THEJOURNAL OF BIOLOGICALCHEMISTRY Vol. 257. No. 5, Issue of March 10. pp. Y335-2342,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 1982 PrintedzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAinzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA US.A.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Biosynthesis of (&)-a-Pineneand (-)-p-Pinene from Geranyl Pyrophosphate by a Soluble Enzyme Systemfrom Sage (Saluia officinalis)*
(Received for publication, August 17, 1981)
Herve Gambliel and Rodney CroteauS From the Institute of Biological Chemistry, and Biochemistry/Biophysics Program, Washington State University, Pullman, Washington 99164
The volatile oil of immature sage (Salvia officinalis) distributed monoterpenes in the plant kingdom and are the leaves was shown to contain (f)-a-pinene and (--)+- major constitutentsof the various turpentines (1). Both opti- pinene as major constituents, and solubleenzyme prep- cal antipodes of a-pinene are natural products, and theymay arations from this tissue catalyzed the divalent cation-co-occur with either isomer predominant (1-3). By contrast, dependent conversion of the acyclic Clo precursor ger- P-pinene almost always occurs as the optically pure (--)-[lS any1 pyrophosphate to a-pinene and P-pinene and to 5S] isomer (I), the (+)-[1R:5R] antipode apparently being of zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA lesser amountsof other monoterpeneolefins. The iden- rather limited distribution (2). Ruzicka et al. (4)proposed an tities of the principalbiosynthetic products as a-pinene ionic scheme fort.he formation of the pinenes (Fig. 2, Scheme and P-pinenewere confirmed by chromatographic I) involving cyclization of the neryl cation to a monocyclic Downloaded from analysis andby the synthesisof crystalline derivatives. a-terpinyl intermediatewhich undergoes internal Markowni- The biosynthetic pinenes derived from [l-3H]geranyl koff addition to a piny1 cation that is subsequently stabilized pyrophosphatewere stereospecifically converted to of vivo camphor, and exchangeof the a-hydrogens of this ke- by two variants proton loss. On the basis of recent in tone established that all the tritium of the original time course studies with Pinuspinaster(5), the acyclic trienes pinane nucleus was located on the methylene bridge cis-ocimene and myrcene were suggested as precursors of a- carbon (C-7), thus demonstratingspecific conversion of and /I-pinene, respectively (Fig.2, Scheme ZI),reviving eariier www.jbc.org the acyclic precursor. Stereospecific conversion of the radical-type cyclization schemes for the construction of the biosynthetic pinenes to fenchone oxime, followed by pinenes (4, 6). Finally, the favorableisomerization of the resolution of thecorresponding diastereomeric (-)- exocyclic to the endocyclic isomer (7, 8) raises the possibility methyloxymethyl ethers, confirmed the stereochemis- of a single cyclization of an acyclic precursor to P-pinene, by guest, on December 7, 2011 try of the original products as (k)-[7-3H]a-pinene and followed by enzymic (or nonenzymic) conversion to 0-pinene (-)-[7-3H]P-pinene. The possibility of isomerization of (Fig. 2, Scheme IZI).The labeling pattern of a-pinene derived P-pinene to a-pinene was eliminated, as was the in- from exogenous [2-’4C]mevalonate in Pinus attenuata (9) is volvement of other free intermediates, thus indicating consistent with any of the above biogenetic schemes and no direct cyclization of geranyl pyrophosphateto themix- experimental evidence to distinguish among the alternatives ture of pinene isomers. NeryI pyrophosphate andIina- is available. While the origin of the pinenes is thus uncertain, loyl pyrophosphate could also function as acyclic pre- the results of chemotaxonomic studies (10) and in vivo tracer cursors of the pinenes and of the other monoterpene studies (11) suggest that the various isomers probably arise olefins, but only geranylpyrophosphate afforded a via distinct routes involving variants both in the stereochem- product distribution comparable thatto formed in vivo. istry of ring formation and in the generation of the double Preliminary examination of the cell-free systemzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA sug- bond in either the endocyclic or exocyclic position. The fact gested the presence of more than one pinene cyclase activity. that positionally isomeric and enantiomeric pinenesco-occur, with the combination (+)-a-pinene and (-)-P-pinene being particularly common (12, 13), furtherimplies that such mul- tiple routes are operativein the same organism. &-Pinene’ and/I-pinene (Fig. 1) are among the mostwidely Crude cell-free extracts from Citrus Zimonum (14), Pinus radiata (15), and common sage (Salvia officinalis) (16, 17) * This investigationwas supported in part by grants from the convert the acyclic C10 precursors GPP and NPP to products National Science Foundation (PCM 78-19417) andHaarmann and chromatographically coincident with a-pinene and ,&pinene. Reimer GmbH. Thisis Scientific Paper6014, Project 0268, College of Agriculture Research Center, WashingtonState University, Pullman, However, the identitiesof the biosynthetic productswere not Washington99164. The costs of publication of this article were confirmed, and, although both tritium atoms from [1-”H2,U- defrayed in part by the payment of page charges. This article must I4C]GPP were shownto be retained in the pinenes inthe latter therefore be hereby marked “advertisement” in accordance with 18 study (17), the labeling pattern and the stereochemistryof the U.S.C. Section 1734 solely to indicate this fact. products were not verified. Furthermore, the possible isom- $ To whom correspondence should be addressed. ’ The abbreviations and trivial names used are: a-pinene, 2,6,6- -__ trimethylbicyclo[3.1.l]hept-2-ene;GPP, pyrophosphate ester of ger- myrcene, 3-methylene-7-methyl-1,6-octadiene;ocimene, 3,7-di- aniol (3,7-dimethylocta-2(trans),6-dienol);NPP, pyrophosphate ester methyl-l,3(cis-or trans-),6-octatriene;a-terpineol, l-methyl-4-(2-hy- of nerol (3,7-dimethylocta-2(cis),6-dienol);LPP, pyrophosphate ester droxyisopropy1)-1-cyclohexene;borneol, 1,7,7-trimethylbicyclo[2.2.1] of linalool (3,7-dimethylocta-l,6-dien-3-ol);P-pinene, 6,6-dimethyl-2- heptan-2-oUendo);1,8-cineole, 1,3,3-trimethyl-2-oxabicyclo[2.2.2]0~- methylenebicyclo[3.1.l]heptane; camphene, 2,2-dimethyl-3-methyl- tane; MES, 2-[N-morpholino]ethanesulfonic acid; Tricine, N-[2-hy- enebicyclo[2.2.l]heptane; terpinolene, l-methyl-4-isopropropyl~ene- droxy-1,l-bis(hydroxymethyl)ethyl]glycine;GLC, gas-liquid chroma- 1-cyclohexene; limonene, I-methyl-4-isopropenyl-1-cyclohexene; tography; TLC, thinlayer chromatography. 2335 2336 Biosynthesis of Isomeric Pinenes from Geranyl Pyrophosphate
zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAverifiedzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA as before (16) by analysis of the products of alkaline phos- phatase-apyrase hydrolysis and acid hydrolysis. (+)-a-Pinene, (-)-a-pinene, and (-)+pinene were obtained from Aldrich Chemical Co., cis- and trans-pinanol were from PCR Re- search Chemicals Inc., (+)-limonene was from ICN Pharmaceuticals, terpinolene was from Haarmann and Reimer GmbH, Holzminden, b West Germany, and the ocimenes were from International Flavors and Fragrances. All other monoterpenes were purchased from Pfaltz a b C and Bauer, Inc. and were purified by distillation or chromatography before use. For isotopic dilution experiments, the monoterpenes were zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAsuspended in water with the aid of Tween 20 (40 pg/pmol of terpene) and sonication (needle probe of a Biosonik 111 at full power). Insoluble polyvinylpyrrolidone (Polyclar AT, GAF Corp.) and Amberlite XAD- 4 polystyrene resin (Rohm and HaasCorp.) were purified by standard procedures for use as adsorbents (23,24). All other biochemicals and reagents were obtained from Sigma or Aldrich. Enzyme Preparation-Sage leaves (2-3 g) were homogenized in a Ten-Broeck homogenizer containing a slurry of an equivalent leaf weight of insoluble polyvinylpyrrolidone in cold 75 mM MES-5 mM sodium phosphate buffer, pH 6.5, containing 20% glycerol, 5 mM f dithioerythritol, 10 mM Na2S205,10 mM sodium ascorbate, 15 mM MgC12. The homogenate was then slurried with an equal tissue weight P of hydrated XAD-4 polystyrene resin and passed through four layers d e of cheesecloth. Centrifugation at 27,000 X g for 20 min (pellet dis- zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAcarded), then at 105,000 X g for 1 h, provided the soluble supernatant
that, after dialysis to assay conditions, was used as theenzyme source. Downloaded from Enzyme Assays-A typical reaction mixture contained 0.2 to 0.5 mg of protein in a total volume of 1 ml of 10 mM MES-5 mM sodium phosphate buffer or 10 mM Tricine-5 mM sodium phosphate buffer, pH 6.5, containing 0.5 mM dithioerythritol, 15 mMMgC12 (or 1 mM MnC12) in a Teflon-sealed screw cap vial. In some experiments, 1 mM sodium ascorbate and 1 mM Na2S205 were included in the assay medium, and the buffer contained 10% glycerol. The reaction was started by the addition of up to10 p~ (10 nmol) of either of the three www.jbc.org acyclic precursors ([1-3H]GPP, [l-”H]NPP, or (*)[1-3H]LPP), and 1 h I ml of pentane was carefully overlaid to trap volatile monoterpene 9 products. The addition of the pentane layer improves the recovery of FIG. 1. Structures of (-)+pinene (a),(+)-&-pinene (b),(-)- monoterpene olefiis by as much as 30% and therefore was used by guest, on December 7, 2011 a-pinene (c),zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA (+)-camphene (d), terpinolenezyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA (e), limonene (f), routinely. After incubation for 1h at 28 “C, the reaction was quenched myrcene (g), cis-ocimene (h),and trans-ocimene (i). by vigorous mixing,and the pentane layer was removed. The aqueous zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAphase was re-extracted with an additional 1 mi of pentane, and the combined extract was eluted through a column (0.5 X 3 cm) of silicic erization of the /3- to thea-isomer was not excluded; thus, no acid (Mallinkrodt, SilicAR CC-7) in a Pasteur pipette, followed by significant conclusions could be drawn from this preliminary washing with another 0.5 ml of pentane. The hydrocarbon fraction work. thus obtained is free of oxygenated products which remain adsorbed on the silicic acid. Internal standards (5 mg each of (+)+pinene, The essential oil of common sage containsa-pinene (-)-/?-pinene, (+)-camphene, myrcene, (*)-limonene, and terpino- (thought to be mainly the (+)-isomer(2)) and &pinene (pre- lene) were then added, and the total volume was adjusted to 3.5 ml. sumed to be the (-)-isomer) as the major terpene hydrocar- A l-ml aliquot was removed for determination of tritium, and the bons, and sage leaf extracts possess the requisite biosynthetic capability (16, 17). This system, which has been utilized for studies on the biosynthesis of p-menthane and bornane mon- oterpenes (18, 19), therefore provided the potential opportu- nity to examine in detailthe biosynthesis of isomeric pinenes. In this communication, we report the conversion of[1-’HI GPP, and other acyclic precursors, to (+)-[7-’H]a-pinene, (-)-[7-’H]a-pinene, and (-)-[7-’H]b-pinene by a soluble en- zyme preparation from sage leaves and describe some of the basic characteristics of this novel biosynthetic system.
EXPERIMENTALPROCEDURES Plants-Sage plants (S. officinalis L.) were grown from seed purchased from GeorgeBall, Pacific Inc. (Sunnyvale, CA) in peat moss/perlite/sand (1:l:l)under long day conditions (16 h; 1800 foot- candles). Rapidly expanding leaves (2-3 cm in length) from the shoot apex of immature plants (3-5 weeks postgermination) were collected and washed with distilled water before use. Substrates and Reagents-The preparation of [l-’H2]GPP (300 SCHEME II Ci/mol) and [1-3H2]NPP (300 Ci/mol) has been described (16, 19). (*)-[l-3H2]Linalool was obtained by mesylation of [l-3H]geraniol in triethylamine (20) followed by solvolysis in water. The product was purified by TLC (Silica Gel G containing 8% AgN03, solvent system A) and itsidentity confirmed by radio GLC. [l-’Hz]LPP (200 Ci/mol) SCHEME Ill was prepared from the alcohol by modification of the Cramer and Bohm method (21, 22) and was purified by ion-exchange chromatog- FIG. 2. Biogenetic schemes for the origin of a-pinene and raphy as described previously (16). The purity of this product was 8-pinene. Biosynthesis of IsomericPinenes fromGeranyl Pyrophosphate 2337 remainder was concentrated under a stream ofzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA nitrogen at 0 "c for M KOH afforded the crude pinonic acid, which was recovered from zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA analysis by radio GLC or TLC (Silica Gel G containing 12% AgN03, the aqueous phase by acidification and re-extraction with ether. This zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAmaterial was taken to dryness under vacuum, and the residue was solvent system B). The oxygenated monoterpene products were recovered by re-ex- subjected to TLC (Silica Gel G, solvent system D) to afford (*)-cis- tracting the original aqueous layer of the incubation mixture with pinonic acid (RF = 0.3) which was recrystallized from hot water ether (1.5 ml), followed byelution of this extract through the original (Fisher Johns; melting point, 104 "c). pentane-washed silicic acid column. Internal standards were added to The hydration of a-pinene (orP-pinene) to therearranged products zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAborneol and fenchol was accomplished by the technique described by the eluent (-3 mg each of 1,8-cineole, (*)-linalool, (It)-a-terpineol, (*)-borneol, geraniol, and nerol) and the volume was adjusted to 3.5 Williams and Whittaker (27) (Fig. 4). Two mmol of[3H]a-pinene (1.55 d,as for the hydrocarbon assay. The tritium contentwas determined zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAmCi/mol) or [3H]P-pinene (3.56 mCi/mol) were dissolved in acetone- on a 1-daliquot, and the remainder was concentrated as before for water (IM, v/v) containing 0.075 N H2S04 and heated at 75 OC in a radio GLC and TLC (Silica Gel G containing 8%AgN03, solvent sealed tube for 12 h. After cooling and neutralization with NaHC03, system E). TLC-purified products to be analyzed further were eluted the reaction mixture was diluted with pentane and thepentane layer from the silica gel with ether. was repeatedly washed with brine. The pentane extract was concen- Examination of the incubation mixture for phosphorylated prod- trated and subjectedto TLC (Silica Gel G, solvent system E) to afford ucts was carried out by adjusting the previously extracted aqueous [3H]endo-fenchol (RF= 0.51; yield, 8-10%) and [3H]borneol (RF= phase to pH 5.0 with 0.1 M sodium acetate buffer,followed by 0.41; yield, 9-12%). Two-phase Cr03 oxidation (28) of the alcohols incubation with 2 units of wheatzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA germ acid phosphatase (Sigma) for provided the respective ketones fenchone and camphor in nearly 1 h at 28 "C. Alcohols thus liberated were extracted with ether (2 X quantitative yield. The ketones were purified by TLC as above and 1.5 ml) and the extract was dried by passage through a column (0.5 their purities were verified to be greater than 95% by GLC. The X 3 cm) of anhydrous Na2S04. Nerol, geraniol, and (&)-linalool (-3 stereospecificity of the rearrangements was examined in separate mg each) were added, and theextract was brought to a volume of 3.5 reactions in which the optical purities of reactants and products were ml with ether. An aliquot was taken for tritium determination, and determined. It was thus demonstrated that the conversion to endo- the remaining sample was concentrated for radio GLC and TLC fenchol proceeds without racemization, while the conversion to bor- analysis as described above for oxygenated monoterpene products. neol occurs with up to 20% racemization [(+)-a-pinene, [a]? + 46.5' Appropriate boiled controls were included in each experiment, and (neat) (+52.4', Ref. 29), affords (+)-fenchone, [a];' + 38.7' (12.4 in in all cases nonenzymic formation of the terpenes of interest was hexane) and (+)-camphor, [a]? + 45.0" (16.7 in hexane), while (-)- Downloaded from negligible. Tween 20 controls, employed in isotopic dilution experi- P-pinene, [a]? - 21.0' (neat) (-22.7', Ref. 29), affords (-)-fenchone ments, showed no effect of this detergenton pinene biosynthesis from [a]? - 39.6" (6.4 in hexane) and (-)-camphor [a]:' - 50.5" (6.7 in GPP. The recovery of labeled products by the above assay procedures hexane)] (Fig. 4). The optical rotations of the fenchone and camphor was routinely 70 f 5%, as determined with authentic [l-3H]limonene thus obtained were compared to reference standards of known purity as standard. Proteinwas routinely estimated by the dye binding assay and toliterature values (2). (Bio-Rad Laboratories). To determine the stereochemical composition of the biosynthetic
Purification and Characterizationof Monoterpene Olefins from 3H-labeled pinenes, the stereospecifically derived fenchones were www.jbc.org Sage Oil-Young sage leaves (0.4 kg) zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAwere exhaustively extractedzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA derivatized tothe crystalline oximes and then converted to the with pentane and an aliquot was analyzed by packed column and diastereomeric (-)-menthyloxymethyl ethers (Fig. 4)which were capillary GLC. The extract was concentrated andsubjected to vacuum separated by TLC as described previously (30). distillation (75 "C at 10 mM Hg), and the distillate was re-examined Chromatography and Determination of Radioactivity-TLC was by GLC. No qualitative or quantitative differences in monoterpene performed on I-mm layers of Silica Gel G or Silica Gel G impregnated by guest, on December 7, 2011 composition were noted between the native and distilled oil. The oil with AgN03 (8% to 15%), which had been activated at 110 "C for 4 h. was diluted with pentane and passed through a column (2 X 50 cm) The solvent systems employed were: (A) hexane/ethyl acetate (73, of silicic acid (Mallinckrodt, SiicAR CC-7) to remove oxygenated v/v); (B) redistilled hexane; (C) benzene/hexane/ether (50:50:1, v/v/ terpenes. The hydrocarbon fraction was recovered, and the individual v); (D) hexane/ether/acetic acid (25255, v/v/v); and (E) hexane/ olefins of interest were isolated by repeated preparative argentation ethyl acetate (8:2, v/v). After development, chromatograms were TLC (Silica Gel G impregnated with 8% AgN03, solvent system B). sprayed with a 0.2% ethanolic solution of 2,7-dichlorofluorescein and Olefin samples for the determination of optical rotation were of viewed under UV light to locate the appropriate components. Ana- chemical purity greater than 96% as determined by GLC. For the lytical capillary GLC was performed with a Perkin-Elmer Sigma 3B examination of the influence of steam distillation on sage terpenes, chromatograph on a 25-meter fused silica column coated with Car- aliquots of the above terpene fractions, as well as the pure olefins, bowax 20M at 50 "C (for hydrocarbons). Packed column GLC was were distilled from 0.1 M HCl, 0.1 M KOH, or water, and thedistillates performed with a Packard 420 chromatograph. The columns were: were examined by GLC. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA25% Carbowax 4000 on 80/100 mesh Gas-Chrome Q, 3.4 m X 3 mm The labeled biosynthetic olefins, obtained from the hydrocarbon outer diameter stainless steel (hydrocarbons were analyzed at 80 "C, fraction of large scale enzyme preparations, were purified for product alcohols at 170 "C); and 10% SE-30 on 80/laO mesh Gas-Chrome Q, identification by TLC (Silica Gel G containing 12% AgN03, solvent 3 m X 3 mm outer diameter stainless steel. Other chromatographic system C). The products were visualized on the chromatogram by conditions are described in the text. Radio GLC was performed with fluorescence quenching of 2,7-dichlorofluoresceinand were recovered a Varian chromatograph attached to a Model 7357 Nuclear Chicago from the scraped gel by pentane extraction. radioactivity monitor calibrated externally with ["Hltoluene. Quan- Chemical Conversions-The stereospecific isomerization of P-pi- titation of data was by integration of radio and mass peaks of the nene to a-pinene was carried out by modification of the procedure of recorder output. Rudakov and Shestaeva (25) (Fig. 4). Two mmol of purified ['HIP- Radioactivity in liquid samples and thin layer fractions was deter- pinene (1.88 mCi/mol) were transferred to an ampoule with 0.2 mmol mined in a counting solution (15 ml) consisting of 0.3% (w/v) Omni- of benzoic acid and 0.1 mmol of hydroquinone (as antioxidant), and fluor (New England Nuclear) dissolved in 30% ethanol in toluene with the ampoule was sealed under N2 and heated at 120 "C for 15 h. The a Packard 3255 LS spectrometer. The counting efficiency for tritium cooled reaction mixture was taken up in pentane, washed with satu- was 37% and all assays were conducted with a standard deviation of rated NaHCO3, decolorized over activated charcoal (Aldrich decolor- less than 3%. king charcoal), and eluted through a silicic acid column (1 X 6 cm) with pentane. The [3H]a-pinene obtained in this fashion (>85% yield) RESULTS AND DISCUSSION was freed from residual starting material by preparative argentation TLC (Silica Gel G containing 8% AgN03, solvent system C). Monoterpene Composition of S. officinalis-Previous and- The oxidation of a-pinene to cis-phonic acid was performed by a yses of the monoterpenes of S. officinalis have been carried variation of the method of Sam and Simmons (26) (Fig. 4). Two mmol out on the steam-distilled oil, and have indicated significant of (f)-[3H]a-pinene (0.51 mCi/mol) were dissolved in 15 ml of dry variation in a- and P-pinene content (16) and in optical purity benzene containing 5.5 mmol of KMn04 and1 mmol of dicyclohexano- 18-crown-6. The reaction mixture was stirred vigorously with coarse of a-pinene (31). As /?-pinene is knownto isomerize to a- sand for 5 h at 20 "C and thenacidified. Mn02formed in the reaction pinene under a variety of conditions (7, 8, 25), the results of was converted to the soluble sulfate with the minimum quantity of these analyses were initially suspect. Samples of sage leaves, NaHS03, and the organic products were extracted with ether which of the age used for the following biosynthetic studies, were was back-washed with brine. Extraction of the ether layer with 0.1 therefore thoroughly extracted with pentane andthe extracts 2338 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBABiosynthesis of IsomericPinenes from Geranyl Pyrophosphate were analyzed directly by capillary and packed column GLC, zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA revealing a monoterpene olefin fraction consisting of a-pinene (27-29%), camphene (19-21%), /?-pinene (38-44%), myrcene (3-5%), limonene (4-6%),sabinene (<3%), terpinolene (
No ["Hlgeraniol or [lH]linalool was detected in the oxygen- To determine thelocation of the tritium in a-pinenederived Downloaded from ated monoterpene fraction derived from [l-:'H]NPP, and no from [l-"H]GPP, thepurified biosynthetic productwas diluted [,"H]geraniol or ['Hlnerol appeared to be formed from [l-:'H] with (f)-a-pinene and treated with acidic-aqueous acetone LPP. Additionally, examination of the aqueous products re- (27) to provide the rearranged monoterpenolborneol (Fig. 4). maining after incubation (by acid phosphatase hydrolysis to The pinane to bornane skeletal rearrangementproceeds with the corresponding terpenols (35,36))provided no evidence for varying degrees of racemization, which under the carefully
the interconversion of the three pyrophosphorylated acyclic controlled conditions was verified at approximately 20% (Le. www.jbc.org precursors. (+)-a-pinene affords 81% (+)-borneol and 19% (-)-borneol).zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Product Identification and Localization of Tritium-Be- Oxidation of borneol to camphor (28), followed by vigorous cause [ l-"H]GPP yielded a radioactive hydrocarbon fraction exchange of both exo- and endo-a-hydrogens (18,39),reduced
zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA by guest, on December 7, 2011
oxime (-1 - menthyloxymethyl ether
cis - PinonicAcid I zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA(+ or -1 KMn04I 2340 Biosynthesis ofIsomeric Pinenes fromGeranyl Pyrophosphate the specificzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA activity of the derived camphorzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA by 76%, from 0.037 - to 0.009 mCi/mol (i.e. the tritium of the racemized product, A predictably, was not exchanged). Similar treatment of [3H]a- T pinene derived from [3H]/3-pinene afforded borneol as before (racemization verified at up to 20%), and this product was oxidized to camphor and subjected to the exchange procedure. The specific activity of the recovered camphor was reduced from 0.078 to 0.019 mCi/mol(i.e. 75% loss of 3H). These findings, when corrected for racemization, demonstrate that essentially all of the tritium of the biosynthetic a-pinene and P-pinene was specifically located on the methylene bridge carbon (C-7),indicating direct derivation from [1-3H]GPP and a consistency with the labeling pattern observed in earlier in vivo studies with [2-'4C]mevalonate as precursor (9). [3H]Camphene derived from [1-3H]GPP was zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBApurified by argentation TLC and subjected to KMn04-NaI04oxidation (40). The major radioactive product obtained was chromato- graphically coincident (radio GLC and TLC) with campheni- lone, confiing the identity of the original product. Hydro- genation of limonene, obtained from incubations with [ 1-3H] NPP, yielded a single radioactive product chromatographi-zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA cally coincident with cis-p-menthane as expected (16). The identities of the other olefins were verified by radio GLC on Downloaded from several columns of widely differingselectivity (Carbowax 4000 and SE-30). zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAFIG.5. Chromatographic resolution of stereoisomers of bio- Stereochemistry of the Biosynthetic Pinenes-To establish synthetic ['Hla-pinene and ['Hlfl-pinene. The diastereomeric the stereochemical composition of the [7-3H]a-pinenederived (-):menthyloxymethyl oxime ethers derived from biosynthetic t3H] from[1-3H]GPP by the cell-free preparation, the labeled a-pmene (A) andr3H]P-pinene (B) wereseparated by thinlayer chromatography on Silica Gel G with (-)$-pinene (97% optically product was diluted with (+-)-a-pineneand treated with acidic- www.jbc.org aqueous acetone as before, and the derived (&)-endo-fenchol pure) as developing solvent. The bar graph represents tritium con- tained in 2-mm sections of gel. The standards indicatedare the (-)- was isolated. This product was oxidized to fenchone and the menthyloxymethyl ethers derived, respectively, from (-)-pinene and crystalline oxime was prepared. Treatment of this material (+)-pinene. Or is the origin, and the right edge of the figure defines with (-)-menthyl chloromethyl ether afforded the diastereo- the solvent front. by guest, on December 7, 2011 meric 0-(-)-menthyloxymethyl ethers of (+)-fenchone oxime (Fig. 4) in 90% yield (30). The stereospecificity of the trans- than [1-3H]GPP, had no effect on the rate of a-pinene for- formation of a-pinene to endo-fenchol as well as theretention mation. From all the evidence, it appears safe to conclude that of absolute configuration of the derived fenchane nucleus no significant P-pinene to a-pinene isomerization occurs under throughout subsequent derivatization has beenpreviously operational conditions of the enzyme assay. documented (30) and was independently verified here with Search for Free Intermediates-To investigate the previ- (+)-a-pinene of known optical purity. TLC with (-)-&pinene ously suggested intermediacy of myrcene and cis-ocimene in as an asymmetric developing solvent (30) provided unambig- the formation of the pinenes (5) (Fig. 2, Scheme II),individual uous separation of the diastereomers (Fig. 5A) and clearly isotopic dilution experiments were performed in the presence demonstrated the enzymic formation of both (+)-[7-3H]a-pi- of a 100-fold excess (relative to [1-3H]GPP) of unlabeled nene and (-)-[7-3H]a-pinene in nearly 2.5:l ratio. myrcene, cis-ocimene, and trans-ocimene, as well as (+-)-cam- The purified [7-3H]a-pinene derived from [7-3H]P-pinene phene and (*)-limonene. Under no circumstances was the was diluted with unlabeled material and converted to endo- conversion of [1-3H]GPP into [7-3H]a-pinene or [7-3H]/3-pi- fenchol as before. Subsequent steps (Fig. 4) followed by TLC nene altered relative to the controls. Furthermore, no radio- separation of the diastereomers fiyestablished that essen- activity above the normal control levelswas found to be tially all of the radioactivity (>95%) of the mixture was associated with the putative intermediates. Similar isotopic associated with the diastereomer derived from (-)$-pinene dilution experiments in the presence of 100-fold excessof the (Fig. 5B). These results thus demonstrate that the cell-free unlabeled monoterpenols (2)-a-terpineol, (f)-cis-pinanol, and synthesis of a-pinene and P-pinene from GPP reflects not only (&)-trans-pinanol, which could conceivably be regarded as the chemical composition, but also the stereochemical com- possible intermediates, did result in decreased pinene forma- position, of the pinenes synthesized in vivo. tion (of up to 45%). However,the relatively high levels(1 mM) Absence of (-)-P-Pinene to (-)-a-Pinene Isomerization- of these alcohols required to reduce product formation, as well The possible in vitro isomerization of (-)-P-pinene was inves- as the inability of these presumptive intermediates to trap tigated by analyzing the products obtained on incubation of detectable radioactivity, suggest that neither ocimenes nor the complete enzyme system with 0.75p~ (-)-[7-3H]P-pinene myrcene, nor any of the other monoterpenes tested, are free (biosynthetically prepared at 300 Ci/mol and assayed without intermediates in the biosynthesis of pinenes from GPP. pentane overlay). No r3H]a-pinenewas observed on radio Properties of the Enzyme System-A preliminary investi- GLC analysis of the isolated olefinic products by a procedure gation of thezyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA properties of the soluble enzyme system was which wascapable of detecting <1%conversion. Additionally, undertaken to assess optimal operating conditions, and to incubation of the enzyme preparation with [ 1-3H]GPPfor 6 h, investigate the likelihood of multiple pinene biosynthetic en- with sequential product analysis (radio GLC) at 30-min inter-zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA zymes. Under the conditions of the assay, the rate of conver- vals, showed that the ratio of a-pinene to P-pinene produced sion of GPP to a-pinene and to P-pinene was constant for up remained constant throughout. Finally, (-)#-pinene, when to 1 h at the 0.5 mg/ml protein level, and all subsequent included in the incubation mixture at levels 100-fold higher zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAmeasurements of enzyme activity were carried out within this Biosynthesis of Isomeric Pinenes from Geranyl Pyrophosphate 2341 TEMPERATURE,('C) zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA linear range. Dialysis of the preparation zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA against the assay zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA buffer, to remove divalent cations present in the extraction 40 45 50 55 medium, reduced a- and P-pinene cyclase activity to negligible I I 1 I 1 levels. However, activity could be fullyrestored by re-addition of 15 m~ MgC12, thus demonstrating the absolute requirement for a divalentzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA cation as with other monoterpene cyclases (12) (the C1- anion was without effect). MnClz (1 mM) was about 85% as effective as MgC12 (15 mM) in restoring P-pinene cyclase activity and roughly 50% as effective as MgC12 in restoring &-pinene cyclase activity under thesesaturating conditions. The observed preference for Mg2' is in contrast to the cation requirement for a- and P-pinene synthesis by C. limonum extracts, in which Mn2+zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAis reportedly preferred (14). In 50 mM Tricine-25 mM sodium phosphate buffer, /?-pinene synthetase activity exhibited a distinct pH optimum near 7.2 (Fig. 6). By contrast, the pH curve for a-pinene synthetase activity was bimodal with optima near 6.0 and 7.2 (Fig. 6), suggesting the presence of two a-pinene cyclase activities. p- Hydroxymercuribenzoate severely inhibited the conversion of GPP to both pinenes, but @pinene cyclase activity was far [INHIBITOR], (pM) more sensitive to thethiol-directed reagent than was a-pinene cyclase (Fig. 7). All monoterpene cyclases thus far examined FIG. 7. Effects ofphydroxymercuribenate (-) and heat are inhibited by thiol-directedzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA reagents (12). Thermal inacti- treatment (- - -) on the cyclization of GPP to a-pinene (0 and Downloaded from vation studies, at temperatures ranging from 45 to 56 "C, A) and to 8-pinene (0and A). For experiments with p-hydroxy- revealed that P-pinene cyclase activity was significantlymore mercuribenzoate, 500 pg of protein were preincubated with the indi- cated concentration of inhibitor for 15 min at 0 "C in 10 mM MES-5 labile than a-pinene cyclase activity (Fig. 7), providing a mM sodium phosphate buffer, pH 6.5, containing 10% glyceroland 15 further distinction between the enzymes involved in the for- m~ MgCl,, followed by the addition of 10 p [1-3H]GPP and incu- mation of these positional isomers. bation for 1h at 28 "C.In thermal inactivation experiments, 500 pg of protein were preincubated at theindicated temperature for 3 min in The enzyme systen~described here is like other monoter- www.jbc.org pene cyclases in general properties and in the ability to utilize 10 m~ MES-5 m~ sodium phosphate buffer, pH 6.5, containing 10% the acyclic precursors GPP,NPP, and LPP with varying glycerol, 15 m~ MgCl,, 1 m~ each of dithioerythritol, sodium ascor- bate, and Naz.9205, followed by the addition of 10 CM [1-3H]GPP and degrees of efficiency (12). A notable feature of the present incubation for 1h at 28 "C. Control values were 0.5 nmolof a-pinene
zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAand 0.4 nmol of P-pinene/h/mg of protein for the inhibitor experi- by guest, on December 7, 2011 ment, and 0.4 nmol of a-pinene and 0.6 nmol of P-pinene/h/mg of protein for the thermal inactivation experiment.
work was the preliminary finding that, of the three possible acyclic precursors examined, only GPP afforded a distribution of monoterpene olefin products which paralleled that found in uiuo. NPP and LPP bothyielded, as major products, olefins which either are minor constituents or are notdetected in the olefin fraction. Until the preparation is fractionated into its constituent activities and the competing phosphatases are removed, the qualitative and quantitative differences in prod- uct formation observed with the various acyclic precursors and cations cannot readily be clarified. The tentative results do suggest, however, thatGPP is the likely precursor of pinenes and other cyclic olefins in uiuo. The fact that GPP was transformed to such cyclic products without detectable conversion to theother acyclic precursors also argues against
\ the need for isomerization to free NPP or LPP asa prelude to b cyclization (13)but does not rule out the intermediacy of the bound equivalents of these compounds. Earlier chemotaxonomic studies (10) and in vivo tracer studies (11) imply the existence of multiple pathways to the pinenes in other species, and in the present case the distribu- tion of isomeric constituents would seem to necessitate the involvement of at least two pinene cyclases. The preliminary PH characterization of the relatively crude system described here, FIG. 6. Effect of pH on the cyclization of GPP to a-pinene which appears to be the fiist report on the cell-free biosyn- (0) and &pinene (0)by the cell-free extract. Each reaction zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAthesis of enantiomeric terpenes, does strongly suggest the mixture, containing 300zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA pg of protein, 10 p~ [1-3H]GPP, 15 mM MgCI,, presence of multiple pinene cyclases. Detailed studies on the 1 mM each of dithioerythritol, sodium ascorbate, and Na2Sz05, in a number and nature of the GPP:pinene cyclases of sage are total volume of 1ml of the appropriate buffer, containing 10% glycerol, now underway. was incubated at 28 "C for 1 h. The buffers were 50 mM Tricine-25 mM sodium phosphate (- - -) and 50 mM MES-50 mM sodium phos- Acknowledgments-We thank Richard Hamlin for raising the phate (-). Rates indicated are in nanomoles of olefin/h/mg of plants, Jerry Winters for technical assistance, and Dr. R. C. Ronald protein. for his helpful discussions. 2342 Biosynthesis of IsomericPinenes fromGeranyl Pyrophosphate
zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAREFERENCES 71,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 775 1. Banthorpe, D. V., and Whittaker, D. (1966) Chem. Reu. 66,643- 22. Cornforth, R. H., and Popjak, G. (1969) Methods Enzymol. 15, zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA656 359-390 2. Guenther, E. (1975) The Essential Oils,Vol. 2 (reprinted),pp. 55, 23. Loomis, W. D. (1974) Methods Enzymol. 31,528-544 420, and 433, Kreiger, Huntington, NY 24. Loomis, W. D., Lile, J. D., Sandstrom, R. P., and Burbott, A. J. (1979) Phytochemistry 18, 1049-1054 3. Plouvier, V. (1966) Phytochemistry 5, 955-967 4. Ruzicka, L., Eschenmoser, A,, and Heusser, H. (1953) Enperientia 25. Rudakov, G. A., and Shestaeva, M. M. (1955) Zh. Obshch. Khim. 9,357-367 25, 2635-2639 5. Gleizes, M. (1976) C. R.Acad. Sci. (D)Paris 283, 97-100 26. Sam, D. J., and Simmons, H. E. (1972) J. Am. Chem. SOC.94, 6. Burwell, R. L., Jr. (1951) J. Am. Chem.Soc. 73,4461-4462 4024-4025 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA27. Williams, C. M., and Whittaker, D. (1971) J. Chem. Soc. B 668- 7. Rudakov, G. A., and Shestaeva, M. M. (1955) Zh. Obshch. Khim. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA672 25, 597 28. Brown, C., and Garg, C. P. (1961) Am. Chem.SOC. 83,2952- 8. Spanninger, P. A,, and Von Rosenberg, J. L. (1969)J. Org. Chem. H. J. 34, 3658-3659 2953 29. Comyns, A. E., and Lucas, H. J. (1957) J. Am. Chem. SOC.79, 9. Banthorpe, D. V., and Le Patourel, G. N. J. (1972) Biochem. J. 4339-4341 130, 1055-1061 30. Croteau, R., Felton, M., and Ronald, R. C. (1980) Arch. Biochem. 10. Zavarin, E. (1970) Phytochemistry 9, 1049-1063 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBABiophys. 200,524-533 11. Valenzuela, P., Cori, O., and Yudelevich, A. (1966)Phytochemistry 31. Guenther, (1974) The Essential Oils, Vol. 3 (reprinted),p. 716, 5, 1005-1011 E. Krieger, Huntington, NY 12. Croteau,R. (1981)in Biosynthesis of IsoprenoidCompounds (Porter, J. W., and Spurgeon, S. L., eds) Vol. 1, pp. 225-282, 32. Valenzuela, P., Beytia, E., Cori, O., and Yudelevich, A. (1966) zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAArch. Biochem. Biophys.113,536-539 Wiley, New York 13. Charlwood, B. V., and Banthorpe, D. V. (1978) Prog. Phytochem. 33. Loomis, W. D. (1967) in Terpenoids in Plants (Pridham, J. B., 5,65-125 ed) pp. 59-82, Academic Press, New York 14. Chayet, L., Rojas, C., Cardemil, E., Jabalquinto, A. M., Vicufia, 34. Attaway, J. A., Pieringer, A. P., and Barabas, L. J. (1966) Phyto-
R., and Cori, 0. (1977) Arch. Biochem. Biophys.180, 318-327 chemistry 5, 141-151 Downloaded from 15. Cori, 0. (1969) Arch. Biochem. Biophys. 135,416-418 35. Croteau, R., and Karp, F. (1979) Arch. Biochem. Biophys. 198, 16. Croteau, E., and Karp, F. (1976) Arch. Biochem. Biophys. 176, 512-522 734-746 36. Croteau, R., Felton, M., and Ronald, R. C. (1980) Arch. Biochem. 17. Croteau, H., and Felton, M. (1981) Arch. Biochem. Biophys. 207, Biophys. 200, 534-546 460-464 37. Hedrick, G. W., and Lawrence, R. V. (1960) Ind. Eng. Chem. 52, 18. Croteau, R., and Karp, F. (1977) Arch. Biochem. Biophys. 184, 853-856
77-86 38. Zweifel, G., and Brown, H. C. (1964) J. Am. Chem. SOC.86, 393- www.jbc.org 19. Croteau, R., and Karp, F. (1977) Arch. Biochem. Biophys. 179, 397 257-265 39. Thomas, A. F., Schneider, R. A,, and Meinwaid, J. (1967) J. Am. 20. Baumann, W. J., and Mangold, H. K. (1964) J. Org. Chem. 29, Chem. Soc. 89,68-70 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 3055-3057 40. Lemieux, R. U., and von Rudloff, E. (1955) Can. J. Chem. 33, 21. Cramer, F., andzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Bohm, W. (1959) Angew. Chem. Int. Ed. Engl. 1701-1709 by guest, on December 7, 2011