A peroxygenase pathway involved in the biosynthesis of epoxy fatty acids in oat ( Avena

While oat has long been known to produce epoxy fatty acids in seeds, synthesized by a peroxygenase pathway, the gene encoding the peroxygenase remains to be determined. Here we report identification of a peroxygenase cDNA AsPXG1 from developing seeds of oat. AsPXG1 is a small protein with 249 amino acids in length and contains conserved heme-binding residues and a calcium-binding motif. When expressed in Pichia pastoris and E. coli , AsPXG1 catalyzes the strictly hydroperoxide-dependent epoxidation of unsaturated fatty acids. It prefers hydroperoxy-trienoic acids over hydroperoxy-dienoic acids as oxygen donors to oxidize a wide range of unsaturated fatty acids with cis double bonds. Oleic acid is the most preferred substrate. The acyl carrier substrate specificity assay showed phospholipid and acyl-CoA were not effective substrate forms for AsPXG1 and it could only use free fatty acid or fatty acid methyl esters as substrates. A second gene, AsLOX2 , cloned from oat codes for a 9-lipoxygenase catalyzing the synthesis of 9-hydroperoxy-dienoic and 9-hydroperoxy-trienoic acids, respectively, when linoleic (18:2-9c,12c) and linolenic (18:3-9c,12c,15c) acids were used as substrates. The peroxygenase pathway was reconstituted in vitro using a mixture of AsPXG1 and AsLOX2 extracts from E. coli . Incubation of methyl oleate and linoleic acid or linolenic acid with the enzyme mixture produced methyl 9,10-epoxy stearate. Incubation of linoleic acid alone with a mixture of AsPXG1 and AsLOX2 produced two major epoxy fatty acids, 9,10-epoxy-12-cis-octadecenoic acid and 12,13-epoxy-9-cis-octadecenoic acid, and a minor epoxy fatty acid, probably 12,13-epoxy-9-hydroxy-10-trans-octadecenoic acid. AsPXG1 predominately catalyzes intermolecular peroxygenation. by the in vitro assay. The results showed that AsPXG1 could only use free fatty acid and fatty acid methyl esters as substrates; whereas phospholipid and acyl-CoA were not effective 12,13-epoxy-18:1-9c was presumably derived from epoxidation of the Δ 12 double bond of linoleic acid using 13-OOH-18:2-9c,11t as co-oxidant. 9,10-epoxy-18:1-12c was derived from epoxidation of the Δ 9 double bond of linoleic acid using 13-OOH-18:2-9c,11t as co-oxidant. 13-OH-9,10-epoxy-18:1-11t was derived from epoxidation of the Δ 9 double bond of 13-OH-18:2-9c,11t . Again, these results confirmed that the epoxidation occurred preferentially at the Δ 12 double bond of linoleic acids.


INTRODUCTION
Peroxygenase is a hydroperoxide-dependent oxygenase that catalyzes the transfer of one oxygen atom from a hydroperoxide to a substrate that is oxidized (Blee et al., 1993). Unlike cytochrome P450 epoxygenase, peroxygenase does not require cofactors such as NAD(P)H and molecular oxygen for substrate epoxidation (Blee et al., 1993;Matsunaga and Shiro 2004).
Peroxygenase activity was first detected in oat (Avena sativa L.) when 9-and 13-hydroperoxyoctadecadienoic acids were reduced to the corresponding alcohols and 9,10-epoxy-13hydroxyoctadecenoic acids or 12,13-epoxy-9-hydroxyoctadecenoic acids was formed (Heimann and Schreier 1970;Heimann and Dresen 1973). However, this reaction was then believed to be catalyzed by a "lipoperoxidase activity" or "hydroperoxide isomerase". In 1977, peroxygenase was first defined by a labeling study using pea microsomes (Ishimaru and Yamazaki 1977) as hydroperoxide-dependent monooxygenase, which differs from typical cytochrome P450 monooxygenase in the oxygen donor. Since then, this type of enzyme activity has been detected in various microorganisms and plants species (Blee et al., 1993;Matsunaga and Shiro 2004). In oat, peroxygenase-catalyzed fatty acid epoxidation/hydroxylation was confirmed using a seed microsomal fraction (Hamberg and Hamberg 1996). In 2006, the first peroxgenase gene was cloned from Arabidopsis thaliana using the sequence information of a partially purified peroxygenase from oat (Hanano et al., 2006). The encoded Arabidopsis peroxygenase shares no sequence similarity with either peroxidase or cytochrome P450, but rather with "caleosin", a small oil body-associated protein with both heme-and calcium-binding motifs in plant seeds.
The peroxygenase activity is strictly dependent on the binding of calcium and ferric-heme (Hanano et al., 2006). Association of the peroxygenase activity with caleosin is surprising. Caleosin was first described as an oil body protein which, like oleosin, was believed to be imbedded in the phospholipid monolayer of lipid particles playing solely a structural role in maintaining the integrity of oil bodies (Chen et al., 1999). However, with the discovery of the biochemical activity, this type of protein has been widely identified in fungi and other plant species and is localized not only in oil bodies, but also in non-lipid storage organelles and tissues (Murphy 2005;Partridge and Murphy 2009).
Oat has been known to produce several epoxy fatty acids in seeds, such as 9,10-epoxy-18:0, 9,10-epoxy-18:1-12c and 12,13-epoxy-18:1-9c together accounting for approximately 0.5 % of the total fatty acids. These fatty acids are mainly present in neutral lipids such as triacylglycerols and diacylglycerols (Leonova et al., 2008;Doehlert et al., 2010). Oat has also been extensively used as an enzyme source to study their biosynthesis, catalyzed by peroxygenase activity (Heimann and Schreier 1970;Heimann and Dresen 1973;Hamberg and Hamberg 1996;Leonova et al., 2008). Although the exact biological function of these fatty acids is not defined, they are perceived to act as precursors for the synthesis of oxylipins, a group of oxygenated fatty acids, in response to biotic and abiotic stress (Feussner and Wasternack 2002;Andreou et al., 2009) or as monomers for the biosynthesis of cutin polymers to cover aerial parts of the plant surface providing the hydrophobic structural barrier to the environment (Lequeu et al., 2003). While the activity of peroxygenase has been well defined in oat in planta, the gene encoding the enzyme remains unknown. Thus, the primary goal of this research is to clone the peroxygenase gene and characterize the peroxygenase pathway in vitro that is involved in the biosynthesis of epoxy fatty acids in oat.

Identification of cDNAs encoding fatty acid peroxygenase and lipoxygenases from A. sativa
To identify a cDNA encoding fatty acid peroxygenase from oat, A. sativa EST databases prepared from leaves, roots and developing seeds at different stages that are publically available from NCBI were searched using Arabidopsis ATS1, an embryo-specific caleosin (Nuccio and Thomas 1999) as a query sequence. The search resulted in identification of ESTs representing six distinct caleosin-like sequences in oat. All these sequences appeared only in EST libraries prepared from developing seeds and were represented by only 1 or 2 ESTs. An exception was a single sequence in an EST collection from developing seeds at the watery stage which was represented by 34 ESTs. The sequence assembly using these highly abundant ESTs gave rise to a contig with 1,073 bps in length (AsPXG1). The sequence analysis of this contig revealed an open reading frame (ORF) with the putative start and stop codons at positions 63 and 810, respectively, and an in-frame stop codon at 30 nucleotides upstream the start codon. To clone the full-length cDNA, total RNA isolated from developing seeds of A. sativa was used as template for RT-PCR amplification with two specific primers containing the putative start and stop codons.
Sequencing of amplified PCR products revealed three types of isomers which differed in length at the 5 ′ ends, giving three different ORFs with 750, 735, and 717 base pairs in length. The reason why three different isoforms were amplified by a single pair of specific primers was not clear. Possibly they were spliced alternatively from the same transcript or transcribed from different loci/alleles that are highly homologous, as oat is a hexaploid. The longest ORF corresponded to the contig from the EST database and coded for a polypeptide of 249 amino acids with a predicted molecular mass of 28.1 kD (AsPXG1). Two conserved histidine residues (His-76 and His-144) were found in AsPXG1 which were believed to be involved in hemebinding. The EF-hand motif for calcium binding was located between position 80 and 93. The sequence of 19 amino acids between position 10 and 29 (AVVVSDAMSSVAKGAPVTAQ) of AsPXG1 was identical to the sequenced fragment of the protein purified from oat lipid particles previously which was used to identify the first peroxygenase gene in Arabidopsis (Hanano et al., 2006) (Figure 1). Sequence comparison showed that AsPXG1 shares high amino acid sequence identity to abscisic acid-induced EFA 27 (66%) from rice, AtPXG1 (At4g26740) (60%) and AtPXG2 (At5g55240) (58%) from A. thaliana ( Figure S1). The hydropathy analysis indicated that AsPXG1, similar to group I caleosins, has a hydrophobic membrane-associated domain in the middle of the protein sequence ( Figure 1) (Hanano et al., 2006).
To identify cDNAs encoding fatty acid lipoxygenase from oat, the same A. sativa EST databases were searched using Arabidopsis lipoxygenase 9-LOX (AtLOX5, Q9LUW0) and 13-LOX (AtLOX2, P38418) (Bell and Mullet 1993;Vellosillo et al., 2007) as query sequences. This resulted in identification of two ESTs from developing seed libraries. One was 700 bps long, while the other was 287 bps in length. To clone the full-length cDNAs, 5`RACE and 3`RACE methods were employed using mRNA isolated from developing seeds as templates. RACE amplifications identified two full-length cDNAs, AsLOX1 and AsLOX2, encoding two distinct putative lipoxygenases. Both cDNAs had ORFs of 2,586 bp long coding for a polypeptide of 862 amino acids with predicted molecular mass of approximately 97 kDa. AsLOX1 and AsLOX2 shared amino acid sequence identity at 74% and exhibited high sequence identity to lipoxygenases from Oryza sativa (P29250) (77% and 78%), Hordeum vulgare (P29114) (73% and 90%) and Zea mays (Q9LKL4) (75% and 74%, respectively) ( Figure S2). Phylogenic analysis indicated both AsLOX1 and AsLOX2 belong to type I lipoxygenase which does not possess any chloroplast transit peptide. Unlike type II lipoxygenase that is associated with chloroplasts and usually introduces a molecular oxygen at the C13 position of polyunsaturated acids, type I lipoxygenase is believed to be extraplastidic and introduce a peroxide group at the C9 position of linoleic and linolenic acids (Feussner and Wasternack 2002).

Functional characterization of the putative peroxygenase in yeast
To functionally characterize the putative peroxygenase, the coding region of AsPXG1 was cloned into the yeast expression vector pPICZ-B under control of the AOX1 promoter. The recombinant plasmid was introduced into Pichia pastoris X33. In presence of cumene hydroperoxide as oxygen donor and oleic acid as a substrate, cell-free extracts of AsPXG1/X33 expressing the putative peroxygenase produced a new fatty acid, compared with the control, with the same chromatographic retention time as 9,10-epoxy stearate (9,10-epoxy-18:0) ( Figure 2A). GC/MS analysis confirmed that the mass spectrum of this fatty acid is identical to that of authentic cis-9,10-epoxyoctadecanoic acid ( Figure 2B) with a molecular ion at m/z 312 and two prominent diagnostic fragments at m/z 199 and m/z 155. The m/z 199 fragment was assigned to the carboxyl-containing ion formed by the cleavage between carbon 10 and 11, while the m/z 155 fragment was the methyl terminal ion formed by the cleavage between carbon 8 and 9. In addition, it was found that formation of the epoxy fatty acid catalyzed by AsPXG1 was strictly hydroperoxide-dependent and sensitive to β-mercaptoethanol, a known peroxygenase inhibitor ( Figure S3). These results indicated AsPXG1 is the long-sought gene encoding a peroxygenase in oat that catalyzes strictly hydroperoxide-dependent monooxygenation of unsaturated fatty acids, giving rise to corresponding epoxy fatty acids.

Enzymatic properties of AsPXG1
Oat peroxygenase is considered to be a membrane-bound monooxygenase associated with either microsomes or lipid particle membranes (Hanano et al., 2006). To confirm whether AsPXG1 is associated with microsomal membrane or lipid particles, the yeast crude extracts of AsPXG1/X33 were fractionated into soluble, lipid particle and microsomal portions by centrifugation and the three subfractions were assayed. The result showed that the lipid particle and microsomal fractions possessed high levels of peroxygenase activities (415.4±30.6 and 50.6±3.2 nmol/min/mg, respectively) which were approximately 20 and 2.5 times that of the soluble fraction (19.5±1.0 nmol/min/mg). In addition, no activity was observed in the microsomal fraction when emulphogene, a detergent that could effectively solubilize the peroxgenase from membrane (Hanano et al., 2006), was added to the lysis buffer. Rather, all the peroxygenase activity was observed in the detergent solubilized supernatant fraction. These results indicate that, consistent with the result from the hydropathy analysis, AsPXG1 is a membrane-associated enzyme that might be located at both lipid particles and endoplasmic reticulum. However, it could be readily solubilized from the membrane by a detergent and the solubilized protein retained enzyme activity.
To further characterize the enzymatic properties of AsPXG1, the crude extract of AsPXG1/X33 was subjected to the activity assay under different temperature and pH buffers. showed that AsPXG1 exhibited activity at a range of pH (5 to 9) and temperature (15°C to 60°C) with an optimal pH value of 7 and an optimal temperature of 45°C ( Figure S4 and S5, respectively).
To identify the acyl carrier substrate specificity of AsPXG1, different types of oleic derivatives such as free fatty acid, fatty acid methyl ester, fatty acyl-CoA and phosphatidylcholine were examined by the in vitro assay. The results showed that AsPXG1 could only use free fatty acid and fatty acid methyl esters as substrates; whereas phospholipid and acyl-CoA were not effective substrate forms (see supplementary materials Figure S6). The ability to use fatty acid methyl ester as substrate by AsPXG1 provides a simple and sensitive way to evaluate the activity by direct chromatographic analysis of reaction products without fatty acid methyl ester derivatization.
To examine whether AsPXG1 is also functional in E. coli, the coding region of AsPXG1 was cloned into the bacterial vector pET15b. The recombinant plasmid was introduced into the Rosetta2(DE3)pLysS strain. The in vitro assay was conducted under the same condition used in oxygen donor ( Figure S7). Functional expression of AsPXG1 in E. coli provides us with the opportunity to reconstitute the two-gene peroxygenase pathway in the bacterium (see below).
www.plantphysiol.org on August 26, 2017 -Published by Downloaded from Copyright © 2011 American Society of Plant Biologists. All rights reserved.

Tissue-specific expression of oat lipoxygenase and peroxygenase genes
As shown in Figure 7, AsPXG1 was exclusively expressed in developing seeds, not in other tissues examined such as roots, leaves, glumes and germinating seeds, whereas AsLOX2 was expressed in all tissues examined with relatively higher expression in developing seeds, roots and germinating seeds.

DISCUSSION
There are three mechanisms underlying the biosynthesis of epoxy fatty acids in plants. The first one was identified by Stymne and colleagues which involves a Δ12 desaturase-like oxygenase epoxidizing the Δ12 double bond of linoleic acid linked to phosphatidylcholine, giving rise to 12,13-epoxy-18:1-9c (vernolic acid) in Crepis palaestina (Lee et al., 1998). The second one was elucidated by Cahoon and colleagues which involves a cytochrome P450-like oxygenase epoxidizing the same substrate, giving the same product in Euphorbia lagascae (Cahoon et al., 2002). The peroxygenase system is the third mechanism involved in the biosynthesis of epoxy fatty acids in plants. Unlike the first two mechanisms that use molecular oxygen as oxidant and phospholipid as acyl carrier substrate, the peroxygenase uses an oxygen atom from fatty acid hydroperoxide as oxidant to epoxidize substrates. Our in vitro study on the acyl carrier substrate specificity showed that AsPXG1 could only use free fatty acid or fatty acid methyl esters as substrates, while phospholipid and acyl-CoA were not effective substrate forms. Free fatty acid is likely the biological substrate for AsPXG1 as fatty acid methyl ester is not a viable form in the cells. Oat has been known to produce several epoxy and hydroxyl fatty acids in seeds synthesized by a peroxygenase pathway, and has thus served as a model system to study enzymatic properties of the peroxygenase (Hamberg and Hamberg 1996). While the oat peroxygenase activity that catalyzes the biosynthesis of these fatty acids has been well characterized in planta, the gene encoding the peroxygenase has not been cloned and characterized in vitro in detail. This study identified a peroxygenase cDNA AsPXG1 from oat developing seeds using the sequence similarity search of EST databases prepared from different tissues of oat plants with an Arabidopsis caleosin as query sequence. AsPXG1 is a small protein with 249 amino acids in length and contains conserved heme-binding residues and calcium-binding motif. Like type I caleosin, AsPXG1 has a hydrophobic domain in the middle of the sequence (Hanano et al., 2006).

Consistent with this observation, activity of the recombinant enzyme AsPXG1 in Pichia was
found to be primarily associated with the lipid particle and microsomal fraction. However, a substantial amount of peroxygenase activity was also found in the soluble fraction. This proportion of activity might result from the process of enzyme preparation, as AsPXG1 contains only a single 20 amino acid membrane-associated domain and the weak association of AsPXG1 with membrane might render it being easily solubilized into the supernatant fraction. This assumption is supported by observation that AsPXG1 is be readily solubilized from the microsomal membrane by the emulphogene detergent.
When expressed in either Pichia or E. coli, AsPXG1 catalyzes strictly hydroperoxide-dependent epoxidation of unsaturated fatty acids giving rise to corresponding epoxy fatty acids with high selectivity for the substrate to be oxidized as well as the co-substrate to be reduced. Under the assay condition used in this study, AsPXG1 significantly prefers hydroperoxy-trienoic acids over hydroperoxy-dienoic acids as oxygen donors to oxidize a very wide range of unsaturated fatty acids with cis-double bonds. Oleic acid is the most preferred substrate to be epoxidized by the peroxygenase. With the broad substrate selectivity, high stereospecificity and low regioselectivity, AsPXG1 can account for the synthesis of all the epoxy fatty acids in oat seeds.
In this report, we have also cloned two lipoxygenase-like genes, AsLOX1 and AsLOX2 from oat seeds. AsLOX1 showed no lipoxygenase activity when expressed in E. coli. Non-function of AsLOX1 might be due to potential mutations already existing in the native sequence or accidently introduced during cloning process. However, other possibility such as a low level of expression or misfolding of the expressed protein in E. coli cannot be excluded. AsLOX2, when expressed in E. coli, catalyzes the synthesis of 9-hydroperoxy-dienoic and 9-hydroperoxy-trienoic acids when linoleic and linolenic acids were used as substrates, indicating it codes for 9-lipoxygenase.
Unlike 13-lipoxygenase, AsLOX2 does not contains a transit peptide for chloroplast targeting; thus, it is presumably extraplastidic. Previous enzymatic assays in planta showed that the peroxygenation in oat seeds is almost exclusively derived from 9-lipoxygenase (Hamberg and Hamberg 1996). Thus, AsLOX2 identified in this report might be the lipoxygenase in oat seeds.
The peroxygenase pathway constitutes one branch of the lipoxygenase pathway where oxygenation of a polyunsaturated fatty acid by lipoxygenase gives rise to corresponding fatty acid hydroperoxide, which is then used by peroxygenase as oxygen donor to oxidize an unsaturated fatty acid (Hanano et al., 2006). This study reconstituted the peroxygenase pathway level of methyl 9,10-epoxy stearate. The initial reaction in this reconstituted peroxygenase pathway catalyzes the formation of 9-hydroperoxy-dienoic acid by AsLOX2, which is then followed by the second reaction that catalyzes epoxidation of oleic acid by AsPXG1 using a 9hydroperoxy-dienoic acid or 9-hydroperoxy-trienoic acid as the co-oxidant. Incubation of a mixture of AsPXG1 and AsLOX2 with linoleic acid alone produced two major epoxy fatty acids, 9,10-epoxy-18:1-12c and 12,13-epoxy-18:1-9c, while structurally unconfirmed 9-OH-12,13epoxy-18:1-10t is only a minor product of the co-oxidation ( Figure 5). 9,10-epoxy-18:1-12c and 12,13-epoxy-18:1-9c are derived from inter-molecular peroxygenation, i.e. epoxidation of ∆ 9 and ∆ 12 double bonds of linoleic acid, respectively, using 9-hydroperoxy-dienoic acid as oxygen donor, while 9-OH-12,13-epoxy-18:1-10t can be derived from either intra-molecular or intermolecular peroxygenation (Figure 6). Intra-molecular peroxygenation occurs when 9hydroperoxy-dienoic acid is epoxidized using oxygen from the hydroperoxy group of the same molecule, while intermolecular peroxygenation occurs when 9-hydroxy-dienoic acid reduced from a 9-hydroperoxy-dienoic acid is epoxidized using oxygen atom from another 9hydroperoxy-dienoic acid molecule. Based on this analysis, we concluded that intermolecular peroxygenation of polyunsaturated fatty acids occurs predominately, if not exclusively, in the oat peroxygenase pathway involved in the biosynthesis of epoxy fatty acids.
The peroxygenase pathway has been perceived to be involved in the biosynthesis of oxylipins in response to environmental stress or in the biosynthesis of cutin polymers in the plant surface area (Hanano et al., 2006). In oat, AsPXG1 is exclusively expressed in developing seeds, not in any other tissues such as leaves, roots and germinating seeds. It is mainly associated with endoplasmic reticulum and lipid particles. AsLOX2, on the other hand, is widely expressed in oat tissues including developing seeds. It has been documented that lipoxylgenase can also be associated with lipid particles (Feussner et al., 1997). Thus, coordinately temporal and spatial expression of the two enzymes in the peroxygenase pathway would result in accumulation of a small amount of epoxy fatty acids in seed triacylglycerols. These stored oxygenated fatty acids could be used as precursors for synthesis of active oxylipins in germinating seeds, whereby inducing defensive systems against abiotic and biotic stress during seed germination and seedling establishment.

Construction of yeast expression plasmid of AsPXG1
The opening reading frame (ORF) of AsPXG1 was amplified by PCR using a forward primer DM194 containing an EcoRI restriction site and a Kozak consensus sequence (5 GCGAATTCATCATGGCGGAGGACGC-3 ′ ), and a reverse primer containing a XbaI site at the . The PCR product with expected size was purified, digested and cloned into pPICZ-B vector (Invitrogen, Carlsbad, CA) using EcoRI and XbaI sites. The resulting plasmid (pDM71) contained an additional sequence encoding 23 amino acid residues (FLEQKLISEEDLNSAVDHHHHHH) at the 3 ′ end of the cloned gene.

Expression of the AsPXG1 cDNA in Pichia pastoris
Ten micrograms of pDM71 containing AsPXG1 was linearized with DraI and transformed into Pichia pastoris X-33 (Invitrogen) by electroporation using an electroporator 2510 (Eppendorf).
The transformed cells were grown at 28°C for 24 hrs with shaking in BMG medium containing 100 mM potassium phosphate, pH 6.0, 1.34% yeast nitrogen base without amino acids, 4×10 -5 % biotin, and 1% glycerol. The cells were then centrifuged and resuspended in a BMM medium containing 100 mM potassium phosphate, pH 6.0, 1.34% yeast nitrogen base without amino acids, 4×10 -5 % biotin, and 1% methanol at an OD 600 of 1.0 to induce the expression. After 60 hrs, the resulting cultures were harvested and washed once with Tris buffer (50 mM Tris-HCl pH 7.5).
The pellet was resuspended in a homogenizing buffer containing 50mM Tris-HCl (pH 7.5) and 0.6 M sorbitol, and the cells were disrupted with glass beads using a Mini-Beadbeater (Biospect, South San Francisco, CA, USA) (three cycles of 1 min each). The homogenate was subfractionated into supernatant, microsome and lipid articles (Kamisaka and Nakahara 1994), and these subfractions were then used as the enzyme sources for the activity assay. Protein concentration was determined according to Bradford using bovine serum albumin as the standard (Bradford 1976).

Peroxygenase activity assays
Unless otherwise stated, peroxygenase assays were carried out in a volume of 200 µl containing 0.1 mg protein, 0.5 mM cumene hydroperoxide, 0.5 mM oleic acid, 50 mM Tris-HCl (pH 7.0) and 10% glycerol. The reaction mixture was incubated at 45°C for 15 min with 500 rpm shaking and stopped by the addition of 500 µl ethyl acetate containing 10 µl of glacial acetic acid. The fatty acid products were extracted twice with ethyl acetate and methylated with diazomethane.
The resulting fatty acid methyl esters (FAMEs) were analyzed by gas chromatography and/or gas chromatography mass spectrometry. Identity of the products was confirmed by comparing GC retention time and MS data with those of standards. For quantifying the products, 3 µg of 18:0 was used as an internal standard. The relative activity of substrates was determined by GC(FID) quantification of products vs substrate using the internal standard for each substrate and then comparison of its value to that of oleic acid substrate. The FAME samples were analyzed on an Agilent 6890N gas chromatography equipped with a DB-23 column (30-m ×0.25-mm) with 0.25-µm film thickness (J&W Scientific). The column temperature was maintained at 160°C for 1 min, and then raised to 240°C at a rate of 4°C/min.
The pH optimum of AsPXG1 was evaluated using 0.5 mM cumene hydroperoxide and 0.5 mM methyl oleate from pH 5.0 to 9.0 using one of three 50 mM buffers (MES, HEPES, and Tris-HCl) in 0.5-unit intervals. The optimal temperature was similarly evaluated from 15°C to 60°C in 5°C intervals at pH 7.0.
Apparent kinetic parameters were determined under optimum conditions. The range of substrate concentrations used to determine the kinetic parameters of 9-HPOD and 9-HPOT was 4-128 µM, while those for 13-HPOD and 13-HPOT were 4-130 µM and 3-107 µM, respectively. The kinetic constants were estimated from Lineweaver-Burke plots using the average of triplicate measurements.

Expression of peroxygenase and lipoxygenase in Escherichia coli
For activity assay of AsPXG in E. coli, coding region was reamplified with two primers  (Hornung et al., 1999) were amplified by RT-PCR using the RNA from cucumber germinating seeds (7 days was first grown at 37°C overnight in 10 ml Luria-Bertani (LB) medium containing either kanamycin (50 µg/mL) or ampicillin (100 µg/mL) along with chloramphenicol (34 µg/mL). The fresh culture was then inoculated into 50 volumes of LB medium containing the same antibiotics.
When the culture was grown at 37°C till OD 600 = 0.5 to 1.0, isopropyl-β-D-thiogalactopyranoside (IPTG) was added to a final concentration of 1mM to induce the expression at 30°C for 4 hrs.
After induction, the cells were harvested by centrifugation (5,000 rpm, 15 min 30 min with 500 rpm shaking. After the reaction, the resulting hydroperoxide products were reduced to their corresponding hydroxy derivatives with 25 mg/ml of Tin (II) Chloride (SnCl 2 ) in ethanol. The mixture was then extracted twice with 2 ml ethyl acetate. The pooled organic phases were evaporated under nitrogen gas and the residue was derivatized with 200 µl of BSA (Aldrich)/pyridine (1:1) at 80 °C for 30 min. The trimethylsilyl esters (TMSs) were analyzed by GC and/or GC-MS.

Tissue-specific expression of AsPXG1 and AsLOX2
Total RNAs were prepared from roots, leaves, developing seeds, glumes and germinating seeds at 5 days using TRIzol reagent (Invitrogen). One microgram of total RNA was treated with DNase I (Invitrogen) and used for cDNA synthesis using SuperScript III RT-PCR system in 20 µl reaction with Oligo(dT) 20 primer according to the manufacturer's instructions. One microliter of the first-strand reaction was then used as a template for 25-µl multiplex PCR reaction using Taq Figure 1. A diagram illustrating the primary structure of the putative peroxygenase AsPXG1 from oat. The black box indicates 19 amino acids between residue 10 and 29 identical to the sequenced fragment of the peroxygenase protein purified from oat lipid particles previously. Two conserved histidine residues are located at position 76 and 144 which are involved in hemebinding. The grey box indicates the EF-hand motif which is involved in calcium binding. The hatched box indicates a hydrophobic membrane-associated domain.

Figure 2.
Functional characterization of oat putative peroxygenase AsPXG1 in P. pastoris in presence of cumene hydroperoxide as oxygen donor and oleic acid as a substrate. A, Gas chromatography analysis of fatty acid methyl esters prepared from enzymatic reactions using extracts of Pichia AsPXG1/X33 and the control X33 cells (the empty vector). B, The mass spectrum of methyl 9,10-epoxy stearate produced by AsPXG1. The fatty acid derivative was prepared from enzymatic reactions using extracts of Pichia AsPXG1/X33.