Enzymatic oxidative activation and transformation of the antitumor agent mitoxantrone

Free Radical Biology & Medicine, Vol. 5, pp. 13-25, 1988 Printed in the USA. All rights reserved.

0891-5849/88 $3.00+ .00 © 1988 Pergamon Press plc

Original Contribution ENZYMATIC

OXIDATIVE ACTIVATION THE ANTITUMOR AGENT

AND TRANSFORMATION MITOXANTRONE

OF

PAWEL KOLODZIEJCZYK, KRZYSZTOF RESZKA, AND J. WILLIAM LOWN* Department of Chemistry, University of Alberta, Edmonton, Alberta, Canada, T6G 2G2

(Received 18 March 1987; Revised 14 August 1987; Accepted 10 September 1987)

Abstract--Ambient temperature incubation of the anticancer agent mitoxantrone with horseradish peroxidase and hydrogen peroxide converts it into a hexahydronaphtho[2,3-f]quinoxaline-7,12-dione in which one side chain has cyclized to the chromophore. The structure of this cyclic metabolite was secured by independent synthesis. This peroxidative conversion of mitoxantrone, the progress of which can be followed spectrophotometrically, is accompanied by formation of a free radical species. The EPR characteristics, and dependence on pH of the latter, suggest it exists as a radical cation. The enzymatic oxidation of mitoxantrone is totally irreversible. The purified cyclic metabolite is a substrate for the peroxidase affording the unstable fully oxidized diimino compound and this reaction is fully reversible upon addition of ascorbate or other biological reductants. Admixture of the fully oxidized diimino product with the reduced cyclic metabolite generates the corresponding radical cation species by disproportionation-comproportionation processes. Independent kinetic studies confirm that reaction of the peroxidase with the cyclic metabolite proceeds more rapidly than with mitoxantrone itself. A derivative of mitoxantrone, in which the side-chain secondary amine functions are acylated, generates a radical cation upon treatment with the peroxidase-H202 system but does not cyclize subsequently. Derivatives without phenolic hydroxyls or those in which the phenolic hydroxyls are blocked also undergo peroxidative reaction. These observations suggest that initial peroxidative attack occurs at the aromatic nitrogens of mitoxantrone. The possible relevance of these results to the anticancer action of mitoxantrone and the implications for suppression of lipid peroxidation in vivo are discussed. Keywords--Antitumor agent, Mitoxantrone, Anthracenediones, Oxidative activation, Metabolism, Horseradish peroxidase, Electron paramagnetic resonance

trone, MXH2) and its congener ametantrone 3'4 which display useful anticancer properties, but which are apparently significantly less cardiotoxic than doxorubicin. 5,6 Although the exact mechanism(s) by which the anthracyclines express their characteristic cardiotoxicity has not been fully elucidated, several lines of evidence implicate certain reactive oxygen species derived from reaction with oxygen following reductive enzymatic activation of these agents. 7-11 The anthracenedionebased drugs are substantially less prone to reductive metabolic activation than the anthracyclines by NADPH-dependent flavoprotein reductases. 12-~6 This may be related to the reduced risk of cardiotoxicity in the case of the anthracenedione agents. ~3-~6 Furthermore, anthracenediones inhibit hepatic microsomal NADPH-cytochrome P450 reductase activity and inhibit the peroxidation of microsomal lipids. 11.17.18The elucidation of the mechanism by which anthracene-

INTRODUCTION

The cumulative dose-dependent risk of cardiotoxicity associated with the clinical application of the widely used anthracycline anticancer agent doxorubicin t'2 has stimulated the development of alternative agents which may possess comparable antitumor efficacy but with diminished cardiotoxicity. Among the more promising new agents is 1,4-dihydroxy-5,8-bis[[2-(2-hydroxyethyl)amino]ethyl]-9,10-anthracenedione, (mitoxan-

*To whom correspondence may be addressed. The abbreviations used are: AA, ascorbic acid; HRP, horseradish peroxidase; MXH2, mitoxantrone, 1,4-dihydroxy-5,8-bis[[2-(2-hydroxyethyl)amino]ethyl]-9,-10-anthracenedione; MH2 and MH22÷, reduced and oxidized forms of the principal oxidation product of MXH2, NDA-MXH2 N,N'-diacetylmitoxantrone, 1,4-dihydroxy-5, 8 - bis[[2 -[(2 - acetoxyethyl)amino]ethyl]amino]-9,10- anthracenedione; HMNDA-MXH2, hexamethyl-N,N'-diacetylmitoxantrone. 13

14

P. KOLODZIEJCZYKet al.

diones inhibit lipid peroxidation may therefore be of potential therapeutic significance. Although mitoxantrone and related structures are evidently resistant to reductive enzymatic activation we recently presented evidence that they are subject to facile oxidative enzymatic action. Z9We herein report the enzymatically initiated transformation of mitoxantrone to a metabolite containing an additional ring and spectroscopic evidence for the existence of a redox system involving mitoxantrone and its oxidized form which is responsible for generation of a free radical intermediate. We also report enzymatic studies with certain mitoxantrone derivatives which serve to establish the structural features necessary for formation of the cyclic metabolite and free radical generation. The possible relevance of these results to the anticancer action of mitoxantrone and the implications for the suppression of lipid peroxidation in vivo are discussed.

M A T E R I A L S AND METHODS

Mitoxantrone was kindly supplied by Dr. K. C. Murdock of Lederle Laboratories, Pearl River, N.Y. Horseradish peroxidase, type VI (EC I . I I . I . 7 ) and catalase (EC 1.11.1.6) were from Sigma Chemical Company (St. Louis, MO). Stock solutions of the drug (2 mg/mL) and H202 (0. I and 0.01 M) were prepared by dissolving the compounds in distilled water. Enzyme stock solutions (0.2 mg/mL) were prepared in 0.1 M acetate buffer pH 5.0. Incubation mixtures containing MXH2, H20: and HRP, were prepared in 0.01 M and 0.1 M acetate buffers for the EPR and spectrophotometric experiments respectively, unless otherwise stated. Thermal denaturation of HRP was achieved by incubation of the enzyme solution at 100°C for 15 min. UV-VIS absorption spectra were recorded with a Hewlett-Packard HP8450A Diode Array Spectrophotometer using standard 1 cm quartz cuvettes. Samples for the EPR studies were introduced into a suprasil quartz flat cell and transferred quickly to the microwave cavity of the Bruker ER-400 EPR spectrometer, operating at 9.5 GHz with 100-kHz field modulation. A g-value was obtained by direct Fieldial measurements. Chromatographic separations were performed using Sephadex LH20-100 from Sigma Chemical Co., St. Louis, MO, or Kieselgel 60 (230-400 mesh) from Merck, Darmstadt, W. Germany. 1H-NMR spectra were obtained on Bruker WH200 and WH400 cryospectrometers. Mass spectra were determined on Associate Electrical Industries (AEI) MS-9 and MS-50 focusing high resolution mass spectrometers. The IR spectra were recorded on a Nicolet 7199 FT spectro-

photometer, and only the principal sharply defined peaks are reported. All experiments were performed at room temperature, unless otherwise stated.

Enzymatic oxidation of mitoxantrone: isolation and structure determination of cyclic metabolite (MH2) A solution of 14.2 mg (28.5 /1moles) of mitoxantrone, hydrogen peroxide (28.5/zmoles) and 0.18 mg of horseradish peroxidase in 500 mL of acetate buffer pH 5.0 was incubated at ambient temperature. An aliquot (1 ml) was withdrawn and the course of the reaction was monitored spectrophotometrically at 20-s intervals. The initial (1) and final (2) spectra are shown in Figure 1. When no further changes in the spectrum were detectable 1 ml of a 57 mM solution of ascorbic acid was added to the incubation mixture and final spectrum (3) (Fig. 1) was recorded. The reaction mixture was lyophilized and the dry residue was dissolved in 1 ml of methanol:water (1 : 1 v/v) and the product isolated by column chromatography on Sephadex LH20 (1 × 80 cm) using the aqueous solvent system as eluent affording 13 mg of MH: (93% yield); ]HNMR (DMSO-d6) ~: 3.08 (m, 2H, CH:); 3.20 (m, 2H, --CH2--),; 3.65 (m, 8H, --CH2--); 3.85 (m, 4H, ---CH2--); 4.5 (b.s., IH, aliphatic OH, D20 exchange); 4.95 (b. s., 1H, aliphatic OH, exchange); 6.18 (s, 1H, aromatic C3); 7.0 (q, 2H, aromatic C6 and C7); 9.0 (s, 2H, NH aliphatic); 10.88 (b.s., 1H, aromatic NH, exchange); 11.27 (b.s., 1H, NH aromatic, exchange); 13.55 (b.s., 1H, OH phenolic, exchange); 14.24 (b.s., 1H, OH phenolic exchange); IR (KBr) "max: 3385; 2975; 1562; 1335 and 1246; MS m/z (relative intensity) 442.1839(29) ]calculated for C22H26N406 442.1852]; 368.1238 (100), [M +1.2

i

i

i

i

i

i

i

1.0

0

0.8 0.6 \

~

/

0.4 0.2 \

0.0

400 '

'

50O '

'

do

'

~t

7

Wavelength (nm) Fig. 1. Absorption spectra resulting from the oxidation of mitoxantrone by HRP/H:O2 system. The incubation mixture contained MXI-I2, H202 and HRP in acetate buffer, pH 5.0. HRP was added after the first scan (1) ( ) and then spectra were measured every 2 0 - 3 0 s until the final spectrum (2) (----) was recorded. Addition of ascorbic acid (120/iM) resulted in spectrum (3) (- + - + - + ). See experimental for details.

Enzymatic oxidation of mitoxantrone

CH2NHCH2CH2OH)]; 355.1164 (78), [(M+H)CH2CH2NHCH2CH2OH]; 324.0974 (90), [(M + H)(CH2)2NH(CH2)2OH -t- CH2OH]; 308.0686 (50), [(M + H)(CH2)2NH(CH2)2OH + CHECH2OH)].

Synthesis of mitoxantrone metabolite (MH2) A solution of 0.11 g (0.4 mmole) of 2,3-dihydro1,4,5,8-tetrahydroxyanthraquinone and 0.125 g (1.2 mmoles) of 2[2(2-aminoethyl)amino]ethanol in 5 ml of pyridine was heated under reflux for 12 h. The solvent was removed in vacuo and the oily residue was submitted to a molecular sieve separation (Sephadex LH20) in a butanol: methanol: water (4: 1 : 1 v/v) solvent system. The product, thus obtained in the free amine form, was treated with 1 ml of 0.1 M methanolic solution of hydrogen chloride and, after removal of the solvents, it was subjected to chromatography on a Sephadex LH20 column (1 x 80 cm) in a methanol:H20 (1:1 v/v) mixture. Removal of the solvents in vacuo afforded 70 mg (34% yield) of MH2 which was identical with that from the enzymatic reaction.

Enzymatic oxidation of metabolite of mitoxantrone (MH2) A mixture containing 57 nmoles of mitoxantrone metabolite (MH2), 57 nmoles of hydrogen peroxide and 0.36/tg of horseradish peroxidase in 1 ml of 0.1 M acetate buffer pH 5.0 was incubated at ambient temperature. The initial absorption spectrum (1) changed within 2 min to the final one (2) (Fig. 2). Addition of ascorbic acid (60/~M) resulted in reversion to spectrum (3) identical with the initial spectrum (Fig. 2). Subsequent addition of further portions of hydrogen peri

1.4

i

i

i

i

i

i

1.2

15

oxide followed by ascorbic acid permitted repetition of this cycle of events several times.

Preparation of N, N '-diacetylmitoxantrone (NDA-MXH2) A solution of 44 mg (0.1 mmole) of mitoxantrone (free amine form) in 5 ml of methanol with 60/11 of triethylamine was set aside at ambient temperature for 2 h, during which time the progress of the reaction was monitored by TLC. When the reaction was completed the solvents were removed in vacuo and the product was purified by column chromatography on silica gel using chloroform:methanol (10: 1) as solvent system, affording pure NDA-MXH2 50 mg (95% yield) m.p. 168-170°C; ;slH-NMR (CDC13) ~: 2.07 (s, 6H, acetyl); 3.4-3.65 (m, 16H, --CH2--); 7.1 (s, 2H, aromatic C2 and Ca); 7.6 (s, 2H, aromatic C6 and C7); 10.5 (s, 1H, phenolic OH), 13.5 (s, 1H, phenolic OH); MS (m/z, relative intensity): 528.2208 (0.9) [calculated for C26Ha2N408 528.2220]; 399.1434 (19), [M + H-CH2CH2N(COCH3)CH2CH2OH]; 296.0804 (38), [(M+H)[CH2CH2N(COCH3)CH2CH:)OH + N(COCH3)CH2CH-2OH]] 283.0720 (100), [C]5HllN204], 270.0642 (52), [C]4H]0N204], 130.0869 (22), [CH:CH2N(COCH3)CH2CH2OH]; UV/VIS (MeOH) kr~x(e) 280 (13100); 620 (19600) and 674 nm (23900).

Enzymatic oxidation of N,N'-diacetylmitoxantrone (NDA-MXH2) A solution of 57 nmoles of the N,N'-diacetylmitoxantrone and 57 nmoles of hydrogen peroxide in 1 ml of acetate buffer pH 5.0 was incubated at ambient temperature. The initial absorption spectrum was recorded. After the addition of 0.36/zg of horseradish peroxidase the spectrum final was recorded within 2 min. Addition of ascorbic acid (60/~M) resulted in almost quantitative reversion to the initial spectrum.

1.0

8 . 0.8

~

0 . 6

\\\


. . . .

Reduction of the oxidized metabolite (MH22+) with its reduced form (MH:) and with mitoxantrone (MXH2)

0.4

012 O.C

400

,500

6OO

70O

Wavelength (nm) Fig. 2. Absorption spectra resulting from the oxidation of mitoxantrone metabolite by HRP/H20~ system in acetate buffer pH 5.0. HRP was added after the first scan (1) ( ), and then spectra were measured every 20-30 s until the final spectrum (2) (---) was obtained. Addition of ascorbic acid (60/zM) resulted in spectrum (3)

(-+-+-+-).

(a) A 1 ml solution containing mitoxantrone (80 nmoles) and hydrogen peroxide (200 nmoles) in acetate buffer buffer pH 5.0 spectrum (1), (Fig. 3a) was treated with horseradish peroxidase (2/tg). Complete oxidation of MXH2 to MH2 +2 was evident within 3 min by the change of the absorption spectrum from (1) to (2) (Fig. 3a). Catalase (10/~g) was then added and the mixture was incubated for an additional 2 min. No change in the spectrum occurred during this incubation. Subsequently 80 nmoles of mitoxantrone was added

P. KOLODZIEJCZYKet al.

16

Isolation of N, N'-diacetylmitoxantrone after enzymatic oxidation and subsequent reduction

2.0 1.8

:2

1.6 il

1.4 ro

-2

a

7-+

1.2 / I .t /1~

1.0

0

,<

~

i

i

/1=% /c\

Jr I

0.8

\ ,~.--.,

0.6

,, /

_~---..

-"/ _--'-k

+ . . . .

0.4

"

0.2 I

0.0 2OO

2.0

1.8

1.4 (3 t-. 1.2

,,¢

~.

III

++

l[ I/

1.6 ~

0

il

.

I

1

I

I

400

500

600

700

.

.

"+

800

. b

+%

' ~

,=, +~_'~

1.0 0.8 0.6 0.4

N,N'-Diacetylmitoxantrone (2 mg, 3.8 kimoles) was dissolved in 10 ml of methanol and diluted with 40 ml of acetate buffer pH 4.0. Then 100 kil of 0.1 M H202 was added and the initial spectrum was recorded. The enzymatic oxidation was initiated by addition of 150 ¢il solution of HRP (2 mg/ml) and spectra were recorded after 15, 30, 45, 60, 75, 90, and 120 s. Reversion to the initial spectrum was observed upon addition of 50 oil of 1 M ascorbic acid. The reaction mixture was concentrated under vacuum and the product was extracted with n-butanol (2 x 5 ml). The butanolic extracts were evaporated in vacuo and the residue was dissolved in methanol and chromatographed on Sephadex LH20 in methanol affording 1.8 mg of product (90% recovery). The molecular ion M.S., m/z 528.2227 (calculated for C26H32N408 528.2220) and fragmentation patterns were identical to those of N,N'-diacetylmitoxantrone.

0.2 0.0 20O

300

400

5130

1.4

II

i/

C: 1.2

*

800

*%

+ ~

7017

* I::',

,

,,<, ~

,

I).8

0.4

~

0.2 0.0

...~ -~ . . . .

,

Synthesis of the methylated analogue of N ,N '-diacetylmitoxantrone (HMNDA-MXH2)

E

*./"

1.6

600

i 3OO

I 4OO

i 500

~

-- ~ ~ N,~.~._.

i 600

i 700

Wavelength (rim) Fig. 3. Absorption spectra resulting from interaction of fully oxidized form of drug A with reduced form of drug B. HRP was added after the first scan (1) ( ) and then afforded spectrum (2) (---); then catalase was added to destroy excess of H202, Upon addition of a drug B spectrum (3) (- + - + - + -) was recorded. (a) drug A and drug B, MXH2; (b) drug A, MXHz; drug B, MH2; (c) drug A and drug B, MHz. and the formation of a new product with absorption maximum at h = 588 nm was observed (Fig. 3a). (b) A similar experiment was performed with the cyclic metabolite (MH2) as reductant instead of mitoxantrone (Fig. 3b). The concentrations of the reactants and the reaction conditions were identical to those described above. (c) A similar experiment as in (a) and (b) was performed employing mitoxantrone metabolite (MH2) both as substrate for enzymatic oxidation and as reductant (Fig. 3c).

A sample of 8 mg (15 mmoles) of N,N'-diacetylmitoxantrone (NDA-MXH2) was dissolved in 1 ml of tetrahydrofuran. Then 5 mg of sodium hydride was added and the mixture was stirred for 10 min. Subsequently 20/xl (300 mmoles) of methyl iodide was added and the progress of the reaction was monitored by TLC using chloroform:methanol (10:1 v/v) as eluent. When the reaction was completed 50 ml of toluene was added and excess of sodium hydride was removed by filtration on sintered glass. The crude product was purified by chromatography on silica in a chloroform-methanol solvent system affording pure HMNDA-MXH2, 3 mg (32% yield); 1H-NMR (CDC13) ~: 2.07 (s, 3H, CH3CO); 2.18 (b.s., 6H, N--CH3); 2.95 (s, 3 H , - - O C H 3 ) , 3.05 (s, 3H, OCH3); 3 . 3 0 - 3 . 6 0 (m, 16H, --CH2); 3.90 (s, 6H, ArOCH3), 7.10 (b.s., 2H, aromatic H, C2 and C3), 7.30 (m, 2H, aromatic H, C6 and C7); MS, m / z (relative intensity) 612.3171 (18) [calculated for C32H44N408 612.3159]; 610.3010 (25), [M-2H]; 482.2293 (20), [M-CH2-N(COCH3)CH2CHzOCH3); 339.1344 (18), [C19H19N406], 144.1026 (100), [CH~CH2N(COCH3]CH2CH2OCH3]; UV/VIS (MeOH) h~ax 232 (3.58 × 103); 396 (6.7 × 103) and 564 nm (4.95 × 103).

Enzymatic oxidation of HMNDA-MXH2 Due to poor solubility of HMNDA-MXH2 in water, a methanol-buffer mixture (1 : 10, v/v) was used as a

Enzymatic oxidation of mitoxantrone

17

Table 1. The kinetic constants of HRP/H~Oz oxidation reaction of MXH2 and MH~ Vmax

K,

(mmole min -m mg -m)

(10-'mole L -t)

Substrate/pH

4.5

5.8

4.5

5.8

Mitoxantrone (MXH2) Mitoxantrone metabolite (MH2)

0.198 5.159

0.176 2.896

1.528 0.660

1.257 0.255

medium for enzymatic oxidation of this derivative. The absorption spectrum of HMNDA-MXH2 in this mixture exhibits maxima at 239 and 329 nm in UV range and a broad peak with maximum at 510 nm in the visible region, so the spectrum is different from the one recorded in methanol (see above). The incubation mixture consisted of 800/~1 of 0.2 M acetate buffer pH 4.0, 100/11 of HMNDA-MXH2 (490/zM in methanol), 50/~1 of 1 mM H202 and, after recording the initial spectrum, 20/tl of HRP (0.2 mg/ml) was added. Spectra was recorded at 20, 40, 60, 80, 100, 120, and 180 s after the enzyme was added. Then 3 0 / d of 1 mM ascorbic acid was added and a spectrum identical with the initial one was recorded.

Kinetic constants of HRP-catalyzed oxidation of mitoxantrone (MXH2) and its metabolite (M]-/2) The enzymatic oxidations of mitoxantrone and its metabolite was assayed using saturation concentrations of enzyme (2.9/zg/ml) and hydrogen peroxide (0.3 raM) and varying concentrations of substrates ( 3 0 / t M 100/tM). The reaction was performed at 25°C and the buffers were 0.2 M sodium acetate pH 4.5 and 0.1 M potassium phosphate, pH 5.8. Initial reaction rates were determined spectrophotometrically following the decrease of absorbance at 682 nm as a function of time. The measurements were taken at 5 s intervals for 1 min. To calculate the rates of consumption of the substrates extinction coefficients at h -- 682 nm were used: 8.36 × 103 M -1 cm -1 for mitoxantrone and 4.34 × 103 M -1 cm -~ for mitoxantrone metabolite. Lineweaver-Burke plots were obtained using linear regression analyses and the values for Vmaxand Km at pH 4.5 and pH 5.8 were calculated. The results are shown in Table 1.

erties and can be reversibly reduced to a stable form MH~. The latter species displays an absorption spectrum different from mitoxantrone (3 and 1 in Fig. 1). In the present work the metabolite MH2 was isolated and its chemical structure was established and secured by independent synthesis. The formation of the metabolite and possible preceeding steps are shown in Scheme 2. The metabolite, MH2 has an additional ring formed to the 2 position of the chromophore and is a substituted hexahydronaphtho [2,3-f]quinoxaline7,12-dione. Treatment of mitoxantrone with hydrogen peroxide and the denatured enzyme has no effect on the drug, confirming the enzymatic nature of the oxidative transformation. Similarly treatment of mitoxantrone with horseradish peroxidase and hydrogen peroxide in the presence of five equivalents of ascorbic acid has no effect on the mitoxantrone (not shown). This "protective effect" of ascorbic acid is attributed to competitive reaction of the peroxidase with the reductant. 2o

Enzymatic oxidation of metabolite (MH2) The cyclic metabolite also proved to be a substrate for the peroxidase. The changes in the absorption spectra resulting from treatment of the metabolite with HRP/H202 are shown in Figure 2. In contrast to the similar treatment of mitoxantrone under these conditions, the reaction with the metabolite is completely reversible upon addition of a reductant, such as ascorbic acid, in the cycle represented by spectra 1 --> 2 --> 3 (Fig. 2). This cycle of events can be repeated several times rapidly and is interpreted in terms of the redox transitions between the reduced (MHz) and the oxidized (MH2 2+) form of the metabolite [Eq. (1)]. However when the metabolite is HRP/H202

MH2 . RESULTS

Oxidative transformation of mitoxantrone The oxidation of mitoxantrone (MXH2) by the HRP/ H202 system was described in our earlier report. 19 It was found that, the then unidentified metabolite (designated as MH22+), possesses strong electrophilic prop-

~,

• MH22+

etc.

(1)

exposed to the oxidizing conditions for several minutes then the quantitative recovery of MH2 is not observed. This implies instability of the oxidized form MH22+ and attempts to isolate this product were unsuccessful owing to polymer formation (vide infra).

18

P. KOLODZIEJCZYKet al.

that initial oxidative attack takes place at the aromatic nitrogen, which is in accord with previous studies of the susceptibility of aromatic amines towards oxidation. The HMNDA-MXH2, hexamethyl derivative of N,N'-diacetylmitoxantrone, in which the hydroxy groups were replaced by methoxy functions, is also reversibly oxidized by the HRP/H202 system (Scheme 2C) which confirms that the aromatic amino groups, but not the hydroxyl functions in the chromophore, are targets for the oxidative attack.

Structural requirements for oxidative enzymatic transformation of mitoxantrone The structural requirements for enzymatic oxidative transformation (i.e. cyclization) of MHX2 were examined with appropriate derivatives. Synthetic N,N'diacetylmitoxantrone (NDA-MXH2), in which the sidechain amine functions are blocked, is a substrate for the HRP/H202 system. The spectral changes, during enzymatic oxidation and subsequent reduction with ascorbic acid, are summarized are similar to those in Figure 2. In this case the initial and final spectra were identical, suggesting that no new structure was formed, and NDA-MXH2 was recovered unchanged as was proven additionally by mass spectrometry. This confirms, as expected, that the presence of a free amine in the side-chain is necessary for cyclization of one of the side-chain amines to the position 2 of the chromophore (Schemes 1 and 2).

Reaction between oxidized and reduced forms of mitoxantrone and its metabolite Mitoxantrone (MXH2) or its metabolite (MH2) were incubated with HRP/H202 under conditions appropriate to accomplish complete conversion to the oxidized forms, (i.e. [H202] ~ [MXH2] or [MH2]). Then catalase (10/tg) was added to decompose any excess of n 2 0 2 . At this stage no EPR signal was detected in the reaction mixture. The addition of either an equivalent of MXH2 mitoxantrone (Fig. 3a) or mitoxantrone metabolite MH2 (Fig. 3b and 3c), resulted in the following changes in the absorption spectra: (a) formation of an absorption band at 588 nm, characteristic of MH2 and (b) increase of the absorbance at 588 nm above the level corresponding to the added amount of MH2 or MXHv No other absorption maxima, which would indicate formation of a charge-transfer complex between

Site of primary oxidation attack A number of alternative sites in the mitoxantrone chromophore, such as amino groups (5,8 position) or 1,4-hydroxy groups, may be considered as possible sites of enzymatic oxidation. Ametantrone, which lacks the hydroxyl functions in the chromophore, is also a good substrate for the HRP/H202 system and affords, among the other products, a cyclic metabolite analogous to MH2 (not shown). These results suggest Scheme 1

HO

H O H-Nj ' ~ ' N ~

HO

O H-N

~

~ O H

HO

0 H-N

~ O H

HO

O H-N

HO

O H-N

H

OH

~

/

~

"

HRP/H202

~-x-HO

0 H-N

N~OH

H

OH

MXH2

",,,÷ HRP/H202.2e~~ HO

O

N~

H ~OH

++ -2H

~-~..~ OH

B HO

O

N MXH2*2

H

OH MH2+2

Scheme 1. Schematic depiction of HRP oxidation of mitoxantrone (MXH2) to a radical cation species MXH2- followed by cyclization to radical cation MHz "+and redox equilibria of the latter with isolated metabolite MH2 and its unstable fully oxidized form (MH:+2).

Enzymatic oxidation of mitoxantrone

19

ll¢lwl~ 2 H

H

HO ~

0 H*N~ N ~ O H ~

HO

O H-N

•

N~ H

~OH

1

HflP/H~O=I,

L"O

OH

O"N

M~I(H2

. ~ N

-2H •

O"J

HO

MXHI+ /

F.o,

",, ~

O H.N

<~

A

.o.

H

OHJ

MHi+

[

A

7

MH2+2 Ac

Ac !

l

HO

O H-N~N~/~"

O~ ~ ~ H . ~

OH

F .o

o ..er'",~"~o.7

.+

MoO

OMeN

"-"~N~/~o.7 ÷+

NDA-I~EH2+2 I

I

OM e - N ~ N ~ ' O M e

~_T.~...~

o

H~p/H202 ND~MIQt~; +

MeO

.o

I

HMNDA-MXH2

Ac

_J

Hlmn~o2 "

.+

÷'

C

HMNOA-MXHi*

Scheme 2. Schematic depiction of HRP oxidation of mitoxantrone (MXH2) to radical cation (MXH/) and cyclic radical cation (MH2.+) and alternative generation of the latter by disproportionation-comproportionation of final cyclic metabolite (MH2) and the corresponding diimino form (MH2+2). The middle part B of the scheme depicts the HRP oxidation of N,N'-diacetylmitoxantrone to the corresponding radical cation and diimino forms without cyclization. The lower part, C, depicts the HRP catalyzed oxidation of hexamethyl N,N'-diacetylmitoxantrone.

the oxidized and reduced forms of the drug, were recorded. However, a mixture of oxidized and reduced forms of the metabolite produces a strong EPR signal (vide infra). From the above observations it was concluded that the reduced forms of MXH2 or MH2 act as electron donors for MH22÷ . EPR studies

The HRP catalyzed oxidation of mitoxantrone metabolite (MH2) and N,N'-diacetylmitoxantrone (NDAMXH2) by hydrogen peroxide was studied with EPR spectroscopy. EPR signals at a g-value of 2.003 were observed for both MH2 and NDA-MXH2 (Fig. 4 a,b) and were similar to those reported previously for mitoxantrone. 19If reduced modulation amplitude was applied an EPR spectrum of NDA-MXH2 radical with poorly resolved hyperfine structure was recorded (Fig. 4c). The EPR signals of MXH2 and MH2 incubated under the same conditions showed no hyperfine structure. The intensity of the EPR signal of MH 2+ depends

on time of incubation and [MH:] / [HzO2] ratio as shown in Figure 5. The signal of maximum intensity was recorded for [MH2]/[H202] ~- 10 and it was stable at this level for approximately 3 min (steady-state conditions) and subsequently decreased. Enzymatic oxidation of NDA-MXH2 produced a stable EPR signal with maximum intensity also for [NDAMXH2]/[HEOz] ~ 10 (data not shown). The intensity of the steady-state EPR signal shows a pH dependence for these two derivatives. This was demonstrated (Fig. 6) by measurements of EPR signals from MH2 and NDA-MXH2 incubated with the HRP/ H202 system at different pH values and at [Drug]/ [H202] = 10. The highest intensity of the EPR signal was recorded at pH 3.0 and it decreased sharply with increase in pH. In order to compare the kinetics of the formation and decay of free radicals of mitoxantrone, with that of its cyclic metabolite and N,N'-diacetylmitoxantrone, the drugs were incubated with H202 and HRP under conditions previously found to be optimal for

P. KOLODZIEJCZYKet al.

20

300

t

200

,3 // \\

_ loo

i'r"

GO

>, Fig. 4. EPR spectra of mitoxantrone metabolite (a) and N,N'-diacetylmitoxantrone (b and c) incubated with HRP/H202 in acetate buffer pH 3.0. [Drug] = 0.7 mM; [H~O2] = 0.07 mM; [HRP] = 7 #g/ml, Instrumental conditions: (a) scan range 150 G; modulation amplitude 5.4 G; gain 1.25 × 105; time constant 0.5 s; microwave power 20 mW; (b) same as above except gain 2.5 × 105, scan time 500 s; (c) same as above except gain 5 × 105 and modulation amplitude 1.7 G. m i t o x a n t r o n e , ~9 i.e. using [Drug]:[H202] = 1 and with the ratio equal 10, o p t i m a l for both MH2 and N D A - M X H 2 (vide supra). As is d e m o n s t r a t e d in Figure 7 the m a x i m u m signal for m i t o x a n t r o n e was reached after about 20 min, w h i l e for the m e t a b o l i t e and N,N'l

I

I

300

~200

\

t~

~q~ GO 0 2

t

L

3

4

5

6

7

8

pH Fig. 6. Dependence of steady-state radical concentration on pH of the incubation mixture of (a) mitoxantrone ( ); (b) N,N'-diacetylmitoxantrone (---) [Drug] = 0.7 mM; ;obH202] = 0.1 mM; [HRP] = 7/zg/ml. d i a c e t y l m i t o x a n t r o n e the m a x i m u m intensities o f the EPR signals were reached in less than 1 min. In the case o f the radical from MH2 incubation the signal d e c r e a s e d to 50% o f its m a x i m u m value within 7 min, and for the radicals from N D A - M X H 2 incubation a stable signal was o b s e r v e d from at least 1 h, whereas the radicals from m i t o x a n t r o n e incubation show intermediate stability.

The generation of free radicals in direct reaction of oxidized and reduced forms of mitoxantrone, mitoxantrone metabolite and N, N ' -diacetylmitoxantrone

I00

(J

rr

I 10

20

Time (min)

30

40

Fig. 5. Enzymatic generation of MH2" radicals:dependence on [H202] and time of incubation. Incubation mixtures consisted of MH2 (0.7 raM), HRP (7/lg/ml) and H202 as follows: 0.02 (-V-V-~'-); 0.04 ( ' A - A ' A ) ; 0.07 ( - I - l - ) ; 0.14 (-0-0-); and 0.18 mM (- • - • - • -) in acetate buffer, pH 3.0.

Solutions containing a drug (0.7 raM), H202 (1.4 m M ) and H R P (14 /zg/ml) in buffer p H 3.0 were incubated for 3 min at 25°C. T h e condffions were chosen to a c c o m p l i s h m a x i m u m p r o d u c t i o n o f the o x i d i z e d form o f drugs. C o m p l e t i o n o f o x i d a t i o n was established by the characteristic changes in the absorption spectra (as d e s c r i b e d above). Then catalase was a d d e d

Enzymatic oxidation of mitoxantrone

21

r-

£ ¢-

._o 100 tO r"

.o

.o_ "10

rr" 0

, , , , I , + ~ , 1 , , , , I

0

10

¢.) to O lOO

20

Time (min)

Fig. 7. Dependenceof radical concentration on the incubation time from MXH2 (-O-O-), MH2 (-(3-0-) and NDA-MXH2 (-&-&-). Incubations in acetate buffer pH 3.0 contained drug (0.7 mM), HRP (14 pg/ml) and H202(0.7 mM) for MXH2or 0.07 mM for MH2and NDA-MXH2. (6.5 gg/ml) to destroy excess of hydrogen peroxide and the mixture was incubated for 1 rain. At this stage no EPR signal was detected in the sample. Addition of an equivalent of the drugs in reduced form (MXH2, MH2, or NDA-MXH2, respectively) produce strong EPR signals.* These signals have a linewidth of 11.75 G and their g-values are 2.003. These EPR signals are very similar to those measured during the enzymatic oxidation of the drugs (Fig. 4a,b) and to the one reported earlier in the enzymatic oxidation of mitoxantrone, t9 The EPR spectrum of the radical from acetylated mitoxantrone shows poorly resolved hyperfine structure and is different from the spectra obtained with MXH2 or MH:. The kinetics of the nonenzymatic generation and decay of free radical EPR signals are presented in Figure 8 and there is evident similarity to those observed for enzymatic free radical formation described above (Fig. 7). These findings clearly indicate that a radical form of the metabolite is not necessarily produced enzymatically but it might also be a result of an electron transfer between fully reduced and fully oxidized forms of the metabolite, presumably by establishing the following equilibrium: MH2 + MH2 +2 .

" 2MH2 "+

(2)

*In a control experiment it was established that: (i) no EPR signal was generated during incubation of a drug with H202 and catalase using the aboveconditions; (ii) subsequent addition of HRP to those mixtures also did not produce EPR signal which indicates that all H~O2was decomposed.

._o n-

o

o

2

4

6

8

Time

lo

12

14

16

18

(rain)

Fig. 8. Generationof free radicals in incubationmixturescontaining oxidized and reduced forms of drugs: dependence on time of incubation. Incubation mixtures consisted of drug A (0.7 mM), H202 (1.4 mM) and HRP (14 #g/ml) in acetate buffer pH 3.0. After 3 min catalase (6.5 #g/ml) was added and mixture was incubated for 1 min. Then drug B (0.7 raM) was added and EPR spectra were recorded. Curves: (-S-O-) A, MXH2 and B, MXH2;(-B-m-) A, MXH2 and B, MH2; (-O-O-) A and B, MH2; (-A-A-) A and B, NDA-MXH2.

A single electron transfer between the oxidized metabolite, MH2 +2 and mitoxantrone is also possible. As was demonstrated above, MH2 ÷2 is produced during enzymatic oxidation of MXH2, therefore the interaction between those two species is very likely and may also contribute to radical formation according to the equilibrium described in Eq. (3). MH2 +2 + MXH2

> (MH2 "+ ... MXH2 +") 2MH2 "+

)

" MH2 +2 + MH2

(3)

DISCUSSION

Previous studies on the metabolism of anthracyclines, anthracenedione-based anticancer drugs and other related xenobiotics have concentrated on the implications of free radical generation by NADPH-cytochrome P450 reductase systems of liver endoplasmic reticulum (microsomes), 9-~5,2~-3° cardiac sarcoplasmic reticulum (sarcosomes) H,22,z~and cardiac mitochondrial NADH dehydrogenase systems. 2+In contrast, the pos-

22

P. KOLODZIEJCZYKet al.

sibility of oxidative metabolism of these agents has been neglected until our recent report. ~9 Our spectrophotometric and EPR studies showed that mitoxantrone is subject to facile oxidative metabolism by the HRP/ H202 system with concomitant generation of free radicals. 19 The structure of the metabolite isolated from this reaction is secured by independent synthesis. An additional ring has been formed to the 2 position of the chromophore and the metabolite is a substituted hexahydronaphtho[2,3-f]quinoxaline-7,12-dione. This type of structure has been reported 31-33 to be a product of reaction of anthraquinones or leucoquinizarin with 1,2diamines. Usually elevated reaction temperatures and the presence of a primary or secondary amine in the side chain is required. 31 Thus, the hexahydronaphthoquinoxalinedione is a by-product in the synthesis of ametantrone when elevated reaction temperatures are applied. 3,4 In this paper we demonstrate that such cyclization can also be catalyzed enzymatically at ambient temperatures. Enzymatic oxidation of phenylenediamines is well established and is known to involve free radical intermediates. 34 The most likely sites of initial enzymatic attack on mitoxantrone are the aromatic amine groups since neither acylation of the aliphatic secondary amines, nor absence nor protection of the phenolic hydroxyls prevents this reaction. The cyclic metabolite (MH2) as well as the N,N'diacetylmitoxantrone, readily undergo enzymatic oxidation concomitant with the generation of free radical species. Their oxidized forms are fully and reversibly reduced with agents such as ascorbic acid, NADPH or sodium borohydride. The data suggest the following series of events in the enzymatic oxidation of mitoxantrone metabolite in which one- and two-electron mechanisms of action of the peroxidase/HzOz 2° are indicated. HRP

H202

) HRP-Compound I

HRP-Compound I + MH2

(4)

>

HRP-Compound II + MH2 "+ (5) HRP-Compound II + MH 2

~ HRP + MH2 +

(6)

also further oxidation can not be excluded: MH2.+

HRP/H20~ MH2 +2

(7)

An alternative, two-electron mechanism, without the free radical intermediate can be written as: MH2

HRP/H202

> MH2 ÷2

(8)

The free radical form, MH2 "+, may be generated by either direct action of peroxidase-compound I are peroxidase-compound II [Eqs. (5) and (6)] and/or in disproportionation-comproportionation reactions [Eq. (2)]. The EPR observations indicate clearly that the radical form of the metabolite is not necessarily produced enzymatically but that it might be the result of an electron transfer between the fully reduced and fully oxidized forms of the metabolite [Eq. (2)]. The increased stabilization of the free radical at low pH's suggests it exists as a radical cation species MH2 "~. The acetylated mitoxantrone, NDA-MXH2, can also participate in a similar electron exchange reaction. The EPR measurements confirmed electron transfer between the fully oxidized and reduced forms of the acetylated drug and can be described by the equation: NDA-MXH2 + NDA-MXH2 ÷2

• 2NDA-MXH2 +" (9)

The EPR spectrum of the radical generated from NDAMXH2 shows poorly resolved hyperfine structure and is therefore different from the spectra of radicals derived from MH2 or MXH2. The EPR signal of NDAMXHz radical is also more stable (no change in the steady-state signal intensity for at least 1 h) in comparison with the signals from incubation containing MXH2 and MH2. The N,N'-diacetylmitoxantrone, by analogy with the mitoxantrone metabolite, undergoes enzymatic oxidation which can be described by equations analogous to Eqs. (5)-(8). This process was found to be fully reversible (Scheme 2B). Upon addition of reducing agent, such as ascorbic acid, to the oxidized sample, the starting material was fully recovered. The spectrophotometric measurements indicated that the enzymatic oxidation-AA reduction cycle do not produce any structural modifications of the mitoxantrone metabolite (MH2) and N,N'-diacetylmitoxantrone (NDA-MXH2) chromophores. Important implications from this are that (i) the presence of a free primary or secondary amine in the side chain, as in mitoxantrone, is necessary to produce the cyclized metabolite upon oxidation of drug and (ii) no cyclization involving the second side chain in the same molecule takes place. The EPR experiments also demonstrate that free radicals observed during HRP/H202 oxidation of mitoxantrone (MXH2), may not be of metabolic origin since single-electron transfer between the oxidized metabolite MH2 ÷2, formed enzymatically, and mitoxantrone itself is also possible. This possibility is described by Eq. (3).

Enzymaticoxidationof mitoxantrone The initial steps in the oxidation of the mitoxantrone can be approximated by oxidation of the acetylated drug, where formation of radicals is readily observed, and cyclization does not take place (Scheme 2B). Comparison of free radical generation for mitoxantrone and mitoxantrone metabolite (Fig. 7) indicates that, for the latter, the maximum concentration of radicals was achieved almost instantaneously whereas for mitoxantrone it was attained after ca. 20 min. This observation may reflect the different degrees of aggregation of MXH2 and MH2 under these conditions. Independent kinetic studies (Table 1) confirm that reaction of the peroxidase with the cyclic metabolite MH2 proceeds more readily than with mitoxantrone. The rates of enzymatic reactions were measured and are summarized in Table 1. The Vm~xvalue for the cyclic form is about 16 to 26 times higher than that of mitoxantrone while the corresponding Km values (depending on the pH) are 2- to 5-fold lower. The observed differences in enzymatic oxidation of MXH2 and MH2 may be a consequence of a multi-step reaction involving both electron transfer and structural modification which takes place in the case of MXH2 while oxidation of MH2 is a simple process which does not include cyclization. The metabolite MH2 is formed after initial activation, presumably by abstraction of a single electron from MXH2 to form MXH2"+which seems to be a transient species and which undergoes internal rearrangement to MH2"+ (see Scheme 1A) and after reduction yields MH2. In fact, one electron mechanisms are most common for HRP catalyzed oxidation of similar systems. 2° Alternatively one can consider if two-electron oxidation of MXH~ yielding MXH2+2 and structure, MXH2 +2, might exist comes from the analogy with the dication NDA-MXH2 +2 which is formed in enzymatic oxidation of NDA-MXH2 as is shown in Fig. 3 and in Scheme 2B. Mitoxantrone or its metabolite may be considered as aromatic p-diamines. Such systems are known to undergo two-electron electrochemical oxidation to diimines. An EPR signal was often observed in such cases and it was attributed to electron exchange between amine and the diimine forms to give a monocation radical. 34-36 The reactions in aqueous solution are complicated by the fact that these compounds can exist in various stages of protonation, depending on the pH of the medium. ~6'37 Although horseradish peroxidase was employed, for our studies due to its simplicity and similarity to physiologically relevant systems,38,39 other heme-containing systems were also considered. For example, cytochrome c, hemoglobin and methemoglobin also catalyze transformation of mitoxantrone, however, the

23

rate o f these reactions is lower by a factor of ca. 100 than for HRP (unpublished results).

Biological relevance and implications of oxidative metabolism of mitoxantrone The transformation of mitoxantrone by the action of enzymes present in body fluids may be relevant to the mode of action of this agent. There are reports indicating that mitoxantrone is metabolized when incubated with human plasma. *°-42 It was also found that ascorbic acid prevents the "oxidative degradation" of the drug. Reynolds et al. al were unsuccessful in their attempts to regenerate oxidized mitoxantrone in plasma by addition of ascorbate but they were able, thereby, to prevent its further metabolism, These findings are consistent with our proposed scheme of irreversible oxidation of mitoxantrone (Scheme 2A) and the protective effect of ascorbic acid due to its competitive reaction with the HRP/H202 system. Peng et al. 42 examined the stability of mitoxantrone and bisantrene in human plasma by HPLC. It was shown that mitoxantrone is less stable than bisantrene in plasma and it was suggested that the instability is related to "more than interaction with plasma proteins" but no further explanation was offered. Recently it has been shown43 that mitoxantrone, in common with several antioxidants, inhibits the synthesis of prostaglandin E2 and collagen-stimulated platelet aggregation. The metastatic spread of tumors is linked to the association of circulating tumor cells with host platelets to form platelet-tumor cell aggregates which adhere to vascular intima. Since mitoxantrone is administered intravenously therefore processes occurring in blood plasma, and the interaction of the drug with blood elements, may be of significance in its clinical application. Neither the isolation nor biological activity data for the cyclic mitoxantrone metabolite have been reported. Therefore, the question arises whether the cyclic metabolite of mitoxantrone (MH2) contributes to the biological activity of this agent. For example, the corresponding cyclic form of ametantrone which was isolated as a by-product of ametantrone synthesis3'4 exhibits approximately 50% of the activity of ametantrone against P388 leukemia.4 Covalent binding to DNA of the antitumor agent N 2methyl-9-hydroxyel-lipticinium initiated in vitro by horseradish peroxidase was recently reported.** Thus the facile generation of reactive species by the action of HRP on mitoxantrone suggests that their interaction with cellular macromolecules should be explored. There are some observations that suggest the need for a closer examination of metabolically generated free radicals from mitoxantrone in microsomal sys-

24

P. KOLODZIEJCZYKet al.

tems. Basra et al. ~2 reported that mitoxantrone free radicals do not result in a significant increase in microsomal superoxide production. Mitoxantrone does not increase N A D P H utilization or O2 ~ formation in rabbit liver microsomes, ~3 and has little effect on N A D P H oxidation in rat liver microsomes. ~ In fact, both m i t o x a n t r o n e and ametantrone were found to inhibit N A D P H - d e p e n d e n t cytochrome P450 reductase activity. 13,15-17 M i t o x a n t r o n e and ametantrone, which were anticipated tO undergo free radical formation and redox cycling, by analogy with doxorubicin, were found to be poor substrates for liver cytochrome P450 reductases. ~2,~6,3° Thus, the oxidative m e c h a n i s m of generation of free radicals from anthracenediones by m i c r o s o m a l mixed function cytochrome P450 complex should be considered. The key c o m p o n e n t of this diversified function system is the hemoprotein cytochrome P450 which acts as the terminal b i n d i n g site as well as the terminal oxidase. Recently it has been shown that cytochrome P450 can accept oxygen from a variety of donors, including organic peroxides and hydrogen peroxide. 45 Cytochrome P450, p r e s u m a b l y because of its ability to b i n d organic substrates, can act as a peroxidase with hydrogen peroxide and organic peroxides. It has also been reported that endogenous lipid peroxides formed by the N A D P H / A D P / F e 3÷ microsomal system may act as a substrate for cytochrome P450. 45 Significantly, the EPR signals reported for mitoxantrone in microsomal systems are similar to those observed with our HRP/H2Oz system ~9 in terms of gvalue and linewidth. It is also noteworthy that incubation of m i t o x a n t r o n e with microsomes and N A D P H requires 20 m i n before the appearance of an EPR signal, 3° in contrast to the rapid generation of daunorubicin s e m i q u i n o n e signal u n d e r similar conditions. Further studies are necessary to confirm whether mitoxantrone interaction with the cytochrome-P450 system can utilize basal or anthracycline-originated lipid peroxides in tissues and thereby constitute a scavenging m e c h a n i s m for these cardiac tissue d a m a g i n g species. The results of studies in these directions will be reported in due course. Acknowledgment--This research was supported by grants (to

J.W.L.) from the National Cancer Institute of Canada and the Natural Sciences and Engineering Research Council of Canada. REFERENCES

1. Bonadonna, G.; Monfardini, S. Cardiac toxicity of daunorubicin. Lancet 1:837-847; 1969. 2. Lenaz, L.; Page, J. A. Cardiotoxicity of adriamycin and related anthracyclines. Cancer Treat. Rev. 3:111-119; 1976. 3. Zee-Cheng, R. K.-Y.; Cheng, C. C. Antineoplastic agents.

4.

5. 6. 7. 8. 9. 10. 11.

12. 13.

14. 15. 16.

17.

18.

19.

20. 21. 22. 23.

Structure-activity relationship study of bis (substituted aminoalkylamino)anthraquinones. J. Med. Chem. 21:291-294; 1978. Murdock, K. C.; Child, R. G.; Fabio, P. F.; Angier, R. B.; Wallace, R. E.; Durr, F. E.; Citarella, R. V. Antitumor agents. 1,4-bis[(aminoalkyl)amino]-9,10-anthracenediones. J. Med. Chem. 22:1024-1030; 1979. Smith, I. E. Mitoxantrone (novantrone): a review of experimental and early clinical studies. Cancer Treat. Rev. 10:103115; 1983. Durr, F. E.; Wallace, R. E.; Citarella, R. V. Molecular and biochemicalpharmacologyof mitoxantrone. Cancer Treat. Rev. 10B:3-11; 1983. Kappus, H. Overview of enzyme systems involved in bioreduction of drugs in redox cycling. Biochem. Pharmacol. 35:16; 1986. Goodman, J.; Hochstein, P. Generationof free radicals and lipid peroxidation by redox cycling of adriamycin and daunorubicin. Biochem. Biophys. Res. Commun. 77:797-803; 1977. Bachur, N. R.; Gordon, S. L.; Gee, M. V. A general mechanism for microsomal activation of quinone anticancer agents to free radicals. Cancer Res. 38:1745-1750; 1978. Handa, K.; Sato, S. Stimulation of microsomal NADPH oxidation by quinone group-containinganticancer chemicals. Gann 67:523-528; 1976. Mimnaugh, E. G.; Trush, M. A.; Ginsburg, E.; Gram, T. E. Differential effects of anthracycline drugs on rat heart and liver microsomal NADPH-dependentlipid peroxidation. Cancer Res. 42:3574-3582; 1982. Basra, J.; Wolf, C. R.; Brown, J. R.; Patterson, L. H. Evidence for human liver mediated free-radical formation by doxorubicin and mitozantrone. Anticancer Drug Design 1:45-52; 1985. Kharasch, E. D.; Novak, R. F. Bis(alkylamino)anthracenedione antineoplastic agent metabolic activation by NADPH-cytochrome P-450 reductase and NADH dehydrogenase:Diminished activity relative to anthraeyclines. Arch. Biochem. Biophys. 224:682-694; 1983. Kharasch, E. D.; Novak, R. F. Anthracenedione activation by NADPH-cytochrome P-450 reductase; comparison with anthracyclines. Biochem. Pharmacol. 311:2881-2884; 1981. Kharasch, E. D.; Novak, R. F. Inhibitory effects of anthracenedione antineoplastic agents on hepatic and cardiac lipid peroxidation. J. Pharmacol. Exp. Ther. 226:500-506; 1983. Kharasch, E. D.; Novak, R. F. Inhibition of adriamycinstimulated microsomal lipid peroxidation by mitoxantrone and ametantrone, two new anthracenedione antineoplastic agents. Biochem. Biophys. Res. Commun. 108:1346-1352; 1982. Kharasch, E. D.; Novak, R. F. Structural and mechanistic differences in quinone inhibition of microsomal drug metabolism; inhibition of NADPH-cytochrome P-450 reductase activity in cytochrome P-450. Developments in Biochemistry 23:623-626; 1982. Kharasch, E. D.; Novak, R. F. Mitoxantrone and ametantrone inhibit hydroperoxide-dependentinitiation and propagation reactions in fatty acid peroxidation. J. Biol. Chem. 260:1064510652; 1985. Reszka, K.; Kolodziejczyk, P.; Lown, J. W. Horseradish peroxidase-catalyzed oxidation of mitoxantrone. Spectrophotometric and electron paramagnetic resonance studies. J. Free Rad. Biol. Med. 2:25-32; 1986. Yamazaki, I. Free Radicals in enzyme-substrate reactions. In: Pryor, W. A., ed. Free radicals in biology, Vol. 3. New York: Academic Press; 1977: 183-218. Bachur, N. R.; Gordon, S. L.; Gee, M. V. Anthracycline antibiotic augmentationof microsomal electron transport and free radical formation. Mol. Pharmacol. 13:901-910; 1977. Doroshow, J. H. Comparative cardiac oxygen radical production by anthracycline antibiotics, mitoxantrone, bisantrene, MAMSA and neocarzinostatin. Clin. Res. 31:67A; 1983. Oldcorne, M. A.; Brown, J. R.; Patterson, L. H. Anthraquinone free-radical formation by mouse liver subcellular fractions. Biochem. Soc. Trans. 12:681-682; 1984.

Enzymatic oxidation of mitoxantrone 24. Kappus, H.; Sies, H. toxic drug effects associated with oxygen metabolism: Redox cycline and lipid peroxidation. Experientia 37:1233-1241; 1981. 25. Doroshow, J. H.; Locker, G. Y.; Myers, C. E. Enzymatic defences of the mouse heart against reactive oxygen metabolites. Alterations produced by doxorubicin. J. Clin. Invest. 65:128135; 1980. 26. Davies, K. J. A.; Doroshow, J. H.; Hochstein, P. Mitochoudrial NADH dehydrogenase-catalysed oxygen radical production by adriamycin and the relative inactivity of 5-iminodaunorubicin. FEBS Len. 153:227-230; 1983. 27. Peters, J. H.; Gordon, G. R.; Kashiwase, D.; Lown, J. W.; Yen, S.-F.; Plambeck, J. A. Redox activities of antitumor anthracyclines determined by microsomal oxygen consumption and assays for superoxide anion and hydroxyl radical generation. Biochem. Pharmacol. 35:1309-1323; 1986. 28. Lown, J. W.; Chen, H. H.; Plambeck, J. A.; Acton, E. M. Further studies on the generation of reactive oxygen species from activated anthracyclines and the relationship to cytotoxic action and cardiotoxic effects. Biochem. Pharmacol. 31:575581; 1982. 29. Kharasch, E. D.; Novak, R. E Inhibition of microsomal oxidative drug metabolism by 1,4-bis[2-[(2-hydroxyethyl)amino]ethylamino]-9,10-anthracenedione diacetate, a new antineoplastic agent. Mol. Pharmacol. 22:471-478; 1982. 30. Sinha, B. K.; Motten, A. G.; Hanck, K. W. The electrochemical reduction of 1,4-bis[2[(2-hydroxyethyl)amino]ethylamino] anthracenedione and daunomycin: Biochemical significance in superoxide formation. Chem. Biol. Interactions 43:371377; 1983. 31. von K611iker, H. P.; Caveng, P. Kondensationen mit aliphatischen und zycloaliphatischen aminen, iminen und diaminen in der anthrachinonreihe. Chimia 20:281-291; 1966. 32. Simon, M. S.; Waller, D. P. 4-[3-Aminoethylamino-l,2-(ethylene-l'2'-diamino)anthraquinone: a condensation product of ethylene diamine and leucoquinizarin. Tetrahedron Left. 16:1527-1529; 1967. 33. Greenhalgh, C. W.; Hughes, N. The reaction of leucoquinizarins with alkylenodiamines. J. Chem. Soc. (C):1284-1288; 1968. 34. Michaelis, L.; Schubert, M. P.; Granick, S. The free radicals of the type of Wurster's salts. J. Am. Chem. Soc. 61:19811992; 1939.

25

35. Cauquis, G.; Delbomme, H.; Serve, D. L'oxydation anodique de quelques N-phenl p-phenylenediamines en milieu acetonitrile. Tetrahedron Lett. 19:1965-1968; 1972. 36. Pierre, L. H.; Ludwig, P.; Adams, R. N. Studies of electrochemically generated radical ion in aqueous solution. Anal. Chem. 34:916-921; 1962. 37. Mark, H. B. Jr.; Anson, F. C. Electro-oxidation of phenylenediamines and related compounds at platinum electrodes. Effects of acid strength. Anal. Chem. 35:722-724; 1963. 38. (a) Griffin, B. W.; Ting, P. L. Mechanism of N-demethylation of aminopyrine by hydrogen peroxide catalyzed by horseradish peroxidase, metmyoglobin and protohemin. Biochemistry 17:2206-2211; 1978. (b) Meunier, B. Horseradish peroxidase: A useful tool for modeling the extrahepatic biooxidation of exogens. Biochimie. 69:3-9; 1987. 39. Chance, M.; Powers, L.; Poulos, T.; Chance, B. Cytochrome C peroxidase compound ES is identical with horseradish peroxidase compound I in iron-ligand distances. Biochemistry 25:1266-1270; 1986. 40. Ehninger, G.; Proksch, B.; Schiller, E. Detection and separation of mitoxantron and its metabolites in plasma and urine by highperformance liquid chromatography. J. Chromatogr. 342:119127; 1985. 41. Reynolds, D. L.; Sternson, L. A.; Repta, A. J. Clinical analysis for the antineoplastic agent 1,4-dihydroxy-5,8-bis{{2-[(2-hydroxy-ethyl}amino]ethyl}amino}9,10-anthracenedione dihydrochloride (NSC 301739) in plasma. J. Chromatogr. 222:225240; 1981. 42. Peng, Y. M.; Ormberg, D.; Alberts, D. S. Improved highperformance liquid chromatography of the new antineoplastic agents bisantrene and mitoxantrone. J. Chromatogr. 233:235247; 1982. 43. Frank, P.; Novak, R. F. Mitoxantrone and bisantrene inhibition of platelet aggregation and prostaglandin Ez production in vivo. Biochem. Pharmacol. 34:3609-3614; 1985. 44. Auclair, C.; Dugue, B.; Meunier, B.; Paoletti, C. Peroxidasecatalyzed covalent binding of the antitumor drug NLmethyl-9hydroxyellipticinium to DNA in vitro. Biochem. 25:1240-1245; 1986. 45. O'Brien, P. J.; Rahimtula, A. D. A peroxidase assay for cytochrome P-450. Methods in Enzymology 52:407-412; 1978.