Linking Chemistry and Genomics for the Study of Secondary Metabolism in Aromatic and Medicinal Plants

The study of the biosynthesis of secondary metabolites and the genes involved in these processes has been greatly facilitated by novel genomic approaches developed during the last years. Many of the biosynthetic pathways dedicated to secondary metabolism, and the enzymes involved in these pathways, have apparently evolved from the much better studied primary biosynthetic pathways. Therefore by exploiting similarities between functionally-related genes, it has been possible to isolate novel genes involved in the formation of unique natural products. To implement this novel approach, appropriate tissues in the proper physiological state, where the compounds of interest are produced in significant levels, is identified. Next, sequence information on large numbers (thousands) of different ESTs (expressed sequence tags) originating in these tissues is obtained. The information obtained is en masse examined using bioinformatic computer algorithms. Predictions on the physiological and biochemical role of individual ESTs are then made based on DNA similarities, and the patterns of expression of individual ESTs. Identity and biochemical function of the particular EST in question can then be confirmed by functional expression experiments. A few examples of such genomic projects aimed at isolating and characterizing genes involved in the formation of key metabolites are reviewed. Some of the genes responsible for the formation of the volatile phenylpropenes prominent in the essential oil of sweet basil and in the formation of the many compounds that compose the fragrance of roses have been identified utilizing this approach. The potential of utilizing genes that code for the formation of volatile compounds, for the improvement of the quality properties of aromatic plants and other agricultural produce, are discussed. INTRODUCTION Plants produce and accumulate an enormous variety of secondary metabolites (often referred as “natural products”). Many of these products have well defined ecological roles in plant defense and in mediating the plant’s interactions with other organisms, while others have yet unknown biological roles. More than 45,000 different chemical structures of natural products have been identified. The largest group of natural products is the terpenes, with more than 25,000 structures elucidated. Additionally, more than 2,000 alkaloids and about 8,000 phenolic derivatives are known (Croteau et al., 2000). Despite the large number of different proven chemical structures, there are an amazingly low number of biochemical pathways by which these compounds are biosynthesized. This apparent inconsistency can be easily explained by the fact that plants produce all their metabolites starting from water, CO2, a few minerals and solar energy. It is seemingly that during the course of evolution, the same basic biosynthetic pathways have developed to allow for the production of different metabolites, by only minor changes in the basic core of the pathways (Pichersky and Gang, 2000). Proc. XXVI IHC – Future for Medicinal and Aromatic Plants Eds. L.E. Craker et al. Acta Hort. 629, ISHS 2004 Publication supported by Can. Int. Dev. Agency (CIDA) 436 CHEMICAL BIOSYNTHESIS Terpenes are biosynthesized by the isopentenoid pathway, that includes two main metabolic branches, the mevalonic acid pathway, by which sesquiand triterpenes are biosynthesized, and the deoxyxylulose diphosphate pathway,by which mono-, di-, and tetra-terpenes are formed (Croteau et al., 2000). Most phenolic derivatives are biosynthesized either by the shikimic acid pathway or through the malonate-acetate pathway and most alkaloids are biosynthesized from amino acids. Monoterpenoids such as linalool, menthol, thymol, and limonene are derived from geranyl diphosphate (GPP, Fig. 1). The different conversions from GPP are catalyzed by a group of enzymes termed monoterpene synthases. These enzymes share many properties, such as cofactor requirements, molecular size and protein sequence similarity, but still they are different enough to allow for the catalysis of the different products according to their specificity (McGarvey and Croteau, 1995). Sesquiand triterpenes, as well as many triterpene-derived metabolites such as the saponins and sterols, are synthesized from the 15-carbon intermediate farnesyl diphosphate (FPP). Similarly to monoterpene synthases, minor changes in the sesquiterpene synthase enzymes (and their genes) are responsible for catalyzing the conversion of FPP to the different sesquiterpenes respectively. Still, at least in angiosperms, sesquiterpene synthase genes are significantly similar to each other, and also similar (to a lesser extent) to angiosperm monoterpene synthases (Bohlmann et al., 2000a). In an analogous way, it seems that the ubiquitous pathway to lignin has been specifically adapted in many plants for the production of unique natural products, again, by small but important modifications of existing genes and enzymes, to allow for the specific conversions (Fig. 2). Thus, phenylpropanoids are mostly derived from Lphenylalanine, not only a precursor of lignin, anthocyanins and other flavonoids, but also, a precursor of t-anethole in anise (Pimpinella anisum) (Manitto et al., 1974a) and in fennel (Foeniculum vulgare) (Kaneko, 1960). L-phenylalanine is also a precursor of cinnamaldehyde in cinnamon (Cinnamomum zeylanicum, Lauraceae Senanayake et al., 1977), and of estragole in sweet basil (Ocimum basilicum, Lamiaceae (Manitto et al., 1974b). GENE CODING At times, the same compound is present in many unrelated organisms. For example, eugenol, a phenylpropanoid compound possessing a very strong pungent aroma reminiscent of cloves, is indeed one of the major components of clove essential oil, but it is also present in lower levels in banana, cinnamon leaf, pimento and other plants (Senanayake et al., 1977, Tucker et al 1991, Jordan et al., 2001). At times, small chemical differences in the metabolites might cause dramatic differences in the biological activity or the aroma of the compound in question. For example, high levels of eugenol are detrimental to the aroma of tomato, rending it reminiscent of cloves. In contrast, its methoxylated derivative methyl eugenol, has a soft scent, reminiscent of cut grass (Fig. 3). Interestingly, the enzyme EOMT (SAM: eugenol O-methyltransferase is able to transfer a methyl group from S-adenosylmethionine (SAM) to release methyl eugenol, converting a pungent compound (eugenol) into a mild-scented one (methyl eugenol) (Fig. 3). This enzyme and its gene have been characterized in sweet basil varieties that accumulate methyl eugenol in their essential oil (Lewinsohn et al., 2000, Gang et al., 2001). The enzyme also accepts other substrates especially para substituted phenols. EOMT can also transfer a methyl group to the p-position of chavicol, releasing estragole, a major constituent of sweet basil essential oil. Thus, it seems that the same enzyme can produce either estragole and/or methyl eugenol, depending on the availability of the corresponding substrate acceptor. The gene that encodes for the EOMT enzyme has been isolated and characterized. It belongs to a relatively large gene family that includes many O-methyltransferases, specific for diverse phenolic and flavonoid substrate acceptors. A very closely related gene, also isolated from sweet basil is 90 % similar to EOMT at the protein level (Gang et