Haplotype-resolved genome assembly provides insights into the evolution of S-locus supergene in distylous Nymphoides indica.

Distyly has evolved independently in numerous animal-pollinated angiosperm lineages. Understanding of its molecular basis has been restricted to a few species, primarily Primula. Here, we investigate the genetic architecture of the single diallelic locus (S-locus) supergene, a linkage group of functionally associated genes, and explore how it may have evolved in distylous Nymphoides indica, a lineage of flowering plants not previously investigated. We assembled haplotype-resolved genomes, used read-coverage-based genome-wide association study (rb-GWAS) to locate the S-locus supergene, co-expression network analysis to explore gene networks underpinning the development of distyly, and comparative genomic analyses to investigate the origins of the S-locus supergene. We identified three linked candidate S-locus genes - NinBAS1, NinKHZ2, and NinS1 - that were only evident in the short-styled morph and were hemizygous. Co-expression network analysis suggested that brassinosteroids contribute to dimorphic sex organs in the short-styled morph. Comparative genomic analyses indicated that the S-locus supergene likely evolved via stepwise duplications and has been affected by transposable element activities. Our study provides novel insight into the structure, regulation, and evolution of the supergene governing distyly in N. indica. It also provides high-quality genomic resources for future research on the molecular mechanisms underlying the striking evolutionary convergence in form and function across heterostylous taxa.