Molecular biology of floral scent evolution: Characterization of linalool synthase (LIS) in diverse species.

Abstract

Flowers of Clarkia breweri as well as many other species emit linalool, an acyclic monoterpene, as part of the scent volatile mixture that attracts pollinators. Surprisingly, it is not uncommon for scented species to evolve from nonscented species. C. breweri, for example, has evolved relatively recently from an extant nonscented species, Clarkia concinna. Despite the importance of floral scent to the survival of many angiosperms, none of the enzymes or genes involved in the production of any floral scent compound had been characterized at the start of this project. Hence, this project focuses on one such enzyme, linalool synthase (LIS), and examines the general question of how nonscented species can give rise to scented species thereby changing the plant's specific pollinator. A cDNA of LIS1, the gene encoding LIS in C. breweri, was isolated, and subsequent Northern blots, in situ hybridizations, and Western blots showed that the strong floral scent of C. breweri is the result of (1) upregulation of a preexisting C. concinna LIS gene at the level of nucleic acid and (2) expansion of the types of tissues that express this gene. Organelle preparations treated with proteinase K, subcellular fractionations, and immunogold localizations together show that leucoplasts are the subcellular location of LIS1. However, although the N-terminus of LIS has the characteristics of a plastid targeting sequence, only a maximum of 8 amino acids are cleaved from the mature protein. Thus, LIS1 appears to have a non-cleavable plastid targeting sequence. LIS and LIS-like (LSL) genes isolated from C. breweri, C. concinna, Oenothera arizonica, and Arabidopsis thaliana encode proteins that are 40-96% identical to each other and have 11 introns in identical positions. Comparisons of these sequences with the sequences of other terpene synthases show that LIS and LSL are composite genes which evolved from a cross-over event between two different types of terpene synthases. Although known to occur, this is the first time such domain swapping has been demonstrated in terpene synthases. This study shows how scent can evolve in a relatively simple way through up-regulation of pre-existing genes and suggests that the combined evolutionary mechanisms of duplication, followed by divergence and/or domain swapping, may explain the extraordinarily large diversity of proteins found in the plant terpene synthase family.