Metabolomic Profiling of Heat Tolerance during Grain Filling in Rice: Comparative Analyses of Panicles and Roots in ‘Fusaotome’ and ‘Akitakomachi’

1. Introduction

Global warming has exerted a profound impact on global food production, and, in particular, high temperatures from the heading stage to grain filling substantially reduce both yield and quality in rice crops. In warm regions of Japan, deterioration of brown rice quality—most notably an increase in chalkiness—has been widely documented, with significant rises in chalky grains reported alongside rising air temperatures [1,2].

Two principal physiological constraints underlie heat-induced grain quality deterioration: (1) a shortage of photoassimilate supply to the endosperm (source limitation) and (2) reduced endosperm capacity for starch biosynthesis (sink limitation), which together lead to insufficient starch accumulation and consequently inferior grain appearance and quality [3,4]. With ongoing climate change, these drivers are likely to intensify, warranting elucidation of the physiological mechanisms that confer tolerance to high temperature during the heading-to-grain-filling period urgently.

To decipher heat stress responses, omics approaches, including metabolomics, have been increasingly employed in recent years. Yamakawa and Hakata [4] showed, through the integration of grain metabolome and transcriptome datasets, that high temperature suppresses the expression of starch biosynthetic genes while inducing amino acid accumulation. Proteomic analyses have likewise indicated conspicuous heat-induced changes in enzymes involved in starch metabolism and stress-responsive proteins [5]. Moreover, high temperature—especially when associated with elevated night temperatures—broadly perturbs central metabolism (glycolysis and the tricarboxylic acid (TCA) cycle), amino acid metabolism, and polyamine-related metabolism [4,6]. Additionally, heat stress increases the production of reactive oxygen species (ROS), triggering oxidative stress. Plants maintain redox homeostasis through antioxidant enzymes and low-molecular-weight antioxidants [7], and enhanced antioxidant enzyme activities may contribute to heat tolerance in rice [8,9].

However, most previous studies have focused on aboveground organs (panicles and leaves), where heat injury is most directly observed. In contrast, less is known about roots, which are metabolically active under environmental stress and indirectly support photosynthesis and starch deposition by supplying water and nutrients. Heat has been reported to impair root system formation, respiration, and water and nutrient uptake in rice [10,11].

To address this gap, we compared the metabolite dynamics in panicles and roots under high-temperature conditions between a heat-tolerant cultivar, ‘Fusaotome’, and a widely cultivated Japanese cultivar, ‘Akitakomachi’. Specifically, we established control and high-temperature treatments, sampled tissues at heading, and conducted comprehensive metabolomic profiling using capillary electrophoresis–mass spectrometry (CE-MS). Our objectives were to clarify (1) the impacts of high temperature on metabolite levels in panicles and roots and (2) cultivar-specific strategies for pathway activation. The outcomes provide a physiological basis for heat tolerance during grain filling and foundational knowledge for developing metabolic markers to support breeding programs and refine cultivation practices.

2. Materials and Methods

2.1. Plant Materials and Cultivation

Experiments were conducted at the experimental field of the Akita Prefectural University (140°05′E, 39°80′N) in 2022. Two cultivars were used: the heat-tolerant ‘Fusaotome’ and the widely cultivated Japanese cultivar ‘Akitakomachi’. Seeds were sown on April 28, 2022, in a growth chamber using Inaho granular soil (Inaho-Kako Co., Ltd., Toyama, Japan) and grown as seedlings. On May 25, 2022, single seedlings were transplanted into 1/5000-a Wagner pots filled with 3 L of Inaho granular soil. Slow-release fertilizers (Ecolong Total 391-70 and 391-140; National Federation of Agricultural Cooperative Associations, Tokyo, Japan) were each added at 0.6 g per pot. Plants were cultivated under flooded conditions. Three biological replicates were prepared for each treatment and cultivar.

2.2. Experimental Treatments

High-temperature treatment began on July 30, 2022, when pots were moved into a vinyl greenhouse located within the experimental field. The treatment was maintained until sampling. High-temperature treatments were conducted inside the greenhouse, while controls were maintained outdoors under ambient conditions. Air temperature in each treatment was recorded hourly using Em5b+RT-1 sensors (Decagon Devices, Pullman, WA, USA).

2.3. Sampling

Panicles and roots were sampled immediately after heading. Sampling dates for ‘Akitakomachi’ were August 9 (high temperature) and August 16, 2022 (control), and August 16 (high temperature) and August 22 (control) for ‘Fusaotome’. Panicles were immediately frozen in liquid nitrogen upon harvest. Roots were carefully washed within 5 min and frozen in liquid nitrogen. Frozen samples were lyophilized, pulverized, and subjected to metabolomic analysis.

2.4. Metabolomic Analysis

Metabolomic analysis was outsourced to Human Metabolome Technologies, Inc. (Tsuruoka, Yamagata, Japan). Approximately 10 mg of lyophilized tissue was extracted with 600 μL of methanol containing 50 μM internal standards, and homogenized on ice using a cell disruptor (BMS-M10N21; Bio Medical Science Inc., Tokyo, Japan) (4 °C, 1500 ×g, 2 min × 2). Chloroform (500 μL) and Milli-Q water (200 μL) were added, followed by centrifugation (2,300 ×g, 4 °C, 5 min). The aqueous phase was transferred to an ultrafiltration tube (Ultrafree-MC PLHCC HMT 5 kDa; Millipore) and centrifuged (9,100 ×g, 4 °C, 120 min). Filtrates were dried and reconstituted in 50 μL of Milli-Q water. Samples were analyzed by CE–TOF–MS (Agilent CE–TOF–MS; Agilent, Palo Alto, CA, USA) as described in Soga, et al. [12].

2.5. Data Analysis

For metabolite datasets, principal component analysis (PCA) and t-tests were performed using JMP 8 (SAS Institute Inc., Cary, NC, USA). Pathway analysis was conducted with MetaboAnalyst 6.0 (https://www.metaboanalyst.ca/), and pathway impact values were calculated [13].

3. Results

3.1. Temperature Conditions in Treatments

Temporal changes in air temperature during the experimental period are shown in Figure 1. During daytime under the treatment, the mean temperatures in ‘Akitakomachi’ were 26.5 °C (control) and 31.8 °C (high temperature), and those in ‘Fusaotome’ were 26.7 °C (control) and 30.5 °C (high temperature). Thus, the high-temperature treatments in ‘Akitakomachi’ and ‘Fusaotome’ were +5.3 °C and +3.8 °C above the controls, respectively (Table 1). The nighttime means were 22.3 °C (control) and 24.6 °C (high temperature) for ‘Akitakomachi’, and 22.1 °C (control) and 24.8 °C (high temperature) for ‘Fusaotome’, i.e., +2.3 °C and +2.7 °C relative to the control, respectively. These differences confirm that the high-temperature treatment was properly implemented.

3.2. Effects of High Temperature on Metabolite Profiles in Panicles and Roots

We detected 312 metabolites in panicles and 262 in roots. PCA was used to visualize treatment- and cultivar-dependent separations (Figure 2). In panicles, control and high-temperature samples separated clearly along PC2, indicating pronounced heat-induced metabolic shifts. Cultivar differences under high temperature were also evident between ‘Fusaotome’ and ‘Akitakomachi’. In the roots, ‘Akitakomachi’ showed clear control vs. high-temperature separation along PC2, whereas ‘Fusaotome’ separated along PC1 but with a smaller magnitude of difference.

3.3. Metabolite-Level Responses to High Temperature

Metabolites differing significantly between control and high temperature were extracted and visualized by volcano plots (Figure 3). Compounds showing significant changes at p < 0.05 with ≥1.5-fold increases or ≤1/1.5-fold decreases are summarized in Table 2 and Table 3.

In the panicles of ‘Fusaotome’, 8 metabolites increased ≥1.5-fold (mainly amino acid metabolism-related), while 11 decreased to ≤1/1.5-fold, mostly associated with amino acid, carbohydrate, and energy metabolism (Table 2). In ‘Akitakomachi’ panicles, 19 metabolites increased ≥1.5-fold, primarily ornithine and aromatic amino acid-related compounds, whereas 6 decreased to ≤1/1.5-fold, many of which were implicated in antioxidant capacity and stress responses.

In the roots, 12 metabolites increased ≥1.5-fold in ‘Fusaotome’, mainly related to nitrogen and amino acid metabolism (Table 3). In ‘Akitakomachi’, nine metabolites increased ≥1.5-fold, including amino acid derivatives, organic acids, and antioxidant-related compounds. No metabolites decreased significantly to ≤1/1.5-fold in the roots of either cultivar.

3.4. Pathway-Level Responses to High Temperature

Based on differential metabolites, pathway analysis using the KEGG database was performed. For each cultivar and organ, relationships between pathway impact and −log₁₀ (p-value) are shown (Figure 4). A pathway “score” was computed as the product of impact and −log₁₀ (p-value), and inter-cultivar differences were evaluated.

In ‘Fusaotome’ panicles, tryptophan metabolism, nicotinate and nicotinamide metabolism, arginine and proline metabolism, glycolysis and the TCA cycle, pyruvate metabolism, and vitamin B6 metabolism exhibited high scores (Table 4). These pathways are associated with antioxidant responses and energy supply, suggesting that ‘Fusaotome’ maintains metabolic homeostasis even under high temperature. In ‘Akitakomachi’ panicles, phenylpropanoid biosynthesis, isoquinoline alkaloid biosynthesis, caffeine metabolism, and ubiquinone/terpenoid–quinone biosynthesis showed high scores (Table 5), largely representing secondary metabolism with an emphasis on defensive responses.

In ‘Fusaotome’ roots, phenylalanine metabolism and phenylpropanoid biosynthesis, tyrosine/tryptophan-related routes, lysine degradation, branched-chain amino acid metabolism (valine, leucine, isoleucine), glycerophospholipid metabolism, and alkaloid biosynthesis showed high scores (Table 6). These pathways are related to antioxidant defense, cell wall reinforcement, carbon–nitrogen reutilization, and membrane stabilization. In ‘Akitakomachi’ roots, alanine, aspartate, and glutamate metabolism; cysteine and methionine metabolism; tryptophan metabolism; arginine and proline metabolism; taurine and hypotaurine metabolism; β-alanine metabolism; and cyanoamino acid metabolism were enriched (Table 7), indicating a bias toward nitrogen metabolism and antioxidant-related pathways.

4. Discussion

We compared metabolite and pathway dynamics in panicles and roots of the heat-tolerant ‘Fusaotome’ and the commonly cultivated ‘Akitakomachi’ under high-temperature conditions.

4.1. Metabolite-Level Responses to High Temperature

Both cultivars showed changes in amino acids and nitrogen-containing compounds, although the direction and magnitude differed markedly. In ‘Fusaotome’ panicles (Table 2), polyamines and amino-acid-related metabolites such as spermidine and N-acetylalanine were elevated, likely contributing to antioxidant defense and membrane stabilization. Polyamines are known to modulate membrane stability and ROS homeostasis, broadly promoting abiotic stress tolerance [14,15]. Exogenous spermidine has been shown to increase antioxidant enzyme activities (SOD, POD) in heat-stressed rice, mitigating oxidative damage—supporting the role of polyamines as ROS scavengers and as inducers of polyamine biosynthetic and antioxidant gene expression [16,17,18]. In contrast, although ornithine and anthranilic acid increased in ‘Akitakomachi’, declines in antioxidant-related precursors, including glutathione derivatives, suggested compromised antioxidant capacity. Given the central role of the glutathione system in plant redox homeostasis [19], ‘Akitakomachi’ may have insufficient antioxidant control under high-temperature stress.

Root responses also differed (Table 3). ‘Fusaotome’ accumulated lysine, arginine, proline, and leucine, with proline playing an important role in osmotic adjustment and ROS quenching under heat and drought conditions [20,21]. Modulations in proline metabolism can affect heat tolerance [22]. Oxidized glutathione (GSSG) accumulated in ‘Akitakomachi’; the increased GSSG levels under oxidative stress [23] indicates reduced antioxidant control during heat stress.

4.2. Pathway-Level Responses to High Temperature

Pathway analysis revealed distinct metabolic strategies between cultivars in both organs. In ‘Fusaotome’ panicles (Table 4), tryptophan metabolism (providing precursors for melatonin), nicotinate/nicotinamide metabolism (supporting NAD⁺/NADP⁺ supply), and arginine/proline metabolism were prominent, facilitating simultaneous antioxidant responses and energy stability [20,21,24,25]. Glycolysis and the TCA cycle were also emphasized, providing a basal metabolic foundation for sustain grain filling under heat stress [4]. In ‘Akitakomachi’ panicles (Table 5), enrichment of secondary pathways (phenylpropanoid and isoquinoline alkaloid biosynthesis) reflected defense-oriented shifts that may be less directly supportive of basal metabolism and sustained grain filling [26].

In the roots, ‘Fusaotome’ exhibited prominent aromatic amino acid metabolism and phenylpropanoid biosynthesis (Table 6), promoting the synthesis of cell wall components (e.g., lignin) and antioxidant compounds [27]. Lysine and branched-chain amino acid degradation may remodel carbon and nitrogen to maintain metabolic homeostasis under stress [28,29]. Moreover, activation of glycerophospholipid metabolism and alkaloid biosynthesis may stabilize membranes and enhance chemical defense [30,31], thereby preserving root function. In ‘Akitakomachi’ (Table 7), nitrogen metabolism-related pathways were predominantly induced, whereas activation of aromatic amino acid metabolism and membrane lipid pathways was limited, indicating a more defensive but less integrative metabolic response at the whole-plant level.

5. Conclusions

This study compared the metabolic responses of the heat-tolerant cultivar “Fusaotome” and the conventional cultivar “Akitakomachi” under high-temperature conditions, revealing distinct metabolic strategies in panicles and roots. These results enhance our understanding of the molecular mechanisms underlying heat tolerance and highlight the potential for developing metabolic markers using specific metabolites or pathways as indicators. Our findings are expected to provide a foundation for breeding and cultivation strategies aimed at ensuring stable rice production under climate change. Future work should focus on quantitatively validating the key metabolites identified in this study and integrating them with transcriptome analysis to construct a more comprehensive model of heat tolerance.