, Li J J, Deng L H, Weng L S, Meng Q C, Yu J H. 2019. High efficiency is a dominant target for current rice breeding., 33(4): 287‒292. (in Chinese with English abstract)

Yang J, Sun K, Li D X, Luo L X, Liu Y Z, Huang M, Yang G L, Liu H, Wang H, Chen Z Q, Guo T. 2019. Identification of stable QTLs and candidate genes involved in anaerobic germination tolerance in rice via high-density genetic mapping and RNA-Seq., 20(1): 355.

Ye N H, Wang F Z, Shi L, Chen M X, Cao Y Y, Zhu F Y, Wu Y Z, Xie L J, Liu T Y, Su Z Z, Xiao S, Zhang H, Yang J C, Gu H Y, Hou X X, Hu Q J, Yi H J, Zhu C X, Zhang J H, Liu Y G. 2018. Natural variation in the promoter of rice calcineurin B-like protein10 (OsCBL10) affects flooding tolerance during seed germination among rice subspecies., 94(4): 612‒625.

Yin X J, Sakata K, Komatsu S. 2014. Phosphoproteomics reveals the effect of ethylene in soybean root under flooding stress., 13(12): 5618‒5634.

Yin X J, Komatsu S. 2015. Quantitative proteomics of nuclear phosphoproteins in the root tip of soybean during the initial stages of flooding stress., 119: 183‒195.

Yu S M, Lee H T, Lo S F, Ho T H D. 2019. How does rice cope with too little oxygen during its early life?, 229(1): 36‒41.

Zhang M C, Lu Q, Wu W, Niu X J, Wang C H, Feng Y, Xu Q, Wang S, Yuan X P, Yu H Y, Wang Y P, Wei X H. 2017. Association mapping reveals novel genetic loci contributing to flooding tolerance during germination inrice., 8: 678.

15 June 2020;

10 November 2020

Copyright © 2021, China National Rice Research Institute. Hosting by Elsevier B V

This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Peer review under responsibility of China National Rice Research Institute

http://dx.doi.org/10.1016/j.rsci.2021.05.009

Sun Jian (sunjian811119@syau.edu.cn); Chen Wenfu (wfchen@syau.edu.cn)

(Managing Editor: Wang Caihong)