630
Chem. Res. Toxicol. 1998, 11, 630-638
Halide Effects in the Hydrolysis Reactions of (()-7β,8r-Dihydroxy-9r,10r-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene Bin Lin,† Lanxuan Doan,† Haruhiko Yagi,‡ Donald M. Jerina,‡ and Dale L. Whalen*,† Department of Chemistry and Biochemistry, University of Maryland, Baltimore County, Baltimore, Maryland 21250, and Laboratory of Bioorganic Chemistry, NIDDK, National Institutes of Health, Bethesda, Maryland 20892 Received September 23, 1997
Rates of reaction of (()-7β,8R-dihydroxy-9R,10R-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene (DE-2) have been determined in 1:9 dioxane-water solutions containing 1.0 M KCl, 0.5 M KBr, and 0.1 M NaI over the pH range 4-13. These pH-rate profiles are more complicated than those for reaction of DE-2 in 0.2 M NaClO4 solutions and are interpreted in part by mechanisms in which halide ion attacks the diol epoxide as a nucleophile at intermediate pH, resulting in the formation of a trans-halohydrin. Reaction of DE-2 in these halide solutions at pH < ca. 5 occurs by rate-limiting carbocation formation, followed by capture of the intermediate carbocation by halide ion. The relative magnitudes of the rate constants for reaction of the intermediate carbocation with halide ions are estimated from product studies. The halohydrins are unstable intermediates and react quickly in subsequent reactions to yield tetrols in a ratio different than that formed from reaction of the carbocation with solvent. Nucleophilic attacks of 1.0 M Cl-, 0.5 M Br-, and 0.1 M I- on DE-2 are the principal reactions in the pH range ca. 6-9, leading to intermediate trans-halohydrins that hydrolyze to tetrols. At pH ca. 9-11, halohydrin formed from attack of halide ion on DE-2 reverts back to epoxide, leading to a negative break in the pH-rate profile. The main product-forming reaction of DE-2 at pH 11.3 is the spontaneous reaction. At pH > 12, the rate of reaction of DE-2 increases due to a second-order reaction of HO- with DE-2.
Introduction The environmental carcinogen benzo[a]pyrene is metabolized in part to a mixture of diastereomeric, bayregion 7,8-diol 9,10-epoxides (1), one with the benzylic C-7 hydroxyl group cis to the epoxide group [(+)-DE-1,1 structure not shown] and the second with the benzylic hydroxyl group trans to the epoxide group [(+)-DE-2]. (+)-DE-2 possesses strong carcinogenic activity and is thought to be the metabolite responsible for the activity of the parent hydrocarbon (2).
The hydrolysis reactions of DE-1 and DE-2 have received considerable attention (3-7). At pH < ca. 7 in water, the hydrolysis of DE-2 is catalyzed by hydronium
ion and yields a mixture of tetrols that are the result of ca. 6% cis and 94% trans hydration.2 This reaction proceeds via an intermediate carbocation, and thus trans attack of solvent on the carbocation is favored over cis attack.2 At pH ca. 7-8, the reaction of DE-2 undergoes change from hydronium ion-catalyzed hydrolysis to spontaneous hydrolysis, which yields ca. 44% cis-tetrol and 56% trans-tetrol (5, 7). It has been reported that halide ions catalyze the formation of tetrol product resulting from cis hydration of DE-2 at pH 7 in 1:9 acetone-water (8). From the observation that tetrol product ratios from hydrolysis of DE-2 in solutions containing increasing chloride ion concentrations change significantly, although there is little accompanying change in reaction rate, it was concluded that chloride ion reacts primarily with an intermediate carbocation 3, formed in a rate-limiting step, to yield chlorohydrin 5 (Scheme 1). The stereochemistry of this chlorohydrin is reported to be trans (9). Subsequent reaction of 5 to yield more cis-tetrol than is formed in the absence of halide ion would then account for the increased cis/trans hydration ratios. Direct SN2 attack of iodide and bromide ions, but not chloride ion,
†
University of Maryland, Baltimore County. National Institutes of Health. Abbreviations: DE-1, 7β,8R-dihydroxy-9β,10β-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene; DE-2, 7β,8R-dihydroxy-9R,10R-epoxy-7,8,9, 10-tetrahydrobenzo[a]pyrene; HEPES, N-(2-hydroxyethyl)piperazineN′-2-ethanesulfonic acid; MOPSO, 3-(N-morpholino)-2-hydroxypropanesulfonic acid; CHES, 2-(N-cyclohexylamino)ethanesulfonic acid; HPLC, high-pressure liquid chromatography. ‡
1
2 The terms cis and trans are used to describe the relative stereochemistry of the C-8 and C-9 groups and also the stereochemistry of addition of solvent or external nucleophiles to the epoxide group of DE-2 such that the resulting hydroxyl group at C-9 and hydroxyl or other group at C-10 in the product have cis and trans stereochemistries, respectively.
S0893-228x(97)00174-4 CCC: $15.00 © 1998 American Chemical Society Published on Web 05/23/1998
Halide Effects in Diol Epoxide Hydrolysis Scheme 1
Chem. Res. Toxicol., Vol. 11, No. 6, 1998 631 Table 1. Summary of Hydrolysis Product Yields from Reaction of DE-2 in 0.2 M NaClO4, 1:9 Dioxane-Watera pHb
cis-tetrol:trans-tetrol
yield of tetrols (%)c
4.00 (4.08) 5.09 (5.73) 8.11 (8.15)d 10.01 (9.82) 11.18 (10.98) 12.0
6:94 6:94 44:56 42:58 36:64 37:63
100 100 86 72 57 53
a All reactions were allowed to proceed at room temperature for 15 h. b The number in parentheses is the pH of the reaction solution after 15 h. c The percent yield of tetrols from reaction of DE-2 at pH > 8 is relative to the yield of tetrols formed from the acid-catalyzed hydrolysis of DE-2 at pH < 5.5. d HEPES buffer (2 × 10-3 M) was used to maintain pH.
at high halide concentrations was indicated by increases in the rates of reaction of DE-2 in solutions containing these halide ions. The reactions of DE-2 with dAMP (9), poly(A) (10), and DNA (11) in solutions containing chloride ion also yield greater amounts of cis adducts than are formed in the absence of chloride ion. It was proposed that chloride ion reacts with the carbocation formed from reaction of DE-2 with H+ to yield a trans-chlorohydrin, followed by SN2 attack of DNA on the trans-chlorohydrin to yield cis adducts. The synthesis of the trans-chlorohydrin and its reaction with deoxyadenosine to give higher yields of cis adducts than are formed from reaction of DE-2 with deoxyadenosine in the absence of chloride ion have recently been reported (9). We have previously reported that rates and products of hydrolysis of indene oxide (12), 9,10-phenanthrene oxide (13), 7-methyl-substituted benz[a]anthracene 5,6oxides (14), styrene oxide, and several vinyl epoxides (15) are dependent on chloride ion concentrations. These results were interpreted in terms of a mechanism in which chloride ion acts as an efficient nucleophile, both toward reactant epoxides and toward intermediate carbocations. To avoid complications associated with the chemical reactivity of chloride ion, sodium perchlorate, a nonnucleophilic salt, was used instead of potassium chloride or sodium chloride to maintain ionic strength in a study of the hydrolysis reactions of DE-1 and DE-2 (5). We have also determined the rates of reaction of DE-2 in solutions containing halide salts as a function of the pH of the solution. Our interest in further characterizing the kinetic parameters and mechanisms of reaction of DE-2 in solutions containing halide salts was stimulated in part by the recent reports on the mechanisms of reaction of DE-2 in solutions containing halide ions, in both the absence and presence of DNA (8-11). We report here detailed rate-pH profiles and product studies for the reaction of DE-2 in solutions containing chloride, bromide, and iodide ions. These rate and product studies confirm reports (8-11) that in the reaction of DE-2 at sufficiently low pH, halide ions effectively capture a carbocation intermediate to yield halohydrins, which then react quickly to yield tetrol products. Rate constants for the reaction of the carbocation with chloride and bromide ions are estimated. Our
results also show that the reaction of DE-2 at pH > ca. 7 yields halohydrins by a different pathway involving the bimolecular attack of halide ion on DE-2. This secondorder reaction becomes the major reaction pathway for reaction of DE-2 at pH ca. 7-10 when [Cl-] > ca. 0.3 M, [Br-] > ca. 0.03 M, and [I-] > ca. 0.003 M. Mechanisms of product formation from reaction of DE-2 as functions of pH and halide ion concentrations are discussed.
Experimental Procedures Materials. (()-DE-2 was prepared by published procedures (3, 16). (Caution: This material is carcinogenic and should be handled with caution.) Dioxane was distilled from sodium prior to use. All other reagents were purchased from commercial sources. Kinetic Procedures. For each kinetic run, approximately 5 µL of a stock solution of DE-2 in dioxane (ca. 1 mg/mL) was added to 2.0 mL of reaction solution in the thermostated cell compartment (25.0 ( 0.2 °C) of a UV-vis spectrophotometer. Reactions were monitored at 348 nm, and pseudo-first-order rate constants were calculated by nonlinear regression analysis of the absorbance vs time data. For kinetic runs at intermediate pH, approximately 10-3 M MES, HEPES, MOPSO, or CHES buffer was used to maintain pH. Products from Reactions of DE-2 in 1:9 (v/v) DioxaneWater Solutions, 0.2 M NaClO4. Aliquots (10.0 µL) of DE-2 in dioxane (ca. 0.5 mg/mL) were added to vials containing 2.0 mL of 0.2 M NaClO4 in 1:9 dioxane-water (v/v) whose pH had been adjusted with either 0.2 M HClO4 in 1:9 dioxane-water or 0.2 M NaOH in 1:9 dioxane-water. After swirling, the vials were capped and allowed to stand at room temperature for 15 h (ca. 8 half-lives for the spontaneous reaction). An aliquot (10.0 µL) of 2-(1-naphthyl)ethanol in dioxane was then added to each vial to serve as an HPLC standard. The pH of each reaction solution was adjusted to ca. 5-8, and they were analyzed by HPLC on a reverse-phase C18 column with 60% methanol-40% water as eluting solvent. Products were monitored by UV detection at 254 nm. The retention times for the trans-tetrol, cis-tetrol, and 2-(1-naphthyl)ethanol standard are 8.7, 10.7, and 12.6 min, respectively. The results of these analyses are summarized in Table 1 and are consistent with earlier product studies of the hydrolysis of DE-2 in solutions containing only 0.1 M NaClO4, for which only tetrol ratios were determined (5). Relative yields of tetrol products were calculated by comparing the area of the tetrol HPLC peaks with that of the standard at each pH. The ratio of cis-tetrol:trans-tetrol from reaction of DE-2 at pH 4.0 and 5.1, where the hydronium ion-catalyzed reaction predominates, is ca. 6:94. HPLC analyses of the tetrol products from reaction of DE-2 at pH 4.0 for 10-15 min or for 15 h at room temperature give identical results. Solutions of tetrol products showed no detectable change when allowed to stand for >12 h at pH 2.1 and 12.0 and reanalyzed by HPLC, thus indicating that no isomerization of the cis- and trans-tetrols had occurred.
632 Chem. Res. Toxicol., Vol. 11, No. 6, 1998
Lin et al.
Figure 1. Plots of log kobsd vs pH for the reaction of DE-2 in 1:9 dioxane-water (v/v) solutions containing 0.1 M NaI, 0.5 M KBr, 1.0 M KCl, and 0.2 M NaClO4, 25 °C. The curves through the data points for NaI, KBr, and KCl are theoretical, based on eq 5 and the kinetic parameters listed in Table 2. The curve for NaClO4 is theoretical, based on eq 1 and the kinetic parameters listed in Table 2.3 Concentrations of hydroxide at pH < 12 were calculated from the equation [HO-] ) KW/[H+], with KW ) 10-14 M2, and were determined directly at pH > 12. Products from Reaction of DE-2 in Solutions Containing Halide Ion. Aliquots (10.0 µL) of DE-2 in dioxane (ca. 0.5 mg/mL) were added to vials containing 2.0 mL of 1:9 dioxanewater (v/v) containing concentrations of NaCl, NaBr, and NaI varying between 0 and 0.2 M. Ionic strength was kept constant at 0.2 by addition of NaClO4. The initial pH of chloride, bromide, and iodide solutions was adjusted with 0.2 M HClO4 in 1:9 dioxane-water to 11 exhibited greater error than those for reaction of DE-2 at lower pH, possibly due to the instability of ketone formed from the spontaneous reaction.
Scheme 2
Figure 2. Plots of kobsd for the reaction of DE-2 in 1:9 dioxanewater (v/v) solutions at pH 5.00 ( 0.02 as functions of varying NaClO4 and KCl concentrations, 25 °C. The slopes of the NaClO4 and KCl plots are (6.9 ( 1.0) × 10-3 and (5.5 ( 0.6) × 10-3 M-1 s-1, respectively.3
reaction of DE-2 in each rate region will be discussed separately. Region A. In this pH range the rate of reaction of DE-2 is proportional to H+ concentration, and acidcatalyzed hydrolysis of DE-2 is the predominant reaction. The reactivity of DE-2 in 1.0 M KCl is only slightly greater than that in 0.2 M NaClO4 solution. Plots of kobsd for reaction of DE-2 in KCl and NaClO4 solutions of varying salt concentrations at pH 5.0 are shown in Figure 2. The slopes of the two plots are very similar, and therefore the small increase in the rate constant for reaction of DE-2 in 1.0 M KCl solution compared to that in 0.2 M NaClO4 solution must be due to a normal salt effect. The fact that DE-2 reacts in KCl, KBr, and NaI
solutions at pH ca. 4 at about the same rate demonstrates that a second-order reaction of DE-2 with halide ion at this pH is small relative to the rate of acid-catalyzed ring opening of DE-2. Since the rate-limiting step of the hydrolysis reactions of DE-2 at pH < ca. 4.5, in both the presence and absence of halide ions, is the formation of carbocation 3 by reaction of H+ with DE-2, the relative reactivities of halide ions with 3 cannot be determined by kinetic measurements. Instead the relative reactivities of halide ions with 3 must be estimated by studying the products of reaction of DE-2 as a function of halide ion concentration. Therefore, tetrol yields from the reaction of DE-2 at pH 3.5-4.5 as functions of varying concentrations of halide ions were determined. In the absence of halide ion, DE-2 reacts at this pH via the acid-catalyzed pathway to yield 6% cis-tetrol and 94% trans-tetrol (5). In solutions of increasing halide ion concentrations at this pH, the reaction of DE-2 gives increasing yields of cistetrol. Plots of the percent cis-tetrol formed from reaction of DE-2 at pH ca. 4 as functions of halide ion concentrations are given in Figure 3. Since there is no detectable increase in the rate of reaction of DE-2 in chloride solutions other than that expected of a normal salt effect, the change in tetrol product ratio with increasing chloride concentrations must be attributed to the capture of an intermediate by chloride ion, after rate-limiting formation of the intermediate, followed by a secondary reaction of the chlorohydrin(s) to yield a tetrol mixture of different composition. This mechanism is also assumed to hold for the reaction of DE-2 in bromide and iodide solutions at pH < 4. These results are readily accommodated by the kH[H+] pathway for reaction of DE-2 in Scheme 2 to yield
634 Chem. Res. Toxicol., Vol. 11, No. 6, 1998
Figure 3. Plots of percent cis-tetrol in the mixture of cis- and trans-tetrols from reaction of DE-2 at room temperature in solutions of varying NaI (pH < 3.5), NaBr (pH < 4.2), and NaCl (pH < 4.4) concentrations. Ionic strength was kept constant at 0.2 with NaClO4. The curves are theoretical, based on eq 4. Values of (% cis)∞ are calculated to be 35.3 ( 1.3, 33.6 ( 0.9, and 34.8 ( 0.2 for KCl, KBr, and KI, respectively. Values of k2/ks are calculated to be 11.8 ( 1.4, 97 ( 17, and (3.4 ( 0.2) × 102 M-1 for Cl-, Br-, and I-, respectively.3
Scheme 3
Lin et al.
reacts to form tetrol products via the free carbocation 3, then there should be no change in the ratio of cis- and trans-tetrols from reaction of DE-2 with increasing halide ion concentrations. The increase in yield of cis-tetrol product observed from reaction of DE-2 in solutions containing halide ion must therefore be due to the incursion of a new pathway other than the ks route for tetrol product formation. The new pathway is labeled k3 in Schemes 2 and 3. From Scheme 3, the ratio of products formed from the ks and k3 pathways is given by the ratio (ks[3])/(k3[7]). However, without knowing the rate ratio k-2/k3 for reaction of each halohydrin 7, the partitioning ratio k2/ ks for reaction of 3 cannot be determined from our product data. However, a lower limit of this ratio can be estimated by assuming (1) that carbocation 3 partitions by reacting with solvent (ks) and by reacting with halide ion (k2) to form halohydrin 7 and (2) that all of the halohydrin 7 that is formed from the k2 reaction goes on to tetrol product via the k3 pathway, e.g., k3[7] . ks[3]. From published data for the hydrolysis of trans-chlorohydrin 5 in NaCl solutions (9), the percent cis-tetrol product formed as a function of NaCl concentration increases somewhat more slowly than the percent cistetrol product from reaction of DE-2 in the same NaCl solutions at low salt concentrations, indicating that k-2 and k3 may be comparable in magnitude. However, the equilibrium between 3 and 7 will shift toward 7 in solutions with increasing halide concentrations, resulting in lower concentrations of 3, and therefore the second assumption may hold reasonably well for the reaction of DE-2 in most of the halide solutions of Figure 3. With the assumption that all of the halohydrin 7 formed as an intermediate in the acid-catalyzed hydrolysis of DE-2 in halide solutions from the k2 step goes on to tetrol product via the k3 pathway, then the mole fraction of tetrol products formed by the ks route is given by eq 2. The percent of cis-tetrol in the product mixture
f(ks route) )
carbocation 3 as the intermediate that is trapped by halide ion. Since the carbocation-forming step (kH) is rate-limiting, in both the presence and absence of halide ion, it can be concluded that carbocation 3 is captured by solvent (ks) faster than it returns to DE-2 and H+. Thus, epoxide ring opening of DE-2 by H+ to yield 3 is irreversible. At pH < ca. 5, most of the tetrol product from reaction of DE-2 in 1.0 M KCl, 0.5 M KBr, and 0.1 M NaI solutions is formed from the intermediate halohydrin, formed from capture of carbocation 3 by halide ion (k2 pathway). In view of the fact that DE-2 reacts in 1.0 M KCl, 0.5 M KBr, and 0.1 M NaI solutions at pH < ca. 5 to yield carbocation 3 mainly by the kH route, then the lowest energy pathway for 3 to return to DE-2 under this condition is by the reverse of the kH step, according to the principle of microscopic reversibility (17). Thus, hydroxide-induced ring closure of 7 (via its conjugate base 6) is precluded as a significant reaction pathway under these conditions. Therefore, the mechanism of hydrolysis of DE-2 in solutions at low pH containing halide ion is given by Scheme 3. If carbocation 3 reacts with solvent faster than it reacts with halide ion, or if halohydrin 7
ks ks + k2[X-]
(2)
will be determined by the relative amounts of tetrol product formed from the ks and k3 pathways and will vary between 6%, when all of the tetrol products are formed via the ks route in the absence of halide ion, and ca. 35%, when all of the tetrol products are formed via the k3 route at high halide concentrations. The mole fraction of tetrol product formed by the ks route is given by eq 3. In eq 3,
f(ks route) )
(% cis)∞ - (% cis)x (% cis)∞ - (% cis)o
(3)
(% cis)x is the percent of cis-tetrol formed from reaction of DE-2 in a solution of a given halide concentration, (% cis)o is the percent of cis-tetrol formed in the absence of halide ion (via the ks route), and (% cis)∞ is the percent of cis-tetrol formed from reaction of DE-2 in solutions of infinite halide concentration (via the k3 route). Combining eqs 2 and 3 provides eq 4, which relates the percent
(% cis)x ) (% cis)∞ -
(% cis)∞ - (% cis)o 1 + (k2/ks)[X-]
(4)
of cis-tetrol product to the concentration of halide ion.
Halide Effects in Diol Epoxide Hydrolysis
Chem. Res. Toxicol., Vol. 11, No. 6, 1998 635
Nonlinear fitting of the percent cis-tetrol vs [X-] to eq 4 yields a lower limit of the value for k2/ks and a value of (% cis)∞, the calculated percent of cis-tetrol formed from reaction of DE-2 in solution with infinite halide concentration.3 Fitting of the data in Figure 3 provides values (lower limits) for k2/ks of 11.8 ( 1.4, 97 ( 17, and (3.4 ( 0.2) × 102 M-1 for reaction of DE-2 in chloride, bromide, and iodide solutions, respectively. Calculated values of (% cis)∞ are 35.3 ( 1.3, 33.6 ( 0.9, and 34.8 ( 0.2 as the percent of cis-tetrol formed from reactions of chlorohydrin, bromohydrin, and iodohydrin, respectively, via the k3 route. From Figure 3, values of the percent cis-tetrol product from reaction of DE-2 at pH 4 in solutions containing [Cl-] > ca. 0.2 M, [Br-] > ca. 0.05 M, and [I-] > ca. 0.02 M are considerably closer to the extrapolated value at infinite halide concentration than they are to the percent cis-tetrol product from reaction of DE-2 in the absence of halide ion (via reaction of 3 with solvent). Several important conclusions can be drawn from these results. First, in solutions containing Br- and I- ions in concentrations equal to or greater than those concentrations listed above and Cl- > ca. 0.4-0.5 M, most of the tetrol product from reaction of DE-2 at pH 4 is formed from halohydrin via the k3 pathway, and not from the free carbocation 3 via the ks pathway (k3[7] > ks[3]). If this were not true, the percent cis-tetrol in the product mixture from reaction of DE-2 would be much closer to that observed for reaction of DE-2 in the absence of halide ion, where the tetrols are formed by reaction of solvent with 3. Second, carbocation 3 must react with halide ion present in these concentrations faster than it reacts with water (k2[X-] . ks). If this were not true, the percent cis-tetrol in the product mixture would not change as a function of halide ion concentration as rapidly as it does. Rate Equation. Although halohydrin 7 may not be a steady-state intermediate in the hydrolysis of DE-2 at pH < ca. 4, where the rate of the second-order reaction of DE-2 with H+ will at some pH exceed the rate of the first-order hydrolysis of 7, it is safe to assume that it will be a steady-state intermediate in the hydrolysis of DE-2 at pH’s greater than that at which halide ion exerts a rate effect by nucleophilic attack on DE-2 (pH > ca. 5 for I- and 6 for Cl-). The trans-chlorohydrin 5 is reported to be considerably more reactive than DE-2 in aqueous buffer solution (9), and the corresponding bromohydrin and iodohydrin are expected to be much more reactive than 5 due to the fact that bromide and iodide are better leaving groups than chloride ion in solvolysis reactions. For the reaction of DE-2 in halide solutions under the conditions of Figure 1, as concluded in the foregoing discussion, k2[X-] . ks and k3[7] > ks[3]. From these conditions and the assumption that intermediates 6 and 7b are steady-state intermediates, the rate expression for reaction of DE-2 is given by eq 5. Weighted nonlinear
kobsd )
kH[H+] + k1[X-] + ko + kOH[HO-] k-1Ka 1+ k3[H+]
(5)
least-squares fit of the rate data to eq 5 yields 3 Curve fitting of data and graphing were done with PRISM, a program available from GraphPad Software Co.
values for kH, ko, kOH, k1, and k-1Ka/k3.3 A summary of these parameters is provided in Table 2. Region B. For halide solutions having pH values in this range and containing sufficient concentrations of halide ion, nucleophilic attack of halide ion on DE-2 becomes the predominant reaction. It is assumed that the second-order, nucleophilic attack of halide ion occurs to yield trans-oxyanion 6. In the pH range of region B, the equilibrium between 6 and its conjugate acid, halohydrin 7b, lies sufficiently in favor of 7b such that all of 6 formed from nucleophilic attack of X- on DE-2 proceeds on to tetrol product via the k3 route. In this pH range k1[X-] > kH[H+] and (k-1Ka)/(k3[H+]) , 1, and the first term of eq 5 reduces to k1[X-]. For 1.0 M KCl, 0.5 M KBr, and 0.1 M NaI, this kinetic term is greater than the spontaneous rate constant and gives a rate plateau in the pH-rate profile that is significantly higher than that due to the spontaneous reaction. For example, the rate constant for reaction of DE-2 in 1.0 M KCl solution at pH 8.5 is 10 times larger than the spontaneous rate constant for reaction of DE-2 in 0.2 M NaClO4 solution at the same pH and ca. 7 times larger than that for reaction of DE-2 at pH 11.2 in the same solvent. Thus, approximately 90% of the reaction of DE-2 in 1.0 M KCl solution at pH 8.5 goes by way of nucleophilic addition of chloride ion to yield chlorohydrin 7b, followed by solvolysis of chlorohydrin via the k3 route to yield tetrols. In 0.5 M KBr and 0.1 M NaI solutions at this pH, an even larger fraction of the reaction of DE-2 hydrolyzes by this mechanism. In 0.2 M NaClO4 solution, the mechanism of reaction of DE-2 changes from acid-catalyzed hydrolysis to spontaneous hydrolysis in the pH range ca. 6.5-8.5. Acidcatalyzed hydrolysis of DE-2 yields only about 6% cistetrol, whereas the spontaneous hydrolysis of DE-2 yields ca. 44% cis-tetrol (5). Thus, the yield of cis-tetrol increases as the hydrolysis mechanism changes from acid-catalyzed to spontaneous, in the absence of halide ion. At pH 4.5, where the acid-catalyzed reaction predominates, there is an increase in yield of cis-tetrol with increasing concentrations of chloride ion, as summarized in Figure 3. However, at pH 8.5, where the spontaneous reaction of DE-2 predominates, there is actually a slight decrease in yield of cis-tetrol brought about by increasing concentrations of chloride ion, from about 44% in the absence of chloride ion to about 30% in 1.0 M KCl. Thus, the presence of halide ion only increases the yields of cistetrol from reaction of DE-2 in that pH range where the acid-catalzyed hydrolysis of DE-2 predominates. At pH 7 in the absence of halide ion, DE-2 reacts partly by acid-catalyzed hydrolysis and partly by the spontaneous reaction. From the rate constants kH and ko (Table 2), it is calculated from eq 1 that 64% of DE-2 reacts at pH 7.0 by the acid-catalyzed route and the remaining 36% by the spontaneous route. Therefore, added halide ion both captures the intermediate carbocation 3 and adds as a nucleophile to DE-2. At this pH, DE-2 reacts 5 times faster in 1.0 M KCl solution than in 0.2 M NaClO4, and thus the rate enhancement due to nucleophilic attack of Cl- on DE-2 is approximately 8 times greater than the rate of acid-catalyzed hydrolysis. Therefore, at this pH 1.0 M chloride ion reacts mainly as a nucleophile in attacking DE-2 and to a lesser extent as a nucleophile in capturing the intermediate carbocation 3 from that portion of the reaction proceeding via acidcatalyzed hydrolysis. The relative amount of DE-2 that
636 Chem. Res. Toxicol., Vol. 11, No. 6, 1998
reacts by each mechanism at pH 7 depends on the halide ion concentration. Our observation that Cl- acts as an efficient nucleophile in reacting with DE-2 at pH 7.0 in 1:9 dioxane-water solutions contrasts to an earlier report that chloride ion does not exert a very large rate effect on the reaction of DE-2 in 10% acetone-water solutions at pH 7.0 containing 10 mM sodium cacodylate buffer (8). This apparent discrepancy may be reconciled if the differences in conditions between the present work and earlier studies in aqueous acetone are taken into account. Part of the difference may be caused by the presence of 10 mM cacodylate buffer in the aqueous acetone solution, which, like phosphate (6, 7), should be an efficient general acid catalyst that would give rise to a larger background rate leading to carbocation 3. Acetone might also act slightly different than dioxane as a cosolvent, resulting in a slight change of the relative rates of acid-catalyzed and spontaneous hydrolysis of DE-2 at pH 7. Regions C and D. Spontaneous Reaction of DE2. Above pH ca. 9.5, the value of (k-1Ka)/(k3[H+]) becomes comparable to or greater than unity. This results in a gradual lowering of the kinetic term in [X-] with increase in pH until it becomes < ko at pH ca. 11.3. Thus, the equilibrium (HX + DE-2 h 7) shifts gradually to the left in favor of epoxide with increase in pH until ko[DE-2] > k3[7] at pH 11.5, where kobsd becomes approximately equal to ko. The larger values of kobsd for reaction of DE-2 at pH 11.3 in 0.5 M KBr and 1.0 M KCl compared to that in 0.2 M NaClO4 most likely reflect a salt effect on ko. The spontaneous reaction of DE-2 (ko) is more favorable than either acid-catalyzed or hydroxide-catalzyed reactions in the pH range ca. 7.5-11.5. Therefore, the principal reactions of DE-2 in halide solutions with pH 8-10 are the spontaneous reaction (ko) and nucleophilic addition of halide ion (k1). The relative amount of reaction proceeding via each pathway will depend on the concentration of halide ion. The cis-tetrol:trans-tetrol ratio from reaction of DE-2 at pH 8-10 in 0.2 M NaClO4 solutions, where the spontaneous reaction predominates, is approximately 44: 56. The yield of tetrol products from the spontaneous reaction of DE-2 in this pH region, however, is found to be only 74-86% of the yield of tetrol products from the acid-catalyzed reaction at pH 4-5. It has been reported that a ketone product [(()-7β,8R-dihydroxy-9-keto-7,8,9, 10-tetrahydrobenzo[a]pyrene] formed in the spontaneous reaction of DE-1 is unstable in water solutions and further reacts to form products that cannot be detected by HPLC (5, 7). A possible explanation for the lower yields of tetrol products from the spontaneous reaction of DE-2 is that some of this same ketone (perhaps 1020%) is formed from this reaction, in addition to tetrol products, but is not stable to reaction conditions. The yield of tetrols from reaction of DE-2 at pH > 10 appears to be reduced somewhat further, and the explanation for this result is not clear. Region E. In region E, kobsd increases with increase in hydroxide ion concentration. For aryl- and vinylsubstituted epoxides, an increase in rate at pH > 12 is due to nucleophilic addition of HO- to the epoxide group, leading to trans-diols (15). However, product studies of the reaction of DE-2 in hydroxide solutions with [HO-] > 0.01 M indicate that the yields of tetrols become lower with increasing hydroxide concentrations. Although there is an increase in the trans-tetrol:cis-tetrol product ratio with increasing hydroxide concentration, there is
Lin et al.
not a good rate-product correlation if it is assumed that the second-order reaction of DE-2 and hydroxide ion is solely due to nucleophilic addition of hydroxide leading to trans-tetrol. Further work is necessary to clarify the mechanism of reaction of DE-2 in highly basic solutions. Stereochemistry of Nucleophile Attack on Carbocation 3. Acid-catalyzed hydrolysis of DE-2 yields 94% trans-tetrol and only 6% cis-tetrol, and this observation has been rationalized by a mechanism in which there is energetically favorable trans, axial attack of solvent on the more stable conformation of carbocation 3 (18, 19). Reaction of DE-2 at pH 4.7 in 1:9 dioxane-water containing sodium azide yields a 1:4 mixture of azide products arising from cis and trans addition of azide ion, respectively, to 3 (20). Thus, reactions of 3 with water and azide ion nucleophiles result in the formation of minor yields of cis products. Therefore, it is very probable that 3 also reacts with halide ion to give some cishalohydrin 7a as a minor product, which may not be easily detected, along with the major trans-halohydrin 7b. Mechanism of Hydrolysis of Halohydrins 7. The acid-catalyzed hydrolysis of DE-2 in solutions containing sufficient concentrations of halide ion, e.g., 0.2 M Cl-, proceeds completely via intermediate halohydrin(s), which lead to tetrol products by the k3 pathway. One possible mechanism for this k3 pathway involves ionization of 7 to a carbocation-halide ion pair, followed by reaction of solvent with this ion-pair intermediate. A second possible mechanism involves the ionization of 7 to a carbocation that is conformationally different from that formed in the reaction of DE-2 by the kH route, followed by reaction of this carbocation with solvent at a rate that is comparable to or greater than that for its conformational isomerization. This mechanism is attractive because it would account for the fact that the ratios of cisand trans-tetrol products from reactions of all three halohydrins via the k3 pathway are almost the same. In solvolysis reactions leading to relatively stable carbocations such as 3, ionization occurs first to an intimate ion pair, then to a solvent-separated ion pair, and finally to fully separated ions, 3 + X- (21). Tetrol product can arise from reaction of solvent with each intermediate along the reaction pathway. Since solvolysis studies of halohydrin 7 in solutions free of external halide salts have not yet been carried out, the amount of tetrol product formed from 7 via a freely solvated ion 3 under such conditions is not known. Even if k-2 is comparable to or greater than k3, capture of 3 by halide ion in halide solutions lowers the steady-state concentrations of 3 in the reaction of DE-2. Consequently, if there are other product-forming pathways for reaction of 7, e.g., k3, then the relative amount of tetrol product from reaction of 7 via a freely solvated ion 3 (ks) in solutions containing halide salts will decrease with increasing halide concentrations. If hydrolysis of 7 occurs by an ionpair mechanism, the relative amount of tetrol forming at the earlier ion-pair stage in the reaction of 7 will increase with increasing halide concentration. Reaction of solvent-separated ion pairs with solvent would also account for the fact that both inversion and retention at carbon occurs. More work is needed to clarify the mechanisms of reaction of these halohydrins. Relative Reactivities of Halide Ions with DE-2 and with Carbocation 3. From the k1 values listed in Table 1, the relative reactivities of Cl-, Br-, and I- toward
Halide Effects in Diol Epoxide Hydrolysis
bimolecular addition to DE-2 are 1:8:96, respectively. This order of nucleophilicity for halide ions in water solution is observed in SN2 reactions (22). In the reaction of halide ion with carbocation 3, the order of nucleophilicity is the same, e.g., I- > Br- > Cl-. However, the relative reactivities for reaction of Cl-, Br-, and I- with 3 are 1:8:28, respectively. Thus, Br- is ca. 8 times more reactive than Cl- toward reaction with both DE-2 and carbocation 3. However, I- is 96 times more reactive than Cl- toward DE-2 but only 28 times more reactive than Cl- in reaction with 3. One possible explanation for this reduced reactivity of I- with 3 relative to that of Br- is that I- reacts with 3 at the diffusional limit and this rate constant cannot be exceeded. Carbocation 3 is also trapped in aqueous solution by azide ion, and the kaz/ks ratio is observed to be 2.5-3.0 × 102 M-1 (23). The kaz/ks ratios for activation-limited attack of N3- on stable carbocations (24, 25) are orders of magnitude greater than that observed for reaction of 3 with N3- (23). Consequently, it has been argued by Jencks and co-workers that low values of kaz/ks for reaction of certain carbocations indicate that the reaction of N3- with the carbocation occurs at the diffusional limit, whereas reaction of the carbocation with solvent (ks) is activation-limited (26). The kaz/ks ratio for reaction of 3 with I- is larger than that for reaction of 3 with N3-, so if N3- reacts with 3 at the diffusional limit, then I- most likely does also. If the diffusion-limited rate constant is assumed to be 5 × 109 M-1 s-1 (26) and kaz assigned this value, then k2 for reaction of halide ions with 3 is calculated to be g(2.4 × 108) M-1 s-1 for Cl-, g(1.8 × 109) M-1 s-1 for Br-, and g(6 × 109) M-1 s-1 for I-.
Summary At pH < ca. 5 in 1.0 M KCl, 0.5 M KBr, and 0.1 M NaI solutions, DE-2 reacts with hydronium ion in a ratedetermining step to yield carbocation 3, which is efficiently trapped by halide ions to yield halohydrins. Relative reactivities of the halide ions with 3 are I- > Br- > Cl-, with I- reacting at or near the diffusional limit with 3. In solutions with these halide concentrations, the halohydrin intermediates hydrolyze mainly by a route that does not involve carbocation 3, but rather they hydrolyze by another mechanism possibly involving attack of solvent on an ion pair or involving a carbocation intermediate that is conformationally different from that formed from reaction of DE-2 and H+. Hydrolysis of the halohydrin by this pathway yields ca. 35% cis-tetrol and 65% trans-tetrol, a mixture enriched in cis-tetrol compared to that from reaction of solvent with 3. When DE-2 is hydrolyzed in solutions at intermediate pH containing sufficient halide ion concentrations (e.g., 1.0 M KCl, 0.5 M KBr, and 0.1 M NaI), nucleophilic attack of halide ion on DE-2 becomes the principal reaction. This reaction leads to a trans-halohydrin intermediate, which hydrolyzes to form a tetrol mixture containing slightly less tetrol than formed from the spontaneous reaction of DE-2 in the absence of halide ion. Thus, the presence of halide ion only increases the yields of cis-tetrol from reaction of DE-2 in that pH range where the acid-catalzyed hydrolysis of DE-2 predominates. At pH greater than ca. 8.59.5, the pH-dependent equilibrium DE-2 + H+ + X- h 7b shifts to the left, resulting in a decrease in the rate of reaction of DE-2, until at pH ca. 11.2 the rate is equal to that of the spontaneous reaction. Kinetic parameters for
Chem. Res. Toxicol., Vol. 11, No. 6, 1998 637
reaction of DE-2 in 1:9 dioxane-water containing 1.0 M KCl, 0.5 M KBr, and 0.1 M NaI in the pH range 4-13 are determined.
Acknowledgment. This work was funded in part by a Special Research Initiative Support Award from the UMBC Designated Research Initiative Fund. Helpful discussions with Dr. Jane Sayer (National Institutes of Health) are greatly appreciated.
References (1) Thakker, D. R., Levin, W., Wood, A. W., Conney, A. H., Yagi, H., and Jerina, D. M. (1984) Stereoselective biotransformation of polycyclical aromatic hydrocarbons to ultimate carcinogens. In Drug StereochemistrysAnalytical Methods and Pharmacology (Wainer, I. W., and Drayer, D. E., Eds.) pp 271-296, Marcel Dekker, New York. (2) Jerina, D. M., Yagi, H., Thakker, D. R., Sayer, J. M., van Bladeren, P. J., Lehr, R. E., Whalen, D. L., Levin, W., Chang, R. L., Wood, A. W., and Conney, A. H. (1984) Identification of the ultimate carcinogenic metabolites of the polycyclic aromatic hydrocarbons: Bay-region (R,S)-diol-(S,R)-epoxides. In Foreign Compound Metabolism (Caldwell, J., and Paulson, G. D., Eds.) pp 257-266, Taylor and Francis Ltd., London. (3) Yagi, H., Thakker, D. R., Hernandez, O., Koreeda, M., and Jerina, D. M. (1977) Synthesis and reactions of the highly mutagenic 7,8diol 9,10-epoxides of the carcinogen benzo[a]pyrene. J. Am. Chem. Soc. 99, 1604-1611. (4) Keller, J. W., Heidelberger, C., Beland, F. A., and Harvey, R. G. (1976) Hydrolysis of syn- and anti-benzo[a]pyrene diol epoxides: stereochemistry, kinetics and the effect of an intramolecular hydrogen bond on the rate of syn-diol epoxide solvolysis. J. Am. Chem. Soc. 98, 8276-8277. (5) Whalen, D. L., Montemarano, J. A., Thakker, D. R., Yagi, H., and Jerina, D. M. (1977) Changes of mechanisms and product distributions in the hydrolysis of benzo[a]pyrene-7,8-diol 9,10epoxide metabolites induced by changes in pH. J. Am. Chem. Soc. 99, 5522-5524. (6) Whalen, D. L., Ross, A. M., Montemarano, J. A., Thakker, D. R., Yagi, H., and Jerina, D. M. (1979) General acid catalysis in the hydrolysis of benzo[a]pyrene 7,8-diol 9,10-epoxides. J. Am. Chem. Soc. 101, 5086-5088. (7) Islam, N. B., Gupta, S. C., Yagi, H., Jerina, D. M., and Whalen, D. L. (1990) Trapping of a carbocationic intermediate in the spontaneous hydrolysis reaction of 7β,8R-dihydroxy-9β,10βepoxy-7,8,9,10-tetrahydrobenzo[a]pyrene: mechanism of the spontaneous and general acid-catalyzed hydrolysis reaction of bayregion benzo[a]pyrene 7,8-diol9,10-epoxides. J. Am. Chem. Soc. 112, 6363-6369. (8) Wolfe, A. R., and Meehan, T. (1994) Halide-catalyzed cis product formation in the hydrolysis of anti-benzo[a]pyrene diol epoxide and its alkylation of poly(A). Chem. Res. Toxicol. 7, 110-119. (9) Meehan, T., Wolfe, A. R., Negrete, G. R., and Song, Q. (1997) Benzo[a]pyrene diol epoxide-DNA cis adduct formation through a trans chlorohydrin intermediate. Proc. Natl. Acad. Sci. U.S.A. 94, 1749-1754. (10) Wolfe, A. R., Song, Q., and Meehan, T. (1996) The effect of temperature on the stereochemistry of adducts and hydrolysis products derived from anti-benzo[a]pyrene diol epoxide in the presence and absence of halide ions. Polycyclic Aromat. Hydrocarbons 10, 203-210. (11) Wolfe, A. R., Yamamoto, J., and Meehan, T. (1994) Chloride ions catalyze the formation of cis adducts in the binding of anti-benzo[a]pyrene diol epoxide to nucleic acids. Proc. Natl. Acad. Sci. U.S.A. 91, 1371-1375. (12) Whalen, D. L., and Ross, A. M. (1976) Specific effects of chloride ion on epoxide hydrolysis. The pH-dependence of the rates and mechanisms for the hydrolysis of indene oxide. J. Am. Chem. Soc. 198, 7859-7861. (13) Whalen, D. L., Ross, A. M., Dansette, P. M., and Jerina, D. M. (1977) Specific effects of chloride ion in the hydrolysis of a K-region arene oxide. J. Am. Chem. Soc. 99, 5672-5676. (14) Nashed, N. T., Balani, S. K., Loncharich, R. J., Sayer, J. M., Shipley, D. Y., Mohan, R. S., Whalen, D. L., and Jerina, D. M. (1991) Solvolysis of K-region arene oxides: substituent effects on reactions of benz[a]anthracene 5,6-oxide. J. Am. Chem. Soc. 113, 3910-3919. (15) Ross, A. M., Pohl, T. M., Piazza, K., Thomas, M., Fox, B., and Whalen, D. L. (1982) Vinyl epoxide hydrolysis reactions. J. Am. Chem. Soc. 104, 1658-1665. (16) Yagi, H., Akagi, H., Thakker, D. R., Mah, H. D., Koreeda, M., and Jerina, D. M. (1977) Absolute stereochemistry of the highly
638 Chem. Res. Toxicol., Vol. 11, No. 6, 1998 mutagenic 7,8-diol 9,10-epoxides derived from the potent carcinogen trans-7,8-dihydroxy-7,8-dihydrobenzo[a]pyrene. J. Am. Chem. Soc. 99, 2358-2359. (17) Carey, F. A., and Sundberg, R. J. (1984) Advanced Organic Chemistry Part A: Structure and Mechanism, p 176, Plenum Press, New York. (18) Gillilan, R. E., Pohl, T. M., and Whalen, D. L. (1982) Substituent effects on rates and product distributions in the hydrolysis reactions of naphthalene tetrahydro epoxides. Models for tetrahydro epoxides of polycyclic aromatic hydrocarbons. J. Am. Chem. Soc. 104, 4481-4482. (19) Sayer, J. M., Yagi, H., Lilverton, J. V., Friedman, S. L., Whalen, D. L., and Jerina, D. M. (1982) Conformational effects in the hydrolyses of rigid benzylic epoxides: implications for diol epoxides of polycyclic hydrocarbons. J. Am. Chem. Soc. 104, 19721978. (20) Lin, B., Doan, L., Yagi, H., Jerina, D. M., and Whalen, D. Quantitative aspects of the effects of halide and azide ions on the acid-catalyzed hydrolysis of (()-7β,8R-Dihydroxy-9R,10R-epoxy7,8,9,10-tetrahydrobenzo[a]pyrene. Polycyclic Aromat. Hydrocarbons, submitted for publication.
Lin et al. (21) Winstein, S., Klinedinst, P. E., Jr., and Robinson, G. C. (1961) Salt effects and ion pairs in solvolysis and related reactions. XXI. Acetolysis, bromide exchange and the special salt effect. J. Am. Chem. Soc. 83, 4986-4989. (22) Swain, C. G., and Scott, C. B. (1953) Quantitative correlation of relative rates. Comparison of hydroxide ion with other nucleophilic reagents toward alkyl halides, esters, epoxides and aryl halides. J. Am. Chem. Soc. 75, 141-147. (23) Lin, B., Islam, N., Friedman, S., Yagi, H., Jerina, D. M., and Whalen, D. L. (1998) Change in rate-limiting step in general acidcatalyzed benzo[a]pyrene diol epoxide hydrolysis. J. Am. Chem. Soc. 120, 4327-4333. (24) Bunton, C. A., and Huang, S. K. (1972) Reactions of the tri-panisylmethyl carbonium ion with nucleophiles. J. Am. Chem. Soc. 94, 3536-3544. (25) Ritchie, C. D. (1975) Cation-anion combination reactions. XIII. Correlation of the reactions of nucleophiles with esters. J. Am. Chem. Soc. 97, 1170-1179. (26) Richard, J. P., Rothenberg, M. E., and Jencks, W. P. (1984) Formation and stability of ring-substituted 1-phenylethyl carbocations. J. Am. Chem. Soc. 106, 1361-1372.
TX9701743
