Citronelloid compounds such as citronellal and citronellol function in plant defense and ecological communication, and they are also widely used in pharmaceuticals and essential oils. Within the Poaceae family, species of the genus Cymbopogon display striking differences in citronelloid accumulation, yet the evolutionary origin and biosynthetic regulation of these compounds have remained poorly understood. The absence of high-quality reference genomes has limited functional gene identification and comparative evolutionary analysis. Moreover, the cellular compartmentation of citronelloid biosynthesis has not been clarified. Based on these challenges, comprehensive genome assembly and integrative multi-omics analysis are urgently needed to dissect the molecular mechanisms underlying citronelloid biosynthesis.

Researchers from Sun Yat-Sen University report the first haplotype-resolved reference genomes of Cymbopogon winterianus and Cymbopogon distans. Using PacBio HiFi and Hi-C sequencing, the team assembled high-quality diploid and tetraploid genomes and performed comparative evolutionary analyses across 21 plant species. Through metabolome–transcriptome association and single-cell RNA sequencing of leaves, the study identifies candidate genes and cell-type-specific expression patterns responsible for citronelloid biosynthesis, offering new insights into terpene evolution within the Poaceae family.

The multi-omics analysis of citronelloid biogenesis pathway in Cymbopogon

The plant division and individual fractions of leaves, shoots, and roots in C. winterianus (a) and C. distans (b) for multi-omics sampling. c. Overview of genomic features in C. winterianus and C. distans. From outside to inside: Track I illustrates genomic landscape of C. winterianus pseudomolecules (left) and C. distans pseudomolecules (right). Tracks II-VI, represented by five concentric rings, illustrate the density distribution of genes, GC content, repeat sequences, LTR/Copia, and LTR/Gypsy elements. Track VII, displays syntenic blocks between C. winterianus and C. distans. d. Analysis of Ks for paralogs and orthologs in C. winterianus, C. distans, and two additional species. The Y-axis represents genes frequency at the respective Ks values. e. Phylogenetic tree constructed from orthogroups across 21 plant genomes (the outgroup Chlamydomonas reinhardtii is not shown). Node support was calculated using 1000 bootstraps, with all nodes exhibiting support values exceeding 0.99. The numbers next to nodes represented the divergence times (million years ago), and the bars indicated the 95% confidence intervals for the divergence times between different clades. Adjacent heat maps illustrate the distribution of terpenoid biosynthesis pathway gene families among species. f. Pie chart illustrating the distribution of various metabolite categories. g. KEGG pathway enrichment of differential metabolites. h. The concentrations of the ten most abundant terpenes in leaf, shoot, and root tissues of C. winterianus were quantified utilizing GC–MS with three biological replicates. i. Phylogenetic analysis of selected TPS genes from C. winterianus, C. distans, and other plant species. The clades were classified according to the AtTPS genes family categorization. The candidate CwTPS genes in C. winterianus that exhibit strong correlations with citronelloid biosynthesis are highlighted in green, whereas the reported genes from other species that participate in the citronelloid biosynthesis pathways are colored red. j. UMAP projection illustrating 12 cell clusters derived from gene expression matrices of high-quality cells. k. Heatmaps show relative expression levels of candidate genes participating in the citronelloid biosynthesis pathway.

The researchers assembled 1.49 Gb and 2.58 Gb genomes for diploid C. winterianus and tetraploid C. distans, respectively, achieving 99.1% completeness. Comparative analyses revealed strong chromosomal collinearity with Sorghum bicolor, indicating shared whole-genome duplication history and divergence approximately 19.2 million years ago.

Gene family analysis showed expansion of early terpenoid pathway genes—including 3-hydroxy-3-methylglutaryl-CoA synthase (HMGS), 3-hydroxy-3-methylglutaryl-CoA reductase (HMGR), and isopentenyl diphosphate isomerase (IDI)—in Poaceae species. Notably, terpene synthase (TPS) genes were markedly expanded in Cymbopogon, consistent with its rich volatile profile.

Metabolomic profiling identified 1,158 volatile compounds, with terpenoids representing the largest fraction. Citronellal, citronellol, geraniol, and related monoterpenes accumulated predominantly in leaves. Integrative correlation analysis pinpointed 19 candidate genes strongly associated with citronelloid content (R > 0.99), including seven TPS, five ADH, two OPR, and five CAR genes.

Single-cell RNA sequencing further revealed spatial specialization: most candidate TPS genes were highly expressed in mesophyll cells, suggesting this cell type as a key metabolic hub. Meanwhile, dehydrogenases and reductases displayed cell-type-specific patterns, implying metabolic compartmentation within leaves. Together, these findings map citronelloid biosynthesis from genome evolution to cellular resolution.

“This is the first time we can trace citronelloid biosynthesis from evolutionary history down to specific leaf cell types,” said the corresponding author. “By integrating genome assembly with metabolomics and single-cell transcriptomics, we were able to connect gene family expansion, pathway diversification, and cellular expression specialization. These results not only clarify how citronella produces its characteristic compounds but also establish a genomic foundation for future functional validation and metabolic engineering.”

The newly generated genomic resources provide a platform for molecular breeding and essential oil optimization in Cymbopogon. Identification of TPS-, ADH-, OPR-, and CAR-related candidate genes enables targeted manipulation of citronelloid composition for pharmaceutical and industrial applications. Furthermore, the discovery of cell-type-specific expression patterns opens possibilities for precision metabolic engineering by modulating pathway compartmentation. Beyond citronella, this study offers a comparative framework for understanding terpene evolution across monocots and aromatic plants. As demand for plant-derived bioactive compounds increases, integrating high-resolution genomics with cellular transcriptomics may accelerate the domestication and metabolic enhancement of other economically important species.

Source – Newswise