Landscape of the lipidome and transcriptome under heat stress in Arabidopsis thaliana

Abstract

Environmental stress causes membrane damage in plants. Lipid studies are required to understand the adaptation of plants to climate change. Here, LC-MS-based lipidomic and microarray transcriptome analyses were carried out to elucidate the effect of short-term heat stress on the Arabidopsis thaliana leaf membrane. Vegetative plants were subjected to high temperatures for one day, and then grown under normal conditions. Sixty-six detected glycerolipid species were classified according to patterns of compositional change by Spearman's correlation coefficient. Triacylglycerols, 36:4- and 36:5-monogalactosyldiacylglycerol, 34:2- and 36:2-digalactosyldiacylglycerol, 34:1-, 36:1- and 36:6-phosphatidylcholine, and 34:1-phosphatidylethanolamine increased by the stress and immediately decreased during recovery. The relative amount of one triacylglycerol species (54:9) containing α-linolenic acid (18:3) increased under heat stress. These results suggest that heat stress in Arabidopsis leaves induces an increase in triacylglycerol levels, which functions as an intermediate of lipid turnover, and results in a decrease in membrane polyunsaturated fatty acids. Microarray data revealed candidate genes responsible for the observed metabolic changes.

Figure 1. Overview of glycerolipid levels in Arabidopsis leaves under heat stress and during recovery.

The 66 lipid species observed from the all environmental conditions (heat and recovery) and ecotypes (Col-0 and Nossen) were shown. Values were normalised to 1d38C data in each experiment, and scaled among measurements as described in Methods. Each column represents an LC-MS measurement. Biological replicates N = 3, 4, 8, or 12. MGDG, monogalactosyldiacylglycerol; DGDG, digalactosyldiacylglycerol; SQDG, sulfoquinovosyldiacylglycerol; PG, phosphatidylglycerol; PC, phosphatidylcholine; PE, phosphatidylethanolamine; PI, phosphatidylinositol; DAG, diacylglycerol; TAG, triacylglycerol. Experimental condition can be found in Supplementary Fig. S1. The normalised data was available in Supplementary Table S1.

Figure 2. Heat map representation of lipid species in Arabidopsis leaves under heat stress superimposed on a glycerolipid metabolic pathway map.

The average changes in lipid species are shown by 10 boxes (2 rows and 5 columns), which contain 2 genotypes: Col-0 (upper row) and Nossen (lower row); and 5 environments: 1d30C, 1d34C, 1d38C, 1d38C + 1d22C, 1d38C + 2d22C (left-to-right). Heat map colours show the average log₂ ratios of fold-changes in each stress condition to the normal 22 °C condition. Due to different induction levels, values of diacyl membrane lipids and TAG are shown by different scales. Samples depicted by boxes with no colour in the Nossen data were not detected. Lipid species belonging to the 10 clusters (Spearman’s correlation coefficients, see Supplementary Fig. S3 for details) are shown with the numbers in parentheses (e.g. “1”, “2”, “6”, “7”, and “10”). Individual experimental data can be found in Supplementary Table S1.

Figure 3. The expression of genes annotated to be involved in glycerolipid metabolism in Arabidopsis leaves under heat stress and during recovery.

The expression of 233 genes (228 probe sets) selected from the literature was analysed by microarray. Four sets of fold-changes were calculated as Heat08 hr/Control08 hr, Heat24 hr/Control24 hr, Recovery08 hr/Control08 hr and Recovery24 hr/Control24 hr and are shown as a heat map from left-to-right. (a) Probe sets increased >1.5-fold under at least one condition and not decreased, (b) probe sets decreased <0.5-fold under at least one condition and not increased, (c) probe sets both increased >1.5-fold under at least one condition and decreased <0.5-fold under at least one condition. Genes were selected from the literature, and classified into 8 groups of metabolic pathways: (1) plastidial fatty acid synthesis; (2) fatty acid elongation, desaturation and export from plastid; (3) plastidial glycerolipid, galactolipid and sulfolipid synthesis; (4) eukaryotic phospholipid synthesis; (5) TAG synthesis; (6) lipid trafficking; (7) TAG degradation, β-oxidation, the TCA/glyoxylate cycles; (8) miscellaneous lipid-related genes, and phytyl ester synthases. Genes that changed significantly (t-test with Benjamini-Hochberg FDR, p < 0.05, N = 3, 6, or 9) among the selected genes are shown. Further information is available in Supplementary Table S2.

Figure 4. Predicted intracellular lipid trafficking. Integrated lipidome and transcriptome investigations suggest an enhanced lipid movement among organelles under heat stress in Arabidopsis leaves.

Seven arrows (‘a’ through ‘g’) show metabolic pathways suggested to be induced by heat stress. FAS, fatty acid synthesis; CP, chloroplasts; TM, thylakoid membrane; PG, plastoglobules; IM, inner envelope membrane; OM, outer envelope membrane; ER, endoplasmic reticulum; LD, lipid droplets; PX, peroxisome; CY, cytosol.