Novel mechanisms for the synthesis of important secondary metabolites in Ginkgo biloba seed revealed by multi-omics data

Abstract

Although the detailed biosynthetic mechanism is still unclear, the unique secondary metabolites of Ginkgo biloba, including ginkgolic acids (GAs) and terpene trilactones, have attracted increasing attention for their potent medicinal, physiological and biochemical properties. In particular, GAs have shown great potential in the fields of antibacterial and insecticidal activities, making it urgent to elucidate their biosynthetic mechanism. In this study, we systematically revealed the landscape of metabolic-transcriptional regulation across continuous growth stages of G. biloba seeds (GBS) based on multi-omics mining and experimental verification, and successfully identified all major types of GAs and terpene trilactones along with more than a thousand kinds of other metabolites. The phenological changes and the essential gene families associated with these unique metabolites were analyzed in detail, and several potential regulatory factors were successfully identified based on co-expression association analysis. In addition, we unexpectedly found the close relationship between large introns and the biosynthesis of these secondary metabolites. These genes with large introns related to the synthesis of secondary metabolites showed higher gene expression and expression stability in different tissues or growth stages. Our results may provide a new perspective for the study of the regulatory mechanism of these unique secondary metabolites in GBS.

Keywords: Ginkgo biloba; gene expression; ginkgolic acids; large intron; secondary metabolites.

Figure 1

Morphological observations of the complete developmental stages of GBS.

Figure 2

Classification of the trend changes in the content of all metabolites over different time periods.

Figure 3

Differential metabolite results; (A) Annotated bubble diagram of metabolites for Group 1; (B) Permutation plot of the OPLS-DA model. The horizontal axis indicates the similarity to the original model, and the vertical axis indicates the value of R2Y or Q2. The blue and red dots represent R2Y and Q2 of the model after Y replacement, respectively, and the dashed line is the fitted regression line. If R2Y and Q2 are smaller than R2Y and Q2 of the original model, then the model could be screened for differential metabolites according to VIP values; (C) Volcanic map of differential metabolites between June and August. The horizontal coordinate represents the fold change of each substance compared in this group, and the size of the scatter point represents the VIP value generated from the OPLS-DA model; (D) Heatmap of all metabolites based on VIP values. (A–C) represent June, August and October, respectively.

Figure 4

The unique secondary metabolites identified in GBS; (A) Venn diagram of all differential metabolites in three time periods; (B) Content variation of five representative GAs in three time periods; (C) Content variation of all types of terpenoids unique to GBS in three time periods; (D) Secondary spectrum of ginkgotoxin and its mirror image of the standard. The upper and lower plots represent the secondary spectrum extracted from the standard and the sample, respectively. The vertical axis represents the mass charge ratio (m/z).

Figure 5

Regulation of GA synthesis in GBS; (A) Clustering and co-expression results of all metabolites based on WGCNA method; (B) Correlation network diagram of GAs and other representative metabolites; (C) Inferred pathway and related genes involved in GAs synthesis in plastid and cytoplasm. Genes of the ACSL family, which are responsible for transmembrane transport, underwent a significant intron length expansion; (D) Co-expression network of candidate essential genes involved in the regulation of GA synthesis based on transcriptome and metabolome. Stars represent metabolites and triangles represent genes. The red and green lines represent positive and negative correlations, respectively.

Figure 6

Analysis results of the ACSL gene family in GBS; (A) Phylogenetic tree and expression pattern of all ACSL gene family members in GBS. The heatmap represents the FPKM values of all samples in three time periods, and the bar chart represents the length of each gene; (B) Length distribution of GbACSL and other GA synthesis-related gene family members; (C) Detailed composition of the intron regions of GbACSL gene members; (D) Violin plot of the expression of GbACSL genes with and without large introns; (E) Expression results of GbACSL genes with large introns validated by RT-PCR experiments; (F) Ka/Ks results for GbACSL genes with and without large introns. **** p value < 0.0001.

Figure 7

Mining results of terpene trilactone co-expression data in GBS; (A) Co-expression heatmap of the terpene trilactones identified in GBS and all genes; (B) Association network of ginkgo terpene trilactones and the essential genes based on Spearman’s correlation coefficient. The red and green lines represent positive and negative correlation, respectively, and the width of the lines indicates the strength of the correlation; (C) Functional annotation results of the extracted essential genes closely related to terpenoid synthesis in GBS; (D) Gene structure diagram of the genes involved in terpene trilactone synthesis in GBS.