Physiological Profiling of Embryos and Dormant Seeds in Two Arabidopsis Accessions Reveals a Metabolic Switch in Carbon Reserve Accumulation

Abstract

In flowering plants, sugars act as carbon sources providing energy for developing embryos and seeds. Although most studies focus on carbon metabolism in whole seeds, knowledge about how particular sugars contribute to the developmental transitions during embryogenesis is scarce. To develop a quantitative understanding of how carbon composition changes during embryo development, and to determine how sugar status contributes to final seed or embryo size, we performed metabolic profiling of hand-dissected embryos at late torpedo and mature stages, and dormant seeds, in two Arabidopsis thaliana accessions with medium [Columbia-0 (Col-0)] and large [Burren-0 (Bur-0)] seed sizes, respectively. Our results show that, in both accessions, metabolite profiles of embryos largely differ from those of dormant seeds. We found that developmental transitions from torpedo to mature embryos, and further to dormant seeds, are associated with major metabolic switches in carbon reserve accumulation. While glucose, sucrose, and starch predominantly accumulated during seed dormancy, fructose levels were strongly elevated in mature embryos. Interestingly, Bur-0 seeds contain larger mature embryos than Col-0 seeds. Fructose and starch were accumulated to significantly higher levels in mature Bur-0 than Col-0 embryos, suggesting that they contribute to the enlarged mature Bur-0 embryos. Furthermore, we found that Bur-0 embryos accumulated a higher level of sucrose compared to hexose sugars and that changes in sucrose metabolism are mediated by sucrose synthase (SUS), with SUS genes acting non-redundantly, and in a tissue-specific manner to utilize sucrose during late embryogenesis.

Keywords: carbon; embryo development; hexoses; metabolites; sucrose; sucrose synthase.

Figure 1

Burren-0 (Bur-0) accession has bigger seeds and mature embryos. (A,B) Mature dried seeds of Arabidopsis thaliana natural accessions (A) Columbia-0 (Col-0) and (B) Bur-0. (C) Seed area. (D) Seed length and width. (E) The ratio of length to width. (F) Seed weight of 100 mature dried seeds (n = 5). (G,H) Mature embryos of (G) Col-0 and (H) Bur-0. Scale, 50 μm. (I) Mature embryo area. Error bars indicate s.d. (n = 20). Statistically significant difference between accessions was calculated using Student’s t-test (NS, not significant; ^***p < 0.001).

Figure 2

Cell division stops at late torpedo and mature stages of embryogenesis. (A–F) RNA in situ hybridization on longitudinal sections through (A,D) early, (B,E) late, and (C,F) mature embryos of (A–C) Col-0 and (D–F) Bur-0 using CYCLINB1;1 (CYCB1;1) as the probe. Scale bars, 100 μm.

Figure 3

Metabolite content of dormant seeds and embryos at late torpedo and mature stages. (A) Principal component analysis (PCA) and (B) clustered heat map of the metabolites measured in late torpedo and mature embryos and dormant seeds of Col-0 (dark gray) and Bur-0 (light gray) wild-type plants grown in long-days (LDs; 16 h light/8 h darkness). Metabolites from biological replicates (n ≥ 3) were averaged and z-score normalized. The percentages of total variance represented by principal component 1 (PC1) and PC2 are shown in parentheses. The loadings of individual metabolites are shown in red.

Figure 4

Metabolite content of late torpedo‐ and mature-stage embryos and dormant seeds of A. thaliana accessions Col-0 (dark gray) and Bur-0 (light gray). Plants were grown in LD conditions (16 h light/8 h darkness). (A) Total protein content. (B) Total amino acids (AAs). (C) Starch. (D) Sucrose. (E) Glucose (Glu). (F) Fructose (Fru). (G) Fumarate. (H) Malate. Note that both fumarate and malate were not detected (n.d.) in embryos of Col-0 and Bur-0 plants. (I) Total carbon (C) accumulated in metabolites measured in embryos and seeds. (J) Sucrose and hexoses (Glu + Fru). (K) Hexose-to-sucrose ratio. Error bars indicate mean ± SEM (n = 3). At each time point, statistically significant difference between the two accessions was calculated using Student’s t-test and is indicated as follows: ^*p < 0.05; ^**p < 0.01; ^***p < 0.001.

Figure 5

Metabolite content of late torpedo and mature embryos. (A) PCA and (B) clustered heat map of metabolites measured in torpedo and mature embryos of Col-0 (dark gray) and Buro-0 (light gray) wild-type plants grown in LDs (16 h light/8 h darkness). Metabolites from biological replicates (n ≥ 3) were averaged and z-score normalized. For the PCA, the percentages of total variance represented by PC1 and PC2 are shown in parentheses. The loadings of individual metabolites are shown in red.

Figure 6

Carbon metabolism is enhanced in mature embryos of Bur-0. (A–D) Transcript abundance of (A,B) SUCROSE SYNTHASE (SUS) and (C,D) cytosolic INVERTASE (CINV) genes in dissected (A,C) late torpedo and (B,D) mature embryos of Col-0 and Bur-0. Error bars indicate ± s.d. (n = 3). At each embryo stage, statistically significant difference between the two accessions was calculated using Student’s t-test, and is indicated as follows: ^*p < 0.05; ^**p < 0.01. (E,F) RNA in situ hybridization using specific antisense probes for SUCROSE SYNTHASE 3 (SUS3, E) and SUS1 (F) genes on longitudinal sections through mature embryos of Col-0 and Bur-0 plants. Arrows indicate the expression of SUS1 at the shoot apical meristem. Sense probes were used as control (right side). Scale bars, 50 μm.

Figure 7

Simplified model describing carbohydrate signatures of embryos during late torpedo and mature stages as well as dormant seeds of A. thaliana Col-0 and Bur-0 accessions. Note that Bur-0 embryos accumulate much higher carbon reserves during late embryogenesis than Col-0 embryos. AA, amino acids.