Metabolomics, genomics, proteomics, and the identification of ...

Metabolomics, genomics, proteomics, and the identification of enzymes and their substrates and products Eyal Fridman and Eran Pichersky A large proportion of the genes in any plant genome encode enzymes of primary and specialized (secondary) metabolism. Not all plant primary metabolites, those that are found in all or most species, have been identified. Moreover, only a small portion of the estimated hundreds of thousand specialized metabolites, those found only in restricted lineages, have been studied in any species. The correlative analysis of extensive metabolic profiling and gene expression profiling has proven a powerful approach for the identification of candidate genes and enzymes, particularly those in secondary metabolism. The final characterization of substrates, enzymatic activities, and products requires biochemical analysis, which has been most successful when candidate proteins have homology to other enzymes of known function. The challenges are to identify new types of enzymes and to develop biochemical techniques that are suitable for large-scale analysis. Addresses Department of Molecular, Cellular and Developmental Biology, University of Michigan, Ann Arbor, Michigan 48109-1048, USA Corresponding author: Pichersky, Eran ([email protected])

Current Opinion in Plant Biology 2005, 8:242–248 This review comes from a themed issue on Physiology and metabolism Edited by Toni Kutchan and Richard Dixon Available online 31st March 2005 1369-5266/$ – see front matter # 2005 Elsevier Ltd. All rights reserved. DOI 10.1016/j.pbi.2005.03.004

Introduction: the significance of identifying enzymes and substrates All cellular processes involve, to a very large extent, the chemical transformation of small and large molecules, which occurs through the action of enzymes. Enzymes constantly carry out reactions to produce energy (e.g. in glycolysis and the TCA cycle), building blocks (e.g. amino acid and fatty acid synthesis) and large-scale structures (e.g. by the synthesis of polynucleic acids, proteins and cell walls). They are also involved in other universal processes of cellular metabolism and in reactions that are specific to individual tissues and cell types (e.g. the production of defense compounds, scent compounds, pigments and so on). To understand how plant cells function, the myriad actors in these metabolic Current Opinion in Plant Biology 2005, 8:242–248

processes — enzymes and their substrates and products — must be elucidated. Studies that begin with the identification of a gene often do not identify the function of this gene at the biochemical level. Often, biochemical function is simply inferred from homology to previously investigated sequences, sometimes to a sequence that is several degrees removed (i.e. homology to an unknown sequence with homology to. . .. . . a gene of known function). In other cases, when a morphological phenotype is observed in a mutant, the gene ‘function’ is simply defined as being responsible for, or involved in, bringing about the structure or developmental process that is defective in the mutant, without any specific biochemical function being assigned to the protein encoded by the gene. A lack of biochemical attention is perhaps understandable when the protein is deemed (on the basis of homology) to be a DNAbinding protein or some other kind of ‘transcription factor’, but these cases are the minority as only a small fraction of the genes in the plant genome are thought to encode such factors [1]. Achieving the goal of cataloguing all of the components of the cell — genes, proteins (of which enzymes are surely the majority), and metabolites — and of elucidating all the causal relationships among them will require a vast effort. Most biological research to date has been the piecemeal elucidation of components and causal relationships within a very small and circumscribed subset of cellular pathways. Several recent approaches, however, have been based on ‘systems biology’ in which a very large number of components are catalogued and statistical methods are used to infer correlations, which in turn suggest further types of investigation. This approach has been most prominent in the sequencing of whole genomes (including those of two plant species, Arabidopsis and rice), followed by computer analysis of the coding information of the genome and by the analysis of the expression of the entire set of genes by means of DNA microarrays. Such analyses might provide some clues regarding the cellular processes that a given gene might be involved in but they cannot positively identify the enzymatic activity of the protein it encodes [2]. Assigning biochemical function(s) to the proteins must clearly be an important next step. The recent addition of metabolic profiling to the systemic biology approach is making a strong contribution toward achieving this goal. Here, we review the main methodological approaches, and highlight some recent successes, in identifying new plant enzymes and their substrates and products. www.sciencedirect.com

Identification of enzymes and their substrates and products Fridman and Pichersky 243

Old and new approaches to matching enzymes with substrates and products — what constitutes a proof? DNA, RNA and proteins each constitute a class of compounds that have some structural properties in common, thus permitting the development of analytical methods that apply to basically all members of the class. As pointed out in an excellent recent review [3], the metabolites within the cell have no shared chemical features on which a general isolation-separation-identification method can be developed. Thus, current metabolic profiling techniques [2,4] allow for the extraction and separation of only a small fraction of plant metabolites, and only a fraction of those have been identified. But even with the current limitations, the large number of recent metabolic profiling studies have resulted in the discovery of many novel compounds from a multitude of wild and domesticated plant species. Some of the identified compounds must be part of primary metabolism, that is, biochemical pathways that are shared by all or most plant species, but many more are specialized metabolites that are found only in restricted lineages. A similar problem of standardization applies to the study of the enzymatic activity of proteins. Each enzyme typically has a narrow range of conditions (e.g. a pH optimum or the presence of cofactors) in which it acts on a distinct substrate. In addition, the products of various enzymatic reactions usually do not have general properties (e.g. absorption spectra) that can be measured in a uniform way. In the future, it is hoped that methods will be developed to test a large number of proteins for enzymatic function, in the same way as protein microarrays are used to test for protein–protein interactions [5]. Some progress has been made in standardizing product analysis, for example, in producing reagents that give uniform color product in all assays of methyltransferases (MT) that use S-adenosyl-L-methionione [6]. At present, however, enzyme assays have to be developed and carried out piecemeal. The ‘classical’ approach to discovering enzymes is to start with a given product and to ask what enzymatic reaction is responsible for its formation, and what is the substrate. Substrates can be hypothesized on the basis of biochemical principles and current knowledge of metabolic pathways and types of enzymes. For example, it could be hypothesized that a newly discovered ester that contains an acyl moiety and an alcohol moiety is formed from the free alcohol and the acyl-CoA by the action of an acyltransferase, because enzymes that produce similar products from similar substrates are already known. Alternatively, labeled precursor-feeding experiments or the analysis of metabolism in mutants might point to a possible substrate, which is then used to develop an enzymatic assay. Approaches that use precursor labeling or even theoretical considerations have often resulted in a www.sciencedirect.com

correct deduction of the pathway, but a given step can only be considered to be ‘proven’ once an enzyme has been identified and shown to catalyze the hypothesized reaction (or, on rare occasions, if the reaction is shown to proceed non-enzymatically). This is an extremely important point because there have been many cases in which precursor labeling has led to erroneous pathway construction, often because low-abundance intermediates were missed or owing to other complications. Biochemical work that combined both enzymatic analysis and protein sequence determination has identified many classes of structurally related enzymes, such as dehydrogenases, methyltransferases and so on. Thus, when a novel protein is found to be homologous to known enzymes, this information can contribute to the elucidation of its potential activity. In specialized (secondary) metabolism, however, a high degree of similarity with a known enzyme is no guarantee that the protein has the same substrate or even the same catalytic function, and highly divergent proteins might catalyze the same reaction in different species [7–9]. A further complication in secondary metabolism is that enzymes can often act on more than one substrate in vitro as well as in vivo, sometimes with incomplete overlap of the two sets of substrates [10,11]. The ‘correct’ identity of the in vivo substrate (or substrates) must therefore rely on several lines of evidence. Such evidence may include the Km value of the enzyme with the candidate substrates in respect of the measured concentrations of these substrates in the cell and, when possible, the results of metabolic profiling of lines in which the gene is mutated or overexpressed.

Recent successes in identifying enzymatic and substrates from genomics, proteomics, and metabolomics data Despite the caveats listed above, several recent studies have made good use of genomic and metabolomic data to unambiguously assign enzymatic function to proteins that are encoded by newly discovered genes. Most of these recent discoveries come from the field of specialized metabolism. The approach to the metabolic profiling portion of these investigations was similar (i.e. a targeted survey approach), as was the final step of the enzymological proof, but there were some differences in the type and source of sequence and gene-expression data used to make the initial correlation among metabolites, genes and enzymes (Figure 1). Below, we describe these different approaches. Genome sequence data combined with metabolic and gene expression profiling

Arabidopsis has been used extensively to study primary metabolism but secondary metabolism is more often studied in ‘exotic plants’. Once genes that encode Current Opinion in Plant Biology 2005, 8:242–248

244 Physiology and metabolism

Figure 1

Plant material: Organ/cell types at specific developmental stage or under specific biotic/abiotic conditions

f no itio ta s i qu da Ac ew n

Metabolic profiling: identification and quantification

Ac

Previously available databases Genome sequence, ESTdatabases, proteomes, established biochemistry

qu ne isitio w da n of ta

EST database, transcription profiling, proteomics: identification and quantification

Correlate expression/protein with metabolite

Deduce hypothetical reaction leading to the synthesis of the metabolite (based on enzyme/substrate similarities)

Select candidate sequence

Produce and purify enzyme

In planta genetic experiments: complementation or suppression, mapping/association analysis

Develop enzymatic assay: procure substrates and methods for identifying and monitoring product formation

In vitro biochemical characterization Current Opinion in Plant Biology

A schematic diagram of the steps involved in the identification of enzymes and their substrates and products.

enzymes that are involved in the biosynthesis of a class of secondary metabolites (e.g. alkaloids) are found in such species, however, sequence comparisons often reveal that the Arabidopsis genome contains related sequences [12]. Such a situation was encountered with terpene synthases (TPSs), a family of enzymes that catalyze the formation of mono-, sesqui-, and diterpenes, a class of volatile and nonvolatile compounds that are involved in many ecological functions. TPSs have been isolated from numerous plant species. Arabidopsis has at least 30 genes with homology to TPS genes [13], yet until recently there were no reports of terpenes (other than gibberellins) being found in or emitted from Arabidopsis plants. In a recent study, Chen and co-workers [14] were able to detect the emission of terpenes from Arabidopsis flowers. These terpenes included three monoterpenes (b-miocene, cymene, and linalool) and many sesquiterpenes (b-caryophyllene being the most abundant). The expression profiles of Current Opinion in Plant Biology 2005, 8:242–248

all of the AtTPS genes showed several of them to be completely or almost completely flower-specific. Identifying the substrates of the floral-specific AtTPSs was facilitated by the fact that all monoterpene synthases use geranyl diphosphate (GPP) as a substrate, whereas all sesquiterpene synthases use farnesyl diphosphate (FPP). Furthermore, monoterpene synthases are localized to the plastids and, therefore, are made initially as precursors that have a transit peptide that directs them to this compartment, and such transit peptides are relatively easy to discern from the gene sequences. The investigators thus expressed these genes in Escherichia coli, after obtaining full-length cDNAs by RT–PCR, and systematically tested each protein with the appropriate substrate. In this way, one monoterpene synthase was identified whose product is linalool, and a second monoterpene synthase was identified whose products include b-myrcene and ocimene as well as several other www.sciencedirect.com


monoterpenes occasionally observed in A. thaliana floral scent. Similarly, the products of two sesquiterpenes have been identified; both make multiple products (a common occurrence with TPSs) that together account for all of the sesquiterpenes emitted from Arabidopsis flowers. A similar biochemical genomics approach was adopted by the same group to identify benzoic acid/salicylic acid methyltransferase (BSMT), the enzyme that catalyzes the methyl esterification of salicylic acid (SA) and benzoic acid (BA) in Arabidopsis [15]. Methylsalicylate (MeSA) and occasionally methylbenzoate (MeBA) are emitted from the leaves of plants that are under attack by insects or pathogens [16,17]. The benzenoid acid MTs that have been identified in other species belong to the SABATH (Type III) MT family [18]. The Arabidopsis genome has 24 genes that belong to this family [18]. Treatments with herbivores and with fungal elicitors were used to elicit detectable levels of emission of MeSA and methylbenzoic acid (MeBA) from Arabidopsis leaves. The expression profiling of all 24 genes during these treatments, as well as under control conditions in which no MeSA and MeBA emission occurred, was conducted by RT–PCR. There was only one AtSABATH MT gene with an unknown function whose expression was always positively correlated with such emission. A cDNA of this gene was expressed in E. coli and shown to have MT activity with SA and BA with physiologically relevant Km values. cDNA libraries or EST databases combined with metabolic and gene expression profiling of specific tissues and treatments

When a complete genome sequence is not available, as is the case for most plant species, cDNA libraries and, increasingly, expressed sequence tag (EST) databases have been used as the source of DNA sequence information. Metabolic profiling performed by Martin et al. [19] on young Norway spruce trees treated with methyljasmonate (MeJA) showed the emission of a large number of monoterpenes and sesquiterpenes, as well as the synthesis of non-volatile diterpenes. This group proceeded to screen a cDNA library of young spruce shoots and leaves by a combination of homology-based PCR and DNA-hybridization techniques, thereby isolating nine TPS cDNAs. Each cDNA was expressed in E. coli and tested with the appropriate substrate (GPP, FPP or geranylgeranyl diphosphate, the substrate of diterpene synthases). The results of these assays indicated that four of the cDNAs encoded monoterpene synthases, three cDNAs encoded sesquiterpene synthases, and two cDNAs encoded diterpene synthases. Each of the enzymes encoded by these cDNAs had a unique product profile, which together accounted for most of the major components of the constitutive and MeJA-induced terpenoids in Norway spruce. An elegant investigation that resulted in the identification of two new sesquiterpene synthases from Cucumis sativus www.sciencedirect.com

(cucumber) was reported by Mercke et al. [20]. Subtractive cDNA libraries were made to identify leaf genes that are up- and downregulated during mite infestation, which induces the synthesis and emission of several terpenes and fatty-acid derivatives. The approximately 700 cDNAs thus identified were placed on a microarray and probed with mRNA derived from leaves that were treated with jasmonic acid, mechanically damaged, or infested with mites. The volatiles emitted from the leaves during these treatments were profiled, and these data were correlated with gene expression. Candidate bcaryophyllene synthase and a-farnesene synthase cDNAs were then sequenced. Full-length cDNAs were obtained from those that had homology with TPSs, these cDNAs were then expressed in E. coli and their activity verified using FPP as a substrate. Roses are known for having a wide range of scent volatiles. Guterman et al. [21] recently characterized very distinct volatile profiles from two rose cultivars, and then proceeded to construct EST databases and to compare the expression profiles of these ESTs in the two cultivars using DNA microarrays. Candidate genes were selected on the basis of correlations between levels of expression and volatile emission, as well as membership of a gene family whose members in other species were known to encode enzymes that synthesize similar compounds. In this way, a cDNA with sequence similarity to sesquiterpene synthases was identified as potentially encoding germacrene D synthase, a cDNA with similarity to acyltransferases was identified as encoding the enzyme potentially responsible for the formation of geranylacetate from geraniol and acetyl CoA [22], and a pair of cDNAs with similarity to Type I MTs were identified as potentially encoding the enzymes that catalyze the two successive methylations of orcinol (3,5-dihydroxy toluene) to make the rose volatile dimethoxytoluene [23]. The orcinol MTs were also identified by another group, using similar methods [24]. All of these enzymes were then produced in E. coli, and the enzymatic activity with the suspected substrates was demonstrated, as was a lack (or low level) of activity with related substrates. Another example of the use an EST database in combination with metabolic profiling and biochemical analysis comes from Stevia rebaudiana [25]. The leaves of this plant make glycosylated diterpenes, called steviol glycosides, that taste intensely sweet to humans. Metabolic profiling showed a large number of such compounds, with the diversity coming from the number of sugars compounds that are bound to the steviol moiety, as well as from additional glycosylations of the glucose moiety itself. Glycosylation in plants is known to be catalyzed by UDP-glucosyltransferases (UGT), but several families of UGTs that have little sequence similarity to each other have been discovered. A bioinformatic analysis identified 17 putative UGTs in the Stevia rebaudiana leaf EST Current Opinion in Plant Biology 2005, 8:242–248


database. A meticulous biochemical analysis of 12 of these enzymes with various substrates identified the unique functions of three UGTs that belong to different UGT families in the biosynthesis of various steviol glycosides. This analysis showed that these UGTs have a very limited but not absolute substrate specificity.

examined, and those cDNAs derived from genes whose expression correlated with alkaloid biosynthesis were expressed in baculovirus expression system and assayed for enzymatic activity with candidate substrates. MTs with physiologically relevant Km values for (R,S)-reticuline and (R,S)-norcoclaurine were thus identified.

EST databases combined with metabolic and gene expression profiling of specialized cell types

Conclusions

While EST databases derived from whole organs and metabolic and gene expression profiling of such organs have proven useful for identifying abundant enzymes of major biochemical pathways, some pathways are confined to only a few specialized cells and specific information pertaining to such metabolic pathways tends to be diluted and therefore difficult to discern. However, when specialized cells can be separated from the rest of the plant, sequence, metabolic, and gene expression information obtained from such cells prove extremely powerful in identifying enzyme function. Above-ground organs of basil contain glands that synthesize and store defense compounds, including methylated phenylpropenes and terpenes. Different cultivars have been bred that have distinct chemical compositions in the glands. Gang et al. [26] and Iijima et al. [27,28] isolated glands from three different cultivars, prepared EST databases from them, and profiled their metabolites. Two unique cDNAs with sequence homology to Type I MTs were found to be highly expressed in the glands of a cultivar that synthesizes the phenylpropenes methylchavicol and methyleugenol, but were not expressed in the glands of the other two cultivars that do not make these compounds. Upon expression in E. coli and in vitro enzymatic assays with potential substrates, one cDNA was shown to encode an enzyme that methylates chavicol to produce methylchavicol, and the other to encode an enzyme that methylates eugenol to make methyleugenol. Similarly, a total of nine TPSs (i.e. five monoterpene synthases and four sesquiterpene synthases) were found to produce specific products in the different cultivars. Proteomics combined with metabolic and gene expression profiling of specialized cell types

Ounaroon et al. [29] identified Type I MTs that are involved in the biosynthesis of alkaloids in opium poppy using a proteomics approach. A wide variety of alkaloids are produced in the latex, which is a cytoplasmic material that is exuded from specialized cells in the capsule. Latex proteins were resolved on two-dimensional gels, and individual protein spots that matched the expected size of MTs were excised and partially sequenced. Peptide sequences from some of the proteins indicated homology to Type I MTs, and oligonucleotides were designed on the basis of these sequences to obtain cDNAs. The expression of the genes encoding these cDNAs was Current Opinion in Plant Biology 2005, 8:242–248

The examples reviewed here demonstrate that, even with all the new methodologies of metabolic and gene expression profiling, the identification of the enzymatic properties of a newly discovered protein still depends on prior biochemical knowledge relating to the family of enzymes to which the candidate protein belongs. Thus, many new methyltransferases, terpene synthases, acyl transferases, and glucosyl transferases have been discovered because we already know representative enzymes of these types and can therefore make an informed guess regarding the potential substrates with which to test candidate proteins. Typically, however, the prototypic member of the family had been characterized by the ‘classical’ biochemical approach, which starts with the development of an enzymatic assays, follows with protein purification, and culminates with protein sequencing. When profiling data implicates a protein with no structural similarity to a known enzyme in the synthesis of a given metabolite, it is much more difficult to identify a substrate and to devise an enzymatic assay. Such inferences should still be possible to make on chemical principles, but we are not aware of a recent project that has employed metabolic profiling to identify the substrate and the reaction of a novel enzyme with no homology to any known protein family. Metabolic profiling of mutants was not discussed in this review, although this is a classical method of enzyme discovery that has had a long and successful record. New methods of inactivating gene expression, such as the insertion of T-DNAs and transposable elements and RNA interference (RNAi) techniques, could be extremely powerful in combination with metabolic profiling techniques. However, gene identification by such techniques must also be followed by biochemical characterization of the protein [30] because the lack of final product in the mutant does not guarantee that the substrate has been correctly identified [31,32]. Gene suppression techniques might also lead to misleading results if several similar genes that encode enzymes with different substrates are suppressed by the same construct. In summary, the combination of the new techniques of metabolic and gene expression profiling with classical techniques of enzymatic analysis (summarized in Figure 1) will allow the identification of the function of the majority of the genes in plant genomes. The challenge is to develop standardized and automatable methods for enzyme analysis. www.sciencedirect.com


Acknowledgements Work in the authors’ laboratory has been funded by National Science Foundation (NSF) grants MCB-0312466 and IBN 0211697, NRICGPUSDA grant 2001-35318-10006, and USDA-BARD grant IS-3332-02C. Eyal Fridman was supported by a Vaadia-BARD postdoctoral fellowship FI-328-2002.

References and recommended reading Papers of particular interest, published within the annual period of review, have been highlighted as: of special interest of outstanding interest 1.

Arabidopsis Genome Initiative: Analysis of the genome sequence of the flowering plant Arabidopsis thaliana. Nature 2000, 408:796-815.

2.

Sumner LW, Mendes P, Dixon RA: Plant metabolomics: large-scale phytochemistry in the functional genomics era. Phytochemistry 2003, 62:817-836.

3.

Trethewey RN: Metabolite profiling as an aid to metabolic engineering in plants. Curr Opin Plant Biol 2004, 7:196-201.

4.

Fernie AR, Trethewey RN, Krotzky AJ, Willmitzer L: Metabolite profiling: from diagnostics to systems biology. Nat Rev Mol Cell Biol 2004, 5:763-769.

5.

Figeys D: Novel approaches to map protein interactions. Curr Opin Biotechnol 2003, 14:119-125.

6.

Hendricks CL, Ross JR, Pichersky E, Noel JP, Zhou ZS: An enzyme-coupled colorimetric assay for S-adenosylmethionine-dependent methyltransferases. Anal Biochem 2004, 326:100-105.

7.

Gang DR, Lavid N, Zubieta C, Chen F, Beuerle T, Lewinsohn E, Noel JP, Pichersky E: Characterization of phenylpropene O-methyltransferases from sweet basil: facile change of substrate specificity and convergent evolution within a plant O-methyltransferase family. Plant Cell 2002, 14:505-519.

8.

Broun P, Shanklin J, Whittle E, Somerville C: Catalytic plasticity of fatty acid modification enzymes underlying chemical diversity of plant lipids. Science 1998, 282:1315-1317.

9.

Pichersky E, Gang DR: Genetics and biochemistry of secondary metabolites in plants: an evolutionary perspective. Trends Plant Sci 2000, 5:439-445.

10. Liu CJ, Dixon RA: Elicitor-induced association of isoflavone O-methyltransferase with endomembranes prevents the formation and 7-O-methylation of daidzein during isoflavonoid phytoalexin biosynthesis. Plant Cell 2001, 13:2643-2658. 11. Humphreys JM, Hemm MR, Chapple C: New routes for lignin biosynthesis defined by biochemical characterization of recombinant ferulate 5-hydroxylase, a multifunctional cytochrome P450-dependent monooxygenase. Proc Natl Acad Sci USA 1999, 96:10045-10050. 12. Facchini PJ, Bird DA, St.-Pierre B: Can Arabidopsis make complex alkaloids? Trends Plant Sci 2004, 9:116-122. 13. Aubourg S, Lecharny A, Bohlmann J: Genomic analysis of the terpenoid synthase (AtTPS) gene family of Arabidopsis thaliana. Mol Genet Genomics 2002, 267:730-745. 14. Chen F, Tholl D, D’Auria JC, Farooq A, Pichersky E, Gershenzon J: Biosynthesis and emission of terpenoid volatiles from Arabidopsis flowers. Plant Cell 2003, 15:481-494. The authors report the first demonstration of volatile terpenes in Arabidopsis, and the identification of some of the TPS enzymes that are involved in their biosynthesis. Gene expression profiling showed that various TPS genes are expressed in all organs of the plant, including roots. The floral scent in Arabidopsis could be involved in attracting pollinators to this largely self-pollinating species. 15. Chen F, D’Auria JC, Tholl D, Ross JR, Gershenzon J, Noel JP, Pichersky E: An Arabidopsis thaliana gene for methylsalicylate biosynthesis, identified by a biochemical genomics approach, has a role in defense. Plant J 2003, 36:577-588. www.sciencedirect.com

16. Shulaev V, Silverman P, Raskin I: Airborne signalling by methyl salicylate in plant pathogen resistance. Nature 1997, 386:738. 17. Van Poecke RM, Posthumus MA, Dicke M: Herbivore-induced volatile production by Arabidopsis thaliana leads to attraction of the parasitoid Cotesia rubecula: chemical, behavioral, and gene-expression analysis. J Chem Ecol 2001, 27:1911-1928. 18. D’Auria JC, Chen F, Pichersky E: The SABATH family of MTs in Arabidopsis thaliana and other plant species. In Recent Advances in Phytochemistry. Vol 37. Edited by Romeo JT. Elsevier Science Ltd; 2003:253-283. 19. Martin DM, Faldt J, Bohlmann J: Functional characterization of nine Norway Spruce TPS genes and evolution of gymnosperm terpene synthases of the TPS-d subfamily. Plant Physiol 2004, 135:1908-1927. A particularly large number of TPS genes and enzymes from one species were identified in this study. 20. Mercke P, Kappers IF, Verstappen FW, Vorst O, Dicke M, Bouwmeester HJ: Combined transcript and metabolite analysis reveals genes involved in spider mite induced volatile formation in cucumber plants. Plant Physiol 2004, 135:2012-2024. The authors combine the statistical methods of correlation analysis and self-organizing maps (SOMs) to link genes with specific metabolites to successfully identify two TPS enzymes and their products. 21. Guterman I, Shalit M, Menda N, Piestun D, Dafny-Yelin M, Shalev G, Bar E, Davydov O, Ovadis M, Emanuel M et al.: Rose scent: genomics approach to discovering novel floral fragrance-related genes. Plant Cell 2002, 14:2325-2338. The authors of this paper and the following two papers [22,23] describe the use of microarray and metabolic profiling techniques in combination with varietal variations in a commercially important species to identify genes and enzymes for consumer-valued traits. 22. Shalit M, Guterman I, Volpin H, Bar E, Tamari T, Menda N, Adam Z, Zamir D, Vainstein A, Weiss D et al.: Volatile ester formation in roses. Identification of an acetyl-coenzyme A:geraniol/ citronellol acetyltransferase in developing rose petals. Plant Physiol 2003, 131:1868-1876. 23. Lavid N, Wang J, Shalit M, Guterman I, Bar E, Beuerle T, Menda N, Shafir S, Zamir D, Adam Z et al.: O-methyltransferases involved in the biosynthesis of volatile phenolic derivatives in rose petals. Plant Physiol 2002, 129:1899-1907. 24. Scalliet G, Journot N, Jullien F, Baudino S, Magnard JL, Channeliere S, Vergne P, Dumas C, Bendahmane M, Cock JM et al.: Biosynthesis of the major scent components 3,5-dimethoxytoluene and 1,3,5-trimethoxybenzene by novel rose O-methyltransferases. FEBS Lett 2002, 523:113-118. 25. Richman AS, Swanson A, Humphrey T, Chapman R, McGravey B, Pocs R, Brandle JE: Functional genomics uncovers three glucosyltransferases involved in the synthesis of the major sweet glucosides of Stevia rebaudiana. Plant J 2005, 41:56-67. The authors carried out an extensive bioinformatic search for candidate UDP-glucosyltransferase (UGT) genes, and biochemically identified three very divergent UGTs that are involved in the same pathway. 26. Gang DR, Wang J, Dudareva N, Nam KH, Simon JE, Lewinsohn E, Pichersky E: An investigation of the storage and biosynthesis of phenylpropenes in sweet basil. Plant Physiol 2001, 125:539-555. 27. Iijima Y, Gang DR, Fridman E, Lewinsohn E, Pichersky E: Characterization of geraniol synthase from the peltate glands of sweet basil. Plant Physiol 2004, 134:370-379. The authors constructed EST databases from glands of three different basil cultivars, and were able to identify a cDNA that encodes a novel TPS that is unique to just one of these cultivars. 28. Iijima Y, Davidovich-Rikanati R, Fridman E, Gang DR, Bar E, Lewinsohn E, Pichersky E: The biochemical and molecular basis for the divergent patterns in the biosynthesis of terpenes and phenylpropenes in the peltate glands of three cultivars of basil. Plant Physiol 2004, 136:3724-3736. Further use of the EST databases from the three basil cultivars described in [27] resulted in the identification of eight additional TPS enzymes and their products, some of which were unique to one cultivar while others were found in more than one cultivar. Current Opinion in Plant Biology 2005, 8:242–248


29. Ounaroon A, Decker G, Schmidt J, Lottspeich F, Kutchan TM: (R,S)-Reticuline 7-O-methyltransferase and (R,S)-norcoclaurine 6-O-methyltransferase of Papaver somniferum — cDNA cloning and characterization of methyl transfer enzymes of alkaloid biosynthesis in opium poppy. Plant J 2003, 36:808-819. A successful proteomics approach resulted in the identification of new proteins and their enzymatic activities. 30. Simkin AJ, Schwartz SH, Auldridge M, Taylor MG, Klee HJ: The tomato CCD1 (CARTENOID CLEAVAGE DIOXGENASE 1) genes contribute to the formation of the flavor volatiles b-ionone, pseudoionone and geranylacetone. Plant J 2004, 40:882-892. Using antisense technology on a candidate gene, the authors demonstrate for the first time that volatile terpenes can be derived from the

Current Opinion in Plant Biology 2005, 8:242–248

degradation of carotenoids in planta. The authors also provide biochemical proof of the activity of the enzyme encoded by the gene by performing in vitro assays and by expressing the gene in an E. coli strain engineered to make carotenoids. They showed that these carotenoids were degraded in E. coli when the plant gene was expressed. 31. Franke R, Humphreys JM, Hemm MR, Denault JW, Ruegger MO, Cusumano JC, Chapple C: The Arabidopsis REF8 gene encodes the 3-hydroxylase of phenylpropanoid metabolism. Plant J 2002, 30:33-45. 32. Schoch G, Goepfert S, Morant M, Hehn A, Meyer D, Ullmann P, Werck-Reichhart D: CYP98A3 from Arabidopsis thaliana is a 3(-hydroxylase of phenolic esters, a missing link in the phenylpropanoid pathway. J Biol Chem 2001, 276:36566-36574.

www.sciencedirect.com