The relevance of enzymatic oxidation by horseradish peroxidase to antitumour potency of imidazoacridinone derivatives

Chemico-Biological Interactions 115 (1998) 1 – 22

The relevance of enzymatic oxidation by horseradish peroxidase to antitumour potency of imidazoacridinone derivatives Zofia Mazerska *, Katarzyna Gorlewska, Agnieszka Kraciuk, Jerzy Konopa Department of Pharmaceutical Technology and Biochemistry, Technical Uni6ersity of Gdan´sk, 80 -952 Gdan´sk, Poland Received 28 July 1997; received in revised form 31 March 1998; accepted 14 April 1998

Abstract The presented study concentrated on the oxidative enzymatic transformation of six imidazoacridinone derivatives exhibiting different antitumour activity. Horseradish peroxidase was applied as an enzymatic model system. The investigations aimed to evaluate: (1) whether the studied compounds can undergo oxidative biotransformation; (2) whether the susceptibility to such biotransformation relates to the structure and antitumour activity of these compounds; and (3) which elements of imidazoacridinone structure are involved in this kind of transformation. The reaction courses were followed by three methods: UV-VIS spectroscopy, electron paramagnetic resonance and high-performance liquid chromatography. It was shown that all the imidazoacridinones studied underwent enzymatic oxidation resulting in the formation of several products, spectra of which revealed that imidazoacridinone chromophore as well as alkylamino side-chain were involved in these biotransformations. The susceptibility to enzymatic oxidation turned out to be well correlated with antitumour activity of these compounds. It was demonstrated that the highly active antitumour 8-hydroxy derivatives underwent oxidative transformation far more readily than the less active 8-methoxy derivatives and analogues without substituent in position 8. The results indicated that the oxidation pathway of 8-hydroxy compounds was different from those observed for the remaining imidazoacridinones studied. It also differed from the pathway * Corresponding author. Tel.: +48 58 3472407; fax: + 48 58 3471516; e-mail: mazerska @altis.chem.pg.gda.pl 0009-2797/98/$-see front matter © 1998 Elsevier Science Ireland Ltd. All rights reserved.

PII S0009-2797(98)00042-8

2

Z. Mazerska et al. / Chemico-Biological Interactions 115 (1998) 1–22

proposed earlier for mitoxantrone. Moreover, it was find out that not only the rate but also the mechanism of horseradish oxidation of 8-hydroxy derivatives depended on the reaction conditions. In the presence of excess of hydrogen peroxide, the drugs were exceptionally reactive giving rise to the mixture of many unstable products, among which compounds with both changed and unchanged chromophore structure were formed. However, the equimolar ratio of drug and hydrogen peroxide led to stable products, which resulted from the oxidation in aminoalkyl side-chain. The possible structures of products of imidazoacridinone enzymatic oxidation are discussed. In conclusion, the results presented in this paper indicate that the oxidative metabolic activation of imidazoacridinones may represent the crucial step in their biological action. © 1998 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Antitumour imidazoacridinones; Enzymatic oxidation; Metabolic activation of drugs; Peroxidase-mediated metabolism

1. Introduction Imidazoacridinones represent a new group of very active antitumour compounds developed in our laboratory [1 –3]. The most valuable are 5-alkylamino imidazoacridinone derivatives which exhibit strong cytotoxicity towards L1210 and HeLaS3 cells in tissue culture. Some of them display significant and clearly differentiated inhibitory properties against 60 human tumor cell lines cultured in vitro included in the National Cancer Institute screening programme (Protocols of the National Cancer Institute, unpublished data). Besides, imidazoacridinones exhibit strong antitumour activity against transplantable tumors in animals: leukaemia P388, melanoma B16 and colon adenocarcinoma 38 and 28 [4]. Human colorectal cancer is also sensitive to these compounds [5]. At the same time, imidazoacridinones display some valuable pharmacological properties. They exhibit only limited mutagenic potential. Cellular transport of these agents occurs rapidly and most of the drug applied accumulates in the nucleus [6]. All these properties make this group of compounds very attractive from clinical point of view; therefore, one imidazoacridinone derivative, denoted C1311 (1), is currently undergoing the phase I clinical studies under the auspices of the European Organisation of Research and Therapy of Cancer (EORTC).

Z. Mazerska et al. / Chemico-Biological Interactions 115 (1998) 1–22

3

Studies on antitumour activity of imidazoacridinones have shown [2,3] that the derivatives containing hydroxyl group in position 8 of imidazoacridinone ring exert remarkably higher activity than those without hydroxyl substituent or with this group attached to another position of heterocyclic core. It has been also demonstrated that the presence of alkylamino side-chain in position 5 of imidazoacridinone ring is crucial for high antitumour potency of these compounds. The most preferable in this regard is diethylaminoethylamino side-chain. In order to explain the benefits associated with the introduction of 8-hydroxyl group into the imidazoacridinone ring and with the presence of 5-alkylamino side-chain, as well as to learn what mechanism is generally responsible for high antitumour potency of imidazoacridinones, studies concerning several aspects of their biological activity have been undertaken. It has been shown that these compounds bind physicochemically to DNA [6,7]. Although this effect seems to be necessary for biological action of imidazoacridinones, it probably has no decisive meaning as it does not correlate with their antitumour potency [7]. Furthermore, imidazoacridinones inhibit the catalytic activity of topoisomerase II and all biologically active derivatives stimulate the formation of cleavable complex in vitro whereas inactive derivatives do not [8]. It has been also found that 8-hydroxyimidazoacridinones induce the arrest of the cell cycle progression in G2 phase [9,10] and apoptosis in tumor cells [11]. The QSAR approach has revealed that lipophilic properties significantly influence the antitumour potency of imidazoacridinones only in the case of 8-hydroxy analogues. The antitumour activity of the remaining derivatives is generally lower and is not correlated with their lipophilicity [12]. Among factors responsible for the antitumour effect of imidazoacridinones also metabolic activation should be considered as an important preliminary step of their biochemical mechanism of action. We assume that metabolic transformation can facilitate or even enable the direct reaction of imidazoacridinones with their targets such as DNA and topoisomerase. The relevance of the metabolic activation to biological activity was evidenced in our laboratory earlier for antitumour drug Nitracrine [13 – 16] and for a number of anthracycline derivatives [17,18]. The studies on the metabolism were also reported for several antitumour agents which resembled imidazoacridinones in some elements of structure. It was shown, for instance, that 9-hydroxyellipticine was transformed by horseradish peroxidase (HRP) [19,20]. The product of such oxidation was more susceptible to attack by nucleophilic agents such as pyrimidine bases in DNA or sulfhydryl groups in proteins than the parent compound [21]. The enzymatic oxidation of mitoxantrone by HRP converted it into a reactive intermediate [22,23] which bound covalently to DNA and RNA [24]. The formation of the oxidation products and their reaction with DNA were also observed during the reaction of mitoxantrone with myeloperoxidase [25] and after metabolic activation in animal and human organisms [26,27]. There were also reports demonstrating that anthrapyrazoles underwent oxidation in rat hepatocytes [28] leading to the metabolic intermediate capable of reaction with a nucleophilic agent glutathione more readily than the parent compound [29]. Taking into account that (1) metabolic activation is suggested to be a preliminary step in biological action of several antitumour drugs, (2) the chemical structure of

Z. Mazerska et al. / Chemico-Biological Interactions 115 (1998) 1–22

4

imidazoacridinone resembles some structural elements of ellipticine, mitoxantrone and anthrapyrazoles, and (3) metabolism occurring in the case of ellipticine, mitoxantrone and anthrapyrazoles is a result of oxidative transformation, we have assumed that oxidative metabolism may represent a probable way of imidazoacridinone activation. Therefore, the presented studies are intended to evaluate whether imidazoacridinones undergo enzymatic oxidation and whether the susceptibility to this transformation relates to biological and antitumour activity of these compounds. We are also interested which elements of imidazoacridinone structure are involved in oxidative enzymatic transformation. For this purpose we selected HRP as the model enzymatic system which is widely used in studies on oxidative activation of antitumour drugs. The compounds chosen for the studies differ in their antitumour potency and contain various structural elements known to be crucial for imidazoacridinone activity (Table 1).

2. Materials and methods

2.1. Chemicals Imidazoacridinone derivatives were synthesized in the Department of Pharmaceutical Technology and Biochemistry of the Technical University of Gdan´sk Table 1 The structures and biological activity [1,2,6] of the studied imidazoacridinone derivatives

No.

Symbol

R8

R5

ECa50 (mM)

ILSb (%)

ODc (mg/kg)

1 2 3 4 5 6

C-1415 C-1330 C-1311 C-1212 C-1374 C-1371

H OCH3 OH H OCH3 OH

CH2CH2N(CH2CH3)2 CH2CH2N(CH2CH3)2 CH2CH2N(CH2CH3)2 CH2CH2CH2N(CH3)2 CH2CH2CH2N(CH3)2 CH2CH2CH2N(CH3)2

0.7 9 0.05 0.78 9 0.02 0.013 90.008 0.77 9 0.05 8.2 92.1 0.014 9 0.005

55 9 5 96 913 93 9 15 25 97 64 913 120 930

175 920 90 910 5 100 150 915 5

a

EC50, drug concentration that inhibits the increment of cellular proteins in the culture of leukaemia L-1210 cells by 50%. b ILS, lifespan increase observed in leukaemia P-388-bearing mice and treated by the drug studied. ILS is calculated as the ratio T/C, where T is the survival time of treated mice and C is the survival time of control untreated mice. c OD, optimal dose determined from the %ILS/dose relationship.

Z. Mazerska et al. / Chemico-Biological Interactions 115 (1998) 1–22

5

(Cholody). Horseradish peroxidase was purchased from Sigma. All other chemicals used were of analytical grade. Methanol and water used in high-performance liquid chromatographic (HPLC) anlysis were HPLC grade (Sigma).

2.2. Procedures Stock solutions of imidazoacridinone derivatives (10 and 20 mM) and enzyme (1 mg/ml and 2 mg/ml) were prepared in 0.05 M phosphate buffer, pH 7.4. Hydrogen peroxide (H2O2) stock solution (100 mM) was prepared in distilled water.

2.3. Spectrophotometric monitoring Incubation mixtures containing imidazoacridinone derivatives (0.1 mM) and H2O2 (0.1 – 0.5 mM) were prepared by dilution of the stocks with 0.05 M phosphate buffer. The final volumes of the samples were 2 ml; experiments were performed at 37°C. Oxidation was initiated by the addition of the proper amount of the stock HRP to a final concentration of 2–10 mg/ml. The reactions were followed by UV-VIS spectroscopy with the aid of the UV-VIS Varian-Techtron 635 Spectrophotometer, using standard quartz cuvettes. The concentrations of the substrates presented in Figs. 2 – 4 were calculated according to the equation: x = (at − a )/ (a0 − a ). where a0 is absorbance of 0.1 mM imidazoacridinone solution at 420 nm, a is absorbance of the incubation solution extrapolated to t , and at is absorbance measured during the reaction after tt.

2.4. Electron paramagnetic resonance (EPR) experiments Incubation mixtures containing imidazoacridinone derivatives (2.5 mM or 5 mM), H2O2 (5 mM, 10 mM or 20 mM) and HRP were prepared in 0.05 M phosphate buffer. The final volume of the samples was 0.2 ml. The samples were kept frozen before addition of the enzyme. Reaction was initiated by the addition of the proper amount of HRP (final concentration was 0.13, 0.26, 0.33 or 0.66 mg/ml). EPR spectra were recorded with the Varian-E4 instrument using capillaries in quartz tubes in a microwave cavity. Varian-E4 instrument was operated at 9.57 GHz with 5 G field modulation.

2.5. HPLC analysis Enzymatic reactions were followed by the reverse-phase HPLC using a Spherisorb ODS2 analytical column (25× 0.4 mm, C18; Jones, UK) as stationary phase and isocratic elution at 1 ml/min with methanol–0.05 phosphate buffer, pH 2.5 (7:3) and 0.01% of diethylamine. A Waters HPLC system including 600K system controller, U6K pump system and 991 multidiode array detector was employed in these experiments.

6

Z. Mazerska et al. / Chemico-Biological Interactions 115 (1998) 1–22

3. Results This study was designed to evaluate whether imidazoacridinones undergo enzymatic oxidation and to clarify two questions concerning this transformation: (1) which of the structural elements are involved in enzymatic oxidation, and (2) whether there exists a relationship between antitumour activity of these compounds and their susceptibility to oxidative transformation. Therefore, the selected derivatives are of two types: they differ either in the substituent in position 8 of imidazoacridinone core or in the structure of aminoalkyl side-chain (Table 1). Thus, not only the highly active 8-hydroxyimidazoacridinones but also derivatives with methoxyl substituent or with hydrogen atom attached in position 8 displaying lower biological activity are included. Methoxy analogues are chosen because such a substituent is characterized by the field electronic properties similar to those of hydroxyl group while it differs in the resonance electronic properties. The diethylaminoethylamino and dimethylaminopropylamino side-chains in position 5 of imidazoacridinone are considered since the presence of the former increases the antitumour activity of imidazoacridinones to higher extent than the presence of the latter. We also intended to learn whether the structure of alkylamino side-chain influences the ability of imidazoacridinones to undergo enzymatic oxidation.

3.1. Spectrophotometric analysis Activation of 5-diethylaminoethylamino derivatives of imidazoacridinone by HRP results in significant changes of the absorption spectra of these compounds (Fig. 1). The changes in absorbance observed during the oxidation of 8-methoxy (C-1330) and unsubstituted in position 8 (C-1415) analogues of imidazoacridinone are similar in the character, whereas the ones observed in the case of 8-hydroxy derivative (C-1311) are clearly distinct. In the latter case, the decrease of the drug absorbance at 425 nm and simultaneous appearance of wide flat band near 500 nm is observed (Fig. 1A). The spectrum of 8-methoxy derivative displays changes at 425 nm and two isosbestic points at 395 and 490 nm (Fig. 1B). Such isosbestic points as well as the significant decrease of absorbance at 425 nm, similarly to 8-methoxy compound appear also in the spectra of unsubstituted analogue (Fig. 1C). The oxidation of 8-methoxy and unsubstituted analogues, contrary to the oxidation of 8-hydroxy derivative, does not lead to the appearance of any new band near 500 nm. The comparison of substrate concentration changes during the reaction for the same compounds, which are considered in the Fig. 1 is presented in Fig. 2. Again, the rate of oxidation of 8-hydroxy derivative is considerably higher compared to that measured for other compounds studied. Transformation to the products is accomplished within several minutes in the case of 8-hydroxyimidazoacridinone, whereas 8-methoxy and unsubstituted derivative did not reach a plateau stage even after two hours. Similar absorbance (Fig. 1) and concentration (Fig. 2) changes as those presented in Figs. 1 and 2 are also observed in the case of 8-hydroxy, 8-methoxy and 8-H derivatives of 5-dimethylaminopropylamino imidazoacridinone (data not shown). Therefore, the above results demonstrate that

Z. Mazerska et al. / Chemico-Biological Interactions 115 (1998) 1–22

7

Fig. 1. The sequence of absorption spectra taken during the oxidation of imidazoacridinone derivatives in the presence of HRP–H2O2 system. The incubation mixture contained 0.1 mM imidazoacridinone derivatives, 5 mg/ml of the enzyme in 0.05 M phosphate buffer, pH 7.4, and H2O2: (A) 0.1 mM; (B) 0.3 mM; (C) 0.5 mM. The spectra were recorded 1, 5, 10, 20, 30, 60, 90 and 120 min after the addition of the enzyme. Arrows indicate the direction of spectrum changes.

contrary to the role of 8-hydroxyl group, the differences in the structure of alkylamino side-chain in position 5 of imidazoacridinone ring have impact on neither the type of biotransformation nor on the rate of HRP oxidation. The rate of enzymatic oxidation of the studied compounds depended on H2O2 concentration as well as on the amount of HRP used. The obtained relationships between the concentration of the drug and time for different concentrations of H2O2 are presented in Fig. 3, whereas Fig. 4 shows how the enzyme concentration influences the rate of the reaction. It is evident that 8-hydroxyimidazoacridinone derivative undergoes oxidation significantly easier under the experimental condi-

8

Z. Mazerska et al. / Chemico-Biological Interactions 115 (1998) 1–22

tions used than the remaining derivatives. This holds both in the case of 5-diethylaminoethylamino (Figs. 3 and 4) as well as of 5-dimethylaminopropylamino side-chains attached to imidazoacridinone core (data not shown).

3.2. Electron paramagnetic resonance (EPR) studies The HRP catalysed oxidations of imidazoacridinone derivatives have been also followed by EPR spectroscopy. Mitoxantrone, an antitumour agent, which has been investigated elsewhere [22,23] was included into these measurements as a reference compound. The EPR spectra obtained during our experiments on oxidation of mitoxantrone resembled those reported by Reszka et al. [23] and are presented in Fig. 5. In the case of mitoxantrone the maximum intensity of the signal has been reached after 7 min of incubation while none of imidazoacridinone derivatives studied (Table 1) exhibited any significant EPR signals under the same conditions (data not shown). Therefore, the reaction was repeated at lower temperature. Fig. 6A shows EPR signals recorded at 4°C for 8-hydroxyimidazoacridinone, C-1311, just after the initiation of the reaction and after 3, 5 and 6 min from the initial point. Contrary to the reaction course detected for mitoxantrone, in the case of 8-hydroxyimidazoacridinone, the intensity of EPR signals dropped from the maximum at the starting point to lower values after several minutes of incubation. Similar result was obtained for another 8-hydroxy derivative (C-1371), which possesses different aminoalkyl side-chain (data not shown). No EPR signals were observed when any component of the incubation mixture was omitted.

Fig. 2. The time course of oxidation of imidazoacridinone derivatives by HRP – H2O2 system: ( +) C-1311, 0.1 mM H2O2; () C-1330, 0.3 mM H2O2; (’) C-1415, 0.5 mM H2O2. The incubation mixture contained 0.1 mM imidazoacridinone derivatives, 5 mg/ml of HRP in 0.05 M phosphate buffer, pH 7.4, and H2O2 as indicated.

Z. Mazerska et al. / Chemico-Biological Interactions 115 (1998) 1–22

9

Fig. 3. The influence of H2O2 concentration on the oxidation of imidazoacridinones studied by HRP–H2O2 system. The incubation mixture contained 0.1 mM imidazoacridinone derivative, 5 mg/ml of HRP and H2O2: (+ ) 0.1 mM; () 0.2 mM; (’) 0.5 mM in 0.05 M phosphate buffer, pH 7.4.

The HRP oxidation of 8-methoxy (C-1330) and unsubstituted in position 8 imidazoacridinone (C-1415) derivatives was also studied at lower temperature. Under such conditions, only slight EPR signals were observed for the both derivatives. These spectra are presented in Fig. 6, part B and C. The similar slight EPR signals were obtained in the case of 8-methoxy (C-1374) and unsubstituted (C-1212) imidazoacridinone derivatives that differ from the former in the structure of aminoalkyl side-chain (data not shown). From the EPR results one can conclude that methoxy and unsubstituted analogues of imidazoacridinone resemble each other in their slight capability of radical formation and they behave in a different manner compared to 8-hydroxy derivatives.

10

Z. Mazerska et al. / Chemico-Biological Interactions 115 (1998) 1–22

Fig. 4. The influence of HRP concentration on the course of oxidation of imidazoacridinones studied: (+ ) C-1311; () C-1330; (’) C-1415. The incubation mixture contained 0.1 mM imidazoacridinone derivatives and 0.2 mM H2O2 in 0.05 M phosphate buffer, pH 7.4.

3.3. HPLC analysis of the oxidation course. The HPLC studies of the reaction catalysed by HRP were limited to three compounds: 8-hydroxy- (C-1311), 8-methoxy- (C-1330) and unsubstituted in position 8 derivative (C-1415), all of which possess the same side-chain in the position 5 (i.e. diethylaminoethylamino group; see Table 1). Nonetheless, the results should hold also for derivatives with other alkylamino chains in position 5 because we did not observe, as it was shown above, any significant influence of different alky-

Fig. 5. EPR spectra taken during the oxidation of mitoxantrone with HRP – H2O2 system. The incubation mixture contained 2.5 mM mitoxantrone, 10 mM H2O2, 330 mg/ml of the enzyme in 0.1 M acetate buffer, pH 7.4. The spectra were recorded 2, 4, 6 and 8 min after the addition of the enzyme. Arrows indicate the direction of spectrum changes. Instrumental conditions: field 3381 G, scan range 100 G, modulation amplitude 5 G, gain 8× 103, scan time 2 min, microwave power 20 mW.

Z. Mazerska et al. / Chemico-Biological Interactions 115 (1998) 1–22

11

Fig. 6. EPR spectra taken during the oxidation of imidazoacridinone derivatives by HRP – H2O2 system. The incubation mixture contained 5 mM imidazoacridinone derivatives, 20 mM H2O2, 600 mg/ml of the enzyme in 0.5 M phosphate buffer, pH 7.4. The EPR spectra were recorded 2, 4, 6 and 8 min after the addition of the enzyme. Arrows indicate the direction of spectrum changes. Instrumental conditions: field 3381 G, scan range 100 G, modulation amplitude 5 G, gain 8 ×103, scan time 2 min, microwave power 20 mW.

lamino side-chains on oxidative tansformation of the compounds studied. The enzymatic oxidation and HPLC analysis of its route were performed using many different experimental protocols and the results obtained under similar conditions for all three compounds investigated are presented in Figs. 7–10. The chosen HPLC chromatograms corresponding to the reaction course observed in the case of 8-hydroxy derivative are presented in Fig. 7. They include a chromatogram of the substrate (Fig. 7A), three chromatograms obtained during the oxidation carried out with the excess of H2O2 (Fig. 7B–D) and one chromatogram

12

Z. Mazerska et al. / Chemico-Biological Interactions 115 (1998) 1–22

Fig. 7. HPLC chromatograms (reversed phase, methanol – 0.05 M phosphate buffer, pH 2.5, 7:3 and 0.01% of diethylamine; detection at 265 nm) taken during the oxidation of 8-hydroxyimidazoacridinone derivative, C-1311, by HRP–H2O2 system. The incubation mixture contained 0.1 mM imidazoacridinone derivative, 5 mg/ml of HRP and H2O2: (A – D) 0.5 mM, (E) 0.1 mM, in 0.05 M phosphate buffer, pH 7.4.

Z. Mazerska et al. / Chemico-Biological Interactions 115 (1998) 1–22

13

Fig. 8. HPLC chromatograms (reversed phase, methanol – 0.05 M phosphate buffer, pH 2.5, 7:3 and 0.01% of diethylamine; detection at 265 nm) taken during the oxidation of 8-methoxyimidazoacridinone derivative, C-1330, by HRP–H2O2 system. The incubation mixture contained 0.1 mM imidazoacridinone derivative, 5 mg/ml of HRP and H2O2: (A – D) 0.5 mM, (E) 0.1 mM, in 0.05 M phosphate buffer, pH 7.4.

14

Z. Mazerska et al. / Chemico-Biological Interactions 115 (1998) 1–22

Fig. 9. HPLC chromatograms (reversed phase, methanol – 0.05 M phosphate buffer, pH 2.5, 7:3 and 0.01% of diethylamine; detection at 265 nm) taken during the oxidation of unsubstituted in position 8 imidazoacridinone derivative, C-1415, by HRP – H2O2 system. The incubation mixture contained 0.1 mM imidazoacridinone derivative, 5 mg/ml of HRP and H2O2: (A – D) 0.5 mM, (E) 0.1 mM, in 0.05 M phosphate buffer, pH 7.4.

Z. Mazerska et al. / Chemico-Biological Interactions 115 (1998) 1–22

15

for the reaction performed using equimolar ratio of the substrates (Fig. 7E). the first chromatogram of the reaction mixture which was taken as soon as it was possible (Fig. 7B) revealed two new products and the substrate whose concentration

Fig. 10. The absorbance spectra taken for chromatogram peaks corresponding to the products of oxidation of imidazoacridinone studied by HRP – H2O2 system: (A) products of oxidation of C-1311 the course of which is presented in Fig. 7; (B) products of oxidation of C-1330 the course of which is presented in Fig. 8; (C) products of oxidation of C-1415 the course of which is presented in Fig. 9. The notations of the products are the same as those used in Figs. 7 – 9.

16

Z. Mazerska et al. / Chemico-Biological Interactions 115 (1998) 1–22

significantly decreased. Five minutes later the chromatogram contained several slightly separated peaks, which is typical for a mixture of many products very difficult to separate. After 15 min of reaction time, three main products were observed and the peak of the substrate became blended with others. It is worthy of notice (Fig. 7B,C) that the heights of the chromatographic peaks of the products (measured as the absorbance level at 260 nm) are lower more than ten times than the height of the initial peak of the substrate. Assuming that the values of molar absorption coefficients of substrate and products are close, it can be suggested that observed products p1, p2 and p3 represent only some of many products, which could not be separated under conditions applied. On the other hand, the equimolar ratio of C-1311 and H2O2 led to another two products, p4 and p5. Products p4 and p5 turned out to be stable within 60 min and, what is more, their concentrations were not as low as in the case of H2O2 excess. The UV-VIS spectra of products formed during the reaction of 8-hydroxy derivative with HRP are presented in Fig. 10A. It can be seen that only one product, p2, displayed spectrum different from the substrate. The remaining products were characterized by the spectra almost identical with that of the parent compound (C-1311). The set of chromatograms presented in Fig. 8A–E was obtained during monitoring of the reaction course for 8-methoxy derivative (C-1330). The HRP oxidation pathway of this derivative significantly differed from the former. First of all, the rate of oxidation was significantly lower and this was observed in the case of reaction carried out both in the presence of fivefold H2O2 excess (Fig. 8B–D) as well as for the equimolar drug:H2O2 ratio (Fig. 8E). The concentration of the substrate dropped a little after 4 h with concomitant formation of three products. The retention times of peaks p1, p2, p3 were lower than the substrate retention time. As can be seen in Fig. 10B, only p3 was characterized by UV-VIS spectrum different from that of the substrate. The compound without substituent in position 8 (C-1415) behaved in the presence of HRP similarly to C-1330. Fig. 9A–D show that reaction proceeded slowly and the character as well as concentration of products obtained with the excess of H2O2 were comparable with those observed for C-1330. The changes in spectra were observed only in the case of one product, p2 (Fig. 10C). However, it can be noticed that the low amount of H2O2 in the reaction environment induced the formation of only two products, p1 and p2 (Fig. 9E), identical with those formed in the presence of H2O2 excess.

4. Discussion In this study we have demonstrated that antitumour imidazoacridinone derivatives undergo enzymatic oxidation induced by HRP. The enzymatic transformation seems to follow different routes depending on the structure of a substituent attached to position 8 of imidazoacridinone core. Taking up the study on metabolic oxidation of imidazoacridinones we considered at least two possible pathways of such biotransformation: oxidation of imida-

Z. Mazerska et al. / Chemico-Biological Interactions 115 (1998) 1–22

17

Scheme 1.

zoacridinone heterocyclic core as well as oxidation of alkylamino side-chain. Our assumptions have been based on the structural similarities of imidazoacridinone (1) and some other antitumour compounds, 9-hydroxyellipticine (2), mitoxantrone (4) and anthrapyrazole (8), for which it has been suggested that metabolic oxidation is an important step in biochemical mechanism of their antitumour activity. It has been shown that hydroxy derivative of ellipticine (2) is oxidized by HRP to its quinone-imine structure (3) (Scheme 1) [19,20]. This type of transformation could be expected also for the studied 8-hydroxyimidazoacridinones because their core resembles the p-hydroxyimino system of ellipticine. On the other hand, the sidechain of mitoxantrone (4), which is similar to the one of imidazoacridinones, undergoes two oxidation steps induced by HRP (Scheme 2). The first results in cyclization (5), which is followed by further oxidation to diimine cation. Finally, the stable product of reversible reduction is formed, and this product has been isolated [22– 25]. Since the structure of the side-chain of the studied imidazoacridinones is similar to that of mitoxantrone, we have suspected that in the case of our compounds cyclization might be also possible. In addition, it has been reported that anthrapyrazoles (6), sharing also some structural similarities with imidazoacridinones, undergo oxidation in rat hepatocytes [28] but no intramolecular cyclic oxidation product has been detected (Scheme 3). Although the postulated metabolic intermediates in the form of diimine-like species (7) have not been isolated, the metabolic products of subsequent intermolecular nucleophilic addition with glu-

Scheme 2.

18

Z. Mazerska et al. / Chemico-Biological Interactions 115 (1998) 1–22

Scheme 3.

tathione have been identified [29]. Therefore, we cannot exclude the possibility that oxidation to p-2,5-diimine derivative might take place also in the case of the studied imidazoacridinones (1). Spectrophotometric monitoring of the reaction catalysed by HRP revealed that the oxidation of the studied compounds (Table 1) proceeds with the changes in their UV-VIS spectra which indicates the significant transformations of the structure of imidazoacridinone chromophore. The type and rate of observed transformations depend on the reaction conditions as well as on the substituent attached to imidazoacridinone ring. 8-Hydroxy derivatives undergo the oxidation markedly faster than their methoxy or unsubstituted analogues. The results of HPLC analysis confirmed and extended the above conclusions. It turned out that enzymatic oxidation of 8-hydroxy derivative (C-1311), which occurred in the presence of the excess of H2O2, gave rise to the mixture of several products immediately after the reaction initiation. Ten minutes later, the majority of unstable products disappeared rendering their HPLC separation impossible. The UV-VIS spectra obtained for the chromatographic peaks revealed that in the reaction mixture there were products with transformed chromophores as well as with the chromophore close to that of the substrate. We presume that the unusually high reactivity of 8-hydroxy derivative observed in the presence of excess of H2O2 is due to the formation of reactive quinone-imine like form of oxidized 8-hydroxyimidazoacridinone. The mesomeric structures of this presumed intermediate of HRP oxidation are presented in Scheme 4. The results of our electrochemical studies suggest that the first mesomeric structure is the most probable. On the other hand, the equimolar amount of H2O2 in the reaction environment induces low rate of the reaction and leads to the stable products. As can be presumed based on UV-VIS spectra, the structure of chromophore of these products appears to be very similar to that of the substrate. This indicates that the

Scheme 4.

Z. Mazerska et al. / Chemico-Biological Interactions 115 (1998) 1–22

19

structure of heterocyclic core of both products obtained under such conditions remains unchanged, whereas changes in the retention time reflect the increase of molecule polarity. Therefore, for equimolar drug:H2O2 ratio, oxidative transformation of 8-hydroxy derivative (C-1311), occurs only in the aminoalkyl side-chain. We suggest that oxidation of the terminal alkylamino group followed by dealkylation seems to be the most probable pathway leading to the formation of the detected products. HRP oxidation of imidazoacridinones devoid of hydroxyl substituent is far more slower than of 8-hydroxy derivatives. The lower rate was observed when the excess of H2O2 was used as well as with equimolar amounts of the substrates. The mixtures of oxidation products contained both: compounds with intact and with transformed chromophore. All products display higher polarity compared to the substrate. Again, we propose the oxidation of the terminal alkylamino group followed by dealkylation as a possible reaction pathway in the case of products with chromophore similar to that of the substrate. However, the absence of easily oxidized hydroxyl substituent in imidazoacridinone core suggests that oxidation of methoxy- and unsubstituted imidazoacridinone derivatives must proceed via another pathway than this proposed for hydroxy analogues. We presume that the oxidation of p-2,5-diimine system is possible in this case as it was postulated for anthrapyrazoles (9) [29]. The cyclization product similar to that evidenced for mitoxantrone (5) seems rather unlikely. Had cyclization occurred, clear changes in the reaction course should have been observed between derivatives with diethylaminoethylamino and dimethylaminopropylamino side-chain which was not the case in our experiments. Detection of EPR signals during the oxidation catalysed by HRP (Fig. 6) indicates that free radicals are formed during this reaction and this confirms that imidazoacridinone molecules undergo oxidative transformation. The absence of hyperfine structure in EPR spectra of imidazoacridinones may indicate that these radicals do not rotate freely in solution. This suggests that radicals exist in an aggregated form as imidazoacridinones dimerize even in very dilute aqueous solutions (J. Dzie˛gielewski, unpublished data). The comparison of EPR signals brought about by different imidazoacridinones have revealed that 8-hydroxy derivatives give rise to significantly higher concentration of radicals than the remaining analogues. It means that 8-hydroxy analogues exhibit the highest reactivity during HRP oxidation. Therefore, EPR results support the substantial role of 8-hydroxyl group in the course of enzymatic oxidation by HRP. Hence, the presented above conclusion that HRP oxidation pathway for imidazoacridinones is different from the one observed for mitoxantrone has been confirmed also by means of EPR. The concentration of radicals decreases fast during the reaction of imidazoacridinone whereas it increases in the case of mitoxantrone. Besides, 8-hydroxyimidazoacridinones have turned out to be more reactive than mitoxantrone as changes in EPR spectra of the former compounds occur faster than of the latter. The results presented in this paper demonstrate that there exists the relationship between biological activity of imidazoacridinones and their susceptibility to enzy-

20

Z. Mazerska et al. / Chemico-Biological Interactions 115 (1998) 1–22

matic oxidation by HRP. The highly active 8-hydroxy derivatives undergo the oxidative transformation far more readily than their methoxy- and unsubstituted analogues. However, the considered structures of aminoalkyl side-chain seem to have no impact on susceptibility to enzymatic oxidation, at least under condition investigated. The same can be said as regards antitumour potency of imidazoacridinones (Table 1). Therefore the oxidative metabolism of imidazoacridinones, especially of compound C-1311, could play very essential role in their biological action. In conclusion, the presented results of enzymatic oxidation catalysed by HRP confirm the hypothesis that antitumour imidazoacridinones undergo oxidative biotransformation. The ability to be transformed is in reasonable correlation with antitumour activity of these compounds. We have demonstrated that the biotransformation of each imidazoacridinone occurs via at least two pathways, in which molecular chromophore as well as alkylamino side-chain are involved. We have also shown that the highly susceptible to oxidation 8-hydroxy derivatives undergo this biotransformation by different mechanism depending on the reaction conditions. We have suggested that the quinone-imine structure of imidazoacridinone ring presented in Scheme 4 could represent the reactive intermediate which is responsible for exceptionally high reactivity of 8-hydroxy derivative when incubated with the excess of H2O2. The oxidation of the terminal alkylamino group which may be followed by dealkylation is proposed as a metabolic pathway during HRP oxidation carried out with equimolar ratio of 8-hydroxyimidazoacridinones to H2O2.

Acknowledgements This work was supported by the Committee for Scientific Research (KBN), Poland, grants no. 4P0504310 (1995) and no. 3 T09A06713 (1997). We are grateful to Dr A. Bartoszek for careful reading of this manuscript.

References [1] W.M. Chołody, S. Martelli, J. Łukowicz, J. Konopa, 5-[(Aminoalkyl)amino]-imidazo[4,5,1de]acridin-6-ones as a novel class of antineoplastic agents. Synthesis and biological activity, J. Med. Chem. 33 (1990) 49–52. [2] W.M. Chołody, S. Martelli, J. Konopa, Chromophore-modified antineoplastic imdazoacridinones. Synthesis and activity against murine leukemias, J. Med. Chem. 35 (1992) 378 – 382. [3] W.M. Chołody, B. Horowska, J. Łukowicz, S. Martelli, J. Konopa, Structure – activity relationship for antineoplastic imidazoacridinones: synthesis and antileukemic activity in vivo, J. Med. Chem. 39 (1996) 1028–1032. [4] H. Kus´mierczyk, W.M. Chołody, J. Łukowicz, C. Radzikowski, J. Konopa, Experimental antitumor activity and toxicity of selected triazolo- and imidazoacridinones, Arch. Immunol. Ther. Exper. 42 (1994) 415–423. [5] A.M. Burger, J.A. Double, J. Konopa, M.C. Bibby, Preclinical evaluation of novel imidazoacridinone derivatives with potent activity against experimental colorectal cancer, Br. J. Cancer 74 (1996) 1369–1374.

Z. Mazerska et al. / Chemico-Biological Interactions 115 (1998) 1–22

21

[6] B. Berger, H. Marquardt, J. Westendorf, Pharmacological and toxicological aspects of new imidazoacridinone antitumor agents, Cancer Res. 56 (1996) 2094 – 2104. [7] J. Dzie˛gielewski, A. Składanowski, J. Konopa, Noncovalent binding of potent antitumor imidazoacridinones to DNA, Proceedings of 8th NCI-EORTC Symposium on New Drugs in Cancer Therapy, Amsterdam, 12–15 March 1996, p. 82. [8] A. Skladanowski, S.Y. Plisov, J. Konopa, K.L. Larsen, Inhibition of DNA topoisomerase II by imidazoacridinones, new antineoplastic agents with strong activity against solid tumors, Mol. Pharmacol. 49 (1996) 772–780. [9] E. Augustin, D.N. Wheatley, J. Lamb, J. Konopa, Imidazoacridinones arrest cell-cycle progression in the G2 phase of L1210 cells, Cancer Chemother. Pharmacol. 38 (1996) 39 – 44. [10] D.N. Wheatley, J. Lamb, Cell killing by the novel imidazoacridinone antineoplastic agent, C-1311, is inhibited at high concentration coincident with dose-differentiated cell cycle perturbation, Br. J. Cancer 74 (1996) 1359–1361. [11] E. Augustin, J. Konopa, Imidazoacridinones induce apoptosis in murine leukemia L1210 cells, Folia Cytochem. Cytobiol. 34 (Suppl. 2) (1996) 56. [12] Z. Mazerska, E. Augustin, J. Dzie˛gielewski, M.W. Chołody, J. Konopa, QSAR of acridines. III. Structure–activity relationship for antitumour imidazoacridinones and intercorrelations between in vivo and in vitro tests, Anti-Cancer Drug Design 11 (1996) 73 – 88. [13] J.W. Pawlak, J. Konopa, In vitro binding of metabolically activated [14C]-Ledakrin, or 1-nitro-914 C-(3%-dimethylamino-N-propylamino)acridine, a new antitumor and DNA cross-linking agent, to macromolecules of subcellular fractions isolated from rat liver and HeLa cells, Biochem. Pharmacol. 28 (1979) 3391–3402. [14] J.W. Pawlak, K. Pawlak, J. Konopa, The mode of action of cytotoxic and antitumor 1-nitroacridines. II. In vivo enzyme-mediated covalent binding of a 1-nitroacridine derivative, Ledakrin or Nitracrine, with DNA and other macromolecules of mammalian or bacterial cells, Chem.-Biol. Interact. 43 (1983) 151–173. [15] J.W. Pawlak, K. Pawlak, J. Konopa, The mode of action of cytotoxic and antitumor 1-nitroacridines. III. In vivo interstrand cross-linking of DNA of mammalian or bacterial cells by 1-nitroacridines, Chem.-Biol. Interact. 43 (1983) 173 – 197. [16] K. Pawlak, J.W. Pawlak, J. Konopa, Cytotoxic and antitumor activity of 1-nitroacridines as an after effect of their interstrand DNA cross-linking, Cancer Res. 44 (1984) 4289 – 4296. [17] A. Składanowski, J. Konopa, Interstrand DNA crosslinking induced by anthracyclines in tumor cells, Biochem. Pharmacol. 47 (1994) 2269 – 2278. [18] A. Składanowski, J. Konopa, Relevance of interstrand DNA crosslinking induced by anthracyclines for their biological activity, Biochem. Pharmacol. 47 (1994) 2279 – 2287. [19] C. Auclair, C. Paoletti, Bioactivation of the antitumor drugs 9-hydroxyellipticine and derivatives by a peroxidase–hydrogen peroxide system, J. Med. Chem. 24 (1981) 289 – 295. [20] G. Meunier, D. Montauzon, J. Bernadou, G. Grassy, M. Bonnafous, S. Cros, B. Meunier, The biooxidation of cytotoxic ellipticine derivatives: a key to structure – activity relationship studies, Mol. Pharmacol. 33 (1987) 93–102. [21] G. Meunier, B. Meunier, Ch. Auclair, J. Bernadou, C. Paoletti, Unexpected regionspecific alkylation of the antitumor agent N 2-methyl-9-hydroxyellipticinium acetate with N, O or S donors, Tetrahedron Lett. 24 (1983) 365–368. [22] K. Reszka, P. Kołodziejczyk, J.W. Lown, Enzymatic oxidative activation and transformation of the antitumor agent mitoxantrone, Free Radic. Biol. Med. 2 (1986) 13 – 25. [23] K. Reszka, P. Kołodziejczyk, J.W. Lown, Horseradish peroxidase-catalysed oxidation of Mitoxantrone: spectrophotometric and electron paramagnetic resonance studies, Free Radic. Biol. Med. 2 (1986) 25–32. [24] K. Reszka, J.A. Hartley, P. Kołodziejczyk, W. Lown, Interaction of the peroxidase derived metabolite of mitoxantrone with nucleic acids. Evidence for covalent binding of 14C-labeled drug, Biochem. Pharmacol. 38 (1989) 4253 – 4260. [25] C. Panousis, A.J. Kettle, D.R. Philips, Myeloperoxidase oxidizes mitoxantrone to metabolites which bind covalently to DNA and RNA, Anti-Cancer Drug Design 10 (1996) 593 – 605.

22

Z. Mazerska et al. / Chemico-Biological Interactions 115 (1998) 1–22

[26] K. Mewes, J. Blanz, G. Ehniger, R. Gebhard, K.P. Zeller, Cytochrome P-450 induced cytotoxicity of mitoxantrone by formation of electrophilic intermediates, Cancer Res. 53 (1993) 5135 – 5142. [27] J. Blanz, K. Mewes, G. Ehninger, B. Proksch, D. Waidelich, B. Greger, K.P. Zeller, Evidence for oxidative activation of mitoxantrone in human, pig and rat, Drug. Metab. Dispos. 19 (1991) 871–880. [28] K. Reszka, P. Kołodziejczyk, J.A. Hartley, W.D. Wilson, J.A. Lown, Molecular pharmacology of anthracenedione-based anticancer agents, in: J.W. Lown (Ed.), Anthracycline and Anthracenedione-based Anticancer Agents, Elsevier, Amsterdam, 1988, pp. 401 – 445. [29] U. Renner, J. Blanz, S. Freund, D. Widelich, G. Ehninger, K.P. Zeller, Biotransformation of CI-937 in primary cultures of rat hepatocytes. Formation of glutathione conjugates, Drug. Metab. Dispos. 23 (1995) 94–101.

.