Engineering the plant cell factory for secondary ... - Springer Link

Transgenic Research 9: 323–343, 2000. © 2000 Kluwer Academic Publishers. Printed in the Netherlands.

323

Engineering the plant cell factory for secondary metabolite production R. Verpoorte1 , R. van der Heijden1 & J. Memelink2 1 Division of Pharmacognosy, Leiden/Amsterdam Center for Drug Research, Leiden University,

PO Box 9502, 2300

RA Leiden, The Netherlands (E-mail: [email protected]) 2 Institute of Molecular Plant Sciences, Leiden University, Leiden, The Netherlands

Key words: Plant secondary metabolism, plant cell factory, metabolic engineering

Abstract Plant secondary metabolism is very important for traits such as flower color, flavor of food, and resistance against pests and diseases. Moreover, it is the source of many fine chemicals such as drugs, dyes, flavors, and fragrances. It is thus of interest to be able to engineer the secondary metabolite production of the plant cell factory, e.g. to produce more of a fine chemical, to produce less of a toxic compound, or even to make new compounds, Engineering of plant secondary metabolism is feasible nowadays, but it requires knowledge of the biosynthetic pathways involved. To increase secondary metabolite production different strategies can be followed, such as overcoming rate limiting steps, reducing flux through competitive pathways, reducing catabolism and overexpression of regulatory genes. For this purpose genes of plant origin can be overexpressed, but also microbial genes have been used successfully. Overexpression of plant genes in microorganisms is another approach, which might be of interest for bioconversion of readily available precursors into valuable fine chemicals. Several examples will be given to illustrate these various approaches. The constraints of metabolic engineering of the plant cell factory will also be discussed. Our limited knowledge of secondary metabolite pathways and the genes involved is one of the main bottlenecks.

Introduction Plants produce a wide variety of so called secondary metabolites. These compounds play a role in the survival of the plant in its ecosystem. Secondary metabolites are thus involved in resistance against pests and diseases, attraction of pollinators, interaction with symbiotic microorganisms, etc. (Harborne, 1978). Although about 100,000 plant secondary metabolites are already known, only a small percentage of all plants species have been studied to some extent for the presence of secondary metabolites. In most cases such studies are also limited to one or only a few classes of secondary metabolites. Based on the database NAPRALERT, it is estimated that about 15% of the ca. 250,000 known plant species has been subject for some sort of phytochemical study, whereas less than 5% have been screened for one or more biological activities (N.R. Farnsworth personal communication, Verpoorte et al. 1998, 1999; Verpoorte, 2000).

Besides the importance for the plant itself, secondary metabolites also are of interest because they determine the quality of food (color, taste, aroma) and ornamental plants (flower color, smell). More recently, various health improving effects and disease preventing activities of secondary metabolites have come forward, such as antioxidative and cholesterollowering properties. Besides these aspects, a number of secondary metabolites are isolated from the plant and are commercially available as fine chemicals, for example, drugs, dyes, flavors, fragrances, and insecticides. Some of these phytochemicals are quite expensive because of the low abundance in the plant. The evolving commercial importance of secondary metabolites, has in recent years resulted in a great interest in secondary metabolism, and particularly in the possibilities to alter the production of secondary metabolites by means of genetic engineering. However, so far the progress in this field is limited. In most cases very little is known about the biosynthesis of these compounds, and often only theoretical con-

324 siderations exist about possible biosynthetic routes. Only for flavonoids and anthocyanins, almost the complete pathway has been mapped at the level of the products, enzymes and genes. For other secondary metabolite pathways, at the best only a few of the enzymes have been identified and their genes cloned. Only few plants have been studied in detail for several different secondary metabolite pathways. Examples are tobacco (anthocyanins/flavonoids, terpenoids, alkaloids), Catharanthus roseus (alkaloids, steroids, brassinolides, flavonoids, and 2,3-dihydroxybenzoic acid), Cinchona (anthraquinones and alkaloids). Here we will discuss various aspects of metabolic engineering of plant secondary metabolism. We will not try to give a complete summary of all results reported, rather we would like to show the perspectives and problems of the different approaches used.

Plant cell cultures For the production of phytochemicals the use of large scale plant cell cultures in bioreactors has been extensively studied (for a review see Verpoorte et al., 1991; 1998; 1999). Only two secondary metabolites have reached a commercial production, shikonin and paclitaxel. From all the research it becomes clear that the technology is feasible. Although some plant cell cultures might be somewhat sensitive for shear forces, most cultures can be grown in large bioreactors without difficulty. A bioreactor size of 60 m3 has been reported for the growth of plant cell cultures (Westphal, 1990), recently also for Taxus species producing the antitumor compound paclitaxel. The major constraints are the slow growth and low productivity of plant cell cultures. Price calculations have been reported in various studies (Drapeau et al., 1987; Goldstein, 1980; 1999). Our calculations with Catharanthus roseus cell cultures producing ajmalicine, gave a price of about US $ 1500 per kg for a compound that is produced at a level of 0.3 g/l per 2 weeks (van Gulik et al., 1988; Verpoorte et al., 1991). A 10-fold increased productivity would result in a price of US $ 430 per kg. Such productivities are possible in plant cell culture, however, for the phytochemicals of interest such productivities are not obtained. Compounds such as morphine, codeine, hyoscyamine, scopolamine, vinblastine and vincristine are not produced at all in cell cultures. In recent years this has resulted in much interest for metabolic engineering, aiming at increasing productivity to levels that makes

plant cell cultures in bioreactors competitive with production of the plants in the field, or further increasing the productivity in plants. Besides the industrial interest, plant cell cultures are a very useful model system for studying biosynthesis of plant secondary metabolites. Cell cultures are suitable systems for feeding experiments, and easy to manipulate for studying the regulation of pathways, for example, by treatment with various effectors of signal transduction pathways.

Aims for metabolic engineering Metabolic engineering has been quite successful for the production of pharmaceuticals in microorganisms (Chartrain et al., 2000), for example, for the increased production of known compounds or for the production of new compounds (recombinatorial biochemistry) by combining different polyketide antibiotic genes (Madduri et al., 1998; Salas & Mendez, 1998). Presently the introduction of new genes into plants has become more or less routine. The particle gun (Leech et al., 2000) and the Agrobacterium tumefaciens-mediated transformation (Hooykaas, 2000) are the most widely used methods. This opens the way for genetic engineering for altering various traits of plants. This can be increased resistance against pests or diseases, improved yields, and improving quality traits of plants (Table 1). Here we will restrict ourselves to secondary metabolism in connection with these goals. Improving resistance against pests or diseases also leads to improved yields. For achieving improved resistance against microorganisms, one can first of all aim at expressing higher levels of the endogenous defense compounds such as phytoalexins and phytoanticipins, but also the production of new compounds for the plant can be considered. Similarly for pest resistance one can improve the production of the endogenous defense compounds, or introduce novel compounds into a plant. Such compounds can be either toxic or repellant for the predator, or they may be attractants for predators of insects attacking the plant. For phytochemicals extracted from the plant, major aims are increasing their levels, or reducing the number of related compounds, thus facilitating the purification of the compound of interest. Examples of improved quality traits, which could be targets for metabolic engineering are new colors for flowers or fruits; improved taste or flavor of food; fragrance

325 Table 1. Goals for genetic engineering of plant secondary metabolism Goal

Type of compounds

Risks

Examples

Improved resistance against insects

Endogenous antifeedants, insecticides

Altered secondary metabolite profile requires studies of toxicity; metabolites; viability plant

Tryptophan decarboxylase

Improved resistance against microorganisms

Endogenous antimicrobial active compounds

Altered secondary metabolite profile requires studies of toxicity; metabolites; viability plant

Salicylate

Lowering level undesired compounds in food or fodder

Toxic compounds, bad taste or flavor

Plants become less resistant against pests or diseases

Glucosinolates in canola

Increasing level of desired compounds in food

E.g. vitamins, antioxidants or other health promoting compounds

Altered secondary metabolite profile requires studies of toxicity; viability plant

Flavonoids in tomato; Vitamin A in rice

Increasing level of desired phytochemical in plant or plant cell

Drugs, flavors, fragrances, dyes, insecticides

Viability plant or plant cells

Scopolamine production in Atropa belladonna

Introduction of new compounds in the plant to increase resistance

Antifeedants, antimicrobials

Altered secondary metabolite profile requires studies of toxicity; viability plant

Resveratrol; Tryptophan decarboxylase

Introduction of new compounds in a plant for production known or new phytochemical

Drugs, flavors, fragrances, dyes, insecticides

Viability plant or plant cells

Ajmalicine in Weigela

Introduction of new compound in a plant for obtaining new trait

Color of flowers or fruits, new flavor or fragrance in food

Altered secondary metabolite profile requires studies of toxicity; viability plant

Petunia with new flower colors

of flowers; lowering levels of undesired (toxic) compounds in food and fodder (e.g. glycoalkaloids in potato, pyrrolizidine alkaloids or glucosinolates in some fodder plants, caffeine in coffee beans); increasing level of known or adding new health promoting/disease preventing compounds in food and fodder (e.g. antioxidants). The possibilities are numerous, but for food and fodder there are certain risks connected with the potential toxicity of the higher levels of secondary metabolites or new metabolites. To assess this risk extensive metabolic profiling of the transgenic food or fodder plants is crucial. However, so far no universal methods exist for making the complete profile of secondary metabolites in a plant. Chromatographic methods such as gas chromatography and high performance liquid chromatography suffer from a too small window of

separation for compounds with very different physical properties. 1 HNMR spectrometry, that is able to detect all compounds with the same sensitivity, suffers from the fact that minor compounds will not be observed. Thus for the analysis of the ‘metabolome’, the total of all low molecular weight compounds present in an organism, it will be necessary to use a set of different methods, each covering another group of compounds. In this field there is a major challenge for phytochemists.

Mapping pathways In order to achieve the above mentioned objectives we need to know the plant secondary metabolite pathways, as well as the role of the secondary metabolites

326 for the plant. Although, much information can be found on secondary metabolite pathways in plants, most of this work is based on feeding experiments using radioactively labeled intermediates, and incorporation in the final products was used as proof for the intermediacy of the fed precursors in the pathway. Problems with this approach is first of all that many intermediates are difficult to synthesize, which leaves many intermediates still unstudied. Secondly the incorporation of a radioactive label may also go through other pathways, for example, after breakdown of the intermediate. Therefore the use of stable isotopes, such as 13 Carbon, of which the site of incorporation in the product can be determined through 13 CNMR spectrometry, is a major step forward in unraveling secondary metabolite pathways. This is probably best illustrated by the recently discovered new terpenoid pathway that does not include mevalonate but deoxyxylulose as an intermediate (2-C-methylD-erythro 4-phosphate- or briefly the MEP-pathway) (for a review see Rohmer, 1999; Lichtenthaler, 1999). Both pathways lead to the general precursors isopentenyl diphosphate (IPP) and dimethylallyl diphosphate (DMAPP), which are interconvertible (Ramos et al., 1997). By using 13 C-labeled early precursors such as glucose, the mevalonate- and the MEP-pathway products can easily be distinguished by determining the site of incorporation in the terpenoids by means of 13 CNMR. It seems that all plastidial terpenoids (e.g. monoterpenoids, diterpenoids, carotenoids) are derived from the MEP pathway, whereas cytosolic terpenoids (e.g. sesquiterpenoids, triterpenoids, steroids) are derived from the mevalonate pathway. However, exchange of the precursors IPP and/or DMAPP between the two pools has been demonstrated. The recent discovery of the involvement of the MEPpathway in the biosynthesis of terpenoid indole alkaloids (Contin et al., 1998) is just one example out of many, which shows us how careful one should be with the interpretation of feeding experiments using radioactively labeled compounds. There are numerous publications showing the incorporation of mevalonate and related precursors into terpenoid indole alkaloids, but the site of incorporation was not determined (for a review see Cordell et al., 1974). That leaves the possibility of channeling the label to the alkaloids via other pathways open. The message is thus that one should first properly establish a secondary metabolite pathway at the level of intermediates by using labeled precursors and confirming the incorporation in the right place, excluding the incorporation of a

precursor via other indirect pathways (e.g. after breakdown of the intermediate into products of primary metabolism). Another point one should take into consideration is that many compounds are not end products, but are continuously produced and catabolized (Barz, 1977). An exogenous molecule of nicotine, for example, in tobacco has a half-life of 7.5 h (Waller & Nowacki, 1978). On the other hand, products stored in specialized tissues can be quite stable (Mihaliak et al., 1992). After having established the pathway at the level of the intermediates, the next step is to establish the pathway at the level of enzymes. This is often an elaborate task, hampered by the fact that the intermediates are not available, the enzyme activities are very low, or that no active enzyme can be extracted from the plant. Using different plant species in parallel for such studies can be very useful, as sometimes the isolation proves to be much more simple from one plant than from another, as for example we experienced with the enzyme isochorismate synthase from Rubia tinctorum and Catharanthus roseus. Two years of work did not result in a useful amount of pure enzyme of the former plant (van Tegelen et al., 1999a), whereas from C. roseus sufficient pure protein could be obtained for amino acid sequencing in a few weeks. Subsequently the gene was cloned and the sequence information of the C. roseus gene was used to clone the gene from other plants (van Tegelen et al., 1999a,b). Similarly, the isolation of secologanin synthase has so far failed from C. roseus whereas from Lonicera japonica it was successful (Yamamoto et al., 2000). For those steps in biosynthetic pathways for which no intermediates are available, and thus no enzyme assay can be developed, or for which no active enzyme can be found, proteomics might give new perspectives for identification of enzymes involved in certain steps of secondary metabolism (Scheme 1). Proteomics concerns the mapping of all proteins of an organism using 2D-gel electrophoresis. By in-situ peptide digestion on the gel followed by mass spectrometry some internal amino acid sequences can be obtained (Jacobs et al., 2000). However, finding the enzymes of interest requires that one has plants, plant tissues or plant cells that can produce the desired compound under certain conditions, whereas controls do not produce these compounds. Comparing the protein profiles (proteomes) of these materials by means of 2D-gelelectrophoresis, proteins connected with the production of the compound of interest can be identi-

327

Scheme 1. Functional genomics concerns the determination of the function of each gene, it thus should lerarn us the relationship between the metabolome and the genes.

fied. Subsequently the amino acid sequences obtained by mass spectrometry can be used for cloning the genes. The final step is proving the function of the cloned genes in the biosynthetic pathway. The latter step is not as self-evident as it seems. Bioinformatics might be useful to determine in general what type of enzyme is involved (Cane, 2000). But the final proof must come from a suitable system for expressing the gene and producing a functional protein. This could be for example Escherichia coli, insect cells, or yeast. Even transforming plant cells by cocultivation with Agrobacterium tumefaciens could be a suitable approach, particularly for enzymes that are dependent on plant specific post-translational modifications, co-factors, substrates or the presence of other enzymes. However, in case of lack of the necessary intermediates, it is not possible to test the function of a protein, requiring the development of other novel methods to determine the function. Knocking out the gene by means of an antisense gene, followed by analysis of the secondary metabolite profile could be an approach. In pathway mapping one should keep in mind that certain steps might just be spontaneous chemical reactions, without the involvement of an enzyme. For example, the conversion of the isoquinoline alkaloid neopine to codeinone in the morphine bio-

synthetic pathway proceeds without the need for an enzyme (Gollwitzer et al., 1993). Our work on the introduction of the microbial salicylate biosynthetic pathway in tobacco showed that the conversion of isochorismate to salicylate also occurs chemically, though not as efficiently as by the microbial enzyme (Verberne et al., 1998). Another point to bear in mind is that certain enzymes might be capable of doing two reactions. The conversion of hyoscyamine via 6-hydroxyhyoscyamine to scopolamine by the enzyme hyoscyamine 6-hydroxylase is such an example (Hashimoto et al., 1993; 1994; Yun et al., 1992). Determining enzyme selectivity is an important aspect of pathway mapping. The enzyme specificity plays for example a role in the biosynthesis of alkaloids in Papaver somniferum. Only the R-isomer of reticuline is accepted as substrate by the cytochrome P450 enzyme that is responsible for the formation of the morphinan skeleton, whereas the S-isomer is the substrate for the berberine bridge enzyme which channels this precursor to the benzophenanthridine alkaloids (Kutchan 1995; 1998; Zenk 1991; 1995). In the biosynthesis of tropane alkaloids in Hyoscyamus niger two reductases play an important role in channeling tropinone into two different pathways. Each of the reductases produced only one stereoisomer of tropinol

328

Figure 1. Biosynthetic pathway of berberine in two different plant species.

(Hashimoto et al., 1992). Tropine is the precursor for the tropane alkaloids hyoscyamine and scopolamine, whereas pseudotropine is channeled to the polyhydroxylated calystegines. Another example is the high substrate specificity of some enzymes in Berberis stolonifera and Coptis japonica. Both plants produce berberine from tetrahydrocolumbamine, but the two steps are in a different sequence, due to the specificity of the enzymes involved (Figure 1) (Galneder et al., 1988). Such networks of related compounds are quite common. Most plants produce similar basic skeletons of secondary metabolites, for example, flavonoids and terpenoids. The subsequent decoration with various functional groups can go parallel or sequential, with high or low substrate specificity of the enzymes. In case of low substrate specificity three enzymes could accomplish all reactions as shown in Figure 2, whereas with high substrate specificity 12 enzymes are needed. For determining a metabolic engineering strategy to increase the level of compound S2,3, it is necessary to know the specificity of each step.

Figure 2. Simplified model for network of secondary metabolite biosynthesis.

329 Finally the role of a secondary metabolite for the plant is an important thing to know. Is it involved in resistance, and thus of interest for increasing levels for improving resistance. Or on the other hand it will be difficult to lower the level if it is involved in resistance? Here, collaboration with plant ecologists and phytopathologists is necessary. In this connection one has to take into account that certain compounds are constitutively expressed as a defense compound. Others are constitutively expressed but need a biochemical activation such as cyanogenic glycosides and strictosidine. From the glycoside an active defense compound is released by an abundant specific glucosidase that is separated in space from the phytoanticipin. Certain compounds, such as phytoalexins, are only produced after induction on gene level through exogenous signals, such as elicitors. In different plant species similar compounds might be produced, but under very different conditions. For example, anthraquinones derived from the chorismate pathway are constitutively produced in Rubia tinctorum roots and in Morinda citrifolia stem bark whereas in Cinchona species the biosynthesis is induced by microbial infection (Wijnsma et al. 1985; Wijnsma & Verpoorte, 1986). In general one can say that most, if not all, plants are capable of making some common basic skeletons (e.g. flavonoids or terpenoids) but differ in the further processing of these basic skeletons into the final products, the species specific secondary metabolites. The enzymes involved might have common ancestors, but are usually specifically tailored for a certain conversion, such as the many highly selective cytochrome P450 enzymes occurring in plants. In biosynthetic pathway mapping, compartmentation should also be studied, as it plays a major role in the regulation of metabolic pathways (De Luca & St-Pierre, 2000). Well-known examples are of course the tropane alkaloids that are produced in the roots of the plant and stored in the green parts. Certain steps of the biosynthesis occur outside the cell producing the basic skeleton of the molecule; for example, scopolamine is formed from hyoscyamine by an epoxidation in the pericycle cells of the roots of Hyoscyamus niger (Hashimoto et al., 1991). The cellular and subcellular compartmentation of the biosynthesis of dimeric indole alkaloids in C. roseus is another example of the complexity of a biosynthetic pathway, requiring different subcellular compartments (for a review see Verpoorte et al., 1997) as well as different cells (St-Pierre et al., 1999; De Luca & St-Pierre, 2000) (Figure 3).

Cloning genes Because of the limited knowledge of plant secondary metabolite biosynthetic pathways, the number of genes known to encode enzymes in these pathways is also very limited. So far most genes have been cloned by the classical approach of identification and purification of the enzyme, followed by cloning of the encoding gene (Scheme 1). Somerville and Somerville (1999) estimated that presently for about 1000 plant genes a function is known. With the genome of several plants becoming known in the near future, further progress in functional genomics will speed up considerably. However, for secondary metabolism a major bottleneck will remain i.e. that this is per definition species specific. Only early parts of most pathways will be common to most plants and thus homology between genes can be used for strategies to clone genes from other plants (e.g. flavonoid and terpenoid biosynthetic pathways). But the genes encoding enzymes involved in the more specific ‘decoration’ of the basic skeletons can only be studied at the level of the producing plant (see also Dixon, 1999). As already mentioned, obtaining the enzymes might be a major bottleneck in this approach. Despite this problem, the number of alternatives is limited. The use of mutants as it is common for microorganisms has several limitations. First of all to make large number of plant mutants will be a major problem, subsequently these have to be analyzed for the mutations of interest, which means one should be able to detect each of the intermediates. However, such analyses might be difficult to develop, as the intermediates are not always available as reference. For easily detectable traits, such as flower color, finding mutants is relatively straightforward. Transposon and T-DNA tagging, offering the advantage of easy cloning of the mutated gene, suffer otherwise from the same problems as chemical mutagenesis, the analysis for the desired trait is very elaborate. Besides using visible traits, the use of selectable traits is another approach, which might be useful in a screening program for metabolic mutants. This approach has for example be used to select cells that have tolerance for the toxic amino acid 4-methyltryptophan (4 mT). In C. roseus this can be due to the overproduction of the enzyme tryptophan decarboxylase (TDC), which converts 4 mT into 4-methyltryptamine, which is not toxic for the cells (Sasse et al., 1983a,b). We used this selection in combination with T-DNA activation tagging. Random integration of a T-DNA carrying a strong constitutive promoter (containing CaMV 35S

330

Figure 3. Compartmentation of alkaloid biosynthesis in Catharanthus roseus. All steps are shown as occurring in a single cells, which in fact is not the case, the last steps of the biosynthesis of the dimeric alkaloids occurs in separate cells (St-Pierre et al., 1999), but it is not yet known what intermediate is transported.

promoter elements) in the C. roseus genome, can activate genes that cause an increased activity of TDC. This can be due to activation of the tdc gene itself, but also to activation of regulatory genes that control tdc expression. In this way we were able to clone an AP2domain transcription factor that is a master regulator of genes in primary as well as secondary metabolism leading to the terpenoid indole alkaloid biosynthesis in C. roseus (van der Fits, 2000). If the secondary metabolite pathway of interest occurs in microorganisms, genes can be cloned through complementation in mutants of the microorganism lacking a step of the pathway. This approach was for example used in cloning the genes of the carotenoid pathway in plants by color complementation in E. coli (Cunningham & Gantt, 1998; DellaPenna, 1999). The yeast two-hybrid system has been used for finding enzymes which interact with each other (Colas & Brent, 1998). In secondary metabolism enzymes might be present in aggregates, also known as meta-

bolic channels (Chapell, 1995; Hrazdina & Wagner, 1985; Rasmussen & Dixon, 1999; Srere, 1987). Genes encoding enzymes of such aggregates could be cloned through the yeast two-hybrid system, once a gene is known in the pathway concerned. The yeast twohybrid system was also applied in an elegant study suggesting that the flavonoid biosynthesis in Arabidopsis occurs in such an aggregate of enzymes (Burbulis & Winkel-Shirley, 1999). The yeast one-hybrid system, using multimers of promoter elements from a terpenoid indole alkaloid biosynthetic gene (STR) coupled to a selectable marker gene, has been used to pick up transcription factors that putatively regulate indole alkaloid biosynthesis in C. roseus (Menke et al., 1999; van der Fits, 2000; van der Fits et al., 2000). Differential screening is another approach for the identification of genes involved in a certain secondary metabolite pathway. This method was for example applied in efforts to clone a gene that encodes the cytochrome P450 enzyme geraniol-10-hydroxylase from

331 C. roseus. Although several clones were obtained using this approach, efforts to obtain evidence for the G10H activity of the encoded enzymes failed (Mangold, 1994; Vetter, 1992). Also PCR-based screening using homology of cytochrome P450 enzymes is no solution, as the plant contains numerous cytochrome P450 enzymes (Meijer, 1993). In this case the final solution came from the purification of the G10H enzyme and sequencing of some internal amino acid sequences (Collu et al., 1999). Lange et al. (2000) reported the use of expressed sequence tags (EST) to identify genes involved in the biosynthesis of essential oils in mint glandular trichomes. As these trichomes, in which 35% of the metabolism concerns the production of terpenoids, can be isolated in pure form, they offer an excellent system for isolation of mRNA specific for terpenoid biosynthesis. By sequencing about 1300 ESTs from the cDNA library and comparing these with known sequences, genes were selected for further functional expression in E. coli. A number of genes could thus be identified, among others from the MEP-pathway and monoterpenoid biosynthesis. With the rapidly expanding databases on plant gene sequences, the use of ESTs will lead to an exponential growth of number of recognized genes. However, also here the bottleneck will be in proving the function of the proteins. Testing of the function requires the availability of the putative substrate and an assay to measure product formation.

Approaches to engineering the plant cell factory We briefly discussed above the objectives of metabolic engineering in terms of possible applications. In terms of effect on the pathways involved one can thus distinguish a number of approaches (see Table 2). Obviously the various approaches require thorough knowledge of the pathway to know where rate limiting steps might occur; whether catabolism is an important factor in limiting accumulation; where branching occurs and thus competition for precursors exists; how the pathway is regulated; and the role of transport and compartmentation. For most of the approaches listed in Table 2 genes of the pathways concerned are needed, however, for certain approaches genes encoding antibodies could be envisaged; for example to reduce precursor availability, or to reduce active protein levels. Even catalytic antibodies (abzymes) have been developed that can be used instead of an enzyme (Haynes et al., 1994).

Table 2. Different approaches for meeting the objectives of metabolic engineering in plants More of a desired compound Increase flux to desired product Overcome rate limiting steps Reduce flux through competitive pathways Overexpress regulatory genes that induce the pathway concerned Inhibit catabolism product Increase number of producing cells New compound Use plant precursors for making new compounds for the plant or completely novel compounds Reduced levels of compound Reduce flux in biosynthetic pathway Increase flux to competitive pathways Block step out of biosynthetic pathway Suppress regulatory genes that upregulate the pathway concerned Increase catabolism

More of a desired compound To increase the flux in a pathway one can look at individual steps of the pathway, competitive pathways and regulatory genes. In principle every enzyme is a key enzyme. However, in several cases it was observed that a single enzyme in a pathway is inducible, and thus acts as an on/off switch for a pathway, for example the production of 2,3-dihydroxybenzoic acid (DHBA) in C. roseus is controlled by the inducible isochorismate synthase (ICS) (Moreno et al., 1994). Constitutive expression of the encoding gene indeed results in the constitutive production of DHBA. On the other hand one has to be aware of the concepts of metabolic networks, in which it is often difficult to redirect carbon fluxes (Stephanopoulos & Vallino, 1991; Cornish-Bowden et al., 1995; Kacser & Acerenza, 1993). A simple example may illustrate this. It concerns the regulation of aromatic amino acid biosynthesis and the importance of feedback inhibition in the biosynthesis of tryptophan. Tryptophan biosynthesis starts with the conversion of chorismate into anthranilate by the enzyme anthranilate synthase, which is feedback-inhibited by tryptophan. At the same time tryptophan induces the activity of chorismate synthase, an enzyme catalyzing the first step in the competitive pathway from chorismate leading to phenylalanine and tyrosine (Bongaerts 1998; Poulsen

332 & Verpoorte, 1991; Romero et al., 1995). Such a metabolic interlock (Stephanopoulos & Vallino, 1991) thus requires the overexpression of a gene encoding a protein that is not sensitive for feedback inhibition. However, to substantially increase tryptophan levels it is probably also required to engineer the competitive pathway. Cascante et al. (2000) compared the organization and regulation of the cell factory with an industrial factory and concluded that in many respects there are quite a few similarities. Many factors, such as co-factor availability, feedback inhibition, transport, compartmentation, pH, thus affect the flux through a pathway, and not only the activity and affinity of the individual enzymes. For the biosynthesis of tryptophan in yeast it has been shown that all five enzymes starting from the intermediate chorismate have to be increased at least 20-fold to obtain an increase in the flux to tryptophan, though the increase of this product was considerably less than the increase of enzyme activities (Cornish-Bowden et al., 1995; Kacser & Acerenza, 1993). It was concluded that all enzymes of the organism have their share in controlling the flux to the product of interest, but the contribution of each individual enzyme cannot be predicted a priori. Engineering single steps In the case of engineering individual steps for increasing enzyme activity, overexpression of the endogenous gene, or the introduction of a more suitable heterologous gene can be considered. The heterologous enzyme might have more favorable properties, such as no feedback inhibition by downstream products, or a higher affinity for the substrate. Such an enzyme might be from another source, but could also be engineered (Chartrain et al., 2000). Despite the above mentioned poor results with the single gene approach for engineering metabolic pathways in microorganisms, much work on engineering single steps has already been performed in plants, with different degrees of success as will be illustrated with some examples. In C. roseus cell suspension cultures the enzyme tryptophan decarboxylase is strongly regulated and thus controls the availability of one of the necessary precursors for terpenoid indole alkaloid biosynthesis (Figure 4) (for a review see Verpoorte et al., 1997). However, overexpression of the tdc gene only resulted in overproduction of the direct product, tryptamine (Canel et al., 1998; Goddijn et al., 1995; Whitmer,

1999), but no increase in indole alkaloid biosynthesis. Apparently the levels of tryptamine produced by the transient activity peak of TDC is sufficient for alkaloid production in the wild type cell culture. Overexpression of the str gene encoding for the enzyme that couples tryptamine with secologanin, however, did seem to result in a number of cases in an increase in alkaloid biosynthesis (Canel et al., 1998; Whitmer, 1999). STR has a more or less constant expression level during the growth cycle of wild type cells. Apparently in transgenic cells with a STR activity up to 50 times higher than in wild type cells, the higher level of the enzyme results in a more efficient channeling of the precursors tryptamine and secologanin towards alkaloids. Expression of the tdc gene in Peganum harmala resulted in the production of tryptamine which is converted into serotonin, but not in an increased production of harman derivatives (Berlin et al., 1993; Berlin & Fecker, 2000). Solving one limiting step immediately teaches us where the next is. However, there are also successful examples of the single-gene approach. To increase the level of flavonoids with an antioxidant activity in tomato, metabolic engineering is being studied as a possible approach. Overexpression of the chalcone isomerase gene from petunia in tomato resulted in a 72-fold increase of flavonol production in the fruits (Verhoeyen et al., 2000). From pharmaceutical point of view, the work of Hashimoto and co-workers (Hashimoto et al., 1993; 1994; Yun et al., 1992) on the cloning of the hyoscyamine-6-hydroxylase (H6H) gene and the subsequent introduction in Atropa belladonna resulting in the production of scopolamine in this plant is an excellent illustration of the great potential of applying metabolic engineering for trimming the plant cell factory. Interestingly the H6h-gene encodes a dioxygenase that first causes the introduction of a hydroxy group at the C-6 position, followed by an epoxidation by the same enzyme. The single-gene approach thus is an excellent way to find where a limiting step occurs in a pathway. In some cases it might be the key to improving productivity of the plant cell factory. As increased levels of secondary metabolites might be toxic for the cells, the use of inducible promoters is an interesting option. Sommer et al. (1998) showed the feasibility of this approach for the production of 4-hydroxybenzoic acid in tobacco cell cultures. Transformation of tobacco cells with the combination of a tetracycline inducible promoter, the E. coli gene enzyme encoding choris-

333

Figure 4. Biosynthesis of terpenoid indole alkaloids.

mate pyruvate lyase and a chloroplast target peptide sequence, resulted in a cell culture in which the 4-hydroxybenzoic acid production could be induced reversibly.

Reduce flux through competitive pathways Decreasing the flux through competitive pathways can be achieved by the introduction of an antisense gene of the competitive enzyme at the branch point. Though there are no actual examples of this approach, the feasibility to block a step in a metabolic pathway has been shown in many examples for both primary and secondary metabolism. However, examples have also been reported in which this approach failed (for a more extensive discussion see below) (Davies 2000; Robbins & Morris, 2000).

Regulatory genes Because of the complexity of the biosynthetic pathways, and the limited success of overexpression of single genes, the use of regulatory genes seems an interesting approach, as these may act as master switches for (part of) a complete pathway (Memelink et al., 2000). The feasibility of using regulatory genes for increasing production of secondary metabolites is illustrated in several studies reporting the ectopic expression of plant Myb and bHLH genes, encoding transcription factors. Their constitutive expression in the original plant or heterologous plant species affects the phenylpropanoid biosynthetic pathways (for a review see Memelink et al., 2000). For instance, Lloyd et al. (1992) showed that anthocyanin production in Arabidopsis and tobacco can be turned on by the maize R (bHLH) and C1 (Myb-

334 like) regulatory genes. Verhoeyen et al. (2000) reported that the overexpression of maize Lc (bHLH) and C1 transcription factors in tomato results in an up to 20-fold increased level of total flavonols in the fruit, apparently in this example the overexpression of a rate limiting enzyme was more effective (see above). In case of the terpenoid indole alkaloid biosynthetic pathway in C. roseus two genes for AP2domain transcription factors were cloned that regulate a number of genes in primary as well as secondary metabolism leading to terpenoid indole alkaloids (Menke, 1999; Menke et al., 1999a,b; van der Fits, 2000). The regulatory genes open interesting perspectives for further studies on the upregulation of biosynthetic pathways through their overexpression. However, plants contain many genes encoding transcription factors, about 13% of Arabidopsis genes are thought to be involved in transcription or signal transduction (Somerville & Somerville, 1999; The EU Arabidopsis Genome project, 1998). To make it even more complex, combinatorial interactions between the transcription factors are involved in the regulation of the expression of the structural genes of a pathway (Singh, 1998; Estelle & Chory, 1999). Genoud and Metraux (1999) described several examples of crosstalk between plant signaling pathways. They conclude that, although signaling networks are structured like neutral networks, Boolean networks are a more simple and logical framework for describing the information flow in cells, which can be used as model for further studies. From these reports it can be concluded that the simple overexpression of a single transcription factor might not always be sufficient, since its effect depends on the (epi) genetic make-up of the transformed cells. Reducing catabolism Although there are several studies on catabolism in secondary metabolism, this phenomenon has not been studied very extensively (Barz, 1977; Waller & Nowacki, 1978; Mihaliak et al., 1992). In plant cell cultures of C. roseus and Tabernaemontana species it was found that the rate of catabolism almost equals the rate of de-novo biosynthesis (Dos Santos et al., 1994; Dagnino et al., 1993; 1994; Schripsema et al., 1994). On the other hand Mihaliak et al. (1992) reported that there is no short term metabolism of monoterpenoids in peppermint leaves, and that in general one should be cautious in the interpretation of experiments

done under non-physiological conditions. The presence of natural storage compartments may play an important role in saving secondary metabolites from catabolism. No enzymes have yet been characterized in catabolism of secondary metabolism, but such knowledge is needed before being able to explore this approach to increase accumulation of a desired compound. Increase number of producing cells In plants often only a small percentage of the cells are actually producing secondary metabolites. Usually these are specialized cells that can produce high levels of secondary metabolites. For example, in C. roseus cell cultures it has been shown that the production of anthocyanins is dependent on the percentage producing cells. All producing cells have about similar levels of anthocyanins (Hall & Yeoman, 1986a,b). In the plant, glandular hairs producing large amounts of essential oils are another clear example. McCaskill and Croteau (1999) reviewed the strategies for bioengineering the development and metabolism in glandular tissues in plants. Lack of knowledge on the regulation of the development of glandular tissues is presently the major bottleneck. The possibility of increasing the number of producing cells is an interesting option, but requires further studies on the regulation of the differentiation process.

Production of new compounds Production of new compounds can be in terms of a new compound for a plant species, or a totally novel compound. Recombinatorial biochemistry for the production of novel compounds has successfully been developed for microorganisms (Chartrain et al., 2000). Particularly, the polyketide pathway is target for such studies, as it is the source of a number of important antibiotics. Unfortunately plants have no operons like microorganisms in which all the genes of antibiotic biosynthesis are clustered, thus facilitating cloning of genes. Currently only a few plant secondary metabolite genes are available which limits the possibilities for recombinatorial biochemistry in plants. However, for the production of new compounds in plants the use of microbial genes is also a feasible strategy. As proof-of-concept, the formation of some benzoic acid derivatives from the abundant precursor

335

Figure 5. Biosynthetic pathways leading to salicylic acid and 2,3-dihydroxybenzoic acid.

chorismate in plants by the introduction of microbial genes, can be mentioned. Heide and co-workers (Heide, 2000; Siebert et al., 1996) achieved the production of 4-hydroxybenzoic acid in tobacco, by overexpression of the ubiC gene from Escherichia coli, encoding chorismate pyruvate lyase in the plastids by using the CaMV 35S or the tetracycline-inducible Triple-Op promoter and the Rubisco small subunit signal peptide sequence. Verberne et al. (1998) introduced two microbial genes into tobacco, one encoding isochorismate synthase (entC gene from E. coli) and one encoding isochorismate pyruvate lyase (orfD gene

from Pseudomonas fluorescence). Under the control of the CaMV 35S promoter and coupled with the Rubisco signal peptide sequences for targeting to the plastids, the enzymes were constitutively expressed in tobacco, which resulted in the biosynthesis of salicylic acid (Figure 5). These tobacco plants were shown to be more resistant against primary infection with tobacco mosaic virus and against fungi. This is an example in which a new pathway is introduced in the plant, leading to a product that the plant is producing normally along a completely different, and much longer pathway. Plants are thought to biosyn-

336 thesize salicylate via the phenylpropanoid pathway, with among others, phenylalanine and benzoic acid as intermediates. The first example for a new compound in a plant using a plant gene concerned the introduction of a gene encoding dihydroflavonol reductase from maize into petunia. This resulted in a new brick red flower color (Meyer et al., 1987). Subsequently a number of examples of changed flower colors through metabolic engineering have been reported (for reviews see e.g. Davies, 2000; Holton & Cornish, 1995; Mol et al., 1995; Mulder-Krieger & Verpoorte, 1994). In connection with plant resistance Hain and coworkers extensively studied the constitutive production of the stilbene phytoalexin resveratrol in plants (Hain et al., 1990; 1993; for a review see Hain and Grimmig, 2000). This was achieved by overexpression of the stilbene synthase (STS) gene from grapevine in other plants, such as tobacco, tomato, potato and oil seed rape. The transgenic plants showed increased resistance against fungal infections. High levels of expression could be obtained by adding a transcriptional enhancer sequence of the 35S promoter to the natural promoter of the STS gene. By insertion of a tetramer of the enhancer 1.5 kb upstream of the transcription start site of the STS gene, a 14-fold increase of STS mRNA could be achieved in tobacco plants transformed via Agrobacterium tumefaciens. A 4-fold increase was obtained in shoots via direct gene transfer to protoplasts. Plants overproducing resveratrol are also of interest from nutritional point of view, as this stilbene is a strong antioxidant, which has among others been proposed as the compound involved in the health promoting effects of wine (French paradox) (Soleas et al., 1997; Jang et al., 1997). Another example concerning improving the nutritional value of food is the introduction of the carotenoid biosynthetic pathway into rice. This was achieved by using a bacterial phytoene desaturase and phytoene synthase in combination with lycopene cyclase from Narcissus pseudonarcissus under the control of the rice glutelin (Gt1) promoter (Ye et al., 2000). The transgenic rice grains contained low levels of carotenoids with provitamin A as major compound, making the grains yellow. Such rice maybe very important to overcome the vitamin A deficiency in many countries in which rice is the major staple food. Terpenoid indole alkaloids are a pharmaceutically important group of alkaloids, with about 15 of the 3000 known alkaloids having interesting pharmacological properties, including the already mentioned

C. roseus alkaloids. These alkaloids are found in about four plant families only, but one of the precursors, the iridoid secologanin is found in many other families. Introduction of the tdc gene, responsible for the production of tryptamine, and the str gene, responsible for the coupling of tryptamine with secologanin, in plants producing secologanin would lead to the production of strictosidine, the general precursor of all terpenoid indole alkaloids (Figure 4). Indeed, after transformation with these genes, Weigela ‘Styriaca’ hairy root cultures produced small amounts of strictosidine and ajmalicine (Hallard et al., 2000). The introduction of the tdc gene into tobacco, resulted in the production of tryptamine. Unexpectedly it was found that these plants cause a reduction of the reproduction of whitefly feeding on these plants (Thomas et al., 1995). On the other hand overexpression of the tdc gene in potato resulted in both reduced levels of tryptophan and phenylalanine and an increased susceptibility for fungal infections (Yao et al., 1995). For all the examples mentioned, the key point is the availability of sufficient amounts of the necessary precursors for the production of the new compounds. Toxicity of the products formed could be a further constraint (see above). In case of strictosidine it was for example observed that tobacco and Morinda citrifolia cells overexpressing the tdc and str genes, upon feeding secologanin excrete the produced strictosidine to the medium, in contrast with C. roseus, where this glucoalkaloid is accumulated in the vacuole. This point to differences in transport of this compound in heterologous systems. Reduced levels of compound Reducing levels of undesired compounds, such as toxic compounds or antifeedants in fodder is an interesting option. Reduction can be achieved by: – antisense gene- or sense gene-suppression of a step of the pathway involved; – the flux of a precursor of the undesired compound can be redirected into other products; – increasing catabolism; – decreasing level(s) of enzyme(s) of the pathway. Particularly antisense strategies are widely used to knock out certain biosynthetic steps, however, not always with success. To start with the failures, an antisense gene for chalcone synthase (CHS) from Phaseolus vulgaris in Lotus corniculatus root cultures did

337 not result in the desired decrease of condensed tannin levels (formed through polymerization of anthocyanins), but in some cases even in an increase (Robins & Morris, 2000). This was explained by the occurrence of a multigene family of CHS encoding genes. Down regulation of one or more members of this gene family by the antisense gene, is compensated for by overexpression of other Chs genes. This phenomenon was named differential modulation of a multigene family (Colliver et al., 1997). Efforts to increase levels of chalcones to alter flower colors, using an antisense chalcone isomerase, the enzyme that channel chalcones into the flavonoid/anthocyanin pathway, were also not successful (for a review see Davies, 2000). Reasons for this could be the lack of suppression due to the genomic location of the target gene; or a chemical conversion replacing the enzymatic conversion of chalcones into flavanones. On the other hand many successes have been reported for the antisense technology. For example an antisense gene for CHS was successfully applied to downregulate CHS in various flowers, resulting in different colors in plants such as rose, chrysanthemum, lisianthus and gerbera (for a review see Davies, 2000). In some cases similar results were obtained by transformation with the sense gene, this so-called cosuppression was first observed in petunia. Cosuppression has been observed in many other cases. Though in our work with C. roseus on the enzymes TDC and STR from the biosynthesis of terpenoid indole alkaloid cosuppression was so far not observed, whereas an antisense tdc gene was an effective means to downregulate the activity of this enzyme and thus alkaloid production in the cell cultures of this plant this (Goddijn et al., 1995). In efforts to reduce lignin levels in wood for the paper industry, transgenic plants with reduced levels of enzymes of the early steps of the phenylpropanoid biosynthesis such as phenylalanine ammonia lyase, cinnamate 4-hydroxylase and 4-coumarate:coenzyme A ligase, have been made. These plants did not only show reduced levels of lignin formation but also other pleiotropic effects on, among others, plant defense (Pallas et al., 1996; Maher et al., 1994). Targeting enzymes further down in the lignin pathway, such as caffeic acid methyltransferase, might be more successful in reducing lignin levels without affecting other functions (Ni et al., 1994). The second mentioned option of redirecting precursors has successfully been applied to decreasing biosynthesis of glucosinolates in canola. Glucosino-

lates acts as antifeedants in this major oil crop plant, and reduces its use as fodder. The glucosinolates are build up from among others tryptophan. By reducing the size of the tryptophan pool by converting it into tryptamine through overexpression of the tryptophan decarboxylase gene from C. roseus, the indole glucosinolate levels in seeds of greenhouse grown transgenic canola plants could be reduced to about 3% of the controls. However, in field trials these results could not be reproduced (Chavadej et al., 1994; De Luca, 2000). For the third option, increased catabolism, to our knowledge no examples have been reported. Besides targeting at a gene to block an enzyme activity also approaches targeting at the mRNA or protein level have been reported. This can be done by means of ribozymes (Rossi, 1995) or antibodies against the target enzyme (Conrad & Fiedler, 1994; Ma & Hein, 1996; Whitelam & Cockburn, 1995; Richardson & Marasco, 1996; Fischer et al., 2000). The use of antibodies to modulate the physiological effects of a regulatory compound formed in plants was reported for abscisic acid (Artsaenko; et al., 1994; 1995; Fischer et al., 2000). Little attention has so far been paid to the effect of blocking a secondary metabolite pathway through cosuppression or antisense genes on the accumulation of other secondary metabolites.

Plant compounds in microorganisms Once the genes of a secondary metabolite pathway are known, one can also envisage the overexpression in microorganisms. Is this feasible? The introduction and functional expression of plant genes in microorganisms is possible. Furthermore it requires that the necessary precursors are present in the microorganism, and the intermediates and the final product should not be toxic for the microbial cells. Lack of precursors can be solved by feeding. Toxicity is much more difficult to deal with, inducible promoters could be a possible solution. The major problem would, however, be the right concertation of the different enzymatic steps in a pathway, in which in the plant compartmentation also plays a role. It means that for short pathways expression in microorganisms could be achieved, but for long pathways it seems not to be a realistic approach. A successful example is the production of strictosidine or cathenamine in yeast. Expression of the C. roseus genes encoding strictosidine synthase

338 (STR) and strictosidine glucosidase (SGD) in yeast, results in a yeast strain which has SGD activity in the cells and excretes STR activity in the medium (Geerlings et al., 1998). Upon feeding secologanin and tryptamine to the transgenic yeast strain, strictosidine is formed in the medium. Grinding the cells with the medium results in the formation of the indole alkaloid cathenamine. By growing the yeast strain on the juice of the berries of Symphoricarpus albus, which contains both a carbohydrate source for the yeast as well as 1% of secologanin and then adding tryptamine, the yeast strain is capable of producing 2 g/l of strictosidine in three days. This productivity is much higher than found in any plant cell culture. Adding genes encoding additional steps of indole alkaloid biosynthesis may lead to an alternative production system for these plant compounds.

Stability An important aspect of the transgenic plants and plant cells is their stability. Gene silencing in transgenic plants is a well-known phenomenon (Finnegan & McElroy, 1994; Vaucheret et al., 1998). However, usually stable transformants have been obtained by extensive selection and crossing programs of the transgenic plants. In case of plant cell cultures of C. roseus, overexpressing the tdc and or str genes, where a continuous selection on the rapid growth occurs, we found that the original high levels of alkaloids gradually went down. However, the levels of the overexpressed enzymes remained much higher than in the controls. Stability of plant cell cultures is a general problem (Deus-Neumann, 1984; Ohta & Verpoorte, 1992) not necessarily connected with the presence of transgenes. The increased production of secondary metabolites in a cell might be a negative factor in the continuous selection process for fast growing cells in plant cell cultures. By using inducible promoters, for example, steroid (Schena et al., 1991), tetracycline (Gatz et al., 1992) or copper (Mett et al., 1993) dependent, growth and production can be separated, and thus stability might be increased.

Conclusions Although we still know very little about plant secondary metabolism, its role and its regulation, this limited knowledge has already been used quite extensively

to explore the possibilities of metabolic engineering. Because of the more or less empirical approach some unexpected results have been found. In general it can be concluded that it is difficult to predict the result of overexpression of a single gene. However, this type of experiments seems useful to perform, as at least it will lead to conclusions about possible (other) limiting steps in a pathway and a better understanding of the regulation of the flux through the pathways (e.g. Bate et al., 1994; Berlin, 2000). Using regulatory genes for upregulation of complete pathways seems a promising approach. The results also show that there is a great potential for a broad range of applications, ranging from improving the production of certain secondary metabolites to the introduction of new pathways in plants. However, we also must conclude that with our present knowledge of secondary metabolism a rational approach to metabolic engineering is in most cases not yet possible. For further developing the full potential of metabolic engineering it is thus necessary to increase our knowledge about plant secondary metabolism, at the level of the intermediates, the enzymes and the genes. But also on the physiology of the pathway, as transport, pH and cellular and subcellular compartmentation also play an important role. As plant secondary metabolism is plant species specific, the genome sequence of model plants, such as Arabidopsis is only of limited value. Sequencing the genome of the plant species concerned would be more useful, with proteomics and metabolomics as important tools for linking the genes with the secondary metabolite pathways (Scheme 1). The use of ESTs in combination with functional expression of selected genes is another approach, which in this connection holds great promise (Lange et al., 2000). The rapidly increasing number of plant gene sequences with a known function will result in an exponential growth of similar genes identified in other plants. For secondary metabolite studies, the major bottleneck in the coming years in functional genomics will be the assays for the enzymes involved in secondary metabolism. These assays require that all the intermediates are known and available for measuring enzyme activities either from the endogenous enzyme, or a protein obtained by heterologous expression of genes identified as being involved in secondary metabolite pathways. Or will we be able to find a way to predict binding of small molecules with proteins, making a similar logical link as exists between genes and proteins?

339 References Artsaenko O, Weiler EW, Muntz K and Conrad U (1994) Construction and functional characterization of a single chain FV antibody-binding to the plant hormone abscicic acid. J Plant Physiol 144: 427–429. Artsaenko O, Peisker M, Zurnieden U, Fiedler U, Weiler EW, Muntz K and Conrad U (1995) Expression of a single-chain FV antibody against abscisic acid creates a wilty phenotype in transgenic tobacco. Plant J 8: 745–750. Barz, W (1977) Catabolism of endogenous and exogenous compounds by plant cell cultures, In: Barz W, Reinhard E and Zenk MH (eds) Plant Tissue Culture and its Biotechnological Application. Springer-Verlag, Berlin. Bate NJ, Orr J, Ni WT, Meromi A, Nadlerhassar T, Doerner PW, et al. (1994) Quantitative relationship between phenylalanine ammonia lyase levels and phenylpropnaoid accumulation in transgenic tobacco identifies a rate determining step in natural product synthesis. Proc Natl Acad Sci USA 91: 7608–7612. Berlin J, Ruegenhagen C, Dietze P et al. (1993) Increased production of serotonin by suspension and root cultures of Peganum harmala transformed with a tryptophan decarboxylase cDNA clone from Catharanthus roseus. Transgen Res 2: 336–344. Berlin J and Fecker LF (2000) Genetic engineering of enzymes diverting amino acids into secondary metabolsim. In: Verpoorte R and Alfermann AW (eds) Metabolic Engineering of Plant Secondary Metabolism. (pp 195–216) Kluwer, Dordrecht. Bongaerts RJM (1998) The chorismate branching point in Catharanthus roseus. Aspects of anthranilate synthase regulation in relation to indole alkaloid biosynthesis. PhD-thesis, University of Leiden. Burbulis IE and Winkel-Shirley B (1999) Interactions among enzymes of the Arabidopsis flavonoid bioynthetic pathway. Proc Natl Acad Sci USA 26: 12929–12935. Cane D (2000) Biosynthesis meets bioinformatics. Science 287: 818–819. Canel C, Lopez-Cardoso MI, Whitmer S, van der Fits L, Pasquali G, van der Heijden R, et al. (1998) Effects of over-expression of strictosidine synthase and tryptophan decarboxylase on alkaloid production by cell cultures of Catharanthus roseus. Planta 205: 414–419. Cascante M, F. Ortega F and Marti E (2000) New insights into our understanding of the regulation and organization of cell factories. TibTech 18: 181–182. Chapell J (1995) Biochemistry and molecular biology of the isoprenoid biosynthetic pathway in plants. Annu Rev Plant Physiol Mol Biol 46: 521–547. Chartrain M, Salmon PM, Robinson DK and Buckland BC (2000) Metabolic engineering and directed evolution for the production of pharmaceuticals. Curr Opin Biotechnol 11: 209–214. Chavadej S, Brisson N and McNeil JN (1994) Redirection of tryptophan leads to production of low indole glucosinolate canola. Proc Natl Acad Sci USA 91: 2166–2170. Colliver SP, Morris P and Robbins MP (1997) Differential modification of flavonoid and isoflavonoid biosynthesis with an antisense chalcone synthase construct in transgenic Lotus corniculatus. Plant Mol Biol 35: 509–522. Conrad U and Fiedler U (1994) Expression of engineered antibodies in plants cells. Plant Mol Biol 26: 1023–1030. Colas P and Brent R (1998) The impactof two-hybrid and related methods in biotechnology. TibTech 16: 355–363. Collu G (1999) Geraniol 10-hydroxylase. PhD-thesis, Leiden University.

Contin A, van der Heijden R, Lefeber AWM and Verpoorte R (1998) The iridoid glucoside secologanin is derived from the novel triose phosphate/pyruvate pathway in a Catharanthus roseus cell culture. FEBS Lett 434: 413–416. Cordell GA (1974) The biosynthesis of indole alkaloids. Lloydia 37: 219–298. Cornish-Bowden A, Hofmeyr JHS and Cardenas ML (1995) Strategies for manipulating metabolic fluxes in bioetchnology. Bioorg Chem 23: 439–449. Cunningham FX and Gantt E (1998) Genes and enzymes of carotenoid biosynthesis in plants. Annu Rev Plant Physiol Plant Mol Biol 49: 557–583. Dagnino D, Schripsema J and Verpoorte R (1993) Comparison of terpenoid indole alkaloid production and degradation in two cell lines of Tabernaemontana divaricata. Plant Cell Rep 13: 95–98. Dagnino D, Schripsema J and Verpoorte R (1994) Terpenoid indole alkaloid biotransformation of suspension cultures of Tabernaemontana divaricata. Phytochemistry 35: 671–676. Davies KM (2000) Plant colour and fragrance. In: Verpoorte R and Alfermann AW (eds) Metabolic Engineering of Plant Secondary Metabolism. (pp. 127–163) Kluwer Academic Publishers, Dordrecht. De Luca V (2000) Metabolic engineering of crops with the tryptophan decarboxylase of Catharanthus roseus. In: Verpoorte R, and Alfermann AW (eds) Metabolic Engineering of Plant Secondary Metabolism. (pp. 179-194) Kluwer Academic Publishers, Dordrecht. Deus-Neumann B and Zenk MH (1984) Instability of indole alkaloid production in Catharanthus roseus cell suspension cultures. Planta Med 50: 427–431. Dos Santos R, Schripsema J and Verpoorte R (1994) Ajmalicine metabolism in Catharanthus roseus cell cultures. Phytochemistry 35: 677–681. DellaPenna D (1999) Nutritional genomics: Manipulating plant micronutrients to improve human health. Science 285: 375–379. Dixon RA (1999) Plant natural products: the molecular genetic basis of biosynthetic diversity. Curr Opin Biotechnol 10: 192–197. The EU Arabidopsis genome project (1998). Nature 391: 485. Drapeau D, Blanch HW and Wilke CR (1987) Economic assessment of plant cell culture for the production of ajmalicine. Biotechnol Bioeng 30: 946–953. Estelle M and Chory J (1999) Signals and pathways: keeping track of what’s going on. Curr Opin Plant Biol 2: 349–351. Finnegan J and McElroy D (1994) Transgene inactivation: Plants fight back! Bio/technology 12: 883–888. Fischer R, Drossard J, Schillberg S, Artsenko O, Emans N and Naehring JM (2000) Modulation of plant function and plant patrhogns by antibody expression. In: Verpoorte R and Alfermann AW (eds) Metabolic Engineering of Plant Secondary Metabolism. (pp. 87–109) Kluwer Academic, Dordrecht. Galneder E, Rueffer M, Wanner G, Tabata M and Zenk MH (1988) Alternative final steps in berberine biosynthesis in Coptis japonica cell cultures. Plant Cell Rep 7: 1–4. Geerlings A, Verpoorte R, van der Heijden R and Memelink J (1998) A method for producing terpenoid-indole alkaloids. International Patent Application PCT/NL99/00733, EP 98204116.2. Genoud T and Metraux JP (1999) Crosstalk in plant cell signalling: structure and function of the genetic network. Trends Plant Sci 4: 503–507. Goddijn OJM, Pennings EJM, van der Helm P, Verpoorte R and Hoge JHC (1995) Overexpression of a tryptophan decarboxylase cDNA in Catharanthus roseus crown gall calli results in in-

340 creased tryptamine levels but not in increased terpenoid indole alkaloid production. Transgenic Res 4: 315–323. Goldstein WE, Lazure LL and Ingle MB (1980) Product cost analysis. In: Staba EJ (ed) Plant Tissue Culture as a Source of Biochemicals. (pp. 191–234) CRC Press, Boca Raton. Goldstein WE (1999) Economic considerations for food ingredients produced by plant cell and tissue culture. In: TJ Fu, G Singh and WR Curtis (eds) Plant Cell and Tissue Culture for the Production of Food Ingredients. (pp 195–214) Kluwer Academic Publishers, Dordrecht. Gollwitzer J, Lenz R, Hampp N and Zenk MH (1993) The transformation of neopine to codeinone in morphine biosynthesis proceeds non-enzymatically. Tetrahedron Lett 34: 5703–5706. Grotewold E, Chamberlin M, Snook M, Siame B, Butler L, Swenson J, et al. (1998) Engineering secondary metabolism in maize cells ectopic expression of transcription factors. Plant Cell 10: 721– 740. Hain R and Grimmig B (2000) Modification of plant secondary metabolism by genetic engineering. In: Verpoorte R and Alfermann AW (eds) Metabolic Engineering of Plant Secondary Metabolism. (pp. 217–231) Kluwer Academic Publishers, Dordrecht. Hain R, Bieseler B, Kindl H, Schroder G and Stocker R (1990) Expression of a stilbene synthase gene in Nicotiana tabacum results in synthesis of the phytoalexin resveratrol. Plant Mol Biol 15: 325–335. Hain, Reif HJ, Krause E, Langebartels R, Kindl H, Vornam B, Wiese W, et al. (1993) Disease resistance results from foreign phytoalexin expression in a novel plant. Nature 361: 153–156. Hall RD and Yeoman MM (1986) Temporal and spatial heterogeneity in the accumulation of anthocyanins in cell cultures of Catharanthus roseus (L.)G. Don. J Exp Bot 37: 48. Hall RD and Yeoman MM (1986) Factors determining anthocyanin yield in cell cultures of Catharanthus roseus (L.)G. Don. New Phytol 103: 33–43. Hallard D, van der Heijden R, Verpoorte R, Lopez-Cardoso I, Pasquali G, Memelink J and Hoge JHC (1997) Suspension cultured transgenic cells of Nicotiana tabacum expressing tryptophan decarboxylase and strictosidine synthase cDNAs from Catharanthus roseus produce strictosidine upon feeding of secologanin. Plant Cell Rep 17: 50–54. Hallard D (2000) Transgenic plant cells for the production of indole alkaloids. PhD thesis, Leiden University, 2000. Harborne JB (ed) (1978) Biochemical aspects of plant and animal coevolution. Ann Proc Phytochem Soc Europe. Vol 15. Academic Press London. Hashimoto T and Yamada Y (1994) Alkaloid biogenesis: molecular aspects. Annu Rev Plant Physiol Plant Mol Biol 45: 257–285. Hashimoto T, Hayashi A, Amano Y, Kohno J, Iwanari H, Usuda S and Yamada Y (1991) Hyoscyamine 6ß-hydroxylase, an enzyme involved in tropane alkaloid biosynthesis, is localized ar the pericycle of the root. J Biol Chem 266: 4648–4653. Hashimoto T, Nakajima K, Ongena G et al. (1992) Two tropinone reductases with distinct stereospecificities from cultured roots of Hyoscyamus niger. Plant Physiol 100: 836–845. Hashimoto T, Yun D-J and Yamada Y (1993) Production of tropane alkaloids in genetically engineered root cultures. Phytochemistry 32: 713–718. Haynes MR, Stura EA, Hilvert D and Wilson IA (1994) Routes to catalysis: structure of a catalytic antibody and comparison with its natural counterpart. Science 263: 646–652. Heide L (2000) Expression of the bacterial ubiC gene opens a new biosynthetic pathway in plants. In: Verpoorte R and Alfer-

mann AW (eds) Metabolic Engineering of Plant Secondary Metabolism. (pp. 233–251) Kluwer Academic Press, Dordrecht. Hibino K and Ushiyama K (1999) Commercial production of Ginseng by plant cell culture technology. In: Fu TJ, Singh G and Curtis WR (eds) Plant Cell and Tissue Culture for the Production of Food Ingredients. (pp. 215–224) Kluwer Academic Publishers, Dordrecht. Holton TA and Cornish EC (1995) Genetics and biochemistry of anthocyanin biosynthesis. Plant Cell 7: 1071–1083. Hooykaas PJJ (2000) Agrobacterium, a natural metabolic engineer of plants. In: Verpoorte R and Alfermann AW (eds) Metabolic Engineering of Plant Secondary Metabolism. (pp. 51–67) Kluwer Academic Press, Dordrecht. Hrazdina G and Wagner G (1985) Metabolic pathways as enzyme complexes: evidence for the synthesis of phenylpropanoids and flavonoids in membrane associated enzyme complexes. Arch Biochem Biophys 237: 88–100. Jacobs DI, van der Heijden R and Verpoorte R (2000) Proteomics in plant biotechnology and secondary metabolism research. Phytochem. Anal. in press. Jang MS, Cai EN, Udeani GO, Slowing KV, Thomas CF, Beecher CWW, et al. (1997) Cancer chemopreventive activity of resveratrol, a natural product derived from grapes. Science 275: 218–220. Kacser H and Acerenza L (1993) A universal method for achieving increases in metabolite production. Eur J Biochem 216: 361–367. Kusnadi A, Nikolov ZL and Howard JA (1997) Production of recombinant proteins in transgenic plants: practical considerations. Biotechnol Bioeng 56: 473–483. Kutchan TM (1995) Alkaloid biosynthesis: The basis for metabolic engineering of medicinal plants. Plant Cell 7: 1059–1070. Kutchan TM (1998) Plant biotechnology and the production of alkaloids. Prospects of metabolic engineering. GA Cordell (ed) The Alkaloids. Vol. 50. (pp. 257–316) Academic Press, San Diego. Lange BM, Wildung MR, Stauber EJ, Sanchez C, Pouchnik D and Croteau R (2000) Probing essential oil biosynthesis and secretion by functional evaluation of expressed sequence tags from mint glandular trichomes. Proc Natl Acad Sci USA 97: 2934–2939. Leech MJ, May K, Hallard D, Verpoorte R, De Luca V and Christou P (1998) Expression of two consecutive genes of a secondary metabolic pathway in transgenic tobacco: molecular diversity influences levels of expression and product accumulation. Plant Mol Biol 38: 765–774. Leech MJ, Burtin D, Hallard D, Hilliou F, Kemp B, Palacios N, et al. (2000) Particle gun methodology as a tool in metabolic engineering. In: Verpoorte R and Alfermann AW (eds) Metabolic Engineering of Plant Secondary Metabolism. (pp. 69–86) Kluwer Academic Press, Dordrecht. Lichtenthaler HK (1999) The 1-deoxy-D-xylulose-5-phosphate pathway of isoprenoid biosynthesis in plants. Annu Rev Plant Physiol Plant Mol Biol 50: 47–65. Lloyd AM, Walbot V and Davis RW (1992) Arabidopsis and Nicotiana anthocyanin production activated by maize regulators R and C1. Science 258: 1773–1775. Ma JKC and Hein MB (1996) Antibody production and engineering in plants. In: GB Collins and RJ Shepherd (eds) Engineering Plants for Commercial Products and Applications. Vol. 792 (pp. 72–81) Annual New York Academic Science. Madduri K, Kennedy J and Rivola G (1998) Production of the antitumor drug epirubicin (4"-epidoxorubicin) and its precursor by a genetically engineered strain of Streptomyces peucetius. Nature Biotechnol 156: 69–74.

341 Maher EA, Bate NJ, Ni WT, Elkind Y, Dixon RA and Lamb CJ (1994) Increased disease susceptibility of transgenic tobacco plants with suppressed levels of preformed phenylpropanoids products. Proc Natl Acad Sci USA 91: 7802–7806. Mangold U, Eichel J, Batschauer A, Lanz T, Kaiser T, Spangenberg G, et al. (1994) Gene and cDNA for plant cytochrome P450 proteins (Cyp 72 family) from Catharanthus roseus and transgenic expression of the gene and a cDNA in tobacco and Arabidopsis thaliana. Plant Sci 96: 129–133. Martin C (1996) Transcription factors and the manipulation of plant traits. Curr Biol 7: 130–138. McCaskill D and Croteau R (1999) Strategies for bioengineering the development and metabolism of glandular tissues in plants. Nature Biotechnol 17: 31–36. McKnight TD, Bergey DR and Burnet RJ (1991) Expression of enzymatically active and correctly targeted strictosidine synthase in transgenic tobacco plants. Planta 185: 148–152. Meijer JJ, Van Gulik WM, Ten Hoopen HJG and Luyben KChAM (1987) The influence of shear stress on growth and morphology of Catharanthus roseus in continuous culture. In: OM Neijssel, RR van der Meer and KChAM Luyben (eds) Proceedings 4th Eur Congr Biotechnol, (p. 409) Vol 2, Elsevier, Amsterdam. Meijer AH, Souer E, Verpoorte R and Hoge JHC (1993) Isolation of cytochrome P-450 cDNA clones from the higher plant Catharanthus roseus by a PCR strategy. Plant Mol Biol 22: 379–383. Memelink J, Menke FLH, van der Fits L and Kijne JW (2000) Transcriptional regulators to modify secondary metabolism. In: Verpoorte R and Alfermann AW (eds) Metabolic Engineering of Plant Secondary Metabolism. Dordrecht, (pp. 111–125) Kluwer Academic Press. Menke FLH (1999) Elicitor signal transduction in Catharanthus roseus leading to terpenoid indole alkaloid biosynthetic gene expression. PhD thesis, Leiden University. Menke FLH, Champion A, Kijne JW and Memelink J (1999a) A novel jasmonate- and elicitor-responsive element in the periwinkle secondary metabolite biosynthetic gene Str interacts with a jasmonate- and elicitor-inducible AP2-domain transcription factor, ORCA2. EMBO J 18: 4455–4463. Menke FLH, Parchmann S, Mueller MJ, Kijne JW and Memelink J (1999b) Involvement of the octadecanoid pathway and protein phosphorylation in fungal elicitor-induced expression of terpenoid indole alkaloid biosynthetic genes in Catharanthus roseus. Plant Physiol 119: 1289–1296. Mett VL, Lochhead LP and Reynolds PHS (1993) Coppercontrollable gene expression system for whole plants. Proc Natl Acad Sci USA 90: 4567–4571. Meyer P, Heidmann I, Forkmann G and Saedler H (1987) A new petunia flower colour generated by transformation of a mutant with a maize gene. Nature 330: 677–678. Mihaliak C, Gershenzon J and Croteau R (1992) In: Petroski RJ and McCormick SP (eds) Secondary Metabolite Biosynthesis and Metabolism. (pp. 229–238) Plenum Press, New York. Mol JNM, Holton TA and Koes RE (1995) Floriculture: genetic engineering of commercial traits. Trends Biotechnol 13: 350–355. Moreno PRH, van der Heijden R and Verpoorte R (1995) Cell and tissue cultures of Catharanthus roseus (L.) G.Don: a literature survey II. Updating from 1988 to 1993. Plant Cell Tiss Org Cult 42: 1–25. Moreno PRH, van der Heijden R and Verpoorte R (1994) Elicitormediated induction of isochorismate synthase and accumulation of 2,3-dihydroxybenzoic acid in Catharanthus roseus cell suspension and shoot cultures. Plant Cell Rep 14: 188–191.

Mulder-Krieger Th and Verpoorte R (1994) Anthocyanins as Flower Pigments. Possibilities for Flower Colour Modification. (154 pp) Kluwer Academic Publishers, Dordrecht. Ni WT, Paiva NL and Dixon RA (1994) Reduced lignin in transgenic plants containing a caffeic acid O-methyltransferase antisense gene. Transgenic Res 3: 120–126. Ohta S and Verpoorte R (1992) Some accounts of variation (heterogeneity and/or instability) in secondary metabolite production by plant cell cultures. Ann Rep Nat Sci Home Econ 32: 9–23. Oosterkamp AJ, Hock B, Seifert M and Irth H (1997) Novel monitoring strategies for xenoestrogens. Trends Anal Chem 16: 544–553. Pallas JA, Paiva NL, Lamb C and Dixon RA (1996) Tobacco plants epigenetically suppressed in phenylalanine ammonia-lyase expression do not develop systemic acquired resistance in response to infection by tobacco mosaic virus. Plant J 10: 281–293. Pen J, Sijmonds PC, van Oooijen AJJ and Hoekema A (1993). Protein production in transgenic crops: analysis of plant molecular farming. In: A Hiatt (ed) Transgenic Plants. (pp. 239–254) Marcel Dekker Inc. Pimm SL, Russell GJ Gittleman JL and Brooks TM (1995) The future of biodiversity. Science 269: 347–350. Ponstein AS, Verwoerd TC and Pen J (1996) Production of enzymes for indutsrial use. In: GB Collins and RJ Shepherd (eds) Engineering Plants for Commercial Products and Applications. Annual New York Academic Science 792: 91–98. Poulsen C and Verpoorte R (1991) Roles of chorismate mutase, isochorismate synthase and anthranilate synthase in plants. Phytochemistry 30: 377–386. Quinn RJ (1999) QPRI’s system for screening natural products. In: Bohlin L. and J.G. Bruhn (eds) Bioassay Methods in Natural Product Research and Drug Development. Proceedings of the Phytochemical Society of Europe. Vol 43. (pp. 151–158) Kluwer Academic Publishers, Dordrecht. Ramos-Valdivia AC, van der Heijden R and Verpoorte R (1997) Isopentenyl diphosphate isomerase: a core enzyme in isoprenoid biosynthesis. A review of its biochemistry and function. Nat Prod Rep 14: 591–604. Rasmussen S and Dixon RA (1999) Transgene-mediated and elicitor-induced perturbation of metabolic channeling at the entry point into the phenylpropanoid pathway. Plant Cell 11: 1537– 1552. Richardson JH and Marasco WA (1995) Intracellular antibodies: development and therapeutic potential. TIBTECH 13: 306–310. Robbins MP and Morris P (2000) Metabolic engineering of condensed tannins and other phenolic pathways in forage and fodder crops. In: Verpoorte R and Alfermann AW (eds) Metabolic Engineering of Plant Secondary Metabolism (pp. 165–177) Kluwer Academic Press, Dordrecht. Rohmer M (1999) The discovery of a mevalonate-independent pathway for isoprenoid biosynthesis in bacteria, algae and higher plants. Nat Prod Rep 16: 565–574. Romero RM, Roberts MF and Phillipson JD (1995) Anthranilate synthase in microorganisms and plants. Phytochemistry 39: 263– 276. Rossi JJ (1995) Controlled, targeted, intracellular expression of ribozymes: progress and problems. TIBTECH 13: 301–306. Salas A and Mendez C (1998) Genetic manipulation of antitumor agent biosynthesis to produce novel drugs. TIBTECH 16: 475– 482. Sasse F, Buchholz F and Berlin J (1983a) Site of action of growth inhibitory tryptophan analogues in Catharanthus roseus cell suspension cultures. Z Naturforsch C Biosci 38: 910–915.

342 Sasse F, Buchholz F and Berlin J (1983b) Selection of cell lines of Catharanthus roseus with increased tryptophan decarboxylase activity. Z Naturforsch C Biosci 38: 916–922. Schripsema J, Dagnino D, dos Santos R and Verpoorte R (1994) Breakdown of indole alkaloids in suspension cultures of Tabernaemontana divaricata and Catharanthus roseus. Plant Cell Tiss Org Cult 38: 301–307. Siebert M., S. Sommer, S.-M. Li, Z. Wang, K. Severin and L. Heide (1996) Genetic engineering of plant secondary metabolism. Accumulation of 4-hydroxybenzoate glucosides results from the expression of the bacterial ubiC gene in tobacco. Plant Physiol 112: 811–819. Singh KB (1998) Transcriptional regulation inn plants: the importance of combinatorial control. Plant Physiol 118: 1111–1120. Soleas TJ, Diamandis EP and Goldberg DM (1997) Resveratrol: a molecule whose time has come and gone. Clin Biochem 30: 91– 113. Sommer S, Siebert M, Bechthold and Heide L (1998) Specific induction of secondary product formation in transgenic plant cell cultures using an iducible promoter. Plant Cell Rep 17: 891–896. Somerville C and Somerville S (1999) Plant functional genomics. Science 285: 380–383. De Luca V and St-Pierre B (2000) The cell and developmental biology of alkaloid biosynthesis. Trends Plant Sci 5: 168–173. St-Pierre B, Vazquez-Flota FA and De Luca V (1999) Multicellular compartmentation of Catharanthus roseus alkaloid biosynthesis predicts intercellular translocation of a pathway intermediate. Plant Cell 11: 887–900. Srere PA (1987) Complex of sequential metabolic enzymes. Annu Rev Biochem 56: 21–56. Stephanopoulos G and Vallino JJ (1991) Network rigidity and metabolic engineering in metabolite overproduction. Science 252: 1675–1681. Thomas JC, Adams DG, Nessler CL, Brown JK and Bohnert HJ (1995) Tryptophan decarboxylase, tryptamine, and reproduction of the whitefly. Plant Physiol 109: 717–720. Van der Fits L (2000) Transcriptional regulation of stress-induced plant secondary metabolism. PhD thesis, Leiden University. Van der Fits L, Zhang H, Menke FLH, Deneka M and Memelink J (2000) A Catharanthus roseus BPF-1 homologue interacts with an elicitor-responsive region of the secondary metabolite biosynthetic gene Str and is induced by elicitor via a JA-independent signal transduction pathway. Plant Mol Biol (in press). van Tegelen LJP, Bongaerts RJM, Croes AF, Verpoorte R and Wullems GJ (1999a) Isochorismate synthase isoforms from elicited cell suspension cultures of Rubia tinctorum. Phytochemistry 51: 263–269. van Tegelen LJP, Moreno PRH, Croes AF, Verpoorte R and Wullems GJ (1999b) Purification and cDNA cloning of Isochorismate synthase from elicited cell suspension cultures of Catharanthus roseus. Plant Physiol 119: 705–712. Van der Heijden R and Verpoorte R (1989) Cell and tissue culture of Catharanthus roseus (L.) G. Don: a literature survey. Plant Cell Tiss Org Cult 18: 231–280. Van Gulik WM, Meijer JJ, van der Heijden R, Verpoorte R and ten Hoopen HJG (1988) Feasibility of raubasine production by cell cultures of Catharanthus roseus. Report of Biotechnology Delft Leiden. Vaucheret H, Beclin C, Elmayan T, Feurebach F, Godon C, Morel JB, et al. (1998) Transgene-induced gene silencing in plants. Plant J 16: 651–659. Verberne M, Verpoorte R, van Tegelen LJP, Moreno PRH, Croes AF, Wullems GJ, et al. (1998) Salicylic acid pathway genes and their use for the induction of resistance in plants. International Patent

Application no 9 62 00 3189, 31/03/98, 6049 US P, 04/04/98 6049 US P2. Verberne MC, Budi Muljono RA and Verpoorte R (1999) Salicylic acid biosynthesis. In: Hooikaas PJJ, Hall MA and Libbenga KR (eds) Biochemistry and Molecular Biology of Plant Hormones. New Comprehensive Biochemistry. Vol. 33. (pp. 295–312) Elsevier, Amsterdam. Verhoeyen M, Muir S, Collins G, Bovy A and de Vos R (2000) Increasing flavonoid levels in tomatoes by means of metabolic engineering. Abstract at the 10th Symposium ALW-Discussion group Secondary metabolism in plant and plant cell. Verpoorte R, van der Heijden R and Moreno PRH (1997) Biosynthesis of terpenoid indole alkaloids in Catharanthus roseus cells. In: GA Cordell (ed) The Alkaloids. Vol. 49. (pp. 221–299) Academic Press, San Diego. Verpoorte R (1998) Exploration of nature’s chemodiversity: the role of secondary metabolites as lead for drug development. Drug Develop Today 3: 232–238. Verpoorte R (1999) Chemodiversity and the biological role of secondary metabolites, some thoughts for selecting plant material for drug development. In: Bohlin L and Bruhn JG (eds) Bioassay Methods in Natural Product Research and Drug Development. Proceedings of the Phytochemical Society of Europe. Vol. 43. Kluwer Academic Publishers, Dordrecht. Verpoorte R and Alfermann AW (1999) Metabolic Engineering of Plant Secondary Metabolism. Kluwer Academic Publishers, Dordrecht. Verpoorte R, van der Heijden R, van Gulik WM and ten Hoopen HJG (1991) Plant biotechnology for the production of alkaloids: present status and prospects. In: A Brossi (ed) The Alkaloids. Vol. 40. (pp. 1–187) Academic Press, San Diego. Verpoorte R, van der Heijden R, Schripsema J, Hoge JHC and ten Hoopen HJG (1993) Plant cell biotechnology for the production of alkaloids: present status and prospects. J Nat Prod 56: 186– 207. Verpoorte R, van der Heijden R and Memelink J (1998) Plant biotechnology and the production of alkaloids. Prospects of metabolic engineering. In: GA Cordell (ed) The Alkaloids. Vol. 50. (pp. 453–508) Academic Press, San Diego. Verpoorte R, van der Heijden R, ten Hoopen HJG and Memelink J (1999) Metabolic engineering of plant secondary metabolite pathways for the production of fine chemicals. Biotechnol Lett 21: 467–479. Verpoorte R (2000) Pharmacognsoy in the new millenium: leadfinding and biotechnology. J Pharm Pharmacol 52: 253–262. Vetter HP, Mangold U, Schroeder G, Marner FJ, Werckreichhardt D and Schroder (1992) Molecular analysis and heterologous expression of an inducible cytochrome P-450 protein from Periwinkle (Catharanthus roseus L.). Plant Pysiol 100: 998–1007. Waller GR and Nowacki EK (1978) Alkaloid Biology and Metabolism in Plants. Plenum Press, New York. Westphal K (1990) Large-scale production of new biologically active compounds in plant cell cultures. In: Nijkamp HJJ, van der Plas LHW and van Aartrijk J (eds), Progress in Plant Cellular and Molecular Biology. (pp. 601–608) Kluwer Academic Publishers, Dordrecht. Whitelam GC and Cockburn W (1996) Antibody expression in transgenic plants. Trends Plant Sci 1: 268–272. Whitmer S (1999) Aspects of terpenoid indole alkaloid formation by transgenic cell lines of Catharanthus roseus over-expressing tryptophan decarboxylase and strictosidine synthase. PhD thesis, Leiden University. Wijnsma R, Go JTKA, van Weerden IA, Harkes PAA, Verpoorte R and Baerheim Svendsen A (1985) Anthraquinones as phytoalex-

343 ines in cell and tissue cultures of Cinchona species. Plant Cell Rep 4: 241–244. Wijnsma R and Verpoorte R (1986) Anthraquinones in Rubiaceae. Fortschritte Chemie Organischer Naturstoffe 49: 79–149. Yamamoto H, Katano N, Ooi A and Inoue K (2000) Secologanin synthase which catalyzes the oxidative cleavage of loganin into secologanin is a cytochrome P450. Phytochemistry 53: 7–12. Yao K, De Luca V and Brisson N (1995) Creation of a metabolic sink for for tryptophan alters the phenylpropanoid pathway and the susceptibility of potato to Phytophthora infestans. Plant Cell 7: 1787–1799. Ye XD, Al-Babili S, Kloti A, Zhang J, Lucca P, Beyer P and Potrykus I (2000) Engineering the provitamin A (beta-carotene)

biosynthetic pathway into (carotenoid-free) rice endosperm. Science 287: 303–305. Yun D-J, Hashimoto T and Yamada Y(1992) Metabolic engineering of medicinal plants; Transgenic Atropa belladonna with an improved alkaloid composition. Proc Natl Acad Sci USA 89: 11799–11803. Zenk MH (1995) Chasing the enzymes of alkaloid biosynthesis. In: Golding BT, Griffin RJ and Maskill H (eds) Organic Reactivity: Physical and Biological Aspects. (pp. 89–109) Royal Society of Chemistry, London. Zenk MH (1991) Chasing the enzymes of secondary metabolism: plant cell cultures as a pot of gold. Phytochemistry 30: 3861– 3863.