AU9455398A - Plant fatty acid hydroxylase genes - Google Patents

Description

WO 99/18224 PCT/IB98/01716 1 PLANT FATTY ACID HYDROXYLASE GENES BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to the identification of plant 5 fatty acid hydroxylase genes; their use in genetic engineering and the modification of the fatty acid content of a cell, preferably a transfected cell; and products thereof, such as nucleic acids, recombinant vectors, polypeptides, host cells, transgenic plants, plant products with altered hydroxylated fatty acid content. 10 2. Description of the Related Art Two laurate hydroxylases exist in plant microsomes that catalyze either the terminal hydroxylation or the in-chain hydroxylation of the fatty acid substrate. These two hydroxylases, which are both cytochrome P-450 enzymes, were not found to coexist in the same plant 15 (Salatin et al., 1982). Lauric acid is hydroxylated in Jerusalem artichoke tubers, tulip bulbs, maize seedlings, and several other plants by an in-chain hydroxylase producing a mixture of o-2, o-3, and o-4 monohydroxylaurates. A laurate omega-hydroxylase is present in other 20 plants, mainly leguminosae, which hydroxylates the methyl terminus of the fatty acid substrate. These two activities are found in different P450 species and did not coexist in 12 plant species that were analyzed (Salatin et al., 1982). The level of cytochrome P450 in most plants is significantly 25 increased by exposure to various xenobiotics, endobiotic substrates, fungal infections, light irradiation, wounding of the tissues and subsequent aging. A synthetic plant hormone, 2,4-dichloro-phenoxyacetic acid (2,4-D), increased the spectrophotometrically detectable amount of cytochrome P450 WO 99/18224 PCT/IB98/01716 2 in Jerusalem artichoke tuber tissues (Adele et al., 1981). Similarly, cytochrome P450 content and, more specifically, the activities of the lauric acid in-chain hydroxylase and omega-hydroxylase were substantially induced by phenobarbitol in various plants (Salailn et al., 1981; 1982). 5 Clofibrate (ethyl 2-[4-chlorophenoxy]-2-methylpropanoate) is a hypolipidemic drug causing a proliferation of mitochondria, smooth endoplasmniic reticulum, and peroxisomes in mammalian liver. Induction of cytochrome P450 and, more specifically, lauric acid omega-hydroxylase activity was observed in liver tissue from clofibrate treated test animals 10 (Gibson et al., 1982). Similar effects are elicited by di-(2-ethylhexyl) phthalate (DEHP), a widely used industrial plasticizer. The induction of cytochrome P450 has allowed the purification of plant fatty acid hydroxylases from enriched sources, determination of a protein sequence from a plant fatty acid hydroxylase, and 15 cloning of a family of hydroxylases as disclosed herein. The identity of the cloned genes as fatty acid hydroxylases is confirmed by functional assay. SUMMARY OF THE INVENTION An object of the invention is to provide cytochrome P450 genes encoding plant fatty acid hydroxylases. In particular, plant genes for 20 terminal (omega or o) hydroxylases having a peptide sequence which is a unique signature of plant fatty acid omega-hydroxylases (for example, cytochrome P450 subtype CYP94) and an in-chain (co-l, co-2, co-3, and co-4 ) hydroxylase (cytochrome P450 subtype CYP81) are provided. Yet another object of the invention is to provide products 25 derived from the plant fatty acid hydroxylase genes. Such products include, for example, nucleic acids, polypeptides, host cells, and transgenic plants.

WO 99/18224 PCT/IB98/01716 3 A further object of the invention is to provide processes of making and using the plant fatty acid hydroxylase genes. In particular, genetic engineering allows making structural and functional variants of the plant fatty acid hydroxylases using the disclosed nucleotide and amino acid 5 sequences. Moreover, plant products with altered hydroxylated fatty acid content are obtained by producing plants with a transgene that affects fatty acid metabolism. In one embodiment of the invention, nucleic acids (e.g., DNA, RNA, variants thereof), recombinant polynucleotides comprised of the 10 nucleic acids (e.g., recombinant and expression vectors), polypeptides encoded by the nucleic acids (e.g., enzymes with fatty acid hydroxylase activity), host cells (e.g., bacteria, yeast, plant) containing the aforementioned molecules, and whole plants containing wild-type/mutant genes (or wild-type/mutant gene products) which are generated by using the 15 disclosed fatty acid hydroxylase sequences are provided. A second embodiment of the invention is a process of making recombinant polypeptide by expressing a plant fatty acid hydroxylase gene sequence. The polypeptide may be isolated from a host cell expressing the gene sequence and used as an enzyme in an industrial process, or the 20 polypeptide may act within a host cell or plant to hydroxylate suitable fatty acid substrates. Variant plant fatty acid hydroxylases could be produced by genetic mutation. Random or site-directed mutation, domain shuffling, rational design based on structural contacts between enzyme and substrate, 25 and correlation between protein structure-enzyme activity are examples of methods to produce variant genes and their cognate proteins. Variant plant fatty acid hydroxylases could be selected for desirable properties such as, for example, modification of substrate specificity. Suitable examples of such WO 99/18224 PCT/IB98/01716 4 modifications include hydroxylation of shorter or longer fatty acid chains, or fatty acids with odd carbon numbers; hydroxylation of FA with in-chain hydroxy or epoxy groups; hydroxylation of thia-FA, which is a FA having a methylene group replaced by a sulfur atom; hydroxylation of an ether-FA, 5 which is a FA having a methylene group replaced by an oxygen atom; hydroxylation of modified fatty acids such as esters or amides, instead of the usual fatty acids with a free carboxylic group. Other desirable properties for selection are substrate affinity, modification of the rate of catalysis, enzyme lability or stability, and cofactor requirements. 10 In a third embodiment of the invention, metabolites of the fatty acid hydroxylases and plant fractions enriched for such metabolites are provided. Processes of making transfected host cells and transgenic plants are provided to increase or decrease specific fatty acid hydroxylases using the disclosed sequences and expression vectors. Such transfected host cells 15 and transgenic plants provide a useful starting source for obtaining the desired metabolites from enriched or depleted fractions. Hydroxylated fatty acids are preferably produced as storage lipids in transgenic seeds. For example, hydroxylated fatty acids are generally present in minor amounts in the phospholipid fractions (cellular membranes) of plants. 20 Overexpression of plant fatty acid hydroxylase genes in a transgenic plant can elevate the content of hydroxylated fatty acids in triglycerides of the transgenic plant. The use of seed-specific promoters may allow accumulation of high amounts of hydroxylated fatty acids as storage lipids. The accumulated fatty acids can be recovered by extracting oil from the 25 transgenic seed and isolating the fatty acids stored therein. A fourth embodiment of the invention are processes providing for identification of additional fatty acid hydroxylase genes by hybridization (e.g., low or high stringency), nucleic acid amplification (e.g., WO 99/18224 PCT/IB98/01716 5 LCR, PCR), or by screening databases using the omega-FA hydroxylase signature defined hereinbelow, and making fatty acid hydroxylase variants by chemical modification of the enzyme or genetic mutagenesis of the hydroxylase sequence (e.g., point mutation, deletion, insertion, domain 5 shuffling). SThe fatty acid substrate may be, for example, capric (C10:0), lauric (C12:0), myristic (C14:0), palmitic (C16:0), oleic (C18:1), linoleic (C18:2 and enantiomers (9E, 12Z) and (9Z, 12E)), and linolenic (C18:3) fatty acid. Capric, lauric, and myristic fatty acids are considered medium 10 chain fatty acids, while palmitic, oleic, linoleic, and linolenic fatty acids are considered long-chain fatty acids. Hydroxylated and epoxidated fatty acids produced by the invention will provide oils with novel properties that may be used for the manufacture of lubricants, anti-slip agents, plasticizers, coating agents, 15 detergents, and surfactants. Apart from industrial considerations such as mass production as storage lipids in seeds, there are other reasons to over or under produce omega-hydroxylated fatty acids in plants. The involvement of the omega hydroxylases in cuticle synthesis suggests that manipulation of the 20 expression of these genes will affect the resistance of plants towards drought, or attack by insects and other pathogens. Furthermore, hydroxylated fatty acids are per se activators (elicitors) which trigger the mechanisms of plant defense against pathogens. BRIEF DESCRIPTION OF THE DRAWINGS 25 Figure 1: Nucleotide sequence of CYP94A1 (Clone A) and deduced protein translation (SEQ ID NOS:3-4, respectively). Nucleotides of WO 99/18224 PCT/IB98/01716 6 the open reading frame are shown in capital letters. The typical heme binding domain, which constitutes the P450 signature, is underlined. Figure 2: Carbon monoxide difference spectrum of yeast microsomes expressing CYP94A1 (Clone A). Microsomes (10 mg 5 protein/ml) prepared as described by Pompon et al. (1996) were diluted 5 fold, reduced with a few grains of sodium dithionite, and divided into two cuvettes. A baseline was recorded between 400 and 500 nm using a Shimadzu MP2000 double-beam spectrophotometer. Carbon monoxide was bubbled into the sample cuvette and the P450-CO complex spectrum was 10 recorded. Based on a millimolar absorbance of 91/cm, the amount of CYP94A1 was 176 pmole/mg protein. Figure 3: Chemical structure of sulfur-containing lauric acid analogs and sulfoxide metabolites. Figure 4: Radiochromatogram of the reaction products formed 15 from capric (C10:0), lauric (C12:0), myristic (C14:0), and palmitic (C16:0) acids by CYP94A1 (Clone A). After incubation of microsomes from transformed yeast with fatty acids, the reaction mixtures were extracted as described and analyzed by TLC. A: in the presence of NADPH; B: in the absence of NADPH; and S: residual substrate. 20 Figure 5: Radiochromatogram of the reaction products formed from oleic (C18:1), linoleic (C18:2), and linolenic (C18:3) acids by CYP94A1 (Clone A). After incubation of microsomes from transformed yeast with unsaturated fatty acids, the reaction mixtures were extracted as described and analyzed by TLC. A: in the presence of NADPH; B: in the 25 absence of NADPH (control); and S: residual substrate. Figure 6: Radiochromatogram of the reaction products formed from 9Z, 12E-octadecadienoic (C18:2-9E,12Z); 9E, 12Z-octadecadienoic (C18:2-9Z,12E); 8-propylsulfinyloctanoic (8S-LAU); and 10- WO 99/18224 PCT/IB98/01716 7 methylsulfinyldecanoic (10S-LAU) acids by CYP94A1 (Clone A). After incubation of microsomes from transformed yeast with unsaturated fatty acids, the reaction mixtures were extracted as described and analyzed by TLC (C18:2-9E,12Z and C18:2-9Z,12E) or HPLC (8S-LAU and 10S-LAU). 5 A: in the presence of NADPH; B: in the absence of NADPH (control); and S: residual substrate. Figure 7: Nucleotide sequence of VAGH811 (incomplete at 5' end), also called Clone B (complete cDNA encoding CYP94A2), and deduced protein translation (c-MAH or CYP94A2) (SEQ ID NOS:5-6, 10 respectively). The consensus heme-binding domain which constitutes the P450 signature is underlined. Figure 8: Carbon monoxide difference spectrum of yeast microsomes expressing o-MAH (Clone B). Microsomes (10 mg protein/ml) prepared as described by Pompon et al. (1996) were diluted 5-fold, reduced 15 with a few grains of sodium dithionite, and divided into two cuvettes. A baseline was recorded between 400 and 500 nm using a Shimadzu MP2000 double-beam spectrophotometer. Carbon monoxide was bubbled into the sample cuvette and the P450-CO complex spectrum was recorded. Based on a millimolar absorbance of 91/cm, the amount of o-MAH was 80 20 pmole/ml microsomes. Figure 9: Radiochromatogram of the reaction product formed from capric acid by o-MAH. After incubation with 14C capric acid, the reaction mixture was extracted as described and analyzed by TLC. A: in presence of NADPH, B: in absence of NADPH. 25 Figure 10: Radiochromatogram of the reaction product formed from lauric acid by o-MAH. After incubation with 14 C lauric acid, the reaction mixture was extracted as described and analyzed by TLC. A: in presence of NADPH, B: in absence of NADPH.

WO 99/18224 PCT/IB98/01716 8 Figure 11: Radiochromatogram of the reaction product formed from myristic acid by o-MAH. After incubation with ' 4 C myristic acid, the reaction mixture was extracted as described and analyzed by TLC. A: in presence of NADPH, B: in absence of NADPH. 5 Figure 12: Radiochromatogram of the reaction product formed from palmitic acid by o-MAH. After incubation with 1 4 C palmitic acid, the reaction mixture was extracted as described and analyzed by TLC. A: in absence of NADPH, B: in presence of NADPH. Figure 13: Radiochromatogram of the reaction product 10 formed from stearic acid by co-MAH. After incubation with " 4 C stearic acid, the reaction mixture was extracted as described and analyzed by TLC. A: in presence of NADPH, B: in absence of NADPH. Figure 14: Radiochromatogram of the reaction product formed from oleic acid by co-MAH. After incubation with 14 C oleic acid, 15 the reaction mixture was extracted as described and analyzed by TLC. A: in presence of NADPH, B: in absence of NADPH. Figure 15: Nucleotide sequence of CYP94A3 (Clone C) and deduced protein translation (SEQ ID NOS:7-8, respectively); compared to Clone B, nine nucleotides were missing at the 5' end. 20 Figure 16: Carbon monoxide difference spectrum of yeast microsomes expressing CYP94A3 (Clone C). Microsomes (10 mg protein/ml) prepared as described by Pompon et al. (1996) were diluted 5 fold, reduced with a few grains of sodium dithionite, and divided into two cuvettes. A baseline was recorded between 400 and 500 nm using a 25 Shimadzu MP2000 double-beam spectrophotometer. Carbon monoxide was bubbled into the sample cuvette and the P450-CO complex spectrum was recorded. Based on a millimolar absorbance of 91/cm, the amount of CYP94A3 was 550 pmole/ml microsomes.

WO 99/18224 PCT/IB98/01716 9 Figure 17: Radiochromatogram of the reaction products formed from capric (C10:0) and lauric (C12:0) acids by CYP94A3 (Clone C). After incubation of microsomes from transformed yeast with fatty acids, the reaction mixtures were extracted as described and analyzed by 5 TLC. A: in the presence of NADPH; B: in the absence of NADPH (control); arid S: residual substrate. Figure 18: Radiochromatogram of the reaction products formed from myristic (C14:0) and lauric (C16:0) acids by CYP94A3 (Clone C). After incubation of microsomes from transformed yeast with fatty 10 acids, the reaction mixtures were extracted as described and analyzed by TLC. A: in the presence of NADPH; B: in the absence of NADPH (control); and S: residual substrate. Figure 19: Radiochromatogram of the reaction products formed from oleic (C18:1) and linoleic (C18:2) acids by CYP94A3 (Clone 15 C). After incubation of microsomes from transformed yeast with fatty acids, the reaction mixtures were extracted as described and analyzed by TLC. A: in the presence of NADPH; B: in the absence of NADPH (control); and S: residual substrate. Figure 20: Nucleotide sequence of CYP81B1 (Clone D) and 20 deduced protein translation (SEQ ID NOS:15-16, respectively). The consensus heme-binding domain which constitutes the P450 signature is underlined. Figure 21: Carbon monoxide difference spectrum of yeast microsomes expressing CYP81B1 (Clone D). Microsomes (10 mg 25 protein/ml) prepared as described by Pompon et al. (1996) were diluted 5 fold, reduced with a few grains of sodium dithionite, and divided into two cuvettes. A baseline was recorded between 400 and 500 nm using a Shimadzu MP2000 double-beam spectrophotometer. Carbon monoxide was WO 99/18224 PCT/IB98/01716 10 bubbled into the sample cuvette and the P450-CO complex spectrum was recorded. Based on a millimolar absorbance of 91/cm, the amount of CYP81B 1 was 202 pmoles/mg protein. Figure 22: TLC analysis of the metabolites obtained after 45 5 min of incubation at 27 0 C of 100 tM ( 14 C)-radiolabeled C10:0 (a), C12:0 (b), C14:0 (c) with microsomes from Helianthus tuberosus (H. tub., 1.2 mg protein) or from transgenic yeast (CYP81B1, 0.1 mg protein) and 600 pLM NADPH. After stopping the reaction with one volume of acetonitrile-acetic acid (99.8/0.2), the incubation medium was directly spotted on TLC silica 10 plates (60 F254, Merck) and developed with a mixture of ether-petroleum benzin-formic acid (70/30/0.2). Reaction products were localized using a radiometer thin-layer scanner (Berthold LB 2723). Figure 23: Radio-HPLC analysis of the metabolites obtained after 45 min of incubation at 27 0 C of 100 [tM C12:0(a), and C14:0(b) with 15 microsomes of transgenic yeast (0.1 mg protein) and 600mM NADPH. After stopping the reaction with one volume of acetonitrile-acetic acid (99.8/0.2, v/v), the incubation medium was extracted twice with ether. Ether was evaporated under argon and the metabolites were separated on a Beckman HPLC ODS 5pm 1.6 mm x 15 cm column using H 2 0-acetonitrile 20 acetic acid (75/25/0.2 by vol for C12:0, 68/32/0.2 by vol for C14:0) as eluent. The flow rate was 2 ml/min. Radioactivity of HPLC effluent was monitored with a computerized on-line solid scintillation counter (Ramona D, Isomess). Figure 24: Nucleotide sequence of CYP94A4 (Clone E) and 25 deduced protein translation (SEQ ID NOS: 9-10, respectively) prepared from tobacco mosaic virus-infected tobacco leaves. Figure 25: Nucleotide sequence of CYP94A5 (Clone F) and deduced protein translation (SEQ ID NOS: 11-12, respectively) prepared WO 99/18224 PCT/IB98/01716 11 from tobacco mosaic virus-infected tobacco leaves. Figure 26: Nucleotide sequence of CYP94A6 (Clone G) and deduced protein translation (SEQ ID NOS: 13-14, respectively) prepared from tobacco mosaic virus-infected tobacco leaves.

WO 99/18224 PCT/IB98/01716 12 DETAILED DESCRIPTION OF THE INVENTION Plants are characterized by the presence of distinct cytochrome P450 isoforms (Salatin and Helvig, 1995). Some of these isoforms appear to be tissue, organ and species specific. Others, such as 5 CYP73 (CA4H = cinnamic acid 4-hydroxylase), are widely distributed in the plant kingdom. As in mammals, the involvement of multiple forms of cytochrome P450s in medium- and long-chain fatty acid (FA) oxidation in plants is well established. Interestingly, there are several similarities between mammals and plants in the catalytic mechanisms and the induction 10 of enzyme activities by various xenobiotics. The plant P450s involved in FA oxidation not been isolated to date because these membrane-bound enzymes are generally present in tissues at very low concentrations. During the past decade, several cytochrome P450 products encoded by the CYP4 gene family (mainly fatty acid hydroxylases) have been purified and cDNAs 15 have been isolated and sequenced from mammalian and insect cDNA libraries. Even though more than fifty cDNAs encoding plant P450s have been sequenced to date, none significantly matches the genes of the CYP4 family from mammals and insects. The biological roles and the substrate specificity of 20 cytochromes P450 isoforms involved in fatty acid and eicosanoid oxidation are poorly understood. Oxygenated FA from plants are mainly found in polar lipids such as triglycerides and phospholipids, and as monomers in polymeric layers. Cutins and suberins are polymers mainly composed of 25 hydroxylated fatty acids, especially omega-hydroxylated fatty acids. They protect plants against water loss, chemical penetration, attack by pathogens (e.g., microbes, insects), and other environmental stresses. Some of these WO 99/18224 PCT/IB98/01716 13 are potent inducers of fungal cutinase and some show anti-fungal properties. These contrasting and apparently opposite effects may be due to the great diversity of defense mechanisms found in plants, and also to strategies developed by the fungi to infect the plant host. 5 On the other hand, and as reported in mammals, hydroxylated fatty acids may play a role in responding to various stresses by giving rise to reactions similar to inflammatory processes, as a defense mechanism. Moreover, the presence of large amounts of hydroxy derivatives of the C18 family in the plant stigma suggests that they may play a role in recognition 10 of the stigma by pollen. Long-chain fatty acid omega and in-chain hydroxylases may play an important role in the synthesis of plant cuticles by generating hydroxy functions which appear essential to polymerization of constitutive cutin monomers. Cuticle monomers are often present as complex mixtures 15 with species-specific profiles. In addition to hydroxylated FA, epoxidated derivatives are also found as monomers of cuticles in a few plant species. Moreover, vicinal diol derivatives resulting from chemical and enzymatic ring-opening of epoxides have not been detected in cuticles from the C 16 fatty acid family, suggesting that an epoxide function is not essential for 20 polymerization of the cuticle matrix. In this case, it was suggested that the introduction of an internal hydroxyl group involves a direct hydroxylation mechanism catalyzed by cytochrome P450-dependent fatty acid hydroxylases. Fatty acids and their derivatives are subjected to many types 25 of oxidation reactions including hydroxylation, epoxidation, dehydration and reduction. Several forms of cytochrome P450 are suspected of being involved in these reactions. For example, previous studies have demonstrated that at least three distinct P450 isoforms are present in WO 99/18224 PCT/IB98/01716 14 microsomes from various plant species when incubated with a model substrate such as lauric acid. An interesting feature is that these P450 systems catalyze alternatively hydroxylation and epoxidation of unsaturated laurate analogs with a regio-specificity strongly dependent on the position 5 and stereo-specificity dependent on the configuration of the double bond in the aliphatic chain. Cytochrome P450-dependent reactions are involved in oxidation of fatty acids and derivatives in plants. The reactions are grouped below according to type of reaction and the position of the carbon attacked. 10 Some examples of induction by chemicals and inactivation by suicide substrates of the cytochrome P450 activities under consideration are discussed. A lauric acid omega-hydroxylase (mo-LAH), producing exclusively 12-hydroxylauric acid, has been described in Pisum sativum, 15 Vicia sativa and other leguminosae. In addition to laurate hydroxylation, the microsomal fraction from clofibrate-treated V. sativa seedlings also catalyzed the omega-hydroxylation of capric (C10:0) and myristic (C14:0) acids. A free carboxyl group appears essential for the binding of substrates to the enzyme. Induction and inhibition studies suggest that a single 20 cytochrome P450 is capable of omega-hydroxylating these fatty acids. To explore the catalytic capabilities of the o-LAH, a series of (1-1 4 C) radiolabeled unsaturated lauric acid analogs (7-, 8-, 9- and 10 dodecenoic acids) was incubated with the microsomal fraction from clofibrate-treated V. sativa seedlings. This subcellular fraction was able to 25 catalyze the omega-oxidation of the analogs when 02 and NADPH were present. The cis and trans forms of the four in-chain unsaturated analogs were 12-hydroxylated with similar efficiency. It is also important to note that allylic oxidation (i.e., 12-hydroxylation of 10-dodecenoate) occurred WO 99/18224 PCT/IB98/01716 15 with complete retention of the stereo-chemistry of the double bond and that allylic transposition was never observed. In contrast, the terminal olefin (11-dodecenoic acid) was epoxidized by the enzyme preparation. The formation of each metabolite was inhibited to the same extent when 5 microsomes were incubated in the presence of CO, anti-cytochrome P450 reductase antibodies and suicide substrates, suggesting that a single P-450 isoenzyme is able to omega-hydroxylate lauric acid, unsaturated analogs with a double bond or 1,4-pentadiene motif and to epoxidize the terminal olefin, 11-dodecenoic acid. The fact that o-LAH activity was not inhibited 10 by oleic acid (C18:1) at a concentration 10 times higher than that of laurate suggests that it is more specific for short- and medium-chain FA. Early work by Soliday and Kolattukudy demonstrated the omega-hydroxylation of palmitic acid (C16:0) by a microsomal fraction from V faba. Inhibition of the reaction by CO suggested the involvement 15 of a cytochrome P-450 monooxygenase but no reversal of CO inhibition by light was obtained. More recently, microsomes from etiolated Vicia sativa seedlings incubated with (1- 14 C)oleic acid (Z9-octadecenoic acid), (1 1 4 C)9,10-epoxystearic acid or (1-_1 4 C)9,10-dihydroxystearic acid catalyzed the NADPH-dependent formation of hydroxylated metabolites. The chemical 20 structure of these compounds was established by GC-MS analysis to be 18 hydroxyoleic acid, 18-hydroxy-9,10-epoxystearic acid and 9,10,18 tribydroxystearic acid, respectively. The reactions were inhibited by CO. Inhibition could be partially reversed by light and all three reactions were inhibited by antibodies raised against NADPH-cytochrome P450 reductase 25 from Jerusalem artichoke. The possibility that a single P450 is involved in the omega-oxidation of both oleic and linoleic acids (C18:2) is suggested by the competitive inhibition of oleic acid hydroxylation by linoleic acid, and vice versa.

WO 99/18224 PCT/IB98/01716 16 In microsomes from Jerusalem artichoke tubers (Helianthus tuberosus), a lauric acid in-chain hydroxylase (IC-LAH) catalyzes hydroxylation of carbons 10, 9 and 8 in a 24:63:13 ratio, respectively. The activity undetectable in dormant tuber tissues, was induced by wounding and 5 exposure to chemicals. Several other plant species, such as maize and tulip, catalyze this type of reaction but, in wheat seedlings, lauric acid is mainly converted to the 11-hydroxy derivative. The lauric acid (O-l)-hydroxylase ((0-l)-LAH) from wheat generates a mixture of monohydroxylaurate in the proportion of 65%, 31% and 4% for 11-hydroxy, 10-hydroxy and 9 10 hydroxylaurate, respectively. Capric and myristic acids were also converted to (c)-1) and (o-2) hydroxylated products. Additional minor metabolites hydroxylated at (0o-3) and (o-4) were also detected when myristic acid was the substrate. Whatever the length of FA (C10 to C14) incubated, no omega-hydroxylated products were detected. In addition, results from our 15 laboratory suggest that the (mo-l)-LAH from wheat catalyzes the hydroxylation of the herbicide diclofop. Biosynthesis of plant cuticles involves distinct P450 systems. The in-chain hydroxylation of omega-hydroxypalmitic acid by V faba microsomes gives rise to 9 (or 10),16-dihydroxypalmitic acid. The reactions 20 have been attributed to a cytochrome P450 which differs from those involved in omega-hydroxylation of palmitic acid by effective reversal by light of CO inhibition. The most abundant constituents found in the cutin of wheat caryopses are omega-hydroxylated oleic and 9,10-epoxystearic acids. 25 Surprisingly, incubation of the microsomal fraction from etiolated wheat shoots (Triticum aestivum L.) with (1- 1 4 C)oleic acid led to the formation of 18-, 17- and 16-hydroxyoleic acids, identified by GC-MS analysis. They were generated in a molar ratio of 1.4:4.6:4, respectively. The involvement WO 99/18224 PCT/IB98/01716 17 of cytochrome P450 was demonstrated by the dependence of these hydroxylations upon 02 and NADPH, and by their light-reversible inhibition by CO. This reaction was selectively inhibited by a suitably designed mechanism-based inhibitor (see below), while lauric acid and cinnamic acid 5 hydroxylations were not affected. The capability of V sativa microsomes to catalyze the oxidation of two sulphur-containing lauric acid analogs has been examined. Two sulphides synthesized in radiolabeled form, (1- 14 C) 10 methylsulphinyldecanoic acid (10S-LAU) and (1- 14 C)8 10 propylsuphinyloctanoic acid (8S-LAU), were incubated with V. sativa microsomes under conditions promoting either P450 or peroxidase reactions. In addition to the expected peroxidative oxidation, both 8- and 10-thio fatty acids were actively converted to the sulphoxide by at least two distinct membrane bound enzymes. Based on the NADPH requirement, reversal of 15 CO inhibition and inactivation of the NADPH-dependent reactions by the mechanism-based inhibitor 11-dodecynoic acid (11-DDYA) targeted to inhibit the o-LAH (see below), it is suggested that the sulphoxidation of 10S-LAU and 8S-LAU were catalyzed by the same or similar P450 forms which hydroxylate lauric acid. The second membrane-bound enzyme which 20 appears to be NADPH-independent was not fully characterized. However, the presence of beta-mercaptoethanol in the incubation medium had no effect on the sulphoxidation of either 8S-LAU or 10S-LAU, suggesting that the peroxidase present in these membranes was not involved. A remarkable property of living organisms is their ability to 25 induce the activity of P450 monooxygenases in response to chemical or physical stresses. Cytochrome P450 activities from plants are induced by light, UV-irradiation, wounding, ripening, fungal infection, elicitors, endogenous compounds and numerous chemicals, including safeners, WO 99/18224 PCT/IB98/01716 18 herbicides, drugs and pollutants. Plant P450 systems involved in omega-hydroxylation of lauric acid (o)-LAH) and oleic acid (o-OAH) are induced by clofibrate in a dose dependent manner. Clofibrate is a well known hypolipidemic drug 5 which induces peroxisome proliferation in both mammals and plants. Clofibrate and related arylphenoxy compounds, such as 2,4 dichlorophenoxy-acetic acid (2,4-D), which selectively induce fatty acid omega-hydroxylase activity, have little or no effect on the activity of IC LAH from H. tuberosus tubers and (o-1)-LAH from wheat seedlings. V. 10 sativa microsomes contain exclusively fatty acid omega-hydroxylases. In chain hydroxylated fatty acids have never been detected in microsomes from either untreated or clofibrate- or phenobarbital-treated Vicia seedlings, although exposure to these xenobiotics produces a dramatic increase of omega-hydroxylase activity: circa 20 times with phenobarbital and over 30 15 50 times with clofibrate. It is noteworthy that in mammalian systems, clofibrate induces the omega-hydroxylase selectively, while phenobarbital enhances (o-1)-hydroxylation of lauric acid. The microsomal (o-1)-LAH activity of etiolated wheat shoot was stimulated by treatment with naphthalic anhydride (NA) or phenobarbital (PB). Coating the seeds with the safener 20 NA resulted in a 4.5-fold increase of (o -)-LAH activity and a 1.5-fold increase in P-450 content, while the activity of cinnamate hydroxylase (CA4H), a P450 involved in lignin synthesis, was reduced. The herbicide metabolizing activity of diclofop arylhydroxylase (DIAH) was stimulated 4 fold. A much higher stimulation of the (mo-I)-LAH and DIAH was 25 observed when the seedlings were aged on a 5 mM PB solution. Coating the seeds with NA and subsequently aging on PB resulted in a synergistic stimulation of (m-1)-LAH and DIAH (20 times) while CA4H activity was strongly depressed. Cytochrome P450 content was increased to about 0.5 WO 99/18224 PCT/IB98/01716 19 nmole/mg, one of the highest levels so far recorded in plants. The relative amounts of 11-, 10- and 9-hydroxylaurates formed remained unchanged under all conditions. Similarly, the (o-1)-oleic acid hydroxylase activity was induced in treated seedlings to the same extent as (Co-1)-LAH, although 5 these P450-dependent reactions were supported by distinct isoforms. A wide range of chemicals has been found to induce the IC LAH activity of tubers and bulbs from Jerusalem artichoke, tulip, and maize seedlings. Activity was induced above the untreated level by wounding slices of Jerusalem artichoke tubers in the presence of 25 mM MnCl 2 and 20 10 mM aminopyrine, but was even more enhanced when tissues were exposed to 8 mM phenobarbital. The mechanism of cytochrome P450 induction in plant systems remains unknown but most of the P450 inducers active in mammals are also effective in induction of plant P450. Recently, experiments 15 demonstrated that induction of cytochrome P450-dependent fatty acid hydroxylases from rodent liver by hypolipidemic drugs, such as clofibrate, and certain physiological conditions involves transcriptional activation of the genes which was mediated by receptors (peroxisome proliferator-activated receptors). The evidence suggested that perturbation of lipid metabolism is 20 the common factor for fatty acid hydroxylase induction by peroxisomal proliferators. Mechanism-based inhibitors (suicide substrates) containing a terminal acetylene are potent irreversible inhibitors of both plant and mammalian fatty acid omega-hydroxylases. Pre-incubation of microsomes 25 from clofibrate-treated V. sativa seedlings with 11-dodecynoic acid (11 DDYA) and NADPH resulted in a pseudo-first-order loss of lauric acid omega-hydroxylation with Ki = 150 yM and a half-life of 2.4 min. The apparent rate constant for inactivation by 11-DDYA was 4.3-4.8 x 10- 3 /sec.

WO 99/18224 PCT/IB98/01716 20 Incubation of microsomes from V sativa with ( 1-1 4 C) 11-DDYA produces a major metabolite, 1,12-decanedioic acid, probably generated by addition of water to a ketene intermediate. This ketene may also interact with nucleophilic residues in the active site leading to a selective chemical 5 labeling of two proteins bands (about 50 kDa). The labeling of microsomal proteins, which correlated well with diacid formation and inactivation of 0) LAHs, increased as a function of incubation time and concentration of (1 1 4 C)11-dodecynoic acid. Based on these results, two potential inhibitors targeted to inactivate the omega-hydroxylation of oleic acid were 10 synthesized. Incubation of microsomes from V sativa with terminal acetylenes, (Z)9-octadecen-17-ynoic acid (17-ODNYA) and the corresponding (Z)9,10-epoxyoctadecan-17-ynoic acid (17-EODNYA), resulted in a pseudo-first-order loss of oleic acid omega-hydroxylation with apparent Ki of 60 yM and 50 pM, respectively. The calculated half-lives of 15 enzyme activity were 6 min and 8 min for saturating concentrations of 17 ODNYA and 17-EODNYA, respectively. Interestingly, these suicide substrates inhibit the omega hydroxylation of oleic acid, epoxide and diol derivatives, and also linoleic acid to a similar extent. To purify and sequence plant cytochrome P450 proteins, a 20 selective covalent binding of P450 apoproteins with labeled mechanism based inhibitors would provide a useful means of following the labeled protein during purification steps. The terminal olefin 11-dodecenoic acid inactivates a P450 from wheat which catalyzes mainly oxidation of the internal carbon (0-l) of 25 laurate. As proposed by Ortiz de Montellano and coworkers, P450 inactivation by a terminal olefin proceeds via an oxidative attack on the internal carbon (o-1) of the double bond leaving a terminal methylene radical free to alkylate the heme unit. In contrast, the plant o)-LAH which WO 99/18224 PCT/IB98/01716 21 exclusively attacks the external position, catalyzed the formation of the 11 12 epoxide without any measurable loss of activity. Acetylenic derivatives of lauric acid are also potent inactivators of (o-1)-LAH from wheat. Incubation of microsomes from etiolated wheat seedlings with 10 5 dodecynoic acid (10-DDYA) produced a dramatic inhibition of lauric acid hydroxylation. The inhibition was dependent upon time and concentration of inhibitor in a process characteristic of mechanism based inhibitors. A half-life of 3 min and an apparent inhibition constant K i of 14 pM were determined from pseudo-first-order kinetic studies of (o-1)-LAH inhibition. 10 Similar results were obtained by incubating microsomes with a terminal acetylene, 11-dodecynoic acid (11-DDYA). In addition, the oleic acid hydroxylase (o-1)-OAH from wheat, oxidizing mainly the (o-1) position, was irreversibly inhibited by a substrate analog displaying an acetylenic function at the (O-l) position. The 15 hydroxylation of oleic acid, but not of lauric acid, was inhibited when microsomes were incubated with cis-9-octadecen-16-ynoic acid (16 ODNYA). These results strongly suggest that at least two different P450 enzymes are involved in the oxidation of oleic and lauric acids. Thus, an internal acetylene exerts a highly destructive effect 20 on P450s catalyzing in-chain oxidation. The mechanism of inactivation remains unknown, but the chemical rearrangement of a putative unstable acetylene epoxide, already suspected in the formation of ketene from terminal acetylene, cannot be excluded. Compared to studies of fatty acid hydroxylases in mammals, 25 understanding of the catalytic mechanism and substrate specificity of plant fatty acid hydroxylases (Table 1) is limited, only two P450 forms catalyzing the dehydration of a fatty acid hydroperoxide have been isolated and cloned to date. Allene oxide synthase, by generating a precursor of jasmonic acid, WO 99/18224 PCT/IB98/01716 22 may be a key enzyme controlling various physiological steps in plant development. In this regard, it will also be of interest to understand the physiological role of the RPP from guayule rubber particles which apparently catalyzes a similar reaction. On the other hand, evidence 5 suggests that long-chain fatty acid hydroxylases (omega and in-chain) play an important role in the biosynthesis of plant cuticles by generating terminal and internal hydroxy functions which appear essential to polymerization of cutin monomers. At least three distinct roles for the plant fatty acid 10 hydroxylases of the present invention are foreseeable: cutin and suberin synthesis, rapid catabolism of free fatty acids (i.e., detoxification), and synthesis of signaling molecules. As discussed above, altering synthesis of cutin and suberin by controlling the activity of plant fatty acid hydroxylases is expected. 15 Therefore, a plant with desirable characteristics (e.g., resistance to drought or chemical penetration) may result from the modification of cutin and suberin production. For example, we envision a null mutant or a hypomorph would be a slow growing plant relative to the wildtype plant. Wounding or other types of stress lead to the activation of 20 phospholipases, drastic liberation of fatty acids, and an oxidative burst (Low et al., 1996). There are several papers showing that elicitors activate phospholipases in plants (see Chandra et al., 1996) whose activation will lead to liberation of free fatty acids. By analogy with the role of P450 fatty acid hydroxylases in animals, one role could be the rapid catabolism of 25 these free fatty acids. Recent data (Tijet et al., 1998) show that CYP94A1 is strongly induced after a few minutes, and up to 400 times after a few hours, in plant tissues exposed to clofibrate, a drug which provokes peroxisomal proliferation in animals and in plants (Palma et al., 1991).

WO 99/18224 PCT/IB98/01716 23 Peroxisomal proliferation is strongly linked to oxidative burst. The work of Schweizer et al. (1996ab) shows that omega hydroxy C16:0 and C18:1 fatty acids are resistance elicitors. By analogy with the arachidonate cascade, other omega-hydroxylated fatty acids might 5 be involved in stress signaling. In addition, increased omega and omega-1 hydroxylation may increase the production of pheromone-like molecules that could enhance insect attraction for pollination. This is based on the fact that many insect pheromones are (or are derived from) omega and (omega 1)-hydroxy fatty acids or alkanes (see Engels et al., 1997). 10 Furthermore, P450s involved in epoxidation of unsaturated fatty acids may also be involved in resistance to disease via the production of hydroxylated and epoxidated fatty acids which have been shown to inhibit the growth of pathogens (i.e., synthesis of signaling molecules). If the role of oxygenated fatty acids in fungal infections is considered, 15 contradictory effects seem apparent, because reports indicate that certain monomers from cutin (i.e., dihydroxy fatty acids and 9,10,18 trihydoxystearic acid) are potent inducers of the cutinase of several pathogenic fungi. Schweizer et al. (1996ab) have shown that cutin monomers, and specifically the omega-hydroxy forms, induce resistance in 20 barley against Erysiphe graminis. The highest effect is found for 9,10,18 trihydroxystearic acid. Pinot et al. (1993) showed that this compound is formed by the action of a 9,10 epoxygenase (not a P450), followed by hydroxylation at position 18 by P450 (CYP94A1, CYP94A4, CYP94A5 and CYP86Al catalyze this reaction), followed by opening of the epoxide by an 25 epoxide hydrolase. Ble et al. (1993) have shown that epoxygenase and epoxidase are extremely active, so that the limiting factor is the P450 catalyzed omega-hydroxylation. Overexpression of this gene, possibly under control of a pathogen-reactive promoter, may result in enhanced resistance.

WO 99/18224 PCT/IB98/01716 24 The oil produced by altering the amount of plant fatty acid hydroxylase activity may exhibit different characteristics from oil produced by the wildtype plant that would be useful for the manufacture of lubricants, anti-slip agents, plasticizers, coating agents, detergents, and surfactants. For 5 example, hydroxylated fatty acids of 10 to 14 carbon length (derived from capric, lauric, or myristic acid) may provide the basis for new detergents and plasticizers. Plastics that can be produced from hydroxylated fatty acids are polyurethanes and polyesters (Weber at al., 1994). It should be noted that cutin itself is a bioplastic constituted almost entirely of oxyfatty acids. 10 Omega hydroxylation is required for the chain-elongation reaction, and in-chain hydroxylation and/or epoxidation is required for reticulation. Plants can be engineered to produce C12 fatty acids by transforming them with the acyl-ACP thioesterase from Umbellularia californica which is specific for lauroyl-ACP. Arabidopsis thaliana 15 transformed with this gene produce up to 25% laurate. If these plants would be further transformed with CYP94A1, they would produce high amounts of omega-hydroxylauric acid. Similarly, plants engineered to produce C14 FA and subsequently transformed with CYP94A2 would produce high amounts of omega-hydroxymyristic acid. To date, the 20 properties and industrial uses of hydroxylated fatty acids have not been elucidated. Lack of commercial use of such fatty acids is due to the fact that these hydroxylated fatty acids do not accumulate in plants under normal conditions. The use of the plant fatty acid hydroxylases of the present invention will allow the mass production of such compounds. Chemical 25 synthesis of omega-hydroxylated fatty acids is difficult and expensive, meaning the production of omega-hydroxylated fatty acids in plants would be of great economic significance. Several o- and in-chain fatty acid hydroxylases have been WO 99/18224 PCT/IB98/01716 25 characterized in higher plants. In microsomes from Helianthus tuberosus tuber the o-2, o-3 and o-4 hydroxylation of lauric acid is catalyzed by one or a few closely related aminopyrine- and MnCl 2 -inducible cytochrome P450(s). To isolate the cDNA and determine the sequences of the(se) 5 enzyme(s), antibodies directed against a P450-enriched fraction purified ++ from Mn -induced tissues were used. Screening of a cDNA expression library from aminopyrine-treated tubers led to the identification of a cDNA (CYP81Bl) corresponding to a transcript induced by aminopyrine. CYP81B1 was expressed in yeast, and a systematic exploration of its 10 function revealed that this enzyme specifically catalyzes the hydroxylation of medium chain saturated fatty acids, namely capric (C 10:0), lauric (C12:0) and myristic (C14:0) acids. The same metabolites were obtained with transgenic yeast and plant microsomes: a mixture of o-1 to o-5 monohydroxylated products was observed. The three fatty acids were 15 metabolized with high and similar efficiencies, the major position of attack depending on chain length. When lauric acid was the substrate, turnover was -1 30.7 ± 1.4 min and Kmap p 788 ± 400 nM. No metabolism of long chain m,app fatty acids, aromatic molecules or herbicides was detected. This new fatty acid hydroxylase is typical from higher plants and differs from those already 20 isolated from other living organisms.

WO 99/18224 PCT/IB98/01716 26 Table 1: Summary of reactions with fatty acids catalyzed by a plant cytochrome P450. Plant species Substrate Product(s) generated or carbon position oxidized Viciafaba palmitic (C16:0) 0o-OH 16-hydroxy C16:0 8, 9 or 10-OH 5 Phaseolus aureus lauric (C 12:0) co-OH Phaseolus vulgaris lauric (C12:0) co-OH Vicia sativa C10:0-C14:0 co-OH Cl2:1 A7-10 co-OH C12:1 Al l 11,12-epoxy C12:1 triple bonds 8-10 co-OH C12:1 triple bond 11 1,12-dicarboxylic + inactivation C18:1 A9, C18:2 A9,12 co-OH 9,10-epoxy C 18:0 co-OH 9,10-diOH C18:0 co-OH Pisum sativum C10:0-C14:0 co-OH 12-oxo-C12:1 A9 12-OH-CI2:1 A9 C18:2 A9,12 9,10-epoxy-C18:1 A12 Glycine max C10:0-C14:0 co-OH 10 Triticum aestivum C10:0-C14:0 (co-3), (co-2), (co-1)-OH (mainly) C12:1 A9 or 10 9,10- or 10,11-epoxy C12:1 Al1 11,12-epoxy + inactivation C18:l A9 (co-2), (co-1) (mainly), co-OH Diclofop ring hydroxylation WO 99/18224 PCT/IB98/01716 27 TABLE 1 (continued) Helianthus tuberosus C10:0-C14:0 (o-2), (co-3) (mainly), (o-4)-OH C12:1 A8 or 9 8,9- or 9,10-epoxy-C12:0 C12:1 A7 or 10 Allylic hydroxy (9-OH) Helianthus annuus C 12:0 (w-2), (o-3) (mainly) or (o-4)-OH Zea mays it 5 Tulipa fosteriana Amaryllis belladonna Spinacia oleracea 18-OH-C 18:1 -CoA 9,10-epoxy-18-OH-C18:0-CoA Euphorbia lagascae linoleyl-PC 12,13-epoxy-C18:1 A9 Parthenium argentantum 13-OOH-C18:2 A9,11 a and y-ketol fatty acids 10 Linus usitanum 13-OOH-C18:2 A9, 11 a and y-ketol, cyclopentenyl product Since 1974, numerous reports have accumulated showing that plant P450 is induced by a great number of physical, physiological and chemical factors. However, in most cases, nothing is known about the identity of the induced isoforms and the actual mechanisms of induction. 15 The results presented hereinbelow show the regulation of three distinct cloned P450 species, CYP73A1, CYP76B1 and CYP94A1, which have been functionally expressed and characterized. CYP73A1 is the cinnamate 4-hydroxylase which catalyzes the first committed oxidation reaction in the general phenylpropanoid pathway, 20 leading to lignins, flavonoids, defense molecules, anti-UV protectants etc. This enzyme was purified to homogeneity using Triton X114 phase partitioning, and cloned using a specific antibody raised against the pure protein (Teutsch et al., 1993). Data suggest that the induction of cinnamate WO 99/18224 PCT/IB98/01716 28 4-hydroxylase activity primarily results from gene activation. Time-course experiments were performed after wounding and aminopyrine treatment. The timing of the induced changes in activity, protein and transcripts confirms that C4H induction results primarily from an increase in CYP73A] 5 mRNA both in wounded and aminopyrine treated tissues. However, post transcriptional mechanisms might also contribute to the regulation of C4H activity. CYP76B1 is an alkoxycoumarin O-dealkylase (Batard et al., 1995), whose true physiological function remains unknown. The protein was 10 purified by the same method employed for CYP73A1, and the gene was cloned using cytochrome P450 primers deduced from the microsequenced peptide (Batard et al., 1998). Determination of the steady-state level of CYP76B1 transcripts after slicing tuber tissues and aging them in water, alone or in the presence of various chemicals, showed that the expression of 15 this P450 was not responsive to mechanical stress, but was strongly induced by chemical treatments. Therefore, CYP76B1 appears to be a good potential marker of chemical stress and of environmental pollution. The data illustrate that the existence of a 'plant P450 induction mechanism' is highly unlikely. The more than 60 P450 20 physiological activities which have been identified, and the several hundreds more expected to be characterized in the near future, are dispersed in the numerous pathways of plant secondary metabolites. As such, induction of plant P450 enzymes will probably be coordinate with that of other enzymes in the pathway. This is particularly true for all the enzymes engaged in the 25 synthesis of defense compounds, which will be triggered by mechanical wounding, infection, and stress situations. The situation with the chemical inducers is more intriguing and very speculative. One can see the treatment of a plant by chemicals as 'painful', constituting a type of stress, that will WO 99/18224 PCT/IB98/01716 29 elicit a signal over one of the different stress signaling chains. On the other hand, some chemicals which induce P450 in animals also induce plant P450 with the same substrate and even regio-selectivity. In animals, plants and also B. megaterium, the fatty acid in-chain hydroxylases are selectively 5 induced by phenobarbital, while in animals and plants the co-hydroxylases are induced by clofibrate. The data suggest that in some instances, regulation mechanisms have been conserved along with the catalytic function during evolution. Finally, it should be stressed that some deviant (from a biochemical, phylogenetic, and probably structural point of view) 10 forms of P450 like the allene oxide synthase (Song et al., 1993) or the benzoate 2-hydroxylase (Leon et al., 1995), catalyze key steps in the synthesis of jasmonate and salicylate, respectively. Clofibrate stimulates efficiently (20-fold) the cytochrome P450-catalyzed activities of lauric and oleic acid omega-hydroxylation in 15 microsomes of Vicia sativa seedlings. DEHP and 2,4-D have a similar stimulating effect on the lauric acid omega-hydroxylase in the same material. Recently, on the basis of an internal peptide sequence, the inventors isolated cDNAs coding for plant fatty acid omega-hydroxylases. After expression in yeast, the omega hydroxylase substrate specificities were 20 characterized: CYP94A1 omega-hydroxylates fatty acids with different chain length (C10 to C18) and different degrees of unsaturation (C18:1, C18:2, C18:3). Northern blot analysis of RNAs from clofibrate-treated Vicia sativa seedlings revealed a very rapid (after 20 min) and large accumulation of the CYP94A1 transcripts, suggesting the involvement of a clofibrate receptor in 25 the signal transduction. To assess the mechanism of regulation of CYP94A1 by clofibrate, and the possible involvement of PPARs in this regulation, a promoter sequence of CYP94A1 was isolated. A search for key regulator elements is currently in progress. In addition, a study of peroxisome WO 99/18224 PCT/IB98/01716 30 proliferation in Vicia sativa in response to clofibrate, at the level of Acyl CoA oxidase transcripts, has been initiated. Furthermore, the inventors recently noted that treatment of etiolated Vicia sativa seedlings with the plant hormone methyl jasmonate 5 (MetJA) led to an increase in cytochrome P450 content, suggesting this plant defense molecule may act via a CYP94A1 pathway (Pinot et al., 1998). Treatment of the seedlings 48 hours in a ImM solution of MetJA stimulated co-hydroxylation of lauric acid 14 fold compared to control samples (153 pmol/min/mg protein versus 11 pmol/min/mg protein). 10 Induction was dose-dependent. The increased activity (2.7 fold) was already detectable after three hours of treatment. Activity increased as a function of time and reached a steady level after 24 hours. Northern blot analysis revealed that the transcripts coding for the a fatty acid o-hydroxylase CYP94A1 accumulated after one hour of exposure to MetJA, with maximal 15 levels accruing between three and six hours. Under the same conditions, a study of the enzymatic hydrolysis of 9,10-epoxystearic acid showed that both microsomal and soluble epoxide hydrolase activities were not affected by MetJA treatment. Thus, regulation of microsomal co-hydroxylation of fatty acids by methyl jasmonate may be a major event in the general 20 mechanism of plant defense. During the past years, the inventors have cloned and characterized at least three fatty acid o-hydroxylases in Vicia sativa. These P450 enzymes are able to introduce an alcohol function on the terminal carbon of fatty acids with different chain lengths and desaturations. These 25 novel enzymes are the first members of the CYP94 family. Mammalian o-hydroxylating enzymes have been extensively studied (Simpson 1997), and they have been classified in the CYP4A family. CYP4A are known to be involved in the metabolism of arachidonic acid WO 99/18224 PCT/IB98/01716 31 leading to the formation of physiologically important metabolites. They also participate in the catabolism of fatty acids (Gibson 1989). In plants, fatty acid o-hydroxylases are implicated in the biosynthesis of cuticle (Kolattukudy 1981). Moreover, w-hydroxy fatty acids have been 5 recently reported to play an important role in plant defense mechanisms (Schweizer 1996), and could also be involved in suberisation in elicitor treated french bean cells (Bolwell 1997). To characterize the physiological role of fatty acid o hydroxylases, the inventors developed a strategy based on the use of 10 transgenic tobacco lines. To date, three new clones belonging to the CYP94 family (CYP94A4, CYP94A5 and CYP94A6) from a tobacco cDNA library have been isolated. To modulate in vivo the activity of these P450s, the two coding sequences were incorporated in sense and antisense orientations in a T-DNA vector and transformed tobacco leaf disks. Expression of a 15 morphological phenotype in those transgenic lines is expected, but the quality and quantity of cuticle and the resistance to pathogens may be affected as well. Arylphenoxy compounds such as the hypolipidemic drug clofibrate and the herbicide 2,4-D are members of a class of chemicals 20 known to induce fatty acid hydroxylase activities and proliferation of peroxisomes. In plants, long-chain fatty acid omega-hydroxylases are believed to play a crucial role in synthesis of cuticles protecting plants from the outer environment. CYP94A1 which is the first P450-dependent fatty acid omega-hydroxylase cloned from a plant was isolated by tagging of the 25 P450 apoprotein with a radiolabeled mechanism-based inhibitor. The functional expression of this novel P450 in S. cerevisiae shows that the methyl end of saturated (from C10 to C16) and unsaturated (C18:1, C18:2 and C18:3) fatty acids was mainly oxidized by CYP94A1. Similar to animal WO 99/18224 PCT/IB98/01716 32 omega-hydroxylases, this plant enzyme was strongly induced by clofibrate treatment. Rapid accumulation of CYP94A1 transcripts was detected less than 20 min after exposure of Vicia sativa seedlings to clofibrate. The rapid induction of CYP94A1 ensures that fatty acids (FAs) are effectively 5 transformed into cutin monomers needed for repair and defense. The possible role of hydroxylated FAs as natural elicitors of plant defense mechanisms opens unexpected perspectives for investigating new regulatory routes of naturally occurring plant defense compounds under chemical or pathogenic stress. 10 All books, articles and patents cited in this specification are incorporated herein by reference in their entirety. The following examples are meant to be illustrative of the present invention; however, the practice of the invention is not limited or restricted in any way by them. 15 EXAMPLES Example I - Clone A (CYP94A1) Clone A encodes the saturated and unsaturated fatty acid (FA) omega-hydroxylase, a microsomal cytochrome P450-dependent hydroxylase which catalyzes the transformation of capric (C10:0), lauric 20 (C12:0), myristic (C14:0) palmitic (C16:0), oleic (C18:1), linoleic (C18:2 and enantiomers (9E,12Z); (9Z,12E)) and linolenic (C18:3) acids into their corresponding omega-hydroxy acids. In addition, two sulfur-containing lauric acid analogs, 10-methylsulfinyl-decanoic acid (10S-LAU) and 8 propylsulfinyl-octanoic acid (8S-LAU), are actively converted to their 25 corresponding 10- and 8-sulfoxides respectively (Figure 3). As previously demonstrated by incubating microsomes from V sativa, a series of unsaturated lauric acid analogs containing a double or WO 99/18224 PCT/IB98/01716 33 a triple bond at carbon positions 8, 9 and 10 should be omega-hydroxylated. In addition, the terminal ethylenic lauric acid analog, 11-dodecenoic acid, should be converted to 11-epoxylauric acid by CYP94A1 (Clone A) (Weissbart et al., 1992; Pinot et al., 1992, 1993; Helvig et al., 1997). 5 Isolation of CYP94A1 (Clone A) Specific peptide sequences of lauric acid omega-hydroxylase from clofibrate-treated V sativa microsomes (Salatin et al., 1986) were obtained by employing a newly developed method based on the alkylation of the P450 apoprotein by radiolabeled (1-1 4 C)11-dodecynoic acid (11 10 DDYA) (Salatin et al., 1988; Helvig et al., 1997). A chemically labeled protein (about 53kDa) was isolated by successive SDS-PAGE analysis and subjected to <<in-gel> V8 proteolysis. Resulting peptides were transferred to a nylon membrane (IMMOBILON T m ) and sequenced by the Edman degradation method. Four peptides were sequenced. Only two showed 15 homology to P450s and were subsequently found in the deduced amino acid sequence of clone A (SEQ ID NO:4). The first peptide contained the 18-20 amino acid hydrophobic domain which is typical of the membrane anchor found in all microsomal P450s. After isolation of the clone, it was confirmed that this peptide 20 corresponds to the N-terminal amino acid sequence of the enzyme: MFQFLLEVLLPYLLPLLLYILPF peptide microsequence MFQFHLEVLLPYLLPLLLLILPT peptide deduced from the sequence of clone A (Residues 1-23 of SEQ ID NO:4). The second peptide had the sequence 25 LMNLYPPVPMMNAKEVVVXVLLXQ. A computer search with this peptide against all known cytochrome P450 enzymes showed partial homology of the beginning with a domain which is found at about 130 residues from the C-terminus in several P450s of family CYP4, the family WO 99/18224 PCT/IB98/01716 34 containing the mammalian fatty acid omega-hydroxylases: peptide p3, rat CYP4A1, rat CYP4B1 and rat CYP4A3. After cloning and sequencing, only eight out of the ten first amino acids, shown here in bold type, are found in the corresponding domain of CYP94A1 (SEQ ID NO:4, residues 5 370-394, SMRLYPPVPMDSKEAVNDDVLPDGW), which implies that peptide was contaminated with another protein. Examination of sequences from 428 cytochrome P450s indicated that two consecutive methionines are never found. The following PCR primer (peptide and nucleic acid sequences contained in SEQ ID NOS: 4 and 3, respectively) was deduced: 10 Tyr Pro Pro Val Pro Met 5' - TAY CCI CCI GTI CCI ATG - 3' This primer was used in association with an oligo (dT) primer to produce a probe of 661 bp by RT-PCR on total RNA from clofibrate-treated V. sativa seedlings. Conditions were the following: 15 denaturation for 5 min at 93 0 C; followed by 30 cycles of denaturation for 1 min at 93oC, 2 min hybridization at 48 0 C, and 3 min elongation at 72oC; and terminated by 10 min elongation at 72oC. A XZAP cDNA library, prepared from poly(A) RNA from 48 hour clofibrate-treated V. sativa seedlings, following the manufacturer's 20 instructions (Stratagene), was screened at high stringency using the 650 bp probe. The probe was random labeled using (C- 32 dCTP) and hybridized for 24 hours at 65 0 C in 5 x SSC, 0.5% SDS, 5 x Denhardt's solution, 100 itg/ml salmon sperm DNA, 2 mM EDTA, and 50 mM sodium phosphate, pH 6.0. After hybridization, the blot was washed twice with 2 x SSC, 0.1% 25 SDS at room temperature for 15 min, and twice with 0.2 x SSC, 0.1% SDS at 55 0 C for 30 min. One clone (1862 bp) VAGH 11 was isolated, sequenced, and found to encode a new cytochrome P450, CYP94A1 (Figure 1).

WO 99/18224 PCT/IB98/01716 35 Heterologous Expression in Yeast Complete sequencing of the genome of Saccharomyces cerevisiae has shown that this yeast has only four P450s, none of which is known to catalyze fatty acid hydroxylation. Furthermore, this yeast can be 5 grown in conditions such that expression of endogenous P450s is minimal (i.e., P450 is spectrophotometrically undectable). Catalytic activity of CYP94A1 (Clone A) was assessed by functional expression in genetically engineered yeast. The system developed by Urban et al. (1990) was used for expression of heterologous P450 enzymes in Saccharomyces cerevisiae. 10 All methods (e.g., subcloning of P450 cDNA, the pYeDP60 shuttle vector, transformation of yeast, and growth conditions allowing the expression of the cloned P450 gene) are described in Pompon et al. (1996). The coding sequence of Clone A (SEQ ID NO:3) was PCR cloned into expression vector pYeDP60 using the BamHI and EcoRI restriction sites as follows. 15 Sense primer: Met Phe Gln Phe His Leu Glu 5' - CGGC GGATCC ATG TTT CAA TTT CAT CTT GGA G - 3' BamHI Antisense primer: 20 Ser Asp Arg Lys Gln Ile 5' - CGGC GAATTC TCA AGA ATC CCT CTT CTG AAT CG - 3' EcoRI stop Stratagene Pfu polymerase was used according to manufacturer's instructions and the amplified sequence was verified to avoid 25 polymerase-generated errors. Yeast strain WAT 11 (Urban et al., 1990) was transformed with the expression vector according to Schiestl and Gietz (1989).

WO 99/18224 PCT/IB98/01716 36 Characterization of CYP94Al (Clone A) Preparation of Microsomes: Yeast strain WAT11 transformed with pYeDP60 harboring CYP94A1 (Clone A) was grown and induced according to Pompon et al. (1996). A culture was started from one isolated 5 colony. After growth, cells were centrifuged 10 min at 7500g at 4°C. The pellet was washed with TEK (2 ml TEK/g cells), and centrifuged 10 min at 7500g at 4oC. The pellet was resuspended in lml TES and glass beads were added up to liquid surface level. Cells were broken by manual shaking for 5 min in the cold room using a 30 ml conical Falcon plastic 10 tube with 0.5 mm diameter glass beads. Greater than 90% of cells were lysed. The homogenate and two 5ml TES washes of the beads were centrifuged 10 min at 7500g at 4oC; the supernatant was centrifuged 45 min at 100,000g at 4 0 C. The resultant pellet was resuspended in 2ml TEG with a loose Potter homogenizer to obtain a fraction designed as microsomes 15 hereinafter. Microsomes can be stored at -20'C for several weeks without loss of activity. WAT11 cells transformed with pYeDP60 expression vector only were subjected to the same procedure for control experiments. TEK: Tris-HCI 50mM pH 7.5; EDTA 1mM; KCl 100mM TES: Tris-HCI 50mM pH 7.5; EDTA ImM; sorbitol 600mM 20 TEG: Tris-HCI 50mM pH 7.5; EDTA ImM; glycerol 20% Measurement of P450: Microsomes were diluted 5-fold with TEG and P450 was measured (Figure 2) with the method of Omura and Sato (1964) using a mmolar absorbance coefficient of 91 cm'mM -1. Measurement of Activity: Enzymatic activities were measured 25 as previously described (Weissbart et al., 1992; Pinot et al., 1992, 1993; Boucher et al., 1996) by following the rates of metabolite formation during incubation of transformed yeast microsomes with radiolabeled substrates. The standard assay contained in a final volume of 0.2 ml, 0.19-0.43 mg of WO 99/18224 PCT/IB98/01716 37 microsomal protein, 20 mM phosphate buffer (pH 7.4), and 100 pM radiolabeled substrate. Omega-hydroxylase activities were measured in the presence of 0.6 mM NADPH plus a regenerating system and 375 pM 3 mercaptoethanol. Ethanol solutions containing radiolabeled substrate were 5 evaporated under a stream of argon before addition of other fractions required for incubation. The reaction was initiated by adding NADPH at 27 0 C and stopped after 10 min incubation with 0.2 ml acetonitrile-acetic acid (99.8/0.2, v/v). After extraction with 2 x 600 pl diethyl ether, the organic phase was spotted on silica thin-layer plates and developed in a 10 mixture of diethyl ether-light petroleum (b.p. 40-60oC)-formic acid (70/30/1, v/v/v for CO10 to C16 substrates, and 50/50/1 for C18 substrates). Plates were scanned with a Berthold thin-layer scanner. For precise rate measurements, radioactive spots were scraped into counting vials and product formation was quantified by liquid scintillation. All the reactions 15 products identified in these experiments have been identified by GC/MS spectroscopy. The activity of CYP94A1 (Clone A) with different fatty acid substrates are shown in Table 2.

WO 99/18224 PCT/IB98/01716 38 Table 2: The activities were measured as described using purified radiolabeled substrates. Substrates (100pM) Vmax Km Product formed (mol/min/mol P450) (pM) (mol/min/mol P450) Capric acid (C10:0) 4.9 ± 0.1 101 ± 7.0 5 Lauric acid (C12:0) 19.95 ± 1.2 14.7 ± 3.0 Myristic acid (C14:0) 24.6 ± 0.13 45.0 ± 5.5 Palmitic acid (C16:0) 7.16 ± 0.3 7.2 ± 0.8 Stearic acid (C18:0) 0 0 Oleic acid (C18:1) 14.3 ± 0.5 38.7 ± 4.3 10 Linoleic acid (C18:2) 9.1 ± 0.9 47.0 ± 1.9 Linolenic acid (C18:3) 24.6 ± 2.3 70.0 ± 3.5 9,10-epoxystearic acid 18.5 1.3 (9R,10S) 9.10-dihydroxystearic acid 8.1 25 15 *C18:2 (9E,12Z) - - 52 *C18:2 (9Z,12E) - - 43 Linoleic (9E,12E) - - 25 *8-propylsulfinyloctanoic - - 40 acid 20 * 10-methylsulfinyldecanoic - - 133 acid *Values are means of triplicate measurements The TLC radiochromatograms are shown in Figure 4 (capric, lauric, myristic, and palmitic acids), Figure 5 (oleic, linoleic, and linolenic 25 acids), and Figure 6 (C18:2-9E,12Z; C18:2-9Z,12E; 8S-LAU and 10S-LAU WO 99/18224 PCT/IB98/01716 39 acids). For each substrate, two chromatograms are shown: with NADPH (A, hydroxylase is active) and without NADPH (B, hydroxylase is inactive). Example II - Clone B (CYP94A2) Clone B encodes co-MAH (omega-myristic acid hydroxylase), 5 a microsomal cytochrome P450-dependent hydroxylase which catalyzes the transformation of myristic (C14:0) acid into 14-hydroxytetradecanoic acid (i.e., the terminal methyl was hydroxylated). Low levels of transformation of lauric (C12:0) and palmitic (C16:0) acids into the corresponding omega hydroxy fatty acids were also observed. 10 Isolation of VAGH811 The clone was obtained by screening a %ZAP cDNA library prepared from clofibrate-treated Vicia sativa seedlings with Clone A, as follows. A XZAP cDNA library, prepared from poly(A) RNAs from 48 hour clofibrate-treated V. sativa seedlings following the manufacturer's 15 instructions (Stratagene), was screened at low stringency using a 661 bp DNA fragment (starting at position 1201 in Figure 1) produced by RT-PCR during the isolation of Clone A. The fragment was 32 P-radiolabeled by random priming. Hybridization was at 55oC overnight in 5 x SSC, 0.5% SDS, 5 x Denhardt's solution, 100 pg/ml salmon sperm DNA, 2 mM 20 EDTA, and 100 mM sodium phosphate, pH 6.0. After hybridization, blots were washed twice with 2 x SSC, 0.1% SDS at room temperature for 15 min, and twice with 0.2 x SSC, 0.1% SDS at 45oC for 30 min. One clone (1437 bp) that was isolated, VAGH811, was sequenced and found to encode a new cytochrome P450 (Figure 7). The clone was incomplete at the 5' 25 end, but the sequence of the incomplete clone was used to synthesize a primer for 5'-RACE with poly(A) RNA .from 96 hour clofibrate-treated WO 99/18224 PCT/IB98/01716 40 Vicia sativa seedlings to obtain the complete coding sequence of VAG811, named Clone B hereinafter (Figure 7). Heterologous Expression in Yeast Catalytic activity of CYP94A2 (Clone B) was assessed by 5 functional expression in genetically engineered yeast, as described hereinabove for clone A. The coding sequence of CYP94A2 (Clone B, SEQ ID NO:5) was PCR cloned into expression vector pYeDP60 using the SmaI and SacI restriction sites as follows (peptide sequences contained in SEQ ID NO:6). 10 Sense primer: Met Glu Leu Glu Thr Leu 5' - GGAT CCCGGG GA ATG GAA CTC GAA ACA TTG - 3' Sma I Antisense primer: 15 5' - AAG AGA AGC CCA CTT GTA TGA - 3' Lys Arg Ser Pro Leu Val stop 3' - CT TCG GGT GAA CAT ACT CTCGAG CTCGCCTA 5' SacI 20 Boehringer HiFi M polymerase was used according to manufacturer's instructions and the amplified sequence was verified to avoid polymerase-generated errors. Yeast strain WAT 11 (Urban et al., 1990) was transformed according to Schiestl and Gietz (1989). Characterization of CYP94A2 (Clone B) 25 Preparation of microsomes: Yeast strain WAT 11 transformed with the pYeDP60 expression vector harboring CYP94A2 (Clone B) was grown and induced according to the method described for Clone A. Measurement of P450: Microsomes were diluted 5-fold with WO 99/18224 PCT/IB98/01716 41 TEG and P450 was measured (Figure 8) with the method of Omura and Sato (1964) using an absorbance coefficient of 91 cm

-

' mM

-

'. Measurement of Activity: Enzymatic activities were measured as previously described for clone B. All the reaction products identified in 5 these experiments have been identified in experiments with plant microsomes, by rechromatography with authentic compounds and by GC/MS spectroscopy. The activities of o-MAH with different fatty acid substrates are shown in Table 3. Table 3: The activities were measured as described using purified 10 radiolabeled substrates. Substrates (100 pM) mole hydroxylated product/min/mole to e0 MAH Capric acid (C10:0) 0 Lauric acid (C12:0) 3.8 Myristic acid (C14:0) 30.4 15 Palmitic acid (C16:0) 4.0 Stearic acid (C18:0) 0 Oleic acid (C18:1) 0 The actual TLC radiochromatograms are shown in Figure 9 (capric acid), Figure 10 (lauric acid), Figure 11 (myristic acid), Figure 12 20 (palmitic acid), Figure 13 (stearic acid), and Figure 14 (oleic acid). For each substrate, two chromatograms are shown: with NADPH (A, hydroxylase is active) and without NADPH (B, hydroxylase is inactive).

WO 99/18224 PCT/IB98/01716 42 Kinetic Parameters and Specific Activity In separate experiments, the apparent Km and Vmax of CYP94A2 (Clone B) for myristic acid were determined to be 3.8 pM and 80 moles 14-hydroxymyristic acid/min/mole o-MAH. 5 Example III - Clone C (CYP94A3) CYP94A3 encodes a cytochrome P450-dependent hydroxylase which catalyzes the methyl terminal oxidation of capric (C10:0), lauric (C12:0), and myristic (C14:0) acids and to a lesser extent the omega hydroxylation of palmitic (C16:0), oleic (C18:1), and linoleic (C18:2) acids. 10 Isolation of CYP94A3 (Clone C) Clone C was obtained by screening a XZap cDNA library prepared from clofibrate-treated Vicia sativa seedlings with a 3' terminal cDNA fragment (300 bp) of CYP94A2 (Clone B) (from the sequence coding for the heme binding domain to the polyA tail) at high stringency. 15 Hybridization was for 24 hours at 65 0 C in 5 x SSC, 0.5% SDS, 5 x Denhardt's solution, 100 jig/ml salmon sperm DNA, 2 mM EDTA, and 50 mM sodium phosphate, pH 6.0. After hybridization, the blot was washed twice with 2 x SSC, 0.1% SDS at room temperature for 15 min, and twice with 0.2 x SSC, 0.1%SDS at 55 0 C for 30 min. The fragment was 32p 20 radiolabeled by random priming. A clone (1600 bp) was sequenced and found to encode a new cytochrome P450 (Figure 15). By sequence comparison with CYP94A2, CYP94A3 was missing nine nucleotides. For heterologous expression, the first nine nucleotides of CYP94A2 were added in front of the incomplete sequence of CYP94A3 (Clone C). The full-length 25 cDNA was isolated since and found to be identical to the one used in these activity experiments. Heterologous Expression in Yeast WO 99/18224 PCT/IB98/01716 43 Catalytic activity of CYP94A3 (Clone C) was assessed by functional expression in genetically engineered yeast as described hereinabove for the preceding clones. The coding sequence of CYP94A3 (Clone C, SEQ ID NO:7) was PCR cloned into expression vector pYeDP60 5 using the Sinal and SacI restriction sites as follows (peptide sequences contained in SEQ ID NO:8). Sense primer: 94A2 94A3 Met Glu Leu Glu Arg Leu Val Ala Trp 10 5' - TCC CCCGGG GT ATG GAA CTC GAA ACA TTG GTT GCA TGG - 3' Sma I Antisense primer: Glu Asp Thr His Ser 5' - ATC CGCTC GAGCTC TTA CTC ATC TGT GTG ACT - 3' 15 SacI stop Boehringer HiFimpolymerase was used according to manufacturer's instructions and the amplified sequence was verified for polymerase-generated errors. Yeast strain WAT 11 (Urban et al., 1990) was transformed according to Schiestl and Gietz (1989). 20 Characterization of CYP94A3 (Clone C) Preparation of Microsomes: Yeast strain WAT11 transformed with pYeDP60 harboring CYP94A3 (Clone C) was grown and induced as described for the preceding clones. Microsomes can be stored at -20 0 C for several weeks without loss of activity. WAT11 cells transformed with 25 pYeDP60 expression vector only were subjected to the same procedure for control experiments. Measurement of P450: Microsomes were diluted 5-fold with TEG and P450 was measured (Figure 2) with the method of Omura and Sato (1964) using a mmolar absorbance coefficient of 91 cm-'mM

-

'.

WO 99/18224 PCT/IB98/01716 44 Measurement of Activity: Enzymatic activities were measured as previously described as described hereinabove for the preceding clones. The activity of CYP94A3 (Clone C) with different fatty acid substrates are shown in Table 4. 5 Table 4: The activities were measured as described using purified radiolabeled substrates. Substrates (100 tM) Mole o-hydroxy-FA/min/mole P450 Capric acid (C10:0) 16.7 Lauric acid (C12:0) 10.8 10 Myristic acid (C14:0) 4.0 Palmitic acid (C16:0) 0.7 Oleic acid (C18:1) 0.7 Linoleic acid (C18:2) 1.6 Values are means of triplicate measurements 15 The TLC radiochromatograms are shown in Figures 17 (capric and lauric acids), 18 (myristic and palmitic acids), and 19 (oleic and linoleic acids). For each substrate, two chromatograms are shown: with NADPH (A, hydroxylase is active) and without NADPH (B, hydroxylase is inactive). 20 Example IV - Clone D (CYP81B1) In-chain Hydroxylase (IC-LAH) of capric, lauric, and myristic acids Clone D encodes a microsomal P450 from Helianthus tuberosus (Jerusalem artichoke), catalyzing the o-2, o-3 and o-4 hydroxylation of capric (C10:0), lauric (C12:0), and myristic (C14:0) acids.

WO 99/18224 PCT/IB98/01716 45 The major metabolite is the co-3-hydroxylated compound. Isolation of CYP81B1 (Clone D) Purification of xenobiotic-inducible 7-ethoxycoumarin O deethylase from H. tuberosus led to the isolation of a P450-enriched 5 fraction containing a mixture of several P450 proteins (Batard et al., 1995). Polyclonal antibodies raised against this P450-enriched fraction were used to screen a XZAPII cDNA library prepared from H. tuberosus tuber tissues sliced and aged 24 hour in presence of 20 mM of aminopyrine. Positive clones (56) were isolated and tested for the presence of a P450 consensus 10 sequence using the PCR technique previously described by Meijer et al. (1993). PCR fragments of expected size were obtained from 15 clones, labeled and hybridized with total RNA prepared from dormant, wounded or aminopyrine-treated tuber tissues. One of the 15 clones corresponded to a 2.2 kb transcript almost undetectable in dormant and wounded tuber, but 15 induced by aminopyrine. Sequencing of its insert showed that it coded for a P450 missing about 150 nucleotides at the N-terminus; rescreening of the library led to the isolation of a longer cDNA missing only five nucleotides. The missing coding sequence was then obtained by 5'-RACE, using poly(A) RNA from 24 hour aminopyrine-treated tuber tissues. The full-length 20 sequence was reconstituted and is named Clone D hereinafter (Figure 20).

WO 99/18224 PCT/IB98/01716 46 Heterologous Expression in Yeast Catalytic activity of CYP81B1 (Clone D) was assessed by functional expression in yeast. A genetically engineered yeast strain providing a suitable environment for plant P450 expression (membrane 5 structures and presence of a plant P450 reductase) was used for this purpose. This strain WAT11, the expression vector, subcloning of the coding sequence, yeast growth, transformation, and preparation of yeast microsomes are described in Pompon et al. (1996). The CYP81B1 (Clone D) cDNA, trimmed of its non-coding sequences, was expressed in the 10 pYeDP60 expression vector under the regulatory control of a galactose inducible promoter (GALO10-CYC1). Subcloning of CYP81B1 (Clone D) for insertion into this vector was performed using Pfu DNA polymerase (Stratagene), and the modified cDNA was checked for PCR-generated errors. 15 P450 content and catalytic activities were measured in microsomes prepared from transformed and control yeast (control = yeast transformed with an empty plasmid). No P450 or fatty acid metabolism was detected in control microsomes. In microsomal membranes from CYP81B1 (Clone D) transformed yeast grown 16 hour in the presence of 20 galactose, the P450 content measured by the method of Omura and Sato (1964) (Figure 21) was about 202 pmoles/mg protein (i.e., about 1% of the microsomal protein). Catalytic activity was tested with more than 20 potential radiolabeled substrates including aromatic compounds, sterols, herbicides, and fatty acids. Fatty acid metabolism was assayed as described 25 by Salatin et al. (1981). C10:0, C12:0 and C14:0 fatty acids were the only molecules metabolized by CYP81B1 (Clone D) (Table 5). Metabolism was dependent on the presence of NADPH.

WO 99/18224 PCT/IB98/01716 47 Table 5: Substrate specificity of CYP81B 1 Activities were measured with radiolabeled substrates. Metabolites (the sum of the three hydroxylated products) were quantified by radio-TLC. No activity was detected in control yeast (transformed with the 5 empty expression vector), or in the absence of NADPH. Substrate (100 p.M) Polar metabolites pmoles.min

-

'.mL

-

' C10:0 51 C12:0 34 C14:0 16 10 C16:0 not detected C18:0 not detected C18:1 not detected C18:2 not detected C18:3 not detected 15 Kinetic Parameters and Specific Activity Apparent Vmax and Km of the reaction were determined in the case of capric and lauric acids. In microsomes from yeast overexpressing Arabidopsis reductase (i.e., the WAT 11 strain), the reaction proceeds with an enzyme turnover of 41 ± 0.8 min

-

' and Km of 903 ± 168 20 nM in the case of capric acid, and with an enzyme turnover of 30.7 ± 1.4 min 1 and Km of 788 ± 400 nM in the case of lauric acid. Characterization of the Metabolites Previous work performed by the inventors (Salatin et al., 1981) has shown that in H. tuberosus tuber microsomes, lauric acid is WO 99/18224 PCT/IB98/01716 48 converted into 8-, 9- and 10-hydroxylated metabolites (25:60:15, respectively). Products obtained with plant microsomes have been characterized by GC-MS. Lauric acid in-chain hydroxylase activitiy was also detected in maize and tulip microsomes, and was induced by 5 aminopyrine, phenobarbital and other xenobiotics (Adel6 et al., 1981; Salatin et al., 1982; Salatin et al., 1986; Fonne-Pfister et al., 1988). TLC profiles in Figure 22 show that CYP81B1 (Clone D) codes for a P450 catalyzing formation of the same metabolites. The (0-2, co-3, o-4 hydroxylated metabolites are generated from the three fatty acid 10 substrates (capric, lauric, and myristic acids) in the same proportions as in plant microsomes. An additional minor product is detected after incubation of lauric acid with the yeast-expressed enzyme; the structure of this metabolite is currently being investigated. In the case of lauric and myristic acids, the presence and proportions of the three metabolites were confirmed 15 by HPLC (Figure 23). Previous work by the inventors also indicates that the same enzyme very likely catalyzes allylic hydroxylation or epoxidation of unsaturated lauric acids (Salatin et al., 1989, 1992, 1993) and sulfoxidation of 9- and 11-thiadodecanoic acids (Bosch, 1992), some unsaturated analogs 20 (in Z conformation) being metabolized with high stereoselectivity (Salatin et al., 1992). Example V Clone E (CYP94A4) CYP94A4 encodes a cytochrome P450-dependent hydroxylase which catalyzes the methyl terminal oxidation of capric (C10:0), lauric 25 (C12:0), myristic (C14:0), palmitic (C16:0), oleic (C18:1), linoleic (C18:2), and linolenic (C18:3) acids, with very contrasted efficiencies. Highest activity is with C14:0 and C12:0.

WO 99/18224 PCT/IB98/01716 49 Isolation of CYP94A4 (clone E) Clone E was obtained by screening a lambda-Zap cDNA library prepared from TMV-infected tobacco leaves (Dr M. Legrand; IBMP Strasbourg) with CYP94Al, A2 and A3 as follows: 5 A XZAP cDNA library, prepared from poly(A') RNAs from tobacco (Nicotiana tabacum var. Samsun NN) leaves infected for 48h with TMV, was screened at low stringency using a mixture of the coding sequences of CYP94Al, A2 and A3 as probe. The probe was 32p radiolabeled by random priming. Fifteen clones greater than 1500 pb were 10 isolated and sequenced. Ten of these clones were full-length and were found to encode for a new cytochrome P450 of the CYP94 family, which was named CYP94A4 (Figure 24). Heterologous expression in yeast. Catalytic activity of CYP94A4 (Clone E) was assessed by 15 functional expression in a ad hoc engineered yeast, as described hereinabove for the preceding clones. The coding sequence of Clone E (SEQ ID NO:9) was PCR cloned into expression vector pYeDP60 using the BamHI restriction site as follows: Sense primers (BamHI) 20 A4 M M I D L E L 5' CG GGA TCC ATG ATG ATA GAC TTG GAG CT 3' Antisense primers (KpnI) A4 GAA AGG AAC GGT ACG GAT ATT TGA A4 E R N G T D I stop 25 3' CC TTG CCA TGC CTA TAA ACT CCA TGG GG 5' WO 99/18224 PCT/IB98/01716 50 Boehringer HIFI' M polymerase was used according to manufacturer's instructions and the amplified sequence was verified for polymerase errors. Yeast strain WAT 1 (Urban et al. 1990) was transformed according to Schiestl and Gietz (1989). 5 Preparation of microsomes Yeast (strain WAT1 1) transformed with pYeDP60 harboring clone E was grown and induced as previously described hereinabove. Untransformed WAT11 cells were subjected to the same procedure for control experiments. 10 P450 measurements Microsomes were diluted 5-fold with TEG and P450 measured (Figure 2) with the method of Omura and Sato (1964) using a mmolar absorbance coefficient of 91.cm-l.mM 1 '. Activity measurements 15 Enzymatic activities were measured as previously described hereinabove. All the reactions products identified in these experiments had been identified before in experiments with plant microsomes, by rechromatography with authentic compounds and by GC/MS spectroscopy. Complete kinetic studies were conducted with each substrate.

WO 99/18224 PCT/IB98/01716 51 Table 6 : The activities for CYP94A4 were measured as described using radiolabeled substrates. Substrates Vmax Km Capric acid (C10:0) 3.5 2.6 1.3 5 Lauric acid (C12:0) 9.8 0.3 30.1 Myristic acid (C14:0) 14.6 2.2 6.6 Palmitic acid (C16:0) 3.1 46.5 0.067 10 Stearic acid (C18:0) 0 0 Oleic acid (C18:1) 1.8 24.8 0.072 Linoleic acid (C18:2) 4.5 47.9 0.094 Linolenic (C18:3) 4.0 95.5 0.042 Example VI Clone F (CYP94A5) 15 CYP94A5 encodes a cytochrome P450-dependent hydroxylase which catalyzes the methyl terminal oxidation of lauric (C12:0), myristic (C14:0), palmitic (C16:0), oleic (C18:1), linoleic (C18:2), and linolenic (C18:3) acids, with very contrasted efficiencies. Highest activity is with C14:0 and C18:2.

WO 99/18224 PCT/IB98/01716 52 Isolation of CYP94A5 (Clone F) Clone F was obtained by screening a lambda-Zap eDNA library prepared from TMV-infected tobacco leaves (Dr M. Legrand; IBMP Strasbourg) with CYP94Al, A2 and A3 as follows. 5 A XZAP cDNA library, prepared from poly(A

+

) RNAs from tobacco (Nicotiana tabacum var. Samsun NN) leaves infected for 48h with TMV, was screened at low stringency using a mixture of the coding sequences for CYP94A1, A2 and A3 as probe. The probe was up_ radiolabeled by random priming. Fifteen clones greater than 1500 pb were 10 isolated and sequenced. Two full-length clones were found to encode for a new cytochrome P450 of CYP94 family, which was termed CYP94A5 (Figure 25 ). Heterologous expression in yeast. Catalytic activity of CYP94A5 (Clone F) was assessed by 15 functional expression in a ad hoc engineered yeast as described for the preceding clones. Reformatting of Clone F The coding sequence of Clone F (SEQ ID NO: 1) was PCR cloned into vector pYeDP60 using the BamHI restriction site as follows 20 (peptide sequences contained in SEQ ID NO:12). Sense primers (BamHI) A4 M M I D L E L 5' CG GGA TCC ATG ATG ATA GAC TTG GAG CT 3' Antisense primers (KpnI) 25 A4 GAA AGG AAC GGT ACG GAT ATT TGA WO 99/18224 PCT/IB98/01716 53 A4 E R N G T D I stop 3' CC TTG CCA TGC CTA TAA ACT CCA TGG GG 5' Boehringer HIFI T M polymerase was used according to manufacturer's instructions and the amplified sequence was verified for polymerase errors. 5 Yeast strain WAT 11 (Urban et al. 1990) was transformed according to Schiestl and Gietz (1989). Preparation of microsomes Yeast (strain WAT1 1) transformed with pYeDP60 harboring clone F was grown and induced as described hereinabove for the preceding 10 clones. Untransformed WAT 11 cells were subjected to the same procedure for control experiments. P450 measurements Microsomes were diluted 5-fold with TEG and P450 measured (Figure 2) with the method of Omura and Sato (1964) using a 15 mmolar absorbance coefficient of 91.cm'.mM

-

. Activity measurements Enzymatic activities were measured as previously described hereinabove. All the reactions products identified in these experiments had been identified before in experiments with plant microsomes, by 20 rechromatography with authentic compounds and by GC/MS spectroscopy. Complete kinetic studies were conducted with each substrate.

WO 99/18224 PCT/IB98/01716 54 Table 7 : The activities for CYP94A5 were measured as described using radiolabeled substrates Substrates Vmax Km Capric acid (C10:0) 0 0 5 Lauric acid (C12:0) 1.5 89.3 0.02 Myristic acid (C14:0) 5.1 50.7 0.1 Palmitic acid (C16:0) 0.8 53.5 0.015 10 Stearic acid (C18:0) 0 0 Oleic acid (C18:1) 1.7 36.5 0.046 Linoleic acid (C18:2) 4.3 36.5 0.12 Linolenic (C18:3) 2.2 17.1 0.13 WO 99/18224 PCT/IB98/01716 55 Table 8 : The relative efficiency of 94A4 versus 94A5 was compared. CYP94A4 CYP94A5 Vm Km Vm/Km Vm Km Vm/Km Eff A4 / Eff A5 C10O 3.5 2.6 1.3 0 0 - C12 9.8 0.3 30.1 1.5 89.3 0.02 1505 C14 14.6 2.2 6.6 5.1 50.7 0.1 66 5 C16 3.1 46.5 0.067 0.8 53.5 0.015 4.5 C18:0 0 0 - 0 0 C18:1 1.8 24.8 0.072 1.7 36.5 0.046 1.5 C18:2 4.5 47.9 0.094 4.3 36.5 0.12 0.78 C18:3 4.0 95.5 0.042 2.2 17.1 0.13 0.32 WO 99/18224 PCT/IB98/01716 56 Example VII Clone G (CYP94A6) CYP94A6 encodes a cytochrome P450. The catalytic activity is presently being assessed. It is expected that it will show fatty acid hydroxylase activity since it displays the characteristic signature sequence 5 for this class of enzymes. Isolation of CYP94A6 (Clone G) Clone G was obtained by screening a lambda-Zap cDNA library prepared from TMV-infected tobacco leaves (Dr M. Legrand; IBMP Strasbourg) with CYP94A1, A2 and A3 as follows. 10 A XZAP cDNA library, prepared from poly(A') RNAs from tobacco (Nicotiana tabacum var. Samsun NN) leaves infected for 48h with TMV, was screened at low stringency using a mixture of the coding sequences for CYP94A1, A2 and A3 as probe. The probe was 32p radiolabeled by random priming. Fifteen clones greater than 1500 pb were 15 isolated and sequenced. One uncomplete clone was found to code for a new cytochrome P450 of the CYP94 family, which was named CYP94A6. The complete sequence for CYP94A6 was obtained by performing inverse-PCR on genomic tobacco (Nicotiana tabacum var. Samsun NN) DNA, using the NdeI restriction site ACATAT at position 594 (Figure 26 ) and sequence 20 specific primers. Heterologous expression in yeast. CYP94A6 has been expressed in yeast as described hereinabove, and the protein produced has been detected by Western blotting. Catalytic activity of Clone G is being assessed by functional 25 expression in a ad hoc engineered yeast.

WO 99/18224 PCT/IB98/01716 57 Reformatting of Clone G The coding sequence of Clone G (SEQ ID NO:13) was PCR cloned into vector pYeDP60 using the BamHI restriction site as follows (peptide sequences contained in SEQ ID NO:14). 5 Sense primer (BamHI) A5 M A L L D L Q 5' CG GGA TCC ATG GCA CTA TTA GAC TTA CAA 3' Anti sense primer (KpnI) A5 GTT ACT ATT GAA GAA AGG ATA TAG 10 V T I E E R I stop 3' A TGA TAA CTT CTT TCC TAT ACT CCA TGG GG Boehringer HIFfI polymerase was used according to manufacturer's instructions and the amplified sequence was verified for polymerase errors. Yeast strain WAT 11 (Urban et al. 1990) was transformed according to 15 Schiestl and Gietz (1989). Preparation of microsomes Yeast (strain WAT 11) transformed with pYeDP60 harboring clone G was grown and induced as described hereinabove. Untransformed WAT 11 cells were subjected to the same procedure for control experiments.

WO 99/18224 PCT/IB98/01716 58 P450 measurements Microsomes were diluted 5-fold with TEG and P450 measured (Figure 2) with the method of Omura and Sato (1964) using a mmolar absorbance coefficient of 91.cm-l.mM

-

'. 5 Obtention of transgenic tobacco plants expressing CYP94A4, CYPA5, CYPA6 Tobacco (Nicotiana tabacum L. var Xanthi) was transformed with the open REPETITION open reading frames of clones E (CYP94A4), F (CYPA5) and G (CYPA6), in sense and antisense orientation. The coding 10 sequences were cloned into pFB8, a custom built vector from our Institute (Atanassova et al. Plant J. 1995, 8, pp 465-477). The transformation was performed using tobacco leaf disks via Agrobacterium (strain LBA 4404) as described by Horsch (Science 1985, 227 pp 1227-1237). Reformatting of CYP94A4, CYPA5, CYPA6 coding sequences 15 The coding sequences were PCR cloned into vector pFB8 using the BamHI and KpnI restriction sites indicated in bold type as follows (peptide and nucleic acid sequences contained in SEQ ID NOS:4 and 3, respectively, for 94A4, and SEQ ID NOS:13 and 12, respectively, for 94A5/94A6 ) 20 a) Sense orientation in pFB8: Sense primers (KpnI) 94A4 M M I D L E L 5' CG CCA TGG ATG ATG ATA GAC TTG GAG CT 3' 94A5/94A6 M A L L D L Q 25 5' CG CCA TGG ATG GCA CTA TTA GAC TTA CAA 3' WO 99/18224 PCT/IB98/01716 59 Antisense primers (BamHI) 94A4 GAA AGG AAC GGT ACG GAT ATT TGA E R N G T D I stop 3' CC TTG CCA TGC CTA TAA ACT GGA TCC GG 5' 5 94A5/94A6 GTT ACT ATT GAA GAA AGG ATA TAG V T I E E R I stop 3' A TGA TAA CTT CTT TCC TAT ACT GGA TCC GG b) Antisense orientation in pFB8: Sense primers (BamHI) 10 94A4 M M I D L E L 5' CG GGA TCC ATG ATG ATA GAC TTG GAG CT 3' 94A5/94A6 M A L L D L Q 5' CG GGA TCC ATG GCA CTA TTA GAC TTA CAA 3' Antisense primers (KpnI) 15 94A4 GAA AGG AAC GGT ACG GAT ATT TGA 94A4 E R N G T D I stop 3' CC TTG CCA TGC CTA TAA ACT CCA TGG GG 5' 94A5/94A6 GTT ACT ATT GAA GAA AGG ATA TAG 94A5/94A6 V T I E E R I stop 20 3' A TGA TAA CTT CTT TCC TAT ACT CCA TGG GG Plants transformed in both sense and antisense directions are growing at present with seeds for TI expected shortly. Signature for fatty acid omega-hydroxylases The inventors have identified a peptide sequence (SEQ ID WO 99/18224 PCT/IB98/01716 60 NO:2), marked by a double underline in Figures 1, 7, 15, 24, 25 and 26, which is a unique signature found in all plant fatty acid omega hydroxylases characterized so far: S(AVS)AL(TVS)WFFWL(LIV) 5 Wherein (AVS) means one of A, V or S; (TVS) means on eof T, V, or S; and (LIV) means one of L, I or V. This signature sequence is present in CYP86A1 (SEQ ID NO:1), CYP86A5, CYP94A1, CYP94A2, CYP94A3, CYP94A4, CYP94A5, and CYP94A6. All but CYP94A6 (characterization under way) have omega-hydroxylase activity. This signature sequence is 10 not present in CYP81B1, the in-chain hydroxylase, as evidenced by sequence alignment. A scan of all plant genes in Genbank for this signature sequence retrieved 12 sequences, all of which are cytochromes P450 isofoms. Some of the sequences are redundant because they originate from 15 different laboratories recloning the same genes. A scan of all sequences contained in the databases (Non redundant GenBank+EMBL+DDBJ+PDB sequences = 364,804 sequences) confirms that this signature is not found in any other gene from any plant, animal or microbial origin. Therefore, any isolated gene presenting this 20 signature is linked to the genes covered by the invention and presents the same type of catalytic activity. REFERENCES Adele, P., Reichhart, D., Salatin, J.P., Benveniste I., and Durst F. (1981) Induction of cytochrome P450 and monooxygenase activity by 2,4 25 dichlorophenoxyacetic acid in higher plant tissue. Plant Sci. Lett., 22, 39-46.

WO 99/18224 PCT/IB98/01716 61 Batard, Y., Zimmerlin A., Le Ret, M., Durst, F., and Werck-Reichhart, D. (1995) Multiple xenobiotic-inducible P450s are involved in alkoxycoumarins and alkoxyresorufins metabolism in higher plants. Plant Cell Environ., 18, 523-533. 5 Batard Y. (1995) Recherche de nouveaux marqueurs biochimiques de pollution: l'induction des cytochromes P450s par les xenobiotiques chez les vegetaux superieurs. Ph.D. Thesis, Universite Louis Pasteur, Strasbourg. Batard Y, LeRet M., Schalk M., Robineau T., Durst F. and Werck Reichhart D. (1998) Molecular cloning and functional expression in yeast of 10 CYP76B1, a xenobiotic-inducible 7-ethoxycoumarin O-deethylase from Helianthus tuberosus. Plant J. 14: 111-120. Benveniste I, Tijet N, Adas F, Philipps G, Salaun JP, Durst F. (1998): CYP86A1 from Arabidopsis thaliana encodes a cytochrome P450-dependent fatty acid omega-hydroxylase. Biochem Biophys Res Commun 243: 688 15 693. Benveniste I., Pinot F., Le Bouquin R. Salatin JP,, Durst F. Clofibrate induction of a cytochrome P450 dependent-fatty acid omega-hydroxylase (CYP 94A1): are PPARs actually present in plants? 12th International Symposium : Microsomes and Drug Oxidations, Montpellier, France, July 20 20-24, 1998. Blde, E., Wilcox, A.L., Marnett, L.J., and Schuber, F. (1993) Mechanism of reaction of fatty acid hydroxyperoxides with soybean peroxygenase. J. Biol. Chem., 268, 1708-1715.

WO 99/18224 PCT/IB98/01716 62 Bosch, H. (1992) Contribution a Itetude de l'oxydation des acides gras par des monooxygnases vegetales a cytochrome P450. Ph.D. Thesis, Universite Louis Pasteur, Strasbourg. Boucher, J-L., Delaforge, Salatin, J-P., Pinot, F., Durst, F., Pflieger, P., and 5 Mioskowski, C. (1996) In vitro hydroxylation and epoxidation of some isomeric lauric acids analogs by rat liver microsomes: Identification of metabolites and effects of clofibrate or phenobarbital pretreatment. Drug Met. Disp., 24, 462-468. Budziszewski, G.J., Croft, K.P.C., and Hildebrand, D.F. (1996) Uses of 10 biotechnology in modifying plant lipids. Lipids, 31, 557-569. Cabello-Hurtado F, Batard Y, Salaun JP, Durst F, Pinot F, Werck-Reichhart D. (1998): Cloning, expression in yeast, and functional characterization of CYP81B1, a plant cytochrome P450 that catalyzes in-chain hydroxylation of fatty acids. J Biol Chem 273: 7260-7267. 15 Cabello-Hurtado F., Batard Y., Salain JP, Durst F., Pinot F., Werck Reichhart D. CYP81B1, a plant P450 which catalyzes in-chain hydroxylation of fatty acids. IVth International Symposium on Cytochrome P450 Biodiversity and Biotechnology. Strasbourg, France. July 12-16, 1998 Chandra, S., Heinstein, P.F., and Low, P.S. (1996) Activation of 20 phospholipase A by plant defense elicitors. Plant Physiol., 110, 979-986. Choi, D., Bostock, R.M., Avdiushko, S., and Hildebrand, D.F. (1994) Lipid derived signals that discriminate wound- and pathogen-responsive isoprenoid WO 99/18224 PCT/IB98/01716 63 pathways in plants - methyl jasmonate and the fungal elicitor arachidonic acid induce different 3-hydroxy-3-methylglutaryl-coenzyme a reductase genes and antimicrobial isoprenoids in Solanum tuberosum L. Proc. Natl. Acad. Sci. U.S.A., 91, 2329-2333. 5 Durst F. Regulation of cytochrome P450 monooxygenases in plants. BRAIN Seminar: Cytochrome P450 and plant genetic engineering. Tsukuba, Japon 30-31 Octobre 1997. Engels, W., Rosenkranz, P., Adler, A., Taghizadeh, T., Lubke, G., and Francke, W. (1997) Mandibular gland volatiles and their ontogenetic 10 patterns in queen honey bees, apis mellifera carnica. J. Insect Physiol., 43, 307-313. Farmer, E.E. (1997) New fatty acid-based signals: A lesson learned from the plant world. Science, 176, 912-913. Fonne-Pfister, R., Simon, A., Salatin, J.P., and Durst, F. (1988) Xenobiotic 15 metabolism in higher plants. Involvement of microsomal cytochrome P450 in aminopyrine N-demethylation. Plant Sci., 55, 9-20. Gibson, G.G., Orton, T.C., and Tamburini, P.P. (1982) Cytochrome P-450 induction by clofibrate. Biochem. J., 203, 161-168. Helvig, C., Alayrac, C, Mioskowski, C., Koop, D., Poullain, D., Durst, F., 20 and Salatin, J-P. (1997) Suicide inactivation of cytochrome P450 by mid chain and terminal acetylenes: A mechanistic study of inactivation of a plant lauric acid omega-hydroxylase. J. Biol. Chem., 272, 414-421.

WO 99/18224 PCTIIB98/01716 64 Le Bouquin R., Benveniste I., Durst F. A transgenic strategy to understand the physiological role of plant fatty acid o-hydroxylases. IVth International Symposium on Cytochrome P450 Biodiversity and Biotechnology. Strasbourg, France. July 12-16, 1998. 5 Low, P.S, and Merida, J.R. (1996) The oxidative burst in plant defense: Function and signal transduction. Physiol. Plant, 96, 533-542. Meijer, A.M., Souer, E., Verpoorte, R., and Hoge, J.H.C. (1993) Isolation of cytochrome P450 cDNA clones from the higher plant Catharanthus roseus by a PCR strategy. Plant Mol. Biol., 22, 379-383. 10 Ohlrogge, J.B. (1994) Design of new plant products: Engineering of fatty acid metabolism. Plant Physiol., 104, 821-826. Omura, T., and Sato, R. (1964) The carbon monoxyde-binding pigment of liver microsomes. J. Biol. Chem., 239, 2370-2378. Palma, J.M., Garrido, M., Rodriguezgarcia, M.I., and Delrio, L.A. (1991) 15 Peroxisome proliferation and oxidative stress mediated by activated oxygen species in plant peroxisomes. Arch. Biochem. Biophys., 287, 68-74. Pinot F., Benveniste I., Salatin JP., Durst F. (1998) Methyl jasmonate induces lauric acid o-hydroxylase activity and accumulation of CYP94A1 transcripts but does not affect epoxide hydrolase activities in Vicia sativa 20 seedlings. Plant Physiol. (in the press) Pinot, F., Bosch, H., Alayrac, C., Mioskowski, C., Vendais, A., Durst, F., WO 99/18224 PCT/IB98/01716 65 and Salaiin, J.-P. (1993) o-Hydroxylation of oleic acid in Vicia sativa microsomes: Inhibition by substrate analogues and inactivation by terminal acetylenes. Plant Physiol., 102, 1313-1318. Pinot, F., Salatin, J.-P., Bosch, H., Lesot, A., Mioskowski, C., and Durst, F. 5 (1992) Omega-hydroxylation of (Z) 9-octadecenoic, (Z) 9,10-epoxystearic and 9,10-dihydroxystearic acids by microsomal cytochrome P450 systems from higher plant. Biochem. Biophys. Res. Commun., 184, 183-193. Pompon, D., Louerat, B., Bronine, A., and Urban, P. (1996) Yeast expression of animal and plant P450s in optimized redox environments. 10 Meth. Enzymol., 272, 51-64. Salatin, J.P., Benveniste, I., Reichhart, D., and Durst, F. (1981) Induction and specificity of a (cytochrome P450)-dependent laurate in-chain hydroxylase from higher plant microsomes. Eur. J. Biochem., 119, 651-655. Salatin, J.P., Benveniste, I., Fonne, R., Gabriac, B. Reichhart, D., Simon, 15 A., and Durst, F. (1982) Hydroxylations microsomales de l'acide laurique catalysees par le cytochrome P450 chez les plantes superieures. Physiol. Veg., 20, 613-621. Salaiin, J.P., Simon, A., and Durst, F. (1986) Specific induction of lauric acid hydroxylase by clofibrate, diethylhexylphthalate and 2,4 20 dichlorophenoxyacetic acid in higher plants. Lipids, 21, 776-779. Salatin, J.-P., Simon, A., Durst, F., Reich, N.O., and Ortiz de Montellano, P.R. (1988) Differential inactivation of plant lauric acid o- and in-chain- WO 99/18224 PCT/IB98/01716 66 hydroxylases by terminal unsaturated fatty acids. Arch. Biochem. Biophys., 260, 540-545. Salaiin, J.P., Weissbart, D., Durst, F., Pflieger P., and Mioskowski C. (1989) Epoxidation of cis and trans A 9 -unsaturated lauric acids by a 5 cytochrome P450-dependent system from higher plant microsomes. FEBS Letters, 246, 120-126. Salafin, J.P., Weissbart, D., Helvig, C., Durst, F., Pflieger, P., Bosch, H., and Mioskowski, C. (1992) Stereochemistry of oxidized fatty acids generated during catalytic oxygenation of lauric acid and unsaturated 10 analogs by plant microsomes. FEBS Letters, 303, 109-112. Salatin, J.P., Weissbart, D., Helvig, C., Durst, F., and Mioskowski, C. (1993) Regioselective hydroxylation and epoxidation of lauric acid and unsaturated analogues by cytochrome P450 in Jerusalem artichoke microsomes. Plant Physiol. Biochem., 31, 285-293. 15 SalaUn, J.P., and Helvig, C. (1995) Cytochrome P450-dependent oxidation of fatty acids. Drug Metabol. Drug Interact., 12, 261-283. Schiestl, R.H., and Gietz, R.D. (1989) High eficiency transformation of intact yeast cells using single stranded nucleic acids as a carrier. Curr. Genet., 16, 339-346. 20 Schneider, M., Schweizer, P., Meuwly, P., and Metraux, J.P. (1996) Systemic acquired resistance in plants. Int. Rev. Cytology, 168, 303-340.

WO 99/18224 PCT/IB98/01716 67 Schweizer, P., Felix, G., Buchala, A., Muller, C., and Metraux, J.P. (1996a) Perception of free cutin monomers by plant cells. Plant J., 10, 331-341. Schweizer, P., Jeanguenat, A., M6traux, J.-P., and M6singer, E. (1996b) Induction of resistance in barley against Erysiphe graminis f.sp. hordei by 5 free cutin monomers. Physiol. Mol. Plant Pathol., 49, 103-120. Tijet N., Helvig C., Pinot F, Le Bouquin R., Lesot A., Durst F., Salaun J.P, Benveniste I.: Functional expression in yeast and characterization of a clofibrate-inducible plant cytochrome P450 (CYP94A1) involved in cutin monomers synthesis. Biochem. J., 332: 583-589. 10 Tijet n., Helvig c., Pinot F., Le Bouquin R., Lesot A., Durst F., Benveniste I., Salatin JP. Arylphenoxy compounds induce a rapid accumulation of transcripts of a plant cytochrome P450 gene involved in the synthesis of cutin monomers. IVth International Symposium on Cytochrome P450 Biodiversity and Biotechnology. Strasbourg, France. July 12-16, 1998. 15 Urban, P., Cullin, C., and Pompon, D. (1990) Maximizing the expression of mammalian cytochrome P450 monooxygenase activities in yeast cells. Biochimie, 72, 463-472. Weber, N., Fehling, E., Mukhergee, K., Vosmann, K., Dahlke, B., Hellbardt, S., and Zech, W. (1994) Hydroxylated fatty acids: Oleochemicals 20 from plant oils. INFORM, 5, 475. Weissart, D., Salatin, J.-P., Durst, F., Pflieger, P., and Mioskowski, C. (1992) Regioselectivity of plant lauric acid co-hydroxylase: Omega- WO 99/18224 PCT/IB98/01716 68 oxygenation of cis and trans unsaturated lauric acid analogs and epoxidation of terminally olefin by plant cytochrome P450. Biochim. Biophys. Acta, 1124, 135-142 While the present invention has been described in connection 5 with what is presently considered to be practical and preferred embodiments, it is understood that the present invention is not to be limited or restricted to the disclosed embodiments but, on the contrary, is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims. 10 Thus, it is to be understood that variations in the described invention will be obvious to those skilled in the art without departing from the novel aspects of the present invention and such variations are intended to come within the scope of the claims below.

Claims (27)

1. An isolated nucleic acid encoding a plant fatty acid hydroxylase selected from the group consisting of omega hydroxylase, in chain hydroxylase, and functional derivatives thereof.

2. An isolated nucleic acid encoding a plant fatty acid hydroxylase, wherein the plant fatty acid hydroxylase is an omega hydroxylase having a peptide sequence of SEQ ID NO:2 or a functional derivative thereof, which hydroxylates a fatty acid substrate at a terminal position.

3. The isolated nucleic acid of Claim 2 wherein the omega hydroxylase is selected from the group consisting of CYP94A1, CYP94A2, CYP94A3, CYP94A4, CYP94A5 and CYP94A6.

4. The isolated nucleic acid of Claim 1 wherein the plant fatty acid hydroxylase is an in-chain hydroxylase (CYP81) or a functional derivative thereof, which hydroxylates a fatty acid substrate at a subterminal position.

5. The isolated nucleic acid of Claim 4 wherein the in-chain hydroxylase is CYP81B1.

6. A recombinant nucleic acid comprising the isolated nucleic acid of any one of Claims 1-5.

7. The recombinant nucleic acid of Claim 6 further comprising a WO 99/18224 PCT/IB98/01716 70 regulatory region which is suitable for expression of the plant fatty acid hydroxylase in a host cell.

8. A host cell comprising the recombinant nucleic acid of Claim 6.

9. The host cell of Claim 8 wherein the host cell is selected from the group consisting of bacterial cell, fungal cell, and plant cell.

10. A transgenic plant comprising the recombinant nucleic acid of Claim 6.

11. A host cell comprising the recombinant nucleic acid of Claim 7.

12. The host cell of Claim 11 wherein the host cell is selected from the group consisting of bacterial cell, fungal cell, and plant cell.

13. A transgenic plant comprising the recombinant nucleic acid of Claim 7.

14. A plant fatty acid hydroxylase encoded by the isolated nucleic acid of any one of Claims 1-5.

15. A composition consisting essentially of the plant fatty acid hydroxlyase of Claim 14.

16. A polypeptide produced by the recombinant nucleic acid of WO 99/18224 PCT/IB98/01716 71 Claim 7.

17. A composition consisting essentially of the polypeptide of Claim 16.

18. A process of isolating additional fatty acid hydroxylase genes from a plant by using the isolated nucleic acid of any one of Claims 1-5.

19. The process of Claim 18 wherein the isolated nucleic acid is used as a labeled probe hybridized to a collection of nucleic acids from the plant to select the nucleic acid encoding the additional fatty acid hydroxylase gene.

20. The process of Claim 18 wherein a primer synthesized according to a conserved nucleotide sequence of the isolated nucleic acid amplifies a collection of nucleic acids from the plant to select the nucleic acid encoding the additional fatty acid hydroxylase gene.

21. An isolated nucleic acid selected by the process of Claim 18.

22. A process of altering fatty acid composition in a plant comprising: introducing the isolated nucleic acid of any one of Claims 1-5 into a plant to produce a transgenic plant; expressing the plant fatty acid hydroxylase in the transgenic plant; and hydroxylating or epoxidating a fatty acid substrate in the transgenic plant. WO 99/18224 PCT/IB98/01716 72

23. The process of Claim 22 wherein the fatty acid substrate is a medium-chain fatty acid.

24. The process of Claim 23 wherein the medium-chain fatty acid is selected from the group consisting of capric fatty acid, lauric fatty acid, and myristic fatty acid.

25. The process of Claim 22 wherein the fatty acid substrate is a long chain fatty acid.

26. The process of Claim 25 wherein the long-chain fatty acid is selected from the group consisting of palmitic fatty acid, oleic fatty acid, linoleic fatty acid, and linolenic fatty acid.

27. The process of Claim 25, wherein the fatty acid substrate is a member selected from the group consisting of fatty acids with odd carbon numbers, fatty acids with in-chain hydroxy groups, fatty acids with in-chain epoxy groups, thia-fatty acids, ether-fatty acids, modified fattyacids having ester linkages and modified fatty acids having amide linkages.