摘要
Guayule plants accumulate large quantities of rubber within parenchyma cells of their stembark tissues. This rubber is packed within discrete organelles called rubber particles composed primarily of a lipophilic, cis-polyisoprene core, small amounts of lipids, and several proteins, the most abundant of which is the Mr53,000 rubber particle protein (RPP). We have cloned and sequenced a full-length cDNA for RPP and show that it has 65% amino acid identity and 85% similarity to a cytochrome P450 known as allene oxide synthase (AOS), recently identified from flaxseed. RPP contains the same unusual heme-binding region and possesses a similar defective I-helix region as AOS, suggesting an equivalent biochemical function. Spectral analysis of solubilized RPP verifies it as a P450, and enzymatic assays reveal that it also metabolizes 13(S)-hydroperoxy-(9 Z,11 E)-octadecadienoic acid into the expected ketol fatty acids at rates comparable with flaxseed AOS. RPP is unusual in that it lacks the amino-terminal membrane anchor and the established organelle targeting sequences found on other conventional P450s. Together, these factors place RPP in the CYP74 family of P450s and establish it as the first P450 localized in rubber particles and the first eukaryotic P450 to be identified outside endoplasmic reticulum, mitochondria, or plastids. Guayule plants accumulate large quantities of rubber within parenchyma cells of their stembark tissues. This rubber is packed within discrete organelles called rubber particles composed primarily of a lipophilic, cis-polyisoprene core, small amounts of lipids, and several proteins, the most abundant of which is the Mr53,000 rubber particle protein (RPP). We have cloned and sequenced a full-length cDNA for RPP and show that it has 65% amino acid identity and 85% similarity to a cytochrome P450 known as allene oxide synthase (AOS), recently identified from flaxseed. RPP contains the same unusual heme-binding region and possesses a similar defective I-helix region as AOS, suggesting an equivalent biochemical function. Spectral analysis of solubilized RPP verifies it as a P450, and enzymatic assays reveal that it also metabolizes 13(S)-hydroperoxy-(9 Z,11 E)-octadecadienoic acid into the expected ketol fatty acids at rates comparable with flaxseed AOS. RPP is unusual in that it lacks the amino-terminal membrane anchor and the established organelle targeting sequences found on other conventional P450s. Together, these factors place RPP in the CYP74 family of P450s and establish it as the first P450 localized in rubber particles and the first eukaryotic P450 to be identified outside endoplasmic reticulum, mitochondria, or plastids. Rubber is a macromolecular polyisoprenoid found in over 2000 plant species (1Archer B.L. Audley B.G. Nord F. Miller L. Phytochemistry. Vol. 2. Van Nostrand Reinhold, New York1973: 310-343Google Scholar, 2Backhaus R.A. Isr. J. Bot. 1985; 34: 283-293Google Scholar). It accumulates in discrete, subcellular organelles called rubber particles that are approximately 1 μm in diameter and are composed of a polyisoprene core (3Gomez J.B. Proceedings of the International Rubber Conference. Vol. 2. Rubber Research Institute of Malaya, Kuala Lumpur1975: 143-164Google Scholar, 4Backhaus R.A. Walsh S. Bot. Gaz. 1983; 144: 391-400Crossref Google Scholar) and associated proteins and lipids (5Hasma H. J. Natl. Rubber Res. 1991; 6: 105-114Google Scholar, 6Hasma H. Subramamiam A. J. Nat. Rubber Res. 1986; 1: 30-40Google Scholar, 7Ho C.C. Subramamiam A. Yong W.M. Proceedings of the International Rubber Conference. Vol. 2. Rubber Research Institute of Malaya, Kuala Lumpur1975: 441-456Google Scholar). In most rubber-producing species, such as Hevea brasiliensis, rubber particles are confined to latex vessels, which are specialized plant cells devoted to latex and secondary product synthesis. However, in the North American desert shrub, guayule (Parthenium argentatum), rubber particles accumulate within ordinary stembark parenchyma cells (4Backhaus R.A. Walsh S. Bot. Gaz. 1983; 144: 391-400Crossref Google Scholar). Consequently, its rubber is produced in otherwise normal cells, presenting a unique opportunity for study. Hevea, the commercial source of virtually all natural rubber, produces particles that contain dozens of proteins. Some of these proteins can trigger severe allergies in people who come in contact with products manufactured from Hevea latex. In contrast, guayule particles possess few proteins (8Backhaus R.A. Cornish K. Chen S.-F. Huang D.-S. Bess V.H. Phytochemistry. 1991; 30: 2493-2497Crossref Scopus (27) Google Scholar, 9Cornish K. Siler D.J. Grosjean O.-K. Goodman N. J. Natl. Rubber Res. 1993; 8: 275-285Google Scholar), and none are known to cause allergies, making them a potential alternate source of latex products for medical applications (10Siler D.J. Cornish K. Proceedings of the International Latex Conference: Sensitivity to Latex in Medical Devices. Vol. 34. Food and Drug Administration, Wash., D. C.1992: 46Google Scholar, 11Siler D.J. Cornish K. Ind. Crops Prod. 1994; 2: 307-313Crossref Scopus (82) Google Scholar A major effort is currently underway to identify the function of rubber-associated proteins in plants. From a biochemical perspective, this is less complex for guayule because its particles contain relatively few proteins. The most abundant is the Mr53,000 rubber particle protein (RPP), 1The abbreviations used are:RPPrubber particle protein13 S-HPOD13(S)-hydroperoxy-(9 Z,11 E)-octadecadienoic acidα-ketol12-oxo-13-hydroxy-9(Z)-octadecenoic acidγ-ketol12-oxo-9-hydroxy-10(E)-octadecenoic acidPAGEpolyacrylamide gel electrophoresisPCRpolymerase chain reactionCHAPS3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acidGC-MSgas chromatography-mass spectrometryOTMStrimethylsilyloxy derivativeAOSallene oxide synthasebpbase pair(s). 1The abbreviations used are:RPPrubber particle protein13 S-HPOD13(S)-hydroperoxy-(9 Z,11 E)-octadecadienoic acidα-ketol12-oxo-13-hydroxy-9(Z)-octadecenoic acidγ-ketol12-oxo-9-hydroxy-10(E)-octadecenoic acidPAGEpolyacrylamide gel electrophoresisPCRpolymerase chain reactionCHAPS3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acidGC-MSgas chromatography-mass spectrometryOTMStrimethylsilyloxy derivativeAOSallene oxide synthasebpbase pair(s). which has been purified (8Backhaus R.A. Cornish K. Chen S.-F. Huang D.-S. Bess V.H. Phytochemistry. 1991; 30: 2493-2497Crossref Scopus (27) Google Scholar). It comprises approximately 50% of the protein in guayule particles and has been implicated as a rubber transferase (12Benedict C.R. Madhavan S. Greenblatt G.A. Venkatachalam K.V. Foster M.A. Plant Physiol. 1990; 92: 816-821Crossref PubMed Scopus (37) Google Scholar), the enzyme that catalyzes the polymerization of thousands of isoprenes into molecules of rubber (1Archer B.L. Audley B.G. Nord F. Miller L. Phytochemistry. Vol. 2. Van Nostrand Reinhold, New York1973: 310-343Google Scholar, 2Backhaus R.A. Isr. J. Bot. 1985; 34: 283-293Google Scholar). rubber particle protein 13(S)-hydroperoxy-(9 Z,11 E)-octadecadienoic acid 12-oxo-13-hydroxy-9(Z)-octadecenoic acid 12-oxo-9-hydroxy-10(E)-octadecenoic acid polyacrylamide gel electrophoresis polymerase chain reaction 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid gas chromatography-mass spectrometry trimethylsilyloxy derivative allene oxide synthase base pair(s). rubber particle protein 13(S)-hydroperoxy-(9 Z,11 E)-octadecadienoic acid 12-oxo-13-hydroxy-9(Z)-octadecenoic acid 12-oxo-9-hydroxy-10(E)-octadecenoic acid polyacrylamide gel electrophoresis polymerase chain reaction 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid gas chromatography-mass spectrometry trimethylsilyloxy derivative allene oxide synthase base pair(s). In this report, we now show that RPP is a member of the CYP74 family of cytochromes P450 and is likely not a prenyltransferase. This is based on the deduced amino acid sequence of RPP cDNAs isolated from a guayule stembark cDNA library. RPP shows significant sequence identity to the P450 known as allene oxide synthase (AOS) (13Song W.C. Funk C.D. Brash A.R. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 8519-8523Crossref PubMed Scopus (269) Google Scholar). Additional biochemical analysis of partially purified RPP solubilized from washed rubber particles reveals a difference spectrum typical of P450s and, like AOS, an ability to rapidly metabolize hydroperoxylinoleic acid into α- and γ-ketol fatty acids (14Song W.C. Brash A.R. Science. 1991; 253: 781-784Crossref PubMed Scopus (230) Google Scholar). RPP and AOS share a number of significant features that distinguish them from conventional P450s. However, RPP also has several unique characteristics that differentiate it from flaxseed AOS. Guayule, P. argentatum Gray, line 11591, was used for all experiments. Tissues were collected from field-grown shrubs at the U. S. Department of Agriculture (Water Conservation Laboratory, Phoenix, AZ). Rubber particles were purified by flotation from homogenized stembark tissues (15Cornish K. Backhaus R.A. Phytochemistry. 1990; 29: 3809-3813Crossref Scopus (100) Google Scholar). Particles were subjected to preparative SDS-PAGE (8Backhaus R.A. Cornish K. Chen S.-F. Huang D.-S. Bess V.H. Phytochemistry. 1991; 30: 2493-2497Crossref Scopus (27) Google Scholar). RPP migrated as a single Mr53,000 band that was purified from the gel by electroelution (model 422 Electroeluter, Bio-Rad) according to the manufacturer's instructions. Purified RPP was subjected to CNBr cleavage by dissolving approximately 1 nmol of lyophilized protein in 150 μl of 70% formic acid followed by 100 μl of 70 μg of CNBr ml−1 in 70% formic acid. The mixture was incubated in the dark at room temperature for 24 h, and the peptide fragments were separated on a 16% acrylamide gel (16Schagger H. Von Jagow G. Anal. Biochem. 1987; 166: 368-379Crossref PubMed Scopus (10469) Google Scholar) and blotted onto polyvinylidene difluoride (Millipore) membranes (17Ploug M. Jensen A.L. Barkholt V. Anal. Biochem. 1989; 181: 33-39Crossref PubMed Scopus (147) Google Scholar). Stained bands were subjected to amino-terminal sequencing, performed at the Protein Structure Lab (U. C. Davis), using an ABI 470A gas phase sequenator. Four peptide fragments yielded the following amino acid sequences: 1, PLTKSVVYESLRIEPPV; 2, EQAEKLGVPKDEAVHNILFAVCFNTFGGVK; 3, LFGYQPFATKDPKVFDRPEEFVPDRFVGDGEALLKY; 4, LKNSSNRVIPQFETTYTELFEGLEA. Total RNA was isolated from guayule stembark tissues using the procedure of Logemann et al. (18Logemann J. Schell J. Willmitzer L. Anal. Biochem. 1987; 163: 16-20Crossref PubMed Scopus (1608) Google Scholar). Tissues were harvested between November and March when rubber synthesis is highest. Poly(A)+ RNA was used to construct the λZAP stembark cDNA library according to Short et al. (19Short J.M. Fernandez J.M. Sorge J.A. Triman K.L. Nucleic Acids Res. 1988; 16: 7583-7600Crossref PubMed Scopus (1080) Google Scholar) following the manufacturer's instructions (Stratagene). The library yielded 3 × 106recombinants before amplification. Approximately 1.5 × 105plaque-forming units from the original cDNA library were screened. Polyadenylated RNAs purified from stembark were used for first-strand cDNA synthesis (20Kawasaki E.S. Innis M.A. Gelfand D.H. Sninsky J.J. White T.J. PCR Protocols. Academic Press, San Diego, CA1990: 21-26Crossref Google Scholar). Reactions were carried out in a 100-μl total volume containing 5 μg of poly(A)+ RNA, 1 m M dNTPs, 0.1 μg of oligo(dT)12–18 (Amersham Corp.), 1 unit of RNase block II, and 100 units of M-MuLV reverse transcriptase (Stratagene). Polymerase chain reactions were carried out in a 100-μl total volume using 1-2 units of Replinase (DuPont NEN) or AmpliTaq (Perkin-Elmer Corp.) with the following components: 1% of first-strand cDNA product, 50 pmol of each primer, 250 μM dNTPs in PCR buffer using 30 cycles (1 min at 94°C, 1 min at 45°C, and 1 min at 72°C). PCR reactions used degenerate primers specific for RPP peptide fragment 3 (sense strand P5, 5′-TTYGGNTAYCARCYNTTYGC-3′; and antisense strand P6, 5′-GCYTCNCCRTCNCCNACRAA-3′). Coding redundancies are: N = C, T, A, or G; H = A, C, or T; S = G or C; R = A or G; B = C, G, or T; K = G or T; Y = C or T; V = A, C, or G; W = A or T; M = A or C; D = A, G or T. The 92-bp product was sequenced according to Kretz et al. (21Kretz K.A. Carson G.S. O'Brien J.S. Nucleic Acids Res. 1989; 17: 5864Crossref PubMed Scopus (111) Google Scholar) to verify that it matched the sequence for peptide fragment 3. A second pair of primers (sense strand P1, 5′-ATHCYNCARTTYGARAC-3′; and antisense strand P9, 5′-TTNACNCCNCCRAANGTRTTRAA-3′) produced a 434-bp product that also matched the upstream portion of RPP. These 92- and 434-bp RPP probes were used to screen the cDNA library. Plaque lifts were prepared using Colony/Plaque Screen (DuPont NEN) and probed with a [α-32P]dATP-labeled PCR fragment (22Sambrook J. Fritch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd Ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY1989Google Scholar). Hybridizations were carried out at 65°C for 16 h in 5 × SSC (1 × SSC is 0.15 M NaCl, 0.015 M sodium citrate), 50 m M sodium phosphate (pH 7.4), 5 × Denhardt's solution (1 × Denhardt's solution is 0.02% Ficoll, 0.02% polyvinylpyrrolidone, 0.02% bovine serum albumin), 0.5% SDS, 10% dextran sulfate. Filters were washed four times as follows: 1) 10 min, 2 × SSC at room temperature; 2) 10 min, 2 × SSC, 1% SDS at room temperature; 3) 30 min, 2 × SSC, 1% SDS at 65°C, and 4) 10 min, 0.1 × SSC, 0.1% SDS at room temperature. The filters were exposed to Kodak x-ray films with intensifying screens at −80°C. Plaques giving the strongest hybridization signals were isolated and rescreened under the same conditions. Phagemids were obtained from purified positive plaques through in vivo excision as described in the manufacturer's protocol (Stratagene). Over 30 RPP cDNAs were isolated to yield four full-length clones, one of which is the clone described herein. All sequencing was by the dideoxy technique (23Kraft R. Tardiff J. Kraauter K.S. Leinwand L.A. Bio Technology. 1988; 6: 544-547Google Scholar) using Sequenase (U.S. Biochemical Corp.). Sequence comparisons to the genetic data base were made using the BLAST algorithm (24Altschul S.F. Gish W. Miller W. Myers E.W. Lipman D.L. J. Mol. Biol. 1990; 215: 403-410Crossref PubMed Scopus (70300) Google Scholar). Rubber particles isolated from stembark homogenates were purified by washing at least three times using the flotation and centrifugation procedure (15Cornish K. Backhaus R.A. Phytochemistry. 1990; 29: 3809-3813Crossref Scopus (100) Google Scholar), except that 50 m M potassium phosphate (pH 7.0) was used as the extraction and wash buffer in place of Tris-HCl. RPP was solubilized by sonicating 3×-washed particles in 0.5% CHAPS for 1 min. This was centrifuged at 14,000 × g for 5 min, and the solution beneath the rubber layer was removed and passed through a Millex-SLGV 0.22-μm filter (Millipore). This eliminated all remaining traces of rubber and yielded a clear solution containing the solubilized RPP. Each step of the protein isolation was monitored by SDS-PAGE using 0.75-mm slab gels of 10% (w/v) polyacrylamide stained with Coomassie Blue. Scanning densitometry of the Coomassie-stained gels indicated that the RPP band comprised about 50% of the total protein in rubber particles. The clarified RPP solution was used for P450 spectral analysis and enzyme assays. The concentration of total protein in this clarified preparation was 363 ng μl−1 based on a comparison against a bovine serum albumin standard. Additional two-dimensional gel electrophoresis revealed that the RPP band contained a single polypeptide and not a mixture of polypeptides. P450 difference spectra were obtained by diluting the CHAPS-solubilized RPP preparation in 50 m M potassium phosphate (pH 7.0) and distributing it to 3-ml reference and sample cuvettes. Both cuvettes were reduced with Na-dithionite, and the sample cuvette was bubbled with CO to produce a difference spectrum with a characteristic peak at 451 nm. The fully developed chromophore appeared after 5 min. The quantity of P450 in the solubilized RPP preparation, determined from the differential absorption of the CO-bound reduced form of P450, using the value of A450-A490 = 91 m M cm−1 (26Omura T. Sato R. J. Biol. Chem. 1964; 239: 2370-2378Abstract Full Text PDF PubMed Google Scholar, 27Omura T. Sato R. J. Biol. Chem. 1964; 239: 2379-2385Abstract Full Text PDF PubMed Google Scholar), was 0.16 pmol μl−1 or 36.7 ng μl−1. Using this value, the content of P450 per total protein was approximately 10%. Comparing this to the densitometry values, we estimated that the range of P450/mg of protein in the particles was between 10-50%. To test whether RPP was capable of metabolizing lipid hydroperoxides, the same filtered, CHAPS-solubilized preparations were used. AOS activity was measured spectrophotometrically (14Song W.C. Brash A.R. Science. 1991; 253: 781-784Crossref PubMed Scopus (230) Google Scholar) using purified 13(S)-hydroperoxy-(9 Z,11 E)-octadecadienoic acid (13 S-HPOD) that was prepared according to Ref. 25Gardner H.W. Newton J.W. Phytochemistry. 1987; 26: 621-626Crossref Scopus (28) Google Scholar. The reaction was initiated by adding 3 μl of CHAPS-solubilized RPP to 3 ml of 50 m M potassium phosphate buffer (pH 7.0) containing 13 S-HPOD at a final concentration of 53.8 μM (1.4 absorbance units at 234 nm). The quantity of P450 in this solubilized RPP preparation was 0.48 pmol based on P450 difference spectra (26Omura T. Sato R. J. Biol. Chem. 1964; 239: 2370-2378Abstract Full Text PDF PubMed Google Scholar, 27Omura T. Sato R. J. Biol. Chem. 1964; 239: 2379-2385Abstract Full Text PDF PubMed Google Scholar). AOS activity was measured by a series of UV spectra taken at 10-s intervals, which showed the rapid degradation of substrate due to the loss of its conjugated diene. One-half of the substrate disappeared after 45 s, resulting in an estimated kcat of 3700 s−1 for this enzyme. A large scale reaction was prepared using 20 μl of CHAPS-solubilized RPP (7.2 μg of protein) and 2 mg of 13 S-HPOD in 6 ml of 50 m M potassium phosphate buffer (pH 7.0) incubated at 22°C for 60 min. The products of the AOS reaction were extracted with 1.5 volumes of a 2:1 (v/v) mixture of chloroform:methanol. The chloroform-extracted products were separated as the free fatty acids on precoated, silica gel TLC plates (T-6270, Sigma) according to the method of Gardner (28Gardner H.W. J. Lipid Res. 1970; 11: 311-321Abstract Full Text PDF PubMed Google Scholar) using isooctane:ether:acetic acid (50:50:1, v/v/v) with detection by 2,4-dinitrophenylhydrazine spray (29Vioque E. Holman R.T. Arch. Biochem. Biophys. 1962; 99: 522-528Crossref PubMed Scopus (79) Google Scholar). For isolation, the products were methyl-esterified by a brief exposure to diazomethane in diethyl ether:CH3OH (9:1, v/v) and separated by TLC (20 × 20 × 0.025 cm, silica 60 F-254 plates, Merck) using development by hexane:ethyl ether (3:2, v/v). Detection was with a non-destructive spray, 0.1% 8-anilino-1-naphthalenesulfonic acid (Na salt) followed by long UV viewing. Fluorescent bands were scraped and extracted with ethyl acetate. The recovered products were examined by nuclear magnetic resonance (NMR), 1H-NMR and/or 13C-NMR, using a Bruker model ARX-400 spectrometer (Karlsruhe, Germany) with the samples dissolved in CDCl3(CDCl3served as internal reference). Gas chromatography-mass spectrometry (GC-MS) was completed by a Hewlett Packard model 5890 (capillary column, Hewlett Packard HP-5MS cross-linked 5% phenylmethyl silicone, 0.25 mm × 30 m, film thickness 0.25 mm) interfaced with a model 5971 mass selective detector operating at 70 eV. The temperature program was from 160 to 260°C at 5°C min−1 with a hold at 260°C for 10 min. The helium flow rate was 0.67 ml min−1. All products were examined by GC-MS as their methyl ester/trimethylsilyloxy (OTMS) derivatives. OTMS derivatives of hydroxy groups were synthesized with trimethylchlorosilane:hexamethyl-disilazane:pyridine (3:2:2, v/v/v). Other derivatives for GC-MS were produced by either NaBH4reduction of the ketone group or reduction of the double bond by H2with a 5% palladium on CaCO3catalyst followed by treatment with OTMS reagent. NH2-terminal Edman sequencing was performed on four CNBr fragments of gel-purified RPP (Table I). RPP cDNA was cloned using a strategy based on these amino acid sequences. Degenerate oligonucleotide primers were used to produce RPP-specific DNA fragments by PCR. These were used as hybridization probes to screen our cDNA library and provided over 30 clones, four of which contained putative full-length sequences.Table I:N-terminal sequence analysis of the CNBr fragments from the guayule rubber particle proteinThe pmol yield for each cycle is indicated. Inconclusive assignments are indicated in parentheses. The pmol yield for each cycle is indicated. Inconclusive assignments are indicated in parentheses. The full-length RPP cDNA contained 1692 nucleotides, including 23 nucleotides of 5′- and 250 nucleotides of 3′-untranslated sequence. It encoded a protein of 473 amino acids, with an NH2-terminal methionine, a COOH-terminal isoleucine, and two internal cysteines (Fig. 1). The deduced Mr53,438 protein contained the four predicted CNBr peptide fragments, and its deduced isoelectric point (pI 6.15) matched the experimental value (pI 6.2) of purified RPP (8Backhaus R.A. Cornish K. Chen S.-F. Huang D.-S. Bess V.H. Phytochemistry. 1991; 30: 2493-2497Crossref Scopus (27) Google Scholar). A comparison of the deduced and measured amino acid composition (8Backhaus R.A. Cornish K. Chen S.-F. Huang D.-S. Bess V.H. Phytochemistry. 1991; 30: 2493-2497Crossref Scopus (27) Google Scholar) was also in agreement. Primer extension analysis revealed that RPP mRNA extended about 35 nucleotides upstream from the 5′-end of the RPP cDNA (data not shown), indicating that the cDNA was a full-length clone for RPP. The sequence flanking the first AUG initiation codon (5′-AAAACAUGG-3′) represented high homology to the consensus AUG start site (5′-TAAACAAUGG-3′ and 5′-AAAA(A/C)AUGG-3′) for translation initiation in plants (30Cavener D.R. Ray S.C. Nucleic Acids Res. 1991; 19: 3185-3192Crossref PubMed Scopus (526) Google Scholar, 31Joshi C.P. Nucleic Acids Res. 1987; 15: 6643-6653Crossref PubMed Scopus (685) Google Scholar), providing additional evidence that this cDNA contained the entire RPP-coding sequence. A search of the genetic sequence data base identified two small regions in RPP that shared significant homology with two small regions of several cytochrome P450s (32Kalb V.F. Loper J.C. Proc. Natl. Acad. Sci. U. S. A. 1988; 85: 7221-7225Crossref PubMed Scopus (84) Google Scholar, 33Nelson D.R. Kamataki T. Waxman D.J. Guengerich F.P. Estabrook R.W. Feyereisen R. Gonzalez F.J. Coon M.J. Gunsalus I.C. Gotoh S. Okuda K. Nebert D.W. DNA Cell Biol. 1993; 12: 1-51Crossref PubMed Scopus (1649) Google Scholar). Those regions aligned with two highly conserved B and C domains (32Kalb V.F. Loper J.C. Proc. Natl. Acad. Sci. U. S. A. 1988; 85: 7221-7225Crossref PubMed Scopus (84) Google Scholar), which are common to all P450s (Fig. 2). Additionally, RPP contained a conserved PPGP tetrapeptide near the amino terminus (Fig. 1) that is also common in P450s and is known to participate in P450 stability and catalysis (34Szczesna-Skorpa E. Straub P. Kemper B. Arch. Biochem. Biophys. 1993; 304: 170-175Crossref PubMed Scopus (40) Google Scholar). However, RPP did not contain the highly conserved A and D domains that are essential components of most P450s. Careful comparison of the RPP sequence with that of CYP73, a plant P450 known as cinnamic acid hydrolase (CA4H) (35Teutsch H.G. Hasenfratz M.P. Lesot A. Stoltz C. Garnier J. Jeltsch J. Durst F. Werck-Reichhart D. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 4102-4106Crossref PubMed Scopus (194) Google Scholar), suggested that the CAG of RPP could act as a putative heme-binding site necessary for P450 action. However, the amino acid sequence surrounding this CAG (Fig. 1) was inconsistent with the known P450 consensus decapeptide of F-(SGNA)- X-(GD)- X-(RHPT)- X-C-(LIVMFAP)-(GAD) (32Kalb V.F. Loper J.C. Proc. Natl. Acad. Sci. U. S. A. 1988; 85: 7221-7225Crossref PubMed Scopus (84) Google Scholar). This, together with the absence of the A domain, suggested that RPP might not act as a true P450. Thus, studies were performed to verify whether RPP possessed the spectral characteristics of P450s. RPP solubilized from washed rubber particles did, in fact, yield a difference spectra for the CO-bound, dithionite-reduced form that clearly indicated that it was a P450 with an absorbance maximum at 451 nm (Fig. 3).FIG. 3Difference spectra of purified RPP. Reference and sample cuvettes containing RPP were reduced with Na-dithionite, and the sample cuvette was bubbled with CO to produce a characteristic peak at 451 nm. The fully developed chromophore appeared after 5 min.View Large Image Figure ViewerDownload Hi-res image Download (PPT) Subsequent to this, we discovered that RPP had significant homology with the amino acid sequence for AOS from flaxseed, whose cDNA was recently described (13Song W.C. Funk C.D. Brash A.R. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 8519-8523Crossref PubMed Scopus (269) Google Scholar). RPP showed over 65% identity and 85% similarity to AOS, which is now known to be an atypical P450, classified as CYP74 (33Nelson D.R. Kamataki T. Waxman D.J. Guengerich F.P. Estabrook R.W. Feyereisen R. Gonzalez F.J. Coon M.J. Gunsalus I.C. Gotoh S. Okuda K. Nebert D.W. DNA Cell Biol. 1993; 12: 1-51Crossref PubMed Scopus (1649) Google Scholar). AOS, noted for its unusual heme-binding site (13Song W.C. Funk C.D. Brash A.R. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 8519-8523Crossref PubMed Scopus (269) Google Scholar), possesses a 15-amino acid peptide sequence that is nearly identical to the putative heme-binding CAG region of RPP (Fig. 4). This sequence, beginning at RPP position 419, shows over 86% identity with the corresponding flaxseed AOS sequence (Fig. 4) and hints that RPP and AOS might catalyze similar reactions. AOS, also previously known as hydroperoxide isomerase (36Zimmerman D.C. Vick B.A. Plant Physiol. 1970; 46: 445-453Crossref PubMed Google Scholar, 37Zimmerman D.C. Biochem. Biophys. Res. Commun. 1966; 23: 398-403Crossref PubMed Scopus (93) Google Scholar), hydroperoxide cyclase (38Zimmerman D.C. Feng P. Lipids. 1978; 13: 313-316Crossref Scopus (120) Google Scholar), fatty acid hydroperoxide dehydrase (39Hamberg M. Biochim. Biophys. Acta. 1987; 920: 76-84Crossref Scopus (103) Google Scholar), and hydroperoxide dehydratase (EC 4.2.1.92), is classified as a hydroperoxide-dependent P450. It metabolizes lipid hydroperoxides into their corresponding allene epoxides by intramolecular oxygen transfer (39Hamberg M. Biochim. Biophys. Acta. 1987; 920: 76-84Crossref Scopus (103) Google Scholar). Enzyme analysis was performed using RPP solubilized from washed rubber particles. SDS-PAGE gels indicated that the preparations were greatly enriched for RPP (Fig. 5). The AOS activity of the CHAPS-solubilized preparation (Fig. 5, lane 3) had an estimated kcat of 3700 s−1, as determined by scanning UV spectroscopy. This is within the same order of magnitude as flaxseed AOS, which has a kcat of approximately 1000 s−1 (14Song W.C. Brash A.R. Science. 1991; 253: 781-784Crossref PubMed Scopus (230) Google Scholar). Analysis of the reaction products by TLC, NMR, and GC-MS also confirmed that RPP had AOS activity. These products consisted of the same α- and γ-ketol fatty acids that are observed with flaxseed (14Song W.C. Brash A.R. Science. 1991; 253: 781-784Crossref PubMed Scopus (230) Google Scholar, 36Zimmerman D.C. Vick B.A. Plant Physiol. 1970; 46: 445-453Crossref PubMed Google Scholar, 37Zimmerman D.C. Biochem. Biophys. Res. Commun. 1966; 23: 398-403Crossref PubMed Scopus (93) Google Scholar, 38Zimmerman D.C. Feng P. Lipids. 1978; 13: 313-316Crossref Scopus (120) Google Scholar) and corn AOS (28Gardner H.W. J. Lipid Res. 1970; 11: 311-321Abstract Full Text PDF PubMed Google Scholar, 39Hamberg M. Biochim. Biophys. Acta. 1987; 920: 76-84Crossref Scopus (103) Google Scholar). The α- and γ-ketols result from the rapid degradation of allene oxide produced by the AOS reaction. The α-ketol, as its methyl ester, was isolated by TLC (RF = 0.34), affording a 56% yield based on the amount of utilized 13 S-HPOD (1.96 mg). 1H-NMR of the α-ketol (methyl ester) furnished the following data in chemical shifts in ppm and number of protons, multiplicity, coupling constants and carbon assignments: 5.62 (1H, dt, J9,10 = 10.8 Hz; J10,11 = 7.3 Hz, C-10 or −9), 5.51 (1H, dt, C-9 or −10), 4.23 (1H, m, C-13), 3.67 (3H, s, OCH3), 3.40 (1H, d, J = 5.0 Hz, OH at C-13), 3.22 and 3.24 (1H each, d, J10,11 = 7.3 Hz, C-11a, b), 2.29 (2H, t, J = 7.5 Hz, C-2), 2.01 (2H, m, C-8), 1.48 and 1.81 (1H each, m, C-14a, b), 1.60 (2H, m, C-3), 1.30 (14H, m, C-4 to −7 and C-15 to −17), 0.88 (3H, t, C-18). The assignments were confirmed by two-dimensional proton correlative spectroscopy. A significant feature of the 1H-NMR spectrum was the coupling J9,10 = 10.8 Hz establishing the (Z) configuration of the double bond. An unusual splitting of the C-13 OH proton was due to hydrogen bonding to the vicinal ketone, which inhibited the normal hydrogen exchange and lack of coupling of hydroxyl hydrogens. The 13C-NMR of the α-ketol (methyl ester) furnished an absorbance