MOLECULAR PHYLOGENETIC TESTS OF FLORAL SCENT EVOLUTION IN THE SOLANACEAE

Independent evolutionary gains of floral scent compound production are common among angiosperms. These independent gains could be due to evolutionary convergence, whereby different species produce a similar volatile through the action of paralogous biosynthetic enzymes. However, it is also possible that these independent gains could be due to parallel evolution, whereby different species attain the ability to produce a compound using orthologous enzymes. The Solanaceae provide an excellent opportunity to investigate the hypothesis of parallel or convergent evolution because methyl salicylate (MeSA) and methyl benzoate (MeBA) likely evolved independently in the floral scent of multiple species. To test these hypotheses in the Solanaceae, we first assessed tissue-specific levels of expression of two paralogous genes that can produce MeSA and MeBA in vitro: salicylic acid methyltransferase (SAMT) and benzoic/salicylic acid methyltransferase (BSMT). SAMT was expressed at high levels in flowers relative to leaves in the MeSA/MeBA-producing species Cestrum nocturnum and Datura wrightii. BSMT was expressed at high levels in flowers relative to leaves in Nicotiana suaveolens and D. wrightii. Heterologous expression of the floral-expressed SAMT sequences from D. wrightii and C. nocturnum, and BSMT from N. suaveolens verified that the proteins encoded by these sequences catalyze the formation of MeSA from salicylic acid and MeBA from benzoic acid. Because paralogous MeSA/MeBA-producing sequences were expressed in C. nocturnum relative to N. suaveolens, the hypothesis of parallel evolution of MeSA production is rejected. Neither hypothesis could be rejected in the case of D. wrightii. This study shows that testing hypotheses of parallel and convergent evolution can benefit from gene expression data but is likely to require a multifaceted approach.