Plant defense in the absence of jasmonic acid: The role of cyclopentenones

Abstract

The Arabidopsis opr3 mutant is defective in the isoform of 12-oxo-phytodienoate (OPDA) reductase required for jasmonic acid (JA) biosynthesis. Oxylipin signatures of wounded opr3 leaves revealed the absence of detectable 3R,7S-JA as well as altered levels of its cyclopentenone precursors OPDA and dinor OPDA. In contrast to JA-insensitive coi1 plants and to the fad3 fad7 fad8 mutant lacking the fatty acid precursors of JA synthesis, opr3 plants exhibited strong resistance to the dipteran Bradysia impatiens and the fungus Alternaria brassicicola. Analysis of transcript profiles in opr3 showed the wound induction of genes previously known to be JA-dependent, suggesting that cyclopentenones could fulfill some JA roles in vivo. Treating opr3 plants with exogenous OPDA powerfully up-regulated several genes and disclosed two distinct downstream signal pathways, one through COI1, the other via an electrophile effect of the cyclopentenones. We conclude that the jasmonate family cyclopentenone OPDA (most likely together with dinor OPDA) regulates gene expression in concert with JA to fine-tune the expression of defense genes. More generally, resistance to insect and fungal attack can be observed in the absence of JA.

A major objective in plant biology is to develop an integrated understanding of how plants survive in their environment and reproduce. Although it has become clear in the last decade that jasmonic acid (JA) is a key regulator in the development, physiology, and defense of plants, the complexity of the signaling network in which JA evolves is just emerging (1). JA is involved in carbon partitioning (2), in mechanotransduction (3), and the ability of plants to synthesize and perceive JA is absolutely essential for the correct development and release of pollen in Arabidopsis (4–7). Highlighting the regulatory importance of JA, a JA-responsive transcription factor, ORCA3, first found in Catharanthus, provides an important link between primary and secondary metabolism (8). There is also strong evidence supporting a central role of JA in plant defense. Exogenous JA powerfully regulates the expression of many defense genes in plants, and its in vivo production and perception seem to be of vital importance in mounting successful defense against insect attackers (9–11). Together with ethylene, JA also plays a crucial role in defense against necrotrophic fungi (12–14) and in induced systemic resistance in response to nonpathogenic rhizobacteria (15). Broader roles of JA in plant stress responses are likely; it is known that the JA biosynthesis pathway is important in gene activation subsequent to UV damage in plants (16), and JA has been implicated in some responses to water stress (17).

The biosynthesis of JA occurs through the octadecanoid pathway (18, 19) and is initiated by the addition of molecular oxygen to linolenic acid (18:3) to form 13-hydroperoxylinolenic acid (13-HPOTrE). This fatty acid hydroperoxide is then dehydrated by allene oxide synthase (AOS) and cyclized by allene oxide cyclase (AOC) to the cyclopentenone 12-oxo-phytodienoic acid (OPDA). Although the chemical nature of OPDA allows four stereoenantiomers, the concerted action of AOS and AOC generate exclusively 9S,13S-OPDA (20), the precursor of active 3R,7S-JA (21). The next step in the formation of JA is the reduction of the pentacyclic ring double bond in 9S,13S-OPDA by the enzyme OPDA reductase 3 (OPR3; ref. 22) to the cyclopentanone 3-oxo-2(2′[Z]-pentenyl)-cyclopentane-1-octanoic acid (OPC:8). Other related enzymes such as OPR1 and OPR2, initially thought to be involved in the biosynthesis of active 3R,7S-JA, have almost no activity against 9S,13S-OPDA (22). Finally, OPC:8 is shortened by three cycles of β-oxidation to yield JA. In parallel, hexadecatrienoic acid (16:3) can be metabolized to dinor oxo-phytodienoic acid (dnOPDA; ref. 23), the 16-carbon structural homolog of OPDA.

Although attention has focused mainly on JA as a signal, the possibility that the JA precursor OPDA could itself be biologically active has been proposed (3, 23–25). However, there is also strong evidence, based on detailed pharmacological studies, that the natural oxylipin capable of eliciting the synthesis of alkaloids like sanguinarine in Eschscholtzia cell cultures is JA (26). Until recently, no tools have been available to perform genetic tests with which to resolve which jasmonate family member(s) is active in vivo. The availability of Arabidopsis plants lacking a functional OPR3 gene (opr3, ref. 7; dde1, ref. 6) has now permitted the genetic separation of OPDA and JA effects in vivo. In the opr3 mutant JA cannot be produced; these plants are male sterile because they produce predominantly inviable pollen and have delayed anther dehiscence. Male sterility in these plants is rescued efficiently by JA but not by OPDA, ruling out an active role for OPDA in pollen development (6, 7). Thus evidence based on well characterized mutants indicates an exclusive role of JA in male gametophyte development in Arabidopsis.

This study addressed the question: Is JA the sole octadecanoid-derived effector necessary for defense in plants? By using the opr3 mutant we were able to investigate the effects of OPDA on gene expression without having to consider OPDA conversion to JA. The results demonstrate that OPDA is itself a potent gene regulator in the wound response and that protection of plants against the attack of an insect or a fungal pathogen is obtained in the absence of JA.

Materials and Methods

Plants, Insects, and Pathogens.

Arabidopsis thaliana were grown under a 9-h [120 μE (1 mol of photons) m⁻² s⁻¹]/15-h light/dark photoperiod. Wild-type (ecotype Wassilewskija, WS), opr3 (in the WS background), coi1–1, and fad3–2 fad7–2 fad8 mutant plants were grown in pasteurized soil (12). Mutants coi1 and fad3 fad7 fad8 are in a Columbia background (Col-0); this ecotype is known to be resistant to attack by Bradysia impatiens larvae (10). B. impatiens (Diptera; Sciaridae) were maintained on wild-type Arabidopsis and, for experiments, adult flies were harvested and used to challenge plants as described in Fig. 1. Alternaria brassicicola (MUCL 20297) was from the Mycothèque Université Catholique de Louvain (Belgium) and was maintained on potato dextrose agar at 26°C. Conidia/conidiospores were collected in distilled water 10 days after subculture, and infection was performed by applying droplets (5 μl) of an aqueous solution containing 5 × 10⁵ spores⋅ml⁻¹ of A. brassicicola onto leaves of four- to six-leaf-stage plants.

Figure 1.

opr3 mutant plants are not susceptible to attack by Bradysia. Results of two independent experiments in which a mixed population of approximately 150 wild-type (WT), fad3 fad7 fad8 triple mutant (tm), coi1, and opr3 plants were grown in a net enclosure and subjected to heavy infestation with B. impatiens larvae. As the seedlings reached the four-leaf stage, adult Bradysia flies (average 50–80) were introduced in the net every other day over a 2-week period. Damage to the petiole or to the roots of the plants by Bradysia larvae resulted in wilting followed by death of the plants. After bolting, the percentage of surviving plants for each genotype was calculated.

cDNA Microarray Analysis and RNA Gel-Blot Analyses.

Established cDNA microarray protocols were used (27). Full details of the cDNA microarray used as well as expressed sequence tags (EST) accession nos. are given in Reymond et al. (28), at www.unil.ch/ibpv, and in Table 1, which is published as supporting information on the PNAS web site, www.pnas.org). Based on previous results (28), the following genes were used as negative controls in wounding experiments: plastocyanin, α-TUB, β-TUB, EF1, and ACT. Wound-inducible positive control genes were JR3 and TSA (COI1-dependent) and GST1, XERO2, and RNS1 (COI1-independent; ref. 28). A 2-fold change in transcript level was regarded as significant (28); however, for key interpretations, genes for which larger changes in transcript level occurred were selected. In brief, mRNA samples (2 μg) from wild-type or opr3 Arabidopsis leaves were converted into fluorescently labeled single-stranded cDNAs by direct incorporation of either Cy3-dCTP (control) or Cy5-dCTP (wounded or treated with either JA or OPDA). These cDNA populations were hybridized simultaneously to cDNA microarrays, and the fluorescent signal for both fluors was recorded and quantitated. Two independent experiments for each wounding time point and three for treatment of opr3 with OPDA or JA were carried out. Cluster analyses (29) were used to display data from time courses or to compare individual experiments. RNA gel-blot analyses were conducted as described (30). For PDF1.2 gel-blot analyses, EST accession no. T04323 was used. The loading control probe was for the Lhb1B2 gene (EST accession no. R89981), encoding chlorophyll a/b-binding protein. Both probes were labeled with digoxigenin (Roche Molecular Biochemicals) by PCR amplification.

Oxylipin Signatures and Treatments.

Oxylipins from Arabidopsis leaves (1 g) were extracted according to Weber et al. (23). Approximately 350 nmol⋅g⁻¹ tissue of OPDA (Cayman Chemicals, Ann Arbor, MI) and JA (Sigma) were sprayed onto plants in an acetone carrier (30), and leaves were harvested after 4.5 h. Control plants were sprayed only with carrier.

Results

opr3 Plants Resist Attack by an Insect and a Fungal Pathogen.

The previous characterization of the Arabidopsis opr3 mutant indicated that JA is the signaling molecule required to induce pollen maturation and release; OPDA could not substitute for JA in this role (7). Several lines of evidence suggest that JA is also a physiological signal for activation of genes involved in the defense of plants against insect attack (9, 10) and infection by some fungal pathogens (12–14). In particular, a triple mutant that lacks the fatty acid precursors of JA synthesis, fad3 fad7 fad8, and the JA-insensitive mutant, coi1, have both been shown to be extremely susceptible to insect attack (ref. 10 and this study). Moreover, almost wild-type level of protection is provided to the fad3 fad7 fad8 mutant by exogenous JA (10, 12). On the basis of these observations, it was expected that the opr3 mutant would also be defective in defense responses. However, casual observation suggested that opr3 plants were not susceptible to insect and pathogen damage. To further investigate this observation, we conducted two independent experiments in which wild-type, fad3 fad7 fad8, coi1, and opr3 plants were grown in mixed stands and challenged with a heavy population of B. impatiens, the common fungal gnat, as they reached four-six leaves (Fig. 1). By the time the plants were flowering, only 4–6% of fad3 fad7 fad8 and 10–11% of coi1 plants had survived. Death occurred as a result of mechanical damage caused by Bradysia larvae chewing on the lower leaves, stems, petioles, and roots of the plants. In contrast, both wild-type and opr3 plants remained largely undamaged, with only one fatality in 146 and 158 plants, respectively. Because opr3 plants looked as healthy as the wild type, and because death rates in fad3 fad7 fad8 and in coi1 were substantial, it can be concluded that opr3 is resistant to Bradysia. To investigate the effects of the opr3 mutation on defense responses of Arabidopsis against a necrotrophic fungi, plants of the same four lines were infected with the fungal pathogen A. brassicicola (14). A. brassicicola caused leaf damage and necrosis of the inoculated leaves in all four lines. However, 3 weeks after infection, both wild-type and opr3 plants remained viable with more than 60% of their leaves being green and showing few or no symptoms. By contrast, over 90% of leaves on fad3 fad7 fad8 and on coi1 plants were dead or severely chlorotic (data not shown). These results show that opr3 plants retain resistance to fungal infection as well as to insect attack.

opr3 Mutant Plants Do Not Synthesize 3R,7S-JA.

Previous results indicated that, in the opr3/dde1 mutant, OPR3 transcript is absent in both reproductive and vegetative tissues (6, 7). The mutant also lacks detectable OPDA reductase activity against 9S,13S-OPDA (6), the main OPDA enantiomer in plants and the precursor of active 3R,7S-JA (21), and all available evidence points to an essential role for OPR3 in OPDA reduction (6, 7, 22). However, because opr3 resists attack, a possibility remains that an as yet uncharacterized OPR isoform is capable of replacing OPR3 in vegetative tissues and producing low levels of 3R,7S-JA. To test the possibility that opr3 vegetative tissues convert OPDA into 3R,7S-JA, the levels of three JA family members (OPDA, dnOPDA, and JA) were measured in nonwounded and wounded leaves. As shown in Fig. 2A, nonwounded leaves of wild-type Arabidopsis contained less than 0.2 nmol/g of fresh weight of JA estimated from both cis 3R,7S and trans 3R,7R-JA enantiomers, and the level increased to 6 nmol/g of fresh weight 3 h after wounding, before declining to an intermediate value after 6 h. By contrast, in opr3 JA levels were close to the detection limit of the mass spectrometric assay used, and increased from 0.03 nmol/g of fresh weight in nonwounded leaves to 0.2 nmol/g of fresh weight after wounding, which corresponds to less than 4% of the level present 3 h after wounding in wild type. Comparatively, the fad3 fad7 fad8 mutant accumulated 0.07 nmol/g of fresh weight of JA and 0.11 nmol/g of fresh weight of OPDA 90 min after wounding (not shown). OPDA and dnOPDA accumulated in response to wounding in both wild-type and opr3 leaves. The accumulation profiles were qualitatively similar in the two genotypes with the slight difference being that, in wounded opr3 leaves, both cyclopentenones reached maximum levels between 90–360 min, whereas in wild-type leaves, the levels of the compounds had partially subsided 360 min after wounding. Clear quantitative differences in cyclopentenone levels were observed. In wounded wild-type leaves, OPDA levels were at their maximum at around 5 nmol/g of fresh weight at both the 90- and 180-min time points. In contrast, at the same time points, OPDA levels reached half of this value (≈2.5 nmol/g of fresh weight) in wounded opr3 leaves. Similar behavior was noted for dnOPDA in wounded opr3 leaves. To investigate further the isomeric composition of the trace level of JA in opr3 leaves, selective ion monitoring of extracts produced 90 min after wounding (Fig. 2B) was effected. In addition to being at extremely low levels relative to wild type, JA in wounded opr3 leaves occurred only in the trans (3R,7R) form, whereas both the cis (3R,7S) and trans (3R,7R) stereoenantiomers were present in the wild type (Fig. 2B) and in fad3 fad7 fad8 (not shown).

Figure 2.

Accumulation of jasmonate family members in wounded Arabidopsis. (A) Oxylipins were extracted from leaves at different times after mechanically wounding the leaf apex. Kinetics of OPDA, dnOPDA, and JA accumulation were followed in unwounded (dashed lines), wounded (solid lines), wild-type (solid symbols), and opr3 (open symbols) plants. Data are the mean ±SE of three determinations. FW, fresh weight. (B) 3R,7S-JA is not detectable in the oxylipin signature of opr3 leaves. Selective ion monitoring for the ion m/z = 224 revealed two methyl jasmonate (MeJA) isomers, 3R,7R-MeJA and 3R,7S-MeJA, in wild-type (WT) Arabidopsis leaf extracts 90 min after wounding but only trace levels of the 3R,7R enantiomer in opr3 extracts.

Wound-Regulated Gene Expression in opr3 Mutant Leaves.

The opr3 mutant is resistant to Bradysia attack and Alternaria infection, whereas both fad3 fad7 fad8 and coi1 plants are highly susceptible. The implication of these observations is that, whereas JA is not necessary to provide full resistance in opr3, another signal molecule(s) must function to activate defense responses in the mutant. To test this hypothesis, the impact of the opr3 mutation on gene expression in wounded Arabidopsis leaves was assessed by using a dedicated cDNA microarray carrying about 150 defense-related genes. Samples were taken at 90, 180, and 360 min after wounding and, for each time point, at least two independent experiments were conducted. An average 2-fold increase in message level in at least 1 time point was taken as an indication of gene activation (28). Whereas the overall pattern of gene expression was very similar between wild-type and opr3 plants after wounding, we observed that gene expression dynamics differed in the two genotypes (see http://www.unil.ch/ibpv and Table 2, which is published as supporting information on the PNAS web site). For example, 90 min after wounding 44 genes were up-regulated in wild-type plants, whereas in opr3 plants, 38 genes were up-regulated. At 180 min after wounding, 42 and 18 genes were up-regulated in wild-type and opr3 plants, respectively. By 360 min, 13 genes were up-regulated in wild-type vs. 3 in opr3 plants. Fig. 3A shows a clustal analysis (29) of 24 representative genes in wounded wild-type and opr3 leaves. Two groups can be distinguished. First, the top section consists of a group of COI1-dependent genes. As reported (28), all of these genes were activated after wounding in wild-type plants. For example, the changes varied from a weak increase in transcript level (2.2-fold for ASB after 180 min) to a strong increase (15.1-fold for VSP after 180 min). Interestingly, in opr3 mutants, the clustal analysis revealed a gradient of induction ranging from genes for which no increase in transcript level was detected (MBP, JIP, VSP) through to genes that were equally activated in both opr3 and wild-type plants, e.g., COMT (90 min after wounding there was an 8.2-fold increase in transcript in opr3 and an 8.6-fold increase in transcript in wild-type leaves) and CCR (for which, 90 min after wounding, there was a 3.6-fold increase in transcript in opr3 and a 3.7-fold increase in transcript in wild type). In between these extremes were genes like JR3. For this gene, 90 min after wounding, transcript increased 2.7-fold in opr3 and 4.8-fold in wild-type leaves. Many other genes (e.g., CYP83B1, HPL, TSA) showed a significant but weaker activation in wounded opr3 leaves. Second, a group of COI1-independent genes (Fig. 3A Bottom) had a highly similar expression behavior in both wild-type and opr3 plants after wounding (e.g., RNS1, OPR1, GST1). Thus, these microarray results show that many genes known to be COI1-dependent were induced by wounding in a mutant lacking JA. As expected (28), several defense-related genes such as PDF1.2 and HEL were not wound-inducible in wild-type or opr3 genotypes (data not shown).

Figure 3.

DNA microarray analysis after mechanical wounding and application of oxylipins. Wound-inducible gene expression in wild-type (WT) and opr3 Arabidopsis plants 90, 180, and 360 min after mechanical wounding (A) was compared with gene expression after treatment of opr3 with OPDA or JA for 4.5 h. (B) A clustal analysis was performed (see Materials and Methods) on a set of genes of interest in this study. Each gene is represented by a single row of colored boxes, and each experiment is represented by a single column. Each experiment is the average of at least two replicates. All gene expression data can be found in Tables 1–3, which are published as supporting information on the PNAS web site, and at http://www.unil.ch/ibpv.

Because the opr3 mutant cannot convert OPDA into JA, this plant offers the possibility to compare the effects of exogenous JA and OPDA on gene expression. The possibility that exogenous OPDA could induce COI1-dependent gene expression in opr3 was then tested. For each treatment at least three independent experiments were conducted. As a positive control, we first verified that exogenous JA was able to induce many of the COI1-dependent genes in opr3 (Fig. 3B Top). Indeed, genes such as MBP, JIP, HPL, JR3, or VSP showed an induction ranging from 3.5-fold (HPL) to 14.1-fold (JIP). When opr3 plants were treated with OPDA, the same genes were induced, with the notable exception of VSP. Surprisingly, OPDA also induced three COI1-independent genes (RNS1, OPR1, and GST1), which in contrast were not up-regulated by JA (Fig. 3B Bottom). Furthermore, two more genes (GST8 and ELI3) were induced only by OPDA and not JA, although they were not significantly induced by wounding (see http://www.unil.ch/ibpv and Table 3, which is published as supporting information on the PNAS web site).

Having shown that the induction of COI1-dependent genes still occurred in wounded opr3 plants and that OPDA was potentially responsible for this induction, we examined whether JA production is necessary for the induction of defense genes after a fungal attack. Defense response against A. brassicicola is also mediated through the COI1 pathway, and expression of PDF1.2, which encodes a protein with demonstrated antifungal properties, is known to be powerfully up-regulated in a COI1-dependent manner in this pathosystem (14). PDF1.2 transcript levels in wild-type and opr3 plants infected with the fungus were monitored by using RNA gel-blot analysis. As shown in Fig. 4, the PDF1.2 transcript is detectable in infected wild-type and opr3 leaves but not in control leaves from either genotype, demonstrating that in this system too, JA is not required for induction of COI1-dependent gene expression.

Figure 4.

PDF1.2 transcript accumulation in wild-type (WT) and opr3 plants challenged with A. brassicicola. RNA gel-blot analysis was performed by using 15 μg of total RNA isolated 48 h after infection. PDF1.2 transcripts levels are shown in control tissue treated with water or inoculated with Alternaria as described (14). The Lhb1B2 gene encoding a chlorophyll a/b-binding protein was used as a loading control.

Discussion

Defense in the Absence of 3R,7S-JA.

The disruption of the JA biosynthetic pathway leads to the extreme susceptibility of plants to attack by some insects (9, 10) and necrotrophic fungi (refs. 12 and 13 and this study). However, which molecule(s) in the pathway need to be synthesized to mount an active defense is unknown. The opr3 mutant offers the unique possibility to test the effects of truncation of the biosynthetic pathway on plant survival from insect attack. Unlike the response of plants defective in the biosynthesis of all jasmonate precursors (fad 3 fad 7 fad 8) or of plants defective in JA perception (col 1), the opr3 plants, which lack the ability to make wild-type levels of JA, survived as well as wild-type plants in the face of attack by caged Bradysia larvae (Fig. 1). Therefore, in the case of defense against this insect, JA production seems unnecessary. This implies that another member of the jasmonate family can replace JA in triggering defense in opr3. The replacement molecules must function by means of the COI1 protein, because the coi1 mutant is susceptible to attack (Fig. 1). Another line of evidence comes from the observation that opr3 and wild-type plants are more resistant to the fungal pathogen A. brassicicola than coi1 and fad3 fad7 fad8 mutants. More generally, the results demonstrate that protection against an insect or a fungal pathogen, and COI1-dependent defense gene expression, can be mediated by OPDA in the absence of JA. This conclusion is in contrast to the case of pollen maturation and anther dehiscence where JA production is an obligate requirement (6, 7).

Jasmonate Family Dynamics in opr3.

Our analysis of the levels of two cyclopentenone jasmonate family members OPDA and dnOPDA, as well as the cyclopentanone JA (23), in wild-type and opr3 plants in response to mechanical wounding (Fig. 2) showed that the pools of these compounds are dynamic, and in the wild type, responded to wounding by a 6-, 13-, and 30-fold increase of the levels of OPDA, dnOPDA, and JA, respectively, 3 h after wounding (Fig. 2A). In striking contrast, the opr3 mutation prevented JA from accumulating after wounding, and the levels observed were slightly above detection limit of the method. The opr3 mutation also resulted in a quantitative alteration in cyclopentenone accumulation. Levels of OPDA and dnOPDA in wounded opr3 leaves reach about half their levels in wounded wild-type leaves. These results imply that cellular cyclopentenone levels are regulated, in part, by JA and/or other cyclopentanones [i.e., the oxo-pentenyl-cyclopentane (OPC) intermediates in JA synthesis]. In wild-type leaves, the amount of cis 3R,7S-JA vs. trans 3R,7R-JA occurred in a 1:3 up to a 1:2 ratio, whereas JA measured in opr3 occurred exclusively in the trans 3R,7R configuration (Fig. 2B). Thermal conversion in the gas chromatograph injector port is responsible for much of the epimerization of 3R,7S-JA into 3R,7R-JA but in wild type and in fad3 fad7 fad8 the former epimer is observed. The absence of detectable 3R,7S-JA in opr3 leaves is highly significant because this isomer is believed to be biologically active unlike the 3R,7R enantiomer (21). Thus, like in floral organs, no biologically active JA is detectable in opr3 leaves after wounding, ruling out the possible involvement of the additional OPR isoforms in active JA biosynthesis. Because both OPR1 and OPR2 were shown to be wound-inducible in Arabidopsis vegetative tissue (31), substrate specificity of OPR1 and OPR2 rather than enzyme availability is the limiting factor in the synthesis of active JA in opr3 (although the possibility that OPR1 and -2 are not expressed in cell types capable of producing JA cannot be ruled out). These observations are strongly supported by the conclusion reached by Schaller et al. (22) that OPR3 rather than OPR1 or OPR2 is the form of OPR required for active JA biosynthesis. The putative fourth and fifth OPRs identified in the Arabidopsis database (GenBank accession nos. AAC33200 and AB010695) are more closely related to OPR1 and OPR2 (7, 19), and it can be assumed that they do not contribute to in vivo production of 3R,7S-JA. In summary, the leaves of opr3 do not contain detectable 3R,7S-JA, making them ideal for the study of the role of JA in defense and gene expression.

OPDA as a Regulator of Gene Expression.

The opr3 mutant was used to investigate the role of JA and its precursors in signal transduction. With this mutant, the following questions were addressed: What role does JA play in the regulation of gene expression in vegetative tissues? Can OPDA (or one of its metabolites) contribute to gene regulation in the absence of JA? Three possibilities exist. Consistent with recent studies (6, 7, 26), JA in free or conjugated form could be uniquely capable of regulating gene expression. A second possibility is that OPDA (and not JA) is the sole in vivo regulator of gene expression and that effects of exogenous JA merely mimic the effects of OPDA or induce its synthesis. A third possibility is that cyclopentanones and cyclopentenones can both contribute to the regulation of gene expression. OPDA has been proposed to be the principal in vivo regulator of tendril coiling in Bryonia dioica (25, 32). Exogenous JA and OPDA induce different patterns of volatile production in lima beans (33). It has been proposed that JA and OPDA/dnOPDA (as well as related compounds) may all be bona fide regulators and that their changing proportions during disease or development would allow the cell to fine-tune gene expression (23, 34). Although this scenario, in which both JA and OPDA are biologically active, has received support (35), the current study allowed a direct genetic test of the hypothesis.

Gene expression profiles in wild-type and opr3 leaves were measured after mechanical wounding (Fig. 3A). We found that transcript levels for many genes were elevated after wounding in both genotypes. This finding is remarkable, because several genes on the microarray that are up-regulated after the mechanical wounding of Arabidopsis leaves require a functional COI1 pathway (28). These genes should, in principle, require both functional JA biosynthesis and signal transduction pathways. However, a closer look at the COI1-dependent genes reveals that there is a gradient of wound induction in opr3. Some genes are powerfully up-regulated in wild-type plants but not in opr3 (MBP, JIP, and VSP, for which there were, respectively, 4.3-, 4.7-, and 15.1-fold increases in transcript leaves 180 min after wounding). These are also strongly induced by JA treatment, suggesting that this group of genes is regulated to a large extent by JA in vivo. Several COI1-dependent genes were up-regulated in both genetic backgrounds but to a lesser extent in opr3 than in wild type (e.g., JR3, HPL, and ACO1). However, most of these genes were induced by treatment with JA or OPDA. This observation indicates that OPDA is a gene regulator per se during wounding and that both JA and OPDA are required for a full response. Decreased amounts of OPDA/dnOPDA in opr3 after wounding as compared with wild type could account for reduced gene expression as could changes in the dynamics of the cyclopentenone pool in opr3 plants. Finally, some COI1-dependent genes are similarly induced in wild type and opr3 (e.g., COMT, CCR). Because these two members of the aromatic amino acid metabolism are not induced by application of JA or OPDA, it is possible that another signal molecule(s) controls the activation of these genes in a COI1-dependent manner. In summary, we show that OPDA is a potent gene regulator. We can exclude the possibility of the conversion of OPDA to JA in opr3 plants, because VSP, which is strongly induced after wounding in wild type and by exogenous JA in opr3, was not induced in opr3 by exogenous OPDA or after wounding (Fig. 3).

These observations were extended by investigating the expression of the COI1- dependent defensin gene PDF1.2 during pathogenesis. PDF1.2 transcript accumulated in both wild-type and opr3 leaves infected with A. brassicicola, indicating that JA synthesis is not necessary for the activation of this gene. However PDF1.2 induction was lower in opr3 than in wild type (Fig. 4), suggesting again that JA in wild-type plants may coregulate or potentiate the activity of OPDA (and perhaps vice versa).

Electrophilic Properties of OPDA.

Further supporting a role for OPDA as a gene regulator is the observation that it can activate genes such as GST1, RNS1, and OPR1. These genes are known to be up-regulated by wounding in a COI1-independent manner (28) but the signal molecule that controls their activation is not known. GST1, RNS1, and OPR1 are induced by wounding in both wild type and opr3 and they are not activated by JA (Fig. 3). It is possible that the expression of these genes depends, at least in part, on a pathway involving the electrophilic effect caused by the chemistry of the cyclopentenone ring in OPDA. OPDA and dnOPDA are reactive electrophilic species possessing α,β-unsaturated carbonyl groups. Molecules containing this feature have been implicated as potent gene regulators in diseased plant tissues where they could also contribute to cell damage (36). Cyclopentenone prostaglandins also possess the α,β-unsaturated carbonyl feature and have been implicated as in vivo regulators (37). The results presented here indicate a new mechanism potentially responsible for the wound induction of several JA-independent genes. However, full demonstration of the relevance of the electrophilic properties of OPDA will require further investigation.

A model summarizing the roles of jasmonates in the control of gene expression in wounded or diseased Arabidopsis leaves is given in Fig. 5. We postulate that JA and OPDA/dnOPDA play distinct but complementary roles in the fine-tuning of gene expression as predicted in the oxylipin signature hypothesis (23). One pathway involving OPDA and JA clearly acts through the COI1 complex, but is it possible that at least one other pathway controlled by means of the electrophilic properties of cyclopentenone jasmonates might exist. This is the first report, to our knowledge, of such properties of cyclopentenones in plants. Interestingly, the chemical structure of the cyclopentanone ring prohibits this function for JA.

Figure 5.

Model for the role of jasmonate family members in the fine control of gene expression in wounded or diseased Arabidopsis leaves. The cyclopentenones OPDA and dnOPDA can regulate genes directly via a COI1-dependent pathway, or via the same signal pathway, after their conversion to cyclopentanones, i.e., JA or its oxo-pentenyl-cyclopentane (OPC) precursors. In this way, JA and OPDA could both regulate specific sets of genes or may act in concert to control the expression of common sets of genes. Another layer of regulation might be provided by the electrophilic properties of cyclopentenone jasmonates.

In conclusion, cyclopentenones and cyclopentanones can act alone or in concert to regulate gene expression. Both JA and OPDA are envisaged to have distinct and important roles in the control of gene expression in plant defense. The exact mechanism of this regulation is not known, but our results illustrate a complex network of signaling molecules in the jasmonate family. The successful genetic dissection of different regulatory activities for lipid-derived signals in plants should pave the way toward a better understanding of how plant cells generate and interpret complex information in the oxylipin signature.

Abbreviations

JA

jasmonic acid

OPDA

12-oxo-phytodienoic acid

OPR

OPDA reductase

dnOPDA

dinor OPDA