Metabolic Dynamics of Developing Rice Seeds Under High Night-Time Temperature Stress

Abstract

High temperature stress during rice reproductive development results in yield losses. Reduced grain yield and grain quality has been associated with high temperature stress, and specifically with high night-time temperatures (HNT). Characterizing the impact of HNT on the phenotypic and metabolic status of developing rice seeds can provide insights into the mechanisms involved in yield and quality decline. Here, we examined the impact of warmer nights on the morphology and metabolome during early seed development in six diverse rice accessions. Seed size was sensitive to HNT in four of the six genotypes, while seed fertility and seed weight were unaffected. We observed genotypic differences for negative impact of HNT on grain quality. This was evident from the chalky grain appearance due to impaired packaging of starch granules. Metabolite profiles during early seed development (3 and 4 days after fertilization; DAF) were distinct from the early grain filling stages (7 and 10 DAF) under optimal conditions. We observed that accumulation of sugars (sucrose, fructose, and glucose) peaked at 7 DAF suggesting a major flux of carbon into glycolysis, tricarboxylic acid cycle, and starch biosynthesis during grain filling. Next, we determined hyper (HNT > control) and hypo (HNT < control) abundant metabolites and found 19 of the 57 metabolites to differ significantly between HNT and control treatments. The most prominent changes were exhibited by differential abundance of sugar and sugar alcohols under HNT, which could be linked to a protective mechanism against the HNT damage. Overall, our results indicate that combining metabolic profiles of developing grains with yield and quality parameters under high night temperature stress could provide insight for exploration of natural variation for HNT tolerance in the rice germplasm.

Keywords: early seed development; grain filling; high night-time temperature; metabolite profiling; rice; starch.

Figure 1

Fresh and dry weight analysis of developing seeds. Florets were marked at the time of fertilization to track the precise developmental timing. One day after fertilization (DAF), plants were either kept in control conditions or moved to greenhouse with adjusted HNT. Only the marked developing seeds at 4, 7, and 10 days after fertilization were harvested and analyzed. (A) Fresh weight was measured immediately after harvesting. (B) For dry weights, the seeds were subjected to drying for 7 days at 40 ˚C prior to measurements. For fresh and dry weight measurements (in mg), sum of ten developing seeds derived from three plants for the respective time-point and treatment was considered. Data is represented as mean ± standard deviation from three independent replicates. Student’s t-test was used for the statistical analysis (***p < 0.001, **p < 0.01, *p < 0.05). HNT, high night-time temperature.

Figure 2

Analysis of mature seeds from control and high night-time temperature treated plants. (A) Percentage of fully developed mature seeds. (B) Weight per seed (in mg). For this, only the marked florets at the time of fertilization were considered (n = 300 – 600) at maturity. One DAF, plants were either kept in control conditions or moved to greenhouse with adjusted HNT until maturity. Data is represented as mean ± standard deviation from eight plants. Student’s t-test was used for statistical analysis (* p < 0.01). HNT: high night-time temperature.

Figure 3

High night-time temperature alters seed quality traits. (A) Mature seed images from six rice genotypes subjected to control and HNT conditions (scale: 1 cm). Light box was used to take these images. (B) Representative images of scanning electron microscopy (SEM) of mature seed from three independent biological replicates (scale: 20 μm). HNT, high night-time temperature.

Figure 4

Gene expression analysis of key starch biosynthesis enzymes under HNT. RT-qPCRs representing expression for selected genes related to starch biosynthesis in developing seeds (4, 7, and 10 DAF) corresponding to four genotypes. The relative expression values indicate ratio of HNT to control for the respective genotype and developmental timepoint. For statistics, paired t-test was used to compare expression levels for each gene under HNT relative to control for the respective genotype and developmental timepoint. Error bars indicate standard deviation from three biological and technical replicates. ***indicates p < 0.001 and **p < 0.01. DAF, days after fertilization; HNT, high night-time temperature.

Figure 5

Principal component analysis (PCA) of the metabolite profiles of rice seed development under control and high night-time temperature conditions. Orange, green, blue, and violet colors represent 3, 4, 7, and 10 days after fertilization (DAF). Circle and triangle represent metabolome for control samples and high night temperature treated samples. Five independent biological replicates were used for the metabolomic analysis. Each point corresponds to one replicate from each of the six rice genotypes. HNT, high night-time temperature.

Figure 6

Temporal trends in the levels of metabolites across early seed development. A heatmap showing the changes in levels of metabolites during the early seed development in the control condition. The ratio of the median of normalized peak area at each time point against the initial time point (3 DAF) was log2 transformed to calculate relative metabolite levels. Hierarchical clustering analysis was performed using Euclidean distance to group metabolites with similar temporal patterns. Directional patterns represent statistically significant changes in metabolite level relative to metabolite level at previous timepoint using generalized linear models, using a cutoff in adjusted p-value at 0.05 ( Table S3 ), which are represented by colored ellipses to the right of the heatmap.

Figure 7

Effects of high night-time temperature on the levels of metabolites across early seed development. A heatmap showing the effects of HNT on the levels of individual metabolites during the early seed development. The ratio of the median of normalized peak area under HNT against control conditions at each time point was log₂ transformed to calculate relative metabolite levels. Hierarchical clustering analysis was performed using Euclidean distance to group metabolites similarly affected by HNT treatment. Statistically significant difference in metabolite levels between HNT and control conditions was determined by meta-analysis, using a cutoff in adjusted p-value at 0.05 ( Table S4 ) and was denoted by a +, ^ indicates a metabolite which was not statistically tested due to inconsistency among genotypes. HNT, high night-time temperature.

Figure 8

Gene expression analysis of genes associated with selected metabolites under HNT. RT-qPCRs representing expression for selected genes: aspartate aminotransferase (AAT), sugar transporter (STP14), branched chain amino acid transaminase 2 (BCAT2), and shikimate kinases (SK1, SK2, and SK3) in developing seeds (4, 7, and 10 DAF) corresponding to four genotypes. The relative expression values indicate ratio of HNT to control for the respective genotype and developmental timepoint. For statistics, paired t-test was used to compare expression levels for each gene under HNT relative to control for the respective genotype and developmental timepoint. Error bars indicate standard deviation from three biological and technical replicates. ***indicates p < 0.001, **p < 0.01 and *p < 0.05. DAF, days after fertilization; HNT, high night-time temperature.

Figure 9

Abundance levels for selected metabolites. Mean of the abundance levels from six rice genotypes at the respective seed developmental stage under control and HNT (other metabolites are shown in Figure S2 ). Data is represented as mean ± standard deviation from five biological replicate for each genotype and treatment. HNT, high night-time temperature.