THE J O U R N A L OF BIOLOGICAL CHEMISTRY 0 1988 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 263, No. 20, Issue of July 15, pp. 9582-9588,1988 Printed in U.S.A.

Purification and Characterization of Chalcone Isomerase from Soybeans*

(Received for publication, October 5 , 1987)

Rodney A. BednarS and John R. Hadcock From the Departments of Pharmacolozical Sciences and Chemistry, State University of New York at Stony Brook, Stony Brook,-New York 11 794-8651 -

Chalcone isomerase from soybean has been purified 11,000-fold over the crude extract. The purification procedure features pseudo-affinity chromatography on an Amicon Matrex Orange A column with selective elution by a product of the enzymatic reaction. The purified enzyme is greater than 99.5% pure and pos- sesses a specificity activity of 340 IU/mg, which is 520-fold greater than previously reported. The appar- ent molecular weight of the chalcone isomerase is 24,000 as determined from sodium dodecyl sulfate- polyacrylamide gels and from size exclusion chroma- tography under native conditions on Sephacryl S-200. The enzyme exists as a monomer that migrates on isoelectric focusing gels with a PI of 5.7. Amino acid analysis indicates that almost 50% of the residues are hydrophobic and yields a partial specific volume of 0.750 ml/g. Chalcone isomerase contains no carbohy- drate moieties and has a blocked N terminus. The purified enzyme catalyzes the conversion of 2’,4’,4- trihydroxychalcone (I) to (2S)-4’,7-dihydroxyflava- none (11) at pH 7.6 with a second order rate constant, kCat/K,,,, of 1.1 X los M-’ min-l and an apparent equi- librium constant, [II]/[I], of 7.6. The rate constant for the conversion of enzyme-bound substrate to the (2s)- flavanone, kcat = 11,000 min”, exceeds the sponta- neous conversion by 36 million-fold. The enzyme cat- alyzes the formation of (2s)-flavanone over 100,000- fold faster than to the (2R)-flavanone, indicating that the enzyme is highly stereoselective, yielding over 99.999% of the (2s)-flavanone.

Chalcone isomerase (EC 5.5.1.6) was first isolated from soybeans (Glycine max) (1, 2) and was shown to catalyze the cyclization of 2’,4’,4-trihydroxychalcone (I) to 4’,7-dihy- droxyflavanone (11). The enzyme, which is part of the flavo- noid pigment biosynthetic pathway of plants (3), is ubiquitous in all plants studied (4-9). Flavonoid pigments function in pollination and seed dispersal, attracting insects and birds by their colors (10). They also absorb UV light and act as sunscreens to protect DNA from UV damage (11). Recent evidence suggests that the enzyme is also involved in the biosynthesis of isofiavonoid phytoalexins (12, 13), which ap- pear to play a role in disease resistance (13, 14). Plants produce phytoalexins (low molecular weight, antimicrobial compounds) when stressed by wounding or infection by fungal pathogens (13,15,16). The production of phytoalexins results

*This research was supported by Grant GM 34832 from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ To whom inquiries and reprint requests should be addressed.

from the coordinate induction of several enzymes involved in the synthesis of flavonoids and isoflavonoids (13, 15, 16). Dixon and co-workers (13, 15, 16) have suggested that the induction results in de m v o protein synthesis of both active and inactive chalcone isomerase. The increase in enzyme activity results from both an activation of pre-existing inac- tive chalcone isomerase and the de m v o synthesis of new enzyme (13, 15, 16). The molecular basis of the proposed activation of pre-existing inactive enzyme is unknown. Mehdy and Lamb (17) have suggested that chalcone isomerase may be a regulatory enzyme in the biosynthesis of flavonoid pig- ments and isoflavonoid phytoalexins. Understanding of the structure, function, and mechanism of chalcone isomerase will require the availability of homogeneous enzyme, which is well characterized at the protein chemistry level.

In this paper, we report the large scale purification of chalcone isomerase and the determination of a number of its physical properties. Kinetic constants for the spontaneous and the enzyme-catalyzed reaction were determined in order to characterize biochemically the magnitude of the rate en- hancement produced by chalcone isomerase. Further, a novel sensitive method which combines kinetics and thermody- namics is used to quantitatively demonstrate the stereoselec- tivity of the reaction catalyzing the formation of (2s)-flava- none by purified chalcone isomerase.

111 1111

EXPERIMENTAL PROCEDURES AND RESULTS’

Spectral Changes on Isomerization-Incubation of 2’,4‘,4- trihydroxychalcone (I) with chalcone isomerase results in a rapid loss of the chalcone absorbance at 390 nm and appear- ance of the 4‘,7-dihydroxyflavanone (11) absorbance at 334 nm (see Fig. 4A). The sharp isosbestic points are indicative of the production of a single product, which is supported by HPLC’ analysis of the reaction mixture. Incubation of the

Portions of this paper (including “Experimental Procedures,” part of “Results,” Figs. 1-3, Table 1, and Refs. 42-60) are presented in miniprint at the end of this paper. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are included in the microfilm edition of the Journal that is available from Waverly Press.

The abbreviations used are: HPLC, high performance liquid chro- matography; naringenin, trivial name for 4’,5,7-trihydroxyflavanone; Bis-Tris, bis(2-hydroxyethyl)iminotris(hydroxymethyl)methane; PIPES, piperazine-N,N’-bis(2-ethanesulfonic acid); PAGE, poly- acrylamide gel electrophoresis; SDS, sodium dodecyl sulfate.

9582

Purification and Characterization of Chalcone Isomerase 9583

5 0. 5 50. D 0

n L 0 M 2 0.

0.

300 Wavelength(nm1

400 500 601

1.21 t ( h r s ) - > A CHALCONE B

-. 20.8 Y

u u C :0.6 L 0 VI

g0.4 v‘ 23 49 6C

0.2

0 . 0 300 400 5 00 601

WavelengthCnm)

FIG. 4. UV-Vis spectra. A, UV-Vis spectra of the enzyme-cata- lyzed conversion of 2’,4’,4-trihydroxychalcone to the 4’,7-dihydroxy- flavanone. 2’,4’,4-Trihydroxychalcone (41 PM) was incubated at pH 7.6 in 50 mM Tris-chloride buffer with chalcone isomerase (0.05 IU/ ml). The absorbance spectra before addition of the enzyme and after 15,30,45,60, and 300 s are shown. The absorbance spectrum of pure 4’,7-dihydroxyflavanone is overlaid onto the other spectra. The chal- cone has an absorbance maximum at 390 nm with an extinction coefficient of 29,400 * 1,000 M” cm“, while the flavanone has no significant absorbance above 370 nm. The absorbance maximum of the flavanone is a t 334 nm with an extinction coefficient of 19,200 * 1.000 M” cm”. Sharp isosbestic points are seen at 265, 278,299, and 343.5 nm with extinction coefficients of 5,820,6,410,7,410, and 15,500 M” cm”, respectively. The equilibrium constant was calculated from these spectral data since the absorbance at 380 nm and above is due entirely to the chalcone and allows an accurate measure of the chalcone concentration at equilibrium (see Table 111). E , UV-Vis spectra of the spontaneous conversion of 2’,4’,4-trihydroxychalcone to the 4’,7-dihydroxyflavanone. 2’,4’,4-Trihydroxychalcone (43 FM) was incubated at pH 7.6 in 50 mM Tris-chloride buffer in the absence of chalcone isomerase. The absorbance spectra are shown at selected times during the course of the reaction. The spectrum of pure 4’,7- dihydroxyflavanone is overlaid onto the other spectra. The isosbestic point at 344 nm is sharp; however, the isosbestic points at 265 and 278 nm are not maintained since the absorbance increases by -0.02 absorbance unit (AU) at these wavelengths over the course of the spontaneous reaction. The absorbance at 390 nm is also higher than expectedbasedontheabsorbanceabove410nmandtheshapeofthe chalcone absorbance spectrum. The inset shows a semilogarithmic plot of the fraction of reaction, ( A ( t ) - A(m))/(A(O) - A(m)), against the hours of incubation.

2’,4’,4-trihydroxychalcone in the absence of enzyme results in a slow spontaneous conversion of the chalcone into the corresponding flavanone. The spectral changes (see Fig. 4B) are similar t o those seen in the presence of enzyme (Fig. 4A).

0.0 ’ ’ ’ I

-0.2 0.0 0.2 0.4 0.6 0 .8 1.0 1.2

1 /[SI 4- FIG. 5. Dependence of the initial velocity of the enzyme-

catalyzed reaction on the concentration of 2’,4’,4-trihy- droxychalcone. The initial velocity of the chalcone isomerase cat- alyzed reaction using 2’,4’,4-trihydroxychalcone as substrate was determined at 25 “C and pH 7.6 in 50 mM Tris-chloride buffer. The data are presented in a double reciprocal format but were fitted by a nonlinear least squares fit to the Michaelis-Menten equation. Kinet- ics constants are given in Table 111.

The absorbance changes observed in the spontaneous reaction follow first order kinetics. Semilogarithmic plots of the frac- tion of reaction, (A(t) - A(w))/(A(O) - A(m)), against time were linear for at least 90% of the reaction (see inset in Fig. 4B). The half-time for the spontaneous reaction is 17.8 f 1.1 h.

Equilibrium Constant-An equilibrium constant for the spontaneous reaction of 15 k 1 was calculated from the end point spectrum (e.g. t = 300 h in Fig. 4B). The spectrum was fit as a linear combination of pure chalcone and pure flava- none. The fit was not good around 390 nm since the actual absorbance was approximately 10% higher than expected. The effect of this small additional absorbance, presumably due to a side product, is expected to be minor as long as it is not mistakenly attributed to chalcone. The equilibrium constant was essentially constant when analyzed at an absorbance of about 410 nm. The validity of this treatment was supported by determination of the equilibrium concentrations by HPLC analysis after 5 and 10 days.

The apparent equilibrium constant for the enzyme-cata- lyzed interconversion 2’,4’,4-trihydroxychalcone and 4‘,7- dihydroxyflavanone at pH 7.6 was determined from the ab- sorbance spectrum of the end point mixture (see t = 300 s in Fig. 4A). The residual absorbance above 380 nm is entirely due to the chalcone since the flavanone product has no absorbance in this region. The spectrum of the end point mixture can be precisely simulated as a mixture of pure chalcone and pure flavanone. When the reaction is started with chalcone, the apparent equilibrium constant is 7.6 f 0.2 (see Table I11 and “Experimental Procedures”). However, use of the enzyme to rapidly produce the chalcone from a racemic (R,S)-flavanone results in a ratio of the flavanone to the chalcone concentration of 16 f 2.

Kinetics of the Enzyme-catalyzed Reaction-The initial ve- locity (v’) of the enzyme-catalyzed reaction as a function of substrate concentration follows the Michaelis-Menten equa- tion with a K,,, of 10 f 2 p ~ . This value is 2.5-fold higher than previously reported at 30 “C (18). Unlike the enzyme from Phaseolus vulgaris, which shows strong substrate inhi- bition (8), no substrate inhibition is seen at concentrations up t o 10 x K,,, (Fig. 5). Additional kinetic and inhibition constants for the enzyme-catalyzed reaction are given in Table 111.

9584 Purification and Characterization of Chalcone Isomerase

DISCUSSION

Enzymatic reactions involving the addition of a nucleophile to a double bond conjugated to a carbonyl group are an important class of biochemical reactions (19). Chalcone iso- merase is unique in this class because it catalyzes an intra- molecular reaction in which the same substrate molecule contains both the nucleophile as well as the double bond (see Equation 1). Unlike most enzyme-catalyzed reactions where the conversion of substrates into products cannot be detected in the absence of the enzyme, the conversion of chalcones to flavanones occurs spontaneously in the absence of the en- zyme. The ability to study both the mechanism of the spon- taneous reaction (20-22) and the enzyme-catalyzed reaction (23, 24) provides a rare opportunity to gain insight into how an enzyme accelerates the rate of a chemical reaction. Unfor- tunately, the enzyme has proven very difficult to obtain in high purity and in the quantities necessary to allow detailed studies of the enzyme structure and how it relates to the mechanism of catalysis. In order to define the rate accelera- tion afforded by chalcone isomerase and to provide a well characterized homogeneous enzyme for mechanistic studies, we have purified the enzyme to homogeneity and character- ized both the spontaneous and enzyme-catalyzed reaction.

Purification-A significant contribution of the present work involves development of a purification strategy which permits the large scale generation of essentially homogeneous (>99.5%) chalcone isomerase. The purification procedure fea- tures the use of an Orange A dye pseudo-affinity column with selective elution by naringenin, a product of the enzyme reaction (Fig. lB, Miniprint). The specific activity of the purified enzyme, 340 IU/mg, is 520-fold greater, and the recovery of enzyme activity is over 7 times greater than previously reported (18). Details of the purification procedure and a number of the physical properties of the enzyme are detailed in the Miniprint.

Molecular Weight-We chose soybeans as the source for chalcone isomerase because its molecular weight was reported to be significantly smaller than the enzyme from all other sources. The small size makes it suitable for analysis by physical methods such as NMR. However, there is a discrep- ancy between the molecular weight of 24,000 determined here under both native and denatured conditions (see Miniprint) and the previously reported value of 15,600 determined by ultracentrifugation (18). The difference may be due to proteo- lytic digestion or to the presence of lower molecular weight impurities in their preparation. Our preparation protocol in- cludes phenylmethylsulfonyl fluoride as a protease inhibitor to minimize possible proteolysis.

The molecular weight reported here is consistent with the subunit and native molecular weight of 27,000 reported for the enzyme from P. vulgaris (15). By comparison, higher native molecular weights have been reported for the enzyme from parsley (50,000 (25)) and from the petals (62,500) and anthers (44,000) of petunia (26). Since the subunit molecular weights for these other sources have not been determined, it is unclear whether the high native molecular weights seen in parsley and petunia are due to dimer formation or to a larger subunit size.

Amino Acid Composition-The first amino acid composition for any chalcone isomerase is reported in Table 11. The protein contains almost 50% hydrophobic amino acid residues, an unusually high value in comparison with most soluble pro- teins, which contain -30-40% hydrophobic amino acid resi- dues (27-29). Only a single cysteine residue was detected as cysteic acid following performic acid oxidation of the protein or by titration of the protein with 5,5’-dithiobis-(2-nitroben-

TABLE I1 Amino acid composition and extinction coefficient of chalcone

isomerase (Glycine mar) A known amount of chalcone isomerase, as determined by its

absorbance at 280 nm, was spiked with an internal standard and hydrolyzed in 6 N HCI, and the quantity of amino acids was deter- mined as described under “Experimental Procedures.” The micro- moles of each type of amino acid residue (Ai) divided by the amount of protein hydrolyzed ( P ) are given in the second column. These results represent the average of 22 HPLC analyses of nine independ- ent hydrolysates of chalcone isomerase. The residues per protein molecule reported in the third column includes data obtained from 12 additional HPLC analyses of two hydrolysates for which the absorbance at 280 nm of the original enzyme sample was not known. These results were normalized using a value of phenylalanine equal to 12.7, in order to obtain an overall molecular weight of 24,000. The partial specific volume of chalcone isomerase in aqueous solution, uz0, was calculated from the amino acid composition using the method of Lee and Timasheff (30). The extinction coefficient reported at the bottom of the second column was calculated from the data in that column and the molecular weights of the amino acid residues ( W;) according to the equation below:

A;/P Residues/pro- tein molecule

Asx Glx SeP GlY His Arg Thr” Ala Pro Tyr Val Met Ile Leu Phe LYS Cys‘ Trpd c2m (cm mg/ml)-’ u70 (ml/el

pmol residues/(Anm rnl/crn) 0.51 f 0.06 1.22 f 0.13 0.92 f 0.04 1.11 f 0.06 0.27 2 0.04 0.38 f 0.05 0.54 f 0.04 1.55 f 0.12 0.61 f 0.06 0.34 f 0.05 1.01 f 0.10 0.16 k 0.03 0.77 f 0.08 1.02 f 0.11 0.69 f 0.07 0.98 f 0.11

0.06 0.12

0.76 f 0.07 0.750

9.3 f 0.8 22.1 f 0.6 17.4 f 1.6 20.2 f 1.1 5.0 f 0.6 7.0 f 0.3 9.8 f 0.4 28.6 f 0.6 11.5 f 0.6 6.2 f 0.2 18.4 f 0.5 2.9 f 0.5 13.9 f 0.7 18.7 f 0.4 12.7 f -b

18.2 f 0.3 1 2

“Values corrected for destruction by dividing by 0.96 (Thr) and - .~~ , ”,

0.91 (Ser) for each 24 h of acid hydrolysis. These values were deter- mined based on hydrolysis of standard amino acids mixtures and from comparison of 24-, 48-, and 72-h hydrolysates of chalcone isomerase.

The results of all 32 HPLC analyses were normalized by the moles of phenylalanine in order to calculate the standard deviations re- ported for the other amino acids.

Determined as cysteic acid following performic acid oxidation of the protein (39) and by titration with 5,5’-dithiobis-(2-nitrobenzoic acid) in 6 M guanidine hydrochloride (40).

Determined spectrophotometrically (41).

zoic acid) in 6 M guanidine hydrochloride (Table 11). Two tryptophan residues were determined spectrophotometrically. The enzyme contains 5 histidine residues, 1 of which has been proposed to function as an active site base (18) or nucleophile (23).

Post-translational Modification-Dixon et al. (16) reported that the equilibrium distribution of the enzyme from P. uul- garis on cesium chloride gradients in density labeling experi- ments indicates that a significant fraction of the elicitor- induced enzyme activity is a result of activation of pre- existing, inactive enzyme. Further, they report that the active enzyme has an unusually low buoyant density of 1.237 g/ml which corresponds to a partial specific volume of 0.808 ml/g

Purification and Characterization of Chalcone Isomerase 9585

(16). Based on the amino acid composition (Table 11), we have calculated (30) a partial specific volume of 0.750 ml/g. This value is typical of many proteins (28). The soybean enzyme should not have an unusual partial specific volume, unless the enzyme is subjected to post-translational modification. I t is conceivable that the activation of enzyme seen by Dixon results from post-translational modification by attachment of, for example, carbohydrate or lipid. No unusual peaks were seen in the amino acid analysis, nor was any attachment of carbohydrate detected; however, the N terminus of the en- zyme is blocked, indicating some post-translational modifi- cation (see “Results”). The molecular basis for the activation of pre-existing inactive enzyme and the unusually low buoyant density of the P. vulgaris enzyme remain unclear.

Stereochemistry and Stereoselectivity of the Enzyme-cata- lyzed Reaction-The equilibrium constant for the enzyme- catalyzed interconversion of chalcone and flavanone appears to depend on the direction from which the equilibrium is approached. Starting with chalcone, the apparent equilibrium constant is 7.6 f 0.2 (Table 111). However, the ratio of the flavanone to the chalcone concentration is 16 f 2 when the enzymatic reaction is started with racemic (R,S)-flavanone. The difference in these numbers and the observed equilibrium constant of 15 f 1 for the spontaneous reaction (Table 111) in either direction can be quantitatively explained if the enzyme rapidly catalyzes the equilibration of the chalcone with only one of the optical isomers of the flavanone. Indeed, the soybean enzyme has been reported to produce a levorotatory

TABLE I11 Conversion of 2‘,4‘,4-trihydroxychalcone to 4’,7-dihydroxyflavanone

at 25 “C and p H 7.6 in 50 mM Tris-chloride buffer The rate constant ( k f ) for the spontaneous conversion of the

chalcone into the flavanone was determined from the initial rate of loss of the chalcone absorbance. The first order rate constant ( k o b d ) was determined from the exponential decay of the chalcone concen- tration over 90% of the reaction (Fig. 4B, inset). The equilibrium constant was determined from the ratio of the flavanone concentra- tion to the chalcone concentration as described under “Experimental Procedures” and “Results.” The steady state kinetic constants for the enzyme-catalyzed reaction were determined from a nonlinear least squares fit to the Michaelis-Menten equation (Fig. 5). The apparent equilibrium constant for the enzyme-catalyzed reaction was deter- mined by starting with the chalcone and adding enzyme to produce a rapid loss of the chalcone (Fig. 4A). Spontaneous reaction

k/ (6.0 f 0.4) X min” N = 7“ kobd (6.5 f 0.4) X min” N = 7 X,, = [(2R,2S)-flava- 15 +- 1 N = 15

none]/[chalcone] Enzyme-catalyzed

reaction k c a t 11,000 +- 1,000 min” N = 6 X 32 k,.t/K, (1.1 f 0.2) X 109M” min” N = 6 X 32 K, 1 0 f 2 p M N = 6 X 3 2 Kq = [(ZS)-flava- 7.6 f 0.2 N = 10

K, for naringenin 48 f 10 KM N = 2 X 2 4 28 & 8 p M b N = 3 X 1 2

none]/[chalcone]

Rate enhancement‘ k c a t l ( 0 . 5 k f ) 3.6 X 10’ k c a t / K m / ( O . B k / ) 3.7 X lo’* M”

The first value is the number of independent experiments used in calculating the standard deviation, while the second value indicates the number of points used in each experiment to determine the indicated constant.

bAt 10 “C and pH 6.5 in 10 mM Bis-Tris-chloride buffer. The inhibition is competitive.

Calculated as the ratio of the rate constants with and without enzyme catalysis for the reaction of the chalcone into the S-isomer of the flavanone.

(-)-flavanone (1). The absolute stereochemistry of levorota- tory 4’,7-dihydroxyflavanone is known to be 2s (31). There- fore, when the reaction is started with chalcone, the apparent equilibrium ratio is dependent on the equilibrium concentra- tion of the (2s)-flavanone which is produced from the chal- cone. When the reaction is started with racemic (R,S)-flava- none, the flavanone concentration in the apparent equilibrium ratio is higher because of the extra contribution due to the initial concentration of the R-flavanone which the enzyme does not rapidly equilibrate with the chalcone. The sponta- neous reaction results in the equilibration of the chalcone with both isomers of the flavanone. The equilibrium constant determined here differs by a factor of 5 from a previously reported value of 37, which was estimated indirectly using the Haldane relationship and the V,,, and K , for the forward and reverse reaction (18).

Circular dichroism measurement of the flavanone product produced by partially purified enzyme from Phaseolus aureus indicated that (-)-(2S)-4’,7-dihydroxyflavanone was pro- duced with an optical purity of only 57 to 74% (24). Since enzyme-catalyzed reactions are expected to be stereospecific (32), we wanted to determine if the reaction catalyzed by chalcone isomerase is stereospecific or if indeed it is only stereoselective.

Treatment of 2’,4’,4-trihydroxychalcone (40 PM) with chal- cone isomerase (2.3 IU/ml, 260 nM) resulted in a loss of 88% of the chalcone absorbance in less than 15 s. Approximately 50% of the remaining chalcone absorbance was lost over several days. The initial rapid phase represents the equilibra- tion of chalcone with the S-isomer, while the slow phase represents the equilibration of the chalcone with the R-isomer. A rate constant for the loss of this additional chalcone absor- bance of 0.0017 -+ 0.0003 min” was determined from initial rate studies from 1 to 30 min after addition of enzyme. This rate constant is only 6-fold larger than would be expected for the spontaneous conversion of chalcone into the R-flavanone (Table 111). To determine if this modest rate enhancement represents true enzymic catalysis, the enzyme was inactivated after the 1st rapid phase by addition ofp-hydroxymercuriben- zoate (1 PM), which completely inactivates the enzyme by titration of an active site ~ulfhydryl.~ The observed rate constant, 0.0013 f 0.0003 rnin”, is only slightly lower and still higher than the spontaneous rate constant, suggesting the bulk of the 6-fold increase is due to nonactive site catalysis or that the reaction to produce the other isomer is not affected by the sulfhydryl modification. The rate constant due to active site catalysis, 0.0004 k 0.0006 min”, is within the experimen- tal uncertainty not significantly different from 0. An estimate4 of the upper limit of the kCat/Km for conversion of the chalcone to the R-flavanone is 10,000 M” min”, indicates that the enzyme produces the S-flavanone at least 100,000-fold faster than the R-isomer. The product of the enzymatic reaction ought to be greater than 99.999% the S-flavanone isomer.

Rate Acceleration Produced by Chalcone Isomerase-A ma- jor question in enzymology involves the quantitation of the factors responsible for the rate acceleration brought about by enzymes. It is clear that a large contribution to the rate acceleration can result from the enzyme bringing the sub- strates together into close proximity for the reaction to occur (33-35). The enzyme utilizes substrate-binding energy to act as an “entropy trap” and formally converts a bimolecular

R. A. Bednar, unpublished experiments. In calculating the concentration of free enzyme available for

reaction, the concentration of total enzyme was reduced by the concentration of enzyme-S-flavanone complex which was estimated using a K, of 18 p M for 4’,7-dihydroxyflavanone (18).

9586 Purification and Characterization of Chalcone Isomerase

process into a unimolecular one (33, 34). Page and Jencks (33, 35) have pointed out that the loss of translational and (overall) rotational entropy on conversion of a bimolecular process into a unimolecular process can lead to effective molarities of 10’ M. Further, the calculation of Page and Jencks and the work of Illuminati and Mandolini (36, 37) indicate that the loss of entropy upon freezing out a single internal rotation of a hydrocarbon chain results in a rate increase of only a factor of 5. Since chalcone isomerase catalyzes an intramolecular addition-elimination reaction, the apparent rate increase expected from entropic contributions should be small, thereby allowing us to better assess other factors (such as general acid-base catalysis (24), nucleophilic catalysis (231, or electrophilic catalysis) which may contribute to the rate acceleration.

The availability of homogeneous enzyme allows us to ac- curately measure the enzyme concentration used in the assays and therefore to calculate a rate constant, kc,, for the reaction of enzyme-bound substrate to be 11,000 min-’ (Table 111). The rate constant for the reaction of free enzyme and sub- strate, kCat/K,,,, is 1.1 X los M” min-’ and is approaching the diffusion-controlled limit (38).

A calculation of the apparent rate acceleration afforded by the enzyme can be obtained by comparing the rate constant for the spontaneous conversion of 2’,4’,4-trihydroxychalcone to the (2S)-4’,7-dihydroxyflavanone (0.5 kt, Table 111) with a rate constant for the enzyme-catalyzed reaction. The ratio of the rate constant for the reaction of free enzyme with the chalcone (kat/K,,,) to the rate constant for the spontaneous reaction is 3.7 x 10” M-’. The significance of this ratio is difficult to interpret. Formally, it indicates that 0.27 pM enzyme will yield a 100% increase in the observed rate at low substrate concentrations. A more meaningful measure of the acceleration due to enzymic catalysis is obtained by using the rate constant for reaction of the enzyme-bound chalcone ( L t )

which has the same dimensions as the spontaneous reaction. Enzyme-bound chalcone is converted into 5’-flavanone 36 million-fold faster than when it is free in solution (Table 111). This is a sizable rate increase considering that this is an intramolecular reaction and little of the observed catalysis will result from entropic factors.

CONCLUSION

The purification strategy detailed in this paper allows the preparation of chalcone isomerase from soybeans in extremely high purity, good recovery, and the quantities necessary for detailed studies of the structure and function of this important and interesting enzyme. We have determined several of the physical properties of the pure protein: amino acid composi- tion, molecular weight (native and denatured), isoelectric point, partial specific volume, carbohydrate analysis, and extinction coefficient. In addition, we have shown that, using pure chalcone isomerase, there is very strong stereoselectivity for production of the (2s)-flavanone, a fact which was not discernible when studying the previously available, impure enzyme. The physical properties reported for the pure protein

as well as the kinetics, thermodynamics, and stereochemistry of both the spontaneous and enzyme-catalyzed reactions con- stitute essential information which will form the basis of any detailed understanding of the mechanism for catalysis by chalcone isomerase. The mechanism of the spontaneous and the enzyme-catalyzed reaction, as well as the factors respon- sible for catalysis, are currently under study.

Acknowledgments-We are grateful to Michael Hartman who per- formed the amino acid analysis and to Bethany Seguerra who deter- mined the K, values for naringenin. We thank Dr. and Mrs. Pilkis for use of their HPLC system for amino acid analysis. We acknowl- edge Drs. Suleiman Bahouth, Thomas Daly, Robert Ehrlich, Steve Rokita, Paul Singer, and William Windsor for their critical review of this manuscript.

REFERENCES

2. Moustafa, E. & Wong, E. (1967) Phytochemistry 6,625-632 1. Wong, E. & Moustafa, E. (1966) Tetrahedron Lett. 26,3021-3022

3. Ebel, J. & Hahlbrock, K. (1982) in The Flavonoids: Advances in Research (Harborne, J. B. & Marbry, T. J., e&) pp. 641-670, Chapman & Hall, London

4. Hahlbrock, K., Wong, E., Schill, L. & Grisebach, H. (1970) Phytochemistry

5. Wlermann, R. (1972) Planta 102,55-60 9,949-958

6. Kuhn, B., Forkmann, G. & Seyffert, W. (1978) Planta 138,199-203 7. Forkrnann, G. & Dangelmayr, B. (1980) Biochem. Genet. 18,519-527 8. Dixon, R. A,, Dey, P. M. &Whitehead, 1. M. (1982) Biochim. Biophys. Acta

9. Dixon, R. A., Dey, P. M. & Lamb, C. J. (1983) Adv. Enzymol. Relat. Areas

10. Goodwin, T. W. & Mercer, E. I. (1982) Introduction to Plunt Biochemistry,

11. Harborne, J. B. (1976) in The Chemistry and Biochemistry of Plant Pig- p. 541, Pergamon Press, Elmsford, NY

12. Hagmann, M. & Grisebach, H. (1984) FEBS Lett. 176 , 199-202 ments (Goodwin, T. W., ed) p. 769, Academic Press, Orlando, FL

13. Dixon, R. A. (1986) Bid. Reu. 61, 239-291 14. Bailey, J. A. & Mansfield, J. W. (1982) Phytoalexins, John Wiley and Sons,

15. Robbins, M. P. & Dixon, R. A. (1984) Eur. J. Biochem. 145,195-202 16. Dixon, R. A,, Gerrish, C., Lamb, C. J. & Robbins, M. P. (1983) Plunta 159,

715,25-33

Mol. Biol. 66, l -121

New York

561-569 17. Mehdy, M. C. & Lamb, C. J. (1987) EMBOJ. 6.1527-1533 18. Boland, M. J. & Wong, E. (1975) Eur. J. Biochem. SO, 383-389 19. Metzler, D. E. (1977) Biochemistry: The Chemical Reactions of Living Cells,

20. Furlong, J. J. P. & Nudelman, N. S. (1985) J. Chem. Soe. Perkin Trans. I I

21. Old, K. B. & Main, L. (1982) J. Chem. Soe. Perkin Trans. I I 1309-1312 22. Furlong, J. J . P., Ferretti, F. H., Pappano, N. B., Debattista, N. B.,

23. Boland, M. J. & Wong, E. (1979) Bioorg. Chem. 8,1-8 24. Hahlbrock, K., Zilg, H. & Grisebacb, H. (1970) Eur. J. Biochem. 15 , 13-

25. Kreuzaler, F. & Hahlbrock, K. (1975) Eur. J. Biochem. 66,205-213 26. van Weely, S., Bleumer, A., S p ~ y t , R. & Schram, A. W. (1983) Planta

27. Lebninger, A. L. (1975) Biochemistry, pp. 101-102, Worth Publishers, New

28. Creighton, T. E. (1984) Proteins: Structures and Molecular Properties, pp.

29. Klapper, M. H. (1977) Biochem. Biophys. Res. Cornmun. 78,1018-1024 30. Lee, J. C. & Timasheff S. N. (1979) Methods Enzymol. 61,49-57 31. Arakawa, H. & Nakaziki, M. (1960) Chem. Ind. (Lond.) 73 32. Rose, I. A. (1972) Crit. Rev. Biochem. 1,33-57 33. Page, M. I. & Jencks, W. P. (1971) Proc. Natl. Acad. Sci. U. S. A. 68,1678-

34. Jencks, W. P. (1975) Adv. Enzymol. Relat. Areas Mol. BioL 43,220-410 35. Page, M. I. & Jencks, W. P. (1987) Gazz. Chim. Ital. 117,455-460 36. Illuminati, G. & Mandolini, L. (1981) Aects. Chem. Res. 14,95-102 37. Mandolini, L. (1986) Adv. Phys. Org. Chem. 22,l-111 38. Fersht, A. (1985) Enzyme Structure and Mechanism, p. 150, W. H. Freeman

p. 399, Academic Press, Orlando, FL

633-639

Borkowski, E. J. & Kavka, J. (1985) An. Quim. 81,199-204

18

169,226-230

York

7,29, W. H. Freeman and Co., New York

1683

39. Hirs, C. H. W. (1967) Methods Enzymol. 11,59-62 40. Riddles, P. W., Blakeley, R. L. & Zerner, B. (1983) Methods EMYWL. 91,

& Co., New York

AQ-An 41. EdiiboEh, H. (1967) Biochemistry 6,1948-1954 Additional references are found on p. 9588.

Purification and Characterization of Chalcone Isomerase 9587

EXPERIMENTAL PROCEDURES

TABLE I

Scheme lor Pvrlflclilon a l Chnlcone Iromrr*l< irom 10 kg of Soybeans (-)a

PURIFICATION SPECIFIC PURIFICATION TOTAL STEP ACTIVITY

TOTAL ENZYME PROTEIN ACTIVITY RECOVERY

1Ui.l fold 'D IU c Crude Extract 0 030 I O 760.000 23.W 100%

Acmd Treatment 0.044 I 5 360.000 16.000 70%

IN11&SO( PPI 0.14 4 1 I10.000 15.000 65%

DEAE-Cellulole 3 I 1 0 0 1,090 9.600 42%

Orange A 61 2 . m 94 JSOO 24%

Sephacryl S-200 340 I I.000 I5 1.100 22%

Purification and Characterization of Chalcone Isomerase

-68 -45

-36

-14.3

200

? D

E x:\

MINIPRINT REFERENCES

42. Gcirrman. T.A. 4 Clinton. R.O. (1946) 68. 691.

43. Wilson. J.M. 4 Wong. E. (1976) Phvlmhemislrv 1% 1313.1341.

44. Scopes. R.K. (1974) AmlAc&m 59. 277-282.

4s. SEOPCS. R.K. (1982) in I ' ' Springer-Verlsy New York. 0. 242.

46. Layne. E. (1917) 3. 447-4s4.

47. Bradford. M.M. (1976) -72. 248-214.

48. Tmford. G . 4 Robem. G.L. (19121 74. 2SW-2sIs.

49. Laemmli. U.K. (1910) Naj!m 227. 680-681.

SO. Dubny, G. 4 Bezard. G . (1982) Anal. Bimhem. 119. 321-329.

S I . Fairbmkr. G.. Sleck. T.L. & Wallach. D.F.H. (I9711 Bimhcmiftw 10. 2606-2616.

12 George. ST.. Ruobo. A.E. 4 Mslbon. C.C (1986) L E M L ~ L 261. 16S59-16564.

S3. Jenskr. W.P. (1969) W w i r in Chemirlrv and Enrvmdom. MsGnw-Hill. New York. p. S62464.

14. Mwre. J.W. 4 Pearson. R.G. (1981) -. 3rd ed.. John Wiley 4 Sons. New York. 0. 70-12.

IS. Hobey. W.D.. Shen. W.-H. 4 GOuin. D. A. (1986) 18. 97-10).

56. Schreiner. W.. KnmEr. M , Krifcher. S. & Langsrm. Y. (1985) May. 110-111.

SI. Lowe. C.R. 4 Pearson. J.C. (1984) &W&Bum& 104. 91-113.

IS. FYIIOII. 5. (1980) in D v e - L i n l n d . Amicon Corpanlion. Lexington. M W . D. C13. C25. C31. C33.

S9. hk ley, B R.. Kiruh. D.R. 4 Morri*, N.R. (1980) &dA&vm 105. 361-363.

60. Manhall. R.D. & Neuberger A. (1912) in Glycmmlp(c ins: Their Csmwrifion.Slruclure miEuajm(Go1uchalk. A. cd.) part B. P I . 141-142.