Convergence and constraint in glucosinolate evolution across the Brassicaceae

Abstract Diversity in plant specialized metabolites plays critical roles in plant–environment interactions. In longer evolutionary scales, e.g. between families or orders, this diversity arises from whole-genome and tandem duplication events. Less is known about the evolutionary patterns that shape chemical diversity at shorter scales, e.g. within a family. Utilizing the aliphatic glucosinolate pathway, we explored how the genes encoding the terminal structural modification enzyme GSL-OH evolved across the Brassicaceae and the genomic processes that control presence–absence variation of its products (R)-2-hydroxy-but-3-enyl and (S)-2-hydroxy-but-3-enyl glucosinolate. We implemented a phylo-functional approach to functionally validate GSL-OH orthologs across the Brassicaceae and used that information to map the genomic origin and trajectory of the locus. This uncovered a complex mechanism involving at least 3 ancestral loci with extensive gene loss across all species, creating unequal retention across the phylogenetic relationships. Convergent evolution in enantiomeric specificity was observed, where several independent species had tandem duplicates that diverged toward producing the R or S enantiomers. To explore potential biological differences between the enantiomers, we performed Trichoplusia ni larval choice assays and tested resistance against Botrytis cinerea in a detached leaf assay. We found that plants with the S-enantiomer were more susceptible to B. cinerea infection than to T. ni larval herbivory, while plants with the R-enantiomer seemed more susceptible to T. ni larval herbivory when compared to B. cinerea. Ultimately, we observed recurrent GSL-OH loss, uncovered a complex origin story for the gene, and measured the bioactivity of the enzyme's metabolic products.