chemlstry vol. 255. no. issue of june 25. pp. 557p-5.585 ... · self-catalyzed inactivation of...

T m ~ O U R N A I . OF HlOLOClCAL CHEMlSTRY

Prinfed l n 1’S.A. Vol. 255. No. 12, Issue of June 25. pp. 557P-5.585. 19Ro

Self-catalyzed Inactivation of Hepatic Cytochrome P-450 by Ethynyl Substrates*

(Received for publication, November 26, 1979, and in revised form, February 20, 1980)

Paul R. Ortiz de Montellano$ and Kent L. Kunze From the Department of Pharmaceutical Chemistry, School of Pharmacy, and Liver Center. University of California, Sun Francisco, California 94143

The following acetylenic substrates have been shown to mediate NADPH-dependent loss of cytochrome P- 450 on incubation with hepatic microsomes from phenobarbital-pretreated rats: 1-ethynylcyclohexanol, 1- ethynylcyclopentanol, 3-methyl-1-pentyn-3-01, norethisterone, (1-methoxycyclohexyl)acetylene, 3-(2,4-dichlorophenoxy)-1-propyne, 3-(4-nitrophenoxy)-l-propyne, 3-phenoxy-l-propyne, 4-phenyl-l-butyne, 3- phenyl-1-propyne, cyclohexylacetylene, acetylene, 3- pentyn-2-01, 4-methyl-2-octyn-4-01, 2-hexyne, p+nyla- cetylene, and N-( 1,1-dimethylpropynyl)-3,5-d1chloro- benzamide. A 10-fold higher nominal concentration of the last two agents was required to obtain the same degree of enzyme loss observed with the other agents at a 1 mM concentration. In vivo administration of acetylene gas and nine of the monosubstituted acetylenes led to accumulation of abnormal hepatic pigments. Similar pigments were not observed in rats treated with disubstituted acetylenes. The pigments obtained with acetylene gas, norethisterone, and six other substrates, after isolation, methylation, and purification, exhibited essentially identical electronic absorption spectra. Field desorption mass spectrometric analysis of these eight pigments has established that each one gives a molecular ion which corresponds to the sum of the molecular weights of the dimethyl ester of protoporphyrin IX plus the substrate plus (probably) an oxygen atom. These results are used to formulate a destructive mechanism in which the enzyme prosthetic heme covalently binds to the substrate during attempted metabolism of the triple bond.

Cytochrome P-450’ monooxygenases, a diverse group of enzymes related by similarities in their catalytic function and by the presence of a heme prosthetic group, dominate the hepatic metabolism of xenobiotic lipophilic substances (1-3). Each of these enzymes catalyzes reductive cleavage of molecular oxygen with concomitant transfer of 1 of the oxygen

* This work has been supported by National Institutes of Health Grants GM-25515 and P-50 AM-18520. The Berkeley Biomedical and Environmental Mass Spectrometry Resource, where mass spectra were obtained, is supported by National Institutes of Health Grant RR 00719. The Computer Graphics Laboratory, used in preparation of Scheme 2, is supported by National Institutes of Health Grant RR 1081-021. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ A Research Fellow of the Alfred P. Sloan Foundation. Cytochrome P-450 generally refers in this paper to the isozymes

induced by phenobarbital, but occasionally is used in a generic sense t.o denote the class of hemoproteins whose reduced carbon monoxide spectra exhibit a maxima at around 450 nm.

atoms to the substrate (1-3), although the observable conse- quences of oxygen transfer depend on the precise nature of the substrate. In general, the established immediate outcome of cytochrome P-450-catalyzed oxygen transfer is ( a ) formation of a hydroxyl derivative by insertion into the bond between a hydrogen and a heavier atom; ( 6 ) formation of an epoxide by addition across a carbon-carbon double bond; or ( c ) formation of a dipolar oxide by combination with the free electron pair of a heteroatom (4, 5). These primary reactions, however, often yield unstable structures which undergo secondary chemical transformations to give final experimentally observed products (4,5). The rearrangement of enzymatically formed aryl epoxides to phenolic derivatives is an example of such a secondary process (6).

Although cytochrome P-450 enzymes involved in physiological biosynthetic pathways usually exhibit high substrate specificity, those forms of the enzyme significantly engaged in xenobiotic metabolism are characterized by broad and over- lapping substrate selectivities (7-10). The effectiveness of this metabolic apparatus, which copes with the wide range of substances to which an individual is environmentally exposed, is optimized through differential induction of cytochrome P- 450 isozymes by lipophilic substrates (10, 11). However, the broad specificity and differential inducibility of these mono- oxygenase enzymes render alterations in metabolism of one agent by another a relatively common phenomenon.

Inhibition of the metabolism of one substance by another in order to enhance pharmacological activity has been com- mercially exploited with the advent of insecticide synergists (12, 13). An impressive fraction of the substances shown to potentiate the action of insecticides is characterized by the presence of a monosubstituted acetylenic function. Among these active agents are aryl propargyl ethers (14-16), N-alkynyl phthalimides (17), alkynyl phosphate esters (12, 181, and alkynyl oxime ethers (12, 19). The mechanism by which these acetylenic substances alter insecticide metabolism and toxicity, however, has remained undefined except for the observation that in vivo administration of 3-(2,4,5-trichloro- phenoxy)-1-propyne and 2-methylpropyl 2-propynyl phenyl- phosphonate to mice causes a measurable decrease in hepatic cytochrome P-450 content (18).

Research in a different biological arena has also provided evidence for altered drug metabolism due, in part, to the interaction of acetylenic groups with cytochrome P-450 enzymes. Synthetic 17-ethynyl sterols, hormonally active con- stituents of most birth control pills (201, have been found to inhibit oxidative drug metabolism (21-25). White and Muller- Eberhard (26) have reported that, at relatively high doses, ethynyl sterols specifically reduce hepatic cytochrome P-450 and heme levels in rats. Norethisterone and ethynylestradiol, the two ethynylsterols examined, selectively affected the phenobarbital-inducible isozymes (27) of cytochrome P-450. The

5578

Inactivation of Cytochrome P-450 by Acetylenes 5579

additional and highly significant observation was made that cytochrome P-450 loss was accompanied by accumulation of a hepatic “green pigment” not unlike that found by other investigators in rats treated with 2-isopropyl-4-pentenamide

Work in this laboratory has established that 2-isopropyl-4- pentenamide inactivates cytochrome P-450 by covalently binding to the heme prosthetic group during catalytic interaction with the enzyme (31-33). In a recent brief communi- cation, we have also reported that radiolabeled norethisterone is covalently incorporated into the “green pigment” produced by treatment of rats with this sterol (34). Covalent attachment of the sterol to prosthetic heme, suggested by the radioisotopic data, was c o n f m e d b y the mass spectrometric molecular ion of the purified pigment, although the observed molecular ion could not be attributed to a specific molecular species (34). We now report an in-depth study of the interaction of acetylenes with cytochrome P-450 which demonstrates the gener- ality of the process, unequivocally establishes the formation of substrate-heme adducts, and provides the basis for formu- lation of a tentative reaction mechanism. Aspects of this study have been presented at several meetings (35-37).

(28-30).

EXPERIMENTAL PROCEDURES

Substrates and Reagents-The following substances, of the high- est grade available, were obtained from the indicated commercial source and were used without further purification: l-ethynylcyclohexanol (A), 1-ethynylcyclopentanol (B), 3-methyl-1-pentyn-3-01 (C), 3-pentyn-2-01 (0). 4-methyl-2-octyn-4-01 (P), 4-phenyl-1-butyne (1)’ and 2-hexyne (Q) (Farchan Division, Story Chemical Corp.); cyclohexylacetylene (L) (Pfaltz-Bauer); phenylacetylene (K) (Aldrich Chemical Co.); and acetylene (M) (Matheson Chemical Co.). Litera- ture procedures were used for the synthesis of (l-methoxycyclo- hexy1)acetylene (E) (38), 3-phenyl-1-propyne (J) (39), 3-phenoxy-l- propyne (H) (14), 3-(2,4-dichlorophenoxy)-l-propyne (F) (14), and 3- (4-nitrophenoxy)-l-propyne (G) (14). Norethisterone (17-hydroxy-19- nor-l7-pregn-4-en-20-yn-3-one, D) was generously provided by Syn- tex Research, Palo Alto, Calif., and N-(l,l-dimethylpropyny1)-3,5- dichlorobenzamide (N, code No. RH-315) by Rohm and Haas Co., Philadelphia, Pa. Solvents used for chromatographic purification of hepatic pigments were analytical grade and were distilled prior to use.

In Vitro Assay of Cytochrome P-450 Inactiuation-The general procedure used, based on that of the Hoffmann-La Roche group (40), has been described (33). Incubation mixtures contained, in addition to substrates, the following: microsomal protein (1 mg/ml), NADPH (1 mM), KC1 (150 mM), and EDTA (1.5 m), all in 0.1 M phosphate buffer (pH 7.4). Acetylenic substrates, added without solvent at a nominal 1 mM concentration, were preincubated with the microsomal suspension for 10 min to allow substrate equilibration before NADPH was added to initiate the reaction. Nominal 10 mM concentrations of phenylacetylene and N-(l,l-dimethylpropynyl)-3,5-dichlorobenzamide were used since these agents caused only a marginal loss of cytochrome P-450 at a 1 mM concentration. In all cases, control incubations were carried out in the absence of added substrates and, for each substrate, in the absence of NADPH. The loss of cytochrome P-450 in the absence of substrates was negligible, demonstrating that lipid peroxidation was effectively suppressed by the added EDTA.

Isolation of Abnormal Hepatic Pigments-Sprague-Dawley male rats weighing 200 to 250 g were injected intraperitoneally once a day ’ for 4 days with an aqueous solution (80 mg/ml) of sodium phenobarbital (80 mg/kg dose). Acetylenic substrates were administered on the 5th day. Norethisterone (25 mg/kg) was injected intraperitoneally in 0.25 ml of trioctanoin after sonication of the mixture to disperse the sterol. AII other solid agents (200 mg/kg) were injected intraperitoneally as a solution in 150 pl of ethanol. Liquid agents ( 5 0 pl/rat) were injected intraperitoneally without dilution in a solvent. Acetylene gas was administered by placing phenobarbital-pretreated rats for 4 h in a chamber through which a stream of air containing 10 to 158 (v/v) acetylene was passed at an approximate rate of 200 to 300 ml/min. Four hours after administration of acetylenic agents, rats were decap- itated and their livers were perfused in situ with ice-cold 0.9% saline solution (100 ml/liver). The pooled livers of rats treated with a common agent were homogenized in 5% (v/v) HrS04/methanol (4 ml/g of liver) and the homogenate was allowed to stand overnight at

4°C in the dark. All manipulations were carried out in the dark since the hepatic pigments are photosensitive. The homogenate was filtered to remove precipitated protein and equivalent volumes of water and chloroform were added. The brown chloroform layer was separated, washed with water until the washes were no longer acidic, and dried over anhydrous Na2S04. After filtration, 2 d of a 0.5% solution of zinc acetate in chloroform was added before solvent removal at a rotary evaporator. The preparation of zinc complexes at this stage prevents the formation of artifactual metal complexes during purification (32, 34). The oily brown residue obtained on solvent removal was chromatographed on preparative Silica Gel GF plates (Analtech, Inc.) using 3:l chloroform/acetone as solvent. The zinc-complexed porphyrin pigments appear as visibly green, red-fluorescing bands with RF values of 0.5 to 0.8. One or two such bands were observed, depending on the substrate (see text). The pigment-containing frac- tions were scraped from the plate and the pigment was extracted with acetone. The isolated zinc-complexed porphyrins were further purified by high pressure liquid chromatography on a Whatman Partisil 10-PAC column using methanol/tetrahydrofuran 4:l (v/v) as solvent (2 to 3 ml/min of flow rate). The column effluent was monitored with a variable wavelength detector set at 431 nm. After removal of an aliquot of the collected pigments for spectroscopic analysis, the puri- tied zinc complexes were demetalated by standing in 1 ml of 5% HzS04/methanol for 15 min. Addition of equal volumes of water and chloroform, separation of the layers, washing of the chloroform fraction with water until the washes were no longer acidic, drying over anhydrous sodium sulfate, and solvent removal yielded pure, rnetal- free porphyrin pigments. The metal-free pigments were analyzed by high pressure liquid chromatography as previously described (31-34). In some instances, metal-free porphyrin pigments were reconverted into zinc complexes by the procedure given above.

Spectroscopic Analysis of Purified Pigments-The electronic (UV-visible) absorption spectra of both the metal-free and zinc-complexed pigments were determined in dilute chloroform solution on a Varian-Cary 118 spectrophotometer.

A Kratos/AEI MS-902 instrument with a field desorption source was utilized for the mass spectrometric studies. Operating conditions were the same as those previously reported (33). Except for norethisterone, a mass spectrum was obtained for each isolated porphyrin both in its free-base and zinc-complexed form. Due to the already high molecular weight of the norethisterone-porphyrin adduct, it was only possible to obtain a mass spectrum for the metal-free form. In those instances where pigments were resolved into two components, electronic absorption and mass spectrometric data were independ- ently obtained for each of the two components.

In Vitro Pigment Formation-The formation of abnormal pigments in vitro was established in a larger scale incubation of norethisterone with the usual microsomal enzyme preparation. The incubation mixture already described was used except for the following alterations in concentration: norethisterone (1 mM), NADPH (4 mM), and microsomal protein (4 mg/ml). The total incubation volume was 40 ml and the incubation time 40 min. At the end of the incubation, the microsomal protein was sedimented by ultracentrifugation (100,OOO X g for 1 h). The microsomal pellet was added to 5 ml of 5% H2S0,/methanol and was allowed to stand overnight. The pigment was extracted as already described. The electronic absorption spectra of the purified pigment, both as the free-base and as a zinc complex, were identical with those of the pigment obtained from in vivo experiments.

RESULTS

Sixteen acetylenic substrates have been incubated in this study with hepatic microsomes from phenobarbital-pretreated rats. A time-dependent decrease in measurable cytochrome P-450 content was caused by each of these substrates, in the presence of NADPH, under incubation conditions in which lipid peroxidation was suppressed (Fig. 1). No cytochrome P- 450 was lost in the absence of added NADPH (data not shown). Although a quantitative structure-activity correlation is not possible due to the heterogeneous nature of the assay system and the differential solubility of the substrates, the data clearly show that all of the structures investigated exhibit the same order of magnitude activity except for phenylacetylene and N-(l,l-dimethyl-2-propynyl)-3,5-dichlorobenzamide. A IO-fold increase in concentration of these two substances

5580

SUBSTRATE

C &*OH

0 &*

Inactivation of Cytochrome P-450 by Acetylenes

PERCENT LOSS OF CYlDCHROM P450 I IN Vlvo

!2f 5 ’ 29f6

23* 4

26f 6

29f 3

34

27i 3

3Ot 6

242 3

30f II

34f2

52t 3

YES

YES

-

YES

-

YES

YES

YES

YES

I FIG. 1. Destruction of cytochrome P-450 on incubation of

acetylenic substrates with hepatic microsomes from phenobarbital-pretreated rats. Incubations were carried out as described under “Experimental Procedures.” The data for norethisterone (D) are from Ref. 34 while the values for acetylene gas (M) are estimated from Ref. 41. The substrate concentration was nominally 1 mM except for phenylacetylene (K) and A”(l,l-dimethyl-2-propynyl)-3,5-dichlorobenzamide (N), which were examined at a 10 mM concentration since they caused only marginal cytochrome P-450 loss at the lower

was required to produce cytochrome P-450 losses equivalent to those observed with the other agents. Analogous data for acetylene gas, extrapolated from the graph reported by White (41), are included in the table for comparison.

Of the 17 active agents in Fig. 1, 13 were selected for in uiuo investigation. The livers of phenobarbital-pretreated rats were examined 4 h after administration of each of these 13 agents for the presence of abnormal (“green”) pigments. Using the extraction and purification methods described under “Exper- imental Procedures,” characteristic abnormal pigments were found in rats treated with each of the 10 substrates bearing a terminal acetylenic function (Fig. 1). In contrast, no abnormal pigments were isolated by the standard procedure from rats treated with the three agents in which the acetylenic moiety was disubstituted, even though t.hese agents caused in vitro loss of cytochrome P-450. In one of these three cases, that of 3-pentyn-2-01, this failure was due to the fact that rats receiving the usual 200 mg/kg dose of the agent died shortly after drug administration. It may be that this exceptional toxicity is related to the unique presence in this substrate of an oxidizable hydroxyl function adjacent to the acetylenic group, a pairing of functionalities which may lead to formation of a reactive carbonyl-conjugated acetylene. Large scale incubation of norethisterone with a microsomal enzyme preparation resulted in isolation of an abnormal pigment identical in its chromatographic and spectroscopic properties with that obtained from in vivo administration of the drug.

Abnormal hepatic pigments were analyzed by thin layer and high pressure liquid chromatography. The free-base form

PERCENT LOSS OF SUBSTRATE

1

20’ 24f 7

2522

23f 4

33

30f 2

33t 10

45t 3

34*5

30’

34* I

27* 2

23*4

40

32f3

35f8

47f I

36t5

IN VIVO

PIGMENT

YES

YES

-

YES

-

lox IC

NO

No

dose. The values given are averages of at least three determinations, with the exception that 3-methyl-1-pentyn-3-01 (C) was only assayed once. Standard deviations are given where applicable. Accumulation in. vivo of an abnormal hepatic pigment in rats treated with the given agents (see “Experimental Procedures”) is indicated by a yes in the column marked “pigment.” A no in this column indicates that pigment was looked for but was not found; a dash indicates that the experiment WIS not performed. Incubation time is gwen in minutes.

of 3 of the 10 pigments was resolved into two components by both of these techniques. The free-bases of the other seven pigments chromatographed as single bands, although the presence of unresolved isomers within these bands is suggested by frequently observed asymmetry in the high pressure liquid chromatography peaks. High pressure liquid chromatographic analysis of the zinc complexes was also performed, although these were not resolved into more than one peak even in those instances where two components were known to be present from the free-base chromatographic data. The high pressure liquid chromatographic analysis of a mixture of metal-free and zinc-complexed forms of the 3-phenoxy-I-propyne pigment, a typical example, is given in Fig. 2. The three pigments which could be resolved into two components were those engendered by norethisterone (34), 1-ethynylcyclohexanol, and I-eth-ynylcyclopentanol. These three substrates were the only ones with a tertiary hydroxyl group adjacent to a terminal ethynyl function. As already reported for norethisterone (341, the two components of each of the three pigments were not interconverted by complexation with zinc acetate and subsequent demetalation in 5% H2S0,/methanol. Only the single component used as a starting material was obtained after this reaction cycle. Nevertheless, even though not easily intercon- vertible, the evidence to be outlined below strongly suggests that the two components of each of these pigments are in fact closely related isomeric structures.

The electronic absorption spectrum of each of the purified pigments, and of each of the two components of a pigment when these were resolved, has been determined both in the


W 0 c 0 ? s n a d

I I I I 0 5 IO 15

Time (min 1 FIG. 2. High pressure liquid chromatographic analysis of a

mixture of the metal-free and zinc-complexed pigment isolated from rats treated with 3-phenoxy-1-propyne. A mixture of the metal-free and zinc-complexed pigment, obtained by addition of zinc acetate to pigment isolated as described under “Experimental Pro- cedures,” was chromatographed on a Whatman Partisil IO-PAC column. The first solvent, hexane:tetrahydrofuran:methanol, 10:3:1, was replaced at the point shown by an arrow with methano1:tetrahydrofuran, 4:l. A 3 ml/min solvent flow rate was used. The detector was a t 417 nm. The first peak is due to uncomplexed product and the second is due to the zinc complex.

metal-free and zinc-complexed state. Within experimental tolerance (approximately f 2 nm), the spectra of all of the uncomplexed pigments were essentially identical with two exceptions. The free-base of the pigment obtained with acetylene gas exhibited a Soret band a t 414 nm rather than at 420 nm as in the spectra of the other pigments. This can be seen by comparing (Fig. 3) the spectrum of the 3-phenyl-1-propyne pigment, a typical case, with that of the acetylene pigment. The second exception is provided by the phenylacetylene pigment, for which a reproducible free-base spectrum has not been obtained due to chemical instability of the chromophore. The spectra of the zinc complexes of the pigments were also essentially identical, in this case with no exceptions (Fig. 3). The zinc complex of the pigment obtained with phenylacetylene, in contrast to the free-base form, exhibited a stable spectrum which could not be differentiated from those of the other zinc-complexed pigments. Correlation of the spectra of both the metal-free and zinc-complexed forms of the pigments with the corresponding spectra of authentic porphyrins (42) leaves no doubt that the basic chromophore in the pigments is that of a porphyrin ring.

Field desorption mass spectra were obtained for the free base of eight of the isolated pigments, including separate spectra for each of the two resolved components in the pigments obtained with 1-ethynylcyclohexanol and l-ethynylcyclopentanol. A mass spectrum was also obtained of the zinc complex of each of the pigments (Table I). The mass spectrum of the pigment produced by treatment with acetylene gas is reproduced in Fig. 4. In general, these spectra are characterized by the presence of a variable ratio molecular ion doublet and by a minimal amount of peaks due to molecular fragmen- tation. The molecular ion doublet is due to desorption of protonated and unprotonated molecular species, the mono- protonated form being favored by higher emitter currents in the mass spectrometer. Protonated and alkali metal-complexed molecular ions are frequently observed in field desorption mass spectrometry (43). A sodium-complexed molecular ion was in fact observed for the norethisterone pigment (34). An instructive exception to the observation of a molecular ion doublet was provided by the 3-(2,4-dichlorophenoxy)-l-propyne pigment, which exhibited the molecular ion cluster ex-

I 1 I 1 I I 1

Wavelength (nm) FIG. 3. Electronic absorption spectra. Electronic (ultraviolet-

visible) absorption spectra in chloroform of (a) the free base of the pigment obtained with 3-phenyl-1-propyne (-); ( b ) the free base of the pigment obtained with acetylene gas (- - -); and ( c ) the zinc complex of the 3-phenyl-1-propyne pigment (.”.).

TABLE I Analysis of isolated hepatic pigments by field desorption mass

spectrometry The observed molecular ion for each purified pigment, after sub-

traction of the mass unit(s) due to the complexed proton or sodium cation (shown in parentheses), is given in the table. These data are given both for the free-base and for the zinc complex obtained on addition of zinc acetate. The principal peak in the zinc complex molecular ion cluster, corresponding to complexation with the major isotope of zinc (”Zn), is listed.

Predicted Observed molecular Ion ( m l e ) Substrate ( M , ) molecular ion -

( m / e ) Free base ”” Zn complex _ _ ~

A (124) B (110)

F (200) H (132) I (130) J (116) M (26)

D (298)

730 716 904

738 736 722 632

a06

730 (H’) 792 (H’) 716 (H’) 778 (H+) 886 (Na’) 806 (H+) 868 (H’) 738 (H’) aoo (H+) 736 (H’) 798 (H+) 722 (H’) 784 (H+) 632 (H’) 694 (H+)

633

I !

450 500 550 600 650

mk FIG. 4. Field desorption mass spectrum of the hepatic pig-

ment isolated from phenobarbital-pretreated rats exposed to acetylene gas. Administration of the agent and isolation of the pigment are described under “Experimental Procedures.” The principal peak observed in the spectrum (m/e 633) corresponds to mon- oprotonated molecular ion.

5582 Inactivation of Cytochrome P-450 by Acetylenes

pected of a mixture of protonated and unprotonated ions of a substance containing 2 chlorine atoms (44). The natural abun- dance ratio of isotopes (45) was also reflected in the molecular ion patterns of the zinc complexes. Again, the zinc complex of the 3-(2,4-dichlorophenoxy)-l-propyne pigment gave an unu- sually complex molecular ion pattern, as expected for a structure containing 1 zinc and 2 chlorine atoms. The molecular ion values tabulated in Table I correspond to the unprotonated molecular species. The values given for the zinc complexes and for the chlorine-containing compound are those of the major peak in the molecular ion clusters and, thus, correspond to the presence of MZn and '''CI. In those cases where a pigment was resolved into two components, both components gave the same mass spectrometric data so that the value in Table I is valid for both components.

DISCUSSION

The importance of the acetylenic group in norethisterone and ethynylestradiol for the destructive interaction of these substrates with hepatic cytochrome P-450 was first noted by White and Muller-Eberhard (26), who reported that destructive activity was lost on replacement of the ethynyl function with a hydrogen atom or an ethyl group. Our subsequent demonstration that the ethynyl function could also not be replaced by a vinyl moiety confirmed the specific role of the carbon-carbon triple bond in the destructive action of these hormonal steroids (34). The observation that nonsteroidal acetylenes also mediated the loss of cytochrome P-450 fur- thermore demonstrated that the sterol framework was not an essential component of the destructive interaction (34, 41). The more extensive structure-activity study described here unequivocally establishes that the potential to interact de- structively with cytochrome P-450 is an intrinsic property of the acetylenic group itself and that although the interaction can be attenuated or suppressed by the framework into which the triple bond is incorporated, no other substrate feature is essential for cytochrome P-450 destruction. Thus, this mono- oxygenase system is efficiently destroyed by monosubstituted acetylenes in which the attached carbon is not only mono- and di-, but also trisubstituted, a result which excludes destructive mechanisms based on formation of a delocalized radical by abstraction of an allylic hydrogen atom. The destructive activity of compounds such as l-ethynylcyclohexane and phenylacetylene clearly shows that a vicinal hydroxyl function is also not required, although involvement of the hydroxyl group in oxidative metabolism of the triple bond in ethynyl sterols (46,47) leaves open the possibility that it may intervene when present (34). Nevertheless, the lack of structural specificity implicit in the observation that all of the substrates listed in Fig. 1 destroyed cytochrome P-450 con- vincingly argues that the nature of the structure bearing the acetylenic moiety is not a primary determinant of destructive activity, although the decreased effectiveness of phenylacetylene and N-(l,~-dimethyl-2-propynyl)-3,5-dichlorobenzamide does show that the surrounding structure can interfere with the destructive process. The intrinsic activity of the acetylenic moiety is perhaps most strikingly demonstrated by the destructive action of acetylene itself (41), whereas the loss of cytochrome P-450 caused by the disubstituted acetylenes 3-pentyn-2-ol,4-methyl-2-octyn-4-o1, and 8-hexyne shows that it is the unsaturated bonds of the acetylenic group and not the acidic acetylenic proton which are responsible for the destructive interaction with cytochrome P-450, since in these substrates the acidic acetylenic proton is absent.

The destruction of cytochrome P-450 by 2-isopropyl-4-pentenamide, ethynyl sterols, and acetylene gas requires

NADPH, oxygen, and catalytically competent enzyme (26,41, 48). However, the most characteristic feature of the inactivation of cytochrome P-450 by these particular agents, the feature which uniquely distinguishes their mechanism, is the accompanying loss of the enzyme prosthetic heme moiety and the accumulation of abnormal hepatic pigments derived from that moiety (26,34,41,48). The destruction of cytochrome P- 450 by monosubstituted acetylenes has been clearly shown by this study to be mediated by this mechanism, both by demonstration that in every instance NADPH is required and, more definitively, that all of the new monosubstituted substrates examined in vivo yielded hepatic pigments indistin- guishable by absorption spectroscopy (Fig. 3) from that previously obtained with norethisterone (34). This identity in electronic absorption spectra, true both for the free-base and zinc-complexed forms of the pigments, was also valid for the two resolved components of the pigments obtained with norethisterone, 1-ethynylcyclohexanol, and l-ethynylcyclopentanol.

The in uitro destruction of cytochrome P-450 was mediated equally well by monosubstituted and disubstituted acetylenes (Fig. 1). In contrast, however, abnormal hepatic pigments were only observed in rats which had been treated in vivo with monosubstituted acetylenes or with acetylene gas. Anal- ogous hepatic pigments were not found in rats which had been treated with disubstituted acetylenes. This difference in the action of mono- and disubstituted acetylenes is an important one since pigment formation is intimately associated with the mechanism which governs destruction of cytochrome P-450 by acetylene gas and monosubstituted acetylenes. In the absence of such pigments, the mechanism by which disubstituted acetylenes destroy the enzyme can not be unambiguously linked to the mechanism defined here (see below) for the action of terminal acetylenes.

In preliminary studies using radiolabeled norethisterone, we have demonstrated that the resulting abnormal pigment is a covalent adduct of the drug with the porphyrin chromophore derived from heme (34). However, little structural information beyond the presence of a covalent link could be inferred from the demonstration that radioactivity remained associated with the chromophore throughout a series of chromatographic procedures. Although a field desorption mass spectrum of the norethisterone adduct was obtained, a clear rationale for the observed molecular ion could not be formulated (34). The limitations of this earlier study have now been transcended by mass spectrometric examination of seven other pigments obtained in this investigation including, where appropriate, separate mass spectra for chromatographically resolved components. The resulting data (Table I) unequivocally confirm that all of these agents act by a common mechanism in which 1:1 covalent alkylation by the causative agent of the protoporphyrin IX skeleton of prosthetic heme leads to inactivation of the enzyme and formation of the abnormal hepatic pigment. In each case, except that of norethisterone itself, the molecular ion observed corresponds exactly to that expected for the sum of the molecular weights of the dimethyl ester of protoporphyrin IX (M, = 590) plus the acetylenic agent plus 16 mass units, these last almost certainly due to an oxygen atom. The dimethyl ester of the porphyrin, rather than the free acid, is expected in this summation since the carboxyl groups are methylated during treatment with H2S04/methanol. Forma- tion of the dimethyl ester has been specifically confmed by deuterium-labeling studies in the case of the porphyrin adduct with 2-isopropyl-4-pentenamide (33). The iron atom is also lost from the adduct during the acidic extraction procedure. A similar sum of structural fragments is required by the


molecular ions of the zinc complexes of the pigments once allowance is made for the expected loss of pyrrolic protons during zinc complexation (Table I). Direct evidence for incor- poration of the intact destructive substrate into the porphyrin adduct is provided by the observation of a molecular ion pattern characteristic for the presence of 2 chlorine atoms in the mass spectrum of the pigment obtained with 3-(2,4-dichlorophenoxy)-1-propyne. These chlorine isotopic data independ- ently c o n f m t h e evidence obtained with radiolabeled norethisterone that the substrate is incorporated into the porphyrin pigment. The presence of the porphyrin, of course, is indicated by the chromophore of the pigment and by its physical properties, including the loss of protons on complexation with divalent zinc. Firm evidence also exists that the porphyrin fragment is derived from heme (26, 48). This independent evidence that both the substrate and protoporphyrin IX are incorporated into the isolated pigments and the observation that the molecular ion of not one but seven of these pigments (adducts) corresponds exactly to the sum of protoporphyrin JX (as the dimethyl ester) plus the substrate plus 16 mass units place beyond reasonable doubt the stoichiometry suggested for the adducts by the mass spectrometric data.

Direct evidence on the identity of the 16 mass unit structural fragment is not yet available. That it is an oxygen atom, however, is suggested by its mass, by the fact that the catalytic function of cytochrome P-450 is to transfer an oxygen atom to its substrate, by the fact that oxygen is required for destruction to occur, and by the fact that the enzyme catalytically partic- ipates in its own destruction. Further evidence, although indirect, is provided by the abnormal molecular ion measured for the norethisterone pigment (34) . This abnormal molecular ion, equivalent to the sum of the dimethyl ester of protoporphyrin IX plus norethisterone less 2 mass units, differs from that of the other seven adducts by 18 mass units (Table I). This difference is most consistent with loss of a molecule of water from the parent norethisterone adduct, probably during mass spectrometric analysis since the higher molecular weight of this adduct required harsher conditions for its desorption from the mass spectrometer emitter surface.

In addition to norethisterone (X), two other pigments, those obtained with 1-ethynylcyclohexanol and l-ethynylcyclopentanol, have been resolved chromatographically into two components. The electronic absorption and mass spectra of both the free-base and zinc-complexed forms of the two components in each of these pigments are identical. It thus appears that the destructive interaction can result in formation of closely related if still undefined isomeric adducts. The occurrence of analogous but unresolved isomers in the other pigments is, thus, a distinct possibility, particularly since no alterations in the spectroscopic data would be caused by their presence.

The evidence suggests that acetylenic substrates are covalently bound to prosthetic heme during ill-fated oxidation of the triple bond by the enzyme. Unfortunately, surprisingly little is known concerning oxidative metabolism of this unsaturated functionality, particularly since in most instances hy- drolytic and oxidative metabolites have not been differentiated. Three of the small number of compounds for which acetylene group metabolites have been reported are included in this study. In early work, phenylacetylene was found to be excreted by rabbits as phenylacetic acid (49). A number of acetylene group metabolites have also been shown to be excreted by rats and cows receiving N-(1,l-dimethylpro- p~~~)-3,5-dicNorobenzamide (50). These include products in which the acetylene has been transformed into 2-carbon al- cohol, acid, a-hydroxy acid, and methyl ketone functions (50). The formation of some or all of these metabolites by gastric

hydrolysis, however, can not be excluded since the agent was orally administered. Clearer evidence for oxidative metabolism is found in the demonstration that 17-hydroxy-17-ethynyl sterols such as norethisterone undergo oxidative expansion of the D-ring (46, 47). Evidence has also recently appeared for oxidative conversion of 4'-ethynyl-Z-fluorobiphenyl to (2-flu- oro-4'-biphenylyl) acetic acid (51). By analogy, this suggests that formation of phenylacetic acid from phenylacetylene is in fact an oxidative process. The metabolites obtained with these acetylenes can be rationalized by oxygen transfer to the acetylenic triple bond, although insertion into the acetylenic carbon-hydrogen bond can also be argued (51). Although it is clear that insertion into the carbon-hydrogen bond is not required for cytochrome P-450 destruction, since disubstituted acetylenes are active in this regard, the failure to isolate hepatic pigments with disubstituted acetylenes makes this an ambivalent result. Further experimentation will, therefore, be required to conclusively establish the unsaturated bonds of the acetylenic group as the focus of the destructive interaction.

Oxidation of a triple bond by oxygen-transfer agents such as peracids and cytochrome P-450, in analogy with the reactions of olefins, might be expected to yield unsaturated epoxides (Scheme 1). Unsaturated epoxides, however, are much less stable than epoxides due both to strain and electronic effects, so that their existence has only been demonstrated by indirect evidence (see Refs. 52 and 53). It has been calculated, in fact, that unsaturated epoxides are thermodynamically disfavored relative to tautomeric ketocarbene structures (54), of which dipolar oxovinyl cation structures are resonance forms (Scheme 1). Experimental evidence exists for the inter- conversion of these reactive structures (55,56). Oxygen transfer by the enzyme to the triple bond, whose occurrence is suggested by the available data on triple bond metabolism, is thus expected to result in the formation of a highly unstable (reactive) species. This species may be stabilized, as in 17- hydroxy-17-ethynyl sterols, by ring expansion or other reactions leading to isolable metabolites. On the other hand, the activated species may react with the prosthetic heme of the enzyme catalyzing its formation, particularly, if as shown in the model mechanism of Scheme 2, the oxyanion remains ligated to the heme iron atom. Retention of the alkoxide oxygen as a ligand is a distinct possibility if the activated oxygen, as shown, is an oxene. iron complex (1, 2) since the oxene would be transformed by acceptance of the n-bond electrons into the alkoxide ligand. It must be noted, however, that the basic alkylative mechanism proposed here does not in itself require either an oxene enzymatic intermediate or an oxygen-iron ligand interaction since in its barest essence, the only requirement is enzymatic conversion of the acetylenic group into the reactive species by oxygen transfer. As a working hypothesis, however, the mechanism illustrated in Scheme 2 is attractive because it rationalizes the catalytic

L SCHEME 1. Reactive species formed by oxidation of acetylenes.

5584 Inactivation of Cytochrome P-450 by Acetylenes

I I

Heme alkylation SCHEME 2. Possible mechanism for alkylation of the prosthetic heme in cytochrome P-450 during attempted metabolism of the acetylenic

function.

participation of the enzyme, the stoichiometry of the resulting prosthetic heme adducts reported here, and the specificity of the process (26).

Demonstration that 3-aryloxy- 1-propynes destroy hepatic cytochrome P-450 through a suicidal interaction with the enzyme clarifies the basis for their action as insecticide and drug synergists (14, 15, 18). The relative structural nonspecif- icity of this destructive interaction, evident on examination of Fig. 1, suggests that a similar destructive event may determine the synergistic activity of propargyl phosphonates (12, 18), propargyl oximes (12,19), alkynyl phthalimides (17), and more recently discovered agents such as the marine natural product dactylyne (57). The interaction of acetylenes with cytochrome P-450 is also of human toxicological importance. Although the physiological importance of the destruction of cytochrome P- 450 by sterols such as norethisterone is hard to evaluate, given the low daily dose commonly administered, the implications are more evident in other cases. This laboratory, for example, has recently obtained evidence that ethchlorvynol, a sedative- hypnotic in clinical use, is a potent inactivator of cytochrome P-450.' Considering the large doses normally utilized, and the even larger doses normally ingested during suicide attempts (58), it would be surprising if cytochrome P-450 destruction did not alter the metabolism and increase the physiological effects not only of ethchlorvynol itself but also of any co- administered drugs.

Acknowledgments-We gratefully recognize the competent experimental assistance of Mr. David McLendon and the generous access to the field desorption mass spectrometer provided by Professor A. Burlingame.

REFERENCES 1. Griffin, B. W., Peterson, J . A,, and Estabrook, R. W. (1979) in

The Porphyrins (Dolphin, D., ed) Vol. 7, pp. 333-375, Academic Press, New York

2. Keevil, T., and Mason, H. S. (1978) Methods Enzymol. 52c, 3-40 3. Lu, A. Y. H., Kuntzman, R., and Conney, A. H. (1976) in Frontiers

of Gastrointestinal Research (Van der Reis, L., ed) Vol. 2, pp. 1-31, S. Karger A. G., Base1

4. Testa, B., and Jenner, P. (1976) Drug Metabolism, Marcel Dek- ker, Inc., New York

5. Hutson, D, H. (1977) in Foreign Compound Metabolism in Mam- mals, Vol. 4, pp. 260-310, The Chemical Society, London

6. Daly, J. (1971) in Handbook of Experimental Pharmacology (Brodie, B. B., and Gillette, J. R., eds), Part 2, vol. 28, pp. 285- 311, Springer-Verlag, Berlin

' P. R. Ortiz de Montellano, J . M. Matthews, B. A. Mico, and K. Kunze, in preparation.

7.

8.

9.

10.

11. 12. 13.

14.

15.

16. 17.

18.

19.

20. 21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

Ryan, D. E., Thomas, P. E., Korzeniowski, D., and Levin, W. (1979) J. Biol. Chem. 254, 1365-1374

Fasco, M. J., Vatsis, K. P., Kaminsky, L. S., and Coon, M. J. (1978) J. Biol. Chem. 253, 7813-7820

Deutsch, J., Leutz, J . C., Yang, S. K., Gelboin, H. V., Chiang, Y. L., Vatsis, K. P., and Coon, M. J . (1978) Proc. Natl. Acad. Sci.

Parke, D. V. (1975) in Enzyme Induction (Parke, D. V., ed) pp.

Conney, A. H. (1967) Pharmacol. Reu. 19,317-366 Casida, J. E. (1970) J. Agric. Food Chem. 18, 753-772 Hodgson, E., and Tate, L. G. (1976) in Insecticide Biochemistry

and Physiology (Wilkinson, C. F., ed), pp. 115-148, Plenum Press, New York

Fellig, J., Barnes, J . R., Rachlin, A. I., O'Brien, J. P., and Focella, A. (1970) J. Agric. Food Chem. 18.78-80

Sacher, R. M., Metcalf, R. L., and Fukuto, T. R. (1968) J. Agric. Food Chem. 16, 779-786

Barnes, J. R., and Fellig, J. J. Econ. Entomol. 62.86-89 Neumeyer, J. I., and Incho, H. H. (December 23 1969) U. S.

Skrinjarid-Spoljar, M., Matthews, H. B., Engel, J. L., and Casida,

Henry, M., Dittrich, U., and Georg, P. (February 15 1973) Swiss

Onken, D., and Heublein, D. (1970) Pharmazie 25, 3-9 OMalley, K., Stevenson, I. H., and Crooks, J. (1972) Clin. Phar-

Field, B., Lu, C., and Hepner, G. W. (1979) Clin. Pharrnacol.

Mackinnon, M., Sutherland, E., and Simon, F. R. (1977) J. Lab.

Mackinnon, A. M., Sutherland, E., and Simon, F. R. (1978)

Einarsson, K., Ericsson, J . L. E., Gustafsson, J.-A., Sjovall, J., and

White, I. N. H., and Muller-Eberhard, U. (1977) Biochem. J. 166,

Botelho, L. H., Ryan, D. E., and Levin, W. (1979) J. Biol. Chem.

Schwartz, S., and Ikeda, S. (1955) in Ciba Foundation Symposium on Porphyrin Biosynthesis and Metabolism (Wolstenhohe, G . E. W., and Millar, E. C. P., eds) pp. 209-228, Little, Brown, and Co., Boston

McDonanh. A. F., Posuisil, R., and Meyer, U. A. (1976) Biochem.

U. S. A. 75,3123-3127

208-271, Plenum Press, New York

Patent 3,485,916

J. E. (1971) Biochem. Pharrnacol. 20, 1607-1618

Patent 531,828

macol. Ther. 13,552-557

Ther. 25, 196-198

Clin. Med. 90,1096-1106

Biochem. Pharrnacol. 27,29-35

Zietz, E. (1974) Biochim. Biophys. Acta 369, 278-293

57-64

254,5635-5640

SOC. Tians. 4, 297-298 DeMatteis. F.. and Unseld, A. (1976) Biochem. SOC. Trans. 4,

I ,

205-209 Ortiz de Montellano, P. R., Mico, B. A., and Yost, G. S . (1978)

Biochem. Biophys. Res. Commun. 83, 132-137 Ortiz de Montellano, P. R., Mico, B. A., Yost, G. S., and Correia,

M. A. (1978) in Enzyme-Actiuated Irreversible Inhibitors (Seiler, N., Jung, M. J., and Koch-Weser, J., eds) pp. 337-352, Elsevier, Amsterdam

Ortiz de Montellano, P. R., Yost, G. S., Mico, B. A., Dinizo, S. E.,


Correia, M. A., and Kumbara, H. (1979) Arch. Biochem. Bio- phys. 197,524-533

34. Ortiz de Montellano, P. R., Kunze, K. L., Yost, G. S., and Mico, B. A. (1979) Proc. Natl. Acad. Sci. U. S. A. 76, 746-749

35. Kunze, K. L., Ortiz de Montellano, P. R., and Yost, G. S. (1979) Fed. Proc. 38,691

36. Ortiz de Montellano, P. R., and Kunze, K. L. (1979) Pharmacol- ogist 21,177

37. Ortiz de Montellano, P. R., Kunze, K. L., Mico, B. A,, and Yost, G. S. (1980) in Microsomes, Drug Oxidations, and Chemical Carcinogenesis (Coon, M. J., Conney, A. H., Estabrook, R. W., Gelboin, H. V., Gillette, J. R., LaDu, B. N., and O’Brien, P. J., eds), in press

38. Hennion, G. F., and Boisselle, A. P. (1961) J. Org. Chem. 26,

39. Gaudemar, M. (1956) Ann. Chim. (Paris) 1, 161-216 40. Levin, W., Lu, A. Y. H., Jacobson, M., Kuntzman, R., Poyer, J.

L., and McCay, P. B. (1973) Arch. Biochem. Biophys. 158,842- 852

2677-2681

41. White, I. N. H. (1978) Biochem J. 174,853-861 42. Buchler, J. W. (1975) in Porphyrins and Metalloporphyrins

(Smith, K. M., ed) pp. 158-231, Elsevier, New York 43. Derrick, P. J. (1977) in Mass Spectrometry (Johnstone, R. A. W.,

ed) VoI. 4, pp. 132-145, The Chemical Society, Burlington House, London

44. Biemann, K. (1962) Mass Spectrometry: Organic Chemical Ap- plications, pp. 205-250, McCraw Hill, New York

45. Handbook of Chemistry and Physics (1972) 53rd Ed, pp. B-267, Chemical Rubber Co. Press, Cleveland

46. Sisenwine, S. F., Kimmel, H. B., Liu, A. L., and Huelius, H. W. (1979) Drug Metab. Disp. 7, 1-6

47. Abdel-Aziz, M. T., and Williams, K. I. H. (1969) Steroids 13,809- 820

48. DeMatteis, F. (1978) in Heme and Hemoproteins, Handbook of Experimental Pharmacology (De Matteis, F., and Aldridge, W. N., eds), Vol. 44, pp. 129-155, Springer-Verlag, New York

49. El Masri, A. M., Smith, J. N., and Williams, R. T. (1958) Biochem. J. 68, 199-204

50. Yih, R. Y., and Swithenbank, C. (1971) J. Agric. Food Chem. 19,

51. Sullivan, H. R., Koffey, P., and McMahon, R. E. (1979) Drug

52. Concannon, P. W., and Ciabattoni, J. (1973) J . Am. Chem. Soc.

53. Franzen, V. (1958) Justus Liebigs Ann. Chem. 614,31-36 54. Strausz, 0. P., Gosavi, R. K., Denes, A. S., and Csizmadia. I. G .

55. Zeller, K. P. (1977) Angeu). Chem. Int. Ed. 16, 781-782 56. Timm, U., Zeller, K. P., and Meier, H. (1978) Chem. Ber. 111,

57. Kaul, P. N., and Kulkarni, S. K. (1978) J. Pharm. Sci. 67, 1293-

58. Teehan, B. P., Maher, J. F., Carey, J . J. H., Flynn, P. D., and

320-324

Metab. Disp. 7, 76-80

95,3284-3289

(1976) J . Am. Chem. Soc. 98,4784-4786

1549-1557

1296

Schreiner, G. E. (1970) Ann. Intern. Med. 72, 875-882

chemlstry vol. 255. no. issue of june 25. pp. 557p-5.585 ... · self-catalyzed inactivation of...

Documents