Abstract
The reductive metabolism of a series of 3-(indol-1-yl)-1,2-benzisoxazoles was examined in vitro using rat liver microsomes. 3-(Indol-1-yl)-1,2-benzisoxazole was reduced to the corresponding amidine (resulting from N–O bond cleavage) under anaerobic conditions. The reaction required viable microsomes and NADPH and was inhibited by carbon monoxide, air, and ketoconazole, suggesting the involvement of cytochrome P450 enzymes. The amidine was subsequently nonenzymatically hydrolyzed to 1-salicylindole, which in turn was hydrolyzed to indole. Addition of electron-withdrawing substituents (Cl-, MeSO2-) at the 6-position of the benzisoxazole ring resulted in a significant increase in the rate of substrate reduction. Introduction of electron-withdrawing substituents on the indole ring likewise increased the rate of substrate consumption but caused a substituent-dependent shift of the site of bond cleavage from the 1,2-isoxazole N–O bond to the C–N bond linking the 1,2-benzisoxazole to the indole moiety. In the case of 3-(2-chloro-3-methanesulfoxylindol-1-yl)-1,2-benzisoxazole, C–N bond cleavage was nearly quantitative, and products resulting from N–O bond reduction were not observed. The overall rates of 3-(indol-1-yl)-1,2-benzisoxazoles reduction were found to be substrate concentration-dependent and observed Michaelis-Menten-type behavior. The apparent Vmax of substrate reduction by rat liver microsomes correlated negatively with the free energy of the lowest unoccupied molecular orbitals (ELUMO) calculated semiempirically using a parameterized model 3 (PM3), and suggested that the initial electron transfer was rate-determining and that the ELUMO could be used as an indication of the susceptibility of 1,2-isoxazoles to undergo reductive metabolism.
The 1,2-isoxazole heterocycle is found in several marketed drugs including the anticonvulsant zonisamide and the antipsychotic iloperidone. One particular feature of the 1,2-isoxazole ring is that it is prone to reductive metabolism resulting in cleavage of the N–O bond (Dalvie et al., 2002). The susceptibility of 1,2-isoxazoles to reductive N–O bond cleavage was first observed when the insecticide isoxathion [O,O-diethyl O-(5-phenyl-3–1,2-isoxazolyl)-phosphorothioate] was administered orally to rats (Ando et al., 1975) and hippuric acid was isolated from urine. In vitro studies indicated that 3-hydroxy-5-phenyl-1,2-isoxazole, a metabolite found in urine as well, was reduced to benzoylacetamide and further metabolized to benzoic acid when incubated with rat liver S9 fractions in the presence of NADPH. More recently, it was shown that zonisamide undergoes extensive reductive metabolism in vivo and in vitro. In the rat, 2-(sulfamoylacetyl)-phenol glucuronide and 2-(1-aminosulfamoylethyl)-phenol, two metabolites identified in rat urine, were most likely formed from a common imine intermediate that resulted from N–O bond cleavage (Stiff and Zemaitis, 1990). It was subsequently shown that metabolism of zonisamide to 2-(sulfamoylacetyl)-phenol was mediated by rat liver microsomes (Nakasa et al., 1992; Stiff et al., 1992) and, to a lesser extent, cytosolic (Sugihara et al., 1996) and intestinal extracts (Kitamura et al., 1997). The microsomal reduction of zonisamide was NADPH-dependent and was inhibited by carbon monoxide and cimetidine, suggesting that the reaction was cytochrome P450-mediated. It was later determined that P450s1 from the 3A family were the main isoforms involved in rat (Nakasa et al., 1993a) and in human (Nakasa et al., 1993b). Other 1,2-isoxazole-containing compounds such as risperidone (Mannens et al., 1993; Meuldermans et al., 1994), iloperidone (Mutlib et al., 1995; Mutlib and Klein, 1998), and leflunomide all exhibit reductive metabolism in vivo. The structurally related heterocycles 1,2-isothiazoles and 1,2,4-oxadiazoles have been reported to undergo similar ring cleavage (Dalvie et al., 2002).
We have recently observed that 3-(indol-1-yl)-1,2-benzisoxazoles were undergoing N-dearylation in rodents and that the extent of this occurrence appeared to be highly dependent upon the nature of the substituent(s) on the substrate. In view of the existing literature on the metabolism of 1,2-isoxazoles, we hypothesized that reductive metabolism could be involved and have sought to investigate the reaction in vitro. We report here that 3-(indol-1-yl)-1,2-benzisoxazoles are metabolized to the N-dearylated indoles by rat liver microsomes under anaerobic conditions via two different, substituent-dependent, pathways involving either reductive N–O bond cleavage followed by hydrolysis or an unprecedented reductive C–N bond cleavage.
Materials and Methods
Chemicals. Chemicals and reagents were obtained from standard commercial sources. NADPH, glucose 6-phosphate, glucose, and glucose-6-dehydrogenase were form Sigma-Aldrich (St. Louis, MO). Catalase and glucose oxidase were from Roche Diagnostics (Indianapolis, IN). Rat liver microsomes were prepared according to a standard protocol (Stiff et al., 1992). The microsomal protein concentration was determined using the BCA reagents (Pierce Chemical, Rockford, IL). The P450 content of the microsomal preparation used throughout was 0.77 pmol/mg as determined by the method of Omura and Sato (1964).
Incubations. Typical anaerobic reaction mixtures consisted of 100 mM potassium phosphate (pH 7.4), 0.1 mM EDTA, 1 mM NADPH, 10 mM MgCl2, 5 mM glucose, 10 mM glucose 6-phosphate, 1 U/ml glucose-6-phosphate dehydrogenase, 30 U/ml catalase, 5 U/ml glucose oxidase, 2 mg/ml rat liver microsomes, and 5 to 500 μM substrate in a total volume of 0.5 or 1 ml. Substrates were added as acetonitrile solutions, and the final concentration of acetonitrile in the incubations did not exceed 1%. Anaerobic incubations were conducted by flushing with and keeping the incubation mixtures under argon. The solutions were preincubated with substrate for 5 min at 37°C before initiating the reaction by addition of NADPH. Aerobic incubations were performed by omitting catalase, glucose oxidase, and argon. For time courses and kinetic analysis, 50-μl aliquots were withdrawn at given time points and quenched with 75 μl of acetonitrile containing oxindole (50 μM) as an internal standard. The samples were kept at 4°C for 20 min and centrifuged. Forty to 80 μl of the supernatant were used for HPLC-UV or liquid chromatography-mass spectrometry analysis.
Analytical. HPLC-UV analysis was carried out on a Shimadzu (Columbia, MD) system equipped with a photodiode-array detector. Liquid chromatography-mass spectrometry was carried out using a Thermo Finnigan (San Jose, CA) LCQ ion trap instrument interfaced with an atmospheric pressure chemical ionization source. The heated capillary was maintained at 200°C while the probe was heated at 450°C. The sheath gas was N2 and kept at 70 psi. Separation was performed using an ACE C18 column (4.6 × 150 mm × 3 μm; MAC-MOD Analytical, Inc., Chadds Ford, PA) and a gradient of solvent B (water/acetonitrile/acetic acid, 5:95:0.1) into solvent A (water/acetonitrile/acetic acid, 95:5:0.1): 0 to 1 min, 10% B; 1 to 13 min, 10 to 100% B; 13 to 18 min, 100% B; 18 to 19 min, 100 to 10%.
Synthesis.General methods. 3-Chloro-1,2-benzisoxazoles were prepared as previously described (Boshagen, 1967). All reagents and solvents were obtained from Sigma-Aldrich and were used without further purification. All reactions involving air-sensitive reagents were performed under nitrogen in glassware that was flame-dried prior to use.
3-(Indol-1-yl)-1,2-benzisoxazole (1a), 3-(indol-1-yl)-6-chloro-1,2-benzisoxazole (1b), and 3-(indol-1-yl-3-carboxaldehyde)-1,2-benzisoxazole (1d). A mixture of indole, the corresponding 3-chloro-1,2-benzisoxazole (1 Eq), and Cs2CO3 (1.2 Eq) was heated in DMF ∼3 ml/mmol at 150°C for 18 h under a nitrogen atmosphere. The reaction was then cooled to room temperature and diluted with ether (100 ml). The organic layer was washed with water and brine, dried over Mg2SO4, filtered, evaporated under vacuum, and purified on SiO2 [Biotage 40S (Biotage Inc., Charlottesville, VA), linear gradient elution with hexanes/ethyl acetate, 5:95–50:50, 750 ml] to give the desired products in 40 to 60% yield: 1a: 1H-NMR (CDCl3) δ: 6.85 (d, J = 3.0 Hz, 1H), 7.28 (dt, J = 1.0, 8.0 Hz, 1H), 7.28 (ddd, J = 1.5, 1.5, 8.0 Hz, 1H), 7.64–7.68 (m, 2H), 7.71 (d, J = 8.0 Hz, 1H), 7.74 (d, J = 3.0 Hz, 1H), 7.92 (d, J = 8.0 Hz, 1H), 8.17 (d, J = 8.5 Hz, 1H); MS: m/z 234 (MH+); 1b: 1H-NMR (CDCl3) δ: 6.88 (d, J = 3.4 Hz, 1H), 7.32 (dt, J = 0.9, 7.1 Hz, 1H), 7.39–7.44 (m, 2H), 7.71–7.74 (m, 2H), 7.86 (d, J = 8.7 Hz, 1H), 8.16 (d, J = 8.5 Hz, 1H), 8.15 (d, J = 8.5 Hz, 1H); MS: m/z 269 (MH+).; 1d: 1H-NMR (CDCl3) δ: 7.48–7.51 (m, 3H), 7.79 (s, 1H), 7.83 (d, J = 8.4 Hz, 1H), 8.05 (d, J = 7.6 Hz, 1H), 8.28 (s, 1H), 8.44 (dd, J = 2.2, 6.1 Hz, 1H), 10.25 (s, 1H); MS: m/z 297 (MH+).
3-(Indol-1-yl)-6-methylsulfonyl-1,2-benzisoxazole (1c).1b (51 mg, 0.19 mmol) and sodium methylthiolate (15 mg, 0.21 mmol) were heated at 100°C in DMF (1.9 ml) under nitrogen for 1 h. The reaction was then cooled to room temperature and then diluted with ether (50 ml). The ether was washed with brine, dried over Mg2SO4, filtered, and evaporated to provide a crude oil that was treated with an excess of peracetic acid to yield 1c as an off-white solid: 1H-NMR (CDCl3) δ: 3.17 (s, 3H), 6.90 (d, J = 3.5 Hz, 1H), 7.32 (t, J = 7.5 Hz, 1H), 7.41 (t, J = 7.5 Hz, 1H), 7.71 (d, J = 3.5 Hz, 1 H), 7.72 (d, J = 7.5 Hz, 1H), 7.99 (dd, J = 1.5 × 8.5 Hz, 1H), 8.15 (d, J = 8.5 Hz, 2H), 8.31 (s, 1H); MS: m/z 313.1 (MH+).
3-(2-Chloro-3-methylsulfoxyindol-1-yl)-6-chloro-1,2-benzisoxazole (1e). Sulfuryl chloride (21 μl, 0.26 mmol) was added to a solution of dimethyl sulfide (27 μl, 0.30 mmol) in 0.5 ml of 1,2-dichloroethane at room temperature. The resulting yellow solution (200 μl) was added to a solution of 1b (17.3 mg, 0.07 mmol) in 500 μl of DMF and stirred overnight at room temperature. The reaction was terminated with water and extracted three times with methyl-t-butyl ether. The organic phases were combined, washed with water and brine, dried over Mg2SO4, filtered, and evaporated to give an oil that was crystallized with methanol to yield 3-(2-chloro-3-methylthioindol-1-yl)-6-chloro-1,2-benzisoxazole as a white solid (13.2 mg, 59%): 1H-NMR (CDCl3) δ: 2.44 (s, 3H), 7.30 (td, J = 7.0 × 1.5 Hz, 1H), 7.33 (td, J = 7.0 × 1.5 Hz, 1H), 7.38 (dd, J = 7.0 × 1.5 Hz, 1H), 7.40 (dd, J = 7.0 × 1.5 Hz, 1H), 7.54 (d, J = 7.0 Hz, 1H), 7.75 (d, J = 1.5 Hz, 1H), 7.79 (dd, J = 7.0 × 1.5 Hz, 1H); MS: m/z 349 (MH+). Treatment with peracetic acid gave 1e in quantitative yield as a white solid: 1H-NMR (CDCl3) δ: 3.16 (s, 3H), 7.37 (m, 3H), 7.42 (dd, J = 7.0 × 1.5 Hz, 1H), 7.51 (d, J = 7.0 Hz, 1H), 7.79 (d, J = 1.5 Hz, 1H), 8.22 (m, 1H); MS: m/z 365 (MH+).
Computations. Kinetics analysis was performed using the kinetic module of SigmaPlot (SPSS Science Inc., Chicago, IL). The energies of the lowest unoccupied molecular orbitals (LUMOs) were computed using the Spartan '02 (Wavefunction Inc., Irvine, CA). The molecular geometries of all the compounds were built using the standard bond lengths and angles within Spartan '02. Semiempirical orbital calculations were performed using the PM3 Hamiltonian.
Results
3-(Indol-1-yl)-1,2-benzisoxazole. Incubation of 1a (Fig. 1) with rat liver microsomes under anaerobic conditions resulted in a time-dependent formation of three main products eluting at 8.8, 10.6, and 12.4 min (Table 1). The compound eluting at 8.8 min had a parent (MH+) ion at m/z 237 that was 2 mass units higher than that of the substrate (m/z 235) and was identified as the amidine 2a, resulting from reductive cleavage of the N–O bond. Its structure was confirmed by comparing it to an authentic standard of 2a prepared by catalytic hydrogenation of 1a (Uno and Kurokawa, 1981). The product eluting at 10.6 had a parent (MH+) ion at m/z 118 and had a UV spectrum and retention time identical to that of an authentic sample of indole 4a. The product eluting at 12.4 min had a parent (MH+) ion at m/z 238 that was 3 mass units higher than that of the substrate. In negative ionization mode, it had a parent ion at m/z 236 and was partially fragmented in the source, giving a daughter ion at m/z 137 consistent with a salicylate ion. It was identified as 1-salicylindole (3a). The system remained catalytically active for several hours, and the time course of the reaction is qualitatively depicted in Fig. 2A. The amidine 2a was the first product formed, followed successively by 3a and indole after a lag period. Salicylic acid, the other product expected from hydrolysis of 3a, was not detected during the initial portion of the time course but was observed after the reaction was allowed to go to completion. This was attributed to poor recovery at low concentrations.
When incubated in reaction buffer in the absence of microsomes, 2a was rapidly hydrolyzed to 3a (Fig. 2A, inset). Hydrolysis of 3a to indole was comparatively much slower. Addition of rat liver microsomes slowed down hydrolysis of 2b but had no apparent effect on the hydrolysis of 3a (not shown). These results suggested that hydrolysis of 2a was nonenzymatic.
The effect of various incubation conditions on the reduction of 1a to 2a was examined (Table 2). No reaction was observed in the absence of microsomes or NADPH or when microsomes were boiled prior to initiating the reaction. The reaction was completely inhibited in the presence of CO and partially inhibited in the presence of ketoconazole (2.5 μM), a potent P450 enzyme inhibitor. Increased activity was observed when liver microsomes from phenobarbital-treated rats were used. Taken together, these data indicate that P450 enzymes are involved in the initial reduction step of 1a.
The rate of 1a reduction was substrate concentration-dependent and followed Michaelis-Menten-type kinetic behavior with Vmax = 0.34 ± 0.01 nmol/min/mg of protein and Km = 118 ± 11 μM (Fig. 3, Table 3).
6-Chloro-3-(indol-1-yl)-1,2-benzisoxazole (1b). Incubation of 1b with rat liver microsomes under anaerobic conditions resulted in the formation of three main products eluting at 10.9, 12.0, and 13.0 (Table 1). The product eluting at 10.9 min was identified as indole (vide supra). The product eluting at 12.0 min had a parent (MH+) ion at m/z 271 that was 2 mass units higher than that of the substrate (m/z 269) and was identified as the amidine 2b. The product eluting at 13.0 min had a weak parent (MH+) ion at m/z 271 that was 3 mass units higher than that of the substrate. In negative ionization mode it had an abundant parent (M - H-) ion at m/z 269 and was identified as the salicylindole 3b. In addition, there was a minor peak that eluted at 9.8 min that had a UV spectrum similar to that of 1,2-benzisoxazole with λmax = 300 nm. In negative ionization mode, it gave a parent (M - H-) ion at m/z 152 with an isotopic ion at 154 in a 3:1 ratio. It was identified as 6-chloro-1,2-benzisoxazole (5a). The time course of the reaction is qualitatively shown in Fig. 2B. The time course of 5a appearance is shown in Fig. 2B inset, and reveals that 5a formation is concomitant with 1b disappearance. The amount of 5a formed was estimated at less than 2% based on its extinction coefficient (Casey et al., 1973). The kinetic parameters of 1b reduction are given in Table 3.
3-(Indol-1-yl)-6-methylsulfonyl-1,2-benzisoxazole (1c). Incubation of 1c with rat liver microsomes under anaerobic conditions gave rise to the formation of three main products eluting at 9.5, 10.6, and 10.9 min (Table 1). The product eluting at 9.5 min had a parent (MH+) ion at m/z 315 that was 2 mass units higher than that of the substrate (m/z 313) and was identified as the amidine 2c. The product eluting at 10.6 min was indole (4a). The product eluting at 10.9 min had a parent (M - H-) ion at m/z 314 and was identified as the salicylindole 3c. The overall time course of the reaction was essentially the same as that observed for 1b except that it was faster and no peak corresponding to 6-methylsulfone-1,2-benzisoxazole could be detected. The kinetic parameters of 1c reduction are given in Table 3.
1-(3-Carboxaldehydeindol-1-yl)-1,2-benzisoxazole (1d). Incubation of 1d with rat liver microsomes under anaerobic conditions resulted in the formation of 4 main products eluting at 7.9, 9.3, 9.9, and 12.5 min (Table 1). The product eluting at 7.9 min had a parent (MH+) ion at m/z 146 and coeluted with an authentic standard of indole-3-carboxaldehyde (4b). The product eluting at 9.3 min had a parent (M - H-) ion at m/z 170 with an isotopic ion at 172 in a 3:1 ratio and was identified as 4-chlorosalicylamide. The product eluting at 9.9 min had a parent (MH+) ion at m/z 152 and was identified as 6-chloro-1,2-benzisoxazole (vide supra). The product eluting at 12.5 min had a parent (MH+) ion at m/z 299 that was 2 mass units higher than that of 1d and was identified as the amidine 2d. The time course of the reaction is qualitatively shown in Fig. 2C. The amount of 5a formed was estimated at 40% of the initial 1d. The kinetic parameters of 1d reduction are given in Table 3.
3-(2-Chloro-3-methylsulfoxyindol-1-yl)-6-chloro-1,2-benzisoxazole (1e). Incubation of 1e with rat liver microsomes under anaerobic conditions resulted in the formation of only two products eluting at 7.0 and 9.9 min (Fig. 4 and Table 1). The product eluting at 7.0 min had a parent (MH+) ion at m/z 214 with an isotopic ion at 216 in a 3:1 ratio and a UV spectrum similar to that of indole with λmax = 271 nm and was identified as 2-chloro-3-methylsulfoxy-indole (4c). The product eluting at 9.9 min had a parent (MH+) ion at m/z 152 and was identified as 6-chloro-1,2-benzisoxazole (5a, vide supra). The time course of the reaction is qualitatively shown in Fig. 2D. Formation of 4c and 5a occurred simultaneously and was concomitant to the disappearance of the substrate 1e.
Structure-Activity Relationship. The ELUMO for 1a-e was computed by a semiempirical method using the PM3 Hamiltonian (Table 2) and plotted against the natural logarithm of Vmax for substrate reduction (Fig. 5). There was a strong negative correlation between ELUMO and ln(Vmax) with the following correlation equation:
Discussion
The reductive biotransformation of zonisamide (1,2-benzisoxazole-3-methanesulfonamide) has been postulated to go through an intermediate imine (Stiff et al., 1992). Although the imine itself has never been isolated, due to its inherent instability and rapid hydrolysis in aqueous media, evidence for this mechanism was provided by Sugihara et al. (1996) when they demonstrated stoichiometric production of ammonia and 2-sulfamoylacetylphenol from zonisamide. We report here that, when incubated with rat liver microsomes under anaerobic conditions, 1a is first metabolized to the amidine 2a followed by a nonenzymatic hydrolysis to 1-salicylindole (3a) (Fig. 2A), providing further evidence that 1,2-isoxazoles are first reductively cleaved to an imine intermediate.
The reduction of 3-(indol-1-yl)-1,2-benzisoxazoles was particularly sensitive to the nature and positioning of substituents (Fig. 6). Electron-withdrawing substituents at the 6-position of the 1,2-benzisoxazole ring caused an increase in the rate of reduction. Electron-withdrawing substituent(s) on the indole moiety likewise increased the rate of reduction but also caused significant changes in the pathway of the reaction with a switch in regioselectivity from N–O bond cleavage to C–N bond cleavage. Although oxidative C–N bond cleavage is quite common, to our knowledge, reductive C–N bond cleavage has not been reported yet.
Reductive metabolism is likely to occur in the venous portion of the liver where the oxygen tension is sufficiently low. However, it is not clear at this time to what extent hepatic cytochrome P450 enzymes are contributing to the N-dearylation of 3-(indol-1-yl)-1,2-benzisoxazoles observed in vivo in the rat. Other biological systems, such as aldehyde oxidase or gut microflora, that are able to catalyze the reduction of 1,2-isoxazole, as has been demonstrated for zonisamide (Sugihara et al., 1996; Kitamura et al., 1997), could be involved as well.
Despite the number of reported cases of 1,2-isoxazole reductive metabolism, little is known about the mechanism by which the isoxazole ring is reduced in biological systems. The overall reaction is equivalent to a hydrogenation, but the involvement of cytochrome P450 suggests that the reaction occurs via two distinct electron transfers to the isoxazole ring. The present study led us to propose a mechanism accounting for the substituent-dependent difference in reaction pathway (Fig. 7). Initial electron transfer to the 1,2-isoxazole ring would yield a radical anion. Electronic consideration would locate the radical on the stabilizing benzylic position with the more electronegative nitrogen bearing the negative charge. Because the N–O bond is particularly prone to homolysis, rearrangement to the phenoxy radical is likely and would yield the imine intermediate. Alternatively, the radical anion can lose the indole group via a heterolytic C–N bond cleavage giving the benzisoxazolyl radical as the other product of the reaction. This second pathway is expected to be favored by electron-withdrawing groups on the indole ring that decrease the pKb of the indole nitrogen, making it a better leaving group.
Reductive desisoxazolylation may be a general feature for 1,2-isoxazoles having a leaving group at the 3-position. The same property may also be true for the sulfur-containing counterparts, 1,2-isothiazoles. For example, the antipsychotic drug ziprasidone contains a 3-piperidyl-1,2-benzisothiazole moiety that undergoes a unique N- desbenzisothiazolylation in humans. A mechanism was proposed that involved hydration of the C=N bond followed by hydrolysis (Prakash et al., 1997a–c). The present study suggests an alternate route with direct cleavage of the C–N bond. Although the piperidinyl anion is a poor leaving group, it is possible that it is actually the protonated form of ziprasidone that undergoes reductive C–N bond cleavage. It is also expected that protonation of the nitrogen linked to the benzisothiazole would considerably lower the ELUMO and, as a consequence, increase the rate of reduction.
We have shown here that there is a significant correlation between the rate of 3-(indol-1-yl)-1,2-benzisoxazole reduction and the ELUMO determined by a semiempirical method (Fig. 5). The negative correlation indicates that when the LUMO energy decreases, the rate of reduction increases, and therefore, the ELUMO could be considered as an index of the susceptibility of 1,2-isoxazoles to accepting an electron. These observations also suggest that the rate-limiting step in the reduction of 1,2-isoxazole by P450 might be the initial electron transfer from the reduced heme Fe2+ to the substrate and that the ELUMO might be used to predict the competency of P450 enzymes to reduce 1,2-isoxazoles within a structural class. For example, the ELUMO of zonisamide calculated using the PM3 Hamiltonian was -24.6 kcal/mol. Using eq. 1, Vmax for the reductive metabolism of zonisamide by rat liver microsomes was calculated to be 0.67 nmol/nmol P450/min, which compared favorably with the published rate of 0.83 nmol/nmol P450/min at 400 μM substrate concentration (Nakasa et al., 1993).
In summary, we have found that 3-(indol-1-yl)-1,2-benzisoxazoles are reductively N-dearylated by rat liver microsomes under anaerobic conditions via two different, substituent-dependent pathways. N–O bond cleavage followed by two hydrolytic steps is predominant with unsubstituted indole, whereas C–N bond cleavage between the indole and 1,2-benzisoxazole group is favored in the presence of electron-withdrawing substituents on the indole moiety. The overall rate of substrate reduction was correlated with the semiempirically calculated ELUMO and suggested that ELUMO may be used to determine the susceptibility of 1,2-isoxazole-containing compounds to reductive metabolism.
Footnotes
-
↵1 Abbreviations used are: P450, cytochrome P450; HPLC, high-pressure liquid chromatography; DMF, dimethylformamide; LUMO, lowest unoccupied molecular orbitals.
- Received January 22, 2003.
- Accepted May 14, 2003.
- The American Society for Pharmacology and Experimental Therapeutics