Introduction

Abstract Introduction Materials & Methods Results Discussion References

Many consider malaria to be the most important infectious disease in the world. The antimalarial quinine still remains to be an important drug of choice for the treatment of severe and complicated malaria. Two recent clinical trials have shown that quinine is as effective as artemether (a qinghaosu derivative), a promising antimalarial that is effective against Plasmodium falciparum malaria (Hien and White, 1993), in children with cerebral malaria (Boele van Hensbroek et al., 1996) and in adults with severe falciparum malaria (Hien et al., 1996). In addition, the resistance of P. falciparum to chloroquine, mefloquine and pyrimethamine/sulfadoxine has been rapidly increasing in endemic regions such as Southeast Asia, South America and East Africa (Moran and Bernard, 1989; Wernsdorfer, 1991; Bradley, 1993), and this has resulted in an increased use of quinine as an alternative drug for treating multidrug-resistant P. falciparum malaria (White, 1996; Tracy and Webster, 1996). Furthermore, quinine is available not only as an oral form but also as an injectable formulation for malaria patients, and it has fewer life-threatening side effects if used correctly and at the normal therapeutic doses (White, 1992; Tracy and Webster, 1996). Thus quinine is considered one of the most effective and convenient drugs for the treatment of malaria.

The primary route of systemic elimination of quinine in the human is via an extensive hepatic biotransformation to hydroxylated metabolites; less than 20% of the drug is excreted unchanged in the urine (White et al., 1982; Tracy and Webster, 1996). However, despite its long history (at least 350 years) in the treatment of malaria, it was not until recently that the detailed metabolism of quinine and the cytochrome P450 (CYP) isoform(s) involved were clarified in humans. The formation of 3-hydroxyquinine is the major metabolic pathway (Wanwimolruk et al., 1995; Wanwimolruk et al., 1996; Zhang et al., 1997) in the human, and the 3-hydroxylation is catalyzed mainly by CYP3A4 (Zhao et al., 1996; Zhang et al., 1997) and to a minor extent by CYP2C19 (Zhao et al., 1996) in human liver microsomes.

So far as we know, the detailed quinine metabolism, the kinetics of quinine 3-hydroxylation and the CYP isoform(s) involved in this metabolic pathway of quinine in animals have been neither elucidated nor compared among different animal species or between an animal species and the human. Thus the aims of this extended in vitro study were 1) to investigate and compare the formation and kinetics of quinine 3-hydroxylation by using different species (i.e., mouse, rat, dog and human) liver microsomes, and to identify the CYP isoform(s) involved in this metabolic pathway of quinine; and 2) to search for a suitable animal model for further study of quinine 3-hydroxylation and quinine-drug interactions in vitro.

	Materials and Methods

Abstract Introduction Materials & Methods Results Discussion References

Chemicals and reagents. Synthetic 3-hydroxyquinine was a generous gift from Dr. P. Winstanley (University of Liverpool, Liverpool, UK). Quinine, TAO, primaquine, doxycycline, chloroquine, diazepam and tetrabutylammonium bromide were purchased from Sigma Chemical Co. (St. Louis, MO). Quinidine, -naphthoflavone, coumarin and p-nitrophenol were purchased from Wako Pure Chemical Industries Ltd. (Osaka, Japan) and sulfaphenazole from Meiji Yakuhin Co. (Tokyo, Japan). Proguanil was kindly supplied by Zeneca Pharmaceuticals (Alderley Edge, UK). Acetonitrile, methanol and other reagents of analytical grade were purchased from Wako Pure Chemical Industries Ltd. NADP⁺, glucose-6-phosphate and glucose-6-phosphate dehydrogenase were obtained from Oriental Yeast (Tokyo, Japan). Racemic mephenytoin was kindly donated by Dr. Küpfer (University of Bern, Bern, Switzerland). S- and R-mephenytoin were separated from racemic mephenytoin on a Chiralcel OJ column (10 µm, 4.6 × 250 mm; Daicel Chemical Co. Ltd., Tokyo, Japan) according to the method of Yasumori et al. (1990). Rabbit polyclonal anti-rat CYP3A2, 2C11 and 2E1 antisera and preimmune serum were obtained from Daiichi Pure Chemicals Co., Ltd. (Tokyo, Japan).

Preparation of microsomal fractions. Human liver samples were obtained from 12 patients who underwent a partial hepatectomy for metastatic liver tumor(s) in the Division of General Surgery, International Medical Center of Japan, Tokyo, as reported previously (Chiba et al., 1993; Zhao et al., 1996). The liver parenchyma of non-tumour-bearing part used for the study was shown later to be histopathologically normal in all cases. Use of human samples for the study had been approved by the Institutional Ethics Committee of the International Medical Center of Japan. Animal liver samples were obtained from 11 male Crj/CD-1 mice, six male Wistar rats and five male beagle dogs (Japan River Co., Yokohama, Japan). Animals were fed a standard diet and had free access to drinking water. The animals were also allowed to acclimate in our animal facility for 1 week before the liver microsomal samples were prepared as described below.

Assay for quinine metabolism with liver microsomes. The assay of quinine metabolism was conducted in the same way as previously described (Zhao et al., 1996; Zhao and Ishizaki, 1997, in press). Microsomal fractions were incubated in the presence of an NADPH-generating system at 37°C for 10 to 15 min in test tubes. The incubation mixture consisted of 0.1 to 0.5 mg/ml microsomal protein, 4 mM MgCl₂, 0.5 mM NADP⁺, 2.0 mM glucose-6-phosphate, 1 IU/ml glucose-6-phosphate dehydrogenase, 100 mM potassium phosphate buffer (pH 7.4), 0.1 mM EDTA and 0.5 to 600 µM quinine, in a final volume of 250 µl. After incubation for 10 to 15 min, the reaction was stopped by addition of 500 µl of ice-cold methanol. The mixture was centrifuged at 1500 × g for 10 min, and the supernatant was injected onto a HPLC apparatus as described below.

Kinetics of the formation of 3-hydroxyquinine. Preliminary results indicated that the formation rates of 3-hydroxyquinine were linear at 37°C for incubation time up to 30 min (humans and dogs) or 10 min (rats and mice) and for microsomal protein concentration up to 0.25 mg/ml (humans and mice) or 1.0 mg/ml (dogs and rats) at the substrate quinine concentration of 50 µM. Accordingly, the kinetic studies were performed at 37°C with an incubation of 15 min (humans and dogs) or 10 min (rats and mice) at a microsomal protein concentration of 0.1 mg/ml (humans, rats and mice) or 0.5 mg/ml (dogs).

Inhibition study. The effects of coincubation of inhibitor or substrate probes specific for different human CYP isoforms on the microsomal metabolism of quinine were studied separately. The specific inhibitors used were -naphthoflavone for CYP1A (Kunze and Trager, 1993), sulfaphenazole for CYP2C9 (Goldstein and de Morais, 1994), quinidine for CYP2D6 (Kobayashi et al., 1989) and ketoconazole and TAO for CYP3A4 (Newton et al., 1995; Watkins et al., 1985). The substrate probes used were coumarin for CYP2A6 (Wrighton and Stevens, 1992), S-mephenytoin for CYP2C19 (Goldstein and de Morais, 1994) and p-nitrophenol for CYP2E1 (Tassaneeyakul et al., 1993). In addition, two so-called activators of CYP3A, -naphthoflavone (Chang et al., 1994; Shou et al., 1994) and diazepam (Andersson et al., 1994; Pearce et al., 1996), were used for testing the possible activation of quinine 3-hydroxylation with the liver microsomes of all species. The concentrations of quinine were chosen according to the respective K_m values obtained from the different species liver microsomes, and quinine was incubated with or without one of the inhibitor or substrate probes, at concentrations ranging from 0.0001 µM to 10 mM, under the incubation conditions described earlier. The effects of each compound on the formation of 3-hydroxyquinine at the respective inhibitor or substrate probe concentrations were compared with the control values determined from the incubation of quinine alone, and the inhibition values were expressed as a percentage of the respective control values. The inhibitor potency of the respective substrates were defined by IC₅₀ (i.e., a 50% inhibition of 3-hydroxyquinine formation compared with the control values). Experiments were performed with microsomal preparations obtained from three or four different livers in each species tested.

Immunoinhibition study. Anti-CYP polyclonal antisera directed against purified rat CYP3A2, 2C11 and 2E1 were raised in rabbits and used in this part of the study. Anti-CYP3A2, 2C11 and 2E1 antisera (50 µl) significantly and selectively inhibited testosterone 6-hydroxylation (>90%), testosterone 16 -hydroxylation (>90%) and aniline hydroxylation (>50%) by male rat liver microsomes, respectively, and cross-reacted with the orthologous human, dog and mouse isoforms, but not with other CYP isoforms (i.e., according to the product instructions of Daiichi Pure Chemicals Co., Ltd. Tokyo, Japan).

Metabolic drug interaction. Several antimalarial drugs, which have been used and may be coadministered with quinine for the treatment and/or chemoprophylaxis of malaria (White, 1988, 1992 and 1996; Bradley, 1993; Tracy and Webster, 1996), were assessed for their possible inhibitory effects on quinine 3-hydroxylation by using rat, dog and human liver microsomes. The drugs included primaquine, doxycycline, tetracycline, chloroquine and proguanil. Other antimalarials (such as mefloquine, sulfadoxine, qinghaosu and its derivatives, halofantrine, amodiaquine, amopyraquine and atovaquone) were not available for the study in Japan, nor were they obtainable as the respective pure chemicals from any chemical or pharmaceutical sources to which the authors had access.

Correlation between immunoquantified microsomal P450 3A and quinine 3-hydroxylation. We studied the correlation between immunoquantified microsomal P450 3A and the rate of quinine 3-hydroxylation by using eight different human liver microsomes, which were used for determining the kinetic parameters in the present study. The P450 3A contents in the eight liver microsomes were quantitated by immunoblotting, using specific antibodies as described previously (Berthou et al., 1994).

Statistics. All values are expressed as the mean ± S.D. The initial statistical analysis to evaluate the differences in the mean data among the different species was conducted by a two-way ANOVA. When this statistical analysis showed a significant difference, a two-sample t test was used for the between-species comparisons within treatments with the same inhibitor/substrate (or possible activator) probes, and a one-sample t test was used to compare the mean base-line and post-treatment values within the same species. Correlation between CYP3A level and 3-hydroxyquinine formation rate was determined by least-squares linear regression analysis. P < .05 was considered statistically significant.

	Results

Abstract Introduction Materials & Methods Results Discussion References

Metabolism of quinine. By using the optimized HPLC conditions described above, we observed similar metabolite formation (i.e., 3-hydroxylation) patterns of quinine within 60 min in all species tested. Quinine was extensively metabolized by the liver microsomes of all species, and the major peak had a retention time (16 min) identical to that of pure 3-hydroxyquinine (fig. 1). The formation of 3-hydroxyquinine was time-, NADPH- and microsomes-dependent (data not shown), which suggests the possible involvement of P450(s) in the metabolism.

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Representative chromatograms of 3-hydroxyquinine (peak 1) and quinine (peak 2) with liver microsomes obtained from mouse (panel A), rat (panel B), dog (panel C) and human (panel D).

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Representative Eadie-Hofstee plots for quinine 3-hydroxylation with microsomes obtained from mouse (), rat (), dog () and human () livers (panel A), and two representative Eadie-Hofstee plots for quinine 3-hydroxylation in mouse liver microsomes showing a biphasic profile (panel B,  = mouse liver 2,  = mouse liver 9). Of the 11 mouse liver microsomes tested, 4 showed such a biphasicity as illustrated in panel B, and the kinetic parameters were analyzed by using a two-enzyme kinetic approach (table 2).

Inhibition study. Two previous in vitro studies (Zhao et al., 1996; Zhang et al., 1997) have demonstrated that CYP3A4 is a principal isoform involved in quinine 3-hydroxylation with human liver microsomes. In order to determine whether the CYP3A subfamily isoform also plays a predominant role in the formation of 3-hydroxyquinine from quinine in other species liver microsomes tested, we used two typical CYP3A inhibitors, ketoconazole and TAO, to perform an inhibition study with microsomes obtained from mouse, rat and dog livers. The effects of coincubation with the inhibitors on quinine 3-hydroxylation are shown in figure 3, together with our published data derived from human liver microsomes (Zhao et al., 1996), in order to facilitate comparison with those obtained from animal liver microsomal samples. All the plots showed that quinine 3-hydroxylation was inhibited in a concentration-related manner. The mean IC₅₀ (± S.D.) values with mouse, rat, dog and human liver microsomes were 0.021 (± 0.005), 0.087 (± 0.012), 0.056 (± 0.010) and 0.026 (± 0.013) µM for ketoconazole and 6.3 (± 1.4), 43.0 (± 4.2), 0.80 (± 0.13) and 29.0 (± 4.7) µM for TAO, respectively. The mean maximum inhibition on quinine 3-hydroxylation in mouse, rat, dog and human liver microsomes was about 94%, 91%, 88% and 90% for ketoconazole and about 85%, 66%, 93% and 70% for TAO, respectively. These observations suggest the possibility that the CYP3A (for rats, dogs and humans) or Cyp3a (for mice) subfamily (Nelson et al., 1996) may be involved in quinine 3-hydroxylation.

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Effects of ketoconazole (panel A) and TAO (panel B) on quinine 3-hydroxylation in microsomes obtained from mouse (), rat (), dog () and human () livers. The plotted data are the mean values of experiments performed with microsomes obtained from three different livers of each species tested.

View larger version (27K): [in this window] [in a new window]   Fig. 4.   Effects of various CYP inhibitor and/or substrate probes on quinine 3-hydroxylation in mouse liver microsomes in which the Eadie-Hofstee plots for the formation of 3-hydroxyquinine showed a biphasic behavior as illustrated in figure 2B. The solid columns represent the use of 10 µM quinine as a substrate concentration, the open columns the use of 120 µM quinine. Data are the mean values of experiments performed with microsomes obtained from three different livers. * P < .05 and ** P < .01 compared with the base-line quinine 3-hydroxylation activity as the control.

View larger version (32K): [in this window] [in a new window]   Fig. 5.   Effects of -naphthoflavone (panel A) and diazepam (panel B) on quinine 3-hydroxylation in microsomes obtained from human (hatched columns), dog (cross-hatched columns), rat (solid columns) and mouse (open columns) livers. Data are expressed as the mean ± S.D. for microsomes obtained from three or four different livers of each species tested. * P < .05 compared with the respective control values within the same species. + P < .05 compared with the human or dog group within the same treatment (i.e., -naphthoflavone or diazepam). § P < .05 compared with the mouse group.

Immunoinhibition. The inhibition of quinine 3-hydroxylation by polyclonal antisera raised against purified rat CYP3A2 is shown in figure 6. The addition of anti-CYP3A antisera (25 µlthat is, 25 µg of IgG per microgram of microsomal protein) inhibited the 3-hydroxyquinine formation by about 96%, 84% and 92% in liver microsomes obtained from mouse, rat and dog, respectively, with no appreciable inhibition of quinine 3-hydroxylation by nonimmune IgG. These results were found to be in good agreement with our previous finding that polyclonal antibodies (10 mg of IgG per milligram of microsomal protein) raised against human CYP3A reduced the 3-hydroxylation activity of quinine by 72% in human liver microsomes (Zhao et al., 1996).

View larger version (16K): [in this window] [in a new window]   Fig. 6.   Immunoinhibition of quinine 3-hydroxylation by anti-CYP3A antisera with liver microsomes obtained from mouse, rat and dog. The open columns represent the use of nonimmune antisera, the solid columns the use of anti-CYP3A antisera. Data are the mean values of two different experiments with each species tested.

Metabolic drug interaction. Metabolic drug interactions between quinine and five other antimalarials that may be coadministered with quinine for the treatment and/or chemoprophylaxis of malaria (White, 1988, 1992 and 1996; Bradley, 1993; Tracy and Webster, 1996) were investigated by using rat, dog and human liver microsomes separately, each of which came from three different livers. The mean results are shown in figure 7, together with our recent data derived from human liver microsomes (Zhao and Ishizaki, 1997, in press) in order to facilitate comparison with those from animal liver microsomes. The results indicated that primaquine, doxycycline and tetracycline substantially inhibited the 3-hydroxyquinine formation (at least >55%) in both animal and human liver microsomes, although the respective IC₅₀ values obtained from rat, dog and human liver microsomes differed somewhat among them (table 3). In addition, chloroquine appreciably reduced the formation of 3-hydroxyquinine from quinine (at least >40%) in rat and dog liver microsomes, but not in human liver microsomes, which suggests an interspecies difference in the interaction potential of these antimalarials with quinine. However, proguanil did not inhibit the 3-hydroxylation activity of quinine in any liver microsomes tested (fig. 7).

View larger version (15K): [in this window] [in a new window]   Fig. 7.   Effects of selected antimalarials on quinine 3-hydroxylation with microsomes obtained from rat (panel A), dog (panel B) and human (panel C) livers.  = primaquine,  = doxycycline,  = tetracycline,  = proguanil and  = chloroquine. Data are the mean values of experiments performed with microsomes obtained from three different livers of each species tested.

Correlation between human CYP3A and quinine 3-hydroxylation. To confirm further the major role of human CYP3A4 isoform in quinine 3-hydroxylation, the CYP3A contents of each of eight human liver microsomes used for determining the kinetic parameters in the present study were measured by Western blot analysis (Berthou et al., 1994). A significant correlation (r = 0.968, P < .001) was found between the CYP3A contents and the 3-hydroxyquinine formation rates in the eight human liver microsomal samples (fig. 8), a result that strongly supports the previous finding that CYP3A is a principal isoform involved in the formation of 3-hydroxyquinine from quinine (Zhao et al., 1996; Zhang et al., 1997).

View larger version (16K): [in this window] [in a new window]   Fig. 8.   Correlation between CYP3A contents and 3-hydroxyquinine formation from quinine in eight human liver microsomal samples. Correlation coefficient (r) value was 0.968, P < .001.

	Discussion

Abstract Introduction Materials & Methods Results Discussion References

Recently, two in vitro studies (Zhao et al., 1996; Zhang et al., 1997) with human liver microsomes obtained from two distinct ethnic groups (Japanese and Caucasians) have revealed that quinine 3-hydroxylation, in both groups, is mediated mainly by human CYP3A4 with the approximate mean K_m and V_max/K_m (i.e., CL_int) values between each other. The current in vitro study is, so far as we know, the first to compare quinine 3-hydroxylation with liver microsomes obtained from different species (i.e., mouse, rat, dog and human) or to search for a suitable animal model resembling human in terms of the in vitro enzyme-kinetic parameters and quinine-drug interaction screenings. The results provided further in vitro evidence that the formation of 3-hydroxyquinine is predominantly mediated via the CYP3A/Cyp3a subfamily in all the species tested and that the dog and/or rat may be more suitable as an animal model for exploring the quinine 3-hydroxylase activity and for screening quinine-drug interactions in vitro, though the results will be at variance with the human liver microsomal data. Nevertheless, whether dogs would be better than rats for further in vitro metabolism research for quinine remains unanswered from the present study, because the quinine-drug interaction data obtained from dog and rat liver microsomes were similarly discrepant with those from the human liver microsomes, as discussed later.

With all species tested, the formation of 3-hydroxyquinine in liver microsomes showed a quite similar chromatogram pattern with 3-hydroxyquinine as the major metabolite of quinine (fig. 1). The rate of quinine 3-hydroxylation was also similar between dog and human and between mouse and rat liver microsomes (table 1). On the basis of these results, one may assume that, from a qualitative point of view, the dog, rat and even mouse could be an appropriate species for conducting an in vitro study on the metabolism of quinine. However, the dog may be quantitatively more suitable for exploring the quinine 3-hydroxylase activity in vitro, because the mean kinetic parameters for quinine 3-hydroxylation (K_m, V_max and V_max/K_m) derived from dog liver microsomes, as compared with the other two rodent species, were closer to those obtained from human liver microsomes (table 1).

We observed a potent and concentration-dependent inhibition of quinine 3-hydroxylation in all species when quinine was coincubated with two CYP3A inhibitors, ketoconazole and TAO (fig. 4), which indicates that CYP3A/Cyp3a is likely to be the major hepatic isoenzyme responsible for the formation of 3-hydroxyquinine. However, a recent study (Eagling et al., 1996) has revealed that ketoconazole, at a relative low concentration (mean IC₅₀ = 1-6 µM), also inhibits other CYP isoforms in rat liver microsomes. This suggests that a differential selectivity of CYP isoform inhibitors such as ketoconazole may exist in the liver microsomes of different species. Nevertheless, in the present study, the mean IC₅₀ values for ketoconazole were less than 0.1 µM among the four species tested (fig. 2), and the mean K_i value for ketoconazole in human liver microsomes was 0.015 µM (unpublished data), a result that supports the previous observations of the selective inhibition of the CYP3A isoform by ketoconazole at a low concentration (<1 µM) (Newton et al., 1995; Bourrie et al., 1996). The results derived from the two chemical inhibitors of CYP3A were also in good agreement with our immunoinhibition results; that is, anti-CYP3A antisera (25 µg of IgG per microgram of microsomal protein) inhibited 3-hydroxyquinine formation by about 96%, 84% and 92% in liver microsomes obtained from mouse, rat and dog, respectively (fig. 6). These findings further confirm the dominant role of the CYP3A/Cyp3a subfamily in the formation of 3-hydroxyquinine from quinine in the animal species tested.

Although ketoconazole and TAO strongly inhibited quinine 3-hydroxylation in all species, we observed a marked interspecies difference in the inhibition potency (e.g., dog vs. human in the case of TAO; fig. 3). This may occur because TAO can function as an "inducer" by stabilizing CYP3A4, and it can act as an inhibitor of CYP3A4 substrates with a low K_m. Thus TAO appears not to be an ideal choice to study as an inhibitor of CYP3A4, compared with another well-known inhibitor of CYP3A4, ketoconazole. This interspecies difference also existed when -naphthoflavone and diazepam, two so-called activators of CYP3A in human liver microsomes (Chang et al., 1994; Andersson et al., 1994; Pearce et al., 1996), were coincubated with quinine to conduct an activation study. As expected, both -naphthoflavone and diazepam tended to activate quinine 3-hydroxylation with human and dog liver microsomes, whereas they substantially inhibited it with mouse and rat liver microsomes (fig. 5). The reason for these discrepant findings is entirely obscure, but they may be explained in part by interspecies differences in CYP3A/Cyp3a isoform(s) that have been observed and characterized among humans (Nelson et al., 1996; Wrighton and Stevens, 1992), dogs (Nelson et al., 1996; Eguchi et al., 1996), rats (Nelson et al., 1996; Soucek and Gut 1992; Nedelcheva and Gut 1994; Funae and Imaoka, 1993) and mice (Nelson et al., 1996; Funae and Imaoka, 1993) with respect to structures, functions and properties. In addition, activation of CYP3A by -naphthoflavone is not always observed; some reactions mediated via CYP3A can be inhibited by this compound (Yun et al., 1992; Berthou et al., 1994). This has been interpreted to indicate an allosteric mechanism (Raney et al., 1992) as well as an interspecies difference in CYP3A isoform(s) (Nelson et al., 1996). In addition, diazepam is also a substrate not only of CYP3A4 but also of 2C19 with human liver microsomes (Andersson et al., 1994), which may inhibit the metabolism of CYP3A4 or 2C19-mediated substrates like quinine. Thus we must exercise caution when we interpret the effects of selective modifiers of P450-catalyzed reactions, particularly their effects on the interaction results obtained from different species, such as those observed in this study.

With mouse liver microsomes in which the formation of 3-hydroxyquinine from quinine showed a biphasic Michaelis-Menten kinetics (fig. 2B), a marked inhibition (>50%) of quinine 3-hydroxylation was observed when quinine was coincubated with p-nitrophenol [a substrate of Cyp2e1 (Forkert et al., 1994)], S-mephenytoin and sulfaphenazole [a substrate and inhibitor of the CYP2C subfamily, respectively (Goldstein and de Morais, 1994)] (fig. 4); this suggests the possible involvement of Cyp2e1 and Cyp2c in quinine 3-hydroxylation in these mouse liver microsomes. However, these findings seem to be inconsistent with our data obtained from the immunoinhibition study: anti-rat CYP2E1 and 2C11 antisera (25 µl) did not appreciably inhibit the formation of 3-hydroxyquinine with the same mouse liver microsomes (data not shown). The reason for this discrepancy remains unknown, but it is possible that the anti-rat CYP 2E1 and 2C11 antisera have a weak ability to cross-react with mouse Cyp2e1 and Cyp2c and/or that p-nitrophenol, S-mephenytoin and sulfaphenazole also inhibit mouse Cyp(s) other than Cyp2e1 and Cyp2c. The latter possibility may be particularly likely with the high concentrations of p-nitrophenol (4 mM and 10 mM) we used. We assume that at such high concentrations, p-nitrophenol might also inhibit Cyp3a and/or other Cyp(s) in mouse liver microsomes, because p-nitrophenol is not solely, though it is largely, metabolized by Cyp2e1 in some animal and human liver microsomes (Monostory and Vereczkey, 1994; Tassaneeyakul et al., 1993). This assumption appears to be supported by the fact that some compounds that are metabolized by one enzyme can also bind to another enzyme and inhibit it. For example, quinidine is a substrate of CYP3A4 (Guengerich et al., 1986) but inhibits CYP2D6 (Kobayashi et al., 1989).

In the quinine-antimalarials interaction study, chloroquine showed a moderate inhibition (about 45%) of quinine 3-hydroxylation in rat and dog liver microsomes, whereas no inhibition was seen with human liver microsomes (fig. 7 and table 3). In addition, the inhibition potency of doxycycline and tetracycline on quinine 3-hydroxylation differed somewhat between animals (i.e., rats and dogs) and humans (fig. 7). The reason for this observation is unclear, but it may be associated with an interspecies difference in substrate affinity (e.g., binding constants to the protein), enzyme activity and/or susceptibility to inhibition. In addition, the possibility cannot be excluded that quinine is more specifically metabolized by different isoforms belonging to the same CYP3A subfamily [e.g., CYP3A3, 3A4, 3A5 and 3A7 in humans, CYP3A12 in dogs, CYP3A1, 3A2 and 3A9 in rats and Cyp3a11, 3a13 and 3a16 in mice (Nelson et al., 1996)] among the species tested, although quinine is 3-hydroxylated mainly via CYP3A4 by human liver microsomes (Zhao et al., 1996; Zhang et al., 1997), and the CYP3A contents correlated significantly with the quinine 3-hydroxylation activities in the eight human liver microsomes assessed in the present study (fig. 8). Furthermore, the possibility of CYP(s)/Cyp(s) other than CYP3A also being involved in quinine 3-hydroxylation with animal liver microsomes cannot be totally ruled out. In this respect, further studies are definitely required.

In conclusion, our data have shown that 3-hydroxyquinine is a main metabolite of quinine and that CYP3A/Cyp3a is a principal isoform involved in this metabolic pathway in all species (rat, dog and human/mouse) tested. Overall, the dog and possibly the rat appear to be qualitatively and quantitatively suitable animal models for exploring the quinine 3-hydroxylase activity and for screening quinine-drug interactions in vitro, though some variance from the human liver microsomal data exists (e.g., an inhibition potency by TAO and discrepant interaction result with chloroquine). Finally, we are tempted to assert that an interspecies in vitro comparison study with animal and human liver microsomes will provide useful information on species differences and similarities in the metabolism of antimalarials such as quinine, thereby helping to identify an appropriate animal species for evaluating and/or forecasting the safety and toxicity of those antimalarial drugs. Liver microsomal samples obtained from such an animal species can be used to assess the possible involvement of candidate CYP(s) in the metabolism and then can be used to evaluate the drug-drug interactions that should be specifically investigated during the early clinical development program of antimalarial drugs. This seems to be particularly important during the early development of a new antimalarial drug in light of the fact that the detailed metabolic profile and involved CYP isoforms of quinine have only recently been clarified (Zhao et al., 1996; Zhang et al., 1997; Zhao and Ishizaki, 1997, in press) despite its 350-year history of clinical use (Tracy and Webster, 1996). Given that CYP isoform(s) involved in the metabolic pathways of many antimalarial drugs has (have) not been known (White, 1992), the search for an animal model, as undertaken by this study, is required for bridging the gap between studies of human and of animal liver microsomal P450.