Cytochrome P450 metabolizing fatty acids in plants: characterization and physiological roles

Authors


F. Pinot, IBMP-CNRS UPR 2357, Institut de Botanique, 28 rue Goethe, F-67083 Strasbourg Cedex, France
Fax: +33 (0)368851921
Tel: +33 3 68 85 19 99
E-mail: franck.pinot@ibmp-cnrs.unistra.fr

Abstract

In plants, fatty acids (FA) are subjected to various types of oxygenation reactions. Products include hydroxyacids, as well as hydroperoxides, epoxides, aldehydes, ketones and α,ω-diacids. Many of these reactions are catalysed by cytochrome P450s (P450s), which represent one of the largest superfamilies of proteins in plants. The existence of P450-type metabolizing FA enzymes in plants was established approximately four decades ago in studies on the biosynthesis of lipid polyesters. Biochemical investigations have highlighted two major characteristics of P450s acting on FAs: (a) they can be inhibited by FA analogues carrying an acetylenic function, and (b) they can be enhanced by biotic and abiotic stress at the transcriptional level. Based on these properties, P450s capable of producing oxidized FA have been identified and characterized from various plant species. Until recently, the vast majority of characterized P450s acting on FAs belonged to the CYP86 and CYP94 families. In the past five years, rapid progress in the characterization of mutants in the model plant Arabidopsis thaliana has allowed the identification of such enzymes in many other P450 families (i.e. CYP703, CYP704, CYP709, CYP77, CYP74). The presence in a single species of distinct enzymes characterized by their own regulation and catalytic properties raised the question of their physiological meaning. Functional studies in A. thaliana have demonstrated the involvement of FA hydroxylases in the synthesis of the protective biopolymers cutin, suberin and sporopollenin. In addition, several lines of evidence discussed in this minireview are consistent with P450s metabolizing FAs in many aspects of plant biology, such as defence against pathogens and herbivores, development, catabolism or reproduction.

Abbreviations
FA

fatty acid

ω

omega position

P450

cytochrome P450

Introduction

Plants produce a wide variety of oxygenated derivatives of fatty acids (FAs) which are involved in various important biological functions ranging from waterproofing to signalling and plant defence. Most oxygenated FAs are the products of enzyme-catalysed reactions. Hydroperoxides of lipid signalling result mainly from the action of lipoxygenases. The hydroxy FAs of some seed oils (e.g. ricinoleic acid of castor bean) and membrane lipids (2-hydroxy FA of sphingolipids) are synthesized by homologues of FA desaturases. But for the vast majority of other oxygenated FA derivatives, the insertion of oxygen atoms in the carbon chain is dependent on cytochrome P450s (P450s).

The involvement of P450s in FA metabolism in plants was first described in the context of cutin and suberin studies [1,2]. These protective biopolymers are made mostly of hydroxyfatty acids and α,ω-dicarboxylic acids esterified to each other and to glycerol. The hydroxyl groups of the FAs can be located at the terminal methyl (ω position) or on various internal positions [3]. Enzymes capable of hydroxylation of FA thus have a key role to play in cutin and suberin synthesis by allowing the production of bifunctional FA derivatives which can serve as monomers for polymerization. The pioneering studies on the biosynthesis of polymers of oxygenated FAs have shed light on the capacity of P450s to catalyse in-chain and ω-hydroxylation of FAs, but also to catalyse the introduction of other functional groups in FAs such as epoxy groups, which are common substituents in cutin or suberin monomers of some plant species [1,2]. Incubation of lauric acid (C12:0) and its unsaturated analogues with microsomal fractions of various plant species confirmed the existence in plants of distinct P450s able to catalyse different reactions [4–6]. More recent studies involving the characterization of plant mutants allowed the identification of additional P450s catalysing FA oxidation (Table 1). The results of the studies presented in this minireview demonstrate the involvement of FAs in the synthesis of plant hydrophobic barriers and suggest their implication in several other biological processes.

Table 1. In vitro activity and physiological roles of characterized plant cytochrome P450s metabolizing fatty acids
Cytochrome P450PlantEnzymatic activityPhysiological roleRef
94A1Vicia sativaω-hydroxylationSignalling? catabolism?[13]
94A2Vicia sativaω- and in-chain hydroxylations?[15]
94A5Nicitiana tabacumω-hydroxylation, diacid formationSignalling? catabolism?[22]
94C1Arabidopsis thalianaω-hydroxylation, diacid formationSignalling? catabolism?[18]
86A1Arabidopsis thalianaω-hydroxylationSuberin synthesis[14]
86A2Arabidopsis thalianaω-hydroxylationCutin synthesis, defence[24]
86A8Arabidopsis thalianaω-hydroxylationCutin synthesis, pleiotropic role[23]
86A4Arabidopsis thalianaω-hydroxylationCutin synthesis[34]
86A22Petunia hybridaω-hydroxylationEstolide synthesis[30]
86A33Solanum tuberosumω-hydroxylationSuberin synthesis[29]
86B1Arabidopsis thalianaω-hydroxylationSuberin synthesis[26]
81B1Helianthus tuberosusin-chain hydroxylationSignalling? defence?[20]
703A2Arabidopsis thalianain-chain hydroxylationSporopollenin synthesis[37]
704B1Arabidopsis thalianaω-hydroxylationSporopollenin synthesis[38]
704B2Oryza sativaω-hydroxylationSporopollenin synthesis[39]
709C1Triticum aestivumin-chain hydroxylationSignalling? defence?[19]
726 A1Euphorbia lagascaeEpoxidationSeed oil synthesis[31]
77A4Arabidopsis thalianaEpoxidation in-chain hydroxylationSignalling? defence?[21]
77A6Arabidopsis thalianain-chain hydroxylationCutin synthesis[34]
92B1Petunia hybridaω-hydroxylationFlowering?[54]
78A1Zea maysω-hydroxylationFlowering?[55]

Identification of the first fatty acid hydroxylases

Typically, FA hydroxylation activities are barely detectable in whole-plant extracts even when using purified microsomal fractions as the enzyme source and radiolabelled FAs as substrates. This is because plant FA hydroxylases are either low in abundance or expressed in specific tissues and/or in response to stress [7]. This has represented a major difficulty to their identification and biochemical analysis. To circumvent this problem, researchers have taken advantage of two major features of P450s acting on FAs: (a) they can be regulated at the transcriptional level by biotic and abiotic stresses, and (b) they can be inactivated by FA analogues carrying an acetylenic function.

In a pioneering work on plant FA hydroxylases [8], abiotic stresses were used to compensate for the low level of enzyme present in plant tissue. Exposure of Vicia sativa seedlings to clofibrate, a hypolipidaemic drug, strongly induced ω-hydroxylase activities [8] and enabled biochemical investigations [9,10]. In microsomes of clofibrate-treated seedlings, inhibition of lauric and oleic acid hydroxylation by a C18 FA carrying a terminal acetylenic function followed different kinetics. This was the first demonstration of the existence of distinct FA ω-hydroxylases in a single species. Weissbart et al. [10] demonstrated that 11-dodecynoic acid irreversibly inhibited ω-hydroxylation of lauric acid in V. sativa microsomes. The mechanism of inhibition has been explored using FA-metabolizing P450s from animals. In these P450s, inactivation results from heme [11] or protein alkylation [12]. Postulating that inhibition of plant P450s followed the same mechanism, inhibition by a FA with an acetylenic function was used to clone the first plant FA ω-hydroxylase (CYP94A1). Incubation of [1-14C]-11-dodecynoic acid with V. sativa microsomes allowed covalent tagging of a protein responsible for ω-hydroxylase activity on lauric acid. An internal peptide sequence from the radiolabelled protein was determined and a full-length cDNA was isolated [13].

Another strategy based on conserved motifs led to the identification in Arabidopsis thaliana of a second plant FA hydroxylase. Using the similarity of an Arabidopsis expressed sequence tag (EST) sequence and the consensus sequence of the fungal CYP52 family of alkane hydroxylases and the mammalian CYP4A family of alkane hydroxylases, a cDNA was cloned and shown to encode a member (CYP86A1) of a new plant P450 family. When expressed in yeast, CYP86A1 displayed ω-hydroxylase activity on a range of saturated and unsaturated C12-C18 fatty acids but not on C16 alkane [14].

Catalytic properties and active site

Plants P450s can catalyse different types of reactions using FAs as substrates. The most typical products of reactions are ω- and in-chain hydroxy fatty acids, but FAs can also be epoxidized, transformed to dicarboxylic FA (Fig. 1) or in the case of hydroperoxides, cleaved to shorter aliphatic compounds (see below). Few studies have addressed what determines substrate specificity and regioselectivity. One has to keep in mind that most of the knowledge concerning these two aspects of FA hydroxylase enzymology comes from biochemical studies performed after heterologous expression. Also, activity measurements were performed with free FA, but natural substrates are very often not known and might be different (acyl-CoAs, glycerolipids, etc.).

Figure 1.

 Metabolization of oleic acid and its derivatives by plant P450s. Reactions presented in this scheme have been described after incubation of oleic acid with microsomes of yeast expressing distinct P450s [18,19,21]. (A) Oleic acid, (B) 18-hydroxyoleic acid, (C) octadec-9-en-18-al-oic acid, (D) octadec-9-en-1,18-dioic acid; (E) 9,10-epoxystearic acid; (F) 17-hydroxyoleic acid.

From a thermodynamic point of view, oxidation of the terminal methyl of a fatty acid is disfavoured in comparison with oxidation of a secondary carbon. This implies that ω-hydroxylases possess a highly structured active site. Using site-directed mutagenesis, Kahn et al. [15] demonstrated that a conservative substitution of Phe494 in CYP94A2 cloned from V. sativa led to a shift in the regiospecificity of lauric acid hydroxylation from the ω- position to the ω-1 position. It was concluded that Phe494 supplies constraints that maintain the terminal methyl of lauric acid near the ferryl oxo species. The aliphatic nature of FAs makes it very likely that hydrophobic interactions are important for the positioning of the substrate in the active site. This was confirmed by Rupashinghe et al. [16] who studied five members of the CYP86 family from Arabidopsis. Using modelling based on the known crystal structure and 3D models of FA-metabolizing P450s in bacteria and animals, the authors confirmed that CYP86 models have a binding site packed with hydrophobic residues (Fig. 2). However, in some cases, the results also suggest the presence of polar residues in the binding site. This is the case for CYP94A1 [17], CYP94C1 [18] and CYP709C1 [19] which metabolize in vitro 9,10-epoxystearic acid with a higher efficiency than C18:1, C18:2 or C18:3. This suggests a strong interaction between the oxiran and a polar residue of the active site. The enantioselectivity of CYP94A1 for 9R,10S-epoxystearic acid supports this hypothesis. The high constraints on the substrate in the active site are also illustrated by the example of CYP77A4, which epoxidizes C18:2 to 12,13-epoxyoctadeca-9-enoic acid presenting a strong enantiomeric excess in favour of the 12S/13R enantiomer representing 90% of the epoxide. Some hydroxylases, e.g. CYP81B1 [20] or CYP77A4 [21], exhibit a regioselectivity depending on the aliphatic chain length. This is likely because of the anchoring of different substrates via interaction of the carboxyl group with the same polar residue of the enzyme. By contrast, some hydroxylases show strict regioselectivity: whatever the chain length, CYP94A1 [13] and CYP709C1 [19] exclusively attack the ω- and ω-1 position, respectively. This clearly indicates that in these cases, the carboxyl group does not interact with a specific polar residue of the active site.

Figure 2.

 Predicted substrate contact residues in all five CYP86A models with predicted conformation for oleic acid binding. Conserved residues are coloured in green, similar residues are coloured in gold, nonconserved residues are in elemental colours. The heme within the catalytic site and oleic acid (C18:0) are shown in space-filling format. From Rapusinghe et al. [16].

The structure of metabolites also depends on the chemical motif present on the substrate. CYP94A5 from tobacco [22] and CYP94C1 from Arabidopsis [18] catalyse the ω-hydroxylation of fatty acids, but oxidation of primary alcohol by these enzymes leads to the formation of dicarboxylic FAs. Using microsomal fraction of V. sativa, Weissbart et al. [10] showed that ω-hydroxylases form epoxides when the terminal carbon is engaged in an unsaturation. The same observation was made with in-chain hydroxylase from microsomes of Heliantus tuberosus [20]. These results were obtained with unsaturated analogues of lauric acid, but recently CYP77A4, an in-chain hydroxylase cloned from Arabidopsis, was shown to be able to epoxidize physiological C18:1, C18:2 and C18:3 [21].

Physiological role

Characterization of the first FA-metabolizing P450 enzymes showed that they exhibited different regioselectivities and substrate specificities and were differently regulated. This suggested that many P450s acting on FAs existed within the same species and acted on distinct substrates in a variety of biochemical pathways. EST and genome sequencing projects demonstrated that P450s acting on FAs belonged to multigenic families and analysis of mutants confirm their involvement in several biological processes (Table 1).

Cutin and suberin biosynthesis

FA ω-hydroxylation

The first plant gene encoding a FA ω-hydroxylase, CYP86A1, was identified in A. thaliana based on sequence homology with ω-hydroxylases from mammals and yeast. The encoded protein was characterized after heterologous expression [14]. A protein from the same subfamily, CYP86A8, was the first FA ω-hydroxylase for which a mutant (lacerata) was isolated [23]. The presence of a maize transposon in the coding sequence of CYP86A8 led to a pleiotropic mutant phenotype with organ fusion, altered cell differentiation, reduced apical dominance and delayed senescence. The organ fusion phenotype observed in lacerata was similar to that of Arabidopsis overexpressing a cutinase. This was consistent with implication of this enzyme in Arabidopsis cutin synthesis. TEM of the epidermis showed that the structure of the cuticle was altered, which suggested that some of these phenotypes were due to a defect in the epidermal cuticle. It was thus proposed that a major role of lacerata was the production of the omega-hydroxy FA constitutive of the cutin polymer matrix of the cuticle. Reverse genetics approaches in Arabidopsis and potato enabled investigation of the putative involvement of other members of the CYP86 family in the synthesis of cutin and the other major FA-based polyester of plants (suberin). The att1 mutant disrupted in CYP86A2 [24] displayed increased sensitivity to the bacteria Pseudomanas syringae, and water loss and a disorganized cuticle structure, which was consistent with a role in cutin biosynthesis. The Arabidopsis horst mutant, impaired in the coding sequence of CYP86A1, showed a total aliphatic root suberin content that was reduced to 60% compared with wild-type [25]. This reduction was because of the strong decrease of C16 and C18 ω-hydroxyacids and corresponding diacids. This observation corroborates in vitro studies performed by Benveniste et al. [14] who showed that microsomal preparations of yeast expressing CYP86A1 metabolized C16 and C18 with high efficiency compared with other FAs tested. The fact that the content in the saturated very long-chain (C22–C24) omega-hydroxy FA of suberin was not affected by in CYP86A1 knockouts indicated that synthesis of these monomers was due mostly or completely to proteins other than CYP86A1.

Coexpression of a second member of the CYP86 family, CYP86B1, using suberin biosynthetic genes and in silico gene expression analysis of its tissue specificity, suggested its possible implication in the syntesis of suberin monomers. This was confirmed by study of the ralph mutant possessing an alteration in the coding sequence of CYP86B1 [26]. This mutation resulted in a strong reduction of C22 and C24 ω-hydroxy saturated FA derivatives in root suberin and seed coat polyesters. Surprisingly, downregulation of CYP86B1 did not impair the water-barrier function of root suberin and seed coat, showing that production of C16 and C18 ω-hydroxy FA by other hydroxylases was sufficient to maintain this major property of polyester-based barriers.

Although CYP86A1 and CYP86B1 seem to be responsible mostly for the synthesis of C16–C18 and C22–C24 ω-hydroxy FAs respectively, in plants co-overexpression of these enzymes with the GPAT5 acyltransferase of suberin synthesis shows that their specificity is probably in part overlapping. Indeed, ectopic overexpression of CYP86A1 in Arabidopsis resulted in the production of C22–C24 ω-hydroxy FA and diacids in stems cutin in addition to a major increase in C16–C18 omega-oxidized monomers [27]. The reverse situation was observed for CYP86B1 [28].

By contrast to CYP86A1, which produced, in vitro, the C16 and C18 monomers missing in the ralph mutant, microsomal incubation of C22 and C24 FAs with microsomes of yeast expressing CYP86B1 did not produce any metabolite [26]. The same observation was made with CYP86A33 involved in potato suberin production [29]. One explanation would be that CYP86B1 does not metabolize free FA, but rather esterified FA (acyl-CoAs, glycerolipids, etc.). In this respect, it is important to note that the demonstration of ω-hydroxylase-metabolizing esterified FA was recently achieved with CYP86A22 from Petunia [30] which ω-hydroxylates saturated and unsaturated acyl-CoA derivatives. In addition, it has also been shown that CYP726A1 from Euphorbia lagascae acts on FAs esterified to phosphatidylcholine to produce epoxy fatty acids [31].

Intracellular localization studies revealed that both proteins encoded by CYP86A1 and CYP86B1 localize in the endoplasmic reticulum, in agreement with the majority of plant FA hydroxylases [32]. It is noteworthy that the endoplasmic reticulum is also the major location of C22 and C24 FA synthesis, which consists of the elongation of FAs exported from chloroplasts.

FA in-chain hydroxylation

Using biochemical assays in Vicia faba, Soliday and Kolattukudy [2,33] demonstrated three decades ago the involvement of at least two distinct FA hydroxylases in formation of the major cutin monomer 10,16-dihydroxypalmitic acid in broad bean. Genetic and biochemical studies allowed Li-Beisson et al. [34] to show that in Arabidopsis CYP77A6 hydroxylated on positions 8, 9 and 10, the 16-hydroxypalmitic acid produced by CYP86A4. The sequential order of the hydroxylation reactions was demonstrated by both cutin monomer profiles in knockout mutants for CYP77A6 and CYP86A4 and the fact that recombinant CYP77A6 expressed in yeast was active on 16-hydroxypalmitic acid, but not on palmitic acid (Fig. 3). Null mutants in the gene coding for CYP77A6 still had 66% of cutin load compared with wild-type, but lacked the typical nanoridges present on the surface of flowers. These specialized structures have been suggested to help attract insect pollinators, giving CYP77A6 an indirect role in reproduction. It is possible that the introduction of in-chain hydroxyl allows cross-linking of cutin, leading to reticulation and strengthening of the polymeric envelope. The discovery of an in-chain hydroxylase producing polyhydroxy FAs possibly important for the formation of nanostructures increases the array of FA-modifying enzymes with biotechnological interest that originate from plant polyester metabolism [35].

Figure 3.

 Synthesis of 10,16-dihydroxypalmitic acid from palmitic acid in Arabidopsis thaliana. Putative pathway based on biochemistry and mutant analysis [34]. In vivo endogenous substrates (free fatty acids, acyl-CoAs and glycerolipids) are still unknown. (A) Palmitic acid, (B) 16-hydroxypalmitic acid, (C) 10,16-dihydroxypalmitic acid.

Sporopollenin biosynthesis

Sporopollenin is a major polymeric component of exine, the outer pollen wall, and represents a protective envelope fundamental for pollen resistance [36]. Its high resistance makes it difficult to study and its exact structure remains to be elucidated. Preservation of ancient pollen grains for millions of years illustrates its stability and its protective properties. The participation of FA in-chain and ω-hydroxylases in sporopollenin synthesis has recently been demonstrated with the study of Arabidopsis mutants. Arabidopsis CYP703A2 [37] and CYP704B1 [38] knockout lines produced non-maturated pollen grain lacking the normal exine layer. Heterologous expression in yeast cells showed that CYP703A2 and CYP704B1 are FA in-chain and ω-hydroxylases, respectively. The first preferentially catalyses the hydroxylation of lauric acid (C12) at position 7, whereas the second hydroxylates the ω position of C18 FAs. A second member of the CYP704B subfamily has been described in rice [39]. One mutant line generated by treatment with 60Co displayed complete male sterility. Exine was absent on the pollen grain and analysis revealed a drastic loss of cutin monomers in cyp704B2 anthers. The capacity of CYP704B2 to ω-hydroxylate C16 and C18 FA was demonstrated after heterologous expression in the yeast system.

Plant defence

Biochemical and genetics evidence indicates the involvement of FA hydroxylases in plant defence. Parker and Köller [40] showed that bean infection by Rhizoctonia solani was decreased when leaves were treated with cutinases releasing ω-hydroxy fatty acids. The protection mechanism remains to be elucidated, but it has been established that pathogen-challenged plants perceive hydroxy FAs as key compounds in the induction of resistance [41,42]. These compounds also induce elicitation of H2O2 production [43]. It is noteworthy that 9,10,18-trihydroxystearic and 18-hydroxy-9,10-epoxystearic acids exhibit the strongest effect in eliciting defence events. These FA derivatives are produced by CYP94A1 from V. sativa suggesting a potential role in plant defence for this enzyme. This hypothesis receives support from experiments showing that treatment of V. sativa seedlings with the stress hormone methyl jasmonate enhanced CYP94A1 at the transcriptional level [44]. Interestingly, clofibrate treatment also enhanced CYP94A1 at the transcriptional level [13] and increased the proliferation of peroxisomes [45] which have an important role in responses to pathogens [46]. Induction of mammalian ω-hydroxylases by clofibrate and peroxisome proliferation occur via activation of a peroxisome proliferator-activated receptor. This peroxisome proliferator-activated receptor can be activated by clofibrate [47] and by FA derivatives such as prostaglandins. Clofibrate produces similar effects in plants and animals [8,45]. Furthermore, there are evident structural analogies between prostaglandins and jasmonates, which are both polyunsaturated FA derivatives involved in response to stress. All these similarities strongly suggest that the mechanisms of ω-hydroxylase regulation by clofibrate and FA derivatives are conserved between plants and animals.

Study of the Arabidopsis att1 mutant confirmed the implication of FA ω-hydroxylases in plant defence [24]. Pseudomonas syringea caused a more severe disease in att1 than in wild-type. This resulted from the induction of type III genes necessary for parasitism, by a still unknown process. In Arabidopsis, expression of five members of the CYP86A subfamily involved in FA ω-hydroxylation was monitored by micro-array and RT-PCR analysis [48]. They were found to be expressed at different constitutive levels and their expression varied with organs. They also responded differently to chemicals and environmental stresses. Sequence analysis of the promoters revealed cis-elements present in the promoters of other plant genes that correlated with gene response.

Implication in plant defence events is not restricted to FA ω-hydroxylases. CYP709C1 is the first subterminal hydroxylase of long-chain FAs characterized in plants [19]. This enzyme exclusively attacks ω-1 and ω-2 carbons, and in the context of plant defence, it is interesting to note that ω-1 hydroxy derivatives of FA have been described. Volicitin the ω-1 hydroxy linolenic acid coupled to glutamine [49] is responsible for the majority of elicitor activity present in the oral secretion of caterpillar species feeding on plants. It is conceivable that products of reactions catalysed by CYP709C1 have eliciting properties or are precursors of molecules with eliciting properties. Studies of substrate specificity showed that among the FAs tested, CYP709C1 metabolized 9,10-epoxystearic with the highest efficiency [19]. Hydrolysis of the resulting 17-hydroxy-9,10-epoxystearic acid by epoxide hydrolase would lead to the formation of a trihydroxy FA with a chemical structure close to that of compounds having antimicrobial properties [19]. The strong and rapid induction of CYP709C1 by methyl jasmonate is also in favour of its participation in plant defence. In Arabidopsis, interplay of the epoxidase CYP77A4 with other enzymes also accounts for the formation of poly(hydroxy FA). CYP77A4 can produce vernolic acid which is converted to diol by epoxide hydrolase [21]. Hydroxylation of this diol by an ω-1 hydroxylase present in Arabidopsis [18] produces 12,13,17-trihydroxyoctadeca-9-enoic acid (Fig. 4), which has been shown to possess antifungal properties [50].

Figure 4.

 Formation of antifungic 12,13,17-trihydroxyoctadeca-9-enoic acid in microsomes of Arabidopsis thaliana. Putative pathway leading to the formation of 12,13,17-trihydroxyoctadeca-9-enoic acid based on distinct enzymes characterized in Arabidopsis [18,21]. (A) Linoleic acid, (B) vernolic acid, (C) 12,13-dihydroxyoctadeca-9-enoic acid, (D) 12,13,17-trihydroxyoctadeca-9-enoic acid.

The importance of P450s metabolizing enzymes in plants is illustrated by the atypical P450 family CYP74. This family has been subjected to a tremendous amount of work. For a detailed discussion the reader in invited to refer to the specific review by Stumpe and Feussner [51]. Briefly, contrary to the majority of P450s, members of this family catalyse oxidative reactions without O2 and NADPH-cytochrome P450 reductase. Three catalytic activities have been assigned to these enzymes: allene oxide synthase, hydroperoxide lyase and divinyl ether synthase (Fig. 5). CYP74A1 with allene oxide synthase activity from Arabidopsis was the first identified and characterized enzyme of the octanoid pathway leading to jasmonates [7]. The so-called oxylipins (jasmonates, aldehydes, divinyl ether, alcohols) generated by CYP74 members are signalling molecules as well as molecules exhibiting antimicrobial and antifungal properties.

Figure 5.

 Metabolism of polyunsaturated fatty acid by lipoxygenase and CYP74. LOX: lipoxygenase; HPL: Hydroperoxide lyase; AOS: Allene oxide synthase; DES: Divinyl ether synthase. Simplified from Stumpe and Feussner [51].

FA catabolism

No direct involvement of plant ω-hydroxylases in FA catabolism has been demonstrated. However, by analogy to knowledge concerning FA ω-hydroxylases in mammals and in microorganisms, it can be assumed that plant ω-hydroxylases could be major actors in this process. This is particularly relevant for ω-hydroxylases that are upregulated by a class of compounds (i.e. clofibrate) [13,48] known to induce peroxisome proliferation in mammals. In mammals, ω-hydroxy derivatives of FA produced by members of CYP4 family can be further oxidized to dicarboxylic acids either by dehydrogenases or by P450s able to perform the complete oxidation of a methyl to a carboxyl group. In both pathways, the resulting dicarboxylic acid can be eliminated by β-oxidation in peroxisome. The key role of ω-hydroxylases in FA catabolism has also been established for yeast belonging to the genus Candida. Members of the CYP52 family enable Candida maltosa to grow on media containing aliphatic hydrocarbons as a sole source of carbon and energy (reviewed in [18]). In plants, besides enhancement of FA hydroxylases at the transcriptional level, clofibrate [13,48] similarly to what is observed in mammals, also induces the proliferation of peroxisomes [45] in which FA β-oxidation is of primary importance for energy production [52]. CYP95A5 from tobacco [22] and CYP94C1 from Arabidopsis [18] are both capable of producing dicarboxylic FAs either by a two-step oxidation of ω-hydroxy FA or by a three-step oxidation starting from a FA (Fig. 1). It is postulated that dicarboxylic FAs are degraded much faster through β-oxidation than monocarboxylic FAs, and it has been proposed that ω-hydroxylases may function in deactivating FA-derived lipid signals and in rapidly turning over free FAs liberated by lipases during stress [52].

Reproduction

As mentioned above, maturation of pollen grains in Arabidopsis and in rice depends in part on FA ω-hydroxylase activity [37,38]. Synthesis of the protective envelope in the rice anther also requires active FA ω-hydroxylase [39]. Downregulation of the enzymes implicated in these processes leads to a male sterility phenotype, giving a key role for FA ω-hydroxylases in reproduction. It has been known for a long time that when present, stigma exudate is of primary importance in pollination events. Di- and triglycerides of Nicotiana tabacum are enriched in ω-hydroxy FA which are believed to be responsible for the recognition of stigma by pollen [53]. Recently, transgenic Petunia expressing CYP86A22-RNAi were produced and the lipid content of stigmas analysed [30]. Downregulation of CYP86A22 was accompanied by a drastic decrease in 18-hydroxyoleic and 18-hydroxylinoleic acids, which can even be lacking in some lines when they represent 96% of total stigma FA in wild-type. This was in agreement with the enzymatic activity determined after heterologous expression in insect cells, which showed that CYP86A22 was able to ω-hydroxylate C16 and C18 fatty acids activated by CoA. Histochemical analysis located CYP86A22 exclusively in the stigma, which is consistent with a specific role in flower development and reproduction. In the context of reproduction, two characterized FA ω-hydroxylases from Petunia CYP92B1 [54] and Zea mays CYP78A1 [55], and the partially characterized Petunia CYP703A1 [56], have also been shown to be preferentially expressed in developing inflorescence.

Phylogeny and evolution

In plants, P450s have duplicated and diverged in order to produce a tremendous array of compounds exhibiting sometimes very similar structures. Fatty acid hydroxylases follow this rule. CYP86A1 and CYP86B1 represent a good example of this evolution: they both hydroxylate the terminal methyl of acyl chains in the suberin biosynthesis pathways and seem to differ only by the chain length preference.

The CYP86 family is found in the genome of the moss Physcomitrella patens in addition to the genomes of Arabidopsis, rice and poplar [57]. It thus seems that this family has played an important role in early plant evolution. This is consistent with the demonstrated role of CYP86As in the biosynthesis of the framework matrix of the plant cuticle, a structure that is thought to have played a great role in the adaptation of early plants to life in a terrestrial desiccating environment. CYP94 family members also duplicated during evolution of land plants while retaining their catalytic activity on FAs. The biological role of the CYP94 family is still unknown, however.

The CYP703 family is also conserved in land plants and typically each plant species contains only one CYP703 [37]. CYP704B1 and CYP704B2 belong to a conserved and ancient P450 family in moss and seed plants. This suggests that reactions catalysed by CYP703 and by members of the CYP704 subfamily represent a key step in exine formation during plant evolution.

The apparent absence of the CYP77 family in the moss genome [58] might indicate that cuticles based on cutin polyesters containing in-chain hydroxylated FAs represent a more recent type of hydrophobic barriers. The selective advantage conferred by cutins rich in polyhydroxy FAs such as dihydroxypalmitates is unknown, but it can be speculated that it is related to the evolution of distinctive epidermal surface structures such as flower nanoridges. Alternatively, it might be that polyhydroxy FA-rich cutins have improved hydrophobic barrier properties.

Conclusion

Plants have developed a highly complex metabolic network using the diversified catalytic properties of the P450s [57]. Among these P450s, FA hydroxylases are major actors involved in many aspects of plant biology. When plants conquered dry land ∼ 400 million years ago, one major problem they had to face was to resist to dessication. They have developed cutin and suberin two biopolymers constituted of FA and FA plus phenolics, respectively. By introducing hydroxyl function in to monomers, FA hydroxylase will allow their condensation and elongation of the polymers. Ensuring reproductive success was also of primary importance for land conquest. Cutin of the anther wall and sporopollenin represent protective envelopes and the key role of FA hydroxylases in their synthesis is now well established. Plants are sessile organisms and rely on a battery of chemicals for survival. Lipid metabolism is a major player in the defence network of plants and it is tempting to speculate that hydroxyls, epoxides of FA as well as dicarboxylic FA have properties similar to those described for FA derivatives generated by members of the CYP74 family implicated in plant defence [51]. Similar to what is known for mammals and microorganisms, FA ω-hydroxylation in plants could be the starting point for their catabolism leading to energy production required for general development and plant defence. More studies are needed to confirm and elucidate the physiological meanings of P450 metabolizing FAs in plants as well as in animals [59] and microorganisms [60]. Concerning the plant kingdom, a major part of this task should be achieved via the study of plant mutants.

Acknowledgements

FP was supported in part by a grant from KBBE 2009 program (grant ANR-09-KBBE-006-001). FB was supported in part by a grant from the 7th European Community Framework Program (Marie Curie International Reintegration Grant 224941).

Ancillary