First haplotype-resolved genome assembly of citral-rich lemongrass Cymbopogon flexuosus var. Krishna

This study presents the first chromosome-scale, haplotype-resolved genome assembly of the citral-rich lemongrass *Cymbopogon flexuosus* var. Krishna, generated using PacBio HiFi and Omni-C technologies, which provides a foundational genomic resource for advancing molecular breeding and metabolic engineering in aromatic grasses.

The Gist

Imagine you have a very complex, double-layered recipe book for making a specific type of perfume. This recipe book belongs to Lemongrass (specifically a super-famous variety called "Krishna"), which is famous for its zesty, lemony scent used in everything from soaps to medicines.

For a long time, scientists knew what this plant did, but they didn't have the complete, high-definition instruction manual (the genome) to understand how it did it. The manual was messy, full of typos, and because the plant has two slightly different sets of instructions (like a double-sided book), it was incredibly hard to read.

Here is what this paper achieved, explained simply:

1. The Challenge: A Messy, Double-Sided Book

Lemongrass is a bit of a genetic troublemaker. Unlike humans or rice, which have very similar copies of their DNA, lemongrass is highly heterozygous.

• The Analogy: Imagine trying to read a book where every page has two different stories written on top of each other. One story says "add salt," and the other says "add sugar." If you just mash them together, you get a confusing mess.
• The Problem: Previous attempts to map this plant's DNA were like trying to read that messy book with a blurry magnifying glass. They couldn't see the details needed to improve the plant or understand how it makes its special oils.

2. The Solution: High-Tech Scanners and a 3D Map

The researchers used two powerful new tools to fix this:

• PacBio HiFi Sequencing (The "High-Definition Scanner"): Instead of taking tiny, blurry snapshots of the DNA, this technology reads long, continuous, crystal-clear strands. It's like switching from a pixelated 1990s photo to a 4K Ultra HD video.
• Omni-C (The "3D Origami Map"): DNA doesn't just float in a straight line; it folds up inside the cell like a ball of yarn. Omni-C technology takes a "snapshot" of how the DNA folds, showing which parts touch each other. This helped the scientists figure out exactly where each piece of the puzzle belongs, like using a 3D map to untangle a knot of headphones.

3. The Result: Two Perfect Books and One Master Copy

The team didn't just make one version of the book; they made three incredibly high-quality versions:

• The "Pseudo-Haploid" (The Master Summary): A single, clean version of the book that combines the best parts of both sides. It's about 798 million letters long and organized into 10 neat chapters (chromosomes).
• The "Haplotype-Resolved" Assemblies (The Two Originals): They successfully separated the two different versions of the book.
  • Book A (Hap1): ~750 million letters.
  • Book B (Hap2): ~726 million letters.
  • Why this matters: Now, scientists can see exactly which version of a gene makes the plant smell more like lemon and which makes it smell less. It's like finally being able to read both sides of the recipe separately to see which one is the "secret ingredient."

4. What's Inside the Book?

The scientists read through the entire manual and found:

• 37,254 Instructions (Genes): These are the specific commands the plant follows to grow and survive.
• The "Citral" Factory: They specifically looked for the instructions on how to make Citral (the chemical that gives lemongrass its smell and is used to make Vitamin A and perfumes). Now that they have the map, they can find the exact "machines" (genes) that build this chemical.
• The "Junk" (Repeats): About 65% of the book is filled with repetitive patterns (like a page that just says "repeat, repeat, repeat"). While this used to be confusing, the new technology sorted this out perfectly, showing exactly where the repeats start and stop.

5. Why Should You Care?

This isn't just about a grassy plant; it's a foundation for the future.

• Better Perfumes & Medicines: With this map, scientists can use "molecular breeding" (like GPS-guided farming) to grow lemongrass that smells even stronger or produces more oil, without needing more land.
• Saving Farmers: This variety ("Krishna") is already a favorite for farmers in India. Having this map helps scientists make it even more resilient to drought or disease, securing income for small farmers.
• A New Era: Before this, studying lemongrass was like trying to fix a car engine with a blindfold on. Now, they have the engine diagram, the parts list, and the manual.

In a nutshell: The researchers took a messy, double-layered genetic puzzle of a super-important plant and solved it with high-tech tools. They produced the first clear, chromosome-by-chromosome map of the plant, separating its two genetic "personalities." This map is the key to unlocking better medicines, better fragrances, and better crops for the future.

Technical Summary

1. Problem Statement

• Economic Importance: Cymbopogon flexuosus (East Indian lemongrass) is a high-value aromatic grass in the Poaceae family, widely cultivated in India for its essential oil rich in citral (78–82%), a precursor for vitamins, flavors, fragrances, and pharmaceuticals.
• Genomic Gap: Despite its commercial significance, high-quality genomic resources for C. flexuosus have been lacking. Existing studies were limited to transcriptomes or draft assemblies of related species (e.g., C. citratus), which lacked the resolution to study genome organization, secondary metabolite gene clusters, or facilitate genomics-assisted breeding.
• Biological Challenge: The species is predominantly outcrossing with a highly heterozygous genome and variable ploidy levels (diploid to hexaploid), making standard assembly difficult. There was no chromosome-scale reference genome available for the elite cultivar 'Krishna'.

2. Methodology

The study employed a multi-platform sequencing strategy to overcome heterozygosity and achieve chromosome-level resolution:

• Plant Material: Young leaves from a single C. flexuosus var. Krishna plant were harvested to minimize polysaccharide contamination. High Molecular Weight (HMW) DNA (>50 kb) was extracted.
• Sequencing Platforms:
  • PacBio HiFi: Generated 32 Gb of high-fidelity reads (40× coverage) using the PacBio Revio platform. Reads had a predicted accuracy ≥0.99.
  • Omni-C (Hi-C): Generated 377 million read pairs (150× coverage) using the Dovetail Omni-C Kit on an Illumina NovaSeq 6000 for chromatin conformation capture and scaffolding.
• Genome Characterization:
  • Genome size estimated via k-mer analysis (Jellyfish/GenomeScope 2.0) and flow cytometry: 0.58–0.80 Gb.
  • Heterozygosity rate: 0.8%; Repetitive content: ~48.9%.
  • Ploidy confirmed as diploid via smudgeplot analysis.
• Assembly Pipeline:
  • Primary Assembly: Hifiasm (v0.16.1) was used for de novo assembly of HiFi reads. Redundant haplotigs were purged using Purge_dups to create a pseudo-haploid assembly.
  • Scaffolding: HapHiC (v1.0.6) utilized Omni-C data to anchor contigs into 10 chromosome-scale scaffolds.
  • Gap Filling: Minimap2 and LR-GapCloser reduced gaps to 38.
  • Haplotype Resolution: Separate haplotype-resolved assemblies (Hap1 and Hap2) were generated from the unitig graph.
• Annotation:
  • Repeats: Annotated using an integrated pipeline (TRF, RepeatModeler, EDTA, RepeatMasker).
  • Gene Prediction: Performed using the MAKER pipeline, integrating ab initio predictions (Augustus trained on Sorghum bicolor), protein homology (from Oryza, Sorghum, Zea, Arabidopsis), and transcriptomic evidence (RNA-seq from leaves/inflorescences).
  • Functional Annotation: Utilized SwissProt, UniProt, NCBI nr, Pfam, InterPro, eggNOG-mapper, and KEGG.

3. Key Results

A. Assembly Statistics

• Pseudo-haploid Assembly:
  • Total Size: ~798 Mb.
  • Contiguity: Organized into 10 chromosomes (matching the haploid number).
  • N50: Scaffold N50 of 64.35 Mb.
  • Completeness: 99.8% BUSCO recovery (Embryophyta dataset: 98.0% complete).
  • Quality: High base-level accuracy confirmed by Merqury QV and Hi-C contact maps.
• Haplotype-Resolved Assemblies:
  • Captured allelic diversity with two distinct haplotypes:
    • Hap1: ~750 Mb.
    • Hap2: ~726 Mb.
  • Together, they represent 95–98% of the pseudo-haploid genome, providing phase-resolved information for gene families.

B. Genomic Features

• Repetitive Elements: Transposable elements (TEs) constitute 65.16% of the genome.
  • LTR Retrotransposons are dominant, specifically Gypsy (25.89%) and Copia (7.23%).
  • Total interspersed repeats: ~504 Mb.
• Gene Content:
  • Total Protein-Coding Genes: 37,254.
  • Functional Annotation:
    • 32,428 genes annotated via UniProt.
    • 31,516 via eggNOG-mapper.
    • 34,681 via InterPro.
    • 26,173 genes mapped to KEGG pathways.
  • Metabolic Pathways: Specific manual curation identified genes involved in monoterpene biosynthesis (citral production).

C. Data Availability

• The assembly is deposited in NCBI GenBank (BioProject: PRJNA1301599, Assembly: JBQKUF000000000).
• Raw sequencing data (HiFi, Omni-C, RNA-seq) is available in the GR-MAP database.

4. Significance and Impact

• First Reference Genome: This represents the first chromosome-scale, haplotype-resolved genome for C. flexuosus, filling a critical gap in the genomic resources of aromatic grasses.
• Breeding and Improvement: The high-quality reference enables marker-assisted selection and genomic selection for traits like citral content, yield stability, and stress tolerance.
• Metabolic Engineering: The detailed annotation of terpene synthase genes and biosynthetic pathways provides a blueprint for metabolic engineering to enhance essential oil production.
• Comparative Genomics: The assembly facilitates comparative studies within the Poaceae family, particularly regarding the evolution of secondary metabolite gene clusters in aromatic grasses versus major cereals.
• Resource for Heterozygous Species: The successful application of PacBio HiFi + Omni-C to resolve the highly heterozygous C. flexuosus genome serves as a methodological model for other complex, outcrossing crop species.