Full text
Rice (Oryza sativa L.) production consumes approximately 80% of the total irrigated freshwater resources in Asia (Fageria, 2003; Tuong & Bouman, 2000). An expected significant increase in rice demand by 2,030 poses the major challenge of producing rice with improved drought tolerance and water use efficiency (Khush, 2005). Rice production mainly depends on both the current assimilates and the redistributed assimilates from reserve pools in vegetative tissues either before or after anthesis (Samonte, Wilson, McClung, & Tarpley, 2001; Venkateswarlu & Visperas, 1987). The redistributed carbohydrates from temporary reserve pools in rice refer to straws, which contribute to approximately one‐third of the grain yield and range from 0% to 40%, depending on cultivar and environmental conditions (Pattanaik & Mohapatra, 1988; Takai, Fukuta, Shiraiwa, & Horie, 2005; Venkateswarlu & Visperas, 1987). In the source–sink context, straws are considered as temporary source organs during grain filling, and grains are considered as sink organs. Previous work has shown that moderate soil drying, particularly when imposed during the middle to late stages of grain filling, could improve the transportation of carbon reserves from straws to grains, accelerate the grain filling rate, and increase grain weight, especially that of inferior grain, in both rice and wheat (Wang, Li, Feng, et al., 2019; Yang & Zhang, 2006; Yang, Zhang, Wang, & Zhu, 2001). The above described farm management practices can not only save the freshwater, but also increase the crop yield.
Spikelets located on apical primary branches usually flower earlier and generate larger and heavier grains, referred to as “superior spikelets” (Mohapatra, Patel, & Sahu, 1993). In contrast, some spikelets located on the proximal secondary branches reach anthesis later and generate smaller or partially filled grains, which are consequently referred to as “inferior spikelets” (Mohapatra et al., 1993). Inferior spikelets usually have partially filled grains that fail to fill completely due to a slow grain filling rate (Yang & Zhang, 2009). Grain filling mainly consists of the transportation of soluble sugars from source tissues (straw and leaf) to the spikelets, followed by starch synthesis and accumulation. Starch accumulation in grains is therefore of great importance, because starch is the major constituent of the seeds (Panigrahi, Kariali, Panda, Lafarge, & Mohapatra, 2019; Yoshida, Forno, & Cock, 1971). The grain filling of inferior grains was significantly correlated with the carbon reserve remobilization from the straws under the moderate soil drying condition (Chen et al., 2013; Wang, Li, Feng, et al., 2019; Zhang et al., 2012). However, the balance control of carbon reserve remobilization between the straws and inferior grains in response to moderate soil drying is poorly understood.
Several studies have been carried out to illustrate carbon reserve remobilization in straws and grain filling in spikelets at phenotypic, physiological, and molecular levels using various approaches, including whole‐genome RNA sequencing and proteomics (Chen et al., 2016; Dong et al., 2014; Wang et al., 2017, Wang, Li, Wang, et al., 2019). These studies mentioned above mainly focused on the differences between superior spikelets and inferior spikelets, carbon reserve remobilization in straws between conventional rice and super rice, and grain filling in inferior spikelets under well‐watered and MD conditions. Furthermore, expression pattern of microRNAs involved in rice grain filling was investigated, which give new insight of how microRNAs regulate grain filling (Peng, Du, et al., 2013; Peng, Sun, et al., 2013; Peng et al., 2014, Peng et al., 2011). Additionally, ETHYLENE RESPONSE2 (ETR2) is responsible for starch overaccumulation in stems and low grain weight (Panda, Badoghar, Sekhar, Shaw, & Mohapatra, 2016; Sekhar et al., 2015; Wuriyanghan et al., 2009). ERS1, ERS2, and ETR3 promote starch accumulation in inferior spikelets enhancing the grain filling (Sekhar et al., 2015; Wuriyanghan et al., 2009). One 14–3–3 protein negatively affects inferior spikelet grain filling in rice (Zhang et al., 2019). These genes are good candidates can be used in the rice breeding to increase grain weight. The balance of carbon reserve remobilization between the temporary source organ, the straws, and the sink organ, the inferior grains, is critical in regulating rice production. However, no study has been carried out by using RNA‐seq to investigate the controlling mechanism of carbon reserve remobilization balance between straws and inferior grains under MD conditions during grain filling. Here, we report an analysis of the dynamic transcriptome profile in straws and inferior grains for the study of the carbon reserve remobilization from source to sink under MD conditions at the transcriptional level. Straws and inferior grains showed distinct responses regarding to carbon flow during grain filling under MD conditions.
This experiment was conducted in a greenhouse at the Chinese University of Hong Kong, Hong Kong, China, during the rice‐growing season (April–August). An inbred indica rice cultivar (Oryza sativa subsp. indica) YD6 (YD) was grown in paddy soil pools, which were transplanted on 2 April at spacing of 0.2 m*0.2 m with one seedling per hill (width 1.5 m, length 5 m). The soil was a sandy loam (Typic Fluvaquents, Entisols, US classification) that contained 24.4 g/kg organic matter, 104 mg/kg alkali‐ hydrolyzable N, 34.1 mg/kg Olsen‐P, and 68.1 mg/kg exchangeable K. The soil water content was monitored by using a tension meter. Seeds were germinated in darkness in Petri dishes with moist filter paper at 28°C for 2–3 days until the roots measured 1 cm. Germinated seedlings were transferred onto a black mesh that was kept floating in Kimura B nutrient solution, and 30‐day‐old seedlings were planted with one seedling per hill. The water level in the pools was maintained at 1–2 cm until 9 days after anthesis (DAA), when the moderate soil drying treatment was initiated.
The water level in the pools was kept at 1–2 cm until 9 DAA. Then, the plant trial was conducted with two different water potential levels by controlling the application of water. Well‐watered control (CK) plants were grown in a water depth of 1–2 cm (soil water potential equal to 0 kPa) in the pools by manually application of tap water, while the plants subjected to the MD treatment were maintained with a soil water potential of –25 kPa. Each treatment was performed for three pools as replicates. The soil water potential in the pools used for the MD treatment was monitored at a soil depth of 15–20 cm using a tensiometer consisting of a 5‐cm‐long sensor (Institute of Soil Science, Chinese Academy of Sciences, Nanjing, China); 4 tensiometers were installed in each pool to maintain an even distribution, and readings were recorded twice daily at 10.00 hr and 16.00 hr. When the reading dropped to the designated value, 40 L of tap water per pool was added manually. The pools were protected from rain by covering with a polyethylene shelter.
Two hundred panicles that headed on the same day were tagged. Straws and inferior grains were chosen among the tagged plants. Thirty tagged straws (only the sheath of flag leaves and stems according to Pattanaik and Mohapatra (1988)) were sampled at 12, 18, and 24 DAA under the CK and MD conditions. The sampled straws were divided into two groups (15 plants each) of subsamples. Fifteen tagged straws (five straws formed a subsample) from each stage were sampled for measurement of the nonstructural carbohydrate (NSC) content, which consists of starch and soluble sugars. The remaining sampled straws (three straws formed a subsample) at each stage were immediately chopped, frozen in liquid nitrogen, and stored at −80°C for further analysis. In addition, the inferior grains locating on the proximal secondary branches without hull were sampled at 12, 18, and 24 DAA under the CK and MD conditions and were frozen in liquid nitrogen and stored at −80°C for RNA extraction.
Straws (sheath and stem) for NSC measurements were immediately dried in a forced‐air dryer at 80℃ to constant weight. The dry straws were ground into powder. The NSC measurement was performed as previously described (Wang et al., 2017).The samples used for measuring the starch and soluble carbohydrate content were ground into a fine powder. In a 15‐ml centrifuge tube, 500 mg of the ground sample was mixed with 10 ml of 80% (v/v) ethanol and kept in a water bath at 80°C for 30 min. After cooling in water, the tube was centrifuged at 8,600 g for 10 min. The supernatant was collected, and the extraction was repeated three times. Then, the extract was diluted to a volume of 50 ml with distilled water. The diluted supernatant was used to measure the soluble carbohydrate content as previously described (Wang et al., 2017). The residue remaining in the centrifuge tube after sugar extraction was dried at 80°C for starch extraction using the HClO4 reagent following the method described by Yang et al. (2001), Wang et al. 2017; Wang, Li, Wang, et al., 2019). The soluble carbohydrate and starch contents were calculated based on the spectrophotometric measurements.
Endogenous ABA levels of straws (sheath and stem) were measured using the methods of Bollmark, Kubát, and Eliasson (1988), He (1993) with modifications. Samples of stems (three biological replicates) were ground in a mortar at 0°C in 10 ml of 80% (v/v) methanol extraction medium containing 1 mM butylated hydroxytoluene as an antioxidant. The extract was incubated at 4°C for 4 hr and centrifuged at 4,800 g for 15 min at 4°C. The supernatants were sequentially passed through Chromosep C18 columns (C18 Sep‐Park Cartridge, Waters Corp) and prewashed with 10 ml of 100% and 5 ml of 80% methanol. The hormone fractions were dried under N2 and dissolved in 2 ml of phosphate‐buffered saline (PBS) containing 0.1% (v/v) Tween‐20 and 0.1% (w/v) gelatine (pH 7.5) for analysis by ELISA. The mouse monoclonal antigen and antibody against ABA and immunoglobulin G–horseradish peroxidase (IgG–HRP) used in the ELISA were produced at the Phytohormones Research Institute, China Agricultural University (He, 1993).
The straw (three biological replications) was ground and extracted (at a ratio of 4 ml of buffer per 1 g of tissue) in cold 0.1 M phosphate buffer (pH 6.5) at 4°C, and centrifuged at 15,000 g for 30 min. Then, the α‐amylase and β‐amylase activities were determined as described previously (Yang et al., 2001). The enzyme activities were calculated based on spectrophotometric measurements.
The total RNA from straws and grains sampled at 12, 18, and 24 DAA under the CK and MD treatments was extracted using an RNeasy Plant Mini Kit (Qiagen). Then, libraries were constructed according to a method previously described (Wang, Li, Feng, et al., 2019). The libraries were then sequenced on an Illumina HiSeq4000 PE101 platform from both the 5′‐ and 3′‐ends on the paired end. The raw image data generated by sequencing were transformed by base calling into sequence data of called raw data/raw reads and were stored in fastq format (Liu et al., 2015). Transcriptome data were analyzed according to Pertea's protocol (Pertea, Kim, Pertea, Leek, & Salzberg, 2016). Briefly, clean data (high‐quality reads) were mapped to the reference genome (
Statistical analysis of the data was performed using ANOVA and Tukey's post hoc test to determine least significant differences using SPSS 19.0 software (SPSS Inc.).
In this study, we measured the soluble sugar and starch contents of straws at 12DAA, 18 DAA, and 24 DAA under both the CK and MD conditions. The straws showed a reduced starch content under MD treatment at 12 DAA and 18 DAA, while the starch content under the MD treatment was lower than that under the CK conditions (Figure 1a). However, starch content was increased at 24 DAA under both MD and CK conditions. The content of soluble sugars fluctuated from 12 DAA to 24 DAA (Figure 1b). The soluble sugar content was higher at 12 DAA and 24 DAA under the MD conditions than that under the CK condition, whereas the situation was opposite at 18 DAA. The NSC that consists of starch and soluble sugars showed a similar changing pattern as the starch content (Figure 1c), suggesting the role of MD treatment in improving carbon reserve flow from straws to grains. However, the content of soluble sugars fluctuated status during carbon reserve remobilization (Figure 1b).
Enlarge this image.1 Figure. Moderate soil drying applied postanthesis alters the NSC, soluble sugar, and starch concentration of straw during grain filling. (a) The starch content under two treatments during grain filling. (b) The soluble sugar content of straw at three grain filling stages. (c) The NSC content of straw at three grain filling stages. Values are means (±SD) of three replicates. Significant differences were determined using ANOVA and Tukey's post hoc test: *p < .05
The application of MD treatment during the middle‐to‐late grain filling stages can significantly reduce the NSC content of the straws (Figure 1). The number of up‐ and downregulated genes in each of the pairwise comparisons was calculated. It is clear that the number of upregulated genes was significantly larger than the number of downregulated genes among the three comparisons (Figure 2a). The number of total DEGs was largest at 18 DAA among the three stages (Figure 2a), with a significantly different transcriptome profile between the CK and MD treatments at this stage. DEGs associated with different stages among the three pairwise comparisons were examined. A total of 229 common DEGs were identified in the three comparisons (Figure 2b), which were significantly enriched in the term “response to stimulus,” “auxin‐activated signaling pathway,” “negative regulation of protein metabolic process,” and “response to auxin” (AgriGO, p < .05).
Enlarge this image.2 Figure. Number of differentially expressed genes of straws during grain filling. (a) Numbers of genes up‐ and downregulated and total number of differentially expressed genes (DEGs) in the comparisons between well‐watered controls (CK) and moderate drying (MD) conditions at 12 days after anthesis (DAA), 18 DAA, and 24 DAA. (b) Venn diagram showing the DEGs in the three pairwise comparisons between CK and MD conditions at the three time points during grain filling
Then, the Kyoto Encyclopedia of Genes and Genomes (KEGG,
Enlarge this image.3 Figure. KEGG analysis of DEGs in straws at 18 DAA. Top enriched KEGG pathways at 18 DAA between CK and MD
According to the RNA‐seq results, genes encoding α‐amylase and β‐amylase, whose expression levels are shown in Figure 4, were significantly altered by MD during grain filling. AMYC2, which is associated with α‐amylase, was upregulated by MD (Figure 4a). The transcript levels of BGIOSGA033092, BGIOSGA031385, and BGIOSGA012615, which encode β‐amylase, were higher than those of AMYC2 and were also significantly promoted by the MD treatment (Figure 4b–d). The expression levels of AMYC2 and β‐amylase genes showed similar expression pattern, which verified in qPCR (Figure S2). In addition, the enzyme activities of α‐amylase and β‐amylase under both the CK and MD treatment during grain filling were measured. It was clear that the activities of α‐amylase were significantly increased by MD treatment during grain filling, especially at 18 DAA and 24 DAA (p < .05) (Figure 4e). β‐amylase activities were also improved by the MD treatment (Figure 4f). Some of the genes encoding debranching enzyme and sucrose phosphate phosphatase were also upregulated under MD treatment (Figure S3a,b).
Enlarge this image.4 Figure. Expression level of genes encoding α‐amylase and β‐amylase under the soil drying treatment. (a) Gene expression level of AMYC2 during grain filling. (b–d) Transcript level of BGIOSGA033092, BGIOSGA031385, and BGIOSGA012615, which encode β‐amylase during grain filling. (e) α‐Amylase activities of straw. (f) β‐Amylase activities of straw. Values are means (±SD) of three replicates. Significant differences were determined using ANOVA and Tukey's post hoc test: *p < .05, **p < .01
The MD treatment significantly increased the ABA content in straws during the grain filling stage (Figure 5h). Then, we found that the ABA catabolism genes, ABA8OX1, ABA8OX2, and ABA8OX3, were expressed in the straws during grain filling. Only ABA8OX1 was predominantly expressed among the three genes during grain filling, suggesting its role in controlling the ABA content in straws (Figure 5a). In addition, the expression level of ABA8OX1 was reduced by MD treatment at 18 DAA (Figure 5a), which was positively associated with the enhanced ABA content at this time point. There are five genes involved in ABA biosynthesis, including NCED1, NCED2, NCED3, NCED4, and NCED5. All of those genes were detected except NCED2 (Figure 5d–g). The expression level of NCED1 showed the highest expression level among all the four genes and was enhanced by MD treatment especially at 18 DAA, indicating its pivotal role in controlling the ABA content in straws (Figure 5d). NCED3 showed a similar expression pattern as NCED1. However, the expression levels of NCED4 and NCED5 were decreased by the MD treatment even despite their relatively low expression levels (Figure 5f,g).
Enlarge this image.5 Figure. Differentially expressed ABA synthesis and catabolism genes among the comparisons between CK and MD at 12, 18, and 24 DAA, and ABA content in straws. (a–c) RPKM value of ABA catabolism genes, ABA8OX1, ABA8OX2, and ABA8OX2. (d–g) RPKM value of ABA synthesis genes NCED1, NCED3, NCED4, and NCED5. (h) ABA content in the straw under CK and MD treatments during grain filling. The values are the means ± SDs of three replicates. Significant differences were determined using ANOVA and Tukey's post hoc test: *p < .05
In straws, SUT1 was highly expressed, followed by SUT2, and SUT4 showed the lowest expression level (Figure 6a). The expression levels of SUT1 and SUT4 were slightly reduced by MD treatment, while SUT2 maintained a similar expression level during grain filling. Four SUT genes were detected in the inferior grains, including SUT1, SUT2, SUT4, and SUT5. Coincidently, those genes in inferior grains exhibited a similar expression pattern compared with that in straws (Figure 6b). Additionally, eleven SWEET genes were detected in both straws and grains. The heatmap of the identified SWEET genes is shown in Figure 6c. In the straws, five SWEET genes, including SWEET1, SWEET6B, SWEET15, SWEET11, and SWEET16, were upregulated across the grain filling stages, and SWEET14 and SWEET3 were downregulated during grain filling. Most of the SWEET genes were downregulated during grain filling in the inferior grains (Figure 6c).
Enlarge this image.6 Figure. Expression level of sucrose transporter genes SUTs and SWEETs in straws and inferior grains under MD treatment. (a) SUTs genes expressed in straws during grain filling. (b) SUT genes expressed in inferior grains during grain filling. (c) Heatmap of SWEETs detected in straws and inferior grains under CK and MD treatments during at 12, 18, and 24 DAA. Stages 1, 2, 3, 4, 5, and 6 refer to YD‐CK‐12, YD‐CK‐18, YD‐CK‐24, YD‐MD‐12, YD‐MD‐18, and YD‐MD‐24, respectively. The maps were plotted using the value of RPKM for each gene in the different samples: Blue indicates low values, and red indicates high values
In the straws, a number of TFs were up‐ or downregulated by MD treatment during grain filling. The number of upregulated TFs in the three comparisons was 101, 107, and 49 at 12, 18, and 24 DAA, respectively (Figure 7a). However, the number of downregulated TFs was lower than the upregulated TFs, at only 14, 25, and 33 at 12, 18, and 24 DAA, respectively (Figure 7c). Furthermore, no TFs were commonly downregulated under MD conditions. A total of 8 upregulated TFs were commonly upregulated under MD conditions among all three comparisons (Figure 7a), which belonged to seven different families, including MYB, ERF, WOX, GATA, GRF, and bHLH (Figure 7b). Then, we investigated the coexpression network of the 8 commonly expressed TFs. MYB, encoded by BGIOSGA008698, was predicted to be coexpressed with MYB30 which has been reported to negatively regulate beta‐amylase genes at the transcriptional level (Figure 7d).
Enlarge this image.7 Figure. Transcription factor (TF) detected in different pair of comparisons and predicted interacting proteins under MD conditions in straws. (a) Venn diagram showed the upregulated TFs in different comparisons. (b) Heatmap showed the common upregulated TFs identified among the three comparisons; the maps were plotted using the value of RPKM for each gene in the different samples: Blue indicates low values, and red indicates high values. (c) Venn diagram showed the downregulated TFs in different comparisons. (d) Predicted interacting proteins of MYB using Interactions Viewer (https://string‐db.org/cgi/input.pl). (e) Expression level of MYB30 during grain filling under MD conditions. Significant differences were determined using ANOVA and Tukey's post hoc test: *p < .05
In this study, genes involved in the trehalose pathway were altered by MD treatment in both the inferior grains and straws. TPS (trehalose 6‐phosphate synthase) genes were mainly downregulated by the MD treatment (Figure 8a). The TPP (trehalose‐phosphate phosphatase) genes, responsible for the conversion of T6P (trehalose 6‐phosphate) into trehalose, were mainly downregulated in the straws following MD treatment. However, TPP genes in inferior grains were either down‐ or upregulated by MD treatment (Figure 8a). In addition, the gene encoding trehalose was downregulated by MD in both inferior grains and straws (Figure 8a). In addition, the expression levels of the TPP and TPS genes, which are responsible for the T6P level, can influence the activity of SNRK1, thus changing the expression level of the SNRK1 marker gene (Zhang et al., 2009). Rice orthologs of Arabidopsis (Arabidopsis thaliana) genes reported to be regulated by SNRK1 were obtained by searching on the BioMart (
Enlarge this image.8 Figure. Heatmap of differentially expressed genes involved in the trehalose synthesis and catabolism pathways. (a) Expression pattern of TPS genes. (b) Expression pattern of TPP genes. (c) Expression pattern of genes encoding trehalase. The maps were plotted using the value of RPKM for each gene in the different samples: Blue indicates low values, and red indicates high values
The MD treatment applied at the middle to late stages of grain filling in rice enhanced the carbon reserve remobilization in the straws (Figure 1) and thus increased the grain weight (both the superior grain and inferior grain), especially that of the inferior grain (Wang, Li, Feng, et al., 2019; Yang & Zhang, 2006; Yang et al., 2001).The transport of prestored assimilates from rice straws to inferior grains in rice was increased, which resulted in increased inferior grain weight and improved rice yield (Wang, Li, Feng, et al., 2019; Yang & Zhang, 2006; Zhang et al., 2012). Therefore, a well‐controlled balance between source and sink under MD treatment during grain filling is pivotal for rice production. However, despite the application of MD treatment during grain filling enhanced the prestored carbon reserve remobilization from straws to grains and increased the grain yield, very little is known about the potential role of gene regulation in the MD‐induced improvement of carbon reserve remobilization in straws and the interaction between the straws and inferior grains at the transcriptional level.
Grain filling in rice under MD treatment was enhanced by the efficient carbon flow from the straws, which was transported to grains to improve sink activity, thus increasing the inferior grain weight (Wang, Li, Feng, et al., 2019). The carbon flow translocated from the straws to grains, mainly referred to as sucrose. Therefore, the process of starch degradation into sucrose in the straws and the sucrose reconversion into starch in the grains is essential for grain filling. The main storage form of carbohydrates in rice straws is starch, which is the main component of NSC. In the straws, starch is first catabolized into glucose, the substrate for sucrose synthesis. Sucrose is recognized as the main substance to be transported from straws to grains (Li et al., 2017). Therefore, the conversion of starch to sucrose in the straws is essential for carbon reserve remobilization. First, starch can be degraded via hydrolysis and phosphorolysis, which are catalyzed by several enzymes, including alpha‐amylase, beta‐amylase, alpha‐glucosidase, and starch phosphorylase (Gallagher, Volenec, Turner, & Pollock, 1997; Preiss, 1982). Then, the sucrose is resynthesized by the enzymes SPS and SPP and mobilized from straws to grains by sucrose transporters or plasmodesmata (Isopp, Frehner, Long, & Nösberger, 2000; Wang et al., 2017; Wardlaw & Willenbrink, 1994). In the straws, the genes encoding the alpha‐amylase and beta‐amylase were enhanced by the MD treatment, which might result in the increased enzyme activities of alpha‐amylase and beta‐amylase to promote starch degradation during grain filling (Figure 4). Besides, the genes encoding the enzymes of SPP were upregulated. The inferior grains exhibited enhanced expression levels of genes encoding SUS, AGPase, StSsase, and SBE, which were responsible for starch synthesis and resulted in starch accumulation (Wang, Li, Feng, et al., 2019). Thus, the integration of upregulated amylase genes and SPP genes in straws, and enhanced SUS, AGPase, StSsase, and SBE genes in inferior grains promoted the conversion of starch and sucrose to increase yields under the MD treatment.
Plant hormones, especially ABA, have been demonstrated to be responsible for the nutrient remobilization and grain filling (Fan et al., 2007; Wang, Li, Feng, et al., 2019). It was documented that ABA can inhibit starch biosynthesis and improve starch degradation in the leaf sheaths of rice (Chen & Wang, 2012). In the present study, ABA8OX1 and NCED1 were predominantly expressed in the straws, which exhibited the down‐ and upregulated transcript levels under MD, respectively (Figure 5a,d). The ABA content in the straws was enhanced by MD treatment (Figure 5h). Both the downregulation of ABA8OX1 and the upregulation of NCED1 were positively associated with the increased ABA content. Considering the enhanced ABA content which caused by the downregulation of ABA8OX2 in the inferior grains under MD (Wang, Li, Feng, et al., 2019), it was concluded that different expression patterns of ABA synthesis and ABA catabolism genes in the straws and inferior grains imposed by MD together controlled the grain filling to increase the grain yield and improve the carbon reserve remobilization in the straws.
Sugar transporters in plants mainly include sucrose transporters (SUTs) and SWEET, which are newly identified transporters comprising seven transmembrane domains with a novel structural model (Chen et al., 2010). Twenty‐one SWEET transporters and 5 SUTs were found in rice, respectively (Aoki, Hirose, Scofield, Whitfeld, & Furbank, 2003; Chen et al., 2010; Durand et al., 2018). The rice sucrose transporters OsSUT1, OsSUT3, and OsSUT4 localize to the aleurone (Bai, Lu, Li, Liu, & Liu, 2016; Ishimaru et al., 2001). Antisense inhibition of OsSUT1 expression causes seed filling defects (Scofield et al., 2002). It was also documented that several sucrose transporting SWEETs contributes to seed filling in Arabidopsis (Chen et al., 2015). In addition, the SWEET11 and SWEET15 were verified to play key roles in seed filling in rice (Yang, Luo, Yang, Frommer, & Eom, 2018). In this study, the gene expression levels of various transporters were altered during grain filling (Figure 6). The expression level of SUTs showed no significant difference between the CK and MD treatments in both the straws and inferior grains, indicating that SUT‐mediated sucrose transportation was not enhanced by MD treatment at the transcriptional level (Figure 6a,b). However, SUT2 in japonica rice was demonstrated to be upregulated by MD treatment at the protein level (Wang, Li, Feng, et al., 2019). Therefore, SUTs in indica rice might also be enhanced by the MD treatment at the protein level, promoting carbon reserve remobilization from the straws to the grains. Furthermore, among the five SUT genes, only SUT1, SUT2, and SUT4 were detected in straws, which is consistent with the previous results in the leaf sheaths and straws of rice (Chen & Wang, 2008; Wang et al., 2017). Additionally, the upregulated SWEET genes especially SWEET11 and SWEET15 in the straw altered by MD might be responsible for the carbon reserve remobilization during grain filling, suggesting that SWEET11 and SWEET15 might play key roles in grain filling not only in the grains but also in the straws (Figure 6c).
Trehalose is an ancient sugar containing two molecular glucose molecules, and it is synthesized by two enzymes: TPS and TPP (Paul, Primavesi, Jhurreea, & Zhang, 2008). The trehalose pathway is an important regulator of sucrose utilization in plants (Schluepmann, Pellny, Dijken, Smeekens, & Paul, 2003). Trehalose‐6‐P (T6P) can regulate starch levels through starch synthesis and degradation (Kolbe et al., 2005; Martins et al., 2013). In addition, T6P acts as an inhibitor of SnRK1 activity (Zhang et al., 2009). In this study, TPS and TPP genes were altered under MD treatment in both straws and grains (Figure 8a), possibly affecting the content of T6P and subsequently regulating gene expression. Rice orthologs of Arabidopsis genes, SNRK1‐induced and SNRK1‐repressed genes, were up‐ or downregulated under the MD condition, which might play a role in mediating carbon reserve remobilization in straw and grains (Figure S4).
In conclusion, the application of MD conditions to rice at the postanthesis stage enhanced the ABA content by reducing the expression level of ABA8OX1 and increasing that of NCED1 in straws, which is different from the downregulation of ABA8OX2 by which the ABA content was increased in inferior grains. The expression of amylase genes as well as amylase activity was enhanced by MD treatment in straws, while in inferior grains, a gene involved in starch synthesis was upregulated. Additionally, downregulated MYB30, which is reported to negatively regulate beta‐amylase genes at the transcriptional level, was consistent with an increase in amylase activity and promoted starch degradation in straws under MD during grain filling. In contrast, NAC plays a key role in activating the expression of starch synthesis genes in inferior grains during grain filling under MD. Therefore, the TFs of MYB30 and NAC cooperated in the source and sink to promote carbon reserve remobilization in response to moderate soil drying.
This work was supported by the Key Research and Development Program of Hunan Province (2018NK1010), the National Natural Science Foundation of China (NSFC31771701), Double First‐Class Construction Project of Hunan Agricultural University (SYL201802013), Hunan Agricultural University excellent talent fund of crop science (ZD2018‐2), and the Hong Kong Research Grants Council (AoE/M‐05/12, AoE/M‐403/16, GRF14122415, 14160516, 14177617).
None declared.
Jianhua Zhang, Nenghui Ye, and Jianchang Yang designed this research; Guanqun Wang, Haoxuan Li, and Yulong Gong performed this research; Guanqun Wang wrote the paper; and Nenghui Ye and Yake Yi revised the paper.
© 2020. This work is published under http://creativecommons.org/licenses/by/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.