Genetics and biochemistry of secondary ... - Semantic Scholar

5 downloads 0 Views 223KB Size Report
synthase (TPS) family (in which exon shuf- fling could have ... epi-aristolochene synthases have shown that the exchange ..... (+)-Bornyl diphosphate synthase (. ).
trends in plant science Perspectives

Genetics and biochemistry of secondary metabolites in plants: an evolutionary perspective Eran Pichersky and David R. Gang The evolution of new genes to make novel secondary compounds in plants is an ongoing process and might account for most of the differences in gene function among plant genomes. Although there are many substrates and products in plant secondary metabolism, there are only a few types of reactions. Repeated evolution is a special form of convergent evolution in which new enzymes with the same function evolve independently in separate plant lineages from a shared pool of related enzymes with similar but not identical functions. This appears to be common in secondary metabolism and might confound the assignment of gene function based on sequence information alone. lants produce an amazing diversity of low molecular weight compounds. Although the structures of close to 50 000 have already been elucidated1, there are probably hundreds of thousands of such compounds. Only a few of these are part of ‘primary’ metabolic pathways (those common to all organisms). The rest are termed ‘secondary’ metabolites; this term is historical and was initially associated with inessentiality but, here, a ‘secondary’ metabolite is defined as a compound whose biosynthesis is restricted to selected plant groups. The ability to synthesize secondary compounds has been selected throughout the course of evolution in different plant lineages when such compounds addressed specific needs (Fig. 1). For example, floral scent volatiles and pigments have evolved to attract insect pollinators and thus enhance fertilization rates2,3. The ability to synthesize toxic chemicals has evolved to ward off pathogens and herbivores (from bacteria and fungi to insects and mammals) or to suppress the growth of neighboring plants4–7. Chemicals found in fruits prevent spoilage and act as signals (in the form of color, aroma and flavor) of the presence of potential rewards (sugars, vitamins and amino acids) for animals who eat the fruit and thereby help to disperse the seeds. Other chemicals serve cellular functions that are unique to the particular plant in which they occur (e.g. resistance to salt or drought8,9). The chemical solutions to a common problem are often different in different plant lineages. For example, the compounds that make up floral scents vary widely from species to species, even when the same class of pollinators (e.g. moths) are attracted to the differing bouquets10. The variety of herbivore-deterring

P

chemicals produced by plants also seems to be vast, and individual plant lineages synthesize only a small subset of such compounds11. Each species contains only a subset of genes for secondary metabolism

Although the pathways that produce most secondary compounds have not yet been elucidated, it is clear that there are possibly hundreds of thousands of different enzymes involved in secondary metabolism in plants. There are many known instances in secondary metabolism in which the synthesis of multiple products can be catalyzed by a single enzyme, either from different substrates12,13 or, more rarely, even from the same substrate14. However, in most cases that have been investigated, the enzymes in plant secondary metabolism are specific for a given substrate and produce a single product. Plant genomes are variously estimated to contain 20 000–60 000 genes, and perhaps 15–25% of these genes encode enzymes for secondary metabolism15,16. Clearly, the genome of a given plant species encodes only a small fraction of all the enzymes that would be required to synthesize the entire set of secondary metabolites found throughout the plant kingdom. This article focuses on the molecular evolutionary mechanisms that are responsible for generating the great diversity of plant secondary metabolites. Gene duplication is not the only mechanism of evolution of new genes in secondary metabolism

It is believed that, at least in primary metabolism, new genes almost always arise by gene duplication followed by divergence17,18. This leaves the organism with one gene that

1360 - 1385/00/$ – see front matter © 2000 Elsevier Science Ltd. All rights reserved. PII: S1360-1385(00)01741-6

maintains the original function and a second copy that is not restricted by natural selection. This second copy can then accumulate mutations until, rarely, it has acquired a new function and might then become fixed in the population. Domain swapping, with or without prior gene duplications, can also create new, composite genes19. How often do genes for secondary metabolism arise by gene duplication and divergence, and how often do they arise by simple allelic divergence? To resolve this issue, comparative analyses of orthologous loci from related species that also include the identification of gene function must be carried out – but such data are not yet available. Obviously, if the original gene had an essential function, as genes of primary metabolism would be expected to have, gene duplication is a necessary prerequisite. However, it is theoretically possible, for example, for a new allele in one of the plant’s genetic loci to be selected for if it encodes the ability to make a new defense compound, whereas the older alleles specify the synthesis of another defense compound that is no longer effective at deterring the plant’s enemies. Thus, in secondary metabolism, there is a potential for new genes to evolve without a prior gene duplication event. In such cases, orthologous genes in related species might encode proteins with different functions. Origin of new genes for secondary metabolism

A gene can be defined as new and distinct from its ancestral gene when: (1) it encodes an enzyme that catalyzes a chemically similar reaction but on a different substrate than the enzyme encoded by its progenitor gene; or (2) the encoded enzyme carries out a different chemical reaction on the same substrate. A single-step change in both the substrate and the type of reaction is much less likely. How often do new genes of secondary metabolism arise from other genes of secondary metabolism, and how often do they arise from genes of primary metabolism? The recent advances in whole-genome sequencing EST databases have provided important information for this question, but no definitive answers. In general, the order of origin of different genes in primary metabolism can be inferred from their level of relatedness to each other (i.e. their level of sequence identity). Current sequencing projects are uncovering many gene ‘families’ whose existence and extent was only suspected before (Table 1). These families are defined by their shared ‘motifs’ in the encoded proteins (which might constitute the active site and/or binding domains of substrates and co-factors). However, because the true functions of most members of plant gene families are not yet October 2000, Vol. 5, No. 10

439

trends in plant science Perspectives

(a)

OH HO

(b)

O

OH O

O

OH

HO

O

OH

O

O

O

O

O

OH OH O

OH

HO

H

H OMe OMe

OH Rutin

Rotenone

(c)

(d) O

OH

O

N OMe OMe

Linalool

(e)

MeO

O

Berberine

OH

(f)

Gain and loss of genes for specific secondary compounds are continuing processes N S

N

O

OH DIMBOA

N H Brassilexin

Fig. 1. Examples of plant secondary metabolites and their proposed function in the plant from which they were isolated. (a) Rutin, obtained from Forsythia intermedia, thought to act as a visual pollinator attractant. (b) Rotenone, obtained from Derris elliptica, thought to act as an insect feeding deterrent. (c) Linalool, obtained from Clarkia breweri, thought to act as an olfactory pollinator attractant. (d) Berberine, obtained from Berberis wilsoniae, thought to act as a defense toxin. (e) DIMBOA, obtained from Zea mays, thought to act as a defense toxin. (f) Brassilexin, obtained from Brassica spp., thought to act as an antifungal toxin.

known, it remains difficult to answer the questions posed above completely. Thus, the large plant gene family of cytochrome-p450s-dependent oxygenases contains only a few members currently recognized to be involved in primary metabolism, such as in steroid and phenylpropanoid biosynthesis20. This large gene family also contains members already identified as being involved in secondary metabolism (e.g. the formation of menthol and carvone 21). A similar situation also exists in the family of genes encoding O-methyltransferase enzymes, which are involved in primary metabolism (e.g. lignin formation) as well as in secondary metabolism (e.g. phenylpropene and alkaloid biosynthesis1,22). Another example is a family of secondary 440

October 2000, Vol. 5, No. 10

possible. For example, some of the acylated anthocyanin derivatives that are synthesized by enzymes belonging to the aforementioned acyltransferase family might be synthesized by all plants, at least under some specific but presently unknown conditions, but this has not yet been ascertained. Even if they are not uniformly found in all plants, the ability to make such compounds might be an ancestral trait that has been lost in various plant lineages; this means that, at one point in time, these compounds were primary metabolites (produced by all plant groups). Thus, although comparisons of available sequence information indicate that many genes of secondary metabolism have evolved directly from other genes known or presumed to be involved in secondary metabolism, it is reasonable to assume that, in most cases, the ultimate (and sometimes the proximate) ancestor was a gene involved in primary metabolism. Indeed, genes of primary metabolism can serve as a pool from which similar genes of secondary metabolism could evolve over and over again.

metabolism glycosyl transferases that also contains the gene for carboxypeptidase of primary metabolism23. By contrast, several large families of genes have recently been identified that contain only a few members with a defined function, all involved in secondary metabolism. For example, a large gene family, with an estimated 70 members in Arabidopsis alone, encodes enzymes for acyl transferases involved in the synthesis of various scent, pigment and defense compounds24. Some members of this gene family might be involved in primary metabolism but none has yet been identified. The distinction between primary and secondary metabolism is difficult to make with our present knowledge and, in some cases, such a distinction is simply not

There are many examples of a specific secondary compound that is restricted to one plant lineage and is not found in related lineages, especially the ancestral one (such an observation should always be considered provisional because it is of course possible that other lineages will later be found to make such a compound). This represents prima facie evidence that the ability to synthesize this compound arose within this lineage. Molecular evidence for the origin of a new gene encoding the enzyme that catalyzes the formation of this compound requires analysis of the presence of the gene in this and related plant lineages, as well as a comparison of its sequence similarity to other related genes. For example, the gene from Clarkia breweri (family Onagraceae) that encodes the enzyme IEMT [which catalyzes the methylation of (iso)eugenol to give (iso)methyleugenol and is involved in floral scent biosynthesis] has been shown to have arisen from the gene encoding the enzyme COMT (which methylates caffeic acid to give ferulic acid and is involved in lignin biosynthesis) some time after the divergence of the order Myrtales22 (Fig. 2). However, such data are rare. The number of changes in the primary sequence of an enzyme that are required to alter its substrate specificity or its mode of action can vary. Sequence comparisons of related extant enzymes do not address this issue directly because enzymes accumulate neutral changes over time, making the

trends in plant science Perspectives amino acid substitutions critical for change of function difficult to identify. This is especially true because, in most cases, the active site and binding site of the enzyme, as well as other functionally important domains, are not well defined22. Nonetheless, examples are known in which pairs of enzymes with different substrates differ at one or a few positions25. In addition, in vitro mutagenesis experiments have shown that the substrate preference of O-methyltransferases and the type of reaction catalyzed by a fatty acid desaturase can be changed by as few as 5–7 amino acid substitutions (turning the fatty acid desaturase into a hydroxylase)22,26. Finally, in the terpene synthase (TPS) family (in which exon shuffling could have been involved in the evolution of some members27), domainswapping experiments with sesquiterpene epi-aristolochene synthases have shown that the exchange of a single small segment could result in new substrate preference or in different products being made from the original substrate28. There are currently not enough data to calculate how frequently such changes have resulted in new enzymes of secondary metabolism. However, several factors seem to facilitate this process. The new substrate (new for the newly evolved enzyme, but not necessarily new to the plant) often closely resembles the old substrate, so that one or a few amino acid substitutions can allow the altered enzyme to recognize the new substrate (while maintaining the same catalytic domain). Sometimes, the enzyme recognizes only a small part of the substrate to begin with, although as long as that part of the molecule is similar between the old and the new substrate, a small change in the substrate-binding site of the enzyme is sufficient. It should be remembered that many enzymes of secondary metabolism can already recognize more than one substrate, although they often have different catalytic rates toward them29. In addition, the existence of large families of enzymes, which are themselves the product of repeated cycles of gene duplications and divergence, increases the probability that a small change in one or another member of the family will result in an enzyme that can carry out the same type of reaction on a new substrate, or carry out a different reaction on an old substrate. This ‘snowball effect’ (the more genes there are in the family, the faster new members arise) can probably explain in part the large size of the O-methyltransferase, terpene synthase, cytochrome p450 and dehydrogenase–reductase gene families (Table 1), to name just a few. Furthermore, because the new product would not be essential for the survival of

Table 1. Selected plant gene families with at least some members that are involved in plant secondary metabolism Enzyme gene family

Example from secondary metabolisma

No. of copies in Arabidopsis

2-Oxoglutarate-dependent dioxygenases

Flavone synthase

.10

Acyl transferases

Acetyl-CoA:benzylalcohol acetyl transferase

.70

Carboxymethyl methyltransferases

S-adenosylmethionine:salicylic acid methyl transferase

.20

Cytochromes p450

DIBOA hydroxylase

.100

Glutathione-S-transferases

Petunia An9 gene

.20

Methylene bridge-forming enzymes

Berberine bridge enzyme

.10

NADPH-dependent dehydrogenases

Isoflavone reductase

.50

O-Methyl transferases

(Iso)eugenol O-methyltransferase

.20

Polyketide synthases

Stilbene synthase

.10

Terpene synthases

Linalool synthase

.20

a

Not from Arabidopsis

the plant, the recently evolved enzyme that catalyzes its formation need not initially be efficient. However, if the production of the new chemical confers a selective advantage to the plant, genetic changes will be selected for over time that favor increased synthesis. Such changes could involve additional amino acid substitutions that increase the catalytic efficiency of the enzyme. However, an alternative way to increase fitness would be to increase the expression of the gene encoding the new enzyme. Indeed, the turnover numbers of many enzymes of secondary metabolism are many orders of magnitude lower than those of enzymes of primary metabolism, even for enzymes from the same gene families. As a consequence, plants often achieve high synthesis rates of some secondary metabolites by expressing their genes at high levels in a given tissue and under given conditions (e.g. following pathogen attack), to the point that the enzymes can constitute 0.1–1.0% (or more) of the total protein in the cell30. New genes are likely to be expressed in specific tissues or cells, or at a specific time

The biosynthesis of secondary metabolites is often restricted to a particular tissue and occurs at a specific stage of development. For a new gene of secondary metabolism to provide an adaptive advantage, it therefore needs to be expressed in a specific tissue or type of cell at a specific time. As described

above, the new enzyme is likely to be a variation of an existing enzyme that uses a similar substrate and catalyzes the formation of a similar product. It is probably more likely to arise from a gene that is already spatially and temporally expressed in the same manner in which production of the new chemical is advantageous, even if the new and old substrates and new and old products are not as structurally similar as they otherwise could have been. This is because descent from an enzyme that recognizes a more similar substrate but is not expressed in the right tissue or at the right time will require mutations in both coding regions and promoter elements (i.e. in two separate parts of the gene). If, on the other hand, modification of only the coding region need occur, genes encoding new enzymes can evolve more rapidly. BEAT (acetyl-CoA–benzylalcohol acetyltransferase, which is involved in floral scent biosynthesis) and GAT4 (anthocyanin 5aromatic acyltransferase, which is involved in floral pigment biosynthesis) might be an example of this24. They are acyltransferases that share significant sequence similarity and show coincident expression in flower petal epidermal cells, although the substrates for the two enzymes differ greatly (small benzenoid versus larger glycosylated flavonoid, respectively). It is therefore not surprising that, in secondary metabolism, there is little discernable correlation between the relatedness of enzymes and the relatedness of the corresponding substrates and products. October 2000, Vol. 5, No. 10

441

trends in plant science Perspectives

subsp. trichocarpa ×

evolved. Instead, the expression pattern of existing biosynthetic genes must have changed (e.g. through an altered promoter or transcription factor). Evolution of new pathways

Myrtales

IEMT

Fig. 2. Phylogenetic tree consisting of COMT (which methylates caffeic acid to give ferulic acid and is involved in lignin biosynthesis) sequences from several species and including the Clarkia breweri IEMT sequence, showing that Clarkia IEMT evolved from Clarkia COMT after the origination of the order Myrtales. Modified from Ref. 22.

Changes in location of new enzymes

If, however, a new enzyme does arise in a cell or organelle separated from where the new reaction can impart benefits to the plant or from where the new substrate is present, there are several possible scenarios that could result in selective advantage. In one scenario, additional mutations in the control region of the gene (including the coding part that specifies subcellular location) or in other genes encoding regulatory proteins could alter the enzyme’s distribution. In a second scenario, additional changes elsewhere in the genome could result either in the de novo synthesis of the substrate in the appropriate location or in the transport of the substrate into the cellular or subcellular location of the new enzyme. The latter changes might explain the observation that biosynthetic pathways for secondary metabolism are sometimes split over more than one subcellular compartment or across two or more types of cells or even tissues (the biosynthesis of alkaloids provides examples of all of these1). Such shuttling of substrates between different cellular compartments or different cells is often cited as examples of a mechanism for the ‘regulation’ of the pathway. However, the possibility that such an arrangement is the result of the contingent nature of the evolution of new enzymes (and therefore of the new pathways) should not be ignored. Thus, although it is possible that a 442

October 2000, Vol. 5, No. 10

regulatory mechanism has evolved post facto to take advantage of the need to shuttle intermediates in the pathway, such a mechanism cannot be assumed ipso facto. Evolution of gene expression

In discussing the origin of new enzymes that catalyze the formation of new products, we should not lose sight of the fact that biochemical ‘pathways’ do not run in parallel, independently of each other, but instead are more accurately represented as interconnected networks of reactions. Although a new reaction in secondary metabolism, possibly more so than in primary metabolism, usually gives rise to an end product that is not further metabolized by the plant, the substrate on which the new enzyme acts could in principle be an intermediate in an existing pathway and not necessarily an end product by itself. Indeed, an ‘instant’ new pathway could be created when a new enzyme converts an intermediate in one pathway into an intermediate of another pathway, thus linking the two (Fig. 3). An example of this concept was recently demonstrated for plant primary metabolism when sweetgum (Liquidambar styraciflua) coniferyl aldehyde 5-hydroxylase (CAld5H) and 5-hydroxyconiferyl aldehyde O-methyltransferase (COMT isoform) were shown to convert coniferyl aldehyde to sinapyl aldehyde via 5-hydroxyconiferyl aldehyde, suggesting that a CAld5H–COMT-mediated pathway to sinapic acid might be functional in some plants31. This is in contrast with the generally accepted route to sinapic acid from ferulic acid through 5-hydroxyferulic acid. A new pathway might be formed even without the creation of any new enzyme simply by

As discussed above, changes in gene expression are often crucial, although not sufficient by themselves, for the evolution of new genes (and new pathways). (a) However, such changes can often be confused with the origin of a new A B D E C gene. For example, C. breweri synthesizes linalool in its petals whereas New enzyme its close relative Clarkia concinna V W X Y Z does not, even though C. concinna possesses the same enzyme respon(b) sible for linalool synthesis, linalool synthase (LIS), as C. breweri does. A B D E However, in C. concinna, LIS is found only in the stigma and at a C much lower level of expression than V W Y Z in C. breweri2,27. Thus, if a plant species is found to synthesize a secondary compound in a particular Fig. 3. Two methods by which new biochemical organ and its relatives do not synthepathways can originate: (a) through the formation size this compound in that same of a new enzyme that links two pre-existing pathorgan, it is important to verify ways; (b) through co-expression in the same comwhether these relatives produce partment of selected enzymes from two pathways such a compound elsewhere in the that share the same intermediate. plant. If they do, this probably means that no new biosynthetic genes have

trends in plant science Perspectives expressing in the same compartment two sets (or partial sets) of enzymes belonging to two different pathways that share at least one intermediate (Fig. 3) but that have not previously operated in the same compartment. As this discussion shows, as long as we continue to think of pathways as linear arrays of reactions, it is difficult for us to determine the sequence of events that gave rise to them. Only a detailed examination of the presence or absence of particular reactions in related plant species whose true phylogeny is known can allow us to determine which reactions came first (based on parsimony analysis). Such analyses are yet to be attempted. Nevertheless, new pathways can also be created by repeated cycles of gene duplication and divergence. Two examples illustrate this point. First, in the pathway leading to the synthesis of the defense compound DIMBOA, four sequential hydroxylation reactions are carried out by similar cytochrome-p450dependent mono-oxygenases32. For such a pathway to evolve over time, the intermediates must have conferred some selective advantage by themselves. Alternatively, the original mono-oxygenase of the DIMBOA pathway might initially have been able to catalyze all (or some) of the four hydroxylation reactions, with the high substrate specificities observed32 in the current enzymes evolving later. A second example is found in the flavonoid biosynthetic pathway, in which two homologous enzymes (flavone synthase and anthocyanidin synthase) are core enzymes in two distinct pathways that diverge from the product of flavanone 3-hydroxylase. All three of these enzymes are 2-oxoglutarate-dependent dioxygenases and share high sequence similarity33. Anthocyanidin synthase is in the pathway leading to the anthocyanins, whereas flavone synthase is a branch-point enzyme, shunting flux to the formation of flavonol glucosides instead. Thus, by multiple duplication events and subsequent divergence of the flavanone 3-hydroxylase or its precursor gene, multiple new enzymes and two pathways evolved.

related to oxidoreductases, some being related to carboxypeptidase and others being of unknown provenance34. Even more intriguing is the observation that many examples of convergent evolution in plant secondary metabolism are of a special case, termed repeated evolution, in which a new genetic function arises independently but from orthologous or paralogous genes27,29. For example, the repeated evolution of the enzyme homospermidine synthase, which catalyzes the committed step in the synthesis of pyrrolizidine alkaloids, from the ubiquitous eukaryotic enzyme deoxyhypusine synthase (which catalyzes the first step in the activation of a translation initiation factor) has been invoked to explain the sporadic occurrence of the pyrrolizidine alkaloids throughout the angiosperms35. In another example, the cyanidin-3-glucoside–gluthathione-S-transferases from maize and petunia each arose independently from paralogous members of the gluthathione S-transferase (GST) family36. Similarly, limonene synthases in both gymnosperms and angiosperms are each more similar to other terpene synthases within their lineages than to each other (Fig. 4), but the terpene synthases from gymnosperms and angiosperms are also related to each other37. This indicates that specific limonene synthase enzymatic activities evolved in plants more than once but, in all cases, they evolved from a member of the terpene synthase family. It appears that the universal presence of several related enzymes in each of the GST

and TPS families that catalyze reactions in primary metabolism or in related branches of secondary metabolism might provide a pool from which new enzymes can evolve, sometimes more than once. An important consequence of repeated evolution is that the catalytic function of a newly described gene or protein cannot be assigned solely on its degree of sequence identity to known enzymes27,37. It is currently a common practice (as can be seen by perusing the sequence databases) to assign function to newly obtained sequences based solely on homology to other sequences in the database, and, less often, on the determination of expression level and cell-type location. These approaches carry a high risk of misidentification of the true role of enzymes involved in plant secondary metabolism. Thus, it was recently shown biochemically that a methyltransferase gene from Arabidopsis that had originally been thought to encode COMT, based on its sequence similarity to other known COMTs, actually encodes an enzyme that methylates quercetin, a flavonol, and is not active with caffeic acid38. Similarly, an Arabidopsis TPS gene initially identified as ‘limonene synthase’ based on sequence comparisons alone has now been shown to encode myrcene synthase39. Clearly, actual biochemical data demonstrating catalytic function will be required to uncover the true function. Experiments yielding biochemical data demonstrating catalytic function are not always straightforward and can be difficult to carry out in practice, especially for large

δ-Selinene synthase (

)

(–)-4 -limonene/(–)-α-pinene synthase ( (–)-4 -limonene synthase (

) )

Myrcene synthase ( Pinene synthase (

Gymnosperms

)

)

(+)-Bornyl diphosphate synthase ( )

(+)-Sabinene synthase ( Convergent and repeated evolution in secondary metabolism

One of the most remarkable observations about the evolution of secondary metabolism in plants is clearly the many cases that appear to represent convergent evolution. For example, cyanogenesis (the release of hydrogen cyanide as a defense compound) appears to have arisen several times during plant evolution. Recently, it has been shown that the enzymes that catalyze the release of HCN from cyanogenic glycosides (hydroxynitrile lyases) have arisen independently several times, with some being

)

1,8-Cineole synthase (

)

(–)-4 -limonene synthase (

)

(–)-4 -limonene synthase (

)

(–)-4 -limonene synthase (

)

Angiosperms

Fig. 4. Phylogenetic tree of terpene synthases from gymnosperms and angiosperms showing that limonene synthases evolved separately in these two plant lineages. Modified from Ref. 37.

October 2000, Vol. 5, No. 10

443

trends in plant science Perspectives numbers of genes with vast numbers of potential substrates. However, if analysis of the sequence homology is coupled to a detailed understanding of the metabolite composition for a given species (‘biochemical genomics’), these experiments do become possible. Future prospects

In future work, researchers will hopefully examine the provenance of new genes of secondary metabolites. This should be done by comparing them with the ancestral genes in related species, using established statistical methods that have been used extensively in taxonomic studies40. As detailed above, many of the questions raised here – how often new genes of secondary metabolism arise, which genes are more likely to serve as the source of new genes and what the specific changes are that occur when new genes of secondary metabolism evolve – can be better addressed when data of such comparisons are available. The increasing availability of plant genome sequences and EST data sets makes this approach feasible, but there is a need to carry out additional EST data acquisition from plant species other than the standard crop plants of the Western world, because it is such nonstandard plants that hold most of the diversity of secondary plant metabolites. Moreover, EST projects that are concerned with secondary metabolism should strive to analyze tissues that are especially active in the synthesis of such metabolites, to increase the probability of obtaining as many sequences as possible from mRNAs that are normally found in low abundance in generalized tissues. The recent report of an EST data set acquired from the peltate glands of mint (specialized tissue for the synthesis of monoterpenes) is a good example41. Finally, there is a pressing need to develop an efficient system to test en masse the catalytic activities of the enzymes whose sequences are being revealed by the ongoing and future mass-sequencing EST projects, because it is clear that, in secondary metabolism, only the demonstration of enzymatic activity can unambiguously identify the function of the protein. It is also clear that, with all the new tools of molecular biology and biochemistry being put to use to address these exciting questions, our understanding of the evolution of plant secondary metabolism is poised for major advances42. Acknowledgements

Research in our laboratory was funded by NSF grant MCB-9974436 and by a Margaret and Herman Sokol Post-doctoral Fellowship in the Sciences to D.R.G. We thank Jonathan Gershenzon, Leslie D. Gottlieb and three anonymous reviewers for their useful comments on the manuscript. 444

October 2000, Vol. 5, No. 10

References 1 De Luca, V. and St Pierre, B. (2000) The cell and developmental biology of alkaloid biosynthesis. Trends Plant Sci. 5, 168–173 2 Dudareva, N. and Pichersky, E. (2000) Biochemical and molecular genetic aspects of floral scent. Plant Physiol. 122, 627–633 3 Mol, J. et al. (1998) How genes paint flowers and seeds. Trends Plant Sci. 3, 212–217 4 Bennet, R.N. and Wallsgrove, R.M. (1994) Secondary metabolism in plant defense – mechanisms. New Phytol. 127, 617–633 5 Harborne, J.B. (1999) The comparative biochemistry of phytoalexin induction in plants. Biochem. Syst. Ecol. 27, 335–367 6 Dixon, R.A. et al. (1996) Metabolic engineering: prospects for crop improvement through the genetic manipulation of phenylpropanoid biosynthesis and defense responses – a review. Gene 179, 61–71 7 Mitchel-Olds, T. et al. (1998) Chemical ecology in the molecular era. Trends Plant Sci. 3, 362–365 8 Trossat, C. et al. (1998) Salinity promotes accumulation of 3-dimethylsulfoniopropionate and its precursor S-methylmethionine in chloroplasts. Plant Physiol. 116, 165–171 9 Nuccio, M.L. et al. (1999) Metabolic engineering of plants for osmotic stress resistance. Curr. Opin. Plant Biol. 2, 128–134 10 Knudsen, J.T. and Tollsten, L. (1993) Trends in floral scent chemistry in pollination syndromes: floral scent composition in moth-pollinated taxa. Bot. J. Linn. Soc. 113, 263–284 11 Ananthakrishnan, T.N. (1999) Induced responses, signal diversity and plant defense: implications in insect phytophagy. Curr. Sci. 76, 285–290 12 Allina, S.M. et al. (1998) 4-Coumarate:coenzyme A ligase in hybrid poplar. Properties of native enzymes, cDNA cloning and analysis of recombinant enzymes. Plant Physiol. 116, 743–754 13 Maury, S. et al. (1999) Tobacco O-methyltransferases involved in phenylpropanoid metabolism. The different caffeoyl-coenzyme A/5-hydroxyferuloylcoenzyme A 3/5-O-methyltransferase and caffeic acid/5-hydroxyferulic acid 3/5-O-methyltransferase classes have distinct substrate specificities and expression patterns. Plant Physiol. 121, 215–223 14 Phillips, M.A. et al. (1999) Monoterpene synthases of loblolly pine (Pinus taeda) produce pinene isomers and enantiomers. Arch. Biochem. Biophys. 372, 197–204 15 Bevan, M. et al. (1998) Analysis of 1.9 Mb of contiguous sequence from chromosome 4 of Arabidopsis thaliana. Nature 391, 485–488 16 Somerville, C. and Somerville, S. (1999) Plant functional genomics. Science 285, 380–383 17 Ohno, S. (1970) Evolution by Gene Duplication, Springer-Verlag

18 Pichersky, E. (1990) Nomad DNA – a model for movement and duplication of DNA sequences in plant genomes. Plant Mol. Biol. 15, 437–448 19 Doolittle, R.F. (1995) The multiplicity of domains in proteins. Annu. Rev. Biochem. 64, 287–314 20 Szekeres, M. and Koncz, C. (1998) Biochemical and genetic analysis of brassinosteroid metabolism and function in Arabidopsis. Plant Physiol. Biochem. 36, 145–155 21 Lupien, S. et al. (1999) Regiospecific cytochrome p450 limonene hydroxylases from mint (Mentha) species: cDNA isolation, characterization, and functional expression of (2)-4S-limonene-3hydroxylase and (2)-S-limonene-6-hydroxylase. Arch. Biochem. Biophys. 368, 181 22 Wang, J. and Pichersky, E. (1999) Identification of specific residues involved in substrate discrimination in two plant O-methyltransferases. Arch. Biochem. Biophys. 368, 172–180 23 Li, A.X. and Steffens, J.C. (2000) An acyltransferase catalyzing the formation of diacylglucose is a serine carboxypeptidase-like protein. Proc. Natl. Acad. Sci. U. S. A. 97, 6902–6907 24 St Pierre, B. and De Luca, V. Evolution of acyltransferase genes. In Recent Advances in Phytochemistry – Evolution of Metabolic Pathways (Vol. 34), Plenum Press (in press) 25 Frickk, S. and Kutchan, T.M. (1999) Molecular cloning and functional expression of O-methyltransferases common to isoquinoline alkaloid and phenylpropanoid biosynthesis. Plant J. 17, 329–333 26 Broun, P. et al. (1998) Catalytic plasticity of fatty acid modification enzymes underlying chemical diversity of plant lipids. Science 282, 1315–1317 27 Cseke, L. et al. (1998) Structure and evolution of linalool synthase. Mol. Biol. Evol. 15, 1491–1498 28 Back, K. and Chappell, J. (1996) Identifying functional domains within terpene cyclases using a domain-swapping strategy. Proc. Natl. Acad. Sci. U. S. A. 93, 6841–6845 29 Wang, J. and Pichersky, E. (1998) Characterization of S-adenosyl-Lmethionine:(iso)eugenol O-methyltransferase involved in scent production in Clarkia breweri. Arch. Biochem. Biophys. 349, 153–160 30 White, W.L.B. et al. (1998) Cyanogenesis in Cassava: the role of hydroxynitrile lyases in root cyanide production. Plant Physiol. 116, 1219–1225 31 Osakabe, K. et al. (1999) Coniferyl aldehyde 5-hydroxylation and methylation direct syringyl lignin biosynthesis in angiosperms. Proc. Natl. Acad. Sci. U. S. A. 96, 8955–8960 32 Glawischnig, E. et al. (1999) Cytochrome p450 monooxygenases of DIBOA biosynthesis: specificity and conservation among grasses. Phytochemistry 50, 925–930 33 Rosati, C. et al. (1999) Molecular characterization of the anthocyanidin synthase gene in Forsythia 3 intermedia reveals organspecific expression during flower development. Plant Sci. 149, 73–79

trends in plant science Perspectives 34 Hickel, A. et al. (1996) Hydroxylnitrile lyases: functions and properties. Physiol. Plant 98, 891–898 35 Ober, D. and Hartmann, T. (1999) Homospermidine synthase, the first pathway-specific enzyme of pyrrolizidine alkaloid biosynthesis, evolved from deoxyhypusine synthase. Proc. Natl. Acad. Sci. U. S. A. 96, 14777–14782 36 Alfenito, M.R. et al. (1998) Functional complementation of anthocyanin sequestration in the vacuole by widely divergent glutathione S-transferases. Plant Cell 10, 1135–1149 37 Bohlmann, J. et al. (1998) Plant terpenoid synthases: molecular and phylogenetic analysis. Proc. Natl. Acad. Sci. U. S. A. 95, 4126–4133

38 Muzac, I. et al. (2000) Functional expression of an Arabidopsis cDNA clone encoding a flavonol 39-O-methyltransferase and characterization of the gene product. Arch. Biochem. Biophys. 375, 385–388 39 Bohlmann, J. et al. (2000) Terpenoid secondary metabolism in Arabidopsis thaliana: cDNA cloning, characterization, and functional expression of a myrcene/(E)-b-ocimene synthase. Arch. Biochem. Biophys. 375, 261–269 40 Rogers, J.S. and Swofford, D.L. (1998) A fast method for approximating maximum likelihoods of phylogenetic trees from nucleotide sequences. Syst. Biol. 47, 77–89 41 Lange, B.M. et al. (2000) Probing essential oil biosynthesis and secretion by functional

Trends in Plant Science Perspectives

evaluation of expressed sequence tags from mint glandular trichomes. Proc. Natl. Acad. Sci. U. S. A. 97, 2934–2939 42 Facchini, P.J. (1999) Plant secondary metabolism: out of the evolutionary abyss. Trends Plant Sci. 4, 382–384

Eran Pichersky* and David R. Gang are at the Biology Dept, University of Michigan, Ann Arbor, MI 48109-1048, USA. *Author for correspondence (tel 11 734 936 3522; fax 11 734 647 0884; e-mail [email protected]).

Trends in Plant Science online — making the most of your personal subscription

Perspectives articles are your opportunity to express



High quality printouts (from PDF files)

personal views on topics of interest to plant scientists.



Links to other articles, other journals and cited

The range of appropriate subject matter covers the breadth of plant biology: historical, philosophical or political aspects; potential links between hitherto

software and databases All you have to do is: •

obtain your subscription key from the address label of your print subscription

disparate fields; or simply a new view on a subject of current importance. Specific issues might include: the



then go to http://www.trends.com/free_access.html



Click on the large ‘Click Here’ button at the bottom of the page

allocation of funding; commercial exploitation of data; the gene sequencing projects; genetic engineering; the

You will see one of the following: (1) A BioMedNet login screen. If you see this, please

impact of the internet; education issues; or public

enter your BioMedNet username and password. If you

perception of science. Current developments in other

are not already a member please click on the ‘Join Now’

disciplines may also be relevant.

button and register. Once registered you will go straight to (2) below. (2) A box to enter a subscription key. Please enter your

Although Perspective articles are more opinionated in

subscription key here and click on the ‘Enter’ button.

tone, they remain subject to stringent peer-review to ensure clarity and accuracy. The decision to publish rests



Once confirmed, go to http://plants.trends.com and view the full-text of Trends in Plant Science

with the Editor. If you wish to write for the Perspectives section, please contact the Editor before embarking on your manuscript, with a 200-word proposal and a list of key recent references ([email protected]).

If you get an error page please contact Customer Services ([email protected]) stating your subscription key and BioMedNet username and password. Please note that you do not need to re-enter your subscription key for Trends in Plant Science, BioMedNet ‘remembers’ your subscription. Institutional online access is available at a premium. If your institute is interested in subscribing to print and online please ask them to contact [email protected]

October 2000, Vol. 5, No. 10

445