Using the knowns to discover the unknowns: MS-based dereplication uncovers structural diversity in 17-hydroxygeranyllinalool diterpene glycoside production in the Solanaceae.

Abstract

Exploring the diversity of plant secondary metabolism requires efficient methods to obtain sufficient structural insights to discriminate previously known from unknown metabolites. De novo structure elucidation and confirmation of known metabolites (dereplication) remain a major bottleneck for mass spectrometry-based metabolomic workflows, and few systematic dereplication strategies have been developed for the analysis of entire compound classes across plant families, partly due to the complexity of plant metabolic profiles that complicates cross-species comparisons. 17-hydroxygeranyllinalool diterpene glycosides (HGL-DTGs) are abundant defensive secondary metabolites whose malonyl and glycosyl decorations are induced by jasmonate signaling in the ecological model plant Nicotiana attenuata. The multiple labile glycosidic bonds of HGL-DTGs result in extensive in-source fragmentation (IS-CID) during ionization. To reconstruct these IS-CID clusters from profiling data and identify precursor ions, we applied a deconvolution algorithm and created an MS/MS library from positive-ion spectra of purified HGL-DTGs. From this library, 251 non-redundant fragments were annotated, and a workflow to characterize leaf, flower and fruit extracts of 35 solanaceous species was established. These analyses predicted 105 novel HGL-DTGs that were restricted to Nicotiana, Capsicum and Lycium species. Interestingly, malonylation is a highly conserved step in HGL-DTG metabolism, but is differentially affected by jasmonate signaling among Nicotiana species. This MS-based workflow is readily applicable for cross-species re-identification/annotation of other compound classes with sufficient fragmentation knowledge, and therefore has the potential to support hypotheses regarding secondary metabolism diversification.