Transcriptomic analysis of wild Cannabis sativa: insights into tissue- and stage-specific expression and secondary metabolic regulation

Abstract

Cannabis sativa is a medicinally and economically significant plant known for its production of cannabinoids, terpenoids, and other secondary metabolites. This study presents a transcriptomic analysis to elucidate tissue-specific expression and regulatory mechanisms across leaves, stems, and roots. A total of 2,530 differentially expressed genes (DEGs) were identified, with key genes such as terpene synthase (TPS) and phenylalanine ammonia-lyase (PAL) exhibiting elevated expression in leaf tissues, emphasizing their roles in terpenoid and phenylpropanoid biosynthesis. Alternative splicing (AS) analysis revealed 8,729 distinct events, dominated by exon skipping, contributing to transcriptomic diversity. Long non-coding RNA (lncRNA) prediction identified 3,245 candidates, many of which displayed tissue-specific expression patterns and co-expression with metabolic genes, suggesting regulatory roles in secondary metabolism. Additionally, 12,314 SNPs and 2,786 INDELs were detected, with notable enrichment in genes associated with secondary metabolite biosynthesis, particularly in leaf tissues. These findings advance the understanding of molecular mechanisms governing secondary metabolism and genetic diversity in C. sativa, providing valuable insights for future metabolic engineering and breeding strategies to enhance cannabinoid production.

Keywords: Cannabis sativa; Alternative splicing; Genetic variation.; Secondary metabolism; Transcriptomics.

Fig. 1

Transcriptomic Patterns Across Tissues and Developmental Stages in C. sativa. (A) Distribution of gene and isoform expression levels (FPKM) across analyzed samples. Each box represents the expression distribution within a specific sample, with the median, interquartile range, and extreme values indicated. (B) Differentially expressed genes (DEGs) between tissue comparisons. Bars represent total DEGs, with upregulated genes in green and downregulated genes in red. The x-axis denotes pairwise comparisons, and the y-axis indicates the number of DEGs

Fig. 2

GO Classification for Differentially Expressed Genes. The y-axis represents the next-level GO terms under the three main GO categories. The x-axis indicates the number of genes annotated to each term, including those associated with its sub-terms. The three classifications represent the primary GO categories: biological process (top), cellular component (middle), and molecular function (bottom)

Fig. 3

KEGG Enrichment for Differentially Expressed Genes. The y-axis represents pathway names, sorted by Q-value in ascending order. The x-axis shows the -log10(Q-value). The size of each dot indicates the number of differentially expressed genes (DEGs) within the pathway, while the color of the dots corresponds to different ranges of the RichFactor

Fig. 4

Classification and Quantification of Alternative Splicing Events. The classification and frequency of alternative splicing (AS) events identified in the C. sativa transcriptome. The x-axis represents ten categories of AS events, defined as follows: 1. SKIP: Exon skipping, the omission of an exon during splicing. 2. MSKIP: Mutually exclusive exon skipping, where one of two alternative exons is included. 3. IR: Intron retention, the retention of an intron in the mature mRNA. 4. MIR: Mutually exclusive intron retention, where one of two introns is retained. 5. AE: Alternative exon ends, variability at the 5’, 3’, or both ends of an exon. 6. XSKIP: Approximate exon skipping, a variant of exon skipping. 7. XMSKIP: Approximate mutually exclusive exon skipping, a variant of mutually exclusive exon skipping. 8. XIR: Approximate intron retention, a variant of intron retention. 9. XMIR: Approximate mutually exclusive intron retention, a variant of mutually exclusive intron retention. 10. XAE: Approximate alternative exon ends, a variant of alternative exon ends. The y-axis indicates the number of AS events in each category. Among these, SKIP (exon skipping) was the most prevalent, followed by IR (intron retention). Less frequent types, such as AE (alternative exon ends) and their approximate variants, also contributed to transcript isoform diversity, albeit to a smaller extent

Fig. 5

Coding Potential Evaluation of Predicted lncRNAs. (A) ROC curve: Performance of the coding potential classifier under different thresholds. (B) PR curve: Precision-Recall relationship for lncRNA identification. (C) Accuracy vs. cutoff: Statistical evaluation of classification accuracy. (D) Two-graph ROC curve: Determination of the optimal cutoff value for distinguishing coding and non-coding RNAs

Fig. 6

Validation of key DEGs involved in secondary metabolism by qRT-PCR. (A) Expression profiles of ten DEGs across root, stem, and leaf tissues at the vegetative stage. (B) Expression profiles of the same DEGs at the flowering stage