OF BIOLOGICAL Vol. 268, of 5, pp. by for and U.S.A. 48 ... · 48-Limonene Synthase from the Oil Glands of Spearmint (Mentha spicata) cDNA ISOLATION, CHARACTERIZATION, AND BACTERIAL

THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1993 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 268, No. 31, Issue of November 5, pp. 23016-23024,1993 Printed in U.S.A.

48-Limonene Synthase from the Oil Glands of Spearmint (Mentha spicata) cDNA ISOLATION, CHARACTERIZATION, AND BACTERIAL EXPRESSION OF THE CATALYTICALLY ACTIVE MONOTERPENE CYCLASE*

(Received for publication, April 2, 1993, and in revised form, July 12, 1993)

Sheila M. ColbyS, William R. AlonsoJ, Eva J. Katahira, Douglas J. McGarvey, and Rodney Croteaull From the Institute of Biological Chemistry, Washington State University, Pullman, Washington 99164-6340

The committed step in the biosynthesis of monoterpenes in mint (Mentha) species is the cyclization of geranyl pyrophosphate to the olefin (-)-4S-limonene catalyzed by limonene synthase (cyclase). Internal amino acid sequences of the purified enzyme from spearmint oil glands were utilized to design three dis- tinct oligonucleotide probes. These probes were subsequently employed to screen a spearmint leaf cDNA library, and four clones were isolated. Three of these cDNA isolates were full-length and were functionally expressed in Escherichia coli, yielding a peptide that is immunologically recognized by polyclonal antibodies raised against the purified limonene synthase from spearmint and that is catalytically active in generating from geranyl pyrophosphate a product distribution identical to that of the native enzyme (principally limonene with small amounts of the coproducts a- and 8- pinene and myrcene). The longest open reading frame is 1800 nucleotides and the deduced amino acid sequence contains a putative plastidial transit peptide of approximately 90 amino acids and a mature protein of about 610 residues corresponding to the native enzyme. Several nucleotide differences in the B‘-untranslated region of all three full-length clones suggest the presence of several limonene synthase genes and/or alleles in the allotetraploid spearmint genome. Se- quence comparisons with a sesquiterpene cyclase, epi- aristolochene synthase from tobacco, and a diterpene cyclase, casbene synthase from castor bean, demonstrated a significant degree of similarity between these three terpenoid cyclase types, the first three examples of this large family of catalysts to be described from higher plants.

Several hundred naturally occurring monoterpenes are

* This investigation was supported in part by National Institutes of Health Grant GM-31354, the Mint Industry Research Council, the McKnight Foundation, and Project 0268 from the Agricultural Re- search Center, Washington State University. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” In accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide sequence(s) reported in thispaper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) LI3459.

$ McKnight Foundation Postdoctoral Fellow. Present address: USDA Forest Service, Pacific Southwest Research Station, P. 0. BOX 245, Berkeley, CA 94710.

Present address: Miles Inc., P. 0. Box 507, Clayton, NC 27520. ll To whom correspondence should be addressed Inst. of Biological

Chemistry, Washington State University, Pullman, WA 99164-6340. Tel.: 509-335-1790; Fax: 509-335-7643.

known, and essentially all are biosynthesized from geranyl pyrophosphate, the ubiquitous Clo intermediate of the isoprenoid pathway (1,Z). Monoterpene synthases, often referred to as “cyclases,” catalyze the reactions by which geranyl pyrophosphate is cyclized to the various monoterpene carbon skeletons. These enzymes have received considerable recent attention because the cyclization process determines the basic character of the monoterpene end products and because the cyclization mechanism is quite complex, involving multiple steps in which many of the carbon atoms of the substrate undergo alterations in bonding, hybridization, and configu- ration (1,2). Research on monoterpene cyclases has also been stimulated by the possible regulatory importance of these enzymes that function at a branch point in isoprenoid metabolism (3), as well as by the commercial significance of the essential oils (4) and aromatic resins (5) and the ecological roles of these terpenoid secretions, especially in plant defense (6, 7).

One of the major classes of plant monoterpenes is the monocyclic p-menthane (1-methyl-4-isopropylcyclohexane) type, found in abundance in members of the mint (Mentha) family. The biosynthesis of p-menthane monoterpenes in Mentha species, including the characteristic components of the essential oil of peppermint (ie. (-)-menthol) and the essential oil of spearmint (ie. (-)-carvone), proceeds from geranyl pyrophosphate via the cyclic olefin (-)-limonene (8) (Fig. 1). The transformation of geranyl pyrophosphate to limonene is seemingly the least complicated terpenoid cyclization (9) in having ample precedent in solvolytic model studies (10-13), and the responsible enzyme has become a prototype for the terpenoid cyclization reaction (14-18). The enzyme that produces the (-)-4S-enantiomer (geranyl pyrophosphate:(-)-4S-limonene cyclase or, simply, (-1-4S-limonene synthase) has been purified from peppermint (Mentha X piperita) and spearmint (Mentha spicata) oil glands (191, and highly specific antibodies directed against this enzyme have been prepared (20). In properties and mechanism of action (18, 21), (-)-4S-limonene synthase is typical of the cyclase class of enzymes (1, 2, 22, 231, and it seemingly catalyzes a slow, possibly rate-limiting, step of monoterpene biosynthesis in Mentha (3).

A detailed understanding of the control of monoterpene biosynthesis and of the cyclase reaction mechanism requires the relevant cDNA clones as tGOlS for evaluating patterns of developmental and environmental regulation and for examining active site structure-function relationships. There have been no previous reports of the cloning of a monoterpene cyclase, although the molecular cloning of a sesquiterpene cyclase of plant origin (epi-aristolochene synthase from tobacco) has been reported recently (24), and the cloning of a

23016

Limonene Synthase Cloning and Expression 23017

A (-)-Menthol

FIG. 1. Outline of the principal pathways of monoterpene biosynthesis in spearmint leading to carvone and in peppermint leading to menthol. After geranyl pyrophosphate is cyclized to limonene, a series of secondary redox transformations convert this olefinic intermediate to other monoterpenes.

diterpene cyclase (casbene synthase from castor bean) has also been accomplished.' In this paper, the isolation and functional expression of a (-)-4S-limonene synthase cDNA from spearmint is described. This first cloning of a monoterpene cyclase was made possible by recently developed methods for isolating in high yield the oil gland secretory cells in which monoterpene biosynthesis occurs (25, 26) and for efficiently purifying the enzyme from this enriched source (23, 27). Comparison between the limonene synthase and other cyclases demonstrated a significant degree of sequence similarity among three different classes of terpenoid cyclases of plant origin.

EXPERIMENTAL PROCEDURES

Plant Materials, Substrates, and Reagents-Spearmint (Mentha spicata L.) plants were propagated and grown as described previously (19). Apical buds and newly emerged rapidly expanding leaves (5-10 mm long) of vegetative stems (3-7 weeks old) were used for nucleic acid isolation and for the preparation of glandular trichome cells for enzyme extraction. [l-3H]Geranyl pyrophosphate (120 Ci/mol) was prepared as described previously (19, 28).

Limonene Synthase Isolation and Assay-The isolation of 4S- limonene synthase by sonic disruption of spearmint (M. spicata) oil gland secretory cells and the purification to homogeneity of this monoterpene cyclase have been described in detail elsewhere (19). In the present instance, minor changes in the protocol were made such that a DEAE-Sepharose Fast Flow anion-exchange column (10 X 100 mm; Pharmacia LKB Biotechnology Inc. fast protein liquid chromatography System) replaced the Mono Q HR 5/5 column, and the dye- ligand chromatography step on Matrex Gel Red A was carried out after (rather than before) the ion-exchange step, to permit the processing of substantially larger amounts of protein (from 50-100-g tissue). The final purification step was carried out by hydrophobic interaction chromatography as before (19), and the purity of the protein was judged to be -99% by SDS-PAGE' (29) with silver staining (30).

The typical assay mixture (1 rnl) consisted of 1-50 pg of native or recombinant protein in 20 mM Mopso buffer (pH 7.0) containing 10% (v/v) glycerol, 15 mM MgC12, 1.0 mM dithiothreitol, and 10 pM [l-3H] geranyl pyrophosphate. All samples were changed to these buffer conditions by dilution, dialysis, or desalting (Econo-Pac lODG desalting column, Bio-Rad) prior to assay for limonene synthase activity by solvent extraction and chromatographic isolation of the product as described previously (19).

' C. J. D. Mau and C. A. West, personal communication. ' The abbreviations used are: PAGE, polyacrylamide gel electro-

phoresis; Mopso, 3-(N-morpholino)-2-hydroxypropanesulfonic acid GLC, gas-liquid chromatography; IPTG, isopropyl-1-thio-P-D-galac- topyranoside.

Amino Acid Analysis and Protein Sequencing-Approximately 25 pg of purified limonene synthase was subjected to acid hydrolysis and subsequent amino acid analysis at the Washington State University Bioanalytical Center. In preparation for proteolysis and peptide sequencing, approximately 75 pg of purified protein was subjected to SDS-PAGE according to the method of Schagger and von Jagow (31) in 10% polyacrylamide vertical slab gels (16 cm X 18 cm X 0.7 mm). Gels were stained with Coomassie Brilliant Blue R-250 (in metha- no1:acetic acidwater (3010:60, v/v/v)) and destained (in metha- nokacetic acidwater (10:10:80, v/v/v)), and the gel bands containing the limonene synthase (at 56 kDa) were excised, washed with water, incubated in SDS sample buffer (29) for 30 min at 30 "C, and then inserted into the sample wells of a 16.5% polyacrylamide vertical slab gel (16 cm X 18 cm X 1.0 mm) for SDS-PAGE as before (31). Each gel slice was overlaid with buffer consisting of 0.125 M Tris (pH 6.8), 1 mM EDTA, 2.5 mM dithiothreitol, and 0.1% (v/v) SDS and containing 2 pg of V8 protease (Sigma) according to the standard practice (32). Samples were electrophoresed until the bromphenol blue dye had traveled about 1 cm to concentrate the protein into the stacking gel, and the power was turned off for 50 min to permit proteolytic digestion. Electrophoresis was then continued for 20 h at a constant voltage of 90 V.

For sequence analysis, polyvinylidene difluoride membranes (Im- mobilon-PSQ, Millipore) were first wetted in methanol and equili- brated in 25 mM Tris, 190 mM glycine (pH 8.3) containing 20% (v/v) methanol and then gel blotted in the same buffer (33). Electroblotting was carried out for 90 min at a constant current of 100 mA using a Hoefer TE 70 semi-dry transfer apparatus. Peptides were sequenced via Edman degradation on an Applied Biosystems 470 sequenator a t the Washington State University Laboratory for Bioanalysis and Biotechnology.

For CNBr cleavage, approximately 50 pg of purified limonene synthase was lyophilized in a 1.5-ml microcentrifuge tube and dissolved in 200 @1 of a degassed solution containing 10 @g/pl CNBr in 70% aqueous formic acid. After thorough mixing and flushing with argon, the mixture was incubated in the dark for 20 h at 20 "C. The sample was dried and excess CNBr and formic acid removed under vacuum (Savant Speed Vac), and the wall of the tube was washed down with 200 pl of 0.001 N NH,OH and the sample redried under vacuum. The sample was then resuspended in SDS buffer (29) and subjected to SDS-PAGE as before (31) in 16.5% polyacrylamide vertical slab gels (16 cm X 18 cm X 1.0 mm), electroblotted to polyvinylidene difluoride (Immobilon-PSQ), and the peptides subjected to Edman degradative sequencing as described above for peptides generated by V8 proteolysis. Alternatively, the CNBr-generated peptides were dissolved in 10% aqueous acetonitrile with 0.1% (v/v) trifluoroacetic acid (starting solvent) and separated by reversed phase high performance liquid chromatography (Rainin, C4-Dynamax 300A column) with a linear gradient from starting solvent to 80% aqueous acetonitrile with 0.07% trifluoroacetic acid. Purified peptides in the lyophilized chromatographic fractions were sequenced as above.

Isolation of Nucleic Acids-Newly emerging spearmint leaves with a very high density of developing epidermal oil glands were used as starting material. Total RNA (yield -1 mg/g fresh tissue weight) was extracted using the procedures of Cathala and associates (34), with the substitution of 300 mM for 50 mM Tris in the tissue homogeni- zation buffer. Poly(A)+ RNA was purified by chromatography on an oligo(dT)-cellulose column (Pharmacia). Plasmid DNA was amplified in Escherichia coli strain XL1-Blue and was prepared by standard procedures (35).

Library Construction and Screening-A spearmint leaf cDNA library was constructed from 5 pg of poly(A)+ mRNA using the ZAP- cDNA synthesis kit with X UniZAP XR vector and was packaged using the Gigapack I1 Packaging Extract according to the manufac- turer's instructions (Stratagene). Using the sequence information obtained from peptides generated by CNBr cleavage and V8 proteolysis of limonene synthase, three degenerate oligonucleotide probes were synthesized (at the Washington State University Laboratory for Bioanalysis and Biotechnology) (see Fig. 2) and end-labeled with 32P using T4 kinase (Life Technologies, Inc.) by standard procedures (36). The three probes were used to screen replicate filter lifts of 2.5 X lo6 primary plaques grown in E. coli PLK-F' using the Stratagene protocols. The hybridization conditions were modified from the Du Pont-New England Nuclear procedure using 6 X SSC (1 X SSC = 0.15 M NaCl in 0.015 M sodium citrate (pH 7.0)) containing 1% SDS,

DNA at 47 "C for 20 h. Blots were washed twice in 6 X SSC for 5 10% dextran sulfate, and 50 pg/ml each of salmon sperm and E. coli

min at 47 "c, thrice in 3 X SSC for 30 min, dried in a vacuum oven

23018 Limonene Synthase Cloning and Expression and, as in all instances described here where 32P-labeled probes were used, autoradiograms were prepared with Kodak XAR-5 film exposed overnight at -80 "C. Plaques affording positive signals with each of the three probes (a total of 20 were picked) were rescreened on E. coli XL1-Blue through four cycles using each probe individually until clones pLC 4.1, pLC 5.2, pLC 8.5, and pLC 10.1, that hybridized to all three probes, were pure. The selected X UniZap XR clones were in uioo excised and recircularized as Bluescript I1 SK(+) phagemids that were then employed to infect E. coli XL1-Blue using procedures provided by Stratagene.

Expression of Limonene Synthase cDNA-E. coli XL1-Blue cells containing pBluescript I1 SK(+), pLC 4.1, pLC 5.2, pLC 8.5, or pLC 10.1 were grown to stationary phase in 5 ml of Luria-Bertani medium with 100 pg/ml ampicillin and used to inoculate 100-ml liquid cultures either without or with 50 pg/ml antibiotic which were incubated at 37 "C with shaking at 350 rpm. When the cultures reached As,,,, = 0.5, either 1 or 10 mM IPTG was added to induce the cultures which were incubated for an additional 2 h. Cells were collected by centrifugation at 2400 X g for 10 min, resuspended in 20 mM Tris (pH 7.5), and centrifuged again, then suspended in 2.5 ml of buffer containing 20 mM Mopso (pH 7.0), 1 mM EDTA, 1 mM dithiothreitol, and 10% (v/ v) glycerol and disrupted by sonication (Braun-Sonic 2000 with microprobe at maximum power for four 1-min bursts a t 0-4 "C). The bacterial homogenates were then clarified by centrifugation at 13,000 x g (the pellets contained negligible limonene cyclase activity), the resulting supernatants dialyzed to standard assay conditions, and the assay for limonene cyclase activity carried out under linear conditions as described above (19). In some instances, portions of these preparations were taken for immunoblotting (described below) or were partially purified by anion-exchange chromatography. In the latter case, the supernatant was desalted into 15 mM potassium phosphate buffer (pH 6.0) containing 10% (v/v) glycerol, 150 mM KCl, and 1 mM dithiothreitol and then applied to a DEAE-Sepharose Fast Flow column (5 X 100 mm, Pharmacia fast protein liquid chromatography system) that was eluted with a linear gradient from 150 to 600 mM KC1 in loading buffer. This limonene cyclase activity expressed in E. coli eluted as a single peak between 365 and 465 mM KCl, as with the native enzyme from Mentha.

To examine the possible production of limonene in transformed E. coli, cultures harboring pLC 5.2 (showing the highest limonene synthase specific activity by in vitro assay) were grown as before to various densities corresponding to Am = 0.3-0.8. IPTG (10 mM) was added, and the induced cultures were incubated for an additional 1- 2 h to provide enzyme activity levels ranging from 10 to 50 nmol/h/ culture; in some instances, a pentane overlay (10 ml/culture) was added to trap the volatile olefinic product. Following incubation, the cultures were chilled in ice, saturated with NaC1, homogenized, and extracted with pentane (3 X 10 ml), and the combined pentane extract was passed over a silica gel column to provide the hydrocarbon fraction, free of oxygenated metabolites. The internal standard was added (10 nmol of p-cymene) and the solvent concentrated under a stream of Nz in preparation for analysis of the hydrocarbon fraction by capillary GLC.

Product Analysis and Other Analytical Methods-Radio-GLC was performed on a Gow-Mac 550P gas chromatograph attached to a Packard 894 gas proportional counter (37) under conditions designed to separate limonene from all other naturally occurring monoterpene olefins (21).

Capillary GLC was utilized for the analysis of limonene production by transformed E. coli cultures (Hewlett-Packard 5890 injector, 220 "C; detector, 300 "C; split ratio, 5:l; HZ carrier at 14 p.s.i. column: 0.25 mm inner diameter X 30 m with 0.25-pm film of AT-1000 (Alltech); programmed from 35 "C (5 min) to 230 "C at 10 *C/min). Based on internal standardization with p-cymene, and the detector response, an amount of limonene equivalent to 0.5 nmol/culture can be easily determined by this method.

Procedures for immunoblotting on nitrocellulose using polyclonal antibodies generated in rabbits against spearmint limonene synthase, and detection by alkaline phosphatase-conjugated goat anti-rabbit IgG, have been described (20). Cross-reacting antibodies to E. coli proteins were removed by subjecting the antiserum to affinity chro-

pBluescript I1 SK(+) linked to CNBr-activated Sepharose (Sigma) matography using a column of soluble proteins from XL-1 Blue/

following established protocols (38). In some instances, horseradish peroxidase-conjugated goat anti-rabbit IgG (1:2000; Bio-Rad) was substituted as the secondary antibody, and positives were detected by enhanced chemiluminescence on Kodak XAR-5 film after treating the filters with the Amersham Corp. enhancing reagents. For quan-

titation, blots were scanned at 633 nm with an LKB 2202 Ultroscan laser densitometer. In most instances, -100 pg of E. coli protein was employed and was dialyzed against 20 mM Tris (pH 6.8) containing 1 mM dithiothreitol and 10% (v/v) glycerol and thermally denatured in the presence of SDS and 0-mercaptoethanol (29) before use. Protein concentrations were estimated by dye binding using the Bio- Rad Protein Assay, and Rainbow molecular weight markers for use in blotting were obtained from Amersham Corp.

For RNA blot analysis, spearmint and peppermint poly(A)+ RNA (3 pg each) were separated on a 1.5% formaldehyde-agarose gel and blotted onto GeneScreen Plus nylon membranes (Du Pont-New Eng- land Nuclear). Probe DNA was prepared by digesting pLC 5.2 with SmaI and ApaI, separated by agarose gel electrophoresis, and elec- troeluted using standard procedures (36). The cDNA insert was then labeled with 3zP via the hexamer reaction (34). Hybridization according to the Du Pont-New England Nuclear protocol was for 16 h at 42 "C in 15 ml of hybridization solution consisting of 5 X SSPE (1 X SSPE = 150 mM NaCI, 10 mM sodium phosphate, and 1 mM EDTA), 50% deionized formamide, 5 X Denhardts, 1% SDS, and 100 pg/ml denatured sheared salmon sperm DNA. Blots were washed twice for 5 min with SSPE at room temperature, thrice at 65 "C for 45 min with 2 X SSPE containing 2% SDS, and finally thrice for 15 min with 0.1 X SSPE at room temperature.

Sequencing double-stranded phagemid DNA of pLC 5.2 and the other clones was by the chain termination method using Sequenase Version 2.0 (United States Biochemical Corp.). Both strands of the pLC 5.2 cDNA were completely sequenced using T7 and T3 primers with additional primers synthesized as needed. DNA fragments were assembled and the sequence analyzed using SEQAID 11. version 3.81 (a public domain program provided by the authors D. D. Rhodes and D. J. Roufa, Kansas State University) and the Genetics Computer Group Packet (39). Searches were done at the National Center of Biotechnology Information using the BLAST network service (40) against the SwissProt, Protein Information Resource, and GenPept data bases.

RESULTS AND DISCUSSION

isolation and Characterization of Limonene Cyclase cDNA- (-)-4S-Limonene synthase, the principle monoterpene cyclase of spearmint and peppermint, is located exclusively in the glandular trichome secretory cells (41, 42) and catalyzes the first step of monoterpene biosynthesis in these essential oil species (8, 43). Newly developed methods for selectively isolating secretory cell clusters from these epidermal oil glands (25, 26) and for purifying this monomeric cyclase ( A I r

-56,000) from these structures (23, 27) were employed to obtain sufficient amounts of protein from spearmint for polyclonal antibody production (20) and amino acid sequence analysis. Since the amino terminus of the cyclase was blocked, internal fragments were generated by CNBr cleavage and V8 proteolysis. Several of these fragments, comprising approximately 15% of the protein, were successfully sequenced and subsequently shown to span about half of the length of the protein starting from near the amino terminus (Fig. 2). From this information three nonoverlapping degenerate oligonucleotide probes (20-, 23-, and 17-mer) were prepared and used to screen a X ZAP cDNA expression library.

Four clones, designated pLC 4.1, pLC 5.2, pLc 8.5, and pLC 10.1, hybridized to all three oligonucleotides and per- sisted through four subsequent rounds of screening and purification. The isolated phagemids were in vivo-excised, circu- larized, and packaged; these phagemids were then used to transfect E. coli. The transfected cultures were induced with IPTG, and the cells from each were harvested, homogenized, and assayed for limonene cyclase activity using [l-3Hlgeran~l pyrophosphate as substrate. Preparations from three of the four transfected cultures (with pLC 4.1,5.2, and 8.5) afforded easily measurable levels of limonene cyclase activity (in the range of 10-50 nmol/culture with specific activities in the range of 0.5-1.0 nmol/h/mg of protein). Each of these preparations also contained a protein that was readily detected by

Limonene Synthase Cloning and Expression

CNBr #1 (99-119) E L E K E T D Q I R Q L E L I D D L Q R M 20 mer GAAAAAGMCAGATCAGAT 128 oligonucleotides G G G G C A

T C

23019

CNBr #2 (273-299) N ~ V V L E L A I L D L N I V Q A Q E Q E E L K E S F 2 3 mer AATATTGTICAAGCICAATTTCA 4R 01 iqonucleotides c c G

~~ ~

G C A

CNBK #3 (342-361) G K V N A L I T V I K D I Y D V Y G T L 17 mer AAGGATATTTACGATGT 4 8 oligonucleotides A C C T C

A FIG. 2. Amino acid sequences of peptides derived by CNBr cleavage and V8 proteolysis of limonene cyclase. The corresponding

degenerate oligonucleotides are also illustrated. The solid line over CNBr #Z indicates the overlapping fragment V8 #I.

0 10 20 30

Time (min.) FIG. 3. Radio-GLC separation of the monoterpene olefins

generated from [ 1 -‘H]geranyl pyrophosphate by recombinant (4s)-limonene synthase. The lower tracing is the thermal conduc- tivity detector response to authentic standards of a-pinene (peak l), P-pinene (peak 21, myrcene (peak 3 ) , and limonene (peak 4 ) . The upper tracing is the radioactivity recorded by the monitor attached to the chromatograph. The inset shows the amplified response for the minor components. Output signal was electronically integrated and gave integration values with an S.E. of less than 10% of the mean. Identical results were obtained when the expressed enzyme was isolated from cultures transfected with pLC 4.1, pLC 5.2, or pLC 8.5. Product isolation and conditions for GLC analysis are described under “Experimental Procedures.”

slot immunoblotting using polyclonal antibodies raised against the purified limonene cyclase from spearmint, and the intensities of the blots were roughly proportional to the observed activity levels (data not shown). E. coli harboring Bluescript phagemid alone (no insert) did not give rise to detectable catalytic activity nor cross-reacting protein, nor did E. coli harboring cDNA pLC 10.1.

To examine in greater detail the product mixture generated by the recombinant limonene cyclases, the proteins encoded by pLC 4.1, 5.2, and 8.5 were partially purified by anion- exchange chromatography on DEAE-Sepharose. A single olefin synthase per clone was observed on fractionation, and preparative scale incubations with each were carried out in the presence of saturating concentrations of divalent metal ion cofactor (Mg“) and [l-3H]geranyl pyrophosphate. The

pentane-soluble reaction products were isolated and passed through silica gel to provide the olefin fraction. Following the addition of authentic monoterpene carriers, this material was analyzed by radio-GLC and shown to contain primarily limonene (-94%) with smaller amounts of other olefins (Fig. 3). Integration of the amplified signal gave 1.8% a-pinene, 2.0% P-pinene, and 1.9% myrcene, which is identical to the levels of these minor coproducts generated by the limonene synthase from spearmint and peppermint (21). Thus, the recombinant proteins produced by pLC 4.5, 5.2, and 8.5 generated, with complete fidelity, the same product mixture as the native enzyme. The production of multiple products, although seemingly common among monoterpene synthases (2, 22), is an unusual enzymatic phenomenon that can be verified by iso- topically sensitive branching experiments (21, 44) but not, with certainty, by either copurification or differential inacti- vation studies (18, 19). The production of multiple products by a recombinant cyclase provides the ultimate proof of this unusual catalytic capability.

Limonene could not be detected as a product of the transformed cultures in uiuo, even when provisions were made to trap this voIatiIe product, under anaIytical conditions whereby about 5% (0.5 nmol) of the minimum production level estimated by in vitro assay could easily have been detected. This result is perhaps not too surprising as E. coli is unlikely to produce the required substrate geranyl pyrophosphate except as a transient bound intermediate of the prenyltransferase reaction en route to farnesyl pyrophosphate and higher pren- 01s (45).

Sequence Analysis-Since the pLC 5.2 cDNA gave the highest apparent specific activity of limonene cyclase expression, this isolate was completely sequenced and examined in detail (Fig. 4). The pLC 5.2 limonene cyclase cDNA is 2182 nucleotides in length and contains a complete open reading frame of 1800 nucleotides. The deduced amino acid sequence indicates the presence of a putative plastidial transit peptide of approximately 89 amino acids and a mature protein of approximately 510 residues. The locations of all four peptide sequence fragments obtained from the native protein have been identified within the open reading frame of pLC 5.2 (Fig. 41, confirming the cDNA to represent a limonene synthase gene.

The transit peptide/mature protein junction and, thus, the exact lengths of both moieties are unknown, because the amino terminus of the mature protein is blocked and has not yet been identified. The approximate size and charge of the leader sequence is appropriate for a plastidial transit peptide;

23020 Limonene Synthase Cloning and Expression

pLC5.2 AGAGAGAGAGAGGAAGGAAAGATTMTC~GCTCTCMAGTGTTMGTGTTGCMCTCAAATGGCGATTCCTAGCMCC 80 pLC4.1 gagag---a------c--------------------------------------------------------------------- pLC8.5 attaatcctagaaaaacat---a------c---------------------------------------------------------------------

~ ~ ~ K ~ & S v a ~ q ~ f i l e ~ ~ L 18

pLC5.2 TMCGACATGTCTTCMCCCTCACACTTCAMTCTTCTCC~CTGTTATCTAGCACTMCAGTAGTAGTCGGTCTCGCCTCCGTGTGTATTGCTCCTC 180 PLC4.1 """""""""""""""""""""""""""""".""""" pLC8.5 --------------------____________________--------------------.----------

I l C C q p s H ~ K S s p K C C p q L W s S S ~ ~ e C ~ v y ~ ~ S 51

pLC5.2 CTCGCMCTCACTACTGAAAGACGATCCGGAMCTACMCCCTTCTCGTTGGGATGTCAACTTCATCCMTCGCTTCTCAGTGACTATMGGAGGACAM 280 - S q c L I ~ e ~ S c ~ y n p p ~ u o v n ~ l ~ S ~ C S o y ~ ~ o ~ 84

pLC5.2 CACGTGATTAGGGCTTCTGAGCTGGTCACTTTGGTGMGATGGMCTGGAGAMGAMCGGATCAAATTCGACAACTTGAGTTGATCGATGACTTGCAGA 380 pLclo.l C""""""""" a""""""""""-t"""""""""""""""""""""""""""

- H ~ ~ ~ ~ S E L V T L V K M E L E K E T D Q I R Q L E L I D D L Q R 118

pLC5.2 GGATGGGGCTGTCCGATCATTTCCAAAATGAGTTCMAGATCTTGTCCTCTATATATCTCGACCATCACTATTACAAGMCCCTTTTCCMMGMGA 480

M G L S D H F Q N E F K E I L S S I Y L D H H Y Y K N P F P K E E 151 pLC10.1 """"""""""""-g"""""""""""""""- - pLC5.2 MGGGATCTCTACTCCACATCTCTTGCATTTAGGCTCCTCAGAGAACATGGTTTTCAAGTCGCACMGAGGTATTCGATAGTTTCMGMCGAGGAGGGT 580

R D L Y S T S L A F R L L R E H G F Q V A Q E V F D S F K N E E G 184

pLC5.2 GAGTTCAMGAAAGCCTTAGCGACGACACCAGAGGATTGTTGCAACTGTATGAAGCTTCCTTTCTGTTGACGGMGGCGMACCACGCTCGAGTCAGCGA 680 E F K E S L S D D T R G L L Q L Y E A S F L L T E G E T T L E S A R 218

pLC5.2 GGGAATTCGCCACCAAATTTTTGGAGGAAAAAGTGAACGAGGGTGGTGTTGATGGCGACCTTTTMCMGMTCGCATATTCTTTGGACATCCCTCTTCA 780 E F A T K F L E E K V N E G G V D G D L L T R I A Y S L D I P L H 251

pLC5.2 TTGGAGGATTAAAAGGCCAMTGCACCTGTGTGGATCGMTGGTATAGGMGAGGCCCGACATGMTCCAGTAGTGTTGGAGCTTGCCATACTCGACTTA W R I K R P N A P V W I E W Y R K R P D M N P V V L E L A I L D L

pLC5.2 MTATTGTTCAAGCACMTTTCAAGAAGAGCTCAAAGAATCCTTCAGGTGGTGGAGAMTACTGGGTTTGTTGAGMGCTGCCCTTCGCMGGGATAGAC N I V Q A Q F Q E E L K E S F R U W R N T G F V E K L P F A R D R L

pLC5.2 TGGTGGMTGCTACTTTTGGMTACTGGGATCATCGAGCCACGTCAGCATGCAAGTGCMGGATMTGATGGGCAMGTCMCGCTCTGATTACGGTGAT V E C Y F W N T G I I E P R P H A S A R I M M G K V N A L I T V I

pLC5.2 CGATGATATTTATGATGTCTATGGCACCTTAGMGMCTCGAACMTTCACTGACCTCATTCGMGATGGGATATAAACTCMTCGACCMCTTCCCGAT D D I Y D V Y G T L E E L E Q F T D L I R R W D I N S I D Q L P D

pLC5.2 TACATGCAACTGTGCTTTCTTGCACTCMCMCTTCGTCGATGATACATCGTACGATGTTATGMGGAGAAAGGCGTCAACGTTATACCCTACCTGCGGC Y M P L C F L A L N N F V D D T S Y D V M K E K G V N V I P Y L R Q

880 284

980 318

1080 35 1

1180 384

1280 418

pLC5.2 MTCGTGGGTTGATTTGGCGGATAAGTATATGGTAGAGGCACGGTGGTTCTACGGCGGGCACAAACCMGTTTGGAAGAGTATTTGGAGMCTCATGGCA 1380 S W V D L A D K Y M V E A R U F Y G C H K P S L E E Y L E N S U Q 451

pLC5.2 GTCGATMGTGGGCCCTGTATGTTAACGCACATATTCTTCCGAGTAACAGATTCGTTCAC~GGAGACCGTCGACAGTTTGTACAMTACCACGATTTA 1480 S I S G P C M L T H J F F R V T D S F T K E T V D S L Y K Y H D L 484

pLC5.2 GTTCGTTGGTCATCCTTCGTTCTGCGGCTTGCTGATGATTTGGGAACCTCGGTGGAAGAGGTGAGCAGAGGGGATGTGCCGAMTCACTTCAGTGCTACA 1580 V R W S S F V L R L A D D L G T S V E E V S R G D V P K S L Q C Y H 518

pLC5.2 TGAGTGACTACAATGCATCGGAGGCGGAGGCGCGGMGCACGTGMATGGCTGATAGCGGAGGTGTGGM~GATGMTGCGGAGAGGGTGTCGAAGGA 1680 S D Y N A S E A E A R K H V K U L I A E V W K K H N A E R V S K D 551

pLC5.2 TTCTCCATTCGGCAAAGATTTTATAGGATGTGCAGTTGATTTAGGAAGGATGGCGCAGTTGATGTACCATMTGGAGATGGGCACGGCACACMCACCCT 1780 S P F G K D F I G C A V D L G R M A Q L M Y H N G D G H G T Q H P 584

pLC5.2 ATTATACATC~CMATGACCAGMCCTTATTCGAGCCCTTTGCA~GAGATGATGACGAGCCATCGTTTACTTACTTAAATTCTACCAAAGTTTTTCG 1880 I I H Q Q M T R T L F E P F A 599

pLC5.2 MGGCATAGTTCGTAATTTTTCAAGCACCAATAAATAAGGAGAATCGGCTCAAACAAACGTGGCATTTGCCACCACGTGAGCACMGGGAGAGTCTGTCG 1980

pLC5.2 TCGTTTATGGATGAACTATTCMTTTTTATGCATGTAATAATTAAGTTCAAGTTCMGAGCCTTCTGCATATTTMCTATGTATTTGAATTTATCGAGTG 2080

pLC5 .2 T G A T T T T C T G T C T T T G G C A A C A T A T A T T T T T G T C A T A T G T G G C A T C T T A T T A T G A T A T C A T A C A G T G T T T A T G G A T G A T A T G A T A C T A T C ~ 2180

pLC5.2 M 2182

FIG. 4. Nucleotide and predicted amino acid sequence of spearmint limonene cyclase clone pLC 5.2 and comparison of the 5'-nucleotide sequence to those of pLC 4.1, pLC 8.6, and pLC 10.1. The start and stop codons and the Shine-Delgarno sequences are underlined. The amino acid sequence comprising the putative transit peptide is singly underlined. The sequences of CNBr fragments 1, 2, and 3 are double underlined. Differences between pLC 4.1, 8.5, and 10.1 and pLC 5.2 are indicated by lowercase letters.


TABLE I 0,

Amino acid composition of native limonene cyclme compared with that of the translated cDNA y % i m m S

The translated sequence includes only those amino acids of the V ~ X a a a a a

7 - m c y l n

; z y y y y putative mature peptide, residues 90-599. Asn/Asp and Gln/Glu could not be distinguished, and tryptophan, methionine and cysteine could not be determined (ND) from the acid hydrolyzed sample. The molecular weight of the native protein was determined by SDS-PAGE and gel permeation chromatography to be about 56,000. The x’ test showed interdependence of the two compositions with p > 0.97. 97.4 ”>

Amino acid Deduced from Acid-hydrolyzed translated cDNA native protein

Ala CYS Asp + Asn Glu + Gln Phe GIY His Ile LY s Leu Met Pro Arg Ser Thr

25 5

56 68 28 24 14 26 21 55 14 17 30 32 23

29 ND 63 14 23 36 12 24 21 58

ND 15 26 28 22

Val 32 31 Trp 13 ND TY r 21 17

Total residues 510

kDa 59.8

the first 89 amino acids are characteristically high in serine and threonine content (-25%), and the Chou-Fasman rules (46) indicate a typical @-sheet adjacent to the putative cleavage site (47-49) which established criteria (50) predict between or near residues AlaR9-Serg0. The location of the CNBr fragment 1 along the deduced amino acid sequence defines the minimum size of the mature peptide, since no fragments preceding MetgR were observed. Translation of the putative mature limonene cyclase cDNA (i.e. residues 90-599) yields a protein of 59.8 kDa, which is within 10% of the 56-kDa mass of the native (mature) enzyme estimated by gel permeation chromatography and SDS-PAGE (19). Additionally, the deduced amino acid composition of the translated pLC 5.2 cDNA open reading frame corresponding to the putative mature peptide (amino acids 90-599) agrees very well with the amino acid composition of the native enzyme (Table I). Finally, ultrastructural immunogold cytochemical studies with anti- limonene cyclase polyclonal antibodies have demonstrated recently the limonene cyclase of mint to be located exclusively in the plastids (leucoplasts) of the glandular trichomes: a finding entirely consistent with the presence of such an or- ganellar targeting sequence in the limonene cyclase cDNA.

Spearmint ( M . spicatu) is a tetraploid and parent of peppermint (M. piperita = Mentha uquaticu x spicatu), a hexap- loid (51). Both contain ostensibly the same limonene synthase (19). RNA blot hybridization of pLC 5.2 insert DNA to spearmint and peppermint poly(A)+ RNA verified the presence of the homologous sequences in both species and provided an estimate of limonene cyclase mRNA transcript size of about 2400 nucleotides (data not shown). Thus, all lines of evidence, including RNA blot hybridization, indicate that the limonene cyclase structural gene from spearmint has been successfully isolated. All peptide sequence fragments obtained

’ J. Gershenzon and R. Croteau, manuscript in preparation.

56 -->

46 -->

30 -->

21.5 -->

FIG. 5. Immunoblot analysis of protein extracts from cultures of transformants. Equivalent amounts of the crude protein extracts were subjected to SDS-PAGE, and the corresponding immunoblots probed with the affinity purified polyclonal antibodies generated against the limonene synthase purified from spearmint. The lanes indicated are the native limonene synthase from spearmint ( S ) and the preparations from E. coli alone (XL-1 Blue) and E. coli transformed with Bluescript containing no insert ( p R l u e ) or the indicated clone ( p L C 4.1, pLC 5.2, pLC 8.5, and pLC 10.1). The protein distributions observed were the same at high (100 pg) and low (20 pg) sample loads. The migration of protein standards (in kilodaltons) is also indicated.

from the native enzyme correspond to sequences identified in the pLC 5.2 open reading frame, which is of the correct size and amino acid composition (excluding the putative transit peptide). The recombinant enzyme when expressed in E. coli is immunologically recognized by polyclonal antibodies raised to the purified limonene cyclase from spearmint and is catalytically active in generating a product distribution identical to that of the native cyclase.

The other full-length clones were partially sequenced and, although the first 100 nucleotides downstream of the translation start were identical in the cDNA inserts of pLC 4.1, pLC 5.2, and pLC 8.5, suggesting that they share a common open reading frame, several nucleotide differences were observed in the 5”untranslated regions of all three (Fig. 4). These differences suggest the presence of several limonene cyclase genes and/or alleles in the allotetraploid spearmint genome. A sesquiterpene cyclase, epi-aristolochene synthase, from tobacco is encoded by a gene family (24). Partial sequencing of the clone pLC 10.1, that was inactive in limonene cyclase expression, revealed a truncated transit peptide and several point mutations in the coding region (Fig. 4). Addi- tionally, the open reading frame of pLC 10.1 was out of frame with the lac2 translation initiation site. However, this com- plication alone is insufficient to explain the lack of expression since all of the other (full-length) cDNAs were also out of frame but nevertheless afforded functional expression of the limonene cyclase.

Based on the length of the full open reading frame, including the putative transit peptide, the transfected cultures should express a protein of a t least 69 kDa; however, immu-

23022 Limonene Synthase Cloning and Expression

Castor Bean 1 MALPSAAMQSNPEKLNLFHRLSSLPTTSLEYGN.NRFPFFSSSAKSHFKK 49

Spearmint 1 ....... MALKVLSVATQMAIPSNLTTCLQPSHFKSSPXLLSSTNSSSRS 43 1 . * . .:. :. I 1 1 : J : :: .. 1 : 11"I :.

Castor Bean

Spearmint

Castor Bean

Spearmint

Tobacco

Castor Bean

Spearmint

Tobacco

50 PTQACLSSTTHQEVRPLAYFPPTVWGNRFASLTFNPSEFESYDERVIVLK

44 RLRVYCSSSQLTTERRSGNYNPSRWDVNFIQSLLSDYKEDKBVIRASELV

100 KKVKDILISSTSDSVETVILIDLLCRLGVSYHFENDIEELLSKIFNSQ..

94 TLVK.MELEKETDQIRQLELIDDLQRMGLSDHFQNEFKEILSSIYLDHHY

1 ..... MLLATGRKLADTLNLIDIIERLGISYHFEKEIDEILDQIYNQN.. 148 .PDLVDEKECDLYTAAIVFRVFRQHGFKMSSDVFSKFKDSDGKFKESLRG

143 YKNPFPKEERDLYSTStAFRLLREHGFQVAQEVFDSFKNEEQEFKESLSD

44 .SN.....CNDLCTSALQFRLLRQHGFNISPEIFSKFQDENGKFKESLAS

. ..: 1 1 . . I . : : I . I: . I :.. .... * I - I

. 1 1 : :. .- . I : : 1 1 1 I / : / : I ~ / : ~ : : . ~ : ~ ~ . ~ : .: I I..: . I : I I I : : I : [ : [ ~ ~ : . ~ : . ~ ~ ~ . 1 1 ::

*

*

.: ....I 1 1 1 ... : . I I : : I : I I 1 . : . : I I . . I I : . : I . 1 1 1 1 1 . : - 1 - 1 1 :...1.11111:111.:..1:1..[.:1:[.I1111..

99

93

147

142

43

196

192

81

Castor Bean 197 DAKGMLSLFEASHLSVHGEDILEEAFAFTKDYLQSSAVE..LFPNLKRHI 244

Spearmint 193 DTRGLLQLYEASFLLTEGETTLESAREFATKFLEEKVNEGGVDGDLLTRI 242

Tobacco 88 DVLGLLNLYEASHVRTHADDILEDALAFSTIHLESAAPH..LKSPLREQV 135

Castor Bean 245

[ . : / : I I : I I I I . . I I . . ( I . I . I . . . : I : ... I : .:I :I I . I I I . I I I I I : I.::..II.I - 1 . 1 1 1 . . . :.:.I ..:

*

Spearmint 243 ... ... Tobacco 136 THALEQCLHKGVPRVETRFFISSIYDKEQSKNNVLLRFAKLDFNLLQMLH 185

Castor Bean 295

Spearmint 292

Tobacco 186 KQELAQVSRWWKDLDFVTTLPYARDRWECYFWALGVYFEPQYSQARVML 235

Castor Bean 345 AKWLLISLIDDTIDAYATMEETHILAEAVARWDMSCLEKLPDYMKVIYK 394

Spearmint 342 GKVNALITVIDDIYDVYGTLEELEQFTDLIRRWDINSIDQLPDYMQLCFL 3 9 1

Tobacco 236 VKTISMISIVDDTFDAYGTVKELEAYTDAIQRWDINEIDRPDYMKISYK 285

: I / I I . : T 1 ( . 1 : 1 : I I . :.: : ~ ~ ~ : . : : : . ~ ~ ~ ~ ~ . : . : . I . . : l . r n l . l I I : . I I I . : I I I . I I I I I . 1 1 . I I I I I . : : :

Castor Bean 395 LLLNTFSEFEKELTAEGKSYSVKYGREAFQELVRGYYLEAWEGKIPS 444

Spearmint 392 ALNNFVDDTSYDVMKEKGVNVIPYLRQSWVDLADKYWVEARWFYOOBKPS 4 4 1 I I ..: . :: I : . I I : . : : I . I :I1 I : I . 1 1

I: ::T: . . . . . . . ... :. ::. .I 1 1 . 1 1 . : 1 . . 1 .

: : : / I I : . ..: I : I I : I : . :..:.I . . . I I : . :.I I1.1.: .... :: I :: :... . :..:.I I ..:: I ::

: I I . : I I . . 1 . 1 1 1 . 1 : . : : : 1 1 1 . : . . . / . . ] I I :..: . : . [ I I : ) I : I I ~ ~ ~ : : : . : : : ~ : ~ . ~ ~ . I . 1 1 . : ...I1

Tobacco 286 AILDLYKDYEKELSSAGRSHIVCHAIERMKEWRNYNVESTWFIEGYMPP 335

Castor Bean 445 FDDYLYNGSMTTGLPLVSTASFMGVQEITGLNEFQWLETNPKLSYASGAF 494

Spearmint 442 LEEYLENSWQSISGPCMLTHIFFRVTDSFTKETVDSLYKYHDLVRWSSFV 4 9 1

Tobacco 336 VSEYLSNALATTTYYYLATTSYLGMKSATE.QDFEWLSKNPKILEASVI1 384

Castor Bean 495 IRLVNDLTSHVTEQQRGHVASCIDCYMNQHGVSKDEAVKILQKMATDCWK 544

Spearmint 492 LRtADDLGTSVEEVSRGDVPKSLQCYMSDYNASEAE-IAEVWK 5 4 1

Tobacco 385 CRVIDDTATYEVEKSRGQIATGIECCMRDYGISTKEAMAKFQNMAETAWK 434

Castor Bean 545

Spearmint 542 KMNAERVSKDSPFGKDFIGCAVDLGRMAQLMY.ENGDGHGTQHPI1HQQM 590

Tobacco 435 DINEGLL.RPTPVSTEFLTPILNLARIVEVTYIHNLDGYTHPEK~KPH1 483

*

...

.:I.: : :..I.:+.. ::[:I:.:: I 1 1 1 1 . . ......... . . . .

Castor Bean 592 KGLFVDPISI 601

Spearmint 5 9 1 TRTLFEPFA . 599

Tobacco 484 INLLVDSIKI 493

. : . : I : .

.. I . : . : FIG. 6. Amino acid sequence comparison of spearmint limonene cyclase (middle) with tobacco epi-aristolochene synthase

(bottom) and castor bean casbene synthase ( top) . The vertical bar marks identical residues. One and two dots indicate that two- and one-point mutations, respectively, are required to produce a match. Comparison was carried out with the GAP program of the Genetics Computer Group Packet (39). Asterisks are marked above conserved histidine and cysteine residues. Putative metal ion-substrate complex binding domains are underlined.


noblot analysis of total cell extracts from all of the functional cultures exhibited no detectable antigen of this molecular mass but rather a single major peptide at 54 kDa (compared with native enzyme a t 56 kDa) (Fig. 5). The most plausible explanation for the origin of this immunopositive 54-kDa peptide (and the larger peptides of lower abundance) is that translation reinitiation or run-through translation occurs. Proteolytic processing of the immature peptide by E. coli enzymes may also take place, much the same way as in spearmint plastids. Several leader peptidases of E. coli are similar to higher plant plastidal transit peptidase (52, 53), and such processing of a recombinant plastid-directed enzyme has been described previously (54). As observed in preliminary immunoblots, E. coli bearing Bluescript I1 SK(+) alone or cDNA pLC 10.1 did not express recognizable cyclase antigen using the purified antibody preparation (Fig. 5).

Homology with Other Terpene Cyclases-Searches of GenPept, Protein Information Resource, and SwissProt data bases (40) revealed no sequences with significant similarity to limonene synthase, nor did the sequence of limonene cyclase resemble those of any of the microbial sesquiterpene cyclases that have been determined recently, including the Fusarium and Gibberella trichodiene synthases (55, 56), the Penicillium aristolochene synthase (57), and the streptomyces pentalenene ~ y n t h a s e . ~ However, sequence comparison (39) with other terpene cyclases from higher plants revealed a significant degree of sequence similarity. The limonene cyclase from spearmint showed 34% identity and 56% similarity (based on conservative amino acid substitutions) to a sesquiterpene cyclase, epi-aristolochene synthase, from tobacco (24), and 31% identity and 53% similarity to a diterpene cyclase, casbene synthase, from castor bean' (Fig. 6). These plant terpenoid synthases represent three different types of cyclases from three diverse families, a monoterpene cyclase from the Lamiaceae, a sesquiterpene cyclase from the Solanaceae, and a diterpene cyclase from the Euphorbiaceae, suggesting a common ancestry for this class of enzymes.

Recent evidence has implicated histidine and cysteine residues at the active site of limonene synthase and of several other monoterpene and sesquiterpene cyclases (18). A search of the aligned sequences of limonene synthase, epi-aristolochene synthase, and casbene synthase revealed the presence of four such conserved residues, positioned at amino acids 124, 167, 251, and 516 of limonene cyclase (Fig. 6) .

The limonene cyclase sequence does not resemble any of the published sequences for prenyltransferases (58-62), a group of enzymes that, like the terpenoid cyclases, employ allylic pyrophosphate substrates and exploit similar electro- philic reaction mechanisms (45). The sequence (I,L,V)XDDXXD occurs in two of the three homologous domains of several prenyltransferases of diverse origin (63), and it has been suggested that these aspartate-rich elements function in binding the divalent metal ion-complexed pyrophosphate moiety of the prenyl substrates (64). The terpenoid cyclases would be expected to exhibit similar substrate binding requirements, and most sesquiterpene cyclase sequences contain this, or a very closely related, motif (24, 55-57). The deduced limonene cyclase peptide sequence contains the ele- ment VIDDIYD at residues 350-356 and the related sequence VDDTSYD a t residues 397-403. The former motif is strongly conserved in the three plant-derived cyclase genes, but the latter is not, and neither is near the 4 conserved histidine and cysteine residues. Cane (65) has urged caution in interpreting the functional role of these aspartate-rich motifs and, based on an observation by T. M. Hohn, has pointed out an addi-

D. E. Cane, personal communicat ion.

tional motif in prenyltransferases and cyclases that is rich in basic amino acids and that has been implicated at the active site of farnesyl pyrophosphate synthase (66). Similar patches of charged amino acids have been described in other terpenoid synthases (67), but neither region has very close analogy in the limonene synthase sequence.

An understanding of the catalytic role of primary sequence elements of limonene cyclase will require detailed study of structure-function relationships, an area of investigation that has heretofore been severely limited by enzyme availability. The isolation of the limonene cyclase cDNA permits the development of an efficient expression system for this functional enzyme with which such detailed mechanistic studies can be undertaken. The limonene cyclase cDNA also provides a useful tool for isolating other monoterpene cyclase genes and for examining the developmental regulation of monoterpene biosynthesis.

Acknowledgments-We thank Greg Wichelns for raising the plants, Chris Steele, John Crock, Jim Tsuru ta , Cheri Peyton, and A1 Koepp for technical assistance and helpful suggestions, and Karen Maer t ens for typ ing the manuscr ip t .

REFERENCES 1. Croteau, R., and Cane, D. E. (1985) Methods Enzymol. 110 , 383-405 2. Croteau, R. (1987) Chem. Reu. 87,929-954 3. Gershenzon, J., and Croteau, R. (1990) in Biochemistry of the Mevalonic

Acid Pathway to Terpenoids (Towers, G. H. N., and Stafford, H., eds)

4. Guenther, E. (1972) The Essential Oils, Vols. 111-VI (reprinted), R. E. pp. 99-160, Plenum Press, New York

5. Zinkel, D. F., and Russell, J. (1989) Naual Stores: Production, Chemistry, Krieger, Huntington, NY

6. Gershenzon, J., and Croteau, R. (1991) in Herbiuores: Their Interactions Utilization, Pulp Chemicals Association, New York

Berenbaum, M. R., eds) pp. 165-219, Academic Press, San Diego, CA with Secondary Plant Metabolites, Vol. I, 2nd Ed. (Rosenthal, G. A,, and

7. Harborne, J. B. (1991) in Ecological Chemistry and Biochemistry of Plant

426, Clarendon Press. Oxford Terpenoids (Harborne, J. B., and Tomas-Barberan, F. A,, eds) pp. 399-

8. Croteau, R. (1991) Planta Med. 57 (suppl.), 10-14 9. Croteau, R., and Satterwhite, D. M. (1989) J. Biol. Chem. 2 6 4 , 15309-

15316

10. Cramir: F., and Rittersdorf, W. (1967) Tetrahedron 2 3 , 3015-3022 11. Haley, R. C., Miller, J. A., and Wood, H. C. S. (1969) J. Chem. Soc. C 264-

12. Kobayashi, S., Tsutsui, M., and Mukaiyama, T. (1976) Chem. Lett. 1137- 268 1 1 0"

13. Vial, M. V., Rojas, C., Portilla, G., Chayet, L., Perez, L. M., and Cori, 0. 1 IS0

(1981) Tetrahedron 37. 2351-2357 14. Cori, 0. (1983) Phytochemistry 22, 331-341 15. Pauly, G., Belingheri, L., Marpeau, A,, and Gleizes, M. (1984) Plant Cell

16. Suga, T., Hirata, T., Izumi, S., Hiraga, Y., and Okamoto, K. (1988) Chem. Rep. 5,19-22

17. Perez, L. M., Pauly, G., Carde, J.-P., Belingheri, L., andGleizes, M. (1990) Lett. 115-118

18. Rajaonarivony, J. I. M., Gershenzon, J., Miyazaki, J., and Croteau, R. Plant Physiol. Biochem. 28,221-229

19. Alonso, W. R., Rajaonarivony: J. I. M., Gershenzon, J., and Croteau, R. (1992) Arch. Biochem. Biophvs. 2 9 9 , 77-82

20. Alonso, W. R., Crock, J. E., and Croteau, R. (1993) Arch. Biochem. Biophys. (1992) J. Biol. Chem. 2 6 7 , 7582-7587

21. Rajaonarivony, J. I. M., Gershenzon, J., and Croteau, R. (1992) Arch. 301 , 58-63

22. Alonso, W. R., and Croteau, R. (1992) in Secondary Metabolite Biosynthesis Biochem. Biophys. 296,49-57

and Metabolism (Petroski, R. J., and McCormick, S. P., eds) pp. 239- 251, Plenum Press, New York

23. Alonso, W. R., and Croteau, R. (1993) Methods Plant Biochem. 9, 239-260 24. Facchini, P. J., and Chappell, J. (1992) Proc. Natl. Acad. Sci. U. S. A . 8 9 ,

25. Gershenzon, J., McCaskill, D., Rajaonarivony, J., Mihaliak, C., Karp, F., 11088-11092

and Croteau, R. (1991) in Modern Phytochemical Methods (Fischer, N. H., Isman, M. B., and Stafford, H. B., eds) pp. 347-370, Plenum Press,

26. Gershenzon, J., McCaskill, D., Rajaonarivony, J. I. M., Mihaliak, C., Karp, New York

27. Lanznaster, N., and Croteau, R. (1991) Protein Expression Purif. 2 , 69-74 F., and Croteau, R. (1992) Anal. Biochem. 2 0 0 , 130-138

28. Croteau, R., and Karp, F. (1976) Arch. Biochem. Biophys. 176 , 734-746 29. Laemmli, U. K. (1970) Nature 227,680-685 30. Wray, W., Boulikas, T., Wray, V. P., and Hancock, R. (1981)Anal. Biochem.

118 . 197-203 31. 32.

33.

34.

35.

Schagger, H., and von Jagow, G. (1987) Anal. Biochem. 166 , 368-379 Cleveland, D. w. , Fischer, S. G., Kirschner, M. W.. and Laemmli, u. K.

Towbin, H., Staehelin, T., and Gordon, J. (1979) Proc. Natl. Acad. sei. (1977) J. Biol. Chem. 2 5 2 , 1102-1106

U. S. A. 76,4350-4354 Cathala, G., Savouret, J.-F., Mendez, B., West, B., Karin, M., Martial, J.,

Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: A and Baxter, J. (1983) DNA ( N . Y . ) 2 , 329-335

23024 Limonene Synthase Cloning and Expression Laboratory Manual, 2nd Ed., Cold Spring Harbor Laboratory, Cold Spring Harbor, NY

Mattick, J. S., Dalrymple, B., Kuramitsu, H., Siroza, T., Foster, T., Clark, W. P., Ross, B., Squires, C. L., and Maurizi, M. R. (1990) Proc. Natl.

36. Titus, D. E. (1991) Promega Protocols and Applications Guide, 2nd Ed., Promega Corp., Madison, WI

Acud. Sci. U. S. A. 87,3513-3517 53. Halpin, C., Elderfield, P. D., James, H. E., Zimmermann, R., Dunbar, B.,

38. Harlow, E., and Lane, D. (1988) Antibodies: A Laboratory Manual, Cold 54. Meadows, J. W., and Robinson, C. (1991) Plant Mol. Biol. 1 7 , 1241-1243 37. Sattemhite, D. M., and Croteau, R. (1987) J. Chromatogr. 452 , 61-73 and Robinson, C. (1989) EMBO J. 8,3917-3921

Spring Harbor Laboratory, Cold Spring Harbor, NY 55. Hohn, T. M., and Beremand, P. D. (1989) Gene (Amst.) 79 , 131-138 39. Devereaux, J., Haeberli, P., and Smithies, 0. (1984) Nucleic Acids Res. 12 , 56. HChE, and Desjardins, A. E. (1992) Mol. Plant-Microbe Interact. 5 ,

40. Altschul, S. F., Gish, W., Miller, W., Myers, E. W., and Lipman, D. J.

41. Gersbenzon, J., Maffei, M., andcroteau, R. (1989) Plant Physiol. 89,1351-

387-395

(1990) J. Mol. Biol. 215,403-410

1.157 42. Mieaskill, D., Gershenzon, J. G., and Croteau, R. (1992) Planta 187,445-

43. Kjonaas, R., and Croteau, R. (1983) Arch. Biochem. Biophys. 2 2 0 , 79-89 44. Wagschal, K., Savage, T. J., and Croteau, R. (1991) Tetrahedron 4 7 , 5933-

45. Poulter, C. D., and Rilling, H. C. (1981) in Biosynthesis of Isoprenoid 5944

Compounds (Porter, J. W., and Spurgeon, S. L., eds) Vol. 1, pp. 161-224, Wiley-Interscience, New York

454

46. Chou, P. Y., and Fasman, G. D. (1978) Adu. Enzymol. 47,45-147 47. Keegstra, K., Olsen, L. J., and Theg, S. M. (1989) Annu. Reu. Plant Physiol.

48. von Heijne, G., Steppuhn, J., and Herrmann, R. G. (1989) Eur. J. Biochem.

49. von Heijne, G., and Nisbikawa, K. (1991) FEES Lett. 2 7 8 , l - 3 50. von Heijne, G. (1983) Eur. J. Biochem. 1 3 3 , 17-21 51. Harley, R. M., and Brighton, C. A. (1977) Bot. J. Linn. SOC. 74 , 71-96 52. Gottesman, S., Squires, C., Pichersky, E., Carrington, M., Hobbs, M.,

Plant Mol. Biol. 40,471-501

180,535-545

57. Proctor, R. H., and Hohn, T. M. (1993) J. Biol. Chem. 268,4543-4548 58. Clarke, C. F., Tanaka, R. D., Svenson, K., Wamsley, M., Fogelman, A. M.,

59. Anderson, M. S., Yarger, J. G., Burck, C. L., and Poulter, C. D. (1989) J. 60. Asbby, M. N., and Edwards, P. A. (1990) J. Biol. Chem. 265,. 13157-13164 61. Carattoli, A,, Romano, N., Ballario, P., Morelli, G., and Maclno, G. (1991)

62. Fujisaki, S., Hara, H., Nishimura, Y., Horiuchi, K., and Nishino, T. (1990) J. Biol. Chem. 266,5854-5859

63. Math, S. K., Hearst, J. E., and Poulter, C. D. (1992) Proc. Natl. Acud. Sci. J. Biochem. (Tokyo) 108,995-1000

64. Ashby, M. N., Spear, D. H., and Edwards, P. A. (1990) in Molecular Biology U. S. A. 89,6761-6764

ojAtherosclerosis (Attie, A. D., ed) pp. 27-34, Elsevier Science Publishers B. V., Amsterdam

65. Cane, D. E. (1993) in Genetics and Biochemistry of Antibiotic Production (Vining, L. C., and Stuttart, C., eds) Butterworth-Heinemann, Stoneham, MA, in press

66. Brems, D. N., Bruenger, E., and Rilling, H. C. (1981) Biochemistry 2 0 ,

67. Chamovitz, D., Misawa, N., Sandmann, G., and Hirschberg, J. (1992) FEES 3711-3718

X4Y-Z3b

and Edwards, P. A. (1987) Mol. Cell. Biol. 7,3138-3146

Biol. Chem. 264,19176-19184

Lett. 2 9 6 , 305-311

Documents

OF BIOLOGICAL Vol. 268, of 5, pp. by for and U.S.A. 48 ... · 48-Limonene Synthase from the Oil Glands of Spearmint (Mentha spicata) cDNA ISOLATION, CHARACTERIZATION, AND BACTERIAL