Investigation of the Mechanistic Basis of N,N ... - ACS Publications

consistent with the finding by Hall and Hanzlik, according to which the intramolecular KDIE on .... W., and Curzon, E. H. (1986) Identification by pro...
0 downloads 0 Views 1MB Size
Chem. Res. Toxicol. 1993,6, 197-207

197

Investigation of the Mechanistic Basis of N,N-Dimethylformamide Toxicity. Metabolism of N'N-Dimethylformamide and Its Deuterated Isotopomers by Cytochrome P450 2El Jaroslav Mrk,+Parmjit Jheeta,i Andreas Gescher,*J Ruth Hyland,t Kenneth Thumme1,S and Michael D. Threadgill11 Mechanisms of Drug Toxicity Research Group, Pharmaceutical Sciences Institute, Department of Pharmaceutical Sciences, Aston University, Aston Triangle, Birmingham B4 7ET, U.K.,National Institute for Public Health, Srob6rova 48, 10042 Prague 10, Czech Republic, Department of Pharmaceutics, School of Pharmacy, University of Washington, Seattle, Washington 98195, and School of Pharmacy and Pharmacology, University of Bath, U.K. Received November 24, 1992

Dimethylformamide (DMF) is an industrial solvent with hepatotoxic properties. The toxicity of DMF has been associated with its metabolism to S-(N-methylcarbamoy1)glutathione(SMG). The major urinary metabolite of DMF is N-(hydroxymethy1)-N-methylformamide(HMMF). HMMF undergoes oxidation in the formyl moiety, possibly via the intermediacy of its hydrolysis product N-methylformamide (NMF), and the reactive intermediate thus generated reacts with glutathione to yield SMG. Experiments were conducted to elucidate enzymatic details of the metabolism of DMF. Generation of HMMF from DMF in microsomes from rats which had received acetone, an inducer of cytochrome P450 2E1, was increased by 175 % over that observed in control microsomes. In liver microsomes from 4 humans the metabolism of DMF to HMMF was inhibited by a monospecific antibody against rat liver P450 2E1, and the metabolic rates were correlated with those of NMF to SMG, a process known to be mediated via P450 2E1. DMF was also metabolized by purified rat liver P450 2E1. The kinetic parameters which characterize the metabolism of DMF or its deuterated isotopomers to the respective HMMF isotopomers, of HMMF to SMG and of NMF to SMG in liver microsomes, were computed from Eadie-Hofstee plots. The affinity of DMF for the metabolizing enzyme in rat liver microsomes is considerably higher (apparent Km = 0.20 mM) than that of NMF (K, = 4.28 mM) or of HMMF (K, = 2.52 mM). The respective values observed with human microsomes are very similar. The apparent K, values for the N-methyl oxidation of Nfl-dimethyldeuterioformamide ( [2H1]DMF)andNfl-bis(trideuteriomethy1)formamide ([2H61DMF)in rat microsomes are 0.14 and 0.21 mM, respectively. The apparent Vm,, for the oxidation of [2HJDMF is similar to that computed for DMF, and the Vm, for [2H61DMFis less than half of that computed for DMF. The kinetic deuterium isotope effect (KDIE) on DMF metabolism was determined in incubations with rat microsomes in three ways: (i) the noncompetitive intermolecular KDIE by the ratio of Vm,/K, for DMF to V&Km for [2H6]DMF, (ii) the competitive intermolecular KDIE as the quotient of metabolic products HMMF to N-(hydroxydideuteriomethy1)-N-(trideuteriomethy1)formamide in incubations of DMF together with [2H6]DMF,and (iii) the intramolecular KDIE as the quotient of the ratio of N-(hydroxymethy1)-N-(trideuteriomethy1)formamide to N-(hydroxydideuteriomethy1)-N-methylformamidegenerated from N-(trideuteriomethy1)-Nmethylformamide ([2H3]DMF). The respective values were found to be (i) 2.4, (ii) 5.0, and (iii) 5.2. DMF inhibited the oxidation of NMF or HMMF to SMG. Deuterium substitution of the DMF methyl hydrogens did not affect the apparent Ki for the inhibition of the oxidation of NMF to SMG. Among a series of 8 formamides and acetamides structurally related to DMF only N,N-diethylformamide and Nfl-dimethylacetamide were equally effective with DMF as inhibitors. The results suggest that (i) hepatic P450 2E1 is an important catalyst of the metabolism of DMF and related low-molecular-weightamides, (ii) DMF inhibits its own metabolic toxification, and (iii) there is a marked KDIE on the metabolic oxidation of DMF.

Introduction N,N-Dimethylformamide [DMF,l (CH&NCHOl is a polar solvent used widely in a variety of industrial processes, among them the manufacture of synthetic fibers, * To whom correspondence should be addressed.

I Aston National Institute for Public Health. University. +

5 11

University of Washington. University of Bath.

leathers, films, and surface coatings. Worldwide production of DMF annually has been estimated to be 2 X 105 tons (2). Gastric irritation and hepatotoxicityare the major effects which have been reported in exposed to DMF (for review see ref 3). For example, an elevation in biochemical indices of hepatotoxicity has been described recently in workers occupationally exposed to DMF under conditions of poor industrial hygiene ( 4 , 5 ) . These reports have been considered of sufficient gravity

0893-228x/93/2706-0197$04.00/00 1993 American Chemical Society

-

198 Chem. Res. Toxicol., Vol. 6, No. 2, 1993 0

0

CD3\

N-C

DMF

HOCD

/p ‘R

CD/

D: [‘HJ

R

DMF

p

CH,

Figure 1. Metabolism of DMF and NMF. The methyl isocyanate ion in brackets is a postulated intermediate. “GS-” denotes glutathionyl.

to warrant the publication of the U S . National Institute for Occupational Safety and Health of a reminder of the protection methods to avert adverse health effects of DMF (6). There is little doubt that the mechanism of hepatotoxicity exerted by DMF and other related formamides involves their metabolism. The major urinary metabolite of DMF in rodents and humans is N-(hydroxymethy1)N-methylformamide (HMMF) (7-9) (Figure 1). Another quantitatively important urinary metabolite of DMF is the mercapturate N-acetyl-S-(N-methylcarbamoy1)cysteine (AMCC) (10, II), the progenitor of which, S-(Nmethylcarbamoy1)glutathione (SMG), is the product of the reaction of a reactive metabolic intermediate, probably the highly toxic methyl isocyanate (CHsNCO), with GSH (12,13) (for structures see Figure 1). AMCC is also the major metabolite of the hepatotoxic experimental antineoplastic agent and industrial solvent NMF, the N-demethyl analog of DMF (10, 14). The elusive metabolic intermediate via which DMF and N-methylformamide (NMF) are thought to mediate toxicity is the product of oxidation at the formamide formyl group (12). The contention that metabolism is a crucial determinant of formamide toxicity is borne out by the finding that of a series of related formamides and acetamides only those which underwent metabolic oxidation in the formyl moiety were found to possess hepatotoxic properties (15). The overall metabolism of DMF to AMCC in vivo was quantitatively much more prominent in humans than in rodents ( I I ) , and we argued therefore that DMF may pose a higher risk to occupationally exposed humans than that which might be inferred from experiments on rodents. Little is known about the enzymeswhich catalyze oxidation of DMF to HMMF and about the steps which lead ultimately to AMCC. We wished to study these issues in order to solve two mechanistic enigmas which are associated with the metabolism of DMF and probably bear upon the expressionof its toxicity. First, AMCC is excreted with the urine in humans exposed to DMF only after a Abbreviations: AMCC, N-acetyl-S-(N-methylcarbamoy1)cysteine; P450 2E1, cytochrome P450 isozyme, also known as CYP2E1, IIE1, ac, j, or 3a, for description of nomenclature see ref 1;DEDTC, diethyldithio-

carbamate; DMF, NJV-dimethylformamide; PH,]DMF, N,N-bis(trideuteriomethy1)deuterioformamide;[2H6]DMF,NJV-bis(trideuteriomethy1)formamide; [2H:~IDMF,N-(trideuteriomethy1)-N-methylformamide; PHjIDMF, N,N-dimethyldeuterioformamide;GLC, gas-liquid chromatography; GSH, glutathione; HMMF, N-(hydroxymethy1)-N-methylformamide; [”HI;]HMMF, N-(hydroxydideuteriomethy1)-N-(trideuteriomethy1)deuterioformamide;[ZHJHMMF, N-(hydroxydideuteriomethy1)N-(trideuteriomethy1)formamide;[ZHJHMMF, N-(hydroxymethy1)-N(trideuteriomethy1)formamide;L2H21HMMF, N-(hydroxydideuteriomethyl)-N-methylformamide; PHIIHMMF, N-(hydroxymethy1)-Nmethyldeuterioformamide; KDIE, kinetic deuterium isotope effect; NMF, N-methylformamide; [*H,]NMF, N-(trideuteriomethy1)deuterioformamide; [*H:JNMF, N-(trideuteriomethy1)formamide;[*HIINMF, N methyldeuterioformamide; SMG, S-(N-methylcarbamoy1)glutathione.

/N-C [‘H,]

R = D:

[‘H,] DMF

[2H6] HMMF //o

/N-C HOCH2

DMF

CH3 \ /N -CH0 CH, \D

\R

CD3\

\H

/p

R = H: [‘H5] HMMF

-

0

CD3\

.:N-C

CD/

R = H: [‘He] DMF

HMMF

Mrhz et al.

[‘H,]

-

\H

HMMF

HOCD2\

+

N-C CH,/

No ‘H

p2]HMMF

CH,\

//O /N -C HOCH, \D

[‘H,]

HMMF

Figure 2. Metabolic oxidation of DMF isotopomera [2H,JDMF and [2H7]DMF(A), r2H3]DMF (B), and [*HJDMF (C).

considerable delay, commencing beyond termination of exposure, whereas the urinary excretion of HMMF occurs within the first hour of exposure (1I,16).In rodents, the delay in urinary AMCC excretion after ip administration of DMF increased with DMF dose (11). The appearance of biochemical indices of hepatotoxicity in rata exposed to DMF at high vapor concentration was similarly retarded (17). The reason for this delay is unclear. Second, the metabolic generation of HMMF from DMF and that of SMG from NMF can conveniently be studied in vitro in preparations of rodent and human microsomes; however, it has not been possible to observe the formation of SMG from DMF under such conditions ( I & , and so far there has been no explanation for this discrepancy. Both phenomena, delayed excretion of AMCC after DMF exposure in vivo and lack of metabolic SMG generation from DMF in vitro, could be caused by inhibition of the metabolic generation of SMG by DMF or one of ita metabolites. This hypothesis was tested in the work described here. Furthermore, it was considered appropriate to characterize the cytochrome P450 isozyme involved in DMF metabolism. The DMF molecule is isoelectronic with the notorious carcinogen N,N-dimethylnitrosamine, which is a good substrate of cytochrome P450 2E1 (19). In view of this similarity and the recent fiiding that the microsomal oxidation of NMF is catalyzed by P4502E1(20),we investigated the possibility that DMF is a substrate of this enzyme. This was indeed found to be the case. Yang et al. reported recently that DMF inhibits the N-demethylation of N,N-dimethylnitrosamine and that this inhibition is subject to a kinetic deuterium isotope effect (KDIE), as adjudged by comparison with N,N-bis(trideuteriomethy1)deuteriofonnamide(L2H,1DMF) (21).We wished to test the hypothesis that the interaction of DMF with P450 2E1, as shown by its oxidation to HMMF and its ability to interfere with the oxidation of NMF, is affected by replacement of the DMF hydrogen atoms with deuterium. To that end, the DMF isotopomers N,N-dimethyldeuterioformamide ( [2HJDMF), N-(trideuteriomethy1)-N-methylformamide( C2H3]DMF),N,Nbis(trideuteriomethy1)formamide ([2HsIDMF),and L2H,1DMF (for structures see Figure 2) were purchased or synthesized and subjected to metabolism studies in rat liver microsomes in vitro.

Microsomal Metabolism of N,N-Dimethylformamide

A

I

1

I

I

2 3 Retention time (min)

I\

Figure 3. HPLC separation of HMMF (retention time 3.9 min) from NMF (retention time 3.4 min) (top trace) and formaldehyde (bottom trace). Formaldehyde content in the fractions of the HPLC eluate was determined using the colorimetric method of Nash (24). Values on the y-axis are arbitrary.

Experimental Procedures Chemicals. DMF, NMF, N,N-dimethylformamide, N,Ndibutylformamide, N-ethylformamide, NJV-dimethylacetamide, NJV-diethylacetamide, N-methylacetamide, formamide, acetamide, and diethyldithiocarbamate (DEDTC) were bought from Aldrich Chemical Co. (Poole, U.K.). DMF and NMF were purified by distillation to >99% purity (GLC) prior to use. [2H& DMF was prepared from [2He]dimethylamine hydrochloride (Sigma Chemical Co. Ltd., Poole, U.K.) and ethylformate essentially according to the method described for the synthesis ofN-(trideuteriomethy1)formamide([*H3]NMF)(22). In analogy [2H1]DMFwas prepared from dimethylamine hydrochloride and [2Hl]methylformate (Aldrich) and PH31DMF from ethylformate and N-(trideuteriomethy1)-N-methylamine hydrochloride, which was kindly provided by Dr. R. P. Hanzlik (University of Kansas). The purity of the synthesized isotopomers was verified by GLC and 'H-NMR spectroscopy. Biochemicals (glucose 6-phosphate, glucose-6-phosphatedehydrogenase, GSH, NADP) were obtained from Sigma. HMMF was prepared according to Gate et al. (23). The crude product of the synthesis contained 6-8% NMF as determined by HPLC analysis (vide infra) and a similar amount of formaldehyde as determined colorimetricallyaccording to Nash (24). The product was purified by semipreparative HPLC to furnish HMMF of suitable purity for in vitro metabolism studies. The HPLC system used for this purpose consisted of a Waters 510 pump, a Waters WISP 710 autoinjector, a Shimadzu SPD6A UV detector (set at X = 233 nm), and a Waters Maxima 820 workstation. Separation was achieved on a Lichrospher 100RP18 column (250 X 4 mm, 5-pm particle size) through which 10 mM phosphate buffer (pH7.4) was pumped at a flow rate of 1.0 mL/min. Under these conditions the contaminants in 20-pL samples containing 10% HMMF were sufficiently separated from the carbinolamide. Figure 3 demonstrates the separation by HPLC of HMMF, NMF, and formaldehyde in a typical sample of crude product. The HMMF content in the collected eluate was measured by GLC as NMF (vide infra); residual NMF was determined by HPLC as described above. SMG was synthesized

Chem. Res. Toxicol., Vol. 6,No. 2,1993 199 by Drs. D. H. Han and P. G. Pearson, University of Washington, Seattle, as described previously (25).PropylN-methylcarbamate was prepared from AMCC and 1-propanol using the procedure for derivatization of AMCC in preparation for GLC analysis (26). All other chemicals and reagents were available in the laboratory. Animals a n d Source of Human Microsomes. Male Sprague-Dawley rats (170-200 g) were purchased from Bantin and Kingmans Ltd. (Hull, U.K.). In some experiments rats were used which had been given acetone as a 50% solution in saline (5 mL k g l ) by single oral gavage 24 h prior to preparation of microsomes (27). Control animals received saline. Excess samples of healthy human liver tissue were obtained after graft reduction of donor liver from the Liver Transplant Unit at the Queen Elizabeth Hospital (Birmingham, U.K.). Tissues originated from four organ donors, two females and two males, aged between 10 and 55 years. Homogenate (25% ) of human or rodent liver was prepared in Tris-buffered (50mM, pH 7.4) KCl solution (154 mM). Human liver microsomes, which were prepared within 3 h after liver became available, were stored in liquid nitrogen for up to 12 months before use. P450 2E1 was purified from streptozotocin-treated rats as described previously (20). Preparation of Liver Microsomes a n d Microsomal Incubations. Microsomes were obtained in the usual way by differential centrifugation of homogenate first a t (9 X 103)gfor 20 min, and then a t 105g for 1 h in a Beckman L8-60M ultracentrifuge. The microsomal pellet was suspended in Tris buffer (50 mM, pH 7.4), recentrifuged at 105g for 1 h, and resuspended in phosphate buffer (50mM, pH 7.4). The rat liver microsomes used in each experiment were pooled from 2-4 animals. All incubations were carried out, with duplicate samples, in 5-mL glass vials under shaking at 37 "C. Incubates contained microsomes (1-2.5 mg of protein-mL-I), a NADPH generating system [glucose 6-phosphate (20 mM), NADP (10 mM), glucose6-phosphate dehydrogenase (4 IU)], GSH (10 mM), phosphate buffer (50 mM, pH 7.4), and substrate with or without inhibitor a t the concentrations decribed below in a final volume of 2 mL. The GSH in the mixture converted metabolically generated N-methylcarbamoylating species to SMG (18). Reactions were initiated by addition of substrate after a 3-min preincubation period a t 37 OC and terminated after incubation for 30 min. Preliminary experiments established that the metabolic generation of HMMF from DMF and that of SMG from NMF or HMMF were linear with time during this time period. The monospecific antibody against rat liver P450 2E1 was prepared and purified as described recently (20). Preparation of Samples a n d Quantitation of Metabolites. In preparation for analysis by GLC samples of the incubate were processed as described by M r b (26)and M r b et al. (28). This procedure involves derivatization of metabolically produced SMG with ethanol in alkali to furnish ethyl N-methylcarbamate and, at the same time, conversion to HMMF to NMF. Likewise metabolically generated N-(hydroxydideuteriomethy1)-N-(trideuteriomethy1)deuterioformamide(r2Hs]HMMF)is converted to N-(trideuteriomethy1)deuterioformamide( [2H4]NMF),N(hydroxydideuteriomethyl)-N-(trideuteriomethyl)for"idd [2H5]HMMF) and N-(hydroxymethy1)-N-(trideuteriomethy1)formamide ([2H31HMMF) to N-(trideuteriomethy1)formamide ( [2H31NMF), N-(hydroxydideuteriomethy1)-N-methylformamide ( [2H2]HMMF)to NMF, and N-(hydroxymethy1)-N-methyldeuterioformamide ( [2H1]HMMF) to N-methyldeuterioformamide U2HllNMF). Metabolic incubations were terminated by transfer of a sample (1.5 mL) of the incubate into a tube containing ethanol (3 mL) with propyl N-methylcarbamate (10 pM), which served as internal standard. For the determination of ethyl N-methylcarbamate (the analytical derivative of SMG) an additional extraction step with ethyl acetate was included (11). Gas chromatographic determination of ethyl N-methylcarbamate and of NMF isotopomers was performed using a HP-5890A gas chromatograph with a nitrogen-selective detector, attached to a fused silica capillary HP-2OM column (25 m X 0.32 mm i.d.; 0.32-pm film thickness). The column oven temperature was 130

Mr&z et al.

200 Chem. Res. Toxicol., Vol. 6, No. 2, 1993

n,7 1OOpm

0

10

20

30

40

Retention time (min)

Figure 4. GLC separation of NMF and [2H3]NMFin extracts of incubates of DMF and [2H6]DMFwith rat liver microsomes. (i), (ii), and (iii) denote peaks with the retention times of DMFi [2H6]DMF,[2H3]NMF,and NMF, respectively. Note that NMF and [2H3]NMFare the products of of hydrolysis of HMMF and [2H5]HMMF,respectively, generated during sample preparation. For details of incubation and metabolite analysis see the Experimental Procedures. “C. Helium was used as carrier gas with an inlet pressure of 0.42 bar leading to a flow rate of 0.7 mlimin. Under these conditions retention times were 3.5 min for DMF, 4.9 min for ethyl N-methylcarbamate, 6.5 min for propyl N-methylcarbamate (internal standard), and 7.5-7.7 min for NMF and its deuterated isotopomers. In order to achieve sufficient resolution between [2H3]NMF and NMF (Figure 4), the column temperature was kept a t 90 OC. For quantitation the height of the peaks caused by ethylN-methylcarbamate or the NMF isotopomers was related to that of propyl N-methylcarbamate. The calibration curve for SMG was derived using authentic compound. NMF isotopomers were quantitated using a calibration curve constructed with NMF. To obtain these curves, SMG or NMF was added to blank microsomal suspensions. It has to be stressed that this method does not discriminate between NMF and HMMF, as HMMF yields NMF by alkaline hydrolysis during sample preparation or by complete thermolysis during GLC analysis. However, we have previously given unequivocal evidence for the contention that the product of DMF metabolism in vivo or in vitro is predominantly, if not exclusively, HMMF (7-9). In the experiments in which NMF was coincubated with either DMF, its deuterated isotopomers, or other amides to inhibit NMF metabolism, these agents or their biotransformation products did not interfere with the analytical quantitation of SMG. The accuracy and reproducibility of the method has been described previously (26,28, 29).

Formaldehyde was determined spectrophotometrically (24) using a Cecil CE 594 instrument. Cytochrome P450 levels were measured spectrophotometrically as described by Gibson and Skett (30). Protein content of microsomal suspensions was determined according to the method of Lowry et al. (31). Enzyme Kinetics. Rates ( u ) of the metabolic generation of HMMF and its deuterated isotopomers from DMF isotopomers or of SMG from NMF and HMMF were plotted against u / s (s = substrate concentration) to furnish Eadie-Hofstee plots (32). Visual inspection of the plots allowed facile discrimination of two linear phases denoting processes characterized by different affinities for the microsomal enzymes. Linearity of the relationship between u and uis for the high-affinity component was confirmed by linear regression analysis which afforded r > 0.98 for each amide. Apparent K , and V,,, values were extrapolated from these plots. In the case of DMF as substrate the highaffinity portion covered substrate concentrations of 0.02 to 5 mM, the low-affinity section 20-100 mM; in the case of NMF and HMMF as substrates the high-affinity segments spanned concentrations of 0.4-10 mM, while the low-affinity portion was measured only for NMF and extended from 40 to 200 mM.

Figure 5. Effect of pretreatment of rats with acetone (A), and of coincubation of microsomes with DEDTC (B)on the metabolism of DMF (10 mM) to HMMF in supensions of rat liver microsomes. The closed and crossed bars in (A) represent metabolism in microsomes from control and acetone-pretreated rats, respectively. Numbers in the bars in (B) indicate DEDTC concentrations. For details of incubation and metabolite analysis see the Experimental Procedures. Values are the mean SD of 4 experiments.

*

The inhibitory influence of amides on the metabolic formation of SMG from NMF was investigated employing two experimental designs: (i) NMF concentration was 1mM, and concentrations of DMF, [2Hs]DMF, or other amides were selected such that product formation was within 20-80% of that observed in the noninhibited reaction. Apparent Ki values were calculatedusing the equation K , = K,Xi/[(u/u’ - 1)(K, + s)], where u and u’ are the reaction rates in the absence and presence, respectively, of inhibitor, s and i are the concentrations of substrate and inhibitor, respectively, and K , is obtained from the noninhibited reaction. This equation was derived by substitution from the two formulas u = Vm,,Xs/(Km s) and u‘ = Vm,,Xs/[Km(l i / K J s] (32). (ii) The effect of DMF on the reaction NMF SMG was also studied using 10 different NMF concentrations between 0.4 and 10 mM and a DMF concentration of 0.1 mM. Here the apparent Ki value was calculated using the formula Ki = K,Xi/(K’, - K,), where i is the inhibitor concentration and K , and K’, were obtained in the absence and presence, respectively, of the inhibitor (32). The influence of DMF on the formation of SMG from HMMF was evaluated a t a HMMF concentration of 5 mM and at DMF concentrations of 0.2,0.5,1, and 2 mM; the apparent Ki was computed as outlined under (i) above. The kinetic experiments were performed with duplicate incubates for each substrate and inhibitor concentration, and each apparent K , and Ki value quoted under Results or in Tables 1-111 is the mean f SD of between 3 and 12 separate experiments. It has to be stressed that in view of the heterogeneous nature of the enzyme source the values of the kinetic constants should be interpreted cautiously, and the constants should be looked at as indicators rather than as definers of the real values or of differences between agents (33). The statistical significance of differences between values was evaluated using Student’s t-test.

+

+

+

-+

Results

Role of P450 2El in DMF Metabolism. In view of the role which P450 2E1 plays in the metabolic N-demethylation of N,N-dimethylnitrosamine (191, we tested the hypothesis that the N-methyl oxidation of DMF, its major metabolic route (7-91, is also catalyzed by this enzyme. Five experimental strategies were employed to investigate this hypothesis. First, metabolism of DMF was measured in microsomes from rat livers in which P450 2E1 had been induced by administration of acetone. Figure 5A shows that the rate at which microsomes from acetone-treated animals oxidized DMF to HMMF was increased by 175% over metabolism in microsomes from control rats. Second,

Chem. Res. Toxicol., Vol. 6, No. 2, 1993 201

Microsomal Metabolism of N,N-Dimethylformamide

Vmax (NMF-SMG)

0

I

I

I

0.5

1 .o

1.5

Vmax (DMF-

HMMF)

[

mg prot. x min. nmo'

Figure 6. Correlation of apparent V,,, of metabolism of NMF with that of DMF in microsomes from 4 human livers. The V,,, values used are the ones which describe the enzyme with high affinity for NMF and DMF. For details of incubation and analysis of metabolites see the Experimental Procedures. Standard linear regression analysis gave a correlation coefficient of r = 0.99.

0

05

15

10

20

vis

Figure 8. Eadie-Hofstee plot of generation of HMMF from DMF in incubates of rat liver microsomes. The individual values are the mean of 2 incubations. They were obtained with substrate concentrations ranging from 0.02 to 100 mM. The plot shown is representative of 4 experiments. Symbols "u" and "s" represent the reaction rate and substrate concentration, respectively. For details of incubation and metabolite analysis see the Experimental Procedures, and for the mean apparent K , and V,,, values see Table I.

0.4

[

]

mgnmol prot.SMG, x mln. 0.3

I\. 1I

0.2

4

z

201

I

0

0 05

0 10

.*

Y

I

0 15

vis

O

O

2 4 6 8 IgG Concentration (mg/nmol P-450)

10

Figure 7. Effect of anti-rat P450 2E1 IgG (closed circles) or preimmune IgG (open squares) on the metabolism of DMF (10 mM) in human liver microsomes. Values, which are expressed as percentage of the rate of metabolic production of HMMF from DMF in incubations without IgG, are the mean of two individual microsomal preparations, each conducted in duplicate. Details of incubation conditions and metabolite analysis are described under Experimental Procedures.

metabolism of DMF was studied in incubates with microsomes in the presence of DEDTC which is a relatively specific inhibitor of P450 2E1 at concentrations in the 10-5-104 M range (34). DEDTC indeed inhibited DMF oxidation potently (Figure 5B). Third, the rate of metabolism of DMF was compared with that of NMF, a P450 2E1 substrate (20),in liver microsomes from four different humans. The plot obtained (Figure 6) shows a good correlation between rates. Fourth, a monospecific antibody against rat P450 2E1 inhibited DMF metabolism in incubations with human liver microsomes substantially (Figure 7). Microsomal N-demethylation of aminopyrine was not affected by the presence of this antibody (result not shown). Fifth, in a preliminary experiment, HMMF was also detected in incubates of DMF (10 mM) with purified rat liver P450 2E1 after reconstitution with cytochrome P450 reductase and cytochrome b5. The mixture was incubated for 1 h in order to maximize metabolite formation. The amount of HMMF generated, 110 nmol.(nmol of cytochrome P450)-', was 5 times that which was found under identical incubation conditions in microsomes from livers of streptozotocin-treated rats, which served as the source of the purified enzyme. The

Figure 9. Eadie-Hofstee plot of metabolic generation of SMG from NMF in incubations of rat liver microsomes. The individual values are the mean of 2 incubations. They were obtained with substrate concentrations ranging from 0.2 to 200 mM. The plot shown is representative of 9 experiments. Symbols "v" and *s" represent the reaction rate and substrate concentration, respectively. For details of incubation and metabolite analysis see the Experimental Procedures. For the mean apparent K , and V,,, values see Table I.

difference in HMMF production between purified enzyme and microsomes mirrors the contribution of P450 2E1, as determined by p-nitrophenol oxidation, to the overall microsomal cytochrome P450 (result not shown). Kinetic Analysis of Formamide Metabolism. In this series of experiments we wished to elucidate the discrepancy in the metabolic generation of N-methylcarbamic acid thioester from DMF between humans or animals in vivo and microsomal incubations in vitro, that is, to ascertain why the microsomal metabolism of DMF yields only HMMF but not SMG (18). To that end, DMF, NMF, and HMMF were incubated with rat and human liver microsomes. The rates of the enzyme-catalyzed reactions (i) DMF HMMF, (ii) NMF SMG, and (iii) HMMF SMG were determined and used to construct EadieHofstee plots. The plots for reactions i and ii are shown in Figures 8 and 9. They consist of two linear segments, which relate to low and high substrate concentrations and describe catalytic reactions of, respectively, high and low affinity for the enzyme. The apparent Kmand Vm, values calculated from the high-affinity sections of the plots are shown in Table I. In the case of the metabolic formation of HMMF from DMF in rat liver microsomes (Figure 8) the high-affinity segment of the Eadie-Hofstee plot is

- -

-

202 Chem. Res. Toxicol., Vol. 6,No. 2, 1993

Mr4z et al.

Table I. Kinetic Parameters for the Metabolic Oxidation of Formamides in Rat and Human Liver Microsomes app V,,, [nmol.(mg of microsomal protein)-'.min-'] app K m (mM) substrate product rat human rat human HMMF 0.20 f 0.06b3C 0.12 f 0.06d 0.54 f 0.2Ot DMF" 0.57 f 0.4ge SMG 4.28 f 1.359 3.92 f 2.11h 0.34 f 0.089 NMFi 0.24 0.17L 1.25 SMG 2.52 f 0.34 HMMFf 0.016 f 0.009 0.033 Substrate concentration range: 0.02-5 mM. * Values are the mean f SD. Details of metabolite analysis and calculation of kinetic constants are described under Experimental Procedures. Number of separate experiments: 12. Number of separate experiments: 4. Individual values: 0.19, 0.14, 0.06, 0.09 mM. e Number of separate experiments: 4. Individual values: 0.23, 1.25, 0.61, 0.19 nmol of HMMF.(mg of microsomal protein)-'.min-'. f Substrate concentration range: 0.4-10 mM. g Number of separate experiments: 10. h Number of separate experiments: 4. Individual values: 6.73, 3.78, 3.55, 1.63 mM. Number of separate experiments: 4. Individual values: 0.10,0.46,0.28,0.09 nmol of SMG.(mg of microsomal protein)-l..min-'. Number of individual experiments: 4.

*

characterized by an apparent K , of 0.20 mM (Table I). Extrapolation of the low-affinity part of the plot furnishes an apparent K , of 23.4 f 3.1 mM and an apparent V,, of 1.32 f 0.34 nmol of HMMF.(mg of microsomal protein)-l-min-l (mean f SD of n = 4). The apparent K , value, 0.12 mM, determined in microsomes from 4 human livers for the DMF oxidation reaction with high affinity for the enzyme, is similar to the value observed in rat microsomes (Table I). The Eadie-Hofstee plot depicting the metabolic generation of SMG from NMF in rat liver microsomes is shown in Figure 9. The apparent K , values characteristic for the high-affinity reaction in rat (4.28mM) and human liver microsomes (3.92 mM) are similar (Table I). The data shown in Figure 9 hints at the possibility that another segment indicative of an even higher affinity than that identified by the K , of 4.28 mM can be observed at NMF concentrations