The giant diploid faba genome unlocks variation in a global protein crop

The global demand for plant-based protein sources is increasing due to concerns about greenhouse gas emissions and biodiversity loss associated with meat-rich diets. Faba bean, a cool-season legume, presents a promising alternative due to its high yield potential and adaptability to various climates. However, its large genome size and limited genomic resources have hindered genetic improvement efforts. Previous research has focused on transcriptome analysis and SNP-based genetic maps, leading to some breakthroughs in understanding specific genes related to traits like tannin content and antinutrient levels. Yet, the absence of a high-quality reference genome has significantly hampered progress. This study aims to address this gap by generating a comprehensive chromosome-scale genome assembly of faba bean and utilizing it to investigate the genetic basis of crucial agronomic traits like seed size and hilum color, ultimately facilitating advancements in faba bean breeding.

Faba bean (*Vicia faba* L.) has been cultivated for over 10,000 years, showcasing remarkable adaptability and nutritional value. However, the lack of a known wild progenitor presents challenges for breeders. Studies have explored the genetic diversity within faba bean, revealing significant variation in seed size and other traits. Early research focused on identifying and characterizing individual genes related to important traits. Although some progress has been made in understanding genes related to tannin content and antinutrient levels, a high-quality reference genome has been absent, hindering a more comprehensive understanding of the genetic architecture of the faba bean genome and the underlying genetic mechanisms controlling agronomically relevant traits.

The study utilized PacBio HiFi long reads to sequence the genome of the inbred line 'Hedin/2' to 20-fold coverage. This resulted in an assembly of 11.9 Gb, with over half represented by contigs longer than 2.7 Mb. A genetic map and Hi-C data were used to scaffold the contigs into chromosome-scale pseudomolecules. Centromere locations were identified using chromatin immunoprecipitation sequencing. The assembly quality was evaluated using Merqury, revealing high completeness and accuracy. Similar procedures were undertaken for the cultivar 'Tiffany'. RNA sequencing data from nine diverse tissues were used for gene annotation. To understand genome size expansion, the study analyzed repeat elements, including long terminal repeat (LTR) retrotransposons and satellite repeats. Gene density, recombination rate, and DNA methylation patterns were investigated. The researchers developed a targeted genotyping assay using single primer enrichment technology (SPET) and applied it to a diversity panel of 197 faba bean accessions. Genome-wide association studies (GWAS) were conducted to identify genetic loci associated with seed size and hilum color. Comparative genomics was conducted by aligning the faba bean genome to those of other legumes. Phylogenetic analysis was performed to establish evolutionary relationships among the PPO (polyphenol oxidase) genes.

The study generated a high-quality chromosome-scale assembly of the 13-Gb faba bean genome, one of the largest diploid plant genomes. Genome size expansion was primarily attributed to the amplification of LTR retrotransposons, particularly the Ogre family, and satellite repeats. Despite its large size, the faba bean genome exhibits uniform gene density and recombination rates across chromosomes, except for heterochromatic regions. The study identified 34,221 protein-coding genes in the Hedin/2 assembly. The GWAS analysis of seed size revealed 15 marker-trait associations across multiple trials, with a prominent signal on chromosome 4 within the *Vfaba.Hedin2.R1.4g051440* gene, homologous to *Arabidopsis CYP78A* genes involved in seed size regulation. Analysis of hilum color revealed an association with a cluster of PPO genes. A 2-kb MITE insertion in the *VfPPO-2* promoter region was identified in the pale hilum genotype 'Tiffany', suggesting a causative mechanism for hilum color variation. DNA methylation was found to be highly efficient in faba bean, with high methylation levels across most genomic regions including transposable elements.

The findings significantly advance our understanding of the faba bean genome and its evolution. The high-quality genome assembly provides a valuable resource for breeders to accelerate genetic improvement. The discovery of genetic loci associated with seed size and hilum color opens avenues for marker-assisted selection. The identification of *Vfaba.Hedin2.R1.4g051440* as a candidate gene for seed size variation and the MITE insertion affecting *VfPPO-2* expression in hilum color variation highlights the potential for gene editing strategies. The uniform distribution of genes and recombination rates suggests that faba bean may be relatively amenable to genetic manipulation compared to some other large-genome species. The study's results demonstrate the power of integrating long-read sequencing, high-throughput genotyping, and GWAS in dissecting complex traits in crops with large genomes.

This research provides a comprehensive chromosome-scale genome assembly of faba bean, significantly advancing the genomics resources available for this important crop. The identified genetic loci associated with seed size and hilum color, along with the insights into the role of transposable elements and DNA methylation, open up opportunities for targeted breeding strategies and gene editing. Future research should focus on expanding the pan-genome to capture more genetic diversity and investigate the function of other candidate genes identified in this study. Deeper exploration of the interplay between transposable elements, DNA methylation, and gene expression is also warranted.

The study focused primarily on two genotypes ('Hedin/2' and 'Tiffany'), limiting the generalizability of some findings. Although the GWAS identified significant loci, environmental factors also influence seed size and hilum color, which were not fully accounted for in the study's model. Further research involving larger panels of diverse accessions and multi-environmental trials would strengthen the conclusions drawn.