HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 2, 1361-1369, January 11, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/2/1361    most recent M109224200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Jez, J. M. Articles by Noel, J. P. Search for Related Content PubMed PubMed Citation Articles by Jez, J. M. Articles by Noel, J. P. Social Bookmarking                      What's this?

Reaction Mechanism of Chalcone Isomerase

pH DEPENDENCE, DIFFUSION CONTROL, AND PRODUCT BINDING DIFFERENCES^*

Joseph M. Jez and Joseph P. Noel^§

From the Structural Biology Laboratory, The Salk Institute for Biological Studies, La Jolla, California 92037

Received for publication, September 24, 2001

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Chalcone isomerase (CHI) catalyzes the intramolecular cyclization of bicyclic chalcones into tricyclic (S)-flavanones. The activity of CHI is essential for the biosynthesis of flavanone precursors of floral pigments and phenylpropanoid plant defense compounds. We have examined the spontaneous and CHI-catalyzed cyclization reactions of 4,2',4',6'-tetrahydroxychalcone, 4,2',4'-trihydroxychalcone, 2',4'-dihydroxychalcone, and 4,2'-dihydroxychalcone into the corresponding flavanones. The pH dependence of flavanone formation indicates that both the non-enzymatic and enzymatic reactions first require the bulk phase ionization of the substrate 2'-hydroxyl group and subsequently on the reactivity of the newly formed 2'-oxyanion during C-ring formation. Solvent viscosity experiments demonstrate that at pH 7.5 the CHI-catalyzed cyclization reactions of 4,2',4',6'-tetrahydroxychalcone, 4,2',4'-trihydroxychalcone, and 2',4'-dihydroxychalcone are ~90% diffusion-controlled, whereas cyclization of 4,2'-dihydroxychalcone is limited by a chemical step that likely reflects the higher pK_a of the 2'-hydroxyl group. At pH 6.0, the reactions with 4,2',4',6'-tetrahydroxychalcone and 4,2',4'-trihydroxychalcone are ~50% diffusion-limited, whereas the reactions of both dihydroxychalcones are limited by chemical steps. Comparisons of the 2.1-2.3 Å resolution crystal structures of CHI complexed with the products 7,4'-dihydroxyflavanone, 7-hydroxyflavanone, and 4'-hydroxyflavanone show that the 7-hydroxyflavanones all share a common binding mode, whereas 4'-hydroxyflavanone binds in an altered orientation at the active site. Our functional and structural studies support the proposal that CHI accelerates the stereochemically defined intramolecular cyclization of chalcones into biologically active (2S)-flavanones by selectively binding an ionized chalcone in a conformation conducive to ring closure in a diffusion-controlled reaction.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Plants use flavonoids for protection against damaging UV light, as floral pigments for attracting pollinators, as inducers of Rhizobium nodulation genes, and as anti-microbial phytoalexins (1-3). Many flavonoids exhibit medicinal properties and are common constituents in human diets (4-5). During the early stages of the biosynthesis of these molecules, chalcone isomerase (CHI, EC 5.5.1.6)¹ catalyzes the intramolecular cyclization of 4,2',4',6'-tetrahydroxychalcone (chalcone) and 6'-deoxychalcone (4,2',4'-trihydroxychalcone), both derived from the upstream enzyme chalcone synthase, into (2S)-naringenin (5,7,4'-trihydroxyflavanone) and (2S)-5-deoxyflavanone (7,4'-dihydroxyflavanone), respectively (6-7) (Fig. 1A). Because chalcones spontaneously cyclize in solution to produce an enantiomeric mixture of flavanones, CHI guarantees formation of biologically active (2S)-flavanones. For example, (2S)-naringenin is the metabolic precursor of anthocyanin pigments, and mutations in the CHI gene are linked to changes in floral pigmentation (8). Recently, introduction of the petunia CHI gene into tomato resulted in fruits with enriched flavanol content (9). Curiously, CHI also appears to be unique to the plant kingdom (10).

View larger version (22K): [in this window] [in a new window]   Fig. 1.   CHI-catalyzed reaction and active site architecture. A, overall reaction catalyzed by CHI. CHI cyclizes 4,2',4',6'-tetrahydroxychalcone (R1, R2, R3 = OH), 4,2',4'-trihydroxychalcone (R1, R3 = OH, R2 = H), 2',4'-dihydroxychalcone (R1 = OH, R2, R3 = H), and 4,2'-dihydroxychalcone (R1, R2 = H, R3 = OH) into 5,7,4'-trihydroxyflavanone (R1, R2, R3 = OH), 7,4'-dihydroxyflavanone (R1, R3 = OH, R2 = H), 7-hydroxyflavanone (R1 = OH, R2, R3 = H), and 4'-hydroxyflavanone (R1, R2 = H, R3 = OH). The C-ring of the flavanone product is highlighted. B, view of the active site hydrogen bond network in the CHI·naringenin complex (17). Hydrogen bond interactions (small spheres) occur within a network centered on two water molecules (red spheres) that contact the flavanone ketone oxygen and through interactions of the 7-hydroxyl moiety of the flavanone product with Asn113 and Thr190. The figure was prepared with MOLSCRIPT (45) and rendered with POV-Ray (persistence of vision ray tracer; www.povray.org). C, proposed cyclization reaction catalyzed by CHI. After nucleophilic attack of the 2'-oxyanion on the ,-unsaturated double bond, a water molecule acts as a general acid to stabilize the enolate, resulting in formation of a flav-3-en-4-ol intermediate that tautomerizes into the expected reaction product.

CHI catalyzes the intramolecular cyclization of 4,2',4'-trihydroxychalcone with a second-order rate constant (k_cat/K_m) that approaches the diffusion-controlled limit and with a turnover rate that exceeds the spontaneous conversion rate by 10⁷-fold (11). Early efforts to understand the reaction mechanism of CHI relied primarily on chemical modification studies and were inconclusive in distinguishing whether CHI used general acid/base catalysis or nucleophilic catalysis for flavanone formation (12-16). Determination of the three-dimensional structure of Medicago sativa (alfalfa) CHI provided clarity by demonstrating the structural basis for CHI-mediated stereochemical control during cyclization and the chemical mechanism promoting flavanone formation (17).

The crystal structure of CHI complexed with (2S)-naringenin revealed that two hydrogen bond networks mediate CHI·flavanone product recognition (17) (Fig. 1B). An extensive hydrogen bond network at the bottom of the binding cleft involving 2 water molecules and 5 amino acids (Thr⁴⁸, Ala⁴⁹, Lys⁹⁷, Tyr¹⁰⁶, and Tyr¹⁵²) provides one set of interactions. This network centers on a water molecule contacting the C-ring ketone of (2S)-naringenin. Asn¹¹³ and Thr¹⁹⁰ provide a second set of hydrogen bonds to the 7-hydroxyl group of (2S)-naringenin, which corresponds to the 4'-hydroxyl group of the chalcone substrate. Based on the position of the water molecule situated between (2S)-naringenin and Tyr¹⁰⁶, a reaction mechanism (Fig. 1C) was proposed in which deprotonation of the substrate 2'-hydroxyl group occurs in solution with subsequent intramolecular attack of the newly formed oxyanion on the ,-unsaturated double bond of chalcone via a Michael addition. In this mechanism, complementary shape and electrostatic features between the active site of CHI and the substrate conformation preceding (2S)-naringenin formation as well as polarization of the ketone of chalcone facilitating the Michael addition reaction, accelerate the reaction rate 10⁷-fold.

In this report we compare the non-enzymatic and CHI-catalyzed cyclization reactions of 4,2',4',6'-tetrahydroxychalcone, 4,2',4'-trihydroxychalcone, 2',4'-dihydroxychalcone, and 4,2'-dihydroxychalcone. Examination of the pH dependence of these reactions supports the proposal that deprotonation of the chalcone 2'-hydroxyl moiety, forming a reactive oxyanion, occurs in solution at physiologic pH. Solvent viscosity experiments establish that CHI catalyzes the cyclization of 4'-hydroxychalcones with a near diffusion-limited rate and demonstrate that the rate-limiting step during flavanone formation changes from a diffusion-limited reaction to a chemically limited reaction with decreasing pH. Finally, the three-dimensional structures of CHI complexed with 7,4'-dihydroxyflavanone, 7-hydroxyflavanone, and 4'-hydroxyflavanone reveal differences in side-chain positions and product binding orientations that may affect the catalytic rate. The reaction mechanism of CHI, much like those of chorismate mutase and catalytic antibodies, appears driven by the inherent reactivity and highly ordered conformation of chalcone in the CHI active site.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- Oligonucleotides were purchased from Operon, Inc. Ni²⁺-nitrilotriacetic acid was from Qiagen. Benzamidine-Sepharose and the Superdex-75 fast protein liquid chromatography columns were from Amersham Biosciences, Inc. Thrombin was obtained from the Sigma. 4,2',4'-Trihydroxychalcone, 2',4'-dihydroxychalcone, and 4,2'-dihydroxychalcone were from Indofine.

Expression Vector Construction and Site-directed Mutagenesis-- M. sativa (alfalfa) CHI cDNA (18) was PCR-amplified using the following primers: 5'-dGCTACCAAAGACCATGGCATGGCTGCATCAATCACCGC-3' as the sense primer (the NcoI site is underlined, and the CHI start site is in italics) and 5'-dCCGGAATTCGGATCCTCAGTTTCCAATCTTGAAAGC-3' as the antisense primer (the BamHI site is underlined, and the stop codon is in italics). The resulting 0.7-kilobase DNA fragment was digested with NcoI and BamHI, gel-purified, and ligated into the pHIS8 expression vector (19). Automated nucleotide sequencing confirmed the fidelity of the expression construct (Salk Institute DNA sequencing facility).

Protein Expression and Purification-- The pHIS8-CHI expression construct was transformed into Escherichia coli BL21(DE3). Cells were grown at 37 °C in Terrific broth containing 50 µg ml¹ kanamycin until A_600 nm = 1.0-1.2. Cultures were shifted to growth at 20 °C for 6 h after induction with 0.5 mM isopropyl-1-thio--D-galactopyranoside. Cells were harvested and re-suspended in 50 mM Tris-HCl (pH 8.0), 500 mM NaCl, 20 mM imidazole (pH 8.0), 10 mM -mercaptoethanol, 10% (v/v) glycerol, and 1% (v/v) Tween 20. After sonication and centrifugation, the supernatant was passed over a Ni²⁺-nitrilotriacetic acid affinity column. The His₈-tagged protein was eluted with lysis buffer minus Tween 20 but containing 250 mM imidazole (pH 8.0). Thrombin was added to the eluant before dialysis overnight at 4 °C against the lysis buffer without Tween 20. Dialyzed protein was re-loaded on a mixed Ni²⁺-nitrilotriacetic acid/benzamidine-Sepharose column and eluted to remove uncleaved CHI and thrombin. Gel filtration of the flow-through fractions on a Superdex-75 fast protein liquid chromatography column equilibrated with 25 mM HEPES (pH 7.5), 100 mM NaCl, and 1 mM D/L-dithiothreitol resulted in purification to homogeneity. Fractions containing CHI were pooled, concentrated to 25 mg ml¹, and stored at 80 °C in 5 mM HEPES (pH 7.5), 25 mM NaCl, and 1 mM D/L-dithiothreitol after buffer exchange.

Kinetic Measurements-- The standard CHI assay was performed at 25 °C in a 0.5-ml reaction volume containing 50 mM HEPES (pH 7.5) and 5% (v/v) ethanol as co-solvent (11). Time-dependent decreases in absorbance were monitored with a Beckman DU-640 spectrophotometer. Initial velocity measurements were conducted in the standard assay system with varied concentrations of either 4,2',4',6'-tetrahydroxychalcone (10-200 µM), 4,2',4'-trihydroxychalcone (2-50 µM), 2',4'-dihydroxychalcone (2-50 µM), or 4,2'-dihydroxychalcone (2-50 µM). The poor solubility of some compounds limited the use of more concentrated substrate solutions. All assays conducted with 4,2',4',6'-tetrahydroxychalcone were corrected for the measurable background reaction rate. Saturation curves were fitted to the Michaelis-Menten equation using Kaleidagraph (Synergy Software). Non-enzymatic cyclization rates for 4,2',4',6'-tetrahydroxychalcone (200 µM), 4,2',4'-trihydroxychalcone (50 µM), 2',4'-dihydroxychalcone (50 µM), and 4,2'-dihydroxychalcone (50 µM) were determined spectrophotometrically under standard assay conditions in the absence of CHI. Determination of pH-rate profiles were performed under standard assay conditions in the presence or absence of CHI using a triple buffer system consisting of 50 mM AMPSO, 50 mM sodium phosphate, and 50 mM sodium pyrophosphate (pH 5.5-8.5) in place of HEPES buffer. The triple buffer system maintains constant ionic strength across a broad pH range (20-21). The pH profiles were fitted to either Equation 1 or 2 (22) using Kaleidagraph.

(Eq. 1)

(Eq. 2)

For solvent viscometric studies, k_cat and K_m values were determined as described above in either HEPES buffer (pH 7.5) or PIPES buffer (pH 6.0) containing varied sucrose concentrations (0-30%, v/v). The solvent viscosity (^rel) of each buffer solution was determined relative to the buffer conditions without sucrose using an Ostwald viscometer and Equation 3, where t is the solvent transit time, and  is the solvent specific gravity.

(Eq. 3)

Three-dimensional Structures of the CHI·Flavanone Complexes-- Crystals of CHI were grown at 4 °C by vapor diffusion using the hanging drop method from a 2-µl drop containing a 1:1 mixture of 25 mg of ml¹ CHI and crystallization buffer (25% glycerol (v/v), 5% ethanol (v/v), 2.0 M ammonium sulfate, and 50 mM PIPES (pH 6.5)) in the presence of either 2 mM 4,2',4'-trihydroxychalcone, 4, 2'-dihydroxychalcone, or 2',4'-dihydroxychalcone. After the addition of protein, the crystallization drop changed color from yellow to clear as the cyclization of chalcones into flavanones proceeded. Crystals of CHI complexed with either 7,4'-dihydroxyflavanone, 7-hydroxyflavanone, or 4'-hydroxyflavanone grew in space group P6₅22 with unit cell dimensions of a = 90.08 Å; c = 352.85 Å; a = 89.85 Å; c = 352.96 Å, and a = 89.66 Å; c = 351.18 Å, respectively, each with 2 molecules per asymmetric unit and a solvent content of 72%. Diffraction data (105 K) for the 7,4'-dihydroxyflavanone and 7-hydroxyflavanone complexes were collected from single crystals at beamline 9-2 of the Stanford Synchrotron Radiation Laboratory (SSRL) equipped with a Quantum 4 CCD detector (Table I). Diffraction data (105 K) for the 4'-hydroxyflavanone complex were collected from a single crystal at SSRL 9-1 using a 34.5-cm MAR imaging plate detector. All images were indexed and integrated using DENZO (23) and the reflections were merged with SCALEPACK (23). For each data set, after rigid-body refinement using the CHI apoenzyme structure (PDB 1EYP (17)) as a starting model in CNS (24), electron density resembling the respective flavanones was apparent, permitting modeling and placement in the CHI active site. After an initial round of simulated annealing refinement and multiple rounds of conjugate gradient minimization, B-factor refinement, and rebuilding in O (25), the R-factors converged to those listed in Table I. Model quality was checked with PROCHECK (26).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Expression, Purification, and Activity of Recombinant CHI-- Recombinant CHI was overexpressed in E. coli as a His₈-tagged protein, the octahistidine-tag was removed by thrombin digestion, and the cleaved protein was purified to homogeneity using a combination of Ni²⁺ affinity and gel filtration chromatography. Purified CHI migrates on SDS-PAGE gels with a molecular mass of 24 kDa and elutes from the gel filtration column as a monomer of approximately the same molecular mass (not shown). Steady-state kinetic values for recombinant CHI (Table II) correspond to those of native CHI purified from alfalfa and soybean (10-11) and, based on comparison of k_cat/K_m values, demonstrate the substrate preference of the enzyme for 4,2',4',6'-tetrahydroxychalcone and 4,2',4'-trihydroxychalcone over the dihydroxychalcones. Moreover, the catalytic efficiency of alfalfa CHI reflects the physiological substrate preference of CHI from legumes for 4,2',4'-trihydroxychalcone over 4,2',4',6'-tetrahydroxychalcone (10).

Rate Enhancement of CHI-catalyzed Chalcone Cyclization over the Non-enzymatic Reaction-- Because the ring-closure reaction converting bicyclic chalcones into tricyclic flavanones occurs in solution with a measurable rate, quantitation of the CHI-catalyzed rate enhancement is possible. In the absence of CHI, the velocity of the chalcone cyclization reaction in solution allows calculation of the uncatalyzed reaction rate according to Equation 4.

(Eq. 4)

With 4,2',4'-trihydroxychalcone and 2',4'-dihydroxychalcone, alfalfa CHI accelerates the rate of cyclization 13 million-fold and 9 million-fold, respectively (Table II). Using 4,2',4',6'-tetrahydroxychalcone and 4,2'-dihydroxychalcone as substrates, the rate enhancement for both compounds is ~2 million-fold. Previously, Bednar and Hadcock (11) reported a 36 million-fold increase in the cyclization rate of 4,2',4'-trihydroxychalcone over the spontaneous reaction rate with soybean CHI.

pH Dependence of Non-enzymatic and CHI-catalyzed Chalcone Cyclization Reactions-- In the proposed CHI reaction mechanism, the monoanionic form of the substrate is essential for catalysis, since the 2'-oxyanion acts as the nucleophile in the ensuing Michael addition to the ,-unsaturated double bond. Determination of the pH-rate profiles of the non-enzymatic and enzymatic reactions evaluated the contribution of the CHI active site to the protonation state of the transition state for the cyclization reaction.

The pH dependence of the uncatalyzed cyclization reactions are consistent with deprotonation of the chalcone substrates and reveals shifts in the pK_a as additional hydroxyl substituents are added to the aromatic ring containing the 2'-hydroxyl moiety (Fig. 2A and Table III). The electron withdrawing effect of multiple hydroxyl groups on the aromatic ring serves to stabilize the 2'-oxyanion. The determined pK_a values of the non-enzymatic cyclization reactions are similar to those observed previously and attributed to deprotonation of the 2'-hydroxyl group of a limited series of substituted 2'-hydroxychalcones (27-29).

View larger version (16K): [in this window] [in a new window]   Fig. 2.   pH dependence of non-enzymatic and CHI-catalyzed flavanone formation. The pH-dependence of the 4,2',4',6'-tetrahydroxychalcone (ovals), 4,2',4'-trihydroxychalcone (squares), 2',4'-dihydroxychalcone (triangles), and 4,2'-dihydroxychalcone (diamonds) cyclization reactions are shown. A, pH dependence of the non-enzymatic cyclization reactions. B, pH dependence of kcat for the CHI-catalyzed reaction. C, pH dependence of kcat/Km for the CHI-catalyzed reaction.

The pH dependence of k_cat and k_cat/K_m during the CHI-catalyzed cyclization reactions was evaluated for each chalcone substrate (Fig. 2, B-C). Overall, the pH rate profiles using these chalcones agree with the reported pH optimums of CHI activity (10). The pK_a values attributed to k_cat and k_cat/K_m decrease with substitution of the aromatic ring containing the 2'-hydroxyl group, as observed in the non-enzymatic reactions (Table III). Because the log (k_cat/K_m) versus pH profiles follow the ionization state of either free enzyme or free substrate and the determined pK_a values approximate those of the non-enzymatic cyclization reactions, the observed breakpoint likely corresponds to titration of the substrate 2'-hydroxyl group, not titration of an active site residue as previously suggested (6, 7, 30). The log (k_cat) versus pH profiles reveal that the pK_a of the enzyme-substrate complex shifts to a slightly more acidic pH (Table III). In these later cases, the perturbation of the pK_a values between the spontaneous and enzyme-catalyzed cyclization reactions results from the influence of the active site environment including specific hydrogen bond interactions and solvent exclusion from the CHI active site (31).

Solvent Viscosity Effects on the CHI Cyclization Reaction-- Because the diffusion constant for a small molecule is inversely proportional to the relative microviscosity of bulk solvent, solvent viscosity versus reaction rate studies can isolate the diffusion-controlled steps in enzymatic pathways (32-35). For the following kinetic mechanism (Equation 5), when the process is diffusion-controlled, increases in microviscosity modulate k₁ and k_-1, but not k₂.

(Eq. 5)

Furthermore, the ratio of K_m/V_max depends on the relative viscosity of the solution according to Equation 6 (32, 34).

(Eq. 6)

Finally, if the rate of product dissociation is kinetically significant, then 1/V_max versus the relative viscosity of the solution will vary according to Equation 7 (36).

(Eq. 7)

In cases where microviscosity changes do not alter the enzymatic rates, the rate-determining step is usually ascribed to a chemical step, not a diffusion-controlled process (37). Because the second-order rate constant of 4,2',4'-trihydroxychalcone cyclization for soybean CHI approaches the diffusion-controlled limit (11), we examined the effect of viscogens on alfalfa CHI at pH 7.5 and 6.0 to isolate the rate-determining step during the cyclization of 4,2',4',6'-tetrahydroxychalcone, 4,2',4'-trihydroxychalcone, 2',4'-dihydroxychalcone, and 4,2'-dihydroxychalcone.

At pH 7.5, increasing concentrations of the microviscogen sucrose affect the rate of CHI-catalyzed cyclization of 4,2',4',6'-tetrahydroxychalcone, 4,2',4'-trihydroxychalcone, and 2',4'-dihydroxychalcone but not the cyclization of 4,2'-dihydroxychalcone (Fig. 3A). This indicates that the rate-determining step for the cyclization of 4,2'-dihydroxychalcone occurs after CHI-substrate binding. For the three altered second- order cyclization reaction rates, as relative solvent viscosity increases, the K_m values for each increases, whereas the values of k_cat remain relatively unchanged. A plot of K_m/k_cat versus relative viscosity (Equation 6) yields values for k₁ and the partition ratio (k_-1/k₂) (Fig. 3B and Table IV). A plot of 1/k_cat versus relative viscosity (Equation 7) provides the values of k₂ and k₃ (Fig. 3C and Table IV), indicating that product dissociation is not rate-limiting. Finally, the quotient of k_cat/K_m and k₁ demonstrates that the cyclization reactions of 4,2',4',6'-tetrahydroxychalcone, 4,2',4'-trihydroxychalcone, and 2',4'-dihydroxychalcone are 84, 86, and 93% diffusion-controlled, respectively.

View larger version (25K): [in this window] [in a new window]   Fig. 3.   Solvent viscosity effects on the cyclization reactions catalyzed by CHI at pH 7.5 and 6.0. A, plot of the relative second-order rate constant as a function of relative viscosity (rel) for cyclization of 4,2',4',6'-tetrahydroxychalcone (ovals), 4,2',4'-trihydroxychalcone (squares), 2',4'-dihydroxychalcone (triangles), and 4,2'-dihydroxychalcone (diamonds) at pH 7.5. The superscript o denotes the absence of viscogen. The upper solid line (slope = 1) represents the theoretical diffusion-controlled reaction, and the lower solid line (slope = 0) represents a reaction without any viscosity effect. The dashed lines from top to bottom are fits of the 2',4'-dihydroxychalcone, 4,2',4',6'-tetrahydroxychalcone, 4,2',4'-trihydroxychalcone, and 4,2'-dihydroxychalcone reactions. B, plot of Km/kcat versus relative viscosity for cyclization of 4,2', 4',6'-tetrahydroxychalcone (ovals), 4,2',4'-trihydroxychalcone (squares), and 2',4'-dihydroxychalcone (triangles) at pH 7.5. C, plot of 1/kcat versus relative viscosity for cyclization of 4,2',4',6'-tetrahydroxychalcone (ovals), 4,2',4'-trihydroxychalcone (squares), and 2',4'-dihydroxychalcone (triangles) at pH 7.5. D, plot of the relative second-order rate constant as a function of relative viscosity (rel) for cyclization of 4,2',4',6'-tetrahydroxychalcone (ovals), 4,2',4'-trihydroxychalcone (squares), 2',4'-dihydroxychalcone (triangles), and 4,2'-dihydroxychalcone (diamonds) at pH 6.0. E, plot of Km/kcat versus relative viscosity for cyclization of 4,2',4',6'-tetrahydroxychalcone (ovals) and 4,2',4'-trihydroxychalcone (squares) at pH 6.0. F, plot of 1/kcat versus relative viscosity for cyclization of 4,2',4',6'-tetrahydroxychalcone (ovals) and 4,2',4'-trihydroxychalcone (squares) at pH 6.0.

In contrast, at pH 6.0, a smaller solvent viscosity effect is observed using 4,2',4',6'-tetrahydroxychalcone and 4,2',4'-trihydroxychalcone (Fig. 3D). The cyclization rate of both dihydroxychalcones remains unperturbed as the microviscosity of the reaction solution increases, thus demonstrating a rate-limiting chemical step during the CHI-catalyzed cyclization of these latter substrates. Plots of K_m/k_cat and 1/k_cat versus relative viscosity (Fig. 3, E-F) yield values of k₁, k₁, k₂, and k₃ (Table IV) for the reactions using 4,2',4',6'-tetrahydroxychalcone and 4,2',4'-trihydroxychalcone. Both the cyclization rate (k₂) and the product release rate (k₃) slow when using 4,2',4',6'-tetrahydroxychalcone as a substrate. The rate of catalysis (k₂) and the product release rate (k₃) are similar using 4,2',4'-trihydroxychalcone as a substrate, thus indicating that the product disassociation rate is limiting. Under these more acidic conditions, the CHI-catalyzed cyclization is 54 and 49% diffusion-controlled while using 4,2',4',6'-tetrahydroxychalcone and 4,2',4'-trihydroxychalcone, respectively.

High molecular weight polymers such as Ficoll alter the macroviscogenic properties of solvents without changing the diffusion rates of small molecules in solution. To ensure that the observed rate changes with solvent viscosity increases resulted from changes in microviscosity not macroviscosity, Ficoll 400 (^rel = 2.2) was used as a macroviscogen under identical assay conditions. As expected, this macroviscogen had no effect on the kinetic constants of CHI at either pH 7.5 or 6.0 (not shown). Finally, because no viscosity effects were observed when using the microviscogen sucrose at either pH 7.5 or 6.0 during assays with 4,2'-dihydroxychalcone as a substrate, the observed viscosity effects on the kinetic constants of CHI can be attributed to changes in microviscosity alone.

Three-dimensional Structures of CHI·Flavanone Complexes-- The 2.1-2.3 Å resolution x-ray crystal structures of CHI complexed with 7,4'-dihydroxyflavanone, 7-hydroxyflavanone, and 4'-hydroxyflavanone (the cyclization products of 4,2',4-trihydroxylchalcone, 4,2'-dihydroxychalcone, and 2',4'-dihydroxychalcone, respectively) were determined to investigate the substrate/product binding sites on CHI and the possible structural underpinnings of the observed kinetic differences for each chalcone. Complexes were generated by co-crystallization with the corresponding chalcones, but after the addition of CHI to the crystallization drops the (2S)-flavanone products were generated in situ. The overall structure of each CHI·flavanone complex is nearly identical to that of the CHI·naringenin complex determined previously (17). The root mean square deviations of the C atoms of the CHI·7,4'-dihydroxyflavanone, CHI·7-hydroxyflavanone, and CHI·4'-hydroxyflavanone complexes compared with the CHI·naringenin complex are 0.22, 0.27, and 0.16 Å, respectively. For the CHI·7,4'-dihydroxyflavanone complex, electron density attributed to the flavanone ligand was readily recognized in both CHI monomers found in the crystallographic asymmetric unit (Fig. 4A). As in the CHI·naringenin complex, the B-factors of the ligand in monomer B were noticeably higher than those of the ligand in monomer A. In the CHI·7-hydroxyflavanone and 4'-hydroxyflavanone complexes, clear electron density for the products was apparent only in monomer A (Fig. 4, B and C).

View larger version (43K): [in this window] [in a new window]   Fig. 4.   Overview of the three-dimensional structures of CHI complexed with flavanones. A, the SIGMAA-weighted  2Fo  Fc  electron density map (1.2 ) for 7,4'-dihydroxyflavanone (green). The ligand from monomer A is shown. B, the SIGMAA-weighted  2Fo  Fc  electron density map (1.2 ) for 7-hydroxyflavanone (gold). C, the SIGMAA-weighted  2Fo  Fc  electron density map (1.2 ) for 4'-hydroxyflavanone (blue). D, ribbon diagram of CHI with the positions of naringenin (aqua), 7,4'-dihydroxyflavanone (green), and 7-hydroxyflavanone (gold) superimposed. Only the side chains of the CHI·naringenin complex are shown for clarity. The figure was prepared with MOLSCRIPT (45) and rendered with POV-Ray (persistence of vision ray-tracer; www.povray.org).

Comparison of the CHI·naringenin, CHI·7,4'-dihydroxyflavanone, and CHI·7-hydroxyflavanone complexes reveals that each product binds in a similar orientation within the CHI active site (Fig. 4D). The positions of two key water molecules located at the bottom of the active site cleft in the CHI·naringenin complex (Fig. 1B) are maintained in the structures of CHI complexed with 7,4'-dihydroxyflavanone and 7-hydroxyflavanone (Fig. 5). Also, in each structure Asn¹¹³ and Thr¹⁹⁰ reside within hydrogen-bonding distance of the flavanone 7-hydroxyl groups. Although the set of intermolecular interactions observed between CHI and both 7-hydroxyflavanones remains largely unchanged, Thr⁴⁸ does adopt a rotamer conformation in each of the newly determined 7-hydroxyflavanone·CHI complexes that differs from the rotamer conformation observed in the CHI·naringenin complex (Fig. 5). Unambiguous electron density for the side chain of this residue in the 1.85 Å resolution CHI·naringenin complex indicated that the threonine hydroxyl group interacted with a water molecule at the active site (Fig. 1B). Electron density for this residue in the 2.3 Å resolution structures of CHI complexed with either 7,4'-dihydroxyflavanone or 7-hydroxyflavanone reveal that Thr⁴⁸ rotates 135° around the C-C bond, allowing the side-chain hydroxyl group to position itself for direct interaction with the C-ring ketone oxygen of each flavanone product.

View larger version (37K): [in this window] [in a new window]   Fig. 5.   Comparison of hydrogen bond interactions in the CHI 7,4'-dihydroxyflavanone (A) and 7-hydroxyflavanone (B) complexes. These views are rotated ~180° around the y axis from the view shown in Fig. 4D and depict the hydrogen bond interactions (dotted lines) at the CHI active site. The figure was prepared with MOLSCRIPT (45) and rendered with POV-Ray (persistence of vision ray-tracer; www.povray.org).

Relative to the 7-hydroxyflavanone complexes described above, the location of 4'-hydroxyflavanone within the CHI active site shifts position due to the absence of a 7-hydroxyl moiety and the presence of an additional water molecule in the active site cleft near the space previously occupied by the 7-hydroxyl group of other flavanone products (Fig. 6). The presence of the extra water molecule, which is hydrogen-bonded to Asn¹¹³ and Thr¹⁹⁰, repositions the chroman-4-one portion of 4'-hydroxyflavanone near Arg³⁶. In turn, this displacement repositions the C-ring ketone of 4'-hydroxyflavanone by 0.97 Å from the position observed in the three 7-hydroxyflavanone complexes described earlier. Movement of the 4'-hydroxyflavanone places the flavanone ketone oxygen 5.43 Å away from the hydroxyl group of Thr⁴⁸. In turn, Thr⁴⁸ adopts the rotamer conformation observed in the 7,4'-dihydroxyflavanone and 7-hydroxyflavanone complexes.

View larger version (87K): [in this window] [in a new window]   Fig. 6.   Product binding differences in the structure of CHI complexed with 4'-hydroxyflavanone. A, stereoview molecular surface representation of (2S)-naringenin (aqua) binding in the CHI active site. Both water molecules are shown as small spheres. B, stereoview molecular surface representation of 4'-hydroxyflavanone (blue) bound at the active site. The three observed active site water molecules are again shown as small spheres. In A and B, the surfaces corresponding to Asn113 and Lys109 have been removed for clarity. C, active site hydrogen bonds in the CHI·4'-hydroxyflavone complex. This view is rotated ~180° around the y axis from the view shown in panel B. The figure was prepared with MOLSCRIPT (45) and GRASP (46) and rendered with POV-Ray (persistence of vision ray-tracer; www.povray.org).

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

This current study comprehensively establishes the importance of substrate reactivity during the chalcone cyclization reaction catalyzed by CHI. Furthermore, this work quantitatively demonstrates using solvent viscosity changes that at physiological pH CHI operates near the diffusion-controlled limit with a physiologically relevant set of chalcone substrates. Finally, the three-dimensional structures of CHI complexed with 7,4'-dihydroxyflavanone, 7-hydroxyflavanone, and 4'-hydroxyflavanone reveal subtle product binding differences in the CHI active site. In total, these studies lay a structural and quantitative foundation for understanding the stereochemistry and chemical mechanism accompanying flavanone formation in plants.

Reactive Anionic State of Chalcone Substrates-- In the overall mechanism of flavanone formation, generation of the 2'-oxyanion is required for the ensuing intramolecular Michael addition to the ,-unsaturated double bond of chalcone substrates. Studies of the cyclization of substituted 2'-hydroxychalcones in solution demonstrate that deprotonation of the 2'-hydroxyl group readily occurs and that electron withdrawing groups enhance the reactivity of the monoanion at this position (27-29). Early biochemical characterization of CHI attributed the pH dependence of the enzyme to titration of an active site residue (16, 30); however, the x-ray crystal structure of CHI revealed that no obvious titratable residues over the pH range used are located within the active site (17).

Before the experiments presented here, no direct comparison of the non-enzymatic and enzymatic cyclization reactions of multiple chalcone substrates had been performed. The observed pH dependence of k_uncat, k_cat, and k_cat/K_m for the spontaneous and CHI-catalyzed cyclization reactions of 4,2',4',6'-trihydroxychalcone, 4,2',4'-trihydroxychalcone, 2',4'-dihydroxychalcone, and 4,2'-dihydroxychalcone demonstrate the importance of substrate deprotonation in the generation of flavanone products. Considering the CHI active site architecture, which lacks a readily identified general base capable of abstracting a proton from the substrate 2'-hydroxyl group, and the facile ionization of the 2'-hydroxyl group at physiological pH, the innate reactivity of chalcone most likely precedes enzyme-catalyzed ring closure.

In other words, the pH dependence of the non-enzymatic and CHI-catalyzed reactions implies that a significant portion of the physiologic substrate pool of 4,2',4',6'-tetrahydroxychalcone and 4,2',4'-trihydroxychalcone, key metabolites for anthocyanin floral pigments and phenylpropanoid phytoalexins, are found in the reactive deprotonated form. As proposed previously (17), the CHI active site locks the monoanionic form of the chalcone substrate into a sterically constrained conformation that positions the reactive 2'-oxyanion within proximity of the ,-unsaturated double bond to form a (2S)-flavanone with a 10⁷-fold acceleration over the spontaneous reaction rate. Given the reactivity of chalcone and the facile nature of the chalcone cyclization reaction in solution, CHI ensures rapid formation of biologically active (2S)-flavanones by operating near the diffusion-controlled limit.

CHI-catalyzed Chalcone Cyclization: Diffusion Versus Chemical Control-- The second-order rate constants (k_cat/K_m) for cyclization of 4,2',4',6'-tetrahydroxychalcone and 4,2',4'-trihydroxychalcone suggested that CHI catalysis is diffusion-controlled (11), but this had not been previously demonstrated. Also, although steady-state kinetic analysis provides a lower limit on the rates of substrate association (k_cat/K_m) and product formation (k_cat), perturbation of the steady-state mechanism in solvent viscosity experiments can provide estimates of the microscopic rate constants (38).

At both pH 7.5 and 6.0, the rate-limiting step in the cyclization of the least reactive substrate, 4,2'-dihydroxychalcone, is a chemical step, not a diffusion-controlled step. The rate-limiting step in the cyclization of 2',4'-dihydroxychalcone shifts from a diffusion-controlled process at pH 7.5 to a chemically limited process at pH 6.0. This trend toward a loss of diffusion control with decreasing pH is also observed with 4,2',4',6'-tetrahydroxychalcone and 4,2',4'-trihydroxychalcone, as evidenced by the shift from a 90% diffusion-controlled process at pH 7.5 to a 50% diffusion-controlled process at pH 6.0. Comparison of the changing rate constants for cyclization of these two chalcones by analysis of the solvent viscosity data shows that as the pH is lowered from 7.5 to 6.0, the rate of catalysis (k₂) decreases 10-20-fold, and the product off-rate (k₃) slows 40-160-fold. These changes contribute to increasing the chemical barrier to CHI catalysis. Furthermore, the viscosity effects are in agreement with the observed effect of pH on the chalcone cyclization reactions and corroborate the importance of the protonation state of the substrate observed in the pH dependence experiments.

Although the reaction catalyzed by CHI is under diffusion control, the k_cat/K_m values of CHI for 4,2',4',6'-tetrahydroxychalcone and 4,2',4'-trihydroxychalcone at pH 7.5 and 6.0 are significantly smaller than the theoretical diffusion limit of 10⁸-10¹⁰ M¹ s¹ for the collision of two molecules (35). For example, with 4,2',4'-trihydroxychalcone, only 4.5% (k_cat/K_m divided by 10⁸ M¹ s¹) of the substrate-enzyme encounters at pH 7.5 and 0.08% of the same collisions at pH 6.0 are catalytically productive ones that result in flavanone formation. Similar observations with other diffusion-limited enzymes, including chymotrypsin and alkaline phosphatase, have been attributed to the need for conformational changes within the active site of the enzyme, dynamic alterations of the substrate as the reaction progresses, or a combination of both mechanisms (34, 39). The three-dimensional structures presented here suggest that such structural changes at both the enzyme and substrate level may occur in CHI.

Structural Variation in Flavanone Binding-- Although many of the amino acid side chains at the CHI active site are observed in similar conformations in the apoenzyme structure and the four flavanone complexes now determined crystallographically, some active site residues, including Thr⁴⁸, adopt different conformations depending on the particular structural determination. This suggests that the CHI active site is a dynamic one and that catalysis may be sensitive to structural variations in the active site geometry. Likewise, differences in how the 7-hydroxyflavanones (Figs. 4 and 5) and 4'-hydroxyflavanone (Fig. 6) fit within the CHI active site imply that changes in substrate conformation do occur. These changes in both enzyme and substrate conformation may influence the percentage of catalytically productive diffusional collisions, as suggested above.

The differences in amino acid side-chain conformations at the CHI active site and the altered positions of flavanones in the binding pocket suggest that further examination of potential hydrogen bond interactions will reveal key residues involved in stabilizing the transition state of the cyclization reaction. Rotation of Thr⁴⁸ in the flavanone complexes presented here positions the side-chain hydroxyl group within hydrogen-bonding distance of the flavanone ketone group. In this conformation (Fig. 5), Thr⁴⁸ can directly interact with the transition state during the cyclization reaction (Fig. 1B). Ultimately, repositioning of the side chain of Thr⁴⁸ may impact catalysis as charge is delocalized from the 2'-oxyanion to the ketone oxygen. Similarly, the extra water molecule in the active site of the CHI·4'-hydroxyflavanone complex implies that rearrangement of solvent occurs within the CHI active site upon ligand binding, since this water is displaced by the 7-hydroxyl group of the 7-hydroxyflavanones.

The three-dimensional structures of CHI complexed with (2S)-naringenin, 7,4'-dihydroxyflavanone, and 7-hydroxyflavanone emphasize the importance of substrate reactivity in the cyclization reaction. Although these structures do not provide a view of substrate binding, the common binding mode observed with each of these flavanones implies that variations in the steady-state kinetic parameters result from differences in the reactivity of the substrate molecule itself. However, the altered position of 4'-hydroxyflavanone in the CHI active site compared with the 7-hydroxyflavanones suggests that both substrate reactivity and binding orientation ultimately contribute to efficient catalysis.

Conclusion-- Our results are consistent with the proposed mechanism for CHI (Fig. 1C). Furthermore, the structural and kinetic studies emphasize the importance of the protonation state of the substrate in the reaction and the requirement for proper orientation of both substrate and active site residues for catalysis. The chemical strategy underlying CHI catalysis, i.e. a facile non-enzymatic reaction enhanced by molding a reactive substrate into a productive binding conformation, resembles the reaction mechanisms of chorismate mutase and assorted catalytic antibodies (40-42). In these systems, the active site provides a precise template for catalysis. This specific three-dimensional contour binds the substrate in a conformation that affords facile substrate turnover by lowering the conformational barrier to catalysis and thus accelerating the reaction rate (43).

Plants have capitalized on this catalytic approach for optimizing production of flavanone metabolites required for the biosynthesis of anthocyanin floral pigments, anti-microbial phytoalexins, and inducers of Rhizobium nodulation genes. A provocative question concerns the evolutionary history surrounding this catalyst, since both the three-dimensional fold and the reaction of CHI appear unique to the plant kingdom. Also, the possibility of further optimizing CHI with a diverse collection of chalcones, including efforts at perfecting CHI (44), i.e. variants with improved collisional efficiencies against both natural and unnatural chalcones, remains to be explored.

	ACKNOWLEDGEMENTS

We thank C. Dana (Virginia Tech) for providing a sample of 4,2',4',6'-tetrahydroxychalcone for these studies, the Pollard lab for use of their spectrophotometer, and E. Dutil, C. Zubieta, and S. Koenig for assistance during data collection at SSRL.

	FOOTNOTES

^* This work was supported by National Science Foundation Grant MCB9982586 (to J. P. N.). Work performed at the Stanford Synchrotron Radiation Laboratory was supported by grants from the National Institutes of Health, National Center for Research Resources, Biomedical Technology Program, and the Department of Energy Office of Biological and Environmental Research.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The atomic coordinates and the structure factors (code 1FM7 (CHI|b17,4`-dihydroxyflavanone), 1FM8 (CHI|b17,4`-dihydroxyflavanone), and 1JEP (CHI|b14`-dihydroxyflavanone).) have been deposited in the Protein Data Bank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/).

^§ To whom correspondence should be addressed: Structural Biology Laboratory, The Salk Institute for Biological Studies, 10010 North Torrey Pines Rd., La Jolla, CA 92037. Tel.: 858-453-4100 (Ext. 1442); Fax: 858-452-3683; E-mail: noel@sbl.salk.edu.

A National Institutes of Health postdoctoral research fellow (Grant CA80396). Current address: Kosan Biosciences, Inc., 3832 Bay Center Place, Hayward, CA 94545.

Published, JBC Papers in Press, November 6, 2001, DOI 10.1074/jbc.M109224200

	ABBREVIATIONS

The abbreviations used are: CHI, chalcone isomerase; AMPSO, 3-[(1,1-dimethyl-2-hydroxyethyl)amino]-2-hydroxypropanesulfonic acid; chalcone, 4,2',4',6'-tetrahydroxychalcone; 4'-hydroxyflavanone, 2-(4-hydroxy-phenyl)-chroman-4-one; 7, 4'-dihydroxyflavanone, 7-hydroxy-2-(4-hydroxy-phenyl)-chroman-4-one; 7-hydroxyflavanone, 7-hydroxy-2-phenyl-1-chroman-4-one; naringenin, 5,7,4'-trihydroxyflavanone; PIPES, piperazine-N,N'-bis-(2-ethanesulfonic acid); SSRL, Stanford Synchrotron Radiation Laboratory.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				L. Ralston, S. Subramanian, M. Matsuno, and O. Yu Partial Reconstruction of Flavonoid and Isoflavonoid Biosynthesis in Yeast Using Soybean Type I and Type II Chalcone Isomerases Plant Physiology, April 1, 2005; 137(4): 1375 - 1388. [Abstract] [Full Text] [PDF]

				S. Hur, Z. E. R. Newby, and T. C. Bruice Transition state stabilization by general acid catalysis, water expulsion, and enzyme reorganization in Medicago savita chalcone isomerase PNAS, March 2, 2004; 101(9): 2730 - 2735. [Abstract] [Full Text] [PDF]

				M. Gensheimer and A. Mushegian Chalcone isomerase family and fold: No longer unique to plants Protein Sci., February 1, 2004; 13(2): 540 - 544. [Abstract] [Full Text] [PDF]

N. Shimada, T. Aoki, S. Sato, Y. Nakamura, S. Tabata, and S.-i. Ayabe A Cluster of Genes Encodes the Two Types of Chalcone Isomerase Involved in the Biosynthesis of General Flavonoids and Legume-Specific 5-Deoxy(iso)flavonoids in Lotus japonicus Plant Physiology,