Integrated Multi-Omics Maps Metabolic Shift During Seed Development Stages

Canavalia gladiata, commonly known as sword bean, is a leguminous crop with high seed protein content (26%–30%) and oil rich in unsaturated fatty acids. Despite its nutritional promise, the species has remained underutilized, largely because a complete genome and the genetic basis of its nutritional traits were lacking.

Now, a research team from Yunnan University and BGI Research (Wuhan) has assembled the first gap-free telomere-to-telomere (T2T) genome of C. gladiata, combined with integrated transcriptomic and metabolomic profiling of developing seeds. The study was published in Journal of Systematics and Evolution.

"The previous genome assembly left centromeric and telomeric regions unresolved," says Yi Wang, first author of the study. "These repetitive regions often hold key information about genome stability and phenotypic variation, so a T2T assembly was essential for functional genomics and breeding."

A Complete Genome, 11 Chromosomes, 22 Telomeres

The team combined Oxford Nanopore ultra-long reads, Hi-C scaffolding, and multiple assembly strategies to produce an 11-chromosome assembly containing all 22 telomeres and 11 centromeres. Centromeric regions were dominated by tandem repeats, with a 504-bp repeat family (TR504) present on six chromosomes and distinct repeat families on the others.

Comparative genomics placed C. gladiata as a basal Phaseoleae species that diverged from soybean approximately 45 million years ago. Ks distribution analysis showed that C. gladiata shares an ancient whole-genome duplication with soybean, but did not undergo the additional recent duplication that occurred in the soybean lineage-making it a valuable reference for reconstructing ancestral legume karyotypes.

Multi-Omics Reveals a Metabolic Switch During Seed Development

Seeds were sampled at three developmental stages: 40 (S1, early), 60 (S2, middle), and 80 (S3, mature) days after flowering. Widely targeted metabolomics detected 2723 metabolites, with flavonoids being the most abundant class. Transcriptomic profiling identified 14 016 differentially expressed genes across the three stages.

The integrated analysis revealed a clear metabolic phase transition. The early stage (S1) was characterized by active starch and sucrose metabolism, branched-chain amino acid (BCAA) degradation, and flavonoid biosynthesis. A central hub metabolite, glucose-6-phosphate (G6P), showed strong correlation with the transcription factor CgKNAT7 (r = 0.99), suggesting a role in carbon allocation. Twelve flavonoid metabolites, including catechin and quercetin, accumulated at S1 alongside coordinated expression of structural genes such as CHS, F3H, DFR, LAR, and ANR.

In contrast, the middle and late stages (S2 and S3) showed redirection of carbon flux toward fatty acid accumulation. Free fatty acids, including 13(S)-HODE and 9(S)-HODE, peaked at S3, consistent with the expression of linoleic acid metabolism genes such as CgLOX5 and CgMMS19.

A Resource for Legume Improvement

The study provides a high-quality reference genome and candidate genes for nutritional quality traits in C. gladiata, offering a foundation for molecular breeding in this and other underutilized legumes. The T2T assembly, transcriptome data, and metabolome data have been deposited in public databases.