Biosynthesis of Phytosterols

Cloned soybean sterol methyltransferase was purified from Escherichia coli to gel electrophoretic homogeneity. From initial velocity experiments, catalytic constants for substrates best suited for the first and second C1 transfer activities, cycloartenol and 24(28)-methylenelophenol, were 0.01 and 0.001 s–1, respectively. Two-substrate kinetic analysis using cycloartenol and S-adenosyl-l-methionine (AdoMet) generated an intersecting line pattern characteristic of a ternary complex kinetic mechanism. The high energy intermediate analog 25-azacycloartanol was a noncompetitive inhibitor versus cycloartenol and an uncompetitive inhibitor versus AdoMet. The dead end inhibitor analog cyclolaudenol was competitive versus cycloartenol and uncompetitive versus AdoMet. 24(28)-Methylenecycloartanol and AdoHcy generated competitive and noncompetitive kinetic patterns, respectively, with respect to AdoMet. Therefore, 24(28)-methylenecycloartanol combines with the same enzyme form as does cycloartenol and must be released from the enzyme before AdoHcy. 25-Azacycloartanol inhibited the first and second C1 transfer activities with about equal efficacy (Ki = 45 nm), suggesting that the successive C-methylation of the Δ24 bond occurs at the same active center. Comparison of the initial velocity data using AdoMet versus [2H3-methyl]AdoMet as substrates tested against saturating amounts of cycloartenol indicated an isotope effect on VCH3/VCD3 close to unity. [25-2H]24(28)-Methylenecycloartanol, [28E-2H]24 (28)-methylenelanosterol, and [28Z-2H]24(28)-methylene lanosterol were prepared and paired with AdoMet or [methyl-3H3]AdoMet to examine the kinetic isotope effects attending the C-28 deprotonation in the enzymatic synthesis of 24-ethyl(idene) sterols. The stereochemical features as well as the observation of isotopically sensitive branching during the second C-methylation suggests that the two methylation steps can proceed by a change in chemical mechanism resulting from differences in sterol structure, concerted versus carbocation; the kinetic mechanism remains the same during the consecutive methylation of the Δ24 bond.

H 3 ]AdoMet to examine the kinetic isotope effects attending the C-28 deprotonation in the enzymatic synthesis of 24-ethyl(idene) sterols. The stereochemical features as well as the observation of isotopically sensitive branching during the second C-methylation suggests that the two methylation steps can proceed by a change in chemical mechanism resulting from differences in sterol structure, concerted versus carbocation; the kinetic mechanism remains the same during the consecutive methylation of the ⌬ 24 bond.
Sterol methyltransferases (SMTs) 1 are ubiquitously represented in plants and fungi (1). Together, these enzymes gener-ate 24-alkyl sterol diversity, which includes formation of singly and doubly C-24-alkylated sterol side chains and olefin variants possessing ⌬ 24 (28) , ⌬ 23 (24) , and ⌬ 25 (27) side chains (2). In most organisms, SMTs catalyze the first committed step in the biosynthesis of phytosterols (3,4) (Fig. 1). The crucial role of these enzymes to generate an essential physiological group in sterol structure has stimulated considerable interest in the stereochemistry and mechanism of the C-methylation reaction (5,6). Several SMTs have been characterized at the molecular level, and their amino acid compositions reveal a highly conserved signature motif that represents the sterol-binding site (7,8). A number of these enzymes have been characterized, and they share similar native molecular masses in the range of 160 -175 kDa and similar properties (1). The methyl transfer reaction catalyzed by SMT is proposed to proceed through a nucleophilic attack by the electrons of the ⌬ 24 double bond on the S-methyl group of AdoMet (9 -11). The reaction can lead to the formation of a high energy intermediate (HEI) possessing a methyl at C-24 and a carbonium ion at C-25. After a hydride transfer from C-24 to C-25, an elimination of a proton at C-28 occurs, giving a 24(28)-methylene sterol. The steric course of the reaction has been hypothesized to proceed by an "X-group" (covalent), carbocation, or concerted mechanism (Scheme 1) (8). As recognized in the steric-electric plug model and X-group mechanism, the conformation of the bound sterol side chain can influence the configuration of the enzyme-generated product at C-24 and C-25 (Scheme 1A) (8).
Recent work on the stereochemistry of phytosterol (24-alkyl sterols) side chain carbon atoms harboring chemically or biosynthetically introduced 13 C label showed that the natural configuration for C-26 and C-27 of ergosterol and sitosterol (C-25 S; 2) is opposite to that considered in the X-group and carbonium ion mechanisms (C-25 R; 2a) (11)(12)(13) and therefore suggested a general catalytic mechanism for sterol C-methylation. Early kinetic studies carried out with microsomal preparations or detergent-solubilized enzyme preparations ruled out a Ping Pong (covalent) mechanism and supported either a concerted or carbocation mechanism for catalysis (9,10). Support for the carbocation mechanism was considered to result from the intermediacy of HEIs in sterol C-methylation catalysis, whereby mimics of the transition state species were effective inhibitors of plant and fungal SMT activity (1). However, the carbocation mechanism was recently eliminated as a possible route in ergosterol synthesis (operating the first C 1 transfer reaction) by work on the stereochemistry of the fungal enzyme-mediated CH 3 to CH 2 reaction and the stereochemistry at C-25 resulting from the 1,2-hydride shift reaction (11)(12)(13).
According to these stereochemical experiments, C-methylation of zymosterol (cholesta-8,24-dienol; native substrate) by yeast SMT is to occur via a noncovalent pathway whereby methyl addition to ⌬ 24 and deprotonation of C-28 give rise to a nucleophilic rearrangement in which H-24 migrates to C-25 on the re-face of the substrate double bond in concert with the initial ionization.
Previous investigators attempted to characterize the chemical and kinetic mechanism of the first and second C 1 transfer reactions catalyzed by plant SMTs by studying each individually (14) (Scheme 1). Thus, differences in pathway sequencing, substrate specificity, inhibitor recognition, and primary structures of the enzyme from different organisms have been interpreted as indicating that two classes of SMTs, SMT1 and SMT2, are responsible for the creation of two types of methylated species (14 -18). One enzyme activity, SMT1, exhibits a marked preference for position-specific ⌬ 24(25) -olefins monomethylating to form ⌬ 24 (28) -sterols in plants and fungi. The second activity, SMT2, was initially described as a 24(28)-methylenelophenol methyltransferase from plants (14), although it is now known to exist in fungi, where it is much more efficient for substrates such as 24(28)-methylenelanosterol (19). SMT2 is considered to monomethylate the product of the first C 1 transfer reaction, forming ⌬ 24 (28) Z-ethylidene sterols exclusively. Nes et al. (20) purified the yeast SMT to homogeneity in 1998 and generated a mutant by site-directed mutagenesis, making it clear that both methyl transfer reactions involving ⌬ 24 (25) -and ⌬ 24(28) -sterols can be catalyzed by a single enzyme and that the second C 1 transfer activity can result from the same active site as the first C 1 transfer activity to give multiple products. However, as of yet, the overall catalytic action of a plant SMT has not been established. The recent cloning of a fusion SMT from soybean (15) and our efforts to prepare a range of sterol substrates from natural sources or through synthesis has made it possible to generate large amounts of native protein for purification and detailed kinetic analysis. We now describe a systematic approach to determine the overall mechanism of SMT catalysis for a plant SMT, which necessitated experiments to establish a stereochemical correlation between the products of the first and second C 1 transfer activities and the predicted high energy intermediate of the reaction. In addition, the steady-state kinetic parameters derived for substrates and inhibitors have been compared in order to reveal topological relations as directed in the steric-electric plug model for SMT action (Scheme 1).

EXPERIMENTAL PROCEDURES
Material-The sources of reagents, authentic substrates, and sterol analogs isolated from nature or prepared synthetically, [ 1. Hypothetical pathway for cycloartenol conversion to sitosterol. SMT1, sterol methyltransferase that catalyzes the first C 1 transfer reaction; SMT2, sterol methyltransferase that catalyzes the second C 1 transfer reaction under physiological conditions. Boxed cycloartenol represents the three main domains of the substrate recognized by the SMT. SCHEME 1. Hypothetical C-methylation pathway. A, steric-electric plug model; B, X-group mechanism, where X is proposed to be an acidic amino acid; C, carbonium ion mechanism. Nu, cycloartenol nucleus. In each case, the isotopically labeled substrate contains a 13 C-26 atom.
Construction of Native SMT-The cDNA encoding soybean SMT (provided by the Noble Foundation) (15) was contained in the plasmid pSERT carrying a FLAG epitope-tagged SMT1 cDNA that generates a fusion protein. The entire open reading frame of the DNA encoding soybean SMT was transferred to a T 7 -based high level expression system, pET23a, and the resulting construct was transformed into BL21 cells for protein expression. The original cDNA was used for PCR amplification using Taq DNA polymerase (Stratagene) with the forward primer GGAATTCCATATGCAAAAAAAAAAAAAAAATCGAA-AC, which substituted the 5Ј terminus of pSERT and installed an NdeI restriction site at the starting methionine, in conjunction with the reverse primer, CGCGGATCCTTAGTTCCTGTCTAAATCAGGC, which introduced a BamHI restriction site following the stop codon to give the native form of the protein. The resulting amplicons were modified to include NdeI and BamHI (sticky ends) overhangs and ligated into TA cloning vector PCR2.1 (Invitrogen) to yield the corresponding vector, SMT/PCR2.1. This construct together with pET23a was doubly digested with NdeI and BamHI. The liberated SMT gene was then ligated into pET23a using the NdeI and BamHI overhangs generated during digestion. The resulting ligation mixture was transformed into BL21(DE3)competent E. coli cells by heat shock (20). Sequencing confirmed that no errors had been introduced by the polymerase reactions (dideoxy terminator sequencing ABI 373 sequencer).
cDNA Expression and Enzyme Purification-A frozen stock of the BL21(DE3) strain harboring the soybean SMT cDNA provided single colonies to inoculate 250 ml of Luria-Bertani medium containing ampicillin (50 g/ml) and grown for 10 h at 30°C in a floor shaker. The culture was used to inoculate 4 liters of Luria Broth divided into four Erlenmeyer flasks containing the same antibiotic. When A 600 of the suspended culture reached ϳ0.5, isopropyl-1-thio-␤-D-galactopyranoside was added to a final concentration of 0.4 mM, and the culture was incubated further at 30°C for 2 h with moderate shaking (200 rpm) to induce SMT expression. The cells were pelleted by centrifugation (4°C, 10,000 x g, 10 min), and the cell paste was snap frozen with liquid nitrogen and stored at Ϫ80°C or used directly. A portion of the cell paste (5 g) was resuspended in 30 ml of buffer A (50 mM Tris-HCl, 2 mM MgCl 2 , 2 mM ␤-mercaptoethanol, 1 mM EDTA, and 5% glycerol (v/v) at pH 8.5), and the suspension was lysed by passage through a French pressure cell at 20,000 p.s.i. Insoluble protein and cell debris were removed by centrifugation (4°C, 100,000 ϫ g, 1 h), and the supernatant (25 ml) was used for enzyme assay or as a source of SMT for further purification. All procedures were carried out at 4°C. In order to reduce contaminants, the 100,000 ϫ g supernatant was applied to a column of Q-Sepharose (ϳ20 nmol of SMT to a 20-ml column volume) pre-equilibrated with buffer A. To the first 25 ml of eluant was added 0.4% emulphogen, and the sample was stirred on ice for 30 min. The material was loaded onto a Q-Sepharose column containing buffer A and emulphogen detergent (to a final concentration of 0.4%). The column was developed with a stepwise gradient from 100 to 300 mM NaCl in buffer A and detergent (0.4%). Each of five fractions (30 ml/15 min) was monitored for SMT activity and by SDS-PAGE gel electrophoresis as described by Laemmli in a 12% polyacrylamide vertical slab preceded by 4% polyacrylamide stacking gel. Gels were stained with 2% (w/v) Coomassie Brilliant Blue 250R in methanol/acetic acid/water (45:10: 45). Protein content during the purification was estimated by the method of Bradford (24) referenced to bovine serum albumin. Fractions corresponding to 150 mM salt containing SMT activity were combined, desalted by washing with fresh buffer A (15 ϫ 5 ml), and concentrated (Amicon ultrafiltration YM-10 membrane) to a final volume of 10 ml. The average yield to this point was about 16 mg of total protein.
The solution from Q-Sepharose was loaded onto an Amersham Biosciences Mono Q TM HR 10/10 column that was previously equilibrated with buffer A containing 0.4% emulphogen. The column was eluted with a 150-ml stepwise gradient from 50 to 300 mM NaCl in the same buffer while monitoring the effluent as before. A set of fractions (2 ml each) in the 150 mM NaCl possessed a homogenous protein by SDS-PAGE (single ϳ41-kDa band). The concentration of SMT at the 90% or greater purity level was 0.3 mg/ml. In some preparations, a doublet was found to chromatograph in SDS-PAGE at the 36 -41-kDa range. The two species were electroeluted from the gel and Edman-sequenced. The lower band was found to possess a sequence identical to the top band except lacking in the first 56 amino acids, suggesting it to be a translational truncation product or partially proteolyzed form resulting from expression in the E. coli host. To generate a pure SMT species, the set of fractions corresponding to SMT activity eluting between fractions 16 and 30 were combined and loaded onto the FPLC again. Fractions (numbers 18 -22) corresponding to the center of the activity/mass peak were found to be pure by SDS-PAGE. These fractions were then used for k cat , molecular weight, and amino acid composition determinations. A second determination of the molecular mass of the pure protein was determined commercially by MALDI-TOF (Scripps Research Institute), giving a mass for the monomer of 41,597 Da. For a precise determination of protein concentration to establish the k cat value, the A 280 was measured using an extinction coefficient of 50,010 M Ϫ1 cm Ϫ1 . Native molecular weight of pure protein from FPLC was determined by gel permeation chromatography. A 0.5-ml sample of the 100,000 ϫ g supernatant in 0.4% emulphogen was loaded onto a calibrated (Bio-Rad reference standards from 1.67 to 670 kDa) Sephacryl S-300 gel filtration column coupled to an FPLC system eluted at a flow rate of 0.5 ml/min. Activity assays were performed on each fraction during the run from 10 to 300 min. A single peak of activity was observed between fractions 135 to 140, which corresponded to a molecular mass of ϳ161 kDa.
SMT Assays and Product Detection-The standard assay for recombinant SMT activity was performed in 600 l of total volume, 5-15 g of pure SMT, or 1-2 mg of total protein in buffer B (50 mM Tris buffer, 2 mM MgCl 2 , 2 mM 2-mercaptoethanol, and 20% (v/v) glycerol, pH 7.5), 50 M sterol substrate, 50 M [methyl-3 H 3 ]AdoMet at 0.6 Ci, and Tween 80 (1%, v/v) to produce 10 4 to 10 6 dpm of product in 45 min at 35°C. The incubation mixture was terminated with 500 l of a solution of 10% methanolic KOH. The methylated sterol product was extracted three times with 2.5 ml each in hexane (Fisher) and mixed on a vortex mixer for 30 s. The resulting organic layer was then transferred to a 7-ml scintillation vial, and the sample was taken for liquid scintillation counting to determine conversion rate (20). For kinetic analysis, standard assays were performed at 10 substrate concentrations from 5 to 200 M sterol with the AdoMet concentration held at saturation of 50 M. The K m and V max for AdoMet were determined from 5 to 200 M with sterol concentration held at 50 M. From the variation of reaction velocity with substrate concentration according to Michaelis-Menten kinetics, an apparent saturation of either substrate approached 50 M (data not shown). Random variations in measured velocities did not exceed Ϯ10%. Reactions up to 3 h were linear with protein concentration up to 2 mg/ml and were carried out such that there was about 70% conversion of substrate using the standard assay at saturating concentrations of cycloartenol and AdoMet. The enzyme preparations were free of contaminating sterol, which can be a problem when 100,000 ϫ g preparations are developed from plant tissues. The initial velocity data were determined using the computer program Sigmaplot 2001 plus the enzyme kinetics module software package. Data were fitted to the equation, using a nonlinear least-squares approach, and the kinetic constants Ϯ S.E. were never greater than 5% of the experimental measurement, and R 2 values were between 0.95 and 0.97. The product distribution generated by the SMT was determined at saturating levels of substrate and with sufficiently large preparations (5 mg/ml of protein) to ensure accuracy by GC/MS peak integration by the total ion current chromatogram (25). An estimate of the conversion of stable isotope-labeled substrates was obtained by summing the relevant ion intensities, followed by background correction, relative to those of the corresponding nonisotopically labeled enzyme-generated product in parallel experiments; isotopic abundance was calculated from the molecular ion(s) following background correction.
The back reaction to establish the equilibrium for the C-methylation reaction was performed with 5 mg/ml protein of partially pure SMT derived from the Q-Sepharose column fractions. The analysis was performed by standard assay, except saturating concentrations of [3-3 H]24(28)-methylenecyloartanol and AdoHcy were added to the reaction mixture for overnight incubation. The resulting tritiated sample was isotopically diluted with 25 g of cycloartenol, and the sterol composition was examined by HPLC-radiocounting (26).
Binding constants for the SMT were determined using equilibrium dialysis and the filter-binding method. Tritiated ligand, sterol or AdoMet, was assayed with pure soybean SMT by a standard protocol (7).
Two-substrate Kinetic Measurements-The two-substrate kinetic analysis was performed at cycloartenol concentrations spanning 10 -50 M and AdoMet concentrations spanning 20 to 100 M, and the data were fitted to the sequential (ternary complex) mechanism equation (Equation 2) using the same software package based on the algorithms of Cleland as discussed by Copeland (27), using a nonlinear least squares approach. Related formulations as discussed by Cleland to establish a Ping Pong (covalent intermediate) mechanism and the equilibrium-ordered mechanism were also examined (data not shown). K ma ϭ K m of AdoMet, K ia ϭ k d for AdoMet (dissociation constant to free enzyme where AdoMet binds prior to sterol), and K mb ϭ K m of sterol.
Steady-state Inhibition-Kinetic data from standard activity assays of product inhibition and dead end inhibition experiments were analyzed with the software package in analogous fashion to the initial velocity data analysis. The constant substrate in these experiments was held at subsaturating concentrations, and the varied substrate was added to the activity assays at several fixed concentrations ranging from 10 to 100 M AdoHcy, from 50 to 150 M 24(28)-methylenecycloartenol, from 5 to 50 M AdoMet, and from 5 to 50 M cycloartenol. Cyclolaudenol and 25-azacycloartanol were used in the dead end inhibition studies assayed in the individual experiments against sterol substrate at several fixed concentrations ranging from 10 to 100 M and from 10 to 100 nM, respectively. To investigate whether competitive (Equation 3), noncompetitive (Equation 4), or uncompetitive (Equation 5) inhibition was observed, the data were fitted to the respective equations based on the algorithms defined by Cleland (27) using nonlinear least squares analysis.
The data for individual experiments with each inhibitor versus varied substrate were fit to all three inhibitor models. Kinetic constants Ϯ S.E. are shown as relevant. V max is the maximum velocity, K m is the Michaelis constant for the varied substrate, S is the concentration of sterol or AdoMet substrate, I is the concentration of the inhibitor, and K i is the dissociation constant (assuming dissociation to free enzyme where AdoMet binds prior to sterol). Choice of kinetic fit was based on a combination of visual inspection and comparison of S.E. values and residual for all three inhibition types applied to the data sets.
Kinetic Isotope Effects-The KIE on overall rate (V) is determined from the ratios of the intercepts (V H /V D ) and the KIE on V/K is determined from the ratio of the slopes ((V/K) H /(V/K) D )). The V and V/K values and KIE are determined by fitting the steady-state data using the computer programs developed by Cleland as described by Cook (28). For determining the kinetic isotope effect in the first C 1 transfer reaction (29), measurement of initial velocities were obtained as a function of AdoMet concentrations ranging from 5 to 100 M and cycloartenol fixed at saturation with [ 2 H 3 -methyl]AdoMet (catalytic amounts of [ 3 H 3methyl]AdoMet was added to the preparation (21) or radioassay) or unlabeled compound. [24-2 H]cycloartenol was varied over the concentration range 5-150 M with [ 3 H 3 -methyl]AdoMet fixed at saturation. Kinetic analyses for substrate dependence of isotope effects in a bireactant sequential mechanism were performed as described by Cook (28). To evaluate the isotope effects associated with the second C 1 transfer reaction, [ AdoMet. Overall reaction rates with each substrate were determined by isolation of the olefin fraction (generated under linear conditions) followed by gasliquid chromatography-MS analysis or HPLC-radiocounting analysis of this material to determine product composition, from which rates for each enzymatic product were calculated. Confirmation of product identities was established by GC/MS and 1 H NMR of compounds isolated from activity assays (5 mg/ml protein) performed overnight with saturating amounts of sterol and AdoMet.
The observed changes in product composition resulting from the second C 1 transfer reaction due to isotopically sensitive branching can be used to assess the magnitude of KIEs for terminal deprotonations in enzyme-mediated catalysis in a similar manner to the KIEs determined for the first C 1 transfer reaction of corn and yeast SMT (10,21). For these types of isotopically branching experiments (30), the kinetics of the terminating deprotonation can be separated from the kinetics of the earlier steps in the reaction pathway by relating the velocity for the formation of one product to the velocity for the formation of the other product(s). The ratios of the velocities are equivalent to the ratios of the corresponding rate constants, because all products experience the same concentration of the branch point intermediate (23). The rate of total product formation catalyzed during the second C 1 transfer reaction can be calculated from Reaction 1, which is related to the minimal mechanism for the generation of three products (P) described as follows, where Ea is the enzyme-substrate complex containing AdoMet, S is the sterol substrate, kЈ 3 is taken as the rate constant for the irreversible C-methylation leading to the branch point intermediate EaSII (either the fucosterol, isofucosterol, or clerosterol cation), and kЈ 4 , kЈ 5 , and kЈ 6 are the rate constants for the terminating protonations (and the 1,2hydride shifts and dissociation steps) leading to the isomeric olefins P1, P2, and P3. Derivation of V max for the formation of P1 according to the steady-state approximation for enzyme catalysis, together with assumptions that the conversion of EaSI to EaSII is essentially irreversible ((kЈ 4 ϩ kЈ 5 ϩ kЈ 6 ) Ͼ Ͼ kЈ Ϫ3 ] and that the conversion of EaSII to products is much faster than C-methylation and that the kЈ 3 step involves a rate-limiting conformation change in the enzyme, leads to Equation 7. For a given enzyme catalytic reaction, in which multiple products are formed from a common intermediate, where V total is the total velocity of the reaction (the sum of the rates of formation of P1, P2, and P3; i.e. V 1 , V 2 , and V 3 , respectively) and is the proportionality constant kЈ 4 /(kЈ 4 ϩ kЈ 5 ϩ kЈ 6 ). Thus, isotopically sensitive branching arises from an increase in the concentration of ESII due to a decrease in kЈ 5 and kЈ 6 , so that kЈ 4 [EaSII], and thus V 1 is enhanced as discussed in terpene metabolism (30). The change in product distribution for the reference and deuterated substrate was used to calculate the observed primary KIE by substitution of the compositional values observed from the activity assays (see "Results") into the following relationship (Equation 8). Therefore, any KIE on formation of one product will give the isotope-sensitive branching effect on the other two. Thus, the observed KIE on forming P1 can be obtained from the equations related to olefin formation.
The observed KIEs on forming P2 and P3 arise from similar equations.

RESULTS
Enzyme Production and Kinetic Parameters-Expression of the soybean cDNA plasmid in E. coli resulted in the production of active soluble native SMT (Ͼ11 mg/liter of culture). We used this source of plant SMT to provide preliminary characterization of the first and second C 1 transfer activities; the enzymegenerated products were characterized by a combination of MS and 1 H NMR (31). Purification of the soybean SMT was accomplished with a modified protocol established previously for the purification of the yeast SMT (20), as summarized in Table I.
The resulting pure protein possessed the predicted molecular mass of the subunit (ϳ41.5 kDa) as determined by MALDI-TOF (data not shown) and SDS-PAGE (Fig. 2).
The native molecular mass obtained from gel filtration experiments was 161 kDa (Fig. 3), suggesting a tetramer. The pure SMT was analyzed for amino acid composition, and the first 13 amino acids at the N terminus of the protein were found to be MQKKKKNRNEVVL, in agreement with the predicted sequence (15).
Substrate specificity studies with a series of sterol analogs indicated that the enzyme can C-methylate both ⌬ 24(25) -and ⌬ 24(28) -sterols. The in vitro substrate specificity of the 100,000 ϫ g soluble soybean SMT was examined with a series of naturally occurring and synthetic ⌬ 24 -sterols. Based on the catalytic competence of a typical set of experiments, the affinities for sterols that bind productively to soybean SMT are similar to one another with an average K m of 30 M, but the rates of product formation differ greatly as a function of the different sterol features (Table II).
We found the enzyme recognizes at least three domains of the physiological substrate corresponding to the nucleophilic features at C-3 and C-24 and the nucleus structure (Fig. 1). Based on the catalytic competence for a series of analogs, the relative affinity of cycloartenol 1 compared with 24(28)-methylenecycloartanol 2 was similar to the substrate pair lanosterol 6 and 24(28)-methylenelanosterol 20 (Fig. 4), suggesting a position-specific recognition of the olefin acceptor molecule. To shed light on the intermediacy of the predicted cationic species that the ammonium analogs were intended to mimic, and thus for the electrophilic nature of this reaction type, we tested 25-azacycloartenol (21 in Fig. 5) for its ability to inhibit the SMT-catalyzed reaction in the direction of ⌬ 24(28) formation. Rahier et al. considered that it was possible to mimic the carbonium ion of 1c in Scheme 1 by replacing the C-25 with a nitrogen atom in the structure of the intermediate (32). The resulting 25-aza derivative, being essentially protonated under physiological conditions, can present certain electronic similarity with the HEI (1c in Scheme 1B) (i.e. a tetrahedral ammonium ion compared with the trigonal cabonium ion) and therefore might inhibit the C-methylation reaction. When 25azacycloartenol was tested against cycloartenol and 24(28)methylenelophenol, the K i for inhibition against either substrate was 40 nM. In both cases, the observed kinetic pattern was noncompetitive. These facts suggested that ⌬ 24(25) -and ⌬ 24(28) -sterol acceptor molecules bind to the same active center, assuming that a conformation change occurs during catalysis to generate the noncompetitive type inhibition (33). In addition, sitosterol, ergosterol, and cholesterol were tested as inhibitors (to a concentration of 150 M) of the first and second C 1 transfer activities; only sitosterol was an inhibitor of SMT activity, exhibiting a competitive-type kinetic pattern and a K i against cycloartenol of 100 M and against 24(28)-methylenelo-phenol of 150 M. The combination of these results suggests the active center is shape-and charge-specific for the natural side chain consisting of anionic site in the vicinity of the 24,25double bond and that the natural membrane insert can downregulate activity.
For the preferred substrate of the first C 1 transfer activity, cycloartenol, the turnover number was established to be 0.01 s Ϫ1 , and the number for the second C 1 transfer activity, 24(28)methylenelophenol, was determined to be 0.001 s Ϫ1 . The k cat  value for the ⌬ 24(25) -sterol acceptor molecule is similar to the value obtained for the yeast SMT (20). The K m and k cat values for AdoMet determined for the first C 1 transfer activity were similar to cycloartenol at 32 M and 0.02 s Ϫ1 .
Preliminary Kinetic Analyses of Soybean SMT-Bisubstrate kinetics utilizing the 100,000 ϫ g supernatant SMT with cycloartenol or AdoMet were performed in the absence of inhibitors to distinguish between a sequential (ordered or random) mechanism and a Ping Pong (covalent intermediate) mechanism. A plot of 1/V versus 1/[cycloartenol] or 1/[AdoMet] resulted in a series of intersecting lines illustrating an intersection between the binding of both substrates (Fig. 6).
These observations suggest that the enzyme has a ternary complex of the enzyme with the two substrates where catalysis involves a sequential kinetic mechanism. Measurement of the initial velocities as a function of deuterium-labeled and unla-  Table II.

FIG. 5. Structures of substrate and high energy intermediate analogs tested with the soybean SMT as described in the text.
beled AdoMet concentration against cycloartenol at saturating concentration gave similar V max effects and revealed that the V max and V/K effects are near unity. A similar kinetic experiment was performed using the standard assay with varying concentrations of [24-2 H]cycloartenol or cycloartenol (5-100 M) paired to a saturating concentration of [methyl-3 H 3 ]AdoMet (100 M). By comparison of activity assays, the apparent isotope effect was V H /V D ϭ 1.1. The lack of an isotope effect in those experiments suggests that C-28 deprotonation is not rate-limiting mechanistically and is predicted by the work of Arigoni (11).
Activity assays performed overnight with saturating amounts of cycloartenol and deuterium atoms incorporated in the enzyme-generated product and chemical reasonableness proves that the deuterium-labeled substrate actually was involved with the catalysis. The position of the deuterium label at C-28 or C-25 in the respective enzyme-generated product was confirmed by 1 H NMR (500 MHz) analysis against reference specimens (13,25). The proposed stereochemical model for the C-methylation of cycloartenol to 24(28)-methylenecycloartanol predicts that the methyl donation to ⌬ 24 occurs from the si (␤)-face of the substrate double bond (Scheme 1A) (8,12). The number of deuterium atoms incorporated into the methylated product from cycloartenol catalysis also indicates the methylation route used during the first C 1 transfer reaction. If there is a 24(28)-un-  (34), where the ␤-methyl of the 24␣/␤-methyl pair of HEI analogs of the first C 1 transfer reaction inhibited yeast SMT action and the stereochemical prediction for ␤-face methyl attack in plant sterol methylation (12).
Product Inhibition of Soybean SMT-To distinguish among the various possible kinetic models for substrate binding and product release, the products of the first C 1 transfer activity were used to inhibit the enzyme activity. To correctly analyze product inhibition data, the reversibility of the SMT-catalyzed reaction was analyzed first. This was accomplished by an activity assay performed overnight using preparative amounts of enzyme and substrate followed by HPLC-radiocounting of the enzyme-generated nonsaponifiable lipid fraction. There was no evidence for conversion of AdoHcy plus [3-3 H]24(28)-methylenecycloartanol to [3-3 H]cycloartenol, in contrast to similar types of experiments with the human sterol 8-isomerase, where the forward and backward isomerizations reactions could be demonstrated using a recombinant enzyme (35). The product inhibitors AdoHcy and 24(28)-methylenecycloartanol were evaluated against the substrates AdoMet and cycloartenol. Initial velocities were determined, and the data were plotted in double reciprocal form with 1/V versus 1/varied substrate at several fixed concentrations. A series of double reciprocal straight line plots intersected on the 1/V abscissa, demonstrating that AdoHcy was a noncompetitive inhibitor versus either AdoMet or cycloartenol as the varied substrate. Alternatively, assay and analysis of double reciprocal plots of 24(28)-methylenecycloartanol against cycloartenol and AdoMet indicated the methylated sterol to be a noncompetitive inhibitor against sterol substrate and a competitive inhibitor against AdoMet (Table III).
Dead End Inhibition Experiments-To further establish that 24(28)-methylenecycloartanol acts as a pure competitive inhibitor, we utilized an isosteric class of inhibitor represented by the olefin analog cyclolaudenol 24 in steady-state inhibition studies. This structural analog to cycloartenol in which the side chain has been modified to contain a 24␤-methyl group and a vinylic substituent in the side chain at the 25(27)-position could theoretically be an alternative SMT substrate to cycloartenol undergoing C-methylation in an analogous fashion to ⌬ 24(25) -or ⌬ 24(28) -sterol acceptor molecules. However, preparative activity assay with cyclolaudenol, a 9,19-cyclopropyl sterol synthesized by algae (36), failed to generate product (data not shown). Because cyclolaudenol was not a substrate, it was evaluated as  (Table III). As expected, cycloaudenol was a linear competitive inhibitor of cycloartenol (Fig. 7A). In studies with cyclolaudenol as the inhibitor and AdoMet as the varied substrate, at subsaturating concentration of cycloartenol there was a clear uncompetitive pattern of inhibition as reflected in the parallel line pattern of Fig. 7B. This pattern indicates that AdoMet must bind earlier than cyclolaudenol for cyclolaudenol inhibition to take place.
A second type of dead end inhibitor, 25-azacycloartanol 21 (37), was tested with the SMT. 25-Azacycloartanol could serve as an isoelectronic high energy intermediate analog of the native intermediate in the catalysis of cycloartenol by SMT. A related 25-aza steroid has been tested with a microsomal yeast SMT, where it was found to be competitive versus AdoMet and uncompetitive versus zymosterol (32) or noncompetitive against sterol or AdoMet (33), whereas in a microsomal corn SMT activity assay, 25-azacycloartanol was noncompetitive against cycloartenol and AdoMet (37). In our studies, noncompetitive inhibition was observed against cycloartenol as the varied substrate and uncompetitive against AdoMet as the varied substrate (Table III). Our results are consistent with the binding of inhibitor to an enzyme-AdoHcy complex such that AdoHcy is the second product released from the enzyme, and the binary complex is trapped by inhibitor prior to dissociation of the product.
Binding Studies-Using equilibrium dialysis, the dissociation constant for AdoMet binding to the SMT was determined. Varying concentrations from 0.25 to 7.5 M of [methyl-3 H 3 ]AdoMet were transferred into the buffer chamber, and 0.57 M SMT was transferred into the sample chamber. After equilibrium, samples were recovered from each chamber and counted by liquid scintillation. The data were analyzed as described elsewhere (7). A representative binding isotherm is shown in Fig. 8. The average value from three separate experiments yielded a k d value 2 Ϯ 0.5 M. Scatchard analysis indicated a single binding site per native enzyme. Efforts to establish binding of [3-3 H]cycloartenol to soybean SMT by either the equilibrium dialysis or filter binding methods were not successful. These binding data are consistent with the steadystate kinetic analyses showing that AdoMet must bind first in order for sterol to bind and establish a sterol-AdoMet-Michaelis complex. In related studies to determine the yeast SMT catalytic mechanism, line patterns of double reciprocal plots of substrates and inhibitors indicated a random mechanism (10). Further studies involving equilibrium dialysis with yeast SMT indicated that sterol and AdoMet can bind to the enzyme independent of each other (7). The latter data are consistent with a random mechanism with no abortive complexes being formed.
Conversion of 28-2 H Substrates to 24-Methylated Olefins-In our earlier investigation of the structure identification of enzyme-generated products of soybean SMT catalysis, we provided gas-liquid chromatography data that showed 24(28)methylenecycloartanol was converted to three products with base-line resolution in the chromatogram represented by the side chain structures 2k, 2l, and 2m in Scheme 2 (31). We now report the stereochemistry of the elimination reaction attending the second C 1 transfer reaction to the structurally related C-methylated sterol olefin sets produced by deuterium labeling of the C-28 trans-and cis-hydrogens of 24(28)-methylenelanosterol as shown in Scheme 2. The purpose of these experiments is 2-fold: 1) to establish whether the same stereochemical restrictions apply in the second C 1 transfer reaction as occur in the first C 1 transfer reaction as a result of C-methylation occurring at the same active site and 2) to provide data on isotopically sensitive branching as it relates to a rate enhancement in the formation of one or more products of a multiple product enzyme caused by a reduction of the rate constant for the formation of a second product due to isotopic substitution in the substrate (30). In the case of the first consideration, we chose C-28 deuterium-labeled sterol specimens (2l and 2f) as substrates to test for syn versus antimechanism in the methylation at C-28 of the sterol side chain by examining the stereochemical product distributions as predicted in Scheme 2.
Following incubation with each deuterated substrate, the overall biosynthetic rates, product distributions, and deuterium content of the products were determined by a combination of HPLC and capillary GC/MS analysis. Parallel incubations were performed with unlabeled substrate to obtain reference values for total rates and product distributions as well as background deuterium abundances. Because cycloartenol is limiting for synthetic purposes, we chose commercially available lanosterol as the starting material to prepare the stereospecifically labeled precursors [28E-2 H]24(28)-methylenelanosterol (2e) and [28Z-2 H]24(28)-methylenelanosterol (2f), and these compounds were used as substrates with the soybean SMT. 24(28)-Methylenelanosterol was found to be a suitable substrate (Table III). However, because it was a weak substrate for the second C 1 transfer activity, a higher concentration of protein was used in the assay, and the time of incubation was extended to 1.5 h for initial velocity experiments.
The steric outcome catalyzed by the second C 1 transfer reaction can be established based on whether Ha (designated as a "cis-process" or a syn mechanism, since the addition of the methyl group and proton loss occur on the same face of the ⌬ 24(28) -bond) or Hb (designated as a "trans-process" or an antimechanism, since the two events occur on opposite faces (22,38)) is eliminated to give isofucosterol and fucosterol. A total lack of stereochemical control would give identical results from either the 28E-2 H-labeled substrate (2e) or 28Z-2 H-labeled substrate (2f). Since the clerosterol side chain 2k should arise from the isofucosterol cation 2g, the resulting olefin 2k should retain deuterium from either progenitor as implied in Scheme 2.
Analysis of the enzyme-generated products with [28E-2 H]24(28)-methylenelanosterol or [28Z-2 H]24(28)-methylenelanosterol as substrate indicated that from each incubation two of the three olefins were labeled (Table IV). When [28E-2 H]sterol was assayed, sterols with clerosterol 2k (24␤-ethyl) and isofucosterol (24(28)-Z-ethylidene) 2m side chains were labeled with two and one 2 H-atoms, respectively. Alternatively, when [28Z-2 H]-sterol was assayed, the clerosterol and fucosterol (24(28)E-ethylidene) 2l side chains contained two and one 2 H atoms, respectively. Taken together, the biosynthetic data for C-methylation of the ⌬ 24 (28) bond to give the product set with side chains 2k, 2l, and 2m (Scheme 2) indicate that the C-24 double bond is formed via processing of the isofucosterol cation 2g by proton removal on the opposite face from AdoMet attack. These predicted results serve to confirm the stepwise sequence illustrated in Scheme 1, in which the isofucosterol cation either undergoes deprotonation at C-28 to generate the isofucosterol side chain structure, or rotation about C-24 (28) by 120°occurs to position the Ha atom adjacent to the same active site base involved with deprotonation of Ha from the isofucosterol cation. In this way, both the isofucosterol and fucosterol side chains can be formed at the same active site involving the same base. Because of the increased steric bulk evolving from the 24-ethyl side chain in the clerosterol cation 2k, some movement of the sterol side chain and the active site structure may occur in the activated complex to juxtapose an active site base normally out of position to deprotonate C-27 to generate the clerosterol side chain. By preparative assay with 5 mg of [25-2 H]24(28)-methylenecycloartanol, synthesized by the soybean SMT from [24-2 H]cycloartenol, the back reaction involving the 1,2-hydride shift of H-25 to C-24 was confirmed.
In each case, all three products contained one deuterium atom as evidenced by the molecular mass of M ϩ 455 atomic mass units (Table IV). The retention of the deuterium atom at C-25 in the HPLC pure 24(28)-olefin products as well as the movement of the deuterium atom to C-24 in the 25(27)-olefin product was confirmed by 1 H NMR (data not shown).
Kinetic Isotope Effects-Changes in the rate of biosynthesis (relative to control) of the product set giving rise to the clerosterol, isofucosterol, and fucosterol side chains occurred with the 24E-2 H-and 24Z-2 H-labeled substrate but not with the 25-2 Hlabeled substrate (Table IV). This result and the fact that we have expressed only one SMT enzyme from the bacteria host clearly demonstrate that 24-alkyl sterol diversity associated with phylogenetics can be generated in the same plant by the action of a single SMT. With [24E-2 H]24(28)-methylenelanosterol as substrate, an overall rate enhancement of 9% was observed and was coupled to a 25 and 29% rate enhancement of the 24␤-ethyl sterol and 24(28)Z-ethylidene sterol, respectively, against a 45% rate reduction in the 24(28)E-ethylidene sterol. Alternatively, with [24Z-2 H]24(28)-methylenelanosterol as substrate, an overall rate reduction of 46% was observed, but the total rate was affected by the 32 and 19% enhanced rate of formation of 24␤-ethyl and 24(28)E-ethylidene sterols compared with the 36% reduced rate of formation of the 24(28)Zethylidene sterol (Table IV). These rate differences arise from primary kinetic isotope effects on the initial, rate-determining 28-H elimination of the 24(28)-methylene sterol substrate. The magnitude of the KIE depends on the location of substitution with the heavy isotope. The change in product distribution between the reference and deuterated substrate was used to calculate the observed primary KIE by substitution of the compositional values reported in Table IV was obtained for the deprotonation step. These kinetic experiments on a native plant SMT and our studies on a site-directed mutant of the pure yeast SMT (25) constitute the first direct evidence concerning the formation of multiple olefins in the biosynthesis of phytosterols from a single enzyme species. DISCUSSION Generally, the kinetic constants of SMT have not been well characterized. The V max values reported seem to refer to enzyme activities and therefore to the corresponding variations in the amounts of the enzyme in different plants or fungi rather than to fundamental kinetic parameters of the enzyme reactions (10,14,19,26,33,34,36,37,39). The inability to properly characterize SMTs in these organisms also stems from not assaying the optimal substrate for a particular isoform, thereby underestimating the number of SMT isoforms present in the plant and complicating interpretation of the activity assays. Previously published kinetics for SMT relied on a microsomebound or soluble enzyme, which we now know can contain a low abundance of enzyme with significant substrate contaminants (9, 10, 20, 33). However, the use of recombinant SMT overexpressed in E. coli (20) has enabled the estimation of actual enzyme concentrations and, thereby, the comparison of exacting kinetic parameters for the first time.
The kinetic behavior of the pure soybean SMT has been shown here to generate two C 1 transfer activities related to the position-specific substrate acceptability of the enzyme. The most distinctive difference between the soybean SMT1 and yeast SMT1 seems to be the operation of a second C 1 transfer activity in the plant isoform with a 1-order of magnitude lower V max for the ⌬ 24 (28) acceptor molecule. Like the yeast SMT, the soybean SMT recognizes a set of strict features of sterol that relate to the nucleophilicity and three-dimensional shape of the molecule and is a rather slow enzyme with a low turnover number, ϳ0.01 s Ϫ1 . From the data presented under "Results," it is possible to construct a general kinetic scheme for soybean SMT-catalyzed consecutive C-methylation of the ⌬ 24 bond of an appropriate sterol acceptor molecule as shown in Scheme 3.
The bisubstrate intersecting line pattern with soybean SMT generated with the native substrates of the first C 1 transfer activity is consistent with a ternary complex mechanism for either the first or second C 1 transfer activity. A Ping Pong mechanism was clearly not operating during catalysis, thereby SCHEME 3. Kinetic scheme for soybean SMT catalyzed C-methylation of ⌬ 24 -bond. Path a and path b refer to the first and second C 1 -transfer activities, respectively. SCHEME 2. Stereochemical model for the formation of phytosterol olefins by C 1 and C 2 activities performed by soybean SMT. Nu, sterol nucleus. In each case, the isotopically labeled substrate contains a 13 C-27 atom.
ruling out an X-group mechanism. A key finding was the inhibition of SMT activity by cyclolaudenol. It was shown that replacement of the ⌬ 24 function of cycloartenol or 24(28)-methylenecycloartanol with a ⌬ 25 bond prevents enzyme-catalyzed C-methyl transfer. The lack of reactivity confirms the nucleophilicity of the ⌬ 24 bond is critical for enzyme-catalyzed reaction, because cyclolaudenol can bind with high affinity to soybean SMT as demonstrated by its inhibitory behavior. As expected, cyclolaudenol proved to be a linear competitive inhibitor of soybean SMT versus varied substrate cycloartenol. It was a clear uncompetitive inhibitor versus AdoMet. These results strongly suggest that AdoMet binds before sterol to the enzyme (i.e. that there is an ordered mechanism). Previous experiments with the corn SMT suggest a different mechanism for the first C-methylation of cycloartenol (37). This difference is inconsistent with the high homology among plant SMTs, especially within the sterol binding domains (8). In addition, the individual rate constants in the first C 1 transfer reaction catalyzed by soybean and yeast SMTs and the operation of a noncovalent mechanism in both cases (8,20) suggest a similar topography in the active site of this class of enzyme. The inhibition data from 25-azacycloartanol tested against cycloartenol and 24(28)-methylenelophenol and the isotopically sensitive branching experiments suggest that all of the enzymegenerated products can be catalyzed at the same active site, consistent with the steric-electric plug model for substrate recognition and catalysis by SMT (12,20). The ordered binding suggests either (i) a conformational change in the protein that causes the sterol binding pocket to become accessible only after AdoMet binds or (ii) sterol undergoes an important, direct noncovalent binding interaction with AdoMet in the enzyme active site. KIEs for C-methylation and the 1,2-hydride shifts in the first C 1 transfer reaction approaching unity support a conformation change of the enzyme that enfolds AdoMet to provide a productive complex for sterol binding and catalysis. The catalytic response of soybean SMT to ⌬ 24(25) -sterol (cycloartenol) and ⌬ 24(28) -sterol (24(28)-methylenelophenol) is different, despite the fact that affinities for ⌬ 24 -position-specific species are similar, suggesting that product release can be excluded as rate-limiting. However, the methylated product of the first C 1 transfer reaction must be released before AdoMet binds again in order to promote a recycling of sterol C-methylation. The low affinity for ⌬ 24(28) -methylene sterols for this SMT (SMT1) and the presence of another SMT (SMT2) in the plant capable of operating the second C 1 transfer reaction preferentially (1), as can occur in Arabidopsis and tobacco plants (16,17,39), suggest that under normal physiological conditions, SMT1 may only function to form cycloartenol, as depicted in the phytosterol pathway (Fig. 1). To date, neither 24␤-ethyl sterols or 24(28)Z-ethylidene sterols have been detected in soybean plants (40,41).
The question arises as to whether the noncovalent/sequential mechanism described here and for the yeast SMT is a general one for SMT enzymes. If so, it would suggest that the overall kinetics have structural implications with regard to the topography of the active site for this class of enzyme. The combination of high sequence homology and chemical labeling evidence in support of a common sterol binding site (1, 7) is thus far in agreement with our hypothesis. Another question is what terminates the consecutive C-methylation. Steric interference resulting from the increased bulk at C-24 may cause a misalignment of substrate in the active site, thereby generating a nonproductive complex for 24-ethyl analogs. Studies to establish the three-dimensional structure of the soybean SMT will provide further insight and are in progress.
The general picture that emerges from these and related studies is one in which the tightly coupled nature of the reaction sequence catalyzed by SMTs has prevented the second C 1 transfer reaction from being detected in a SMT1 isoform. In our earlier study on soybean SMT, we found that the Z methyl group on cycloartenol corresponding to C-27 is transformed into the R methyl group at C-25 on 24(28)-methylenecycloartanol by migration of the hydrogen atom at C-24 to C-25 from the re-face of the substrate double bond, suggesting a 24␤-alkyl carbenium ion (1a in Scheme 2) as intermediate for the first C 1 transfer reaction (31). These results and the finding presented herein that 24␤-methylcycloartanol but not 24␣-methylcycloartanol inhibits soybean SMT activity are consistent with the observation of Arigoni for an S N 2 reaction mechanism catalyzed by the yeast SMT (11). Chemical rationalization of a si-face mechanism with regard to structure of the clerosterol 2k (Scheme 1) product generated from the second C 1 transfer activity, as determined from our earlier study of soybean SMT (31), led to the proposal that the isofucosterol cation (2g in Scheme 2) is formed as a catalytic intermediate (Scheme 2). The chemical evidence from the activity assay with the [28E-2 H]24(28)methylenelanosterol and [28Z-2 H]24(28)methylenelanosterol as substrates determined in this study established that C-28 elimination during the second C 1 transfer reaction occurs by an antimechanism. The inverse value established by the isotopically sensitive branching experiments for the second C 1 transfer reaction is strong evidence that the proton transfer from the substrate to the active site base has come to equilibrium prior to the ratedetermining transition state and supports the proposal of a discrete carbocation intermediate. The chemical mechanisms associated with the first and second C 1 transfer activities differ in the order the two bonds are cleaved in the ⌬ 24(25) -and ⌬ 24(28) -substrates. In the carbocation mechanism, the bond to the leaving group is broken first, whereas in the concerted mechanism, both bonds are cleaved simultaneously without intervention of an intermediate. There can be much variability in the timing of steps in the nucleophilic rearrangements. Thus, the first C 1 transfer reaction operates a nonstop concerted mechanism, whereas the second C 1 transfer reaction operates a stepwise carbocation mechanism. The change in reaction mechanism with the binding of distinct position-specific olefins suggests that the reactants occupy different subsites in the active center.
The remarkable fidelity in olefin product formation by SMTs from less and more advanced organisms may reflect the inherent positional selectivity of the normal deprotonation step of the reaction and serve to fingerprint a highly conserved base(s) that catalyzes the regiospecific elimination of H-28 and H-27. It is interesting that ⌬ 25(27) -olefins are not generated by the first C 1 transfer activity, although precedent exists for both ⌬ 24(28)and ⌬ 25(27) -olefins to be formed by a plant SMT1 (29). A relatively small number of SMTs may be required to generate the product diversity of phytosterols produced by a given species. However, the great variety of structural types encountered in plants (as many as 60 sterols have been isolated from corn) (23) suggests that multiple SMT isoforms may be synthesized with individual isoforms synthesized at specific times during plant development. The most important advances in this study were the evidence that demonstrated plant SMT1 can catalyze the consecutive methylations, acting as if it were a SMT2; evidence showing an overall kinetic scheme for catalysis; the generation of product diversity from a single active site, particularly involving the second C 1 transfer reaction; and the altered reactivity between the ⌬ 24 (25) and ⌬ 24 (28) acceptor molecules that show mechanistic differences in the chemical steps catalyzed by an SMT enzyme. Such discoveries raise the intriguing possibility that understanding the fundamental issues of molecular recognition and catalytic control in phytosterol synthesis will lead to species-specific antifungals and engineering of modified plant sterol compositions that provide value-added traits, such as insect resistance.