Free Radicals Produced during the Oxidation of Hydrazines by

Free Radicals Produced during the Oxidation of Hydrazines by Hypochlorous Acid ... Publication Date (Web): November 27, 1996 ... The reaction between ...
0 downloads 7 Views 152KB Size
Download PDF
Chem. Res. Toxicol. 1996, 9, 1333-1339

1333

Free Radicals Produced during the Oxidation of Hydrazines by Hypochlorous Acid Douglas C. Goodwin, Steven D. Aust, and Thomas A. Grover* Biotechnology Center, Utah State University, Logan, Utah 84322-4700 Received July 1, 1996X

Hypochlorous acid (HOCl) derived from activated neutrophils and monocytes has been implicated in the activation of hydrazine-containing drugs to toxic intermediates. However, reactive intermediates formed during the reaction between HOCl and these drugs have not been identified. We investigated the oxidation of the hydrazine derivatives isoniazid, iproniazid, and hydralazine by HOCl. The reaction between HOCl and all three hydrazines resulted in O2 consumption, indicating that free radicals were produced, but the rate and extent of O2 consumption were different for each hydrazine. Moreover, reduction of nitroblue tetrazolium (NBT) was observed only during the reaction between HOCl and isoniazid, suggesting that different radical species may be produced from HOCl reaction with each hydrazine. The oxidation of iproniazid by HOCl in the presence of the radical trap 5,5-dimethyl-1-pyrroline N-oxide (DMPO) resulted in the formation of a carbon-centered radical adduct. In contrast, the reaction between HOCl and hydralazine resulted in the formation of a nitrogen-centered DMPO radical adduct. The oxidation of isoniazid by HOCl resulted in the formation of two oxygen-centered radical adducts, DMPO-OOH and DMPO-OH. Myeloperoxidase-catalyzed oxidation of these hydrazines in the presence of Cl- and H2O2 produced radical species that were identical to those observed with HOCl. Thus, some of the toxic side effects of these drugs may be the result of the production of free-radical intermediates from reaction with neutrophilderived oxidants, such as HOCl. The types of radicals produced and the consequences of generating these reactive species are discussed. The oxidation of isoniazid and other hydrazines to reactive intermediates has been the subject of intense investigation (1-7). Isoniazid is a compound used in the treatment of tuberculosis (3-8). It has been proposed that oxidation of isoniazid by mycobacterial catalase/ peroxidases results in the antibiotic effects of this medication (3, 4). Indeed, it has been shown that a large proportion of antibiotic resistant strains of Mycobacterium tuberculosis lack the catalase/peroxidase responsible for activation of isoniazid (8). Other hydrazines such as hydralazine and iproniazid have also been used as medications in treatment of other clinical disorders such as hypertension and depression (9). Hydrazine-containing drugs produce a number of toxic side effects (9-11). Many of these hydrazines are known to cause severe liver damage, and some have been implicated as carcinogenic agents in animals (12, 13). In addition to these hepatotoxic and possible carcinogenic side effects, many hydrazines are known to cause a relatively high incidence of drug-induced lupus (10, 11, 13-16). Jiang et al. have suggested that activated neutrophils use myeloperoxidase to transform a number of lupus-inducing drugs to cytotoxic intermediates (16). Myeloperoxidase is known to produce HOCl in the presence of H2O2 and Cl- (17). The involvment of myeloperoxidase in the toxicity of hydrazine-containing drugs suggests that HOCl may be one neutrophil-derived oxidant responsible for activation of these drugs to potentially cytotoxic intermediates. The reaction of HOCl with isoniazid and other hydrazines, however, has not been well characterized. Thus, * To whom correspondence should be addressed. Phone (801) 7972710; FAX (801) 797-2766; E-mail [email protected]. X Abstract published in Advance ACS Abstracts, November 1, 1996.

S0893-228x(96)00108-7 CCC: $12.00

while end products of these reactions have been determined, reactive intermediates produced by such reactions have not yet been identified (10, 11, 15). Although these intermediates remain unidentified, recent work with hydralazine by another group suggests that a diazene and a diazonium salt may be reactive intermediates formed during hydralazine oxidation by HOCl (18). The oxidation of hydrazine-containing drugs in a number of other biological systems results in the formation of a number of free-radical species (1, 2). It is reasonable to suggest, therefore, that free radicals may also be produced during oxidation of these chemicals by HOCl. We present evidence which indicates that a number of free radical species are produced as a result of the oxidation of hydrazines by HOCl. Moreover, these radicals are also produced when HOCl is replaced by myeloperoxidase/Cl-/H2O2. This suggests that these chemicals may be oxidized to free radicals as well as other reactive intermediates by activated neutrophils in vivo. Free radicals formed in this manner may participate in the myeloperoxidase-mediated toxic side effects of isoniazid and other hydrazine-containing drugs.

Materials and Methods Caution. Some of the chemicals used in this study are hazardous and should be handled carefully. Proper protection should be used while these chemicals are handled. Chemicals. Hypochlorous acid was purchased from Mallinckrodt (Paris, KY). Diethylenetriaminepentaacetic acid (DETAPAC),1 superoxide dismutase (SOD) from bovine erythrocytes, 1 Abbreviations: NBT, nitroblue tetrazolium; DMPO, 5,5-dimethyl1-pyrroline N-oxide; SOD, superoxide dismutase; DETAPAC, diethylenetriaminepentaacetic acid; MNP, 2-methyl-2-nitrosopropane; NAC, N-acetylcysteine.

© 1996 American Chemical Society

1334 Chem. Res. Toxicol., Vol. 9, No. 8, 1996 myeloperoxidase, nitroblue tetrazolium (NBT), isoniazid, hydralazine, iproniazid, DMPO, and 2-methyl-2-nitrosopropane (MNP) were purchased from Sigma (St. Louis, MO). All of these chemicals were used without further purification except DMPO. Impurities in DMPO were removed using activated charcoal according to the method described by Thornalley and Bannister (19). Stock solutions of MNP were prepared by adding 5 mg of MNP to 500 µL of acetone. This solution was then diluted by addition of 500 µL of acetate buffer, pH 5.0. All solutions were prepared using purified water (Barnstead NANOpure II system; specific resistance 18 Mohm/cm). Buffers and KCl were treated with Chelex 100 (Bio-Rad, Richmond, CA) prior to use. Concentrations of HOCl were determined spectrophotometrically at 292 nm using an extinction coefficient of 350 M-1 cm-1 (20, 21). Similarly, concentrations of DMPO were measured at 234 nm using an extinction coefficient of 7700 M-1 cm-1 (19). Myeloperoxidase activity was determined according to the method of Desser and co-workers (22). Oxygen Consumption. Changes in O2 concentration during reactions between each hydrazine and HOCl were monitored using a Gilson 5/6 oxygraph (Middleton, WI) equipped with a Clark type O2 sensitive electrode. All reactions were carried out at room temperature (24 °C) in a water jacketed, 1.8 mL reaction chamber. All reactions were initiated by the addition of HOCl. Other reaction conditions are indicated in the figure legends. Nitroblue Tetrazolium Reduction. Reduction of NBT to monoformazan during reaction between HOCl and each hydrazine was monitored at 560 nm using an extinction coefficient of 1.5 × 104 M-1 cm-1 (23). Reactions were carried out using a three-syringe stopped-flow spectrophotometer from KinTek Instruments (State College, PA). The instrument was in the single mixing mode such that HOCl and each hydrazine derivative were mixed just prior to entering the observation cell. HOCl was placed in the first syringe. The second syringe contained buffer, and the third syringe contained NBT and either isoniazid, iproniazid, or hydralazine. SOD, when present, was placed in the third syringe. ESR Spin-Trapping. All spin-trapping was performed using DMPO or MNP. When MNP was used as a spin trap, its final concentration was 0.5 mg/mL (2.5% final acetone concentration). All reactions were carried out at room temperature. For experiments with HOCl, each reaction was initiated by addition of HOCl, while experiments with myeloperoxidase were initiated with H2O2. Reactions used to study myeloperoxidasecatalyzed iproniazid oxidation contained 0.25 U/mL myeloperoxidase and were allowed to incubate for 2 h prior to collection of spectra. This long incubation time was not feasible with hydralazine and isoniazid due to the relatively rapid decay rate of the DMPO spin adducts obtained during oxidation of these two compounds. Thus, experiments involving these two chemicals were performed with higher myeloperoxidase concentrations (2 U/mL) and little or no incubation time. There was no incubation for reactions involving HOCl. A Bruker ECS-106 spectrometer (Billerica, MA) was used for all ESR experiments. The spectrometer was operated at 9.72 GHz with a 50 kHz modulation frequency. Other spectrometer settings were as follows: gain, 2 × 104; microwave power, 20 mW; modulation amplitude, 0.946 G; time constant, 328 ms; sweep rate, 35.9 G/min; sweep width, 100 G (3430-3530 G).

Results and Discussion The reactions between HOCl and each of the hydrazines (hydralazine, iproniazid, and isoniazid) resulted in differing rates and extents of O2 consumption (Figure 1). The most rapid and extensive O2 consumption was observed with iproniazid. Much less O2 consumption was observed with the other hydrazines. No O2 consumption was observed when hydrazines were excluded from the reaction. These data suggested that free radicals may be produced as a result of the reactions of HOCl with

Goodwin et al.

Figure 1. Change in O2 concentration during HOCl reaction with various hydrazine-containing drugs. Reactions contained 1 mM HOCl, 50 mM acetate buffer, pH 5.0, and 1 mM hydralazine (open squares), 1 mM isoniazid (closed squares), 1 mM iproniazid (open triangles), or no hydrazine (closed circles). All reactions were initiated with HOCl at the time indicated by the arrow.

various hydrazine-containing drugs. Furthermore, the differences in rate and extent of O2 consumption suggested that different types of radicals may be produced by HOCl oxidation of each hydrazine. During O2 consumption experiments, we also observed that oxidation of hydrazines resulted in the production of gas bubbles. Because O2 was consumed during all of these reactions, it seemed unlikely that O2 gas was evolved. This gas did not appear to oxidize I-, excluding Cl2 as a possibility. It has been proposed that oxidation of hydrazines in other systems results in production of N2 (1, 24). This suggests that N2 gas may also be evolved during the oxidation of hydrazine-containing drugs by HOCl. The reduction of NBT was observed in reactions of HOCl with isoniazid, but no NBT reduction was detected with iproniazid or hydralazine (Figure 2). Furthermore, no NBT reduction was observed when either HOCl or isoniazid was omitted (data not shown). Addition of SOD to reactions containing NBT, isoniazid, and HOCl had no effect on NBT reduction. This would suggest that the NBT reduction observed in this reaction was due primarily to the initial formation of some reducing organic radical as opposed to O2•-. The O2•--independent reduction of NBT is often observed during the free-radical oxidation of isoniazid by peroxidases (5, 7). Our results here indicate that a similar radical may be produced during isoniazid oxidation by HOCl. The lack of NBT reduction during HOCl reaction with iproniazid and hydralazine indicated that the organic radicals produced in these reactions were not reducing in nature. HOCl oxidation of iproniazid in the presence of the spin trap DMPO produced a six-line ESR spectrum indicative of a carbon-centered radical (Figure 3A). The splitting constants obtained from the spectrum (aN ) 16.3 G; aHβ

Oxidation of Hydrazines by HOCl

Chem. Res. Toxicol., Vol. 9, No. 8, 1996 1335

Figure 2. Change in absorbance at 560 nm due to reduction of NBT during reactions between HOCl and various hydrazines. Reactions contained 100 µM HOCl, 1 mM NBT, 50 mM acetate buffer, pH 5.0, and 100 µM isoniazid (a), 100 µM hydralazine (b), 100 µM iproniazid (c), or 100 µM isoniazid and 50 µg of SOD (d). No change in absorbance was observed if isoniazid, HOCl, or NBT was excluded from the reaction.

Figure 4. ESR spectra of free-radical intermediates trapped by DMPO during HOCl reaction with hydralazine. All reactions contained 100 mM DMPO and 1 mM DETAPAC. Spectrum A was obtained with 1 mM HOCl and 1 mM hydralazine in the presence of 50 mM acetate buffer, pH 5.0. Spectrum B was obtained under conditions identical to those used for spectrum A except that 50 mM phosphate buffer, pH 7.0, was used in place of acetate buffer. Spectrum C was obtained under identical conditions as those used for spectrum A except that HOCl was excluded from the reaction.

Figure 3. ESR spectra of free-radical intermediates detected by spin-trapping during HOCl oxidation of iproniazid. All reactions contained 100 mM DMPO, 1 mM DETAPAC, and 50 mM acetate buffer, pH 5.0. Spectrum A was obtained with 1 mM HOCl and 1 mM iproniazid. Spectrum B was obtained under the same conditions as spectrum A except that MNP (0.25 mg) was used in place of DMPO. Spectrum C was obtained under identical conditions to spectrum A except that iproniazid was excluded. Similarly, spectrum D was obtained as was spectrum A except that HOCl was excluded.

) 24.7 G) were in close agreement with those obtained by Sinha for oxidation of iproniazid by horseradish peroxidase and H2O2 (1). Sinha proposed that the radical trapped during iproniazid oxidation by horseradish peroxidase was an isopropyl radical (1). This suggests that isopropyl radical may be one of the radicals produced during iproniazid oxidation by HOCl. This was confirmed by the ESR spectrum obtained when the spin trap MNP was used in place of DMPO (Figure 3B). The

“triplet of doublets” signal is indicative of a carboncentered radical with a single β hydrogen. Moreover, the hyperfine splitting constants (aN ) 16.8 G; aHβ ) 2.0 G) were consistent with that previously obtained for isopropyl radical (1). Production of a carbon-centered radical such as this would explain the remarkable rate and extent of O2 consumption and the apparent lack of NBT reduction observed in reactions containing HOCl and iproniazid. The inclusion of SOD in the HOCl/iproniazid reaction mixture with DMPO had little effect on the resulting ESR spectrum (data not shown). This indicated that formation of this radical was not the result of a chain reaction initiated by O2•-. The carbon-centered DMPO radical adduct was not observed in the absence of HOCl or iproniazid (Figure 3C,D). When HOCl was excluded, no signal was observed. When iproniazid was absent, a weak uncharacterized spectrum was obtained. The width of the signal suggested that it was not a carbon-centered adduct. Moreover, the width of the signal was too broad to be that of DMPO-X. It may be the result of DMPO oxidation by HOCl; however, we were unable to determine the identity of this adduct from previously observed DMPO oxidation products (25). The reaction between hydralazine and HOCl resulted in the formation of a nitrogen-centered DMPO radical adduct (Figure 4). The splitting constants obtained (aN ) 14.3 G; aHβ ) 18.5 G; aNβ ) 3.2 G) were similar to those calculated for DMPO adducts detected during hydralazine oxidation by horseradish peroxidase at pH 5.0 (1). The addition of SOD had little effect on the intensity and shape of this spectrum (data not shown). This indicated that O2•- was not involved in the formation of this

1336 Chem. Res. Toxicol., Vol. 9, No. 8, 1996

Goodwin et al.

that •OH was being generated during the reaction between HOCl and isoniazid. To determine if the DMPO-OOH component of the spectrum in Figure 5A was the result of O2•- production or formation of an alkylperoxyl radical, SOD was added to the reaction. Addition of SOD resulted in the loss of the DMPO-OOH component of the spectrum, leaving only a weak DMPO-OH spectrum (Figure 5C). Conversely, addition of boiled SOD to the reaction had no effect, as both DMPO-OOH and DMPO-OH were still observed (Figure 5D). These data suggested that O2•- was trapped by DMPO as opposed to an alkylperoxyl radical. The formation of •OH during isoniazid oxidation by HOCl is most puzzling. Transition metals are known to catalyze the formation of •OH through the Fenton reaction. However, the presence of DETAPAC in each of our experiments should prevent these metal-catalyzed reactions (26). Alternatively, it has been proposed that the one-electron reduction of HOCl results in formation of • OH (20). Some investigators have suggested that O2•may reduce HOCl to yield •OH and Cl-, as shown in the following reaction (20, 28, 29):

HOCl + O2•- f •OH + Cl- + O2 Figure 5. ESR spectra of free-radical intermediates trapped by DMPO during HOCl reaction with isoniazid. All reactions were carried out in 50 mM acetate buffer, pH 5.0, with 100 mM DMPO and 1 mM DETAPAC. Spectrum A was obtained by reacting 1 mM HOCl with 1 mM isoniazid. Conditions for all other spectra (B-E) were identical to those used for spectrum A with the following exceptions: ethanol (10%) was included (B); SOD (50 µg) was included (C); boiled SOD (50 µg) was added (D); HOCl was excluded (E).

nitrogen-centered radical. The hyperfine splitting constants (aN ) 15.1 G; aHβ ) 16.7 G; aNβ ) 2.6 G) calculated from the spectrum obtained at pH 7.0 (Figure 4B) were considerably different than those calculated from the spectrum collected at pH 5.0 (Figure 4A). Sinha also observed similar spectral changes when hydralazine was oxidized by horseradish peroxidase at various pH (1). These spectral changes are likely the result of protonation of the DMPO spin adduct. Recent investigations by Hofstra and Uetrecht into hydralazine oxidation by HOCl indicate that a diazene intermediate may be formed during this reaction. These investigators utilized Nacetylcysteine (NAC) as a trapping agent. The primary intermediate trapped using NAC was 1-phthalazylmercapturic acid (18). Together with our results, this suggests that a number of potentially harmful radical and nonradical intermediates are generated upon hydralazine oxidation by HOCl. When isoniazid was oxidized by HOCl in the presence of DMPO, two radical adducts were detected (Figure 5A). The spectrum appeared to be a mixed signal of DMPOOH and DMPO-OOH. It has been well established that DMPO-OOH can decompose to form DMPO-OH (26, 27). Thus, to determine whether the DMPO-OH signal was due to the formation of •OH or decomposition of DMPOOOH, a control reaction was performed in the presence of 10% ethanol. Under these conditions, the DMPO-OH component was no longer detected. Instead, a mixed spectrum of DMPO-OOH and a carbon-centered radical adduct was observed (Figure 5B). The hyperfine splitting constants for the carbon-centered component (aN ) 15.8, aHβ ) 22.8) were identical to those reported previously for DMPO-ethanol radical adducts (25). This indicated

(1)

Because O2•- is produced during isoniazid oxidation by HOCl, it is possible that this O2•- may react with HOCl to form •OH. Thus, it appears that several radical species may be produced during HOCl oxidation of isoniazid. From the spin-trapping studies, it is clear that both O2•- and •OH are produced. However, the reduction of NBT during isoniazid oxidation suggests the formation of a third radical species. It seems clear that most of the observed NBT reduction is not due to the formation of O2•-, but rather to an unidentified organic radical. Similarly, other investigators have noted the O2•--independent reduction of NBT during oxidation of isoniazid by a number of peroxidases (5-7). It has been proposed that the initial free-radical intermediate formed upon oxidation of isoniazid by peroxidase is that of a nitrogen-centered radical (1, 3, 4). Previous attempts to trap such a radical in these oxidation systems have been unsuccessful (1). Our results also suggest that this radical is highly unstable and/or does not react efficiently with DMPO as no nitrogen-centered DMPO adduct was observed. Conversely, the nitrogen-centered radical of hydralazine appears to be much more amenable to spin-trapping by DMPO. With both compounds, however, the decomposition of the initial nitrogen-centered radical appears to give rise to O2•-. Free radicals were also produced upon oxidation of these hydrazines by myeloperoxidase in the presence of Cl- and H2O2. Oxidation of iproniazid by myeloperoxidase/ Cl-/H2O2 in the presence of DMPO (Figure 6A) resulted in a six-line ESR spectrum with identical hyperfine splitting constants to those obtained with HOCl (Figure 3A). Similar to results obtained with HOCl (Figure 3), the carbon-centered radical was not detected in the absence of iproniazid (Figure 6B). However, a weak unidentified spectrum similar to that shown in Figure 3C was detected. When both iproniazid and Cl- were omitted (Figure 6C), no signal was observed. Moreover, detection of the carbon-centered radical required myeloperoxidase (Figure 6D). As with HOCl, oxidation of iproniazid by myeloperoxidase/Cl-/H2O2 resulted in production of the isopropyl radical as determined by spin-

Oxidation of Hydrazines by HOCl

Figure 6. ESR spectra of free-radical intermediates detected by spin-trapping during myeloperoxidase-catalyzed oxidation of iproniazid. Spectrum A was obtained using 0.25 U/mL myeloperoxidase, 500 µM H2O2, 100 mM KCl, 100 mM DMPO, 1 mM iproniazid, 1 mM DETAPAC, and 50 mM acetate buffer, pH 5.0. Conditions for all other spectra (B-F) were identical to spectrum A with the following exceptions: iproniazid was excluded (B); iproniazid and KCl were excluded (C); myeloperoxidase was excluded (D); MNP (0.5 mg/mL) replaced DMPO (E); MNP replaced DMPO, and myeloperoxidase was excluded (F).

trapping with MNP (Figure 6E). No MNP-isopropyl radical adduct was observed upon omission of myeloperoxidase (Figure 6F). It is important to mention that isopropyl and other radicals were detected in the absence of Cl- (data not shown), but the signal intensity was much greater when Cl- was present. This indicates that myeloperoxidase can carry out the direct oxidation of iproniazid. However, indirect oxidation of iproniazid through production of HOCl appears to be much faster. This is consistent with myeloperoxidase-catalyzed oxidation of a number of organic compounds. Isoniazid oxidation by myeloperoxidase/Cl-/H2O2 resulted in the formation of DMPO-OH and DMPO-OOH (Figure 7A). The observed spectrum was identical to that obtained during isoniazid oxidation by HOCl (see Figure 5). A similar but much weaker signal was observed in the absence of Cl- (Figure 7B). Consistent with results obtained with HOCl, the addition of ethanol or SOD to the myeloperoxidase/Cl-/H2O2 reaction mixture confirmed the generation of hydroxyl and O2•- radicals. In the presence of ethanol, the DMPO-OH component of the ESR spectrum was not observed, but was replaced with a carbon-centered radical adduct spectrum (Figure 7C). In the presence of SOD, the DMPO-OOH adduct spectrum was not observed (Figure 7D). No radicals were detected in the absence of myeloperoxidase (Figure 7E). Oxidation of hydralazine by myeloperoxidase/Cl-/H2O2 resulted in production of a nitrogen-centered radical (Figure 8A). This DMPO-radical adduct was identical to that observed during hydralazine oxidation by HOCl (see Figure 4A). The same radical was trapped in the absence of Cl- (Figure 8B). Interestingly, the greatest

Chem. Res. Toxicol., Vol. 9, No. 8, 1996 1337

Figure 7. ESR spectra of free-radical intermediates detected by spin-trapping during myeloperoxidase-catalyzed oxidation of isoniazid. Spectrum A was obtained using 2 U/mL myeloperoxidase, 500 µM H2O2, 100 mM KCl, 100 mM DMPO, 1 mM isoniazid, 1 mM DETAPAC, and 50 mM acetate buffer, pH 5.0. All other spectra (B-E) were collected under identical conditions to A with the following exceptions: KCl was excluded (B); ethanol (5%) was included (C); SOD (50 µg) was included (D); myeloperoxidase was excluded (E).

Figure 8. ESR spectra of free-radical intermediates trapped by DMPO during myeloperoxidase-catalyzed oxidation of hydralazine. Spectrum A was obtained using 2 U/mL myeloperoxidase, 500 µM H2O2, 100 mM KCl, 100 mM DMPO, 1 mM isoniazid, 1 mM DETAPAC, and 50 mM acetate buffer, pH 5.0. KCl was excluded from the reaction used to obtain spectrum B, and myeloperoxidase was excluded from the reaction to collect spectrum C.

signal intensity was observed when Cl- was absent from the reaction. This indicates that hydralazine is oxidized

1338 Chem. Res. Toxicol., Vol. 9, No. 8, 1996

to free radical intermediates by myeloperoxidase in the presence or absence of Cl-, but direct oxidation by myeloperoxidase appears to be more efficient than indirect oxidation using HOCl. Production of these radicals by HOCl or myeloperoxidase/Cl-/H2O2 may have several implications for the mechanism of immunotoxicity of these drugs. It is known that the incidence of drug-induced lupus is higher with hydralazine than with isoniazid (16, 17, 19). It would appear that iproniazid is not associated with this particular side effect. Other investigators have previously observed that the presence of a free hydrazyl group is an important factor in the ability of hydralazine to cause drug-induced lupus and other disorders (10, 13). Indeed, individuals who are able to rapidly acetylate hydralazine have much lower plasma concentrations of this drug and do not suffer from drug-induced lupus (30, 31). Conversely, those who acetylate hydralazine slowly have higher plasma hydralazine concentrations, and these individuals develop hydralazine-induced lupus (29, 30). The high incidence of drug-induced lupus observed with hydralazine therapy may be due to the production of a relatively stable hydrazyl radical. Modification of proteins [e.g., major histocompatibility complex class II or histones (18)] or other cellular components by hydralazyl radical or a subsequent intermediate may result in the production of autoantibodies, a key event in the progression of drug-induced lupus (14). Interestingly, it has been suggested that production of hydralazyl radical in other systems results in DNA strand breaks and modification of DNA bases (32, 33). This has been suggested as the reason for the potential carcinogenicity of hydralazine. Thus, modification of DNA and other cellular components by hydralazyl radicals or other intermediates produced during oxidation by HOCl could be a factor in a number of the toxic side effects of hydralazine. While isoniazid is also oxidized to a hydrazyl radical, the short lifetime of this species may lessen the ability of this chemical to cause the cellular damage that leads to drug-induced lupus. Conversely, iproniazid does not have a free hydrazyl group. Thus, as suggested by Sinha, any nitrogen-centered radical formed from this compound would quickly rearrange to yield an isopropyl radical, possibly through an oxygen-centered radical intermediate (1).

O C

N

• NHNCH(CH )

3 2

•

O HCN

NCH(CH3)2

N O CH

+ • CH(CH3)2 + N2 N

Although generation of an isopropyl radical would result in cellular damage, it is likely that the deleterious effects would be different than those caused by a hydrazyl radical. These observations suggest that hydrazyl radical intermediates formed by HOCl oxidation of hydralazine and isoniazid may be responsible for the observed immunotoxicity of these drugs.

Goodwin et al.

In conclusion, it is clear that the reaction of HOCl with various hydrazine-containing drugs results in the formation of reactive free-radical species. Moreover, these same radicals can be generated by myeloperoxidase in the presence of Cl- and H2O2. It is also evident that the differing structures of each of the hydrazines give rise to different free radical species. It is possible that oxidation of these drugs by HOCl derived from myeloperoxidase of activated neutrophils may lead to production of these and other free radicals in vivo. Production of these reactive free-radical intermediates by these cells may explain, in part, many of the toxic side effects of these medications, including drug-induced lupus.

Acknowledgment. This work was supported by NIEHS Grants ESO4922 to S.D.A. and T.A.G. and ESO5056 to S.D.A. The authors thank Curtis Takemoto and Randy Booth for excellent technical assistance, and Terri Maughan for excellent secretarial assistance.

References (1) Sinha, B. K. (1983) Enzymatic activation of hydrazine derivatives: a spin-trapping study. J. Biol. Chem. 258, 796-801. (2) Sinha, B. K., and Motten, A. G. (1982) Oxidative metabolism of hydralazine: evidence for nitrogen centered radicals formation [sic]. Biochem. Biophys. Res. Commun. 105, 1044-1051. (3) Johnsson, K., and Schultz, P. G. (1994) Mechanistic studies of the oxidation of isoniazid by catalase peroxidase from Mycobacterium tuberculosis. J. Am. Chem. Soc. 116, 7425-7426. (4) Johnsson, K., King, D. S., and Schultz, P. G. (1995) Studies on the mechanism of action of isoniazid and ethionamide in chemotherapy of tuberculosis. J. Am. Chem. Soc. 117, 5009-5010. (5) Shoeb, H. A., Bowman, B. U., Jr., Ottolenghi, A. C., and Merola, A. J. (1985) Peroxidase-mediated oxidation of isoniazid. Antimicrob. Agents Chemother. 27, 404-407. (6) Shoeb, H. A., Bowman, B. U., Jr., Ottolenghi, A. C., and Merola, A. J. (1985) Evidence for the generation of active oxygen by isoniazid treatment of extracts of Mycobacterium tuberculosis H37Ra. Antimicrob. Agents Chemother. 27, 404-407. (7) Hillar, A., and Loewen, P. C. (1995) Comparison of isoniazid oxidation catalyzed by bacterial catalase-peroxidases and horseradish peroxidase. Arch. Biochem. Biophys. 323, 438-446. (8) Zhang, Y., Heym, B., Allen, B., Young, D., and Cole, S. (1992) The catalase-peroxidase gene and isoniazid resistance of Mycobacterium tuberculosis. Nature 358, 591-593. (9) Juchau, M. R., and Horita, A. (1972) Metabolism of hydrazine derivatives of pharmacologic interest. Drug Metab. Rev. 1, 71100. (10) Uetrecht, J. (1990) Drug metabolism by leukocytes and its role in drug-induced lupus and other idiosycratic drug reactions. Crit. Rev. Toxicol. 20, 213-235. (11) Uetrecht, J. P. (1992) The role of leukocyte-generated reactive metabolites in the pathogenesis of idiosyncratic drug reactions. Drug Metab. Rev. 24, 299-366. (12) Toth, B. (1978) Tumorigenic effect of 1-hydrazinophthalazine hydrochloride in mice. J. Natl. Cancer Inst. 61, 1363-1365. (13) Shaw, C. R., Butler, M. A., Thenot, J. P., Haegle, K. D., and Matney, T. S. (1979) Genetic effects of hydralazine. Mutat. Res. 68, 79-84. (14) Perry, H. M. (1973) Late toxicity to hydralazine resembling systemic lupus erythematosus or rheumatoid arthritis. Am. J. Med. 54, 58-72. (15) Hofstra, A. H., Li-Muller, S. M. A., and Uetrecht, J. P. (1992) Metabolism of isoniazid by activated leukocytes: possible role in drug-induced lupus. Drug Metab. Dispos. 20, 205-210. (16) Jiang, X., Khursigara, G., and Rubin, R. L. (1994) Transformation of lupus-inducing drugs to cytotoxic products by activated neutrophils. Science 266, 810-813. (17) Harrison, J. E., and Schultz, J. (1976) Studies on the chlorinating activity of myeloperoxidase. J. Biol. Chem. 251, 1371-1374. (18) Hofstra, A. H., and Uetrecht, J. P. (1993) Reactive intermediates in the oxidation of hydralazine by HOCl: the major oxidant generated by neutrophils. Chem.-Biol. Interact. 89, 183-196. (19) Thornalley, P. J., and Bannister, J. V. (1985) The spin trapping of superoxide radicals. In The Handbook of Methods for Oxygen Radical Research (Greenwald, R. A., Ed.) pp 133-136, CRC Press, Boca Raton, FL.

Oxidation of Hydrazines by HOCl (20) Folkes, L. K., Candeias, L. P., and Wardman, P. (1995) Kinetics and mechanisms of hypochlorous acid reactions. Arch. Biochem. Biophys. 323, 120-126. (21) Morris, J. C. (1966) The acid ionization constant of HOCl from 5 to 35°. J. Phys. Chem. 70, 3798-3805. (22) Desser, R. K., Himmelhoch, S. R., Evans, W. H., Januska, M., Mage, M., and Shelton, E. (1972) Guinea pig heterophil and eosinophil peroxidase. Arch. Biochem. Biophys. 148, 452-465. (23) Auclair, C., and Voisin, E. (1985) Nitroblue tetrazolium reduction. In Handbook of Methods for Oxygen Radical Research (Greenwald, R. A., Ed.) pp 123-132, CRC Press, Boca Raton, FL. (24) Ator, M. A., and Ortiz de Montellano, P. R. (1987) Protein control of prosthetic heme reactivity: reaction of substrates with the heme edge of horseradish peroxidase. J. Biol. Chem. 262, 15421551. (25) Buettner, G. R. (1987) Spin-trapping: ESR parameters of spin adducts. Free Radical Biol. Med. 3, 259-303. (26) Buettner, G. R., and Mason, R. P. (1990) Spin-trapping methods for detecting superoxide and hydroxyl free radicals in vitro and in vivo. In Methods in Enzymology (Packer, L., and Glazer, A. N., Eds.) Vol. 186, pp 127-133, Academic Press, New York. (27) Rosen, G. M., and Rauckman, E. J. (1984) Spin trapping of superoxide and hydroxyl radicals. In Methods in Enzymology (Packer, L., Ed.) Vol. 105, pp 198-209, Academic Press, New York.

Chem. Res. Toxicol., Vol. 9, No. 8, 1996 1339 (28) Long, C. A., and Bielski, B. H. J. (1980) Rate of reaction of superoxide radical with chloride-containing species. J. Phys. Chem. 84, 555-557. (29) Candeias, L. P., Patel, K. B., Stratford, M. R. L., and Wardman, P. (1993) Free hydroxyl radicals are formed on reaction between the neutrophil-derived species superoxide anion and hypochlorous acid. FEBS Lett. 333, 151-153. (30) Perry, H. M., Tan, E. M., Carmody, S., and Sakamoto, A. (1970) Relationship to acetyl transferase activity to antinuclear antibodies and toxic symptoms in hypertensive patients treated with hydralazine. J. Lab. Clin. Med. 76, 114-125. (31) Drayer, D. E., and Reidenberg, M. M. (1977) Clinical consequences of polymorphic acetylation of basic drugs. Clin. Pharmacol. Ther. 22, 251-258. (32) Yamamoto, K., and Kawanishi, S. (1991) Free radical production and site specific DNA damage induced by hydralazine in the presence of metal ions or peroxidase/hydrogen peroxide. Biochem. Pharmacol. 41, 905-914. (33) Runge-Morris, M., Wu, N., and Novak, R. F. (1994) Hydrazinemediated DNA damage: Role of hemoprotein, electron transport, and organic free radicals. Toxicol. Appl. Pharmacol. 125, 123-132.

TX960108L