1,2-dithiole-3-thione (Oltipraz) - ACS Publications - American

1000 Hilltop Circle, Baltimore, Maryland 21250, and Department of Environmental Health. Sciences, Johns Hopkins University School of Public Health, 15...
0 downloads 0 Views 71KB Size
AUGUST 2001 VOLUME 14, NUMBER 8 © Copyright 2001 by the American Chemical Society

Communications Kinetic Constraints for the Thiolysis of 4-Methyl-5-(pyrazin-2-yl)-1,2-dithiole-3-thione (Oltipraz) and Related Dithiole-3-thiones in Aqueous Solution Kieran A. Carey,† Thomas W. Kensler,‡ and James C. Fishbein*,† Department of Chemistry and Biochemistry, University of Maryland Baltimore County, 1000 Hilltop Circle, Baltimore, Maryland 21250, and Department of Environmental Health Sciences, Johns Hopkins University School of Public Health, 15 North Wolfe Street, Baltimore, Maryland 21205 Received February 15, 2001

This report summarizes an investigation of the reactions of biological and other thiols with the cancer chemopreventive oltipraz and other dithiolethiones. Analysis of the kinetics of reaction of 4-methyl-5-(pyrazin-2-yl)-1,2-dithiole-3-thione (oltipraz) 1 with monothiols and dithiols in the range of 0.75-20 mM in aqueous 15% ethanol, at pH 7.5 (0.1 M Tris buffer) and at 37 °C has been undertaken. A plot of kobsd against [thiol] shows that reactions of monoand dithiols are first order in thiol concentration. The dependence on pH of these reactions shows that the active species is the thiolate anion. Specific second-order rate constants, k2 (M-1 s-1) for reaction of the thiolate anions with oltipraz have been determined to be cysteine, 0.040 ( 0.001; 2-mercaptoethanol, 2.0 ( 0.02; glutathione, 0.099 ( 0.001; mercaptoacetic acid anion, 4.0 ( 0.01; dithiothreitol, 1.33 ( 0.02; 1,3-propanedithiol, 10 ( 0.5; 1-mercaptopropane3-ol, 6.5 ( 0.1; 1-mercaptopropane-2,3-diol, 1.26 ( 0.05. A plot of pKa against log k2 for monothiols shows a linear dependence of k2 on pKa, βnuc 1.1 ( 0.07, which accounts for most of the reportedly enhanced reactivity of dithiols over monothiols. The pseudo-first-order rate constant for the solvolysis of oltipraz has been measured as 2.2 ((0.2) × 10-8 s-1. The kinetics of reaction of three other dithiole-3-thiones with glutathione has also been studied for comparison with oltipraz. The specific second-order rate constants, k2 (M-1 s-1) are 5-phenyl1,2-dithiole-3-thione, 4.7 × 10-4; 5-(4-methoxyphenyl)-1,2-dithiole-3-thione, 4.1 × 10-4; and 1,2-dithiole-3-thione 0.08. Important implications for the mode of biological action of these compounds and the nature of the putative biological targets of the compounds are discussed.

Introduction It has been known for decades that the carcinogenic potency of some chemicals can be mitigated by previous * To whom correspondence should be addressed. E-mail: jfishbei@ umbc.edu. † Department of Chemistry and Biochemistry. ‡ Department of Environmental Health Sciences.

or concurrent exposure to various natural products and drugs, commonly referred to as cancer chemopreventive agents (2). Some of these agents appear to act by elevating, mainly through transcriptional activation, the levels of phase 2 detoxication enzymes such as glutathione-S-transferases, quinone reductase, epoxide hydrolases, and glucuronyl transferases, among others. In

10.1021/tx0100340 CCC: $20.00 © 2001 American Chemical Society Published on Web 07/04/2001

940

Chem. Res. Toxicol., Vol. 14, No. 8, 2001

1988, the Talalay group observed that a common functionality among a number of such phase 2 enzyme inducers was the presence of a “Michael acceptor”san electron deficient carbon-carbon double bond (3). Subsequently, a much larger group of phase 2 enzyme inducers was recognized including, as well as classical Michael acceptors, hydroquinones, isothiocyanates, peroxides, heavy-metal containing salts, and dithiolethiones (4). In most cases, transcriptional activation by these compounds is mediated by a 5′-upstream 41-bp enhancer termed the antioxidant response element (ARE) (5). A common chemical property of all these compounds was notedsthe ability to react either covalently or through redox reactions with thiols. It was hypothesized that such agents might act to enhance the levels of phase 2 enzymes through interaction with thiols, possibly a thiol-containing protein mediating transcription. The Keap 1/Nrf2 complex represents a candidate for such control (6, 7). The transcription factor Nrf2, which binds to the ARE, appears to be essential for the induction for many phase 2 genes (5). Nrf2 is chiefly a cytosolic protein that is bound to the cysteine-rich chaperone Keap 1. This complex is disrupted by certain sulfhydryl-reactive inducers with subsequent localization of Nrf2 in the nucleus. Keap 1 thus represents a putative target for thiol mediated transcriptional activation by phase 2 enzyme inducers. In summary, there is considerable circumstantial evidence for the importance of thiols as mediators of the action of phase 2 enzyme inducers. The nature of the chemical processes by which thiols might mediate the action of phase 2 enzyme inducers is thus presently of considerable interest. While this communication was under review, a report appeared that concluded that there was a correlation between reactivity toward thiols and phase 2 enzyme inducing activity of some aryl vinyl ketones (Michael acceptors) (8). Hence, compounds 1, hydroxylated, in either ring, in the ortho

position exhibit, compared to para- and unsubstituted compounds, both enhanced chemical reactivity toward thiols, such as glutathione and dithiothreitol, and enhanced biological activity with respect to induction of quinone reductase and glutathione transferase in cultured cells. Such correlations are consistent with concepts summarized above. However, the details of chemical mechanism remain uncertain, even in this case (vide infra). These details bear on the issue of generality regarding the mechanisms by which other inducers act, and this question of generality impacts on thinking with respect to development of other clinically useful cancer chemopreventive agents. Thus, the chemistry of the reactions of thiols with phase 2 enzyme inducers is of ongoing importance. Presently, we are interested in the thiolytic chemistry of cyclic 1,2-dithiole-3-thiones, a subclass of phase 2 enzyme inducers. One such compound, oltipraz, 4-methyl5-(pyrazin-2-yl)-1,2-dithiole-3-thione (2), is currently in Phase II clinical trials as a chemopreventive in individuals at high risk for aflatoxin-induced hepatocellular carcinoma (9-12).

Communications

Dithiolethiones are a subclass of the phase 2 enzyme inducing chemicals described above. Similarly, the phase 2 enzyme inducing ability of oltipraz and other dithiolethiones is mediated, at least in part, through the ARE (13), and targeted disruption of the Nrf2 gene leads to lower constitutive expression of phase 2 genes and near complete loss of their inducibility by 1,2-dithiole-3-thiones (5). Moreover, intrinsic sensitivity to carcinogenesis is increased, and the chemopreventive efficacy of oltipraz is lost, in nrf2-deficient mice (14). The chemical basis by which dithiolethiones stimulate phase 2 enzyme induction is unclear. Reactions of dithiolethiones with biological (possibly protein) thiols have been suggested to be important in initiating the phase 2 enzyme induction, based on the generality from other phase 2 enzyme inducers, above, and the following additional observations. (a) Disulfides in general are chemically unreactive in physiological media except to thiols, and the disulfide pharmacophore of dithiolethiones, which is essential for inducer activity (10, 15), is thus suggestive of a thiol target. (b) Investigations into the nature of the antischistosomal activity of oltipraz revealed it to be a covalent inhibitor of parasite glutathione reductase. Inhibition involves the formation of a mixed disulfide that could be cleaved by addition of thiols (16). (c) Oltipraz has been purported to be much more reactive toward dithiols than monothiols, possibly implicating a dithiol target similar to the redox sensitive dithiol-containing transcription factors in bacterial systems (OxyR and SoxR) (17, 18) (15). These observations led to the suggestion that the efficacy of dithiolethiones might lie in their ability to bypass the glutathione pool (1-10 mM in many mammalian cells) thus preferentially targeting a protein dithiol which could stimulate reduction to form a disulfide and a presumably electrophilic Michael acceptor, A, as in the upper pathway of Scheme 1 (15). (The unreactivity of monothiols, according to this mechanism, could be accounted for by the reversibility of the first step and the intermolecularity of the second step, kbm, in the lower pathway of Scheme 1, compared to the intramolecularity of the second step, kbd, in the upper pathway of Scheme 1.) Either the original dithiol-containing protein could be the ultimate target, the transformation above giving rise to a protein disulfide bridge, or it is conceivable that the Michael acceptor, A, so-formed could react with the ultimate target. Except for the work of Fleury in which some reaction products in alkaline ethanol were extensively characterized, there has been relatively little in the way of aqueous solution chemistry reported for this important class of compounds (16, 19-21). An alternative reaction of thiols with dithiolethiones has recently been elucidated in which thiols, oxygen, metal ions and dithiolethiones react to give rise to DNA cleaving speciesspresumably reactive oxygen species (22). Such reactive oxygen species are known to effect the activation of certain transcription factors (18, 23, 24). The mechanism of Scheme 1 is easily tested directly by analysis of the kinetics of the reaction. We further

Communications

Chem. Res. Toxicol., Vol. 14, No. 8, 2001 941

Scheme 1. Proposed Mechanism for Reaction of Oltipraz with thiols (lower) and dithiols (upper) (qua ref 15)

sought to quantify the purported advantage of dithiols over monothiols in reactions with dithiolethiones and delineate the structural requirements for these effects. We report here results of a study of the kinetics of reactions of thiols with some dithiolethiones, compounds 2, 3 [5-(4-methoxyphenyl)-1,2-dithiole-3-thione], 4 (5-

phenyl-1,2-dithiole-3-thione), and 5 (1,2-dithiole-3-thione). These results (a) rule out the mechanism of Scheme 1 in aqueous solutions; (b) delineate important considerations in the nature of putative target biological thiols; (c) detail surprising comparisons regarding chemical versus biological reactivity that are in contrast with what has been most recently reported in the case of vinyl ketone inducers.

Experimental Section Materials. Oltipraz and 1,2-dithiole-3-thione were provided by the Chemoprevention Branch, National Cancer Institute, Rockville, MD, and 5-phenyl-1,2-dithiole-3-thione and 5-(4methoxyphenyl)-1,2-dithiole-3-thione were gifts of Latima, Paris, France. Other reagents were used as obtained from commercial sources and were typically ACS grade or better. Kinetic Methods. Generally, stock solutions were purged with argon prior to making up reaction solutions. Some reactions with oltipraz were run without argon purging (discussed further in the Results). Reactions were initiated by injection of an ethanolic dithiolethione-containing solution, into a temperature equilibrated (37.0 °C) vial containing aqueous reaction solution with at least a 10-fold molar excess of thiol over dithiolethione. Final reaction solutions were 15 vol % ethanol, typically pH 7.5 using Tris buffer (0.1 M, 20% base form) and contained EDTA (1 mM). The disappearance of the dithiolethione was monitored by HPLC using a Waters 715 ultra Wisp Chromatograph fitted with Waters 501 pumps, a Waters 486 variable wavelength

detector and a Phenomenex Luna 5 µm C18 (2) column, 25 × 4.6 cm. The concentration of thiol in each sample was assayed at the beginning and end of each kinetic run using a literature procedure (25, 26) and was found to change by less than 10%. The values of kobsd were determined by the method of initial rates or by fitting to a single-exponential decay with zero intercept. Reactions of glutathione with 5-phenyl-1,2-dithiole-3-thione and 5-(4-methoxyphenyl)-1,2-dithiole-3-thione proved exceedingly sluggish resulting in extensive thiol oxidation even with solutions initially purged with argon. Ultimately, oxidation was obviated ( 0.995) indicating that the reactions are kinetically first order in total thiol. Similarly, all plots of kobsd against thiol concentration for reactions with oltipraz and other dithiolethiones that contained a minimum of 4 points were linear (r2 > 0.995). At a given total thiol concentration, the value of kobsd increased proportionately with the thiolate concentration. In the case of the reaction of glutathione with oltipraz, a plot (containing six points) of kobsd against percent thiolate, at constant total glutathione concentration is linear with an intercept that is within experimental error equal to zero, thus indicating that the observed reaction is a function of the thiol anion concentration.1 Similarly, a two point plot for the reaction of 1,2-dithiole-3-thione also exhibits a y-intercept that is within experimental error zero and indicates that a 3.4-fold increase in percent thiolate yields a 3.5-fold increase in kobsd. These observations require the rate law of eq 1,

kobsd ) ko + k2[thiolate]

(1)

in which ko represents the thiol-independent rate constant and k2 is the specific second-order rate constant for reaction of thiolate ion. The values of ko for oltipraz, 1,2dithiole-3-thione, 5-phenyl-1,2-dithiole-3-thione, and 5-(4methoxyphenyl)-1,2-dithiole-3-thione were determined by initial rate methods under a single set of conditions (pH 7.5, 0.1 M Tris buffer, 0.001 M EDTA) to be ko ) 2.2 × 10-8, 1.46 × 10-6,