On the Reduction of Dithiolethiones and Dithiolylium Ions by NADPH and Glutathione Reductase

Archives of Biochemistry and Biophysics Vol. 382, No. 2, October 15, pp. 189 –194, 2000 doi:10.1006/abbi.2000.2022, available online at http://www.idealibrary.com on

On the Reduction of Dithiolethiones and Dithiolylium Ions by NADPH and Glutathione Reductase Be´ne´dicte Levron, Gwe´nola Burgot, and Jean-Louis Burgot 1 Laboratoire de Chimie Analytique, Faculte´ de Pharmacie, 2 avenue du Pr Le´on Bernard, 35043 Rennes Cedex, France

Received February 25, 2000, and in revised form July 18, 2000

Results of in vitro experiments carried out in water at 25°C and at pH 7.56 proved that NADPH in the presence of yeast glutathione reductase did not react with 1,2-dithiole-3-thiones and 1,2-dithiole-3-ones. On the other hand, 3-methylthiodithiolylium ions did react in these conditions. The reaction was identified and methyl 3-mercaptopropenedithioate resulting from a two-electron reduction process was obtained. A kinetic scheme consisting in a biordered mechanism has been found (K m ⴝ 2.6 10 ⴚ5 mol 䡠 l ⴚ1). All these results raise the question of a possible in vivo methylation (or alkylation) of dithiolethiones occurring prior to any other reductive biochemical process they may undergo. They also raise the question of the very existence (or in any case the generalization) of a reductive metabolism of dithiolethiones. © 2000 Academic Press Key Words: dithiolethiones; GSH; dithiolylium ion; antioxidant activity; cancer chemoprevention.

1,2-dithiole-3-thiones (dithiolethiones 1)

offer considerable promise as cancer chemopreventive and antioxidant drugs (1 and references therein). De1 To whom correspondence and reprint requests should be addressed. Fax: 02 99 33 68 88. E-mail: jean-louis.burgot@univ. rennes1.fr.

0003-9861/00 $35.00 Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.

rivatives of 1 that have been the most studied in this respect are Oltipraz (35972 RP) (R 4 ⫽ methyl, R 5 ⫽ 2-pyrazinyl), originally developed for its activity against schistosoma mansoni infections (2), and anetholtrithion (Sulfarlem) (R 4 ⫽ H; R 5 ⫽ p-methoxyphenyl), which has been marketed for 50 years first for its choleretic properties (3) and later for stimulation of salivary secretion (4). The chemopreventive activity of some other dithiolethiones has, however, also received attention and preliminary qualitative and quantitative structure–activity relationships have been published (1, 5, 6). The mechanisms of chemoprevention by dithiolethiones are not fully understood (1). So far, the activity of dithiolethiones has been mainly investigated from the standpoint of an enhancement or a loss of some enzymatic activities. Several studies, indeed, have evidenced the fact that these compounds are potent inducers of phase-II detoxication enzymes (7). Their action on cytochrome P 450 levels and on other phase-I enzymes is more subtle. Oltipraz enhances moderately phase-I enzyme activities (8) but appears to be, in a first stage, a transient inhibitor of cytochrome P 450 (9). The schistosomicidal activity of Oltipraz is correlated, according to Moreau et al. (10), with a loss of the glutathione reductase activity in schistosoma mansoni, which can likely be ascribed to a metabolite rather than to Oltipraz itself. Surprisingly, the simple hypothesis of a reduction of the disulfide bridge of dithiolethiones 1 by NADPH in the presence of glutathione reductase (GR) in water to give the corresponding 3-mercaptopropenedithioic acid (or its salts according to the pH value of the medium) has not been checked:

189

190

LEVRON, BURGOT, AND BURGOT

SCHEME 1

It is true that the hypothesis of such a reaction becomes less evident to formulate when one takes into account the fact that the disulfide bridge of these derivatives is embodied in the dithiole ring, which exhibits some aromatic character (11). This implies that the bond between the sulfur atoms 1 and 2 exhibits some double bond character, which means that the disulfide bridge of dithiolethiones and dithiolones is not a typical one. Nevertheless, the 1,2 bond of these derivatives can be opened by different purely chemical means (12 and references therein). To have a complete view of the different aspects of the cancer chemopreventive and antioxidant activities of dithiolethiones, we investigated systematically the different mechanisms that could explain their effectiveness. We present here the preliminary results of a study on the reactivity of some dithiolethiones, one dithiolone and one dithiolylium ion with NADPH in aqueous solution at 25°C at near neutral pH values in the presence of yeast glutathione reductase. The studied derivatives are the parent compound 1a (R 4 ⫽ R 5 ⫽ H) and the [5-methyl-1,2-dithiole-3-thione-4]-ylethanoic acid 1c (R 5 ⫽ CH 3 , R 4 ⫽ CH 2 COOH), the parent dithiolone 2a (R 5 ⫽ R 4 ⫽ H) and the 3-methylthio-5-phenyl-1,2-dithiolylium ion sulfomethylate 3b:

SCHEME 2

MATERIALS AND METHODS Derivatives 1a, 1c, and 2a were chosen for their solubility in water. 1a is the less lipophilic (unfunctionalized) dithiolethione [log P (water/n-octanol) ⫽ 1.58] (13). Its aqueous solubility (⬇ 2 䡠 10 ⫺4 mol l ⫺1) is sufficient to follow its concentration decrease in the course of a chemical or biochemical reaction by UV-visible spectrophotometry. This is also the case with dithiolone 2a. 1c was an original dithiolethione purposefully designed to be perfectly soluble in water at near neutral pH values by formation of the carboxylate ion owing to its pK a value: 4.14. It can be inferred from some physicochemical studies (13) that the dithiolethione nucleus in 1c very likely retains its typical reactivity of 1,2-dithiole-3-thione since the methylene frag-

ment in position 4 precludes the strong interaction between the dithiole nucleus and the carboxylic group, which has already been evidenced (14). (The synthesis and the analytical data of 1c will be given elsewhere). The choice of studying a dithiolone was dictated by the fact that the metabolites of the two dithiolethiones (Oltipraz (15) and Anetholtrithion (16)) are known, and the only metabolite of the same kind found in both derivatives was the corresponding dithiolone. 2a was prepared according to (17). The choice of studying a 3-methylthiodithiolylium ion was suggested both by the good water solubility of these derivatives and by their possible formation in vivo. To our knowledge, only the dithiolylium ion of Oltipraz has been studied from a biochemical and pharmacological standpoint (10). In this preliminary work, attention was paid first to ion 3b with a 5-phenyl substituent assumed to be a good representative of these derivatives from a pharmacochemical standpoint. This is probably not the case of 3-methylthiodithiolylium ions resulting from Anetholtrithion and Oltipraz, which seem to be somewhat atypical dithiolethiones as the structures of their metabolites suggest (see discussion below). 3b was prepared according to (18). ● The obtained compound 5 (methyl 3-mercapto-3-phenylpropene dithioate) is of noticeable instability. It was synthetized by action of sodium borohydride on dithiolylium ion 3b (the detailed preparation will be given elsewhere). Analytical data; 1H N.M.R. [CDCl 3, ␦ ppm/ TMS]:2,65 [s, 3 H]; 7.40 [m, 6 H]; 10.25 [1, 1 H]; Mass spectrum (electronic impact 70 eV): (M-H) ⫹; th: 224.9866, found: 224.9875: (M-CH 3) ⫹; th: 210.9710 found 210.9729. ● Yeast (Saccharomyces cerevisiae) glutathione reductase was a commercial product (Randox, Antrim, UK) presented under the form of a suspension for immediate use (99% purity and specific activity 197 U/mg). Reduced glutathione was a Sigma-Aldrich product (98% purity) and was used as such. NADPH tetrasodium salt (98% purity) came from Boehringer-Mannheim. The reactants were dissolved in phosphate buffer, pH 7.56 (at 70 ml M/15 Na 2HPO 4 (2H 2O) was added 30 ml KH 2PO 4 M/15). The pH value of the medium was systematically recorded before and after reactions. In this preliminary work no other pH value was chosen. Water was ultrapure grade (Elga Co.) and prepared just before use. ● Mass spectra were recorded at the Centre de mesures physiques de l’Ouest, Universite´ de Rennes I. UV-visible spectra were recorded with a Biotek-Kontron-Uvikon 930 or 922 apparatus in the following conditions: T ⫽ 25°C, quartz cell (l ⫽ 1 cm) and scan speed 1.000 nm min ⫺1. The fluorometric apparatus was a Perkin–Elmer LS 50B used with quartz cells (l ⫽ 1 cm), excitation slit: 10 nm, emission slit 5 nm, ␭ ex: 340 nm, ␭ em: 470 nm and scan speed 1.500 nm min ⫺1. ● The study of the enzymatic reaction by UV-visible spectrophotometry was carried out directly inside the cuvet of the spectrophotometer. The base line was matched with a solution prepared by mixing 800 ␮l of phosphate buffer (pH 7.56) and 300 ␮l of ultrapure water. The ratio between the volumes of these two solutions as well as the overall volume (1.1 ml) of the solution in the cell were kept systematically constant throughout the enzymatic studies. Glutathione reductase and NADPH were dissolved into the phosphate buffer and the studied compound into water. Finally added in the working cell: 500 ␮l of the glutathione reductase solution (0.817 U ml ⫺1) in buffer, 300 ␮l of the NADPH solution (concentration range 9.3 10 ⫺5 mol 䡠 l ⫺1 to 4.6 10 ⫺4 mol l ⫺1) in buffer, and 300 ␮l of the aqueous

REDUCTION OF DITHIOLETHIONES AND DITHIOLYLIUM solution (1.83 10 ⫺4 mol l ⫺1) of the studied product. The final concentration of the studied compound (dithiolethione, dithiolone, or dithiolylium ion) was 5 䡠 10 ⫺5 mol l ⫺1. After mixing, the absorbance was recorded in the 220 – 600 nm region every 5 min for 30 min, the necessary time for the reaction to be complete. The conditions chosen for the fluorometric study of the NADPH fate were: ␭ ex ⫽ 340 nm, ␭ em ⫽ 465 nm, determinations every 5 min for 30 min. The calibration curve was constructed from the results of six solutions (NADPH concentration range: 1.04 䡠 10 ⫺6 to 1.66 䡠 10 ⫺5 mol 䡠 l ⫺1) prepared in exactly the same conditions as those used for UV-visible spectrophotometry. The absence of fluorescence of other reactants and products had previously been checked. In this preliminary work 25°C was the only temperature used and no consideration of ionic strength was taken into account. ● The kinetic scheme (see Results) led, after application of the stationary state principle to the concentration of the intermediary ternary complex, to the mathematical model: dA k 2 E 0 S 0 ⑀ p2 ⫺ k 2 E 0 ⑀ p A ⫽ [1] dt 共K m ⫹ S o兲 ⑀ p ⫺ A Using this model was legitimate for ␭ ⬎ 450 nm, wavelengths at which the product P was the only one to absorb with the molar absorption coefficient ⑀ p . In relation [1], A was the measured absorbance, E o the initial enzyme concentration, and S o the total concentration of the dithiolylium ion. Experimental absorbances

191

results were used to obtain with relation [1] the ⑀ p and K m values by a nonlinear least-squares procedure used in our laboratory (19). The derivative dA/dt was approximated by the ratios of finite variations ⌬A/⌬t, where ⌬A was the absorbance difference for an interval time ⌬t ⫽ 17 s.

RESULTS AND DISCUSSION

Dithiolethiones 1a, 1c, and dithiolone 2a did not react in water at 25°C with NADPH in the presence of yeast glutathione reductase at pH 7.56. No variation with time of the UV-visible spectra was detected when these derivatives were mixed with NADPH and glutathione reductase. The spectra were only the invariable superimposition of the NADPH and the studied derivative spectra, whichever relative concentrations were used. From a methodological standpoint, this result could not be attributed to inappropriate experimental conditions since NADPH reacted normally in the same conditions with its natural substrate GSSG. On the other hand, dithiolylium ion 3b reacted in the same conditions. The reaction was identified to be:

SCHEME 3

This is a 2e-redox reaction which frequently happens between NADPH and a disulfide bridge (20, 21). Preliminary experiments performed with 10 ions 3 at our disposal indicated that the reaction mentioned in scheme [3] seems to be general. It is worth noting that dithiolylium ions 3a and 3c, corresponding to dithiolethiones 1a and 1c, which did not react, exhibited such a reduction process. The 3-methyldithiolylium ion derived from Oltipraz, however, gave a derivative exhibiting a UV-visible spectrum, which differed markedly from those obtained with other ions. These results deserve further studies. The assessment of the reactional scheme [3] was based on the following arguments: ● At time t ⫽ 0, the UV-visible spectrum in the wavelength range 280 –390 nm corresponded to the superimposition of the NADPH and 3b spectra. For ␭ ⬎ 390 nm, absorbance was only that of 3b. At times ⬎ 0 (Fig. 1) continuous evolution was noticed and for ␭ ⬎ 390 nm the spectrum obtained was the superimposition of the spectra of the remaining ion 3b and of the

product since NADPH and NADP ⫹ do not absorb in this region. The appearance of an isosbestic point at exactly 400 nm was a strong argument for the existence of only one product (aside NADP ⫹); ● The proof of structures 5 (and 5ⴕ) followed the synthesis in our laboratory of the product: the methyl 3-mercapto-3-phenyl propene dithioate 5 (see Materials and Methods). Despite the noticeable instability of 5, already mentioned in the literature for its only two known analogous derivatives (22), its RMN 1H, IR, and mass spectra could be recorded. They were in full agreement with the given structure (see Materials and Methods). Furthermore, 5 gave by methylation the already known compound 6 (22).

SCHEME 4

192

LEVRON, BURGOT, AND BURGOT k2

¡ NADP ⫹ E ⫹ P, NADPH-E-S O

FIG. 1. Evolution of the UV-visible spectrum during reaction [3]—Wavelength range 320 – 620 nm, 25°C, NADPH in the presence of yeast glutathione reductase. Absorbances recorded every 3 min for 21 min (t 1 ⫽ 0; t 6 ⫽ 21). ● The synthetic compound 5 in the pH and concentration conditions of the enzymatic reaction gave the same UV spectrum as that obtained at the end of the enzymatic reaction. Moreover, reconstituted solutions of 5 and 3b in the same range of concentrations as that chosen for the enzymatic reaction gave rise to an isosbestic point at exactly 400 nm. (In scheme [3] the existence in the solution of base 5ⴕ must also be considered owing to the acidic character of the thiol group of 5, the pK a value of which is still unknown. The occurrence of the mixture 5 ⫹ 5ⴕ is not in contradiction with the existence of an isosbestic point because the mixture 5 ⫹ 5ⴕ can be assimilated to a unique product since the ratio of their concentrations is constant for a given pH value). ● The part played by glutathione reductase was evidenced by the fact that 3b and NADPH did not react without glutathione reductase in the same experimental conditions. Likewise, mixing the enzyme and 3b did not induce any reaction. The part played by NADPH in the reaction was also evidenced by spectrofluorometry. The emission peak given by NADPH at ␭ ⫽ 465 nm after excitation at 340 nm (conditions in which glutathione reductase and 3b do not fluoresce) fully disappeared during the reaction (Fig. 2). As a result, the enzyme reaction mentioned in scheme [3] was considered as quasicomplete.

The kinetics of the enzymatic reaction are consistent with the biordered mechanism (23)

where P is the product (5 ⫹ 5ⴕ) with the substrate S ⫽ 3b. NADPH-E-S is a very transient species with a lifetime negligible compared to that of the binary complex NADPH-E. In the above scheme the limiting reaction is the dissociation of the ternary complex. Handling experimentally determined absorbances according to the mathematical model corresponding to the kinetic scheme with the help of a non linear least squares procedure (see Materials and Methods) gave the values of K m, ⑀ p , and V max (Table I). The ⑀ p value ( ⑀ p ⫽ 18000 ⫾ 3000 at pH 7.56 and ␭ ⫽ 450 nm) obtained indirectly through this methodology applied to several experiments performed in different relative NADPH and 3b concentration conditions was in fair agreement with that ( ⑀ p ⫽ 16200) obtained by direct determinations performed with the synthetic compound 5 (Table I). It is worth noting that the methodology was applied to several experiments performed in

FIG. 2. Evolution of the fluorescence spectrum of NADPH at 465 nm (after excitation at 340 nm) with time (conditions in which glutathion reductase and compound 3b do not fluoresce):

fast

NADPH ⫹ E O ¡ NADPH-E k1

NADPH-E ⫹ S | L ; NADPH-E-S k ⫺1

Spectrum Spectrum Spectrum Spectrum Spectrum Spectrum

1 2 3 4 5 6

Time (min)

Fluorescence (arbitrary units)

5 10 15 20 25 30

83.1 48.6 24.3 7.6 1.7 1.7

193

REDUCTION OF DITHIOLETHIONES AND DITHIOLYLIUM TABLE I

⑀ p and K m Values Obtained through Mathematical Treatment of the Kinetic Results at pH 7.56 and ␭ ⫽ 450 nm for the Reaction of Dithiolylium 3b in Water at 25°C with NADPH in the Presence of Yeast Glutathione Reductase [NADPH] 10 ⫺5 䡠 M in cell

2.5

2.5

5.1

5.1

⑀ p dm ⫺3 䡠 mol ⫺1 䡠 cm ⫺1 16400 16400 18900 19500 K m 10 ⫺5 䡠 mol 䡠 l ⫺1 1.5 1.3 3.2 6.6 ⫺6 ⫺1 ⫺1 V max 10 䡠 mol 䡠 l min 19.2 19.2 16.8 16.2 ⑀ p ⫽ 18000 dm ⫺3 mol ⫺1 䡠 cm ⫺1 V max ⫽ 17.3 10 ⫺6 mol 䡠 l ⫺1 min ⫺1

different relative NADPH and 3b concentration conditions. The coherence of the results obtained through the kinetic determination with those obtained directly is a strong argument in favor of the accuracy of the kinetic model and of the results of the experiments. A K m value in the 10 ⫺3–10 ⫺5 mol 䡠 l ⫺1 range means that the substrate exhibits a medium affinity for the enzyme (24). The K m value corresponding to the system GSSG-glutathione reductase (range: 6.1 䡠 10 ⫺5 mol 䡠 l ⫺1– 6.5 䡠 10 ⫺5 mol 䡠 l ⫺1) (25, 26) is somewhat higher than that found in this work for the dithiolylium ion 3b (2.6 䡠 10 ⫺5 mol 䡠 l ⫺1). It is interesting to note that the 3b dithiolylium ion reacted in a purely chemical way with reduced glutathione at pH 7 to give a product the structure of which is still unknown but is not 5. A competition reaction between GSSG and 3b for NADPH in the presence of glutathione reductase was studied. The result was that GSSG was reduced by the enzymatic way and that the formed GSH reacted subsequently chemically with the dithiolylium ion. The striking result of this work is that 3-methylthiodithiolylium ions reacted in the above conditions while the corresponding dithiolethiones and dithiolones did not. It is very likely that it is also the case in vivo because dithiolethiones must be confined to the lipidic phases owing to their lipophilic character, which makes reaction of the scheme 1 even more difficult than in the in vitro conditions from a thermodynamic standpoint. The difference of reactivity between dithio-

5.1

12.7

19800 18200 1.1 1.3 15.6 17.4 K m ⫽ 2.6 䡠 10 ⫺5 mol 䡠 l ⫺1

12.7 18800 3.3 16.8

lylium ions and dithiolethiones must undoubtedly be accounted for by the enhanced positive charge brought by the dithiole nucleus of the ions 3. In this respect, an estimate of both total and ⌸ electronic charges brought by the atoms constituting the molecules of the dithiolethione 1b and of the dithiolylium ion 3b (Fig. 3) is impressive. The given values were calculated within the framework of the MNDO-PM3 method (MOPAC 6,0 program (27)), which provides satisfactory estimates of the dipolar moments of some dithiolethiones (28) (i.e., most probably, satisfactory estimates of their electronic charges and geometrical parameters). If one admits that the reduction of the ions results from the same mechanism as that admitted for glutathione (29) (i.e., a mechanism involving at the end of the catalytic cycle a nucleophilic attack of the disulfide bridge by the thiol function of the cysteine 58 of the enzyme protein followed by the attack of the intermediary asymmetrical disulfide by the thiolate function of cysteine 63), it becomes evident that, in the 1,2-dithiole series, the magnitude of the electrical charge brought by the disulfide bridge is determining. Dithiolylium ions are obviously good substrates for nucleophilic reactions owing to their full positive charge (Fig. 3). Michael’s reaction is one of them and it is very interesting to mention, at this point, that according to Talalay et al. (30) a common feature of phase-II enzyme inducers is that they are Michael-reaction acceptors. All these results lead unavoidably to the question of a necessary prior in vivo methylation or, more gener-

FIG. 3. Total and ⌸ electronic charges of dithiolethione 1b and of its corresponding 3b methylthiodithiolylium ion calculated within the framework of the MNDO-PM3 method (MOPAC 6.0 program).

194

LEVRON, BURGOT, AND BURGOT

ally, alkylation of dithiolethiones before they undergo a reductive metabolic pathway if this is actually their fate. It is worth noting that this hypothetical prior alkylation may be performed with the help of S-transferases and, once again, it must be remembered that most phase-II enzyme inducers are also substrates for glutathione-S-transferases (30). ACKNOWLEDGMENT We thank Mr. G. Bouer for his assistance.

REFERENCES 1. Mascuitenko, Y. Y., Libby, A. H., Howard Joyner, H., Curphey, T. J., MacMillan, D. L., Kensler, T. W., and Roebuck, B. D. (1998) Carcinogenesis 19, 1609 –1615. 2. Leroy, J. P., Barreau, M., Cotrel, C., Jeannart, C., Messer, M., and Benazet, F. (1978) Curr. Chemother. 1, 148 –150. 3. Halpern, B. N., and Gaudin, O. (1950) Arch. Int. Pharmacol. 83, 49 –79. 4. Ha¨usler, R., and Ritschard, J. (1979) Praxis 68, 1063–1068. 5. Egner, P. A., Kensler, T. W., Prestera, T., Talalay, P., Libby, A. H., Howard Joyner, H., and Curphey, T. J. (1994) Carcinogenesis 15, 177–181. 6. Burgot, G., Bona, M., Christen, M. O., and Burgot, J. L. (1996) Int. J. Pharm. 129, 295–299. 7. Ansher, S. S., Dolan, P., and Bueding, E. (1986) Food Chem. Toxicol. 24, 405– 415. 8. Dimitrov, N. V., Bennett, J. L., MacMillan, J., Perloff, M., Leece, C. M., and Malone, W. (1992) Invest. New Drug 10, 289 –298. 9. Morel, F., Langoue¨t, S., Mahe´o, K., and Guillouzo, A. (1997) Cell Biol. Toxicol. 13, 323–329. 10. Moreau, N., Martens, T., Fleury, M. B., and Leroy, J. P. (1990) Biochem. Pharmacol. 40, 1299 –1305. 11. Mayer, R., Hartmann, H., Fabianand, J., and Melhorn, A. (1967) Z. Chem. 6, 209 –216. 12. Pedersen, C. Th. (1995) Sulfur Reports 16, 173–221.

13. Bona, M., Boudeville, P., Zekri, O., Christen, M. O., and Burgot, J. L. (1995) J. Pharm. Sci. 84, 1107–1112. 14. Chollet, M., and Burgot, J. L. (1998) J. Chem. Soc., Perk. T. 2, 387–391. 15. Bieder, A., Decouvelaere, B., Gaillard, C., Depaire, H., Heusse, D., Ledoux, C., Lemar, M., Le Roy, J. P., Raynaud, L., Snozzi, C., and Gregoire, J. (1983) Arzneimittel-Forsch. 33, 1289 –1297. 16. Dansette, P. M., Sassi, A., Christen, M. O., and Mansuy, D. (1988) Communication at Workshop on Drug Metabolism. Konstanz, Germany. 17. Meinetsberger, E., Scho¨fer, A., and Behringer, H. (1977) Synthesis - Stuttgart 802– 803. 18. Bo¨ttcher, B. (1948) Chem. Ber. 81, 376 –377. 19. Zekri, O., Boudeville, P., Genay, P., Perly, B., Braquet, P., Jouenne, P., and Burgot, J-L. (1996) Anal. Chem. 68, 2598 –2604. 20. Kolthoff, I. M., Stricks, W., and Tanaka, N. (1955) J. Am. Chem. Soc. 77, 4739 – 4742. 21. Veine, D. M., Arscott, L. D., and Williams, C. H., Jr. (1998) Biochemistry 37, 15575–15582. 22. Maignan, J., and Vialle, J. (1973) Bull. Soc. Chim. Fr. 6, 1973– 1978. 23. Pelmont, J. (1995) Enzymes: Catalyseurs du Monde Vivant. Presses Univ. Grenoble. 24. Penasse, L. (1974) Les Enzymes: Cine´tique et Me´canisme d’Action. Masson, Paris. 25. Mavis, R. D., and Stellwagen, E. (1968) J. Biol. Chem. 243, 809 – 814. 26. Krauth-Siegel, R. L., Arscott, L. D., Schonleben-Jonas, A., Schirmer, R. H., and Williams, C. H., Jr. (1998) Biochemistry 37, 13968 –13977. 27. Quantum Chemistry Program Exchange (1990) No. 455, Indiana Univ. 28. Botrel, A., Illien, B., Rajczy, P., Ledoux, I., and Zyss, J. (1993) Theor. Chim. Acta 87, 175–194. 29. Pai, E. F., and Schulz, G. E. (1983) J. Biol. Chem. 258, 1752– 1757. 30. Talalay, P., De Long, M. J., and Prochaska, J. (1988) Proc. Natl. Acad. Sci. USA 85, 8261– 8265.