Comparative transcriptomic analysis provides insights into the genetic networks regulating oil differential production in oil crops

Background Plants differ more than threefold in seed oil contents (SOCs). Soybean (Glycine max), cotton (Gossypium hirsutum), rapeseed (Brassica napus), and sesame (Sesamum indicum) are four important oil crops with markedly different SOCs and fatty acid compositions. Results Compared to grain crops like maize and rice, expanded acyl-lipid metabolism genes and relatively higher expression levels of genes involved in seed oil synthesis (SOS) in the oil crops contributed to the oil accumulation in seeds. Here, we conducted comparative transcriptomics on oil crops with two different SOC materials. In common, DIHYDROLIPOAMIDE DEHYDROGENASE, STEAROYL-ACYL CARRIER PROTEIN DESATURASE, PHOSPHOLIPID:DIACYLGLYCEROL ACYLTRANSFERASE, and oil-body protein genes were both differentially expressed between the high- and low-oil materials of each crop. By comparing functional components of SOS networks, we found that the strong correlations between genes in “glycolysis/gluconeogenesis” and “fatty acid synthesis” were conserved in both grain and oil crops, with PYRUVATE KINASE being the common factor affecting starch and lipid accumulation. Network alignment also found a conserved clique among oil crops affecting seed oil accumulation, which has been validated in Arabidopsis. Differently, secondary and protein metabolism affected oil synthesis to different degrees in different crops, and high SOC was due to less competition of the same precursors. The comparison of Arabidopsis mutants and wild type showed that CINNAMYL ALCOHOL DEHYDROGENASE 9, the conserved regulator we identified, was a factor resulting in different relative contents of lignins to oil in seeds. The interconnection of lipids and proteins was common but in different ways among crops, which partly led to differential oil production. Conclusions This study goes beyond the observations made in studies of individual species to provide new insights into which genes and networks may be fundamental to seed oil accumulation from a multispecies perspective. Supplementary Information The online version contains supplementary material available at 10.1186/s12915-024-01909-x.


Figure S2
Figure S2 Spearman's rank correlation of sample transcript expression profiles of S. indicum.

Figure S3
Figure S3 Spearman's rank correlation of sample transcript expression profiles of B. napus.

Figure S4
Figure S4 Spearman's rank correlation of sample transcript expression profiles of G. hirsutum.

Figure S5
Figure S5 Spearman's rank correlation of sample transcript expression profiles of G. max.

Figure
Figure S6 Spearman's rank correlation of sample transcript expression profiles of Z. mays.

Figure S7
Figure S7 Principal component analysis (PCA) of sample transcript expression profiles.(a) The PCA of the oil crop and maize samples.(b) The PCAs of the individual species.

Figure S8
Figure S8 Analysis flow of multispecies comparative transcriptomic analysis based on coexpression networks.

Figure S10
Figure S10 Ratios and functional enrichments of the DEGs between the high-oil and low-oil materials.Red box on behalf of the up-regulated DEGs in the high-oil materials were enriched in the corresponding pathways, and the blue represents down-regulated DEGs.

Figure S11
Figure S11GO terms on biological processes shared between high-and low-oil materials of different species for DEGs between the adjacent developmental stages.Red circles indicated enrichment of up-regulated genes, blue indicated downregulated.

Figure S12
Figure S12DEGs in 'Fatty Acid Synthesis' that were up-regulated in the latter developmental stages compared to the former in each crop.Blank boxes indicated that the gene family was absent in the species.The median of log2FC of the up-regulated genes represents the log2FC of the gene family.

Figure S13
Figure S13DEGs in 'Triacylglycerol Biosynthesis' that were up-regulated in the latter developmental stages compared to the former in each crop.Blank boxes indicated that the gene family was absent in the species.The median of log2FC of the up-regulated genes represents the log2FC of the gene family.

Figure S14 A
Figure S14A full pathway model of oil accumulation in plant seeds and expression profiling of these genes in the ovules of the four oil crops.

Figure S15
Figure S15 Hierarchical clustering trees showing coexpression modules identified using WGCNA.Modules correspond to branches and are labeled by colors as indicated by the color band "Module" underneath the tree.

Figure S16
Figure S16Hierarchical clustering dendrograms of the eigengenes of each module in which the dissimilarity of eigengenes EI, EJ is given by 1-cor(EI; EJ).The shorter the height of the clustering tree, the higher the correlation (absolute value) of the two modules.The modules with red background are the modules related to SOS identified.

Figure S17
Figure S17 Ve nn diagram of KEGG ko2 pathways associated with the hubs of SOS modules in each oil crop (GSEA P ＜ 0.05, FDR Q ＜ 0.1).

Figure S18
Figure S18 Comparison of expression levels between pathway genes in the low-oil materials.(a) Comparison of expression levels between FAS genes in SOS modules and FAS genes not in the SOS modules in low-oil materials of each crop.(b) Comparison of expression levels between FAS genes in SOS modules and 'Glycolysis / Gluconeogenesis' genes in the SOS modules in low-oil materials of each crop.ns P value > 0.05; * P value ≤ 0.05; ** P value ≤ 0.01; *** P value ≤ 0.001; **** P value ≤ 0.0001, Wilcoxon rank-sum test.

Figure S19
Figure S19 Venn diagram of the protein domain PFAM annotations of 692 aligned genes in conserved networks and the top 20 PFAM annotations in counts in each species.

Figure S20
Figure S20 The PK network in maize and its expression profile.(a) The top 30 genes in weights connected to the PK.The node color represents the kME value of the corresponding gene in the network.(b) Heatmap of the PK and the top 30 genes in weights connected to the PK.Red arrows point to the PK in maize.