Available online at www.sciencedirect.com

Plant Physiology and Biochemistry 46 (2008) 302e308www.elsevier.com/locate/plaphy

Review

Molecular mechanism of enzymatic allene oxide cyclization in plants

Eckhard Hofmann a,*, Stephan Pollmann b

a Biophysics, Department of Biology and Biotechnology, Ruhr-University Bochum, Universitaetsstrasse 150, D-44801 Bochum, Germanyb Plant Physiology, Department of Biology and Biotechnology, Ruhr-University Bochum, Universitaetsstrasse 150, D-44801 Bochum, Germany

Received 14 November 2007

Available online 31 December 2007

Abstract

Jasmonates, a collective term combining both jasmonic acid (JA) and related derivatives, are ubiquitously distributed in the plant kingdom.They are characterized as lipid-derived signal molecules which mediate a plethora of physiological functions, in particular stress responses, malefertility, and a multitude of developmental processes. In the course of JA biosynthesis, the first oxylipin with signal character, cis-(þ)-12-oxo-phytodienoic acid (OPDA), is produced in a cyclization reaction catalyzed by allene oxide cyclase (AOC). This enzyme-catalyzed ring closure isof particular importance, as it warrants the enantiomeric structure at the cyclopentenone ring which in the end results in the only bioactive JAenantiomer, cis-(þ)-JA. In this review, we focus on the structural and molecular mechanisms underlying the above mentioned cyclization re-action. In this context, we will discuss the crystal structure of AOC2 of Arabidopsis thaliana with respect to putative binding sites of the instablesubstrate, 12,13-epoxy-9(Z ),11,15(Z )-octadecatrienoic acid (12,13-EOT), as well as possible intermolecular rearrangements during the cycliza-tion reaction.� 2007 Elsevier Masson SAS. All rights reserved.

Keywords: Allene oxide cyclase; Allene oxide synthase; Jasmonate; 12-Oxo-phytodienoic acid; Oxylipins; X-ray structure

1. Introduction

Besides brassinosteroids and oligopeptides with hormone-like functions, the jasmonates are among the most recentlyidentified signal molecules with phytohormone properties,and are widespread throughout a variety of different plantphyla [25]. Although the jasmonic acid methyl ester (MeJA)was demonstrated to be a constituent of the essential oil of

Abbreviations: ACS, acyl-CoA synthase; AOC, allene oxide cyclase; 12,13-

EOT, 12,13(S )-epoxy-9(Z ),11,15(Z )-octadecatrienoic acid; 12,13-EOD,

12,13(S )-epoxy-9(Z ),11-octadecatrienoic acid; AOS, allene oxide synthase;

CESG, Center for Eucaryotic Structural Genomics; CTS/PXA1, ABC trans-

porter for OPDA or OPDA-CoA import; HPOD, 13(S )-hydroperoxy-

9(Z ),11(E )-octadecadienoic acid; JA, jasmonic acid; MeJA, jasmonic acid

methylester; OPC-8:0, 3-oxo-2(20(Z )-pentenyl)-cyclopentane-1-octanoic acid;

LA, a-linolenic acid; OPDA, 12-oxo-phytodienoic acid; OPR, 12-oxo-phytodie-

noic acid reductase; 13(S )-HPOT, 13(S )-hydroperoxy-9(Z ),11(E ),15(Z )-octa-

decatrienoic acid; 13-LOX, 13-lipoxygenase.

* Corresponding author. Tel.: þ49 234 32 24463; fax: þ49 234 32 14238.

E-mail address: eckhard.hofmann@bph.ruhr-uni-bochum.de (E.

Hofmann).

0981-9428/$ - see front matter � 2007 Elsevier Masson SAS. All rights reserved.

doi:10.1016/j.plaphy.2007.12.007

Jasminum grandiflorum in the early 1960s [7], it took nearlyanother twenty years until the first physiological effects ofMeJA and the occurrence of the free acid were described[4,33]. To date, jasmonic acid and its derivatives are associatedwith diverse physiological functions. The most prominent oneis the involvement of JA in wound response and pathogenesis[8,18,37]. Additionally, jasmonates play a crucial role in re-production [9], metabolic regulation [35], and as a signaltransducer in mechanotransduction [31,38,39]. JA is also re-quired for protection from ozone damage [26,27], and hasa pivotal role in the production of protective secondary metab-olites in cell cultures of Eschscholtzia californica [2,3].

The pathway of jasmonic acid biosynthesis is shown inFig. 1. Jasmonic acid and its octadecanoid precursors are syn-thesized from a-linolenic acid (a-LA) which is found in greatextent in plastidial membranes. From there, a-LA is suggestedto be released by the action of lipases, e.g. the phospholipaseA1 DAD1 [19]. The subsequent oxygenation of a-LA at theC-13 position is catalyzed by 13-lipoxygenase [1]. The resulting13-hydroperoxide, 13(S )-hydroperoxy-9(Z ),11(E ),15(Z )-octadecatrienoic acid (13-HPOT), is further dehydrated with

Fig. 1. Pathway of jasmonic acid biosynthesis in plants. Intermediates are ab-

breviated as: 13-HPOT, 13(S )-hydroperoxy-9(Z ),11(E ),15(Z )-octadecatrie-

noic acid; 12,13-EOT, 12,13(S )-epoxy-9(Z ),11,15(Z )-octadecatrienoic acid;

OPDA, cis-(þ)-12-oxo-phytodienoic acid; OPDA-CoA, cis-(þ)-12-oxo-phyto-

dienoic acid-coenzyme A; OPC8:0, 3-oxo-2(20(Z )-pentenyl)-cyclopentane-1-

octanoic acid. The enzymes are indicated as: LIP, lipase; LOX, lipoxygenase;

AOS, allene oxide synthase; AOC, allene oxide cyclase; OPR, oxo-phytodie-

noic acid reductase; CTS/PXA1, comatose, ABC transporter for OPDA or

OPDA-CoA import; ACS, acyl-CoA synthase.

303E. Hofmann, S. Pollmann / Plant Physiology and Biochemistry 46 (2008) 302e308

the help of allene oxide synthase [29,30], providing the unstableintermediate, 12,13-epoxy-9(Z ),11,15(Z )-octadecatrienoicacid (12,13-EOT). Allene oxide cyclase [14,32] catalyzes thereaction within the octadecanoid pathway which guaranteesenantiomeric specificity, by converting 12,13-EOT to 12-oxo-10,15(Z )-phytodienoic acid (OPDA). OPDA is then trans-ferred from the chloroplast to the peroxisome where it is furthermetabolized by reduction of the D10-double bond catalyzedby oxo-phytodienoic acid reductase [28], yielding 3-oxo-2(20(Z )-pentenyl)-cyclopentane-1-octanoic acid (OPC-8:0). Due to radiotracer experiments [36], it is generally agreedthat OPC-8:0 undergoes three consecutive cycles of b-oxidationwhich results in the production of bioactive JA with (3R,7S )-configuration, i.e. (þ)-7-iso-JA.

Allene oxide cyclase (AOC) has been described for the firsttime from Zea mays [14,41], followed by the cloning of thecorresponding genes from tomato [42], Arabidopsis [32], andbarley [24]. While in tomato AOC is encoded as a singlegene, in A. thaliana four isogenes can be found, which mostlikely evolved from one ancestral isoform by gene duplicationevents. With respect to functional differences of the four isoen-zymes, it has been shown that especially AOC2 mRNA accumu-lates in the case of local as well as systemic wound response,whereas AOC1 mRNA seems to be preferentially transcribedin systemic wound response. By reason that allene oxide syn-thase (AOS) transcription is also systemically induced afterwounding [20,21], a specific interaction of AOS and AOC1might be supposable in systemically responding leaves. Furtherevidence for functional differences of the AOCs is emphasizedby the occurrence of dinor-oxo-phytodienoic acid (dnOPDA)which is synthesized from hexadecatrienoic acid. Possibly,in this context, the isoenzymes possess diverse substratespecificities [32]. Unfortunately, investigations of either en-zyme kinetic or substrate specificity of the AOCs have beenhampered by the instability of their substrate. So far, all activityassays utilized a coupled test system, determining the combinedactivity of both AOS and AOC.

The AOCs from Arabidopsis contain a predicted plastidialtarget sequence which facilitates the import of the enzymesinto the chloroplast. Functional import of AOC into the chloro-plast, investigated by immunocytochemical means, has alreadybeen described [32]. However, differentiation between the indi-vidual isoforms was not possible, suggesting that a more de-tailed examination of the import of the single isoforms ofAOC is needed. Intriguingly, the expression of ArabidopsisAOCs has been shown for all plant organs, including roots[6]. This finding is in contrast to that obtained from tomatowhere AOC expression is described to be restricted to floralorgans and vascular bundles [15]. However, there are stillmany open questions which mark challenges for future work.One of the most urgent topics, after unraveling the molecularmechanism of AOC catalyzed 12,13-EOT cyclization, is theelucidation of the functional interconnection of AOS andAOC. Although a covalent interaction of AOS and AOC hasbeen described as unnecessary [40], the close vicinity of thetwo proteins seems to enhance their combined activity(P. Zerbe, personal communication).

304 E. Hofmann, S. Pollmann / Plant Physiology and Biochemistry 46 (2008) 302e308

2. Structures of AOC2

The structure of AOC2 from Arabidopsis thaliana has beendetermined by X-ray crystallography independently in twodifferent labs. Due to these efforts five different structuresare available from the protein databank. Selenomethioninelabeled protein has been crystallized in orthorhombic spacegroups and solved to a resolution of 1.7 A and 1.5 A by theCenter for Eucaryotic Structural Genomics (CESG) (1Z8K,Wesenberg et al., unpublished) and by a group from theRuhr University Bochum (2BRJ, [16]), respectively. The inde-pendently determined structures superpose extremely wellwith an overall root mean square deviation of only 0.31 Afor 173 Ca atoms [16]. In both crystal packings one stabletrimer of AOC2 is observed per asymmetric unit. As a controlthe Bochum group also solved the structure of the unlabelledprotein in a monoclinic spacegroup at 1.8 A with two trimersin the asymmetric unit (2GIN, [16]). Of functional importanceis the result of soaking experiments of orthorhombic crystalswith a competitive inhibitor, which led to the coordinates ofthis molecule inside the proposed catalytic site of the enzyme(2DIO, [16]). Reevaluation of the original data of the CESGwith an improved refinement protocol resulted in the deposi-tion of the fifth coordinate set (2Q4I, [22]).

In the following review we will use the highest resolutioncoordinates 2BRJ to introduce the overall architecture of theenzyme. As noted above, in all crystal forms observed sofar, AOC2 was found to form a trimeric quarternary struc-ture. In Fig. 2 this trimer is shown in ribbon representation

Fig. 2. Structure of AOC2 from Arabidopsis thaliana. Shown is the complete

trimer found in the asymmetric unit of the crystal (accession code 2BRJ). Each

monomer is individually colored from blue (N-terminus) to red (C-terminus)

and labeled with the chain ID (A,B,C). The position of the threefold non-

crystallographic axis is marked by the black triangle. Figure produced with

Pymol [5].

looking along the threefold axis; in Fig. 3 the monomer isshown.

The main structural feature of AOC2 is the central 8-stranded antiparallel b-barrel. It has a slightly elliptical crosssection with axes of about 14 and 18 A and walls of height be-tween 11 and 30 A [16]. The barrel is not filled completely byside chain atoms but rather forms an elongated hydrophobiccavity reaching deep into the protein. Notably the highest re-gions of the wall (strands 3e5) which show the most extendedh-bonding network and are therefore expected to be the moststable areas of the structure interact with the neighboringmonomers in the trimer interface. This trimerization interfacecovers roughly 2000 A2 of a monomer surface [16]. While thebarrel is formed by residues 17e147, the remaining 41 C-terminal residues (colored salmon in Fig. 3) form a mixtureof helical and random coil structures, which cover the bottomof the barrel and the sides not involved in trimerization. Thefirst 16 N-terminal residues are not visible in the structuredue to disorder. They consist of the engineered His6-tag with

Fig. 3. The AOC2 monomer. Shown is the protein in ribbon representation

(2BRJ, chain A). The C-terminal residues 148e188 are colored in salmon.

The view is rotated about 90� with respect to Fig. 2. The b-strands of the barrel

are labeled S1eS8, the termini of the model are labeled N and C.

Figure produced with Pymol [5].

O

O

O

OO

O

O

O

O

O

O

O

12,13-EOT

OPDA

12,13-Epoxyoctadecadienoic acid

AOC2

12,13-EOD

Fig. 4. Structural formulas of productive substrate 12,13-EOT, not cyclizable

analogs 12,13-EOD and the competitive inhibitor vernolic acid (12,13 epox-

yoctadecadienoic acid) together with the product OPDA.

305E. Hofmann, S. Pollmann / Plant Physiology and Biochemistry 46 (2008) 302e308

some additional linker residues and the first two residues afterthe predicted signal peptidase processing site. The equivalentresidues are missing in all structures in the database so far, re-gardless of the expression constructs used.

Based on topological arguments and on the results ofa DALI-search [17], AOC has been tentatively assigned tobe a member of the lipocalin family [10,16]. Intriguingly, par-allels exist both in the overall architecture and the substrateclass bound in the central cavity. Most notably prostaglandinD synthase is an enzyme involved in the prostanoid synthesisin mammals and has been found to share a lipocalin fold [34].Nevertheless key sequence features of lipocalins [10] are miss-ing in the AOC2 structure and the observed similarity mightwell be a result of convergent evolution instead of a commonancestor [16]. Based on sequence data, the four other knownlipocalins from plants have been classified to be members ofa divergent subclass of lipocalins termed outlier lipocalins[13]. Therefore the proposal should be critically reevaluatedonce more lipocalins from plants are structurally known.

The finding that AOC2 forms trimers in the crystal struc-tures led to a reevaluation of biochemical data. For cornAOC a dimeric form had been proposed based on size exclu-sion chromatography [41]. Similar results have been reportedfor AOC2 [16], but have been interpreted as being compatiblewith trimers. The observation of SDS-stable trimers both inplant extracts and in purified enzyme led to the conclusionthat the retardation on the size exclusion column might behigher than expected by the molecular weight. A more system-atic biochemical analysis is underway in our laboratory buthas not yet resulted in clear evidence for the oligomeric statein solution (S. Pollmann, unpublished results).

3. The reaction mechanism

Analysis of the AOC2 structure already suggests that theactive site is located inside the barrel cavity. While it is mostlylined by hydrophobic and aromatic residues, three patches ofmore polar residues are noteworthy. One patch is formed bya proline (P32), serine (S31), and two asparagine residues(N25, N53) which coordinate a water molecule found in allavailable AOC structures so far. On the opposite side of thebarrel wall a cysteine (C71) is located. Finally at the bottomof the barrel a glutamate residue (E23) is found. All of theseresidues are strictly conserved in known AOC sequences.

To obtain experimental evidence for the location of the ac-tive site the competitive inhibitor vernolic acid has been used(Fig. 4). It differs from the substrate EOT by the absence of theC11eC12 double bond and can therefore not undergo the cy-clization reaction. Vernolic acid has been shown to be a potentinhibitor of the corn AOC [41], but the inhibitory effect is notas pronounced in the case of Arabidopsis AOC1-4 (P. Zerbe,personal communication).

In the structure of AOC2 in complex with vernolic acid(2DIO; Fig. 5A), the inhibitor is bound with the acyl chainburied deep inside the pocket. The carboxylic moiety is nottightly bound, but rather flexible on the protein surface. Theepoxy group of vernolic acid is found in hydrogen-bonding

distance to the conserved water molecule W75. In the complexstructure no induced fit of the protein is observed (Fig. 5A).

Based on the structural evidence for the binding of vernolicacid, both substrate 12,13-EOT and product OPDA were mod-eled into the binding pocket to facilitate discussion of possiblereaction mechanisms (Fig. 5B) [16]. On this basis the follow-ing scheme was postulated: 12,13-EOT binds with the u-endinto the hydrophobic pocket of AOC2. The epoxide oxygenis coordinated by a conserved water molecule, which is inturn tightly bound by the surrounding protein residues P32,S31, N25 and N53. The carboxylic moiety of 12,13-EOT rea-ches out of the pocket to the surface of the protein. Cyclizationis initiated by the delocalization of the C15 double bond whichin turn is triggered by the essential E23 (Fig. 6A). This leads tothe epoxide opening by anchimeric assistance. The importanceof the C15 double bond for enzyme-controlled cyclization hasbeen demonstrated by several studies and would be rational-ized by this proposal. After epoxide opening the resulting oxy-anion is readily stabilized by the conserved water. Hofmannet al. [16] proposed the formation of a classical pentadienylcation. To reach a productive conformation a transecis isom-erization around the C10eC11 bond has to be assumed, whichcould be driven by the hydrophobic effect due to a better burialof the hydrocarbon chain in the pocket. The resulting non-planar dienyl-like pentadienyl cation would be stabilized bypep interactions of nearby aromatic residues; also the con-served C71 was postulated to stabilize the delocalized positivecharge (Fig. 6B). The final step of the reaction would be a con-rotary pericyclic ring closure along the lines of the Wood-wardeHoffmann rules (Fig. 6C). Stereoselectivity would bemainly achieved by the geometrical control of the transecisisomerization by the hydrophobic parts of the pocket.

While the proposed mechanism builds upon the concept ofanchimeric assistance proposed for the oxylipin cyclization by

Fig. 5. The AOC2 binding pocket. (A) Inhibitor vernolic acid (VA, salmon stick model) bound in the barrel cavity. Shown are both the ligand-free (2BRJ, grey) and

the ligand-bound (2DIO, yellow) structure superposed. Residues closer than 4 A to either vernolic acid or the conserved water molecule W75 (red sphere) are

shown as sticks. (B) Substrate EOT (green) and reaction product OPDA (purple) modeled into the binding pocket of AOC2. The conserved water molecule

W75 is shown as a red sphere, surrounding sidechains are shown as sticks and labeled. Figure produced with Pymol [5].

306 E. Hofmann, S. Pollmann / Plant Physiology and Biochemistry 46 (2008) 302e308

Grechkin and coworkers [11], and assumes therefore a similarreaction initiation, it differs in the assumed mechanism forring closure. In solution a strong dependency of the cyclizationreaction of 12,13-EOT on the pH has been observed [12],which cannot be easily explained by a classical pericyclicring closure. Instead Grechkin et al. [12] propose a dipolar an-nulation mechanism, which includes an intermediate witha carbanion located at C-9 and a carbocation located at C-13. This concept is needed to accommodate the observedpH-dependency and has been generally accepted.

For the enzyme-controlled reaction the situation is slightlydifferent. In the rather tight hydrophobic pocket no free sol-vent molecules are available to act as stabilizing counter-charges for intermediate ionic states in the course of thereaction. Instead stabilization has to be facilitated by proteinside chains or tightly bound water molecules. So far thestructural evidence does not present a likely candidate to sta-bilize the proposed carbanion localized at C-9. Therefore themore classical scheme with a delocalized charged intermedi-ate seems to be more favorable. In addition the schemewould be in agreement with a theoretical study on vinylallene oxides, suggesting a stepwise mechanism includingstrong enantiofacial torque selectivity enforced by the protein[23].

Nevertheless, one has to bear in mind that there is no realstructure of the substrateeenzyme complex, which mightwell show detailed structural features in favor of the annula-tion mechanism. In addition analysis of point mutations inthe active site area might allow discrimination between thetwo concepts and will clarify the situation. Clearly, morework is needed for a full understanding of the reactionmechanism.

4. Comparison with other AOC structures

A new structure of AOC2 has been deposited recently by theCESG (2Q4I, [22]). Here the original crystallographic datahave been reevaluated with an ensemble refinement method.AOC2 has been one of 50 structures used in this methodolog-ical study and is not separately discussed in the paper. In thedeposited coordinates 8 different conformers have been refinedin parallel to allow a better representation of the inherentflexibility in the protein. The rationale behind the method isthe observation that this kind of model fits the experimentaldata better and should therefore be more informative than theclassical model with isotropic temperature factors for singleatoms modeling the inherent disorder. Discussion of this ap-proach is beyond the scope of this review, but a brief analysisof the ensemble model confirms the rigid nature of the barrel,while the exterior loops show a somewhat higher flexibility.Most notably the backbone and side chains of residues aroundthe active site do not show any significant shifts in the differentmembers of the ensemble.

In the target list of the CESG all four isoforms of AOC havebeen included (http://targetdb.pdb.org/). In addition to thestructures of AOC2 mentioned above, the structure of AOC1has also been solved at a resolution of 1.8 A and deposited inthe PDB (1ZVC). Interestingly, AOC1 also crystallizes as a tri-mer, but in a completely different crystallographic environ-ment. This supports the hypothesis of the trimer beinga relevant aggregation state of AOC1 and AOC2 in vivo. Acomparison between the two structures showed no significantstructural deviations [16]. Both proteins are very similar anddiffer only in 10 positions in sequence, most of which representconservative changes (Fig. 7). None of these mutations are

Fig. 6. Proposed cyclization reaction mechanism. Substrate, product and inter-

mediates are shown as structural formulas, important amino acids are shown as

rendered sticks. For details see text.

Fig. 7. Superposition of AOC1 (1ZVC) and AOC2 (2BRJ). Both proteins are

shown in ribbon representation (AOC2 in gray, AOC1 in blue). The sidechains

of residues of AOC2 which are mutated in AOC1 are shown as stick models

and labeled. Figure produced with Pymol [5].

307E. Hofmann, S. Pollmann / Plant Physiology and Biochemistry 46 (2008) 302e308

found to be located close to the active site, which is consistentwith the finding that the enzymes show very similar activitiesand substrate specificities (P. Zerbe, personal communication)[6]. If of any physiological relevance, the small structural dif-ferences observed between the two enzymes might play a rolein proteineprotein interactions with e.g. AOS [16].

While for AOC1 and AOC2 structures have already beendeposited, work on AOC4 has been stopped. For AOC3 thestructure elucidation progressed to ‘‘crystallized’’ and we canhope for this structure to be available soon. While we do not

expect any large structural differences to AOC1 or AOC2,the comparison will improve our understanding of the criticalregions of the enzyme.

Acknowledgements

We thank Dr Florian Schaller and Dr Philipp Zerbe for theirdiscussions and help with figure preparation. This work wasfunded by grants from the Deutsche Forschungsgemeinschaft(DFG), Bonn SCHA939 and HO2600 for EH, SFB480, projectA-10 for SP.

References

[1] E. Bell, R.A. Creelman, J.E. Mullet, A chloroplast lipoxygenase is re-

quired for wound-induced jasmonic acid accumulation in Arabidopsis,

Proceedings of the National Academy of Sciences of the United States

of America 92 (1995) 8675e8679.

[2] S. Blechert, W. Brodschelm, S. Holder, L. Kammerer, T.M. Kutchan,

M.J. Mueller, Z.Q. Xia, M.H. Zenk, The octadecanoic pathway e signal

molecules for the regulation of secondary pathways, Proceedings of the

National Academy of Sciences of the United States of America 92 (1995)

4099e4105.

308 E. Hofmann, S. Pollmann / Plant Physiology and Biochemistry 46 (2008) 302e308

[3] S.Y. Byun, Re-elicitation with methyl jasmonate in Eschscholtzia cali-fornica cell suspension cultures, Journal of Microbiology and Biotech-

nology 10 (2000) 107e110.

[4] W. Dathe, H. Ronsch, A. Preiss, W. Schade, G. Sembdner, K. Schreiber,

Endogenous plant hormones of the broad bean, Vicia faba L. (�)-

jasmonic acid, a plant-growth inhibitor in pericarp, Planta 153 (1981)

530e535.

[5] W.L. DeLano, J.W. Lam, PyMOL: A communications tool for computa-

tional models, Abstracts of Papers of the American Chemical Society

230 (2005) U1371eU1372.

[6] C. Delker, I. Stenzel, B. Hause, O. Miersch, I. Feussner, C. Wasternack,

Jasmonate biosynthesis in Arabidopsis thaliana e enzymes, products,

regulation, Plant Biology 8 (2006) 297e306.

[7] E. Demole, E. Lederer, D. Mercier, Isolement et determination de la

structure du jasmonate de methyle, constituant odorant caracteristique

de lessence de jasmin, Helvetica Chimica Acta 45 (1962) 675e685.

[8] E.F. Farmer, C.A. Ryan, Octadecanoid precursors of jasmonic acid acti-

vate the synthesis of wound-inducible proteinase-inhibitors, Plant Cell 4

(1992) 129e134.

[9] B.J.F. Feys, C.E. Benedetti, C.N. Penfold, J.G. Turner, Arabidopsis mu-

tants selected for resistance to the phytotoxin coronatine are male-sterile,

insensitive to methyl jasmonate, and resistant to a bacterial pathogen,

Plant Cell 6 (1994) 751e759.

[10] D.R. Flower, A.C.T. North, C.E. Sansom, The lipocalin protein family:

structural and sequence overview, Biochimica et Biophysica Acta: Pro-

tein Structure and Molecular Enzymology 1482 (2000) 9e24.

[11] A.N. Grechkin, Cyclization of natural allene oxide fatty-acids e the an-

chimeric assistance of beta, gamma-double bond beside the oxirane and

the reaction-mechanism, Biochimica et Biophysica Acta: Lipids and

Lipid Metabolism 1213 (1994) 199e206.

[12] A.N. Grechkin, I.R. Chechetkin, L.S. Mukhtarova, M. Hamberg, Role of

structure and pH in cyclization of allene oxide fatty acids: implications

for the reaction mechanism, Chemistry and Physics of Lipids 120

(2002) 87e99.

[13] J. Grzyb, D. Latowski, K. Strzalka, Lipocalins e a family portrait, Jour-

nal of Plant Physiology 163 (2006) 895e915.

[14] M. Hamberg, Biosynthesis of 12-oxo-10,15(Z )-phytodienoic acid: iden-

tification of an allene oxide cyclase, Biochemical and Biophysical

Research Communications 156 (1988) 543e550.

[15] B. Hause, I. Stenzel, O. Miersch, H. Maucher, R. Kramell, J. Ziegler,

C. Wasternack, Tissue-specific oxylipin signature of tomato flowers:

allene oxide cyclase is highly expressed in distinct flower organs and vas-

cular bundles, Plant Journal 24 (2000) 113e126.

[16] E. Hofmann, P. Zerbe, F. Schaller, The crystal structure of Arabidopsis

thaliana allene oxide cyclase: insights into the oxylipin cyclization

reaction, Plant Cell 18 (2006) 3201e3217.

[17] L. Holm, C. Sander, Searching protein-structure databases has come of

age, Proteins: Structure Function and Genetics 19 (1994) 165e173.

[18] G.A. Howe, J. Lightner, J. Browse, C.A. Ryan, An octadecanoid pathway

mutant (JL5) of tomato is compromised in signaling for defense against

insect attack, Plant Cell 8 (1996) 2067e2077.

[19] S. Ishiguro, A. Kawai-Oda, J. Ueda, I. Nishida, K. Okada, The DEFEC-

TIVE IN ANTHER DEHISCENCE1 gene encodes a novel phospholi-

pase A1 catalyzing the initial step of jasmonic acid biosynthesis,

which synchronizes pollen maturation, anther dehiscence, and flower

opening in Arabidopsis, Plant Cell 13 (2001) 2191e2209.

[20] I. Kubigsteltig, D. Laudert, E.W. Weiler, Structure and regulation of the Ara-

bidopsis thaliana allene oxide synthase gene, Planta 208 (1999) 463e471.

[21] D. Laudert, E.W. Weiler, Allene oxide synthase: a major control point in

Arabidopsis thaliana octadecanoid signalling, Plant Journal 15 (1998)

675e684.

[22] E.J. Levin, D.A. Kondrashov, G.E. Wesenberg, G.N. Phillips, Ensemble

refinement of protein crystal structures: validation and application, Struc-

ture 15 (2007) 1040e1052.

[23] C.S. Lopez, O.N. Faza, D.M. York, A.R. de Lera, Theoretical study of the

vinyl allene oxide to cyclopent-2-en-1-one rearrangement: mechanism,

torquoselectivity and solvent effects, Journal of Organic Chemistry 69

(2004) 3635e3644.

[24] H. Maucher, I. Stenzel, O. Miersch, N. Stein, M. Prasad, U. Zierold,

P. Schweizer, C. Dorer, B. Hause, C. Wasternack, The allene oxide

cyclase of barley (Hordeum vulgare L.) e cloning and organ-specific

expression, Phytochemistry 65 (2004) 801e811.

[25] A. Meyer, O. Miersch, C. Buttner, W. Dathe, G. Sembdner, Occurrence

of the plant-growth regulator jasmonic acid in plants, Journal of Plant

Growth Regulation 3 (1984) 1e8.

[26] K. Overmyer, H. Tuominen, R. Kettunen, C. Betz, C. Langebartels,

H. Sandermann, J. Kangasjarvi, Ozone-sensitive Arabidopsis rcd1 mutant

reveals opposite roles for ethylene and jasmonate signaling pathways in

regulating superoxide-dependent cell death, Plant Cell 12 (2000) 1849e

1862.

[27] M.V. Rao, H. Lee, R.A. Creelman, J.E. Mullet, K.R. Davis, Jasmonic

acid signaling modulates ozone-induced hypersensitive cell death, Plant

Cell 12 (2000) 1633e1646.

[28] F. Schaller, E.W. Weiler, Molecular cloning and characterization of 12-

oxophytodienoate reductase, an enzyme of the octadecanoid signaling

pathway from Arabidopsis thaliana e structural and functional relation-

ship to yeast old yellow enzyme, Journal of Biological Chemistry 272

(1997) 28066e28072.

[29] W.C. Song, A.R. Brash, Purification of an allene oxide synthase and

identification of the enzyme as a cytochrome-P-450, Science 253

(1991) 781e784.

[30] W.C. Song, C.D. Funk, A.R. Brash, Molecular-cloning of an allene oxide

synthase e a cytochrome-P450 specialized for the metabolism of fatty-

acid hydroperoxides, Proceedings of the National Academy of Sciences

of the United States of America 90 (1993) 8519e8523.

[31] B.A. Stelmach, A. Muller, P. Hennig, D. Laudert, L. Andert, E.W. Weiler,

Quantitation of the octadecanoid 12-oxo-phytodienoic acid, a signalling com-

pound in plant mechanotransduction, Phytochemistry 47 (1998) 539e546.

[32] I. Stenzel, B. Hause, O. Miersch, T. Kurz, H. Maucher, H. Weichert,

J. Ziegler, I. Feussner, C. Wasternack, Jasmonate biosynthesis and the

allene oxide cyclase family of Arabidopsis thaliana, Plant Molecular

Biology 51 (2003) 895e911.

[33] J. Ueda, J. Kato, Isolation and identification of a senescence-promoting

substance from Wormwood (Artemisia absinthium L.), Plant Physiol.

66 (1980) 246e249.

[34] Y. Urade, O. Hayaishi, Biochemical, structural, genetic, physiological,

and pathophysiological features of lipocalin-type prostaglandin D syn-

thase, Biochimica et Biophysica Acta: Protein Structure and Molecular

Enzymology 1482 (2000) 259e271.

[35] L. van der Fits, J. Memelink, ORCA3, a jasmonate-responsive transcrip-

tional regulator of plant primary and secondary metabolism, Science 289

(2000) 295e297.

[36] B.A. Vick, D.C. Zimmerman, The Biosynthesis of jasmonic acid ea physiological-role for plant lipoxygenase, Biochemical and Biophysi-

cal Research Communications 111 (1983) 470e477.

[37] P. Vijayan, J. Shockey, C.A. Levesque, R.J. Cook, J. Browse, A role for

jasmonate in pathogen defense of Arabidopsis, Proceedings of the Na-

tional Academy of Sciences of the United States of America 95 (1998)

7209e7214.

[38] E.W. Weiler, T. Albrecht, B. Groth, Z.Q. Xia, M. Luxem, H. Liss,

L. Andert, P. Spengler, Evidence for the involvement of jasmonates

and their octadecanoid precursors in the tendril coiling response of Bryo-

nia dioica, Phytochemistry 32 (1993) 591e600.

[39] E.W. Weiler, T.M. Kutchan, T. Gorba, W. Brodschelm, U. Niesel,

F. Bublitz, The Pseudomonas phytotoxin coronatine mimics octadeca-

noid signaling molecules of higher-plants, Febs Letters 345 (1994) 9e13.

[40] P. Zerbe, E.W. Weiler, F. Schaller, Preparative enzymatic solid phase

synthesis of cis-(þ)-12-oxo-phytodienoic acid e physical interaction of

AOS and AOC is not necessary, Phytochemistry 68 (2007) 229e236.

[41] J. Ziegler, M. Hamberg, O. Miersch, B. Parthier, Purification and charac-

terization of allene oxide cyclase from dry corn seeds, Plant Physiology

114 (1997) 565e573.

[42] J. Ziegler, I. Stenzel, B. Hause, H. Maucher, M. Hamberg, R. Grimm,

M. Ganal, C. Wasternack, Molecular cloning of allene oxide cyclase e

the enzyme establishing the stereochemistry of octadecanoids and jasm-

onates, Journal of Biological Chemistry 275 (2000) 19132e19138.