Comparative Proteomic Analysis Provides New Insights into Improved Grain-filling in Ratoon Season Rice

Abstract

Grain-filling of rice spikelets (particularly for the later flowering inferior spikelets) is an important characteristic that affects both quality and yield. Rice ratooning technology is used to cultivate a second crop from dormant buds that sprout from stubble left after the first harvest. This study used two rice varieties, the conventional indica rice ‘Jinhui 809’ and the hybrid indica-japonica rice ‘Yongyou 1540’, to assess the impact of rice ratooning on grain-filling. The results indicated that the grain-filling process in inferior spikelets of ratoon season rice (ISR) showed significant improvement compared to inferior spikelets of main crop (late season) rice (ISL). This improvement was evident in the earlier onset of rapid grain-filling, higher seed-setting percentage, and improved grain quality. A label-free quantitative proteomic analysis using mass spectrometry identified 1724 proteins with significant abundance changes, shedding light on the molecular mechanisms behind the improved grain-filling in ISR. The functional analysis of these proteins indicated that ratooning stimulated the metabolic processes of sucrose-starch, trehalose, and hormones in rice inferior spikelets, leading to enhanced enzyme activities related to starch synthesis, elevated concentrations of trehalose-6-phosphate (T6P), indole-3-acetic acid (IAA) and zeatin riboside (ZR) during the active grain-filling phase. This research highlighted the importance of the GF14f protein as a key regulator in the grain-filling process of ISR. It revealed that GF14f transcriptional and protein levels declined more rapidly in ISR compared to ISL during grain-filling. Additionally, the GF14f-RNAi plants specific to the endosperm exhibited improved quality in inferior spikelets. These findings suggest that the enhancement of starch synthesis, increased levels of IAA, ZR, and T6P, along with the rapid decrease in GF14f protein, play a role in enhancing grain-filling in ratoon season rice.

Introduction

Rice ratooning is the practice of harvesting a first crop (referred to as main crop) and then obtaining a second crop (referred to as ratoon crop) from the stem nodes on the rice stubbles remaining after the harvest of the main crop (Plucknett et al. 1970). The utilization of the rice ratooning technology presents numerous benefits, such as a shortened growth cycle, decreased labor and cost demands, improved yield and efficiency, and reduced reliance on pesticides and water usage (Peng et al. 2023; Wang et al. 2020a). This technology has been adopted in various countries, including the United States and Japan, with China being the earliest adopter dating back to 1,700 years ago (Dong et al. 2017; Xu et al. 2021; Wang et al. 2021; Nakano et al. 2020). In South China, rice ratooning has gained popularity as a method of increasing grain yield through multiple cropping. This practice is particularly relevant in regions where light and temperature conditions are suitable for only one seasonal crop but insufficient for two seasons. According to statistics, more than 1 million hectares (Mha) of ratoon rice is grown annually in China (Zhang et al. 2022).

Previous studies have shown that the quality of rice grains in the ratoon season is significantly improved compared to those of the main crop (early season). Deng et al. (2021) demonstrated that ratoon season rice exhibited higher levels of starch and amylose, along with enhancements in the amylose to amylopectin ratio, average molecular weight of starch granules, and proportions of fa (DP 6–12) and fb1 (DP 13–24) chains in amylopectin, when compared to main crop (early season) rice. Additionally, Wu (2005) observed a notable decrease in the percentage of rice chalky grains in the ratoon season compared to the main crop (early season). However, several studies have shown that the rice grain quality of the ratoon season is consistently superior to that of the main crop (late season) when heading time is synchronized, irrespective of seasonal variations such as temperature fluctuations that may impact grain-filling. Huang et al. (2020) and Lin et al. (2022) found ratooning significantly reduced grain chalkiness in the comparison between ratoon season rice and main crop (late season) rice, and this difference was particularly pronounced in inferior spikelets. Therefore, in specific temperate regions, advocating for the adoption of the ratoon cropping system is beneficial for optimizing land resource utilization and is essential for achieving high-yield and high-quality rice production.

Grain-filling is a complex and systematic physiological and biochemical process regulated by various factors and intricately connected to grain quality. To explore the impact of ratooning on rice grain-filling, researchers often compare ratoon season rice and main crop (late season) rice with synchronized heading times to minimize the effects of seasonal climate variations. Previous studies have extensively examined grain-filling characteristics and source-sink traits in both types of rice, revealing that ratoon season rice shows a more efficient transfer of dry matter from source to sink organs compared to main crop (late season) rice. This enhanced transfer benefits the grain-filling process, leading to a faster grain-filling rate in ratoon season rice (Wu et al. 2023; Huang et al. 2020). However, the molecular mechanism underlying the improved grain-filling of ratoon season rice remains unclear. Omics technology is a widely used tool for identifying and analyzing gene expression or protein abundance alterations in particular plant tissues or regions, enabling a more comprehensive understanding of specific biological processes. Proteomic analysis has been widely employed in the study of grain-filling due to the crucial role of protein as the final product of gene expression and a direct representation of physiological phenomena. Xian et al. (2021) performed a proteomic analysis to compare the protein profiles of grains from autotetraploid rice AJNT-4x and its diploid counterpart AJNT-2x. The study revealed an upregulation of proteins B8AM24, B8ARJ0, B8AQM6, and A2ZCE6 in AJNT-4x grains of superior quality, known to play roles in lysine biosynthesis, cell growth, and endosperm development. Luo et al. (2021) improved the grain-filling of inferior spikelets by removing the superior spikelets and revealed an upregulation of several key proteins, including Cytochrome b6-f complex iron-sulfur subunit, F-type H⁺-transporting ATPase subunit epsilon, and photosystem I subunit II, whose heightened expression provides sufficient energy for the grain-filling process following the removal of superior spikelets. These studies identified key proteins impacting grain-filling through proteomic analysis and provide novel perspectives for better understanding the grain-filling process. However, to date, no study has elucidated the inherent mechanisms underlying improved grain-filling by rice ratooning at the proteomic level.

This study compared grain quality and grain-filling of ratoon season rice and main crop (late season) rice with synchronized heading times, using two varieties: conventional indica rice ‘Jinhui 809’ and hybrid indica-japonica rice ‘Yongyou 1540’. Quantitative proteomics strategies were utilized to elucidate the molecular mechanism underlying the changes in grain-filling induced by rice ratooning. This study aimed to identify the vital proteins involved in the enhanced rice grain-filling of ratoon season, which is essential for understanding the molecular basis of the improvement in ratooning-induced grain-filling.

Materials and Methods

Plant Material and Sampling

The indica-type rice ‘Jinhui 809’ and the indica-japonica-type hybrid rice ‘Yongyou 1540’ were grown at the Experiment Station (119.280E, 26.080 N) of Fujian Agriculture and Forestry University, Fuzhou, China, in 2021 and 2022. Each variety was planted in three plots measuring 3 × 7 m² each, where both ratoon rice and main crop (late season) rice were cultivated. Main crop (early season) rice was sown on March 22, 2021, and transplanted on April 28, 2021, with harvest on August 1, followed by the ratoon season. Main crop (late season) rice was sown on June 7, 2021, and transplanted on July 7, 2021, with heading occurring simultaneously with the ratoon season rice. The heading date, when 80% of the plants headed, was September 5, 2021, for ‘Jinhui 809’ and September 1, 2021, for ‘Yongyou 1540’ in the ratoon season. One rice seedling of each rice variety was planted with a spacing of 15 cm × 15 cm per hill in loamy sand soil, which contained 1.54 g·kg^− 1 nitrogen, 2.31 g·kg^− 1 phosphorus, 3.83 g·kg^− 1 potassium, and 2.21% organic substances. The pH value of the soil was 5.70. The fertilization regimen for the main crop (early season) rice consisted of a ratio of N: P:K = 1:0.5:0.8, with nitrogen applied at a rate of 225 kg N·ha^− 1 in four separate applications (basic, tillering, booting, and grain-filling fertilizers in a ratio of 4:2:3:1). Phosphorus was applied solely as basal fertilizer, while potassium was applied in two applications (basal and booting fertilizer in a ratio of 1:1). The fertilization regimen for the main crop (late season) rice mirrored that of the main crop (early season) rice. The fertilization regimen for the ratoon season rice involved applying 180 kg N·ha^− 1 of nitrogen in two separate applications, which included root-preserving fertilizer and tiller-promoting fertilizer in a 1:1 ratio. The root-preserving fertilizer was applied ten days before harvesting the main crop (early season) rice, while the tiller-promoting fertilizer was applied three days after the harvest. The stubble height was maintained at 25 cm during harvesting. This experiment was repeated at the same test site from March to November 2022.

The transgenic plant material (GF14f-RNAi) and its wild-type counterpart (Jinhui 809) were grown in the same agricultural environment and time frame as the main crop (late season) rice of 2021, using the same planting densities and fertilization methods. The transgenic material (GF14f-RNAi) is available at our research facility, and its creation process was described in a previous publication by Zhang et al. (2019).

Each panicle that headed on the same day was tagged. The flowering date and spikelet position were recorded. Panicles were sampled every five days for 5–35 days after flowering (DAF) and stored at − 80 °C. Superior spikelets and inferior spikelets were distinguished according to a previous report (Ishimaru et al. 2003). The superior spikelets refer to those located on the first and second primary rachis branches from the apex of the panicle, while inferior spikelets are situated on the base of secondary rachis branches of the lowest primary branch.

Measurements of Grain Yield and Grain-filling

The yield performance was assessed by measuring the panicle number per square meter, grain number per panicle, seed-setting percentage, and 1000-grain weight at the ripening stage of rice. Panicle numbers per square meter were recorded for each cropping type of rice in every plot. Mature rice grains were collected from each cropping type of rice in every plot and categorized as filled, partially filled, and unfilled grains. The seed-setting percentage was calculated as the ratio of filled grains to the total number of filled, partially filled, and unfilled grains. Filled grains were dried at 80 °C until a constant weight was achieved and then weighed to determine the 1000-grain weight. The theoretical grain yield was calculated using the formula: theoretical grain yield (kg·ha^− 1) = panicle number per square meter × grain number per panicle × seed-setting percentage (%) × 1000-grain weight (g) × 10^− 4. Furthermore, the duration from seed emergence to maturity for the main crop (late season) rice and from the main crop (early season) rice harvest to ripening for ratoon season rice was documented based on the rice developmental staging system by Counce et al. (2000). This growth duration was used to calculate the theoretical daily yield as follows: theoretical daily yield = theoretical grain yield / growth duration.

In the assessment of grain-filling performance, both the 1000-grain weight and seed-setting percentage of superior and inferior spikelets were evaluated separately across different cropping systems. Grains from superior and inferior spikelets of each rice cropping type were sampled at various stages of grain-filling (ranging from 5 DAF to 35 DAF). Some grain samples were dehulled to observe morphological developmental changes, while others were dried at 80 °C to a constant weight and then weighed. The grain weights at different DAF were analyzed using the logistic equation (W = A(1 + Be ^− kt)⁻¹) to create a grain-filling curve, following the methodology described by Kato (1986). The goodness of fit was assessed using the coefficient of determination (R²), where a higher R² value indicates a better fit, with a value closer to 1 representing a stronger fit. By applying the logistic equation, the time reaching the peak filling stage (d) was determined as (lnB) / k.

Grain Quality Determination

The mature filled grains from the inferior spikelets of each cropping type of rice were taken and subsequently air-dried in a well-ventilated area for three months. After the physiological characteristics of the grains were stabilized, the grain quality was determined. A total of 50.0 g of grains were used for the analysis. The glumes were removed from the grains using the motorized huller (TR-250, KETT, Tokyo, Japan) to obtain brown rice grains. Subsequently, the brown rice grains was milled using the rice milling machine (Pearlest, KETT) to remove the outer paste layer, resulting in milled rice grains. The weights of the brown and milled rice grains were recorded. The milled rice grains were then scanned for appearance and shape analysis using the Rice Appearance and Quality Inspection Analyzer (SC-E, Wanshen, Hangzhou, China). The obtained indices included the degree of chalkiness, chalky grain percentage, and broken rice rate. According to the method described by Liu et al. (2023), the brown rice yield, milled rice yield, and head rice yield were calculated as follows: brown rice yield (%) = (the weight of brown rice grains) / (the weight of grain samples)× 100%, milled rice yield (%) = (the weight of milled rice grains) / (the weight of grain samples) × 100%, head rice yield (%) = milled rice yield × (1-broken rice rate). The nutritional composition of milled rice grains was analyzed using the near-infrared multifunctional quality analyzer (NIRS^TMDS2500F, FOSS, Hillerød, Denmark), and the indices included moisture content, protein content, amylose content, fat content, and taste value.

Scanning Electron Microscopy Analysis of the Endosperm Structure

According to the method described by Li et al. (2014), milled rice grains were cut in half and coated with gold under vacuum. The endosperm structure in the belly (the side where the embryo is located), chest (the center part of the cross-section), and back (the opposite side of the belly part) of the endosperm was examined using a scanning electron microscope (Phenom ProX, Phenom-World, Eindhoven, Netherlands) at 10 kV accelerating voltage.

Protein Sample Preparation and LC-MS/MS Analysis

Three fresh grain samples were used for protein sample preparation and LC-MS/MS analysis. For protein extraction, a grain sample was ground to a fine powder using liquid nitrogen and transferred into a tube. Then, the lysis buffer (4% sodium dodecyl sulfate, 100 mM Tris-HCl (pH 8.5), and protease inhibitors) was added to the tube. The solution was subjected to ultrasonication at 30% power for 5 min on ice at 9-s intervals and then centrifuged at 15,000 g for 10 min at 4 °C. The supernatant was collected and combined with pre-cooled methanol at − 20 °C in a 5-fold volumetric ratio. The resulting mixture was settled at − 20 °C for a minimum of 2 h, followed by centrifugation at 18,000 g for 15 min to remove the supernatant. Methanol was used for protein extraction twice, and the remaining residue was settled at room temperature for 5 min to remove methanol. Subsequently, lysis buffer (8 M urea, 100 mM Tris-HCl, and protease inhibitors) was added to the residue. The total protein concentration was determined using the Bicinchoninic Acid Protein Assay Kit (Lablead, Beijing, China). Additionally, the protein was alkylated with iodoacetamide and reduced with dithiothreitol. The peptides underwent enzymatic digestion using the filter-aided sample preparation method, following the protocol outlined by Winiewski et al. (2009). These digested peptides were desalted using a C-18 column and dried using a CentriVap concentrator (Labconco, Kansas City, MO, USA) (Li et al. 2016).

Liquid chromatography-mass spectrometry/mass spectrometry (LC-MS/MS) analysis was performed according to previously described methods with minimal modifications (Li et al. 2017). In summary, desalted peptide samples were separated using the EASY-nano-LC system (Thermo Scientific, Wilmington, MA, USA) at a flow rate of 1200 nL/min. The separation gradient consisted of the following steps: mobile phase B (0.1% formic acid in acetonitrile) increased from 5 to 12% from 0 to 6 min; mobile phase B increased from 12 to 30% from 6 to 72 min; and the mobile phase B increased from 30 to 45% from 72 to 80 min. The solubility increased to 95% within 5 min and then remained constant for the next 5 min. A Q Exactive HF mass spectrometer (Thermo Scientific) was used for data-independent acquisition (DIA) based database construction analysis (Wang et al. 2020b). The full scan was set to a resolution of 120,000 over a range of 398–1002 m/z. DIA scans were performed at a resolution of 30,000. The normalized collision energy was set to 27%, and the maximum injection time was set to automatic. A total of 45 variable DIA windows were employed for data acquisition. The acquired DIA data were processed using Spectronaut 17.4 software (Biognosys, Zurich, Switzerland) (Martinez Del Val et al. 2021). The protein database used for comparison contained the proteome of Oryza sativa L. The false discovery rate at the peptide and protein levels was set to 1%.

Assay of the T6P, IAA, and ZR Content

The Plant T6P ELISA Kit (Zhenke, Shanghai, China) was used for the trehalose-6-phosphate (T6P) content assay. The fresh samples were ground with liquid nitrogen. After precisely weighing a 0.1 g sample, it was transferred to a tube. Then, 900 µL of PBS buffer (0.27 g KH₂PO₄, 1.42 g Na₂HPO₄, 8 g NaCl, 0.2 g KCl in 1 L deionized water, pH = 7.4) was added to the tube. The solution was incubated at room temperature for 10 min and then centrifuged at 4000 g·min^–1 for 5 min. The supernatant was collected for the T6P content assay according to the manufacturer’s instructions.

High performance liquid chromatography (HPLC) was used for the quantification of indole-3-acetic acid (IAA) and zeatin riboside (ZR). The extraction procedure began by pulverizing the fresh samples into a fine powder using liquid nitrogen. Subsequently, 1.0 g of the powdered sample was precisely measured and combined with 80% chromatographic-grade methanol in a pre-cooled tube. After vigorous shaking, the tube was placed on a 4 °C shaker in darkness for a minimum of 16 h for extraction. The resulting supernatant was carefully collected and stored under low light at a low temperature. The residue underwent three additional extractions using the same method. The combined extracts were then subjected to purification following the protocol outlined by Chen & Yang (2005) and filtered through a 0.22 μm nylon membrane. The extracted solution was injected into an HPLC system (Shimadzu, Kyoto, Japan) equipped with a pulseless pump (LC-20 A, Shimadzu) operating at a flow rate of 1.0 mL·min^–1 and a refractive index detector (SPAD-20 A, Shimadzu). The mobile phase consisted of solvent A (0.1% glacial acetic acid) and solvent B (acetonitrile). The elution program used for separation was as follows: 10% B for 0–5 min; 40% B for 5–25 min; 85% B for 25–35 min; 85% B for 35–40 min; 30% B for 40–50 min; and 10% B for 50–55 min. A shim-pack column (GWS C18, 250 mm×4.6 mm) was employed to separate IAA and ZR. The column temperature was set at 35 ℃, and the detection wavelength was 254 nm. The external calibration curve method was utilized with chromatographic standard reagents purchased from Sigma-Aldrich, including ZR and IAA, for quantification purposes.

Determination of Starch Content

The samples used to measure the starch content were killed at 105 ℃ and dried to constant weight at 80 ℃. The dry samples were ground into a fine powder using a 100-mesh sieve. The powder (100 mg) was mixed with ethanol (10 mL, 80% v/v) and heated in a water bath at 80 ℃ for 30 min. After cooling, the sugar-containing supernatant was removed from the mixture by centrifuging it thrice. The remaining residues were dried, and distilled water (2 mL) was added. The tube was shaken in a boiling water bath for 25 min, and 9.36 M HClO₄ (2 mL) was added after it cooled down. After 15 min of shaking, the extract was made up to approximately 10 mL and then centrifuged. The supernatant was collected, and 4.68 M HClO₄ (2 mL) was added to the residue, thereby repeating the extraction process. A volume of 50 mL was achieved by combining the supernatants with distilled water. Finally, starch was analyzed as described by Pucher et al. (1948).

Enzyme Activity Assays

The enzyme activities of trehalose-6-phosphate synthase (TPS) and trehalose-6-phosphate phosphatase (TPP) were measured using TPS (Bioss, Beijing, China) and TPP (Tongwei, Shanghai, China) Activity Assay Kits, respectively, in accordance with the manufacturer’s instructions. Enzyme activities of all samples were quantified in nmol NADH·g^–1 FW·min^–1 for TPS and µg glucose·g^–1 FW·h^–1 for TPP.

The sucrose synthase (SUS) extraction was performed as follows: 1 g of sample was mixed with 5 mL of extraction buffer (50 mM Hepes-NaOH, pH 7.5; 10 mM MgCl₂; 20 mM βME; 2 mM EDTA, 2% (w/v) EG). The mixture was ground and homogenized in a pre-cooled mortar. The extraction process lasted 30 min at 4 ℃. The mixture was then centrifuged for 10 min at 12,000 g and 4 ℃. The supernatant was separated for the activity assay of SUS according to the method described by Doehlert et al. (1988). The activity of SUS was expressed as µg sucrose·g^–1 FW·min^–1.

The extraction methods for ADP-glucose pyrophosphorylase (AGPase), soluble starch synthase (SSS), and starch branching enzyme (SBE) were similar to those for SUS. The only differences were in the composition of the extraction buffer (100 mM Tricine-NaOH, pH 7.5; 8 mM MgCl₂; 2 mM EDTA; 12.5% (v/v) glycerol; 1% (w/v) PVP-40; 50 mM βME) and the centrifugal speed (30,000 g). Activity assays of AGPase, SSS, and SBE were performed according to the method outlined by Nakamura et al. (1989). The activities of AGPase, SSS, and SBE were expressed as U·g^–1 FW·min^–1.

Real-time Reverse Transcription PCR (qRTPCR)

Total RNA extraction from dehulled grains was performed using TRIzol reagent (Invitrogen, Carlsbad, CA, USA). The removal of gDNA and synthesis of cDNA was performed using the TransScript One-Step gDNA Removal and cDNA Synthesis SuperMix (TransGen Biotech, Beijing, China). Specific primer pairs were designed using NCBI resources and were listed in Table S1. Gene expression was assessed using the Mastercycler EP Realplex RT-PCR system (Eppendorf, Hamburg, Germany) and PerfectStartTM Green qPCR SuperMix (TransGen Biotech) according to the manufacturer’s protocol. The Actin1 gene was used as an internal control, and the 2^−ΔΔCT (cycle threshold) method was used to quantify the gene expression levels.

Protein Extraction and Western Blotting Analysis

Proteins were extracted using the standard trichloroacetic acid acetone method (Zhang et al. 2014). The Bicinchoninic Acid Protein Assay Kit (Lablead, Beijing, China) was used to determine the protein concentration. The protein expressions of GF14f and AGPS2 were detected using a GF14f antibody (BPI, Beijing, China) and an AGPS2 antibody (Orizymes, Shanghai, China), respectively. Horseradish peroxidase-conjugated goat anti-rabbit antibody (Sigma-Aldrich, St. Louis, MO, USA) was used as the secondary antibody. Heat shock protein 81 − 2 was used as the internal control. Western blotting was performed after the enhanced chemiluminescence (ECL) reaction using a ChemiDoc MP Imaging System (Bio-Rad, Hercules, CA, USA).

Data Analysis

Statistical analyses were conducted using analysis of variance (ANOVA). A p-value of less than 0.05 was deemed statistically significant for all data. Data organization and analysis were performed using Excel 2021 and DPS 9.01 statistical software, while data visualization was accomplished with Origin 2024 and MapMan 3.5.1.

Results

Analysis of Yield Performance and Grain-filling Characteristics

The present study examined the effects of ratooning technology on grain yield and its components utilizing the large-panicle conventional rice variety ‘Jinhui 809’ and the hybrid rice variety ‘Yongyou 1540’. Results from the two-year study indicated notable variations in multiple parameters (Table S2). In comparison to main crop (late season) rice, ratoon season rice demonstrated a higher effective number of panicles and seed-setting percentage, but a lower number of grains per panicle and 1000-grain weight, ultimately leading to a decreased theoretical grain yield. Nevertheless, upon examination of growth duration, ratoon season rice displayed a higher theoretical daily yield. During the ratoon seasons of 2021 and 2022, the theoretical daily yield of ‘Jinhui 809’ showed significant increases of 7.03% and 18.64%, respectively, compared to the main crop (late season), with corresponding increases of 48.24% and 44.09% observed for ‘Yongyou 1540’. The results from this two-year study indicate that, despite the distinct differences in yield performance among the rice varieties under the ratoon cropping system, ratoon season rice continues to demonstrate high-yield characteristics.

Additionally, this two-year study found that inferior spikelets were consistently more significantly affected by rice ratooning in terms of seed-setting percentage compared to superior spikelets (Table 1). During the ratoon seasons of 2021 and 2022, the seed-setting percentage of inferior spikelets of ‘Jinhui 809’ exhibited a notable increase of 12.41% and 9.32% higher, respectively, in comparison to the main crop (late season), with corresponding increases of 5.52% and 6.17% observed for ‘Yongyou 1540’. Notably, both superior and inferior spikelets demonstrated a reduction in 1000-grain weight by rice ratooning.

Table 1 1000-grain weight and seed-setting percentage of superior and inferior spikelets of ratoon season rice and main crop (late season) rice

The logistic equation was utilized to effectively model the rice grain-filling process in both superior and inferior spikelets of ratoon season rice and main crop (late season) rice in 2021 (Fig. 1B-C; Table S3). Analysis of the grain-filling curve revealed that the grain weight of superior spikelets of ratoon season rice (SSR) and main crop (late season) rice (SSL) increased rapidly between 5 and 15 DAF. In contrast, there was a noticeable disparity during the same timeframe between inferior spikelets of ratoon season rice (ISR) and main crop (late season) rice (ISL). ISR exhibited accelerated grain weight growth compared to ISL between 5 and 15 DAF, particularly in the ‘Jinhui809’ variety. In the case of ‘Jinhui809’, ISL experienced a rapid increase in grain-filling around 15 DAF, whereas ISR initiated this process approximately 5 days earlier than ISL. The observed variations in grain weight were consistent with developmental changes in the morphology of the superior and inferior spikelets of the two varieties during the same year (Fig. 1A). Furthermore, the logistic regression analysis demonstrated that ratoon season rice exhibited an earlier peak grain-filling rate compared to main crop (late season) rice, particularly for inferior spikelets as shown in Table S3. The SSR of ‘Jinhui 809’ and ‘Yongyou 1540’ reached their maximum grain-filling rates 0.67 and 0.83 days earlier than their SSL counterparts, respectively. Similarly, the ISR of ‘Jinhui 809’ and ‘Yongyou 1540’ achieved their peak grain-filling rates 5.75 and 2.92 days earlier than their ISL counterparts.

Fig. 1

Differences of grain-filling process and endosperm structures between grains of ratoon season rice and main crop (late season) rice. A Developmental changes in grains of ratoon season rice and main crop (late season) rice; B-C Grain-filling curve of ratoon season rice and main crop (late season) rice; D-E Scanning electron microscopy images of the cross-sections in endosperm for inferior spikelets of ratoon season rice and main crop (late season) rice at their maturity stage. SSR, superior spikelets of ratoon season rice; SSL, superior spikelets of main crop (late season) rice; ISR, inferior spikelets of ratoon season rice; SSL, inferior spikelets of main crop (late season) rice

Grain Quality Analysis

The grain-filling process significantly impacted grain quality, with inferior spikelets exhibiting more pronounced changes in ratoon season compared to superior spikelets, as evidenced by our results. A comprehensive examination of grain quality was conducted for ISR and ISL samples collected in 2021 and 2022, as presented in Table 2. Throughout the two-year study, consistent disparities in grain quality were observed between ISR and ISL, encompassing processing, appearance, and nutritional characteristics. In the case of the ‘Jinhui 809’ variety, ISR showed superior performance compared to ISL in brown rice, milled rice, and head rice yields, with margins of 3.29%, 4.16%, and 6.4% in 2021, and 2.67%, 2.77%, and 7.32% in 2022, respectively. Similarly, for the ‘Yongyou 1540’ variety, ISR demonstrated higher yields of brown rice, milled rice, and head rice than ISL by 3.25%, 5.48%, and 5.27% in 2021, and 4.67%, 8.76%, and 8.13% in 2022. Notably, ISR displayed a significantly lower chalkiness degree and chalky grain percentage compared to ISL. However, ISR exhibited lower protein and amylose contents compared to ISL, while fat content and taste value were notably higher across both varieties in the two-year study.

Table 2 Grain quality of inferior spikelets of ratoon season rice and main crop (late season) rice

The cross-sectional endosperm structures of ISR and ISL samples collected in 2021 were examined using scanning electron microscopy (500 ×, 100 μm) (Fig. 1D-E). Significant differences were observed in the endosperm structure between ISR and ISL, particularly in the belly, chest, and back regions. The endosperm structure of ISL appeared to be less densely packed compared to that of ISR, displaying noticeable gaps.

Differential Proteomic Analysis of Inferior Spikelets between Ratoon Season Rice and Main Crop (Late Season) Rice with Synchronized Heading Times

To probe the molecular physiology mechanism by which rice ratooning technology improve grain-filling, the label-free quantitative proteomic study was conducted. Grain samples collected in 2021 were utilized for proteomic analysis and subsequent analyses based on proteomic results. The ‘Jinhui 809’ variety was specifically focused on, whose grain-filling of inferior spikelets showed more noticeable changes by rice ratooning compared to the ‘Yongyou 1540’ variety (Fig. 1A-C; Table 1). A differential protein abundance analysis was conducted on the ISR and ISL of ‘Jinhui 809’ at 20 DAF, a stage where they both experienced rapid grain-filling and displayed significant differences in grain weight and size (Fig. 1A-B). The analysis employed a label-free quantitative proteomic approach based on mass spectrometry, successfully quantifying a total of 6433 grain proteins. In total, 1724 differentially expressed proteins (DEPs) were identified using the criteria: P-value < 0.05 and fold change > 1.5 for up-regulated DEPs, and fold change < 0.667 for down-regulated DEPs. Of these DEPs, 987 were up-regulated, and 737 were down-regulated in ISR compared to ISL (Fig. 2A-B). Around 75.3% of the DEPs were associated with 33 known pathways using MapMan annotation (Fig. S1). Notably, a significant proportion (79.7%) of the annotated DEPs were linked to metabolic and regulatory processes, including pathways closely related to grain-filling, such as carbohydrate metabolism, hormone metabolism, and signal transduction (Fig. 2C-D). These findings indicate that the improved grain-filling observed in the ratoon season involves a complex molecular ecological process, highlighting the need for further in-depth analysis of these pathways.

Fig. 2

Quantitative identification and MapMan visualization of differentially expressed proteins (DEPs) in inferior spikelets of ratoon season rice (ISR) compared with inferior spikelets of main crop (late season) rice(ISL). A Volcano plot of the proteins identified. The X-axis specifies the Log2(Fold change) with the fold change between ISR and ISL, and the Y-axis specifies the –Log10(P-Value); up-regulated (up), down-regulated (down), and non-significant (not) proteins are shown as red, blue, and black dots, respectively; vertical and horizontal dotted lines reflect the filtering criteria (fold change > 1.5 or < 0.667 and P-Value < 0.05). B The number of total proteins and DEPs. C Metabolism overview. D Regulation overview. MapMan visualization is shown based on the Log2 ratio of protein levels, and each BIN or subBIN is represented as a block, where each transcript is displayed as a square. Proteins are colored red if they are up-regulated in ISR (Log₂ Fold Change > 0), or blue if they are down-regulated in ISR (Log₂ Fold Change < 0)

Carbohydrate Metabolism in Rice Inferior Spikelets is Regulated by Ratooning

Carbohydrates play a crucial role in the process of grain-filling, which mainly involves starch synthesis and accumulation. Our proteomic analysis identified 14 DEPs related to sucrose-starch metabolism, with 11 proteins showing up-regulation and 3 proteins showing down-regulation in ISR compared to ISL (Fig. 2C; Table S4). The expression profiles of key proteins involved in sucrose or starch metabolism, including soluble starch synthase 2 − 1 and 3a (SSII-1, SSIIIa), sucrose synthase 1 and 4 (SUS1, SUS4), glucose-1-phosphate adenylyltransferase small subunit 2 (AGPS2), 1,4-alpha-glucan-branching enzyme (SBE1), sucrose-phosphate synthase 2 (SPS2), cytosolic invertase 1 (CINV1), cell-wall invertase (CIN1), alpha-amylase (Amy4A), and beta-amylase (BAM7) were vilified by qRT-PCR at the transcriptional level, and the results showed these proteins exhibited significant changes in ISR compared to ISL during grain-filling. During the period between 10 and 20 DAF, which corresponds to the active grain-filling phase of ISR (Fig. 1A-B), our study identified several proteins related to starch synthesis. Proteins such as SSII-1, SSIIIa, SUS1, SUS4, AGPS2, and SBE1 were found to be significantly up-regulated at varying degrees at the transcriptional level between 10 and 20 DAF and at the protein abundance level at 20 DAF, in ISR compared to ISL (Fig. 3A; Table S4). Particularly, AGPS2 showed the most pronounced increase in gene expression. Western blot analysis confirmed higher levels of AGPS2 protein of ISR compared to ISL between 5 and 20 DAF (Fig. 3F). Enzyme activity assays were also conducted on key enzymes such as sucrose synthase (SUS), glucose-1-phosphate adenylyltransferase (also known as ADP-glucose pyrophosphorylase, AGPase), soluble starch synthase (SSS), and 1,4-alpha-glucan branching enzyme (also known as starch branching enzyme, SBE) to investigate the physiological changes in inferior spikelets by rice ratooning. The activities of SUS, AGPase, SSS, and SBE were generally higher in ISR than in ISL between 10 and 20 DAF (Fig. 3B-E). Subsequent starch content assays on ISR and ISL showed a faster and more substantial accumulation of starch in ISR during grain-filling compared to ISL (Fig. 3G). These findings collectively indicate that ratooning enhances starch biosynthesis in rice inferior spikelets.

Fig. 3

Comparison of sucrose-starch metabolism between inferior spikelets of ratoon season rice (ISR) and inferior spikelets of main crop (late season) rice (ISL). A Heatmap of expression of sucrose metabolism related and starch metabolism related proteins at transcriptional level in ISR and ISL during grain-fillling. Red color indicates up-regulated genes in ISR (high Log₂ Fold Change), while blue color indicates down-regulated genes in ISR (low Log₂ Fold Change). The mRNA level in 5 DAF ISL was used as a control. B-E Enzyme activities of SUS, AGPase, SSS, and SBE in ISR and ISL during grain-filling. F Western blot analysis of AGPS2 protein abundance in ISR and ISL during grain-fillling, with heat shock protein 81 − 2 (HSP81-2) as internal control. G Starch content in ISR and ISL during grain-filling. The data are the means of three biological replications ± SD, consisting of 4 technical replications in each biological replication. Post hoc comparisons were tested using Student’s t test, *P < 0.05 or **P < 0.01

Interestingly, our proteomic analysis revealed a significant difference in trehalose metabolism, a subset of minor carbohydrate metabolism, between ISR and ISL. Specifically, two TPS family proteins (TPS1, TPS6) were up-regulated, while one TPP family protein (TPP3) was found to be down-regulated in ISR compared to ISL (Fig. 2C; Table S4). Subsequent assessment of the transcriptional levels of these proteins using qRT-PCR showed distinct expression patterns for TPS1 and TPP3 (Fig. 4A). TPS1 exhibited a notable up-regulation in ISR compared to ISL between 10 and 20 DAF, whereas TPP3 displayed the opposite trend. Enzyme activity assays revealed higher activity in ISR compared to ISL between 10 and 20 DAF, along with lower TPP activity in ISR relative to ISL during the grain-filling process (Fig. 4B-C). TPS is responsible for catalyzing the conversion of UDP-glucose and glucose-6-phosphate into T6P, while TPP functions to dephosphorylate T6P into trehalose. These findings suggest a potential variation in the T6P content of inferior spikelets in ratoon season rice. To confirm this, an ELISA-based assay was conducted, revealing higher T6P content in ISR than in ISL between 5 and 20 DAF but significantly lower T6P content in ISR compared to ISL after 25 DAF (Fig. 4D).

Fig. 4

Comparison of trehalose metabolism between inferior spikelets of ratoon season rice (ISR) and inferior spikelets of main crop (late season) rice (ISL). A Heatmap of expression of trehalose metabolism related proteins at transcriptional level in ISR and ISL during grain-fillling. Red color indicates up-regulated genes in ISR (high Log₂ Fold Change), while blue color indicates down-regulated genes in ISR (low Log₂ Fold Change). The mRNA level in 5 DAF ISL was used as a control. B-C Enzyme activities of TPS and TPP in ISR and ISL during grain-filling. D T6P content in ISR and ISL during grain-filling. The data are the means of three biological replications ± SD, consisting of 4 technical replications in each biological replication. Post hoc comparisons were tested using Student’s t test, *P < 0.05 or **P < 0.01

Rice Ratooning Influences the Indole-3-acetic acid and Cytokinin Metabolism in Inferior Spikelets

Hormones are essential for the growth and development of rice grains, affecting the processes of grain-filling and grain quality formation. This research demonstrated notable variances in hormone metabolism between ISR and ISL. A total of 33 DEPs associated with various hormones, including abscisic acid, indole-3-acetic acid (IAA), cytokinin (CTK), ethylene, gibberellin, jasmonate, and salicylic acid, were identified (Fig. 2D). Particularly noteworthy were the changes observed in IAA and CTK metabolism In ISR. Alterations in the expression levels of six proteins involved in the metabolism of IAA were observed when compared to ISL (Table S4). Specifically, IAA-amino acid hydrolases ILL9 and ILL7 exhibited up-regulation, whereas IAA-amido synthetase GH3.8 displayed down-regulation. These proteins were identified as key regulators of free IAA levels (Hayashi et al. 2021; Ding et al. 2008; Mao et al. 2019). Additionally, cytokinin dehydrogenase 11 (CKX11), responsible for the degradation of cytokinins (Zhang et al. 2020), was found to be down-regulated in ISR (Table S4). Transcriptional analysis revealed lower expression levels of GH3.8 and CKX11 in ISR compared to ISL after 10 DAF. In comparison, ILL7 and ILL9 showed higher expression levels in ISR between 5 and 20 DAF (Fig. 5A). Moreover, we also examined the expression of TAR2, a crucial gene responsible for IAA synthesis and regulated by T6P (Meitzel et al. 2021). The results indicated a greater expression of TAR2 in ISR compared to ISL between the 5–20 DAF (Fig. 5B). Analysis of hormone content through HPLC revealed a significant increase in IAA content between 5 and 20 DAF, followed by a decline after 20 DAF in ISR compared to ISL (Fig. 5C). Similarly, there was a notable increase in ZR content between 5 and 15 DAF in ISR compared to ISL, with no significant difference after 15 DAF (Fig. 5D).

Fig. 5

Comparison in indole-3-acetic acid and cytokinin metabolism between inferior spikelets of ratoon season rice (ISR) and inferior spikelets of main crop (late season) rice (ISL). A Heatmap of expression of indole-3-acetic acid metabolism related and cytokinin metabolism related proteins at transcriptional level in ISR and ISL during grain-fillling. Red color indicates up-regulated genes in ISR (high Log₂ Fold Change), while blue color indicates down-regulated genes in ISR (low Log₂ Fold Change). The mRNA level in 5 DAF ISL was used as a control. B qRT-PCR analysis of TAR2 in ISR and ISL during grain-filling. The mRNA level of TAR2 in 5 DAF ISL was used as a control. C-D indole-3-acetic (IAA) and zeatin riboside (ZR) levels in ISR and ISL during grain-filling. The data are the means of three biological replications ± SD, consisting of 4 technical replications in each biological replication. Post hoc comparisons were tested using Student’s t test, *P < 0.05 or **P < 0.01

Impact of GF14f Protein on Grain-filling and Grain Quality in Ratoon Season rice

Signal transduction proteins are essential in regulating plant growth and development. This study identified 63 DEPs involved in the signaling pathway, with 38 up-regulated and 25 down-regulated in ISR compared to ISL (Fig. 2D). These DEPs include proteins related to sugar and nutrient physiology, receptor kinases, calcium regulation, phosphoinositide, G-proteins, MAP kinases, 14-3-3 proteins, and light. Notably, two 14-3-3 family proteins, GF14f and GF14h, were down-regulated in ISR (Table S4). Our previous research indicated that rice 14-3-3 protein, specifically GF14f, has a negative impact on grain-filling (Zhang et al. 2019). qRT-PCR and Western Blot analyses showed that both gene expression and protein abundance of GF14f decreased more rapidly in ISR than in ISL as grain-filling progressed (Fig. 6A-B), suggesting that the lower level of GF14f in the ISR may contribute to improved grain-filling.

Fig. 6

Comparison in GF14f expression in inferior spikelets of ratoon season rice (ISR) compared with inferior spikelets of main crop (late season) rice (ISL) and comparison in endosperm structures in inferior spikelets of GF14f-RNAi (ISG) compared with inferior spikelets of ‘Jinhui 809’ (ISJ). A Heatmap of expression of 14-3-3 proteins at transcriptional level in ISR and ISL during grain-fillling. Red color indicates up-regulated genes in ISR (high Log₂ Fold Change), while blue color indicates down-regulated genes in ISR (low Log₂ Fold Change). The mRNA level in 5 DAF ISL was used as a control. B Western blot analysis of GF14f protein abundance in ISR and ISL during grain-fillling, with heat shock protein 81 − 2 (HSP81-2) as internal control. C Scanning electron microscopy images of cross-sections in the endosperm of ISG and ISJ at their maturity stage. The data are the means of three biological replications ± SD, consisting of 4 technical replications in each biological replication. Post hoc comparisons were tested using Student’s t test, *P < 0.05 or **P < 0.01

To investigate the impact of GF14f protein on the grain quality of ISR, transgenic plants (GF14f-RNAi) were utilized, which were obtained from a previous study conducted by our group (Zhang et al. 2019). A thorough assessment of grain quality was carried out, as shown in Table 3. Scanning electron microscopy analysis was conducted on the inferior spikelets of both the transgenic material and its wild-type counterpart (Jinhui 809) (Fig. 6C). In terms of processing quality, the brown rice yield of inferior spikelets of GF14f-RNAi (ISG) was 3.90% higher than that of inferior spikelets of ‘Jinhui 809’ (ISJ). Similarly, the milled rice and head rice percentages in ISG increased by 4.35% and 5.72%, respectively, compared to ISJ. In terms of appearance quality, ISG exhibited a chalkiness degree 2.58% lower than ISJ, along with an 11.24% decrease in chalky grain percentage. Furthermore, ISG demonstrated superior taste value in terms of nutritional quality compared to ISJ, with a 1.14% decrease in protein content, a 0.55% decrease in amylose content, and a 0.25% increase in fat content. Notably, significant disparities were observed in the endosperm structure between ISG and ISJ, with the endosperm structure of ISJ displaying a loose arrangement in the belly, chest, and back regions, characterized by distinct gaps.

Table 3 Grain quality of inferior spikelets of GF14f-RNAi and ‘Jinhui 809’

Discussion

The rice ratoon cropping system, also known as a ‘one planting and two harvests’ cultivation mode, involves regenerating a crop from dormant buds in the rice stubbles of the harvested main crop. This system can impact both the grain yield and grain quality of rice. In this research, two different rice varieties, ‘Jinhui 809’ and ‘Yongyou 1540’, were chosen due to their distinct yield performances under the ratoon cropping system. This study aimed to analyze the changes in grain-filling that occur as a result of rice ratooning. To eliminate the influence of environmental variables on grain-filling characteristics, ratoon season rice and main crop (late season) rice with synchronized heading times were used as experimental materials. Furthermore, the conventional rice variety ‘Jinhui 809’ was subjected to proteomic analysis to uncover the molecular mechanisms behind these changes. The conclusions of this study are as follows.

The Grain-filling of Rice Spikelets, Especially for Inferior Spikelets, can be Improved by Ratooning Technology

This research conducted a comprehensive examination of the morphology and grain-filling dynamics of ISR and ISL. Findings revealed that ISR began rapid grain-filling and achieved peak rate sooner compared to ISL. This disparity was particularly pronounced in the rice variety ‘Jinhui 809’. Additionally, rice ratooning was found to enhance seed-setting percentage in ISL, particularly in ‘Jinhui 809’. Nevertheless, this practice also led to a reduction in the 1000-grain weight for both ISR and ISL. In terms of physiological characteristics, ratoon season rice, in comparison to main crop (late season) rice, has a lower leaf area index and grain sink capacity but a higher grain number to leaf area ratio (Huang et al. 2020). Moreover, ratoon season rice exhibits a faster aging rate of the entire plant during the late grain-filling stage compared to main crop (late season) rice (Wu et al. 2023). These physiological traits are advantageous in improving the photosynthate remobilization ability in ratoon season rice, thereby promoting a rapid and consistent grain-filling process.

The present study also investigated the difference in grain quality between ISR and ISL. Compared to ISL, the endosperm structure in ISR was found to be more tightly arranged and exhibit less chalkiness. The presence of chalkiness is primarily attributed to inadequate starch granule accumulation, leading to a lower level of compactness. This deficiency can result in decreased head rice yield and have a detrimental impact on the appearance quality and taste of the rice (Fei et al. 2023; Fitzgerald et al. 2009; Lisle et al. 2000). The findings of this study suggest ratoon season rice exhibits higher grain-filling efficiency compared to main crop (late season) rice, despite a decrease in final grain weight. Furthermore, nutritional analysis demonstrated that ISR exhibited lower protein and amylose levels but higher fat content compared to ISL. The reduction in protein may lead to reduced stickiness, while the decrease in amylose could enhance cooking quality, and the increase in fat may improve flavor (Deng et al. 2021; Hamaker and Griffin 1993; Liu et al. 2021; Zhang et al. 2009). Notably, prior research has shown an increase in amylose content in grains of ratoon season rice in comparison to that of the main crop (early season) rice (Deng et al. 2021). However, our results showed a decrease in amylose content in grains of ratoon season rice compared to that of main crop (late season) rice, possibly due to environmental differences between the two comparisons. The previous research demonstrated a notable enhancement in the grain quality of ratoon season rice relative to the main crop (early season) rice (Deng et al. 2021; Wu 2005). In specific rice-growing regions suitable for ratoon rice cultivation, such as southern China, the daily mean temperature during the grain-filling stage of the main crop (early season) is consistently higher than that of the ratoon season, which may impose a significant temperature stress on grain-filling and potentially hinder the formation of grain quality (Lin et al. 2022; Shen et al. 2022). In the present study, we compared ratoon season rice with main crop (late season) rice that had synchronized heading times to eliminate environmental influences. The results demonstrated that the grain quality of ratoon season rice also showed improvement compared to main crop (late season) rice, particularly in terms of head rice yield and chalkiness. This suggests that the ratooning technology can significantly enhance the grain quality of rice.

Rice Ratooning Improves Grain-filling and Grain Quality through Changes in Starch Synthesis and Trehalose Metabolism

Starch serves as the primary energy source in rice grains, with its synthesis and accumulation being crucial processes during grain-filling. This research identified several key proteins associated with starch synthesis, such as SUS1, SUS4, AGPS2, SSII-1, SSIIIa, and SBE1, which exhibited significant up-regulation at both the transcriptional and protein abundance levels in inferior spikelets of ratoon season rice compared to main crop (late season) rice. Among these proteins, AGPS2 displayed the most notable up-regulation in ISR, supported by both qRT-PCR and western blot analysis. AGPS2 is known to play a vital role in starch synthesis and can result in shriveled endosperms with low expression (Lee et al. 2007). Moreover, it was observed that during the active grain-filling phase of ISR (10–20 DAF), the enzyme activities of SUS, AGPase, SSS, and SBE were generally higher in ISR compared to ISL, providing additional evidence for the up-regulation of these key proteins. These four enzymes play critical roles in sucrose-starch conversion. Following transportation to sink organs (grains), sucrose undergoes initial degradation by invertase or SUS. Sucrose degradation in sink organs, such as grains, is initiated by invertase or SUS enzymes. SUS plays a critical role in the early stages of sucrose conversion to starch in grains, and its activity is closely linked to the rate of grain-filling (Huang et al. 2020; Kato 1995; Sun et al. 1992; Wang et al. 1993). AGPase is responsible for catalyzing the formation of ADP-glucose, a key molecule in rice starch synthesis. The enzymes SSS and SBE are crucial for determining grain quality, especially in the synthesis of starch in rice endosperm amyloplasts (Naoko et al. 2009; Li et al. 1997). The heightened enzymatic activity of these four enzymes facilitates starch synthesis and accumulation while concurrently inhibiting chalkiness formation, thereby enhancing the grain-filling process and ultimately leading to increased seed-setting percentage and improved grain quality in ratoon season rice.

Furthermore, a significant alteration in trehalose metabolism was observed in this study. Numerous studies, including research conducted by our team, have demonstrated the positive impact of trehalose on promoting grain-filling in rice (Min et al. 2021; Huang et al. 2023). Nevertheless, a growing body of evidence suggests that T6P, the precursor metabolite of trehalose, plays a more crucial role in influencing rice grain-filling. In this study, two proteins related to TPS, TPS1 and TPS6, showed up-regulation, while one protein associated with TPP, TPP3, exhibited down-regulation in rice inferior spikelets of the ratoon season relative to the main crop (late season). Previous studies have highlighted the functional roles of TPS1 and TPP3 genes in trehalose metabolism in rice (Zang et al. 2011; Ye et al. 2023). The findings regarding TPS1 and TPP3 gene expressions, as well as TPS and TPP activities, were consistent with the variations in protein abundance of TPS1 and TPP3. TPS synthesizes T6P from glucose-6-phosphate and UDP-glucose, which TPP can then dephosphorylate to form trehalose. These results led to further investigation into the changes in T6P levels. Researchers have characterized T6P as a key signaling molecule for monitoring sugar levels in plants, playing a role in regulating the utilization of energy sources. A study by Lunn et al. (2006) has shown that T6P is associated with the redox activation of AGPase pyrophosphorylase and increased rates of starch synthesis. Similarly, Martinez-Barajas et al. (2011) demonstrated that elevated levels of T6P facilitate the efficient conversion of sucrose into starch during the early grain-filling stage. The results of our study revealed a significant increase in T6P content in ISR compared to ISL during the active grain-filling phase of ISR. Therefore, we propose that the heightened T6P level in inferior spikelets of ratoon season rice during its active grain-filling phase may play a crucial role in enhancing grain-filling by modulating starch synthesis rates.

Influence of rice Ratooning on Indole-3-acetic acid and Cytokinin Metabolic Pathways in Inferior Spikelets

Hormones are involved in regulating the growth, development, and stress responses of plants. They also play a crucial role in the grain-filling process of rice, serving as key regulatory factors in rice yield formation (Chen et al. 2022; Ma et al. 2023; Waadt et al. 2022). In this study, the IAA and CTK metabolic pathways of rice inferior spikelets were significantly influenced by ratooning. Previous research has indicated that ILR1/ILL (ILR1-like) genes play a role in releasing free IAA from its storage forms (IAA-Asp and IAA-Glu), whereas GH3s genes are responsible for converting free IAA back into its storage forms in Arabidopsis thaliana. These genes work together to regulate the balance of IAA levels (Hayashi et al. 2021). This study revealed the up-regulation of ILL9 and ILL7 proteins and the down-regulation of GH3.8 protein in rice inferior spikelets of ratoon season relative to the main crop (late season), consistent with gene expression results. GH3.8 is known to catalyze the synthesis of IAA-amino acid conjugates, inhibiting free IAA accumulation in rice (Ding et al. 2008; Mao et al. 2019). These findings led to the hypothesis that reduced GH3.8 protein levels in ISR could increase IAA levels, supported by elevated IAA content in ISR during its active grain-filling phase. Additionally, the down-regulation of the CKX11 protein in ISR compared to ISL was observed. CKX11 is known to be responsible for the degradation of various cytokinins in rice (Zhang et al. 2020). Similarly, lower gene expression of CKX11 and higher ZR (a prevalent type of CTK in plants) content in ISR was detected during its active grain-filling phase.

IAA and CTK are known to influence grain-filling by regulating the formation or proliferation of endosperm cells. The activities of SUS, AGPase, and SSS are significantly enhanced by exogenous IAA or CTK treatments (Chen et al. 2022). Our results suggest that increased levels of IAA and CTK in inferior spikelets of ratoon season rice during its active grain-filling phase positively impact grain-filling and promote starch synthesis, ultimately enhancing grain quality. Changes in the abundance of GH3.8 and CKX11 proteins lead to variations in IAA and CTK contents, respectively. Additionally, alterations in the abundance of ILL9 and ILL7 proteins may also play a role in modifying IAA content.

Moreover, T6P is closely linked to IAA, with Meizel et al. (2021) demonstrating that T6P activates the IAA biosynthetic pathway through the regulation of TAR2 expression. Teng et al. (2022) observed increased T6P and IAA levels, as well as up-regulated TAR2 expression in inferior spikelets during grain-filling after moderate soil drying. This study also highlighted higher T6P, IAA, and TAR2 levels during the active grain-filling phase of ISR. However, our proteomic results did not detect the TAR2 protein, suggesting a need for further research to explore its response to rice ratooning.

GF14f Protein is an Essential Regulator for the Improvement of Grain-filling of Ratoon Season Rice

In this study, GF14f, a distinct member of the 14-3-3 protein family, was identified as exhibiting a more pronounced decrease in protein abundance in ISR than in ISL during the grain-filling process. Prior studies have indicated the involvement of the 14-3-3 family in the regulation of grain-filling and grain size in rice (Li et al. 2023; Song et al. 2023; Zhang et al. 2019). Our previous research has also demonstrated that GF14f can interact with key enzyme proteins involved in starch synthesis, including granule-bound starch synthase 1, sucrose synthase 4, and glucose-1-phosphate adenylyltransferase small subunit 1. The reduction in GF14f expression in the grain endosperm results in increased activities of SSS, SUS, and AGPase, ultimately enhancing starch synthesis in the grain (Zhang et al. 2019). In this study, the activities of starch synthetic enzymes (SUS, AGPase, SSS, and SBE) in ISR were found to be higher than those in ISL, potentially due to the swift decline in GF14f protein levels in the ISR. Furthermore, GF14f-RNAi plants specific to the endosperm were utilized to evaluate grain quality and the results revealed improvements in the processing, appearance, and nutritional quality of inferior spikelets in GF14f-RNAi plants when compared to wild-type rice. Scanning electron microscopy analysis showed a more tightly packed endosperm structure in GF14f-RNAi grains. These findings suggest that the rapid reduction in GF14f protein levels in ISR plays a significant role in improving the grain quality of ratoon season rice.

Our previous study on GF14f-RNAi plants revealed a notable increase in the TPS1 protein abundance, which was identified as a potential client of the GF14f protein (Zhang et al. 2019). Surprisingly, our research showed that the TPS1 protein had a higher abundance in ISR compared to ISL. Considering the expression trend of GF14f in ISR, we hypothesize that GF14f is a significant factor contributing to the increased abundance of TPS1 protein expression in the grains of the ratoon season relative to the main crop (late season). Previous research has indicated that among the 11 TPS family members identified in rice, only TPS1 encodes an active TPS (Zang et al. 2011). Additionally, several studies have demonstrated that 14-3-3 proteins regulate key enzyme activities in plant metabolism through a phosphorylation-binding mechanism (De Boer et al. 2013). For instance, Bachmann et al. (1996) showed that nitrate reductase is phosphorylated by SnRK1 kinase in the dark, leading to its interaction with 14-3-3 protein and a subsequent decrease in enzymatic activity. Studies on Arabidopsis thaliana have revealed that the Ser22 and Thr49 sites of the TPS5 member of the TPS protein family can bind with the 14-3-3 protein upon phosphorylation by AMP-activated protein kinase (AMPK) and SNF1-related protein kinase (SnRK1) (Harthill et al. 2006). Therefore, it is hypothesized that the GF14f protein may negatively regulate TPS enzyme activity by interacting with the TPS1 protein through a phosphorylation mechanism, thereby influencing T6P content in ratoon season grains. However, further verification is still needed.

Conclusions

In summary, this study revealed a significant change in the proteomic profile of inferior spikelets induced by rice ratooning, accompanied by increased levels of starch synthesis, T6P, IAA, and ZR during the active grain-filling phase, as well as a decrease in the abundance of GF14f protein during almost the grain-filling process. These changes collectively represent a complex regulatory mechanism intended to enhance the grain-filling process of inferior spikelets by rice ratooning, with GF14f identified as a key player in this process (Fig. 7). For future rice breeding, manipulating the abundance of key proteins in grain endosperm might be exploited to improve nutritional quality and yield. However, an in-depth molecular mechanism associated with this process needs to be further investigated.

Fig. 7

The regulatory framework about the improved grain-filling of inferior spikelets by rice ratooning. The contents in boxes represent proteins. The proteins in the green boxes may be influenced by the protein GF14f. The content on the lines represent genes. Red color indicates increased levels of protein abundance, gene expression, or substance content, while blue indicates the opposite. Lines with arrows indicate positive effects, lines with slanted dashes indicate negative effects, and the dotted line indicates an effect that lacks additional confirmation. Big arrows indicate the final result