Co-expression analysis screened out the set of terpene-related
genes
To investigate the transcriptomic differences in leaves, stems, and
roots of S. rosmarinus , three biological replicates were
collected from each organ to ensure accuracy. Raw data underwent
filtering, resulting in 7.02-7.10 million 150 bp paired-end reads (Table
S13), which were mapped to our assembly at the rate of
67.95%– 92.40% (Table S14). Differentially expressed genes
(DEGs) were identified using DESeq2 (Varet, Brillet-Gueguen, Coppee, &
Dillies, 2016), with a stringent threshold of
Log2|FoldChange| >1 and p-value
< 0.05. We identified a total of 16,052 DEGs, with 8,833
up-regulated and 7,219 down-regulated genes in the root vs. Leaf
comparison, 11,982 DEGs (6,496 up-regulated, 5,486 down-regulated) in
the root vs. Stem comparison, and 15,598 DEGs (8,241 up-regulated, 7,357
down-regulated) in the leaf vs. Stem comparison (Figure S14). In
addition, 3,475 genes were differentially expressed in all three groups
(Figure 3e).
To gain insight into the metabolic processes involved in different
organs of S. rosmarinus , KEGG enrichment analysis was performed.
The DEGs between roots and leaves were significantly enriched in
”Biosynthesis of other secondary metabolites” and ”Metabolism of
terpenoids and polyketides” (p-value< 0.05) (Figure S15, Table
S15). While the DEGs between roots and stems were significantly enriched
in ”Metabolism of terpenoids and polyketides” and ”Terpenoid backbone
biosynthesis” (p-value< 0.05) (Figure S15, Table S15).
Notably, the DEGs between stems and leaves were significantly enriched
in ”Biosynthesis of other secondary metabolites,” ”Flavonoid
biosynthesis,” and ”Phenylpropanoid biosynthesis” (p-value <
0.05) (Figure S15, Table S16). These pathways provided a
transcriptomic-level understanding of the metabolic processes in
different organs of S. rosmarinus .
To construct the WGCNA co-expression network, we used all DEGs and
differential metabolites. By calculating Pearson correlation
coefficients (PCCs) between contents of components and gene expression
modules (Figure S16, Tables S11, S12), we identified a set of genes
associated with monoterpenoids and diterpenoids (PCCs >
0.6). Our results suggested that SrCYP71D8 , SrWRKY2 (with
p-value of 0.0315 and 0.0237, respectively) were highly correlated with
the accumulation of monoterpene in S. rosmarinus organs. We also
screened key genes, including SrHMGR (with p-value located at
0.0019– 0.0401) and limonene synthase, which was found to be
involved in the biosynthesis process of monoterpene. Moreover, we
identified SrCYP81Q32 , and SrMYB1 (with p-values of 0.0004
and 0.0004, respectively) as highly correlated with the accumulation of
diterpenoids. Finally, we found that SrGGPPS (with p-values of
0.0004– 0.0475) was highly correlated with diterpenoids
biosynthesis (Table S21). These findings provide important evidence to
support further exploration of terpenoid biosynthesis in S.
rosmarinus .