Differential Induction of Sesquiterpene Metabolism in Tobacco Cell Suspension Cultures by Methyl Jasmonate and Fungal Elicitor

Archives of Biochemistry and Biophysics Vol. 381, No. 2, September 15, pp. 285–294, 2000 doi:10.1006/abbi.2000.1961, available online at http://www.idealibrary.com on

Differential Induction of Sesquiterpene Metabolism in Tobacco Cell Suspension Cultures by Methyl Jasmonate and Fungal Elicitor 1 Alejandra Mandujano-Cha´vez,* ,† ,2 Mark A. Schoenbeck,* Lyle F. Ralston,* Edmundo Lozoya-Gloria,† and Joseph Chappell* ,3 *Plant Physiology/Biochemistry/Molecular Biology Program, Agronomy Department, University of Kentucky, Lexington, Kentucky 40546-0091; and †Departamento de Ingenierı´a Gene´tica de Plantas, CINVESTAV-IPN Unidad Irapuato, P.O. Box 629, CP 36500 Irapuato Gto., Me´xico

Received April 12, 2000, and in revised form June 6, 2000

Jasmonates are well documented for their ability to modulate the expression of plant genes and to influence specific aspects of disease/pest resistance traits. We and others have been studying the synthesis of sesquiterpene phytoalexins in elicitor/pathogen-challenged plants and have sought to determine if methyl jasmonate (MeJA) could substitute for fungal elicitors in the induction of capsidiol accumulation by tobacco cell cultures. The current results demonstrate that MeJA does in fact induce phytoalexin accumulation, but with a much more delayed induction time course than elicitor. While elicitor treatment induced strong but transient changes in key enzymes of sesquiterpene biosynthesis, sesquiterpene cyclase, and aristolochene/deoxy-capsidiol hydroxylase, MeJA did not. Instead, MeJA caused a protracted induction of cyclase activity and only a low level of hydroxylase activity. MeJA induced the expression of at least two sesquiterpene cyclase genes, including one that had not been observed previously in elicitor-induced mRNA populations. Only a small portion of the total sesquiterpene cyclase mRNA induced by MeJA was associated with polysomal RNA, suggesting that the MeJA treatment imposed both transcriptional and posttranscriptional regulation in tobacco cells. These results are not consistent with MeJA playing a role in orchestrating defense responses in elicitor-treated tobacco cells, but do provide evidence that MeJA in1

The nucleotide sequences reported in this study have been deposited with GenBank; the Accession No. for the EAS4 cDNA is L04680 and for the EAS-MeJA cDNA it is AF272244. 2 Permanent address: Departamento de Ingenierı´a Gene´tica de Plantas, CINVESTAV-IPN Unidad Irapuato, P.O. Box 629, CP 36500 Irapuato Gto., Me´xico. 3 To whom correspondence should be addressed. Fax: (859) 2577125. E-mail: [email protected]. 0003-9861/00 $35.00 Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.

duces a subset of genes coding for the biosynthesis of sesquiterpene phytoalexins. © 2000 Academic Press Key Words: methyl jasmonate-defense response; phytoalexins.

Jasmonates are well documented for their ability to modulate the expression of plant genes and to influence specific aspects of disease/pest resistance traits (1–3) and plant growth and development (4). Induction of proteinase inhibitor (Pin) 4 gene expression in wounded tomato leaves was one of the earliest examples of these observations (5). Farmer et al. (6) demonstrated that exogenous jasmonates were capable of directly inducing Pin gene expression and correlated the release of jasmonates from wounded leaf tissue with the subsequent appearance of proteinase inhibitor proteins/activity in nonwounded leaves located some distance from the wounded leaf. Numerous reports have followed that document how plants respond to elicitors or physical treatments by a transient accumulation of jasmonates (7–9), and how exogenous application of jasmonates can induce expression of defense genes (10, 11), and genes for various biosynthetic pathways (12– 14) and the accumulation of compounds representing a wide range of chemical classes (9, 13, 15, 16), including phytoalexins (17–20). Together this type of information has provided strong support for the suggestion that jasmonates act as signal molecules within a transduc4

Abbreviations used: EAS, 5-epi-aristolochene synthase; HMGR, 3-hydroxy-3-methylglutaryl CoA reductase; JA, Jasmonic acid; MeJA, methyl jasmonate; FPP, farnesyl diphosphate; GGPP, geranylgeranyl diphosphate; DIECA, diethyldithiocarbamic acid; Pin, proteinase inhibitor. 285

286

´ VEZ ET AL. MANDUJANO-CHA

SCHEME I. The biosynthetic pathway for capsidiol. The cyclization of farnesyl diphosphate to 5-epi-aristolochene is catalyzed by a sequiterpene cyclase, 5-epi-aristolochene synthase (EAS). The order for the two subsequent hydroxylations of 5-epi-aristolochene is not known (33, 50) and therefore depicted by broken arrows.

tion cascade coordinating or orchestrating defense responses in plants (21, 22). The role of jasmonates in controlling gene expression and secondary metabolism is, however, complicated by a number of observations, especially with regard to isoprenoid metabolism. Choi et al. (23) reported that jasmonates differentially regulate HMG-CoA reductase genes in potato. At low concentrations of MeJA, HMG1 gene expression and steroid glycoalkaloid accumulation in wounded potato tuber disks were moderately enhanced, but suppressed by higher concentrations. In contrast, the induction of HMG2 gene expression and sesquiterpene phytoalexin accumulation occurred only in elicitor-induced tuber disks and this induction was suppressed by high concentrations of MeJA. Similarly, Maldonado-Mendoza et al. (24) isolated several HMGR genes from Camptotheca acuminate and demonstrated differential expression of these genes in relationship to the accumulation of camptothecin, a terpene indole alkaloid. Like the potato HMG1 gene, the wound inducibility of the C. acuminate HMG1 gene was suppressed by exogenous MeJA, but expression of this gene was not correlated with the accumulation of camptothecin under normal conditions. While such results have not supported a role of jasmonates in orchestrating the biosynthesis of these particular compounds, they have suggested the possibility of differential expression of members within gene families for the purpose of assembling metabolic channels dedicated to the biosynthesis of the specific endproduct metabolites (23–25). In contrast to the reports noted above, several recent investigations have reported that jasmonates do play a role in the induction of terpene metabolism. For example, Catharanthus roseus seedlings (26) and cell suspension cultures (27) accumulate terpene indole alkaloids in response to MeJA. The accumulation of these mixed alkaloids was furthermore correlated with the

MeJA-inducible expression of two of the corresponding biosynthetic genes (22, 28). Hefner et al. (16) recently demonstrated that MeJA induced the accumulation of the diterpene taxol in Taxus canadensis cell cultures and that this accumulation was correlated with the induced expression of two genes coding for downstream biosynthetic enzymes, GGPP synthase, and taxadiene synthase. Singh et al. (29) also reported on the ability of MeJA to induce terpene accumulation by root cultures of Hyoscyamus muticus. Interestingly, wounded H. muticus root cultures did not accumulate any sesquiterpene phytoalexins until treated with MeJA or a fungal elicitor. However, elicitor-treated root cultures accumulated predominately lubimin while the MeJAtreated cultures accumulated solavetivone, a putative precursor of lubimin (30). Last, MeJA acted synergistically with low concentrations of elicitor to stimulate the accumulation of relatively high levels of solavetivone. We and others have been studying synthesis of sesquiterpene phytoalexins, primarily capsidiol, in tobacco in response to pathogens (31, 32) and fungal elicitors (33, 34) and in pepper in response to UV irradiation and arachidonic acid (35, 36). Previous work demonstrated that the elicitor-induced accumulation of capsidiol was correlated with the induction of 5-epiaristolochene synthase (EAS), a putative branch point sesquiterpene cyclase diverting FPP from the general isoprenoid biosynthetic pathway to the dedicated synthesis of sesquiterpene phytoalexins (Scheme I) (37). The induction of this particular enzyme activity has been correlated with the expression of one member of a 12- to 15-member gene family (38, 39). In the current work, we sought to determine if MeJA could substitute for fungal elicitors in the induction of capsidiol accumulation in tobacco cell cultures. If so, our intent was to compare the MeJA induction mechanism to that of elicitor, assessing the possibility that MeJA might, in part, mediate elicitor-induced responses. While our re-

INDUCTION OF SESQUITERPENE METABOLISM IN TOBACCO CELL SUSPENSIONS

sults indicate that MeJA does induce sesquiterpene metabolism in tobacco cell cultures, it apparently does so via a mechanism(s) different from that induced by fungal elicitor. MATERIALS AND METHODS Cell cultures and induction treatments. Cell suspension cultures of Nicotiana tabacum, cultivar Kentucky 14, were maintained and subcultured in Murashige–Skoog media. Cultures in their rapid phase of growth (3 days old) were used for all the experiments described herein (40). Induction treatments were performed by the addition of cellulase from Trichoderma viride (Calbiochem), methyl jasmonate (Aldrich Chem Co) dissolved in ethanol, or parasiticein, an elicitin protein (41), at the concentrations indicated. Parasiticein was purified from Escherichia coli cells overexpressing recombinant parasiticein protein containing a carboxy terminal histidine purification tag (Back and Chappell, unpublished). Appropriate control treatments included cultures treated with a corresponding volume of water or ethanol. All the experiments were replicated in several independent experiments. While the absolute values presented may have varied between experiments by as much as 50%, the trends and time courses were consistent. Detection of phytoalexins. Detection and quantification of capsidiol in the extracellular culture media was performed according to Chappell et al. (34). Essentially, chloroform extracts of the culture media were concentrated and the level of capsidiol in the concentrated samples quantified by GC using hexadecane as an internal standard. In some instances, the culture media was evaluated for antimicrobial compounds using a Cladosporium cucumerinum bioassay (42). The fungus was grown on standard V8 media and the spores harvested in a 2.5 mM KPO 4, pH 6.0, solution containing (per liter): 20 g glucose, 2 g casamino acids, 2 mg ZnCl 2, 2 mg MnSO 4 䡠 4H 2O, 110 mg KCl, and 1.5 g MgSO 4 䡠 7H 2O. Filtered spore suspensions were sprayed onto silica TLC plates containing media extracts separated in acetone:cyclohexane (1:1). The plates were incubated in a moist, dark environment at room temperature and visually inspected for zones inhibiting spore germination 5–10 days later. Cell death assays. Cell death was determined according to the dye-exclusion assay of Glazener et al. (43). Aliquots of cells were stained for 15 min in 0.1% Evans Blue (Sigma) and then transferred to a disposable column and washed three to five times with water to remove excess stain. Washed cells were ground in 0.5% SDS, centrifuged for 5 min, and the absorbance of the resulting supernatant determined at 600 nm. The extent of cell death was assessed relative to control cells, which did not receive any treatment. Enzyme assays and immunodetection. Sesquiterpene cyclase enzyme assays were performed with total protein extracts according to Vo¨geli and Chappell (44). Cells were homogenized in 80 mM potassium phosphate buffer, pH 7.0, 20% glycerol, 10 mM sodium metabisulfite, 15 mM MgCl 2, 10 mM sodium ascorbate, 1% PVP (MW 40,000), and 5 mM ␤-mercaptoethanol. After centrifugation, the conversion of 10 –20 ␮M radiolabeled [1- 3H]FPP (American Radiolabeled Chemicals, Inc.) by the supernatant fraction to hexane extractable products was determined. Verification of the radiolabeled products as 5-epi-aristolochene was accomplished by comparative argentation TLC using authentic 5-epi-aristolochene. For immunodetection of cyclase protein, total protein samples of 15 ␮g were separated by SDS–PAGE, transferred to nitrocellulose membranes and the cyclase protein detected with a monoclonal antibody prepared against the native cyclase protein (45). Protein concentration was determined by the Bradford (Bio-Rad) assay. 5-Epi-aristolochene hydroxylase activity was measured as the incorporation of [ 3H]-5-epi-aristolochene into capsidiol by intact cells. [ 3H]-5-Epi-aristolochene was produced by incubating an excess of [1- 3H]FPP (1 ␮M, 20.5 Ci/mmol) with recombinant 5-epi-aristolochene synthase enzyme purified to homogeneity (46, 47). The

287

hexane extractable radioactivity from these incubations was treated with a small amount of silica gel before quantifying the yield of radioactive 5-epi-aristolochene. Control and elicitor-treated cells were then incubated with [ 3H]-5-epi-aristolochene (approximately 100,000 DPM at 2.5 nM) for 3-h periods at various points during an induction time course before collecting the cell and media samples. The media samples were extracted with organic solvents as described above, and the amount of radioactivity incorporated into capsidiol determined. For these determinations, samples were separated by TLC and the zones corresponding to capsidiol (R f of 0.3) scraped from the plates for liquid scintillation counting. Detection of cyclase mRNA. Total RNA was extracted from control and induced tobacco cells using the Trizol reagent according to the manufacturer’s suggestions (Gibco-BRL). RNA samples (10 ␮g) were size-separated by electrophoresis in 1.2% agarose-formaldehyde gels before being transferred to a nylon (Z-probe, Bio-Rad) membrane by capillary transfer in 10⫻ SSC overnight (48). The nylon membrane was hybridized with a sesquiterpene cyclase probe corresponding to the full-length EAS4 cDNA (GenBank L04680) described by Facchini and Chappell (38). Radioactive labeling of the DNA fragment was performed with [␣- 32P]dCTP using the Random Primer Kit of Stratagene. Hybridizations were performed at 42°C in 50% formamide, 0.12 M NaHPO4, pH 7.2, 0.25 M NaCl, 7% SDS, 1 mM EDTA. Membranes were then washed with 0.2⫻ SSC, 0.1% SDS at 42°C for 25 min, and twice at room temperature with 2⫻ SSC, 0.1% SDS (48). Hybridization was detected using a Phosphorimager (Molecular Dynamics). Polysomal RNA isolation. Polysomal RNA was extracted from control or elicited tobacco cells 24 h after induction treatments. Ten grams of cells was ground to a fine powder with liquid nitrogen, homogenized with 20 ml polysomal extraction buffer (200 mM Tris– HCl, pH 8.0, 200 mM KCl, 250 mM sucrose, 300 mM MgCl 2, 2.5 mM dithiothreitol, 10 mM EGTA), and centrifuged at 3000g for 5 min at 4°C. The supernatant was incubated with 1% Triton X-100 (v/v) on ice for 10 min before centrifuging the samples at 20,000g for 10 min at 4°C. Four milliliters of the 20,000g supernatant was layered onto 1 ml of a 50% (w/v) sucrose cushion in 40 mM Tris–HCl, pH 8.0, 20 mM KCl, 10 mM MgCl 2 and centrifuged at 100,000g for 120 min in a SW 55Ti rotor at 4°C. The polysomal pellet was dissolved in 200 ␮l of 40 mM Tris–HCl, pH 8.0, 20 mM KCl, 10 mM MgCl 2, phenol– chloroform extracted and the polysomal RNA precipitated with ethanol. The sesquiterpene cyclase mRNA was detected within the polysomal RNA (10 ␮g) as described above. Amplification of cyclase mRNAs. A reverse transcription–polymerase chain reaction (RT-PCR) was carried out with total RNA isolated from tobacco cell suspension cultures induced with 500 ␮M MeJA for 36 h using the one-step RT-PCR system of Gibco-BRL. The reverse primer was oligo(dT) 27 V (V representing A, T, G, and C). Specific forward primers were used (F1, 5⬘-TCTGCACTTCAATTTCGATTGCTC-3⬘; and F2, 5⬘-CAAGTGACACATGCCCTTGAGCAA-3⬘) which correspond to conserved regions within the second and third exons of the sesquiterpene cyclase gene family within tobacco (Schoenbeck and Chappell, unpublished). The RT-PCRs were carried out according to the manufacturer’s instructions using annealing and elongation conditions of 2.5 min at 50°C and 2 min at 72°C, respectively. The reaction products were ligated into the pCR2.1 vector (Original TA Cloning Kit, Invitrogen) and sequenced using the dideoxy nucleotide termination method with an ABI 310 DNA sequencer (PE Applied Biosystems). DNA sequences were verified by sequencing both strands of DNA and by sequencing three or more independently isolated clones.

RESULTS

Capsidiol accumulation is induced by MeJA and cellulase elicitation. Exogenously applied jasmonic acid (JA) and its methyl ester, MeJA, are capable of inducing defense proteins and secondary defense metabo-

´ VEZ ET AL. MANDUJANO-CHA

288 TABLE I

Dose-Dependent Response of Tobacco Cell Cultures to MeJA Treatment

Treatment

Capsidiol production (ng/ml)

Control Cellulase 0.001 mM MeJA 0.01 mM MeJA 0.1 mM MeJA 0.5 mM MeJA 1.0 mM MeJA 2.0 mM MeJA

0.0 1837.0 6.7 8.5 77.8 51.2 35.4 12.0

The decrease in capsidiol concentration observed after 48 h of cellulase treatment (Fig. 1A) has been reported previously by us (34, 37, 40) and other investigators (49). Threlfall and Whitehead (33) ascribed the

Note. Tobacco cell cultures were treated with cellulase (0.5 ␮g/ml) or the indicated concentrations of MeJA for 24 h before collecting the culture media. The media samples were extracted with chloroform, concentrated, and aliquots analyzed for capsidiol by GC.

lites in a wide range of plant species (6, 9, 17–20). Because of our interest in determining the mechanisms regulating sesquiterpene phytoalexin biosynthesis in solanaceous plants (30, 33, 37), we tested the ability of MeJA to induce capsidiol production in tobacco cell suspension cultures. Previous work had documented the accumulation of capsidiol in the culture media of tobacco cell suspension cultures in response to several fungal elicitors including Phytophthora parasitica cell wall fragments, cellulase from Trichoderma viride, and elicitin proteins from Phytophthora species (32, 40). The capsidiol secreted by elicitor-treated cells can be conveniently extracted from the culture media using organic solvents and qualitatively examined by TLC or quantified by GC. Significant capsidiol accumulation was observed at 10 ␮M MeJA with maximum accumulation occurring at 100 to 500 ␮M in a dose-response assay for capsidiol production after 24 h of MeJA treatment (Table I). Cell cultures treated with concentrations of MeJA above 1 mM accumulated lower levels of capsidiol. Nonetheless, the levels of capsidiol accumulating (77 ng/ml) within 24 h of optimal concentrations of MeJA were at least 23 times lower than those observed in cellulase (elicitor)-treated cell cultures (1837 ng/ml). The MeJA dose–response experiments were done initially assuming a similar time course of events as for elicitor-treated cell cultures. To validate this assumption, the accumulation of capsidiol by cell cultures over a 72-h period was determined at a concentration of 0.5 mM MeJA (Fig. 1A). As noted for the dose–response experiments, significant but low levels of capsidiol were detected within 24 h of MeJA addition. However, a much more dramatic accumulation of capsidiol occurred 48 to 72 h after addition of MeJA to the cultures. This contrasts with the cellulase-treated cultures, which showed maximal accumulation between 6 and 48 h after elicitor addition to the cultures.

FIG. 1. Induction time course for capsidiol accumulation, sesquiterpene cyclase enzyme activity, and sesquiterpene hydroxylase activity in cellulase- (0.5 ␮g/ml) (䊐), MeJA- (0.5 mM) (Œ) or controltreated (■) cell cultures. Extracellular sesquiterpenes were extracted from the cell culture media samples collected at the designated time points and the level of capsidiol in these samples determined by GC (A). Sesquiterpene cyclase enzyme activity was determined in extracts prepared from cells collected at the indicated time points (B). Sesquiterpene hydroxylase activity was determined using an indirect assay. Cell cultures were incubated with [ 3H]-5epi-aristolochene for 3 h, ending at the indicated time points before quantifying the incorporation of radioactivity into extracellular capsidiol, a dihydroxylated form of aristolochene (C). Each panel represents an independent experiment.

INDUCTION OF SESQUITERPENE METABOLISM IN TOBACCO CELL SUSPENSIONS TABLE II

Dose-Dependent Induction of Sesquiterpene Cyclase Enzyme Activity in MeJA-Treated Tobacco Cell Cultures

Treatment

Sesquiterpene cyclase activity (nmol/mg prot 䡠 h)

Control Cellulase Ethanol 0.001 mM MeJA 0.01 mM MeJA 0.1 mM MeJA 0.5 mM MeJA 1.0 mM MeJA 5.0 mM MeJA

0.3 86.4 1.7 8.4 4.8 11.2 34.6 28.7 0.6

Note. Tobacco cell suspension cultures were incubated with cellulase (0.5 ␮g/ml) or the indicated concentrations of MeJA for 48 h before determining the sesquiterpene cyclase enzyme activity associated with the cells.

net decrease of capsidiol in elicitor-treated cell cultures to an acetylation of the capsidiol at one or both of its hydroxyl substituents. We have not measured this activity directly in the current experiments, but note that the turnover or catabolism of capsidiol by whatever means seems limited in the MeJA-treated samples. Another important comparison between the two differently treated cell cultures is the accumulation rate of capsidiol. The maximum accumulation rates of capsidiol by the cellulase-treated cells and that by the MeJA-treated cells were roughly equal: 0.1 ␮g of capsidiol/ml of cell culture per hour for the cellulasetreated cultures and 0.125 ␮g/ml h ⫺1 for the MeJAtreated cultures. Integration of this accumulation rate over the respective time period translates into the absolute amount of capsidiol accumulating in the MeJAtreated cultures well within the range observed for the elicitor-treated cell cultures. The authenticity of the phytoalexin produced after MeJA treatment was verified in comparison to authentic capsidiol by GC-MS (data not shown) and in TLC-bioassays with Cladosporium cucumerinum (42), demonstrating the antimicrobial action of the phytoalexin accumulating 48 to 72 h after MeJA treatment (data not shown). Differential induction of sesquiterpene cyclase and hydroxylase activities in MeJA and elicitor-treated cells. Capsidiol biosynthesis entails the diversion of FPP, from the general isoprenoid biosynthetic pathway via the action of a sesquiterpene cyclase, EAS, followed by the action of two hydroxylating activities (33, 45, 50). Accumulation of capsidiol in elicitor-treated cells has previously been correlated with the induction of EAS activity (37). To determine if a similar association also held for the MeJA induction of capsidiol, the dosedependent induction of the EAS activity was examined (Table II). Similar to the induction profile for capsidiol

289

accumulation, low but significant levels of enzyme activity were induced by concentrations of MeJA below 100 ␮M with maximal induction occurring at 500 ␮M MeJA. In contrast to the induction pattern for capsidiol, the maximum cyclase activity induced by MeJA was within twofold of that induced by elicitor treatment in the 48-h incubation period. Capsidiol accumulation under these conditions was 10 times greater for the elicitor treatment than for the MeJA treatment (see Fig. 1A). The induction time courses of cyclase activity in MeJA and fungal elicitor-treated cultures were determined as a further means of correlating this enzyme activity with the accumulation of capsidiol (Fig. 1B). A rapid induction of cyclase activity was noted within 6 h of fungal elicitor addition to the cell cultures, which increased up to a maximum by 36 h before declining approximately 15% over the next 36 h. The induction pattern for cyclase activity by 500 ␮M MeJA resembled that induced by elicitor up to 36 h. However, the rate at which cyclase activity accumulated was only 50% of that induced by elicitor treatment. A slight but significant accumulation of activity in the MeJA-treated cells continued over the second 36-h period, which was not evident in the elicitor-treated cells. As reported previously (36, 43), a reasonable correlation between the induction of cyclase activity and capsidiol accumulation was evident in the elicitortreated cell cultures (comparing Figs. 1A and 1B). That is, the induction of cyclase activity preceded the accumulation of capsidiol up to 48 h after elicitor addition, but the subsequent decrease of capsidiol occurred when relatively high levels of cyclase activity were evident. This inconsistency at the later time points could arise from changes in the turnover rate or acetylation of capsidiol (33), aspects which were not accounted for in the current work, as well as limitations in the aristolochene and deoxy-capsidiol hydroxylating activities for sesquiterpene biosynthesis (33, 50). The discrepancy between the induction of cyclase activity and capsidiol accumulation was even greater for the MeJAtreated cells. Within 24 h of the MeJA treatment, cyclase activity was induced to 60% of its maximum although little capsidiol accumulated during this period. However, maximal accumulation of capsidiol occurred 48 to 72 h after MeJA addition and at a time when cyclase activity was also at its highest level. This comparison, like that for the elicitor treatment data, suggested the necessary involvement of other activities like aristolochene and deoxy-capsidiol hydroxylating activities and those responsible for capsidiol turnover in controlling the accumulation and stability of this compound in the cell cultures. Hoshino et al. (50) and Threlfall and Whitehead (33) have previously described 5-epi-aristolochene and deoxy-capsidiol hydroxylating activities and provided evidence for the involvement of these activities in the

290

´ VEZ ET AL. MANDUJANO-CHA

biosynthesis of capsidiol. To examine the hydroxylase activities in tobacco cells, a modified form of these assays was developed in which 3H-labeled 5-epi-aristolochene was fed to cells for a 3-h period before determining the amount of label converted to labeled capsidiol. Unlike the cyclase enzyme activity (Fig. 1B), the in vivo hydroxylase activities following cellulase treatment were transiently induced with a maximum occurring 12 h after elicitor treatment (Fig. 1C). The fold induction of this activity was minimally 30-fold within the first 12 h, followed by a 3- to 4-fold decrease in activity over the next 12-h period. In contrast, the aristolochene to capsidiol hydroxylating activities in the MeJA-treated cells were slowly induced over the entire time course of 72 h, resulting in an approximate 10-fold induction of these activities. Molecular regulation of sesquiterpene cyclase gene expression in MeJA and elicitor-treated cells. Previous work not only correlated the acumulation of capsidiol in elicitor-treated cells with the induction of sesquiterpene cyclase enzyme activity (45), but several lines of evidence demonstrated that the induction of the cyclase activity was primarily controlled by the relative transcription rate of the cyclase gene (39, 44). To determine if the MeJA induction of cyclase activity were similarly regulated, the absolute levels of the cyclase protein and mRNA in MeJA-treated cells were measured and compared to those in elicitor-treated cells. As reported earlier (44), the cyclase polypeptide was absent from control cells as measured by immunodetection (Fig. 2A). Slight amounts of the polypeptide were detected at 6 h after elicitor treatment, increased to a maximum by 36 h, and remained at those elevated levels for the next 48 h. This accumulation of the cyclase protein directly paralleled the induction of cyclase enzyme activity (compare with Fig. 1B), suggesting that little if any posttranslational control of the cyclase enzyme activity occurred in elicitor-treated cells. The cyclase mRNA, measured in RNA blot hybridizations, was also undetectable in control cell cultures but transiently induced upon elicitor treatment (Fig. 2B). The cyclase mRNA level increased strongly within 6 h of elicitor treatment, reached a maximum by 12 to 24 h, and then decreased rapidly over the next 12 h to levels 10- to 20-fold lower than the maximum. The cyclase mRNA levels returned to near control levels by 48 h after the inductive treatment. Important to note, the induction profile of the cyclase mRNA, with its maximum 12 to 24 h after elicitor treatment, corresponded to that time when the accumulation rate of the cyclase protein (Fig. 2A) and enzyme activity were greatest (Fig. 1B). MeJA treatment induced very low and difficult to detect levels of the cyclase protein within the first 12 h of treatment (Fig. 2C). However, significant levels of

FIG. 2. Induction time course for sesquiterpene cyclase protein and mRNA in elicitor (cellulase)- and MeJA-treated cells. Total protein extracts prepared from cells at the indicated times of the respective induction treatments were size separated by SDS–PAGE, transferred to nitrocellulose membranes and the cyclase protein immunodetected using a 5-epi-aristolochene synthase monoclonal antibody (45) (A and C). Total RNA was extracted from an independent set of cell samples, size fractionated by agarose gel electrophoresis under denaturing conditions, and transferred to a nylon membrane before probing with the full-length EAS4 cDNA (B and D). Equal loading of the RNA samples was confirmed by hybridization with a 28S rRNA probe (data not shown).

the cyclase polypeptide were observed by 24 h and reached a maximum and relatively stable level by 36 h. Like the correlation noted above for elicitor-treated cells, the level of cyclase protein in the MeJA treated cells corresponded quite closely with the induction profile of enzyme activity (compare Fig. 2C with Fig. 1B), consistent with no posttranslational control of the cyclase enzyme activity in the MeJA-treated cells. Detectable levels of cyclase mRNA were measured within 6 to 9 h after MeJA treatment and rose to a maximum by 36 h after treatment (Fig. 2D). Unlike the transient induction in the elicitor-treated cells (Fig. 2B), the level of the cyclase mRNA in the MeJA-induced cells remained relatively constant for the remainder of the experimental time. Posttranscriptional control of the cyclase mRNA. The above experimental data were suggestive of some sort of posttranscriptional control of the cyclase activity in the MeJA-treated cells. Assuming that the cyclase mRNA present at 36 to 72 h of MeJA treatment is translationally competent, synthesis and accumulation of the cyclase protein and enzyme activity would be

INDUCTION OF SESQUITERPENE METABOLISM IN TOBACCO CELL SUSPENSIONS

FIG. 3. Viability of control tobacco cells (■) and cells treated with fungal elicitin (330 nM)(F), cellulase (0.5 ␮g/ml) (䊐), or MeJA (0.5 mM) (Œ). Cells incubated with the indicated additions were collected at the designated time points and assayed for their ability to exclude the Evans Blue dye. Dye exclusion is a measure of cell viability.

expected during this time period. This was not observed (compare Fig. 2D with Figs. 1B and 2C). One possible explanation for such an observation would be that the cells were simply metabolically compromised at these later times of MeJA treatment. Several reports have suggested that MeJA might in fact induce programmed cell death (51), a response which could distort interpretations concerning any mRNA species or level. The ability of MeJA to induce cell death was therefore determined using an Evans blue exclusion assay (43) (Fig. 3). Consistent with results of Ricci et al. (52), cell death was induced by the addition of 330 nM elicitin protein to the cultures. However, the Evans blue dye was excluded from control cell cultures as well as those receiving concentrations of MeJA up to 500 ␮M, indicating that the MeJAtreated cells were as viable as the control cells. The discrepancy between the high levels of cyclase mRNA without obvious accumulation of the cyclase protein at the latter times of MeJA treatment could also be due to a differential recruitment of the cyclase mRNA into polysomes for protein synthesis. To assess this possibility, total and polysomal RNAs were extracted from cells induced with cellulase or MeJA for 24 h and the level of cyclase mRNA in each determined by hybridization with a full-length cyclase cDNA (Fig. 4). While the level of cyclase RNA associated with polysomal RNA from elicitor-treated cells was 42% of that detected in the total RNA sample (lane 5 relative to lane 2), less than 12% of the cyclase mRNA in the total RNA sample from the MeJA-treated cells was found associated with polysomal RNA (lane 6 relative to lane 3). MeJA induces expression of multiple sesquiterpene cyclase genes. Jasmonates are known to influence gene expression at many different levels including transcription, RNA processing, transcript stability and translation efficiency (53). For example, MeJA induces

291

a 35 base shift in the 5⬘ end of the mRNA coding for the large subunit of ribulose-bisphosphate carboxylase/oxygenase (RUBISCO) which is correlated with a decrease in the translatability of this new mRNA species (54). The 35-base elongation at the 5⬘ end of this mRNA is thought to create additional secondary structural features that hinder the translatability of this mRNA. The results in Fig. 4 also suggested that MeJA might induce a cyclase mRNA that might not be actively engaged in the synthesis of additional cyclase protein. To begin assessing the nature of the cyclase mRNA(s) induced by MeJA, a RT-PCR strategy was employed. The rationale behind this strategy was that of the approximate 12–16 different EAS-like genes found in the tobacco genome, we have the sequence for 8 of these genes (Newman, Schoenbeck and Chappell, unpublished) and have found them to be highly conserved within the coding domains, especially within exons 2 and 3. Moreover, these genes can be differentiated from one another only by short stretches of noncoding sequences found within the putative noncoding 3⬘ regions that appear in the respective mRNA transcripts. Using an oligo(dT) primer for first-strand cDNA synthesis as the reverse primer in PCRs and nondegenerate primers to conserved regions within exon 2 and 3 as forward primers, two unique cDNA fragments were observed, cloned, and the respective nucleotide sequences determined (Fig. 5). The amino acid sequences predicted from the open-reading frames of these two cDNA clones unk1 and unk2 were nearly identical, with unk1 matching the sequence previously published for EAS4 (Fig. 5B). Comparison of the 3⬘ nontranslated region of these genes also indicated that unk1 was identical to EAS4, but the cDNA sequence for unk2 was distinctly different from EAS4 (Fig. 5A) and subsequently designated EAS-MeJA cDNA. DISCUSSION

The results presented here do not support the notion that MeJA is the key regulatory element in an elicitorinduced signal transduction cascade leading to phytoa-

FIG. 4. Differential association of the sesquiterpene cyclase mRNA with polysomal RNA in elicitor- or MeJA-treated cells. Total (lanes 1, 2, and 3) and polysomal (lanes 5 and 6) RNAs were isolated from control cells (lane 1) and cells treated with cellulase (0.5 ␮g/ml) (lanes 2 and 5) or MeJA (0.5 mM) (lanes 3 and 6) for 24 h. Equal amounts of the RNA samples (10 ␮g) were size-fractionated by agarose gel electrophoresis under denaturing conditions, transferred to nylon membranes, and probed with the full-length EAS4 cDNA. Molecular weight standard RNAs were run in lane 4. Equal loading of the RNA samples was confirmed by hybridization with a 28S rRNA probe (data not shown).

292

´ VEZ ET AL. MANDUJANO-CHA

FIG. 5. DNA and deduced amino acid sequence comparisons between the 3⬘ nontranslated region and the partial coding regions for two sesquiterpene cyclase cDNAs isolated from MeJA-treated cells. Total RNA isolated from MeJA-treated cells was amplified in reverse transcription-polymerase chain reactions using an oligo(dT) reverse primer and specific forward primers corresponding to highly conserved domains within the second and third exons of all sesquiterpenes genes isolated from tobacco to date. The 3⬘-NTR of EAS4 starts at nt 1651 with the TGA stop codon indicated in bold (A). Identical nucleotides are indicated by dashes, missing nucleotides indicated by a slash mark, and the site of nucleotide insertions by vertical ( |) lines. Alignment of the amino acid sequences deduced from the EAS4 and unknown cDNAs (B). Dash lines denote identity between the two sequences while small letters indicate differences and a missing amino acid is indicated by a slash mark. The stop codon (*) is also indicated. Numbering for both nucleotide and amino sequences is according to that previously determined for the full-length EAS4 cDNA GenBank Accession No. L04680.

lexin biosynthesis in tobacco cell suspension cultures. Although we observed an accumulation of sesquiterpene phytoalexins in MeJA treated cell cultures, the time course for this accumulation as well as for the respective biosynthetic enzymes were significantly delayed relative to those observed in elicitor-treated cells (Fig. 1). If MeJA were the key regulatory component, then MeJA would be expected to induce time course changes superimposable or more rapid than those caused by elicitor-treatment. Our results, however,

cannot rule out an accessory or component role of jasmonates in the elicitor-induced responses. Rapid, transient induction of jasmonate accumulation in elicitorand pathogen-challenged tobacco tissues has been documented (51, 56), and the current results demonstrate that MeJA can induce phytoalexin biosynthesis in the tobacco cell suspensions. One alternative is that elicitor treatment induces a transient accumulation of jasmonates along with other signal molecules which work together to evoke the rapid induction of phytoalexin biosynthesis. The accumulation of momilactone A, a diterpene phytoalexin, in elicitor-treated rice cell suspension cultures features some of these characteristics (17). JA alone induced a slower accumulation of momilactone A than did elicitor treatment, resulting in the accumulation of 20% as much phytoalexin. However, momilactone A accumulation was dramatically reduced in elicitor-treated cells by the addition of inhibitors of JA biosynthesis, and this inhibition was relieved by the exogenous application of jasmonate. Our attempts to assess a component role of MeJA in elicitor-induced phytoalexin accumulation in tobacco cell cultures in a similar way have been inconclusive and equivocal (data not shown). For example, preincubation of tobacco cells with n-propyl gallate (57), a putative inhibitor of jasmonate biosynthesis, did reduce the induction of sesquiterpene cyclase activity in response to elicitor treatment. DIECA (28), another inhibitor of jasmonate biosynthesis, however, did not. More critical, the inhibition of elicitor-inducible sesquiterpene cyclase activity by n-propyl gallate could not be restored by exogenous application of MeJA. Although our general impression of these preliminary findings is that MeJA may not mediate the induction of phytoalexin biosynthesis in elicitor-treated tobacco cell cultures, this interpretation is tempered by issues of inhibitor specificity and questions about their potential for secondary sites of action. Rickauer et al. (56) came to a similar conclusion about the lack of MeJA’s role in the phytoalexin elicitation process in tobacco, but with results quite different from ours. These investigators reported that while mRNAs for proteinase inhibitor II, chitinase, glucanase, and hydroxyproline-rich glycoprotein were induced by 890 ␮M MeJA, sesquiterpene cyclase mRNA was not. Our results suggest that although the cyclase mRNAs might be difficult to detect within 8 h of MeJA treatment, a clear induction of this mRNA was evident by 24 h (Fig. 2D). The induction of cyclase mRNA was also corroborated in the current experiments by a corresponding induction of cyclase enzyme activity and the cyclase polypeptide within this time period (Figs. 1B and 2C). Although the differences between the results of Rickauer et al. (56) and ourselves may reflect differences in the plant cultivars used (Wisconsin 38 versus KY14), under the condition of the cultures used experimentally (cells in stationary phase versus rapid

INDUCTION OF SESQUITERPENE METABOLISM IN TOBACCO CELL SUSPENSIONS

growth) and the concentration of MeJA used (890 ␮M versus 500 ␮M), these differences seem minor and not sufficient to account for the induction of cyclase mRNA in our experiments but not in the work of Rickauer et al. (56). More direct comparisons using the two different cell cultures and identical inductive treatments will be needed to resolve these discrepancies. The current results provide support for the differential expression of phytoalexin biosynthetic genes in response to elicitor and MeJA and perhaps the differential expression of additional genes coding for isoprenoid biosynthetic activities. While elicitor treatment induced a rapid accumulation of capsidiol, phytoalexin accumulation in MeJA-treated cells occurred relatively late. Correspondingly, elicitor treatment induced a rapid and transient induction of sesquiterpene cyclase and hydroxylase activities, while MeJA induced a more protracted induction time course of cyclase activity and only a low level of hydroxylase activity. Also, while both treatments induced sesquiterpene gene expression, MeJA induced the expression of at least one cyclase gene not previously documented in elicitor-induced mRNA. Based on sequence alignment with the 5-epi-aristolochene synthase amino acid sequence and our current understanding of those residues participating directly in catalysis (55), the EAS-MeJA mRNA appears to encode for a typical aristolochene synthase activity like the EAS4 mRNA. In subsequent sequencing efforts, one cDNA corresponding to the EAS-MeJA sequence has been observed in a cDNA library prepared against mRNA from elicitor-treated cells (data not shown). We do not, however, have a measure of how this particular mRNA contributes to the overall synthesis of the sesquiterpene cyclase protein in MeJAtreated cells, nor what contribution the EAS-MeJA mRNA makes to the total cyclase mRNA induced by MeJA. The results presented in Fig. 4 do suggest that there is approximately four times less cyclase mRNA engaged in polysomes of MeJA-treated cells than for elicitor induced cells. Determining whether the EASMeJA mRNA represents a unique cyclase gene subject to regulation by various inducers or only MeJA, and whether there are sequences within either EAS4 or EAS-MeJA mRNA that effect their translatability like that observed for the large subunit of RUBISCO (54) will require additional molecular analysis. This might include sequence comparisons between the full-length cDNAs from MeJA-treated cells and the corresponding genomic clone(s), and tests relying on the development of transgenic lines containing reporter gene constructs composed of the EAS-MeJA promoter and various portions of the EAS-MeJA and EAS4 5⬘ NTR leader sequences. Jasmonates have been shown to effectively induce the accumulation of a wide range of plant secondary compounds within and outside the context of plant– pathogen interactions. The list of compounds accumu-

293

lating in cells responding to jasmonates include isoprenoids (16, 23, 29), alkaloids (9, 28) and phenylpropanoid derivatives (12). The ability of jasmonates to induce isoprenoids has, however, been accompanied by several intriguing observations. Using potato tuber discs, Choi et al. (23) reported that low concentrations of MeJA act synergistically with wounding to induce steroidal glycoalkaloid accumulation while high concentrations of MeJA inhibited this accumulation. MeJA alone did not induce sesquiterpene phytoalexin accumulation in the wounded tuber discs and therefore appears to moderate a wound response rather than an elicitor- or pathogen-induced defense response in potato. Singh et al. (29) reported an analogous finding with Hyocyamus muticus root cultures, except that the root cultures responded synergistically to MeJA and wounding by the accumulation of the sesquiterpene phytoalexin solavetivone. Elicitor-treated H. muticus roots accumulated lubimin, a sesquiterpene phytoalexin derived from solavetivone (30). These latter observations by Singh et al. (29) suggest that although MeJA can induce isoprenoid biosynthetic activities up to the initial cyclization of farnesyl diphosphate, the ability of MeJA to induce latter steps in the sesquiterpene biosynthetic pathway are rather limited. Similar to the differential accumulation of solavetivone and lubminin in MeJA- and elicitor-treated H. muticus roots (29), the modest accumulation of capsidiol in the MeJA-treated tobacco cells appears to be associated with a low level induction of the aristolochene/deoxycapsidiol hydroxylase activities (Fig. 1C). A logical conclusion from these observations is that MeJA has a limited ability to induce the expression of the corresponding hydroxylase gene(s) or other downstream catalytic steps in sesquiterpene metabolism. Such a suggestion can easily be tested as probes for the corresponding genes become available. ACKNOWLEDGMENTS This work was supported by a grant from the National Science Foundation and the Kentucky Agricultural Experiment Station. A.M.C. is recipient of a Doctoral fellowship (No. 96382) from CONACyT (Me´xico).

REFERENCES 1. McConn, M., Creelman, R. A., Bell, E., Mullet, J. E., and Browse, J. (1997) Proc. Natl. Acad. Sci. USA 94, 5473–5477. 2. Staswick, P. E., Yuen, G. Y., and Lehman, C. C. (1998) Plant J. 15, 747–754. 3. Thomma, B. P. H. J., Eggermont, K., Penninckx, I. A. M. A., Mauch-Mani, B., Vogelsang, R., Cammue, B. P. A., and Broekaert, W. F. (1998) Proc. Natl. Acad. Sci. USA 95, 15107–15111. 4. Creelman, R. A., and Mullet, J. E. (1997) Annu. Rev. Plant Physiol. Plant Mol. Biol. 48, 355–381. 5. Farmer, E. E., and Ryan C. A. (1992) Plant Cell 4, 129 –134. 6. Farmer, E. E., Johnson, R. R., and Ryan, C. A. (1992) Plant Physiol. 98, 995–1002.

294

´ VEZ ET AL. MANDUJANO-CHA

7. Mueller, M. J., Brodschelm, W., Spannagl, E., and Zenk, M. H. (1993) Proc. Natl. Acad. Sci. USA 90, 7490 –7494. 8. Gundlach, H., Mu¨ller, M. J., Kutchan, T., and Zenk, M. H. (1992) Proc. Natl. Acad. Sci. USA 89, 2389 –2393. 9. Blechert, S., Brodschelm, W., Ho¨lder, S., Kammerer, L., Kutchan, T. M., Mueller, M. J., Xia, Z., and Zenk, M. H. (1995) Proc. Natl. Acad. Sci. USA 92, 4099 – 4105. 10. Creelman, R. A., Tierney, M. L., and Mullet, J. E. (1992) Proc. Natl. Acad. Sci. USA 89, 4938 – 4941. 11. Howe, G.A., Lightner, J., Browse, J., and Ryan, C. A. (1996) Plant Cell 8, 2067–2077. 12. Ellard-Ivey, M., and Douglas, C. J. (1996) Plant Physiol. 112, 183–192. 13. Horvath, D. M., and Chua, N. H. (1996) Plant Mol. Biol. 31, 1061–1072. 14. Suzuki, G., Ohta, H., Kato, T., Igarashi, T., Sakai, F., Shibata, D., Takano, A., Masuda, T., Shioi, Y., and Takamiya, K. (1996) FEBS Lett. 383, 83– 86. 15. Yukimune, Y., Tabata, H., Higashi, Y., and Hara, Y. (1996) Nature Biotech. 14, 1129 –1132. 16. Hefner, J., Ketchum, R. E. B., and Croteau, R. (1998) Arch. Biochem. Biophys. 360, 62–74. 17. Nojiri, H., Sugimori, M., Yamane, H., Nishimura, Y., Yamada, A., Shibuya, N., Kodama, O., Murofushi, N., and Omori, T. (1996) Plant Physiol. 110, 387–392. 18. Tamogami, S., Rakwal, R., and Kodama, O. (1997) FEBS Lett. 412, 61– 64. 19. Mu¨hlenweg, A., Melzer, M., Li, S., and Heide, L. (1998) Planta 205, 407– 413. 20. Kuroyanagi, M., Arakawa, T., Mikami, Y., Yoshida, K., Kawahar, N., Hayashi, T., and Ishimaru, H. (1998) J. Nat. Prod. 61, 1516 –1519. 21. Vijayan, P., Shockey, J., Le´vesque, C. A., Cook, R. J., Browse, J. (1998) Proc. Natl. Acad. Sci. USA 95, 7209 –7214. 22. Menke, F. L. H., Champion, A., Kijne, J. W., and Memelink, J. (1999) EMBO J. 18, 4455– 4463. 23. Choi, D., Bostock, R. M., Avdiushko, S., and Hildebrand D. F. (1994) Proc. Natl. Acad. Sci. USA 91, 2329 –2333. 24. Maldonado-Mendoza, I. E., Vincent, R. M., and Nessler, C. L. (1997) Plant Mol. Biol. 34, 781–790. 25. Chappell, J. (1995) Annu. Rev. Plant Physiol. Plant Mol. Biol. 46, 521–547. 26. Aerts, R. J., Gisi, D., Carolis, E. D., Luca, V., Baumann, T. W. (1994) Plant J. 5, 635– 643. 27. Gantet P., Imbault, N., Thiersault, M., and Doireau, P., (1998) Plant Cell Physiol. 39, 220 –225. 28. Menke, F. L. H., Parchmann, S., Mueller, M. J., Kijne, J. W., and Memelink, J. (1999) Plant Physiol. 119, 1289 –1296. 29. Singh, G., Gavrieli, J., Oakey, J. S., and Curtis, W. R. (1998) Plant Cell Reports 17, 391–395. 30. Stoessl, A., Stothers, J. B., and Ward, E. W. B. (1976) Phytochemistry 15, 855– 872. 31. Ku´c, J. (1995) Annu. Rev. Phytopathol. 35, 275–297.

32. Keller, H., Czernic, P., Ponchet, M., Ducrot, P. H., Back, K., Chappell, J., Ricci, P., and Marco, Y. (1998) Planta 205, 467– 476. 33. Threlfall, D. R., and Whitehead, I. M. (1988) Phytochemistry 27, 2567–2580. 34. Chappell, J., Nable, R., Fleming, P., Andersen, R. A., and Burton, H. R. (1987) Phytochemistry 26, 2259 –2260. 35. Back, K., He, S., Kim, K. U., and Shin, D. H. (1998) Plant Cell Physiol. 39, 899 –904. 36. Hoshino, T., Chida, M., Yamaura, T., Yoshizawa, Y., and Mizutani, J. (1994) Phytochemistry 36, 1417–1419. 37. Vo¨geli, U., and Chappell, J. (1988) Plant Physiol. 88, 1291–1296. 38. Facchini, P. J., and Chappell, J. (1992) Proc. Natl. Acad. Sci. USA 89, 11088 –11092. 39. Yin, S., Leng, M., Newman, J., Back, K., and Chappell, J. (1997) Plant Physiol. 115, 437– 451. 40. Chappell, J., and Nable, R. (1987) Plant Physiol. 85, 469 – 473. 41. O’Donohue, M. J., Gousseau, H., Huet, J-C., Tepfer, D., and Pernollet, J-C. (1995) Plant Mol. Biol. 27, 577–586. 42. Ward E. W. B., Unwin, C. H., and Stoessl, A. (1974) Can. J. Bot. 52, 2481–2488. 43. Glazener, J. A., Orlandi, E. W., and Baker, C. J. (1996) Plant Physiol. 110, 759 –763. 44. Vo¨geli, U., and Chappell, J. (1990) Plant Physiol. 94, 1860 –1866. 45. Vo¨geli, U., Freeman, J. W., and Chappell, J. (1990) Plant Physiol. 93, 182–187. 46. Back, K., Yin, S., and Chappell, J. (1994) Arch. Biochem. Biophys. 315, 527–532. 47. Rising, K. A., Starks, C. M., Noel, J. P., and Chappell, J. (2000) J. Am. Chem. Soc. 122, 1861–1866. 48. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. 49. Guedes, M. E. M., Ku´c, J., Hammerschmidt, R., and Bostock, R. (1992) Phytochemistry 21, 2987–2988. 50. Hoshino, T., Yamaura, T., Imaishi, H., Chida, M., Yoshizawa, Y., Highashi, K., Ohkawa, H., and Mizutani, J. (1995) Phytochemistry 38, 609 – 613. 51. Kenton, P., Mur, L. A. J., Atzorn, R., Wasternack, C., and Draper, J. (1999) Molecular Plant-Microbe Interact. 12, 74 –78. 52. Ricci. P., Trentin, F., Bonnet, P., Vernard, P., Mouton-Perronnet, F., and Bruneteau, M. (1992) Plant Pathol. 41, 298 –307. 53. Reinbothe, S., Mollenhauer, B., and Reinbothe, C. (1994) Plant Cell 6, 1197–1209. 54. Reinbothe, S., Reinbothe, C., Heintzen, C., Seidenbecher, C., and Parthier, B. (1993) EMBO J. 12, 1505–1512. 55. Starks, C. M., Back, K., Chappell, J., and Noel, J. P. (1997) Science 277, 1815–1820. 56. Rickauer, M., Brodschelm, W., Bottin, A., Ve´rone´si, C., Grimal, H., and Esquerre´-Tugaye´, M. T. (1997) Planta 202, 155–162. 57. Pen˜a-Corte´s, H., Albrecht, T., Prat, S., Weiler, E. W., and Willmitzer, L. (1993) Planta 191, 123–128.