(Received for publication, September 27, 1996, and in revised form, May 27, 1997)
From the Pharmazeutisches Institut, Christian-Albrechts-Universität zu Kiel, Gutenbergstr. 76, 24118 Kiel, Germany
Drugs containing strong basic nitrogen functional groups can be N-oxygenated to genotoxic products. While the reduction of such products is of considerable toxicological significance, most in vitro studies have focused on oxygen-sensitive reductase systems. However, an oxygen-insensitive microsomal hydroxylamine reductase consisting of NADH, cytochrome b5, its reductase, and a third unidentified protein component has been known for some time (Kadlubar, F. F., and Ziegler, D. M. (1974) Arch. Biochem. Biophys. 162, 83-92). This report describes the isolation and identification of all of the components required for the reconstitution of an oxygen-insensitive liver microsomal system capable of catalyzing the efficient reduction of primary N-hydroxylated structures such as amidines, guanidines, amidinohydrazones, and similar functional groups. In addition to cytochrome b5 and its reductase, the reconstituted system requires phosphatidylcholine and a P450 isoenzyme that has been purified to homogeneity from pig liver. The participation of cytochrome b5 and NADH cytochrome b5 reductase in cytochrome P450-dependent biotransformations has previously only been described for oxidative processes. The data presented suggest that this system may be an important catalyst in the reduction of genotoxic N-hydroxylated nitrogen components in liver. Their facile reduction by cellular NADH may be the reason why N-hydroxylated products can be missed by studies in vivo. Furthermore, the enzyme system is involved in the reduction of amidoximes and similar functional groups, which can be used as prodrug functionalities for amidines and related groups.
The metabolism of nitrogen-containing functional groups has become a topic of considerable interest since the early discovery that N-hydroxylated intermediates are often responsible for the toxic and/or carcinogenic properties of aromatic amines, hydrazines, and amides (1). On the other hand, the more facile N-oxygenation of secondary and tertiary alkylamines to hydroxylamines and N-oxides was considered as a route for detoxication (2, 3), and it was generally assumed that the strongly basic nitrogen compounds were metabolically stable. However, we have demonstrated that even the protonated hydrophilic amidines (4-6), as well as diamidines such as pentamidine and diminazene and also guanidines and amidinohydrazones, are capable of undergoing metabolic N-oxygenation by liver microsomal cytochrome P450 monooxygenases (7, 8). The N-oxygenation of these functional groups produces more reactive metabolites, and the genotoxic properties of benzamidoxime are well known (9). During investigations of the metabolic fate of strongly basic N-hydroxylated xenobiotics, we observed that they were readily reduced both in vivo and in vitro by a microsomal system present in all mammalian species (rats, rabbits, pigs, and humans) tested to date (10-13).
Preliminary experiments indicated that this system (10) had many of the
characteristics of the microsomal O2-insensitive hydroxylamine reductase described by Kadlubar et al. (14,
15), which required NADH-cytochrome b5
reductase, cytochrome b5, and a third
unidentified protein component. In this report, we describe the
isolation, purification, and characterization of this component from
pig liver microsomes and show that the system catalyzing the
O2-insensitive reduction of benzamidoxime requires
NADH-cytochrome b5 reductase, cytochrome
b5, a cytochrome P450, and phospholipid (Fig.
1).
Purification
Pig Liver MicrosomesPig liver microsomes were prepared by fractional acid precipitation according to the procedure of Ziegler and Pettit (16) with slight modifications (7).
Cytochrome b5, NADH-Cytochrome b5 Reductase and NADPH-Cytochrome P450 ReductaseCytochrome
b5, NADH-cytochrome b5
reductase, and NADPH-cytochrome P450 reductase were separated by
modifications of the procedure described by Kling et al.
(17). Thesit (Boehringer Mannheim) was used in place of Emulgen 913 to
solubilize the microsomal proteins and in the elution buffers. All
purification steps were performed at 4 °C. The solubilized pig liver
microsomes were applied to an octyl-Sepharose CL 4B (Pharmacia,
Freiburg, FRG) column (inner diameter, 3.8 cm; length, 38 cm), and
enzymes were eluted by a stepwise increase in the detergent
concentration and a decrease in the salt concentration in elution
buffers (buffer A: 10 mM potassium phosphate, pH 7.4, 1 mM EDTA (Serva, Heidelberg, FRG), 1 mM
dithiothreitol, 20% (w/v) glycerol, 0.5% (w/v) sodium cholate, and
0.5 M NaCl; buffer B: 10 mM potassium
phosphate, pH 7.4, 1 mM EDTA, 1 mM
dithiothreitol, 20% (w/v) glycerol, 0.44% (w/v) sodium cholate, 0.2%
(w/v) Thesit, and 0.5 M NaCl; buffer C: 10 mM
potassium phosphate, pH 7.4, 1 mM EDTA, 1 mM
dithiothreitol, 20% (w/v) glycerol, 0.2% (w/v) sodium cholate, and
2% (w/v) Thesit). The flow rate was 50 ml/h; the volume of the
fractions collected was 8.5 ml; and the eluate was monitored at 280 nm
(Fig. 2). The fractions containing
NADPH-cytochrome P450 reductase activity (Fig. 2, peak 2)
and those containing NADH-ferricyanide reductase activity (Fig. 2,
peak 4) were collected. In addition, the fractions with the
highest absorbance at 417 nm (cytochrome b5)
were combined (Fig. 2, peak 3).
NADH-Cytochrome b5 Reductase
NADH-cytochrome
b5 reductase (Fig. 2, peak 4) was
purified to homogeneity by affinity chromatography on 5-AMP-Sepharose
4B (Pharmacia) similar to the procedure described for the purification of NADPH-P450 reductase (18). The fractions containing the highest NADH-ferricyanide reductase activity were combined and concentrated, followed by gel filtration (NAP 10, Pharmacia). The specific activity of the purified reductase was 27 units/mg, and only one band was detectable on SDS-PAGE1 (data
not shown).
Cytochrome b5, present in the highest concentration in peak 3 (Fig. 2), was purified on a DEAE-cellulose column (inner diameter, 2.5 cm; length, 30 cm) equilibrated with buffer D (10 mM potassium phosphate, pH 7.4, 0.1 mM EDTA, 0.1 mM dithiothreitol, 20% (w/v) glycerol, 0.5% (w/v) sodium cholate). After equilibration, elution was performed with a linear gradient from 0 to 1 M KCl in buffer D, collecting 10-ml fractions, and those with the highest absorbance at 417 nm were combined (data not shown) and further processed by gel filtration on Sephadex G-100 (inner diameter, 1.5 cm; length, 45 cm), equilibrated in buffer D, at a flow rate of 15 ml/h. The fractions containing cytochrome b5 were concentrated by ultrafiltration (Amicon, Witten, FRG; exclusion size, 10 kDa), and the buffer was changed to buffer D, lacking sodium cholate, by gel filtration on NAP 25 columns (Pharmacia). The final cytochrome b5 fraction contained 10.2 nmol of cytochrome b5/mg of protein and exhibited one band at 16 kDa after SDS-PAGE (data not shown).
NADPH-Cytochrome P450 ReductaseThe NADPH-cytochrome P450
reductase (Fig. 2, peak 2) was further purified by affinity
chromatography on 2,5
-ADP-Sepharose 4B (Pharmacia) and subsequently
by gel filtration chromatography on acrylamide-agarose (AcA 44 Ultrogel; Pharmacia), as described by Yasukochi and Masters (18). The
preparation also appeared homogenous on SDS-PAGE (data not shown).
Preliminary measurements indicated that in addition to cytochrome
b5 and its reductase, further components present
in fractions 88-101 (Fig. 2, peak 4) were also required for
reconstitution of amidoxime reductase. The combined fractions 88-101
were concentrated by ultrafiltration (exclusion size, 30 kDa; Amicon),
desalted by gel filtration (NAP10, Pharmacia), and chromatographed on
an anion exchange column by preparative HPLC with a conventional HPLC
system (L-6210 Intelligent Pump, 655 A-22 UV detector, D-2500 integrator; Merck/Hitachi, Darmstadt, FRG). 1 mg of protein was applied
to a semipreparative Fractogel EMD TMAE 650(S) column (inner diameter,
10 mm; length, 150 mm; particle size, 25-40 µm; Merck), which had
been equilibrated with buffer E (10 mM Tris acetate, pH
7.4, 20% (w/v) glycerol, 0.5% (w/v) Thesit, 0.1 mM EDTA,
0.1 mM dithiothreitol). After equilibration, fractions were eluted with a one-step increase in the salt concentration up to 500 mM in buffer E at a flow rate of 1 ml/min. The fractions
with the highest absorbance at 280 nm (Fig.
3) were pooled and concentrated by
pressure dialysis, tested for NADH cytochrome b5
reductase activity, and tested for benzamidoxime reductase activity in
the reconstituted system. TM1 showed only one band on SDS-PAGE (Fig. 4).
Fraction TM1
Fraction TM1 (Fig. 3, peak 1), which contained the third component required for benzamidoxime reductase activity in the reconstituted system and the UDP-glucuronyltransferase (both Mr 50,000), was applied to a UDP-hexanolamine-Sepharose (Sigma) column (inner diameter, 1.2 cm; length, 4.5 cm) equilibrated with 10 mM Tris acetate buffer, pH 7.4, and washed with 50 ml of buffer F, consisting of 50 mM KCl and 40 µM phosphatidylcholine in 10 mM Tris acetate buffer, pH 7.4. Activity was recovered quantitatively in the effluent, which showed only one band (50 kDa) on SDS-PAGE (like TM1 in Fig. 4). This fraction is named the third protein or the third component.
Detergents were removed from the purified enzymes by shaking the concentrated fractions for 4 h with Calbiosorb (Calbiochem, La Jolla, CA) at 4 °C.
Analytical Procedures
MaterialsBenzamidoxime was synthesized by a published method (19). Benzamidine was obtained from Aldrich-Chemie (Steinheim, FRG), dextromethorphan was from Sigma, and dextrophan-D-tartrate was from ICN Biochemicals Inc. NADPH and NADH were purchased from Merck.
Spectrophotometric measurements were performed with a Kontron Uvicon 930 spectrophotometer.
For HPLC analysis, a conventional system was used: L-5000 controller, 655 A-11 HPLC gradient pump, L-422 UV detector, AS-2000 autosampler, and D-2500 integrator (Merck/Hitachi).
Amino acid sequence analyses on peptide fragments of the purified proteins were carried out by TopLab (Munich, FRG).
Protein ConcentrationsAll protein concentrations were measured using the method described by Smith et al. (20) with bicinchoninic acid (BCA reagent kit; Pierce).
SDS-PAGEThe SDS-PAGE analyses were carried out by the
method of Laemmli (21), with a 3% stacking gel. For cytochrome
b5 preparations a 12% separation gel was used,
while all other preparations were separated on an 8% gel (1.5-mm
thickness). Staining was performed with Coomassie Brilliant Blue R250
(Serva, Heidelberg, FRG). Standards and samples were pretreated with
-mercaptoethanol for 5 min at 90 °C. The following proteins were
used as standards (high molecular weight calibration kit; Pharmacia):
myosin (Mr 205,000),
-galactosidase (Mr 116,000), phosphorylase b
(Mr 94,000), albumin (Mr
67,000), ovalbumin (Mr 45,000), and carbonic
anhydrase (Mr 29,000).
Cytochrome
b5 concentration was determined by recording the
reduced minus the oxidized spectrum (absorbance 185 mM1 cm
1) as described by
Estabrook and Werringloer (22).
Cytochrome P450
concentrations were determined by measuring the carbon monoxide
difference spectra after reduction with dithionite as described by
Omura und Sato (23). In the case of TM1 protein, cytochrome P450 was
also quantified by CO difference spectra after reduction (50 µg of
protein) with 3 units of NADPH-P450 reductase, 1.5 mM NADPH
and 40 µM L--dilauroyl phosphatidylcholine
(DLPC) (Sigma) in a total volume of 1 ml of 100 mM
potassium phosphate, pH 7.4. After preincubation for 5 min at 25 °C,
the cuvettes were saturated with CO, and the difference spectra were
recorded.
NADH-cytochrome b5 was measured by its NADH-ferricyanide reductase activity (1 unit = 1 µmol of reduced ferricyanide/min) as described by Mihara and Sato (24).
NADPH-Cytochrome P450 Reductase ActivityNADPH-cytochrome P450 reductase activity was measured by following the reduction of cytochrome c (type 4 horse heart; Sigma) as described (25) with minor modifications (18) (1 milliunit = 1 nmol of reduced cytochrome c/min).
Assay for the Reduction of BenzamidoximeThe incubation
mixture for the reconstituted enzyme system consisted of 50 pmol of
cytochrome b5, 0.5 unit of NADH cytochrome b5 reductase, 5 µg of fraction TM1 (Fig. 3,
peak 1) or 5 µg of third protein, 7.5 µg of DLPC (final
concentration, 40 µM; Sigma), and 150 nmol of
benzamidoxime (final concentration, 0.5 mM) in a 100 mM potassium phosphate buffer, pH 6.3. After a 5-min
preincubation period at 37 °C, the reaction was initiated by the
addition of NADH (final concentration, 1 mM) to a total
volume of 300 µl. After 20 min, the reaction was terminated by adding
300 µl of methanol on ice. Precipitated proteins were sedimented by
centrifugation at 10,000 × g for 5 min, and the
supernatant was analyzed by HPLC (26). Aliquots of 10 µl were
injected into a LiChrospher RP-Select B column (125 × 5 mm, 5 µm; Merck) with an RP-Select B precolumn (4 × 4 mm; Merck). The
separation was carried out at room temperature with 3 mM
1-octanesulfonic acid, pH 2.5 (adjusted with 85% phosphoric acid)/acetonitrile (88:12, v/v) as the mobile phase, at a flow rate of
0.7 ml/min. The effluent was monitored at 229 nm. The retention time
for benzamidine was 16.5 ± 0.4 min and 14.3 ± 0.2 min for
benzamidoxime when injected as controls (Fig.
5). The rate of product formation was
linear for about 60 min, and the limit of detection of benzamidine was
10 pmol/injection. Standard curves with 1.0, 2.5, 5.0, 7.5, 10.0, 15.0, 20.0, 25.0, and 50.0 µM benzamidine were constructed and
found to be linear over this range with correlation coefficients
>0.9993. The recovery of benzamidine from incubation mixtures was
98.5 ± 1.3% (n = 36) of that obtained using
samples that contained the same amount of benzamidine dissolved in
phosphate buffer.
Microsomal Reductase Activities
The reductase activities of pig and human liver microsomes were measured by following the reduction of benzamidoxime, as described above for the reconstituted system, in 300 µl of 100 mM phosphate, pH 6.3, containing 500 µM benzamidoxime, 1 mM NADH, and 0.05-0.1 mg of human liver microsomes or 0.2 mg of pig liver microsomes. Ten different samples of human liver microsomes obtained from Human Biologics, Inc. (Phoenix, AZ) were tested. Various enzymatic activities related to specific human P450 subfamilies had been determined by Human Biologics, Inc.
O-Demethylation of DextromethorphanThe O-demethylation of dextromethorphan was carried out essentially as described (27) in 100 mM phosphate, pH 7.4, containing 0.3 units of NADPH-cytochrome-P450 reductase, 5 µg of third protein component or 5 µg of fraction TM1 (Fig. 3), 40 µM DLPC, and 3.3 mM MgCl2 in a final volume of 300 µl.
NADH ConsumptionThe reconstituted assay for the reduction of benzamidoxime was carried out with 0.1 mM NADH (final concentration), and the oxidation of NADH during the incubation was monitored by absorbance at 340 nm. Further procedures were performed as described above.
The data summarized in Table I demonstrate that reduction of benzamidoxime by NADH requires three proteins purified to apparent homogeneity from pig liver microsomes. The fastest rates were consistently observed with mixtures containing 0.5 units of NADH-cytochrome b5 reductase, 50 pmol of cytochrome b5, and 5 µg of the third protein isolated from peak 4 (Fig. 2) along with DLPC in phosphate buffer. DLPC was the least essential ingredient, and omitting this phospholipid decreased activity by only 30%. On the other hand, omitting either cytochrome b5 or its reductase virtually abolished activity, and omission of the third protein component decreased activity by almost 90% compared with rates obtained with the complete system (Table I). Replacing NADH with NADPH at the same initial concentration (1 mM) reduced activity by almost 60%. The consumption of NADH was linear during the incubation with the complete system. The molar ratio of benzamidine production and depletion of NADH was 1. In the absence of the third protein component, the decrease of absorbance was the same as the blank (without substrate).
|
Benzamidoxime reductase activity of the reconstituted system was insensitive to catalase and cyanide. A complete inhibition of the benzamidoxime reductase activity was observed upon the addition of 0.5 mM N-methylhydroxylamine (data not shown). Superoxide dismutase (500 units/ml) increased the rate of reduction with some preparations as much as 50%.
We also investigated the addition of NADPH-cytochrome P450 reductase. The presence of NADPH-cytochrome P450 reductase in the reconstituted system decreased the benzamidoxime reductase activity by 35% (Table II). In addition to this, added NADPH (1 mM) resulted in a further 25% inhibition compared with the complete system. The inhibitory effect of NADPH-P450 reductase was reversed by the addition of superoxide dismutase. The replacement of NADH-cytochrome b5 reductase by NADPH-P450 reductase in presence of superoxide dismutase diminished activity by about 80%; in the absence of superoxide dismutase, however, there was no detectable activity (Table II).
|
The addition of dextromethorphan (final concentration, 0.5 mM), NADPH-P450 reductase (0.5 unit), and NADPH (final concentration, 1 mM) to the reconstituted benzamidoxime reductase system diminished the appearance of the benzamidine product by about 30% in presence of superoxide dismutase (500 units/ml) (data not shown).
In addition to benzamidoxime, the reconstituted system also catalyzed the reduction of guanoxabenz and N-hydroxydebrisoquine (data not shown). Although the reduction of these N-hydroxylated substrates by NADH required the same components, the optimal pH for reduction of N-hydroxydebrisoquine (near pH 7) was higher than that for the reduction of guanoxabenz (pH 6) or benzamidoxime (pH 5.5).
Composition of Third Protein ComponentSpectra of this
protein purified to apparent homogeneity from peak TM1 (Fig. 3) were
similar to those of other cytochrome P450 isoenzymes and, like the
latter, gave typical reduced minus oxidized difference spectra in the
presence of CO (data not shown). The amino acid sequences of three
peptides isolated from a proteolytic digest of the purified protein are
shown in Table III. The sequence of
peptide 2 is identical to that of a sequence in P450 2D described by
Tsuneoka et al. (28), and the sequences of peptides 1 and 3 are similar to sequences of P450 2D (peptide 1, 66.7%; peptide 3, 83.3%) and P450 2D6 (peptide 1, 73.3%; peptide 3, 75%) (28, 29). The
sequence of peptide 3 also shows similarity to P450 isoenzymes of the
subfamilies CYP2A (57.8%), CYP1A (
57.8%) and CYP2J (
56.3%),
while the sequence of peptide 2 shows a slight similarity to the
subfamily CYP3A.
|
The data summarized in Table IV demonstrate that the third protein component can function as a monooxygenase in the system reconstituted with the NADPH-cytochrome P450 reductase. The results show that the protein purified to homogeneity from fraction TM1 also has the catalytic activity expected of this class of P450-dependent monooxygenases.
|
Activity measurements with human liver microsomes show that these preparations also catalyze oxygen-insensitive reduction of benzamidoxime by NADH. The activity of 10 samples of human microsomes varied between 0.2 and 0.7 nmol of benzamidine/min/mg of protein. The stoichiometric ratio of cytochrome b5 to NADH-cytochrome b5 reductase to third protein component in the reconstituted system from pig liver was 10:2:1. However, there was still an increase in amidoxime reductase activity by the addition of cytochrome b5 to this reconstituted system, while the stoichiometric ratio of NADH-cytochrome b5 reductase to third protein component was optimal. Since the relative composition of the three protein components of benzamidoxime reductase varies in all human samples and each of these proteins can be rate-limiting (data not shown), no clear correlation between benzamidoxime reductase activity and the content of the various P450 isoenzymes was observed.
Microsomal preparations from rabbit and rat liver (10), as well as those from pig and human livers catalyze the NADH-dependent reduction of benzamidoxime to benzamidine. The reduction of other N-oxygenated basic nitrogen-containing functional groups by these liver preparations has also been described for N-hydroxylated guanidine, N-hydroxydebrisoquine (12), and amidinohydrazone guanoxabenz (30). Previous studies (4, 7) have also shown that liver microsomal preparations can also catalyze NADPH-dependent N-hydroxylation of these compounds, but the rates of N-hydroxylation were about 2 orders of magnitude lower than rates of reduction, which suggests that the latter reaction may prevent accumulation of toxic N-hydroxylated products. The in vivo relevance of such processes has been confirmed previously with studies on live animals (13).
Preliminary studies on the biochemical properties of the benzamidoxime reductase suggested that it shared characteristics reported by Kadlubar and Ziegler (15) for pig liver microsomal hydroxylamine reductase. Both were completely oxygen-insensitive, preferentially use NADH as reductant, and are most active in slightly acidic media. The report by Kadlubar and Ziegler (15) also indicated that this hydroxylamine reductase was a multicomponent system consisting of cytochrome b5, cytochrome b5 reductase, and a third rather unstable uncharacterized component.
The present report clearly demonstrates that the oxygen-insensitive benzamidoxime reductase purified and reconstituted from pig liver microsomes requires cytochrome b5, its reductase, and a third electrophoretically pure protein (Fig. 4 and Table I). This protein also has characteristics of a P450 enzyme. A comparison with all known cytochrome P450 isoenzymes shows that its amino acid sequence is very close to isoenzymes of the subfamily 2D from other species. Similarities with other isoenzymes are less striking. The assumption that the isolated enzyme is a P450 isoenzyme was confirmed by the CO difference spectra. The identity of the homogenous third component as a cytochrome P450 was also confirmed by its ability to catalyze the demethylation of dextromethorphan upon reconstitution with NADPH-P450 reductase and phospholipid (Table IV). This reaction is a characteristic marker activity of the P450 2D subfamily (31). The addition of dextromethorphan and the components, which demethylate dextromethorphan, to the benzamidoxime reduction system diminished the benzamidoxime reductase activity. This indicates that benzamidoxime and dextromethorphan are both substrates for the purified enzyme. Furthermore, the mass of the purified third component, 50 kDa, is consistent with the size of other P450 isoforms isolated from other species. Cytochrome P450 2D isoenzymes from pig have not been described previously. Attempts to reconstitute the benzamidoxime reductase from human liver and to identify the responsible human P450 isoenzyme are in progress. Human hepatic microsomes catalyzed the reduction of benzamidoxime, and the conversion rates observed were of an order of magnitude comparable with those determined for microsomes from porcine liver. We can therefore expect that the results of these investigations will be comparable with those from experiments with isolated human enzymes.
The participation of cytochrome P450 and NADPH-cytochrome P450 reductase in reductive metabolic processes has mainly been demonstrated by in vitro experiments under anaerobic conditions (1). Cretton et al. (32) pointed out that P450 and P450 reductase participate in the microsomal reduction of azides, which they observed under normal O2 partial pressure. However, the reduction was enhanced under anaerobic conditions. The cytosolic aldehyde oxidase-catalyzed benzamidoxime reduction (33) has only been observed under anaerobic conditions in vitro. However, the relevance of anaerobic in vitro reactions to well oxygenated liver parenchymal cells in vivo is questionable. The reaction described in the present work proceeds not only in vitro but also in vivo in the presence of oxygen.
The participation of cytochrome b5 and NADH cytochrome b5 reductase in cytochrome P450-dependent biotransformations has previously only been described for oxidative processes. It has already been demonstrated that using the appropriate substrate cytochrome b5 may efficiently donate both electrons to cytochrome P450 (34). It is, therefore, possible that N-oxygenated derivatives such as benzamidoxime can compete with oxygen for the sixth binding site of the heme cation of certain P450 isoforms. Thus, the electrons are donated to N-hydroxylated structure and not to oxygen. These mechanistic speculations have to be clarified in further studies.
The addition of N-methylhydroxylamine to the reconstituted system inhibited the reductase activity, as was also described for the microsomal reaction (12). It has been reported by Cribb et al. (35) that sulfamethoxazole hydroxylamine is reduced aerobically by human liver microsomes. The authors employed the same microsomal liver preparations (10 different samples of human liver microsomes provided by Human Biologics, Inc.). The reaction rates obtained in this study on the reduction of benzamidoxime are very similar to those described for the reduction of sulfamethoxazole hydroxylamine (correlation coefficient, 0.9835), strongly suggesting that the same enzyme system that reduces amidoximes is also capable of reducing hydroxylamines.
It has been shown that the presence of the three protein components, cytochrome b5, its reductase, and cytochrome P450, as well as DLPC and NADH is required for the N-reduction of benzamidoxime (Table I). The observation that the absence of DLPC in reconstituted systems does not lead to a pronounced lowering of the conversion rates can be explained by the possibility that traces of detergent are still present in the enzyme fractions. The role of phospholipids in reconstituted P450 enzyme systems can be partially fulfilled by a low concentration of nonionic surfactants (36).
The presence of superoxide anions evidently gives rise to inhibiting reactions, as demonstrated by the stimulation of benzamidoxime reduction by superoxide dismutase (Table I). This observation also demonstrates that a nonenzymatic, iron-catalyzed reduction of benzamidoxime by superoxide does not contribute to product formation.
The electron transport of the benzamidoxime reduction could be partially effected by the NADPH-cytochrome P450 reductase, although only in the presence of superoxide dismutase. Furthermore, the conversion rate is only 20% of that of the complete system. The reduction in activity upon the addition of this flavoprotein to the complete reconstituted system, which is further accentuated by the addition of NADPH, can be explained by the formation of superoxide (Table II).
The reconstitution of the three protein components is essential for the electron transfer, as was also shown by measuring the consumption of NADH.
The evidence presented suggests that the reductase activity of the reconstituted system is solely enzymatic. The proteins cytochrome b5, NADH-cytochrome b5 reductase, and cytochrome P450, isolated to electrophoretic purity, in a reconstituted system under aerobic conditions are highly efficient as a catalyst for the reduction of N-hydroxylated compounds, such as N-oxygenated amidines, guanidines, and amidinohydrazones. We expect that this reduction also occurs with weakly basic N-hydroxylated substances that often exhibit a genotoxic potential, as is the case for the above mentioned compounds with strongly basic functional groups. These biotransformations can be considered as a physiological detoxification reaction.
The facile reduction of amidoximes and similar functional groups in vivo by the microsomal enzyme system described here and possibly by other reductive systems is responsible for the pharmacological activity of N-hydroxylated compounds such as the amidoximes of pentamidine (37). Thus an amidoxime group can serve as a prodrug functionality for an amidino group. The same concept can be applied to guanidines and amidinohydrazones and their N-oxygenated derivatives (12, 30). Therefore, strong basic nitrogen-containing functional groups, which are protonated under physiological conditions and are not absorbed as cations, can be made orally available by introducing a hydroxy group, which lowers the pKa values significantly. The amidoximes and similar functionalities are absorbed as the free bases and reduced in the liver to the active principles by the enzyme system described here. Meanwhile, this prodrug approach has not only been applied to pentamidine but also to orally active fibrinogen receptor antagonists (38).
The studies summarized in this report completely characterize for the first time all of the components of the microsomal enzyme system catalyzing the reduction of a wide variety of N-oxygenated metabolites. The observation that human liver microsomal benzamidoxime and sulfamethoxazole hydroxylamine reductase (35) are similar to the purified pig liver system suggests that this multicomponent reductase may have a vital role protecting humans and other mammals against accumulation of potentially genotoxic N-hydroxylated metabolites.
We are indebted to Dr. Friedrich Lottspeich, in whose laboratory the amino acid sequence analyses were done, and express our thanks to Dr. Thomas Kunze for valuable discussions.