Cannabis Pangenome Scaffolded Genomes

Abstract

Cannabis sativa is a globally significant seed-oil, fiber, and drug-producing plant species. However, a century of prohibition has severely restricted legal breeding and germplasm resource development, leaving potential hemp-based nutritional and fiber applications unrealized. Existing cultivars are highly heterozygous and lack competitiveness in the overall fiber and grain markets, relegating hemp to less than 200,000 hectares globally¹. The relaxation of drug laws in recent decades has generated widespread interest in expanding and reincorporating cannabis into agricultural systems, but progress has been impeded by the limited understanding of genomics and breeding potential. No studies to date have examined the genomic diversity and evolution of cannabis populations using haplotype-resolved, chromosome-scale assemblies from publicly available germplasm. Here we present a cannabis pangenome, constructed with 181 new and 12 previously released genomes from a total of 156 biological samples from both male (XY) and female (XX) plants, including 42 trio phased and 36 haplotype-resolved, chromosome-scale assemblies. We discovered widespread regions of the cannabis pangenome that are surprisingly diverse for a single species, with high levels of genetic and structural variation, and propose a novel population structure and hybridization history. Conversely, the cannabinoid synthase genes contain very low levels of diversity, despite being embedded within a variable region containing multiple pseudogenized paralogs and distinct transposable element arrangements. Additionally, we identified variants of acyl-lipid thioesterase (ALT) genes² that are associated with fatty acid chain length variation and the production of the rare cannabinoids, tetrahydrocannabinol varin (THCV) and cannabidiol varin (CBDV). We conclude the Cannabis sativa gene pool has only been partially characterized, and that the existence of wild relatives in Asia remains likely, while its potential as a crop species remains largely unrealized.

1. Nions, U. Commodities at a glance: Special issue on industrial hemp. Commod Glance (2023) doi:10.18356/9789210019958.

2. Pulsifer, I. P. et al. Acyl-lipid thioesterase1-4 from Arabidopsis thaliana form a novel family of fatty acyl-acyl carrier protein thioesterases with divergent expression patterns and substrate specificities. Plant Mol. Biol. 84, 549–563 (2014).

Pangenome assembly and scaffolding

All genomes labeled Hifiasm_HiC, Hifiasm_Trio_RagTag, Hifiasm_RagTag, and Hifiasm (Supplementary Table 1) were assembled using Hifiasm v0.16.1¹. When available, HiC data and HiFi parental trio data were also incorporated into the assembly process defining the Hifiasm_HiC and Hifiasm_Trio_RagTag types respectively. CLR (continuous long reads) assemblies were generated using FALCON Unzip from PacBio SMRT Tools 9.0 Suite ¹⁵ and CCS (circular consensus sequencing) labeled genomes were assembled with HiCanu v2.2 ¹⁶. After assembly, HiC reads were aligned to the Hifiasm_HiC contigs using the Juicer v1.6.2 pipeline² followed by ordering and orientation utilizing version 180922 of the 3D-DNA pipeline³. The scaffolded assemblies were then manually corrected using Juicebox v1.11.08⁴. Hifiasm_RagTag and Hifiasm_Trio_RagTag assemblies were scaffolded using the split chromosomes of the 24 HiC scaffolded genomes and error checked with yak-0.1 (github.com/lh3/yak). Sourmash v4.6.1¹⁷ was used to generate a Jaccard similarity matrix between the chromosomes and each un-scaffolded assembly, and the most similar version of chromosome 1 through X was concatenated to generate a reference for scaffolding via RagTag v2.1.0¹⁸. If the similarity matrix identified the Y chromosome as the best match, the assembly remained un-scaffolded. BUSCO v5.4.3 ²¹ with the eudicots_odb10 dataset and assembly-stats v1.0.1 (https://github.com/sanger-pathogens/assembly-stats) were used on all assemblies to measure completeness and contiguity.

1. Cheng, H., Concepcion, G. T., Feng, X., Zhang, H. & Li, H. Haplotype-resolved de novo assembly using phased assembly graphs with hifiasm. Nat. Methods 18, 170–175 (2021).

2. Durand, N. C. et al. Juicer Provides a One-Click System for Analyzing Loop-Resolution Hi-C Experiments. Cell Syst 3, 95–98 (2016).

3. Dudchenko, O. et al. De novo assembly of the Aedes aegypti genome using Hi-C yields chromosome-length scaffolds. Science 356, 92–95 (2017).

4. Durand, N. C. et al. Juicebox Provides a Visualization System for Hi-C Contact Maps with Unlimited Zoom. Cell Syst 3, 99–101 (2016).

15. Chin, C.-S. et al. Phased diploid genome assembly with single-molecule real-time sequencing. Nat. Methods 13, 1050–1054 (2016).

16. Nurk, S. et al. HiCanu: accurate assembly of segmental duplications, satellites, and allelic variants from high-fidelity long reads. Genome Res. 30, 1291–1305 (2020).

17. Titus Brown, C. & Irber, L. sourmash: a library for MinHash sketching of DNA. J. Open Source Softw. 1, 27 (2016).

18. Alonge, M. et al. Automated assembly scaffolding using RagTag elevates a new tomato system for high-throughput genome editing. Genome Biol. 23, 258 (2022).

21. Smit, A. F. A., Hubley, R. & Green, P. RepeatMasker Open-4.0. 2013--2015. Preprint at (2015).