Chem. Res. Toxicol. 1993,6, 578-583
578
Reaction of (E,@-Muconaldehyde and Its Aldehydic Metabolites, (&E) -6-OxohexadienoicAcid and (E,@-6-Hydroxyhexa-2,4-dienal, with Glutathione Stanley A. Kline,* Qian Xiang, Bernard D. Goldstein, and Gisela Witz UMDNJ-Robert Wood Johnson Medical School, Department of Environmental and Community Medicine, Piscataway, New Jersey 08854, a n d Environmental a n d Occupational Health Sciences Institute, Piscataway, New Jersey 08855 Received February 9, 1993
(E,E)-Muconaldehyde (muconaldehyde) has been identified as a hematotoxic metabolite of benzene in vitro. It is metabolized in mouse liver cytosol to oxidized and reduced derivatives including the a,P-unsaturated aldehydes (E,E)-6-oxohexadienoic acid and (E,E)-&hydroxy2,4-hexadienal. In this study we have synthesized the aldehydic metabolites of muconaldehyde. The reaction of glutathione with muconaldehyde and its aldehyde metabolites was investigated. Reactions were bimolecular and stoichiometric in aldehyde and glutathione in the initial phases. Second-order rate constants were determined, and the rates were in the order muconaldehyde > (E,E)-6-hydroxy-2,4-hexadienal> (E,E)-6-oxohexadienoic acid. Further investigation of the reaction of muconaldehyde with glutathione showed that the bimolecular reaction is reversible but the initial product decomposed irreversibly to two or more products, one of which had a red-shifted UV spectrum. Rate constants for these subsequent reactions were determined. The results are discussed in terms of the toxicity of muconaldehyde at tissues distal from the liver, where it is believed to be formed from the metabolism of benzene.
Introduction Benzene is a hematotoxin (I) and human leukemogen (2). Its toxicity is believed to be mediated by metabolites
formed in the liver (3). Metabolism of benzene results in the formation of ring-hydroxylated as well as ring-opened products, primarily truns,truns-muconicacid (4). We have hypothesized that reactive ring-opened intermediates play an important role the toxicity of benzene (5). Our laboratory has identified one such compound, the a,& unsaturated dialdehyde (E,E)-muconaldehyde (8) (muconaldehyde),l in mouse liver microsomal incubations of benzene (6). It has been suggested that @,E)-muconaldehyde is itself derived from the corresponding cis,cisisomer (7).Muconaldehyde is hematotoxicwhen injected in mice (8). In vitro, it is rapidly metabolized, first to the carboxylic acid/aldehyde (E,E)-6-oxohexadienoicacid (7) and the alcohol/aldehyde (E,E)-6-hydroxy-2,4-hexadienal (9),and ultimately to (E,E)-6-hydroxy-2,4-hexadienoic acid and truns,trum-muconic acid (9, 10). In this study we have examined the reactions of muconaldehyde and its metabolites, (E,E)-6-oxohexadienoicacid (7) and (E,E)6-hydroxy-2,4-hexadienal(9), with the biologically important cellular nucleophile glutathione.
Experimental Section Instrumentation. NMR spectra were determined with a Varian Model VXR 4000spectrometer. UV spectra were recorded on a Perkin Elmer Model 552 vis/UV spectrophotometer. Kinetic experiments were performed using a Perkin Elmer X 3B vis/UV spectrophotometer. Electron impact mass spectrometry (EI-
* To whom correspondence should be addressed at the Toxicology Division, Environmental and Occupational Health Sciences Institute, 681 Frelinghuyaen Rd., P.O. Box 1179,Piscataway, NJ 08855-1179; 908932-0233.
MS) was performed on a Finnigan MAT Model 8230 mass spectrometer. Negative-ion discharge mode mass spectrometry was performed using a Vestec Model 201 mass spectrometer. Reverse-phase HPLC was carried out using a Hibar Cl8 column (EM Science, Darmstadt, FRG), 0.46 cm X 25 cm. The HPLC system consisted of a Kratos Spectroflow 400 pump, Spectroflow 430 gradient maker, and Spectroflow 773 variable-wavelength detector along with a Shimadzu Model C-R3A Chrompac integrator. Chemicals. Reduced GSH and 5,5’-dithiobis(2-nitrobenzoic acid) (DTNB)were obtained from Sigma ChemicalCo. (St. Louis, MO). (E$)-Muconaldehyde (8) was custom-synthesized by Calbiochem (San Diego, CA) using a procedure of Kossmehl and Bohn (11) previously used in our laboratory (5). A simpler synthesis of (&E)-muconaldehyde and its 2,Z-isomer has been described by Golding et al. (7). @,E)-Muconaldehyde-1,6-W was custom-synthesized by Moravek Biochemical (Brea, CA). This compound was determined to be >97% radiochemically pure by reverse-phase HPLC. Chemical Synthesis. Crotonaldehyde diethyl acetal (2) was prepared from crotonaldehyde (1) and tetraethoxysilane (12). Ethyl glyoxylate (4) was synthesized by periodic acid oxidation of diethyl tartrate (13). (A) (E)-1-Ethoxybutadiene (3) was prepared using a modification of a method of Nazarov et al. (14). A stirred mixture of2 (71.5g,0.5mol)and2.0gofa”oniumdihydrogenphosphate was heated in a flask equipped with a 10-cm Vigreaux column and distillation condenser. The reaction bath temperature started at 160 OC and was allowed to rise to 180 “C during the course of the reaction. During the reaction, the crude product was collected by distillation over a temperature range of 110-117 “C. The distillation temperature was not allowed to rise above 120 “C. The distillate was washed once with water and once with 5% sodium bicarbonate and dried over anhydrous potassium carbonate. The pale yellow oil was redistilled under reduced pressure, affording 17 g (35%)of a clear mobile liquid bp 94-98 1 Abbreviations: DTNB, 5,6’-dithiobis(2-nito~~oic acid); ELMS, electron impact mass spectrometry; muconaldehyde, (E,E)-muconalde-
hyde.
0 1993 American Chemical Society 0893-228~/93/2706-0578$04.00/0
GSH Reactivity of Muconaldehyde a n d Metabolites
Chem. Res. Toxicol., Vol. 6, No. 4, 1993 579
OC/l60 mmHg [lit. bp 65-67 OC/150 mmHg (14)l; lH NMR (CDCls) 6 1.29 (t, 3H), 3.80 (q, 2H), 4.80 (d, lH), 4.98 (d, lH), 230 5.56 (dd, lH), 6.21 (ddd, lH), 6.56 (d, 1H);UV (ethanol) A, nm (e = 26 OOO). (B)2-Ethoxy-6-(ethoxycarbonyl)-6,6-dihydro-2H-pyran (5) was prepared according to Shavrygina and Makin (15).3 (18.8 g, 0.194 mol) and 19.8 g (0.194 mol) of 4 were sealed in two Pyrex test tubes under an argon atmosphere and heated at 128130 OC for 4 h. The resulting oil was distilled under reduced pressure, and 85.1 g (43%) of a pale yellow oil boiling at 96-98 OC/O.45 mmHg was collected [lit. bp 79-80 OC/1.5 mmHg (1511. (C)(E)-2-Hydroxy-6-oxohex-4-enoic Acid (6). 5 (3.0 g, 15 mmol) was dissolved in 3 mL of acetone and 3 mL of water, and 300 mg of Dowex AGX-8-50 resin (H+ form) was added. The mixture was heated for 7-8 h at 90 OC with rapid stirring under an argon atmosphere. The reaction was terminated when the UV absorbance at 220 nm showed no further increase in optical density. After filtration of the resin, the water was removed by lyophilization. After lyophilization about 80% of the UV absorption at A, = 220 nm was lost, presumably due to internal hemiacetal formation. This did not affect the subsequent reaction, and the orange semisolid was used without further purification. (D) (E,E)-6-Oxohexadienoic Acid (7). Compound 6 (1.1g, 7.5 mmol) dissolved in 50 mL of glacial acetic acid was heated at reflux until the UV absorbance appearing at 265 nm showed no further increase in optical density (about 13 h). The acetic acid was removed under reduced pressure, and the residue was recrystallized from hot ethyl acetate, yielding 760 mg (80%) of pale yellow crystals: mp 191-192 OC [lit. mp 192-193 OC (IS)]; UV (ethanol) A, 265 nm (e = 22 800) [lit. A, 265 nm (e = 25 800) (IS)]; 'H NMR (CDC13) 6 6.32 (d, lH, Hs, 523 = 15.4 Hz), 6.50 (dd, lH, Hs, J- = 7.7, JM= 15.7 Hz), 7.20 (dd, lH, Hd, Jw = 11.1,JM= 15.7 Hz), 7.50 (dd, lH, H3, Jw = 11.1Hz, JM 15.4 Hz), 9.70 (d, lH, He, J- = 7.7 Hz); positive-ion MS ( m l ) / e = 127. (E)(E,E)-6-Hydroxy-2,4-hexadienal (9). (E,E)-Muconaldehyde (8) (0.500 g, 4.46 mmol) was dissolved in 4 mL of ethanol and further diluted with 4 mL of water. After the solution was deoxygenated with an argon stream, 2.675 mL (1.34 mmol) of 0.50 M aqueous sodium borohydride (freshly dissolved) was added. The solution was stirred under an argon atmosphere for exactly 5 min, after which it was acidified with 112 pL of concentrated HC1. Analysis of the mixture by Cla reverse-phase HPLC (methanovwater = 1090,flow rate = l.OmL/min, analysis at X = 270 nm) revealed a ratio of 8 (retention time = 7.7 min) to 9 (retention time = 10.3 min) = 1.25. Additional sodium borohydride solution (0.5 M, freshly dissolved), equivalent to 0.45 mmol, was added, and the mixture was stirred under argon for 3 min, after which it was rapidly acidified with minimum concentrated HC1. The mixture was analyzed by HPLC as described above. If necessary, addition of 0.45 mol of borohydride and reaction for 3 min under argon followed by acidification were repeated until HPLC analysis showed a peak area ratio of 923 > 5. UV spectroscopy indicated little overreduction to trans,trans-2,4-hexadiene-l,6-diol (A, = 227 nm). The reaction was extracted with ethyl acetate, dried (MgSOd), and concentrated, yielding 0.345 g of crude product, which was purified by flash chromatography on silica gel (30-60 "C pet ether/ethyl acetate = 6040) and recrystallized from ether/pet ether, yielding 0.126 g (25%) of 9 as pale yellow needles: mp 44-45.5 OC. [lit. mp 268 nm (e = 22 800);1H NMR 39-40 OC (IO)]; UV (ethanol) A, (CDCla) S 4.53 (d, 2H, He, Ju = 4.9 Hz), 6.18 (dd, lH, H2, Jz-1 8.5, Ju = 15.7 Hz), 6.36 (dt, lH, Hs, Ju = 15.3,J u = 4.9 Hz), 6.58 (dd, lH, H4, Ju = 11.1Hz, JM= 15.3 Hz), 7.14 (dd, lH, 11.1Hz), 9.58 (d, lH, Hi, J 1 - 2 = 8.5 Hz); Ha, J s . 2 15.7, JEI-MS m/z = 112. Kinetics Procedures. Concentrations of unreacted aldehydes 7-9 were determined by measurement of optical density at the appropriate A, using calculated molar extinction coefficients. eM of 8 was determined to be 31 500 at 277 nm. Unreacted GSH (measured as reduced thiol) was determined by reaction with
+
DTNB (0.5 mg/mL in 0.1 M potassium phosphate, pH 7.4) and measurement of optical density at 420 nm after 1 min at room temperature. Reduced GSH concentration was read from a standard curve of absorbance a t 420 nm us GSH concentration. Where necessary, solutionswere diluted in 0.1 M phosphate buffer (pH 7.4) to achieve a concentration of (1-5) X 1odM in aldehyde or glutathione prior to analysis. Where analytical solutions to the rate equations could not be obtained, rate constants were determined using the Simulsolv program with a Gear algorithm (Dow Chemical Co., Midland, MI) for integration of stiff differential equations. (A) Addition of Glutathione to Aldehydes 7-9. For the reaction of 8 with glutathione, reactant solutions consisting of 2.0 X lod M 8 and 6.0 X lod M glutathione in 50 mM sodium phosphate (pH 7.4) were freshly prepared and deoxygenated by purging with an argon stream for 15 min. For the reaction of 7 and 9 with glutathione, reactant solutions consisted of 2.0 X 10-9 M aldehyde and 6.0 X 10-9 M glutathione. An aliquot of each solution was removed prior to reaction for determination of aldehyde and reduced glutathione concentration as described above. For kinetic determinations, equal volumes of aldehyde and glutathione solutions were mixed and the reactions were repurged with a 1min argon stream, stoppered, and incubated at 37 "C at pH 7.4. Aliquots (1mL) were removed at various times over 3 h. Controls consisted of either aldehyde or glutathione only in phosphate buffer. Second-order rate constants were determined from the slope of ln([Aldlt[GSH1o[Ald]o[GSH],) us time in the linear portion of the curve, where [AldIt and [GSH], were concentrations of aldehyde and glutathione at time = t, and [Aldlo and [GSHIo were the respective initial concentrations. Reported values are the mean of two experiments. In order to measure the rate of addition of a second molecule of glutathione to muconaldehyde, 0.10 M 8 (100 pL) was added to a deoxygenated solution consisting of 0.028 M glutathione in 0.1 M phosphate (pH 7.4) (900 pL). A control consisting of 0.025 M glutathione in buffer was used to monitor glutathione oxidation. Aliquots were removed over 20 min, and reduced glutathione was determined. (B)Reactionof Muconaldehydewith theDTNBReaction Product. A 5.0 X 1WM solution of 2-carboxy-4-nitrothiophenol (10) was prepared from glutathione (5.0 X lW M) and DTNB (1X 10-9 M) in 50 mM phosphate buffer (pH 7.4). Two hundred microliters of 5.0 X 1W M 10 was added to 800 pL of a deoxygenated solution of 8 (6.25 X 10-9 M) in 50 mM phosphate buffer (pH 7.4), and the optical density at 420 nm was recorded at room temperature over 1 h. The pseudo-first-order rate constant for the addition of 10 to 8 was calculated from the initial linear portion of the slope of ln(At/Ao)us time, where At and A0 are optical densities at 420 nm at times t and 0, respectively. (C)Reverse Reaction and Formation of Secondary Reaction Products of Glutathione and Muconaldehyde. One hundred microliters of 8 (0.1 M in ethanol) was added to 900 pL of a deoxygenated solution of glutathione (0.0133 M) in 50 mM phosphate, pH 7.4. After reaction for exactly 1 min at room temperature, 10 pL of the reaction was diluted into a quartz cuvette containing 990 FL of 5 X 10-9 M DTNB in 50 mM phosphate (pH 7.4). Optical density at 420 nm was recorded over 4 h at room temperature. Reaction of [W]Muconaldehyde with Glutathione. One hundred microliters of [14C]muconaldehyde(0.01 M in ethanol, 58 mCi/mmol) was added to 900 pL of a deoxygenated solution of glutathione (1.33 X 10-9 M) in 50 mM phosphate (pH 7.4) and incubated at room temperature for 3 h. Eighty microliters of the reaction mixture was chromatographed on a Claanalyticalcolumn (methanol/l% acetic acid = 2080, flow rate = 1.0 mL/min, analysis at A = 340 nm). Fractions were collected every 30 s over 20 min, and radioactivity was determined by scintillation counting. Eighty-four percent of injected counts were recovered in all fractions.
580 Chem. Res. Toxicol., Vol. 6,No. 4, 1993
Kline et al.
Scheme I. Synthetic Pathways for (E,E)-6-Oxohexadienoic Acid (7) and (E,E)-6-Hydroxy-2,4-hexadienal(10)
1
P
Q
Table I. Reactivity of a,@-UnsaturatedAldehydes toward Glutathione second-order R.: CHO
WH
(M-W) 11.2 f 1.6
half-life in the presence of 10 mM glutathione 6.2 s
H
8
*
-HZ
H
0.231 f 0.009
5.0 min
0.0979 f 0.018
11.8 min
H
10 -O@+H H
7 a
37
OC,
pH 7.4. Average of two determinations.
Results Syntheses of Muconaldehyde Analogs. The syntheses of the carboxylic acidlaldehyde (7) and the alcohol1 aldehyde (9) are summarized in Scheme I. Compound 7 was prepared using amodification of a procedure of Makin et al. (15). The starting materials for this synthesis were (E )-l-ethoxybutadiene (3) and ethylglyoxylate (4). Compound 4 was conveniently prepared by oxidation of diethyl tartrate. Compound 3 was synthesized by a liquid-phase pyrolysis of acetal 2. The choice of the acid catalyst was important in this reaction. Attempts to use p-toluenesulfonic acid instead of ammonium phosphate resulted in rapidly exothermic polymerization. Dihydropyran 5 was synthesized by a Diels-Alder reaction of 3 and 4 (15). Hydrolysis of 5 to the unsaturated aldehyde 6 was catalyzed by cation-exchange resin in the H+ form. Upon removal of water, 6 cyclizedto a hemiacetal. Attempts to dehydrate 6 by reaction with p-toluenesulfonic acid in refluxing benzene as reported in the literature (15)did not yield 7 in our hands. Rather, another product with an UV ,A, = 253 nm resulted. This compound was not further characterized. Compound 6 was effectively dehydrated to 7 by heating in refluxing glacial acetic acid. NMR of 7 confirmed the trans stereochemistry of both dienes and
was in complete agreement with literature values (15). The alcohollaldehyde 9 was prepared from 8 by controlled reduction with sodium borohydride at room temperature. The reaction was controlled by successive addition of limited borohydride followed by a short reaction and subsequent acidification to stabilize the unsaturated aldehyde and also to decompose unreacted reagent. Under these conditions, minimal overreduction to the l,6-diol occurred. Compound 9 was recently reported by Goon et al. (9),who prepared it from @,E)2,4-hexadiene-1,6-diol by oxidation using pyridinium chlorochromate. Attempts to repeat this oxidation in our hands resulted only in production of 8. However, the NMR for compound 9 was essentially identical to that reported by Goon et al. (9). Reaction of Aldehydes with Glutathione. Secondorder forward rate constants for the reaction of aldehydes 7-9 with GSH were determined at 37 "C (pH 7.4). Initial reactant concentrations were adjusted such that 60% of aldehyde was reacted within 2-4 h. Loss of GSH thiol due to oxidation was determined to be less than 3% during the period of measurement. All reactions exhibited second order kinetics up to 30 min. Stoichiometry was 1:l in aldehyde and glutathione thiol consumption throughout the course of the reaction. Bimolecular rate constants, kl, for the addition reaction were calculated from the slopes of the linear portions of the kinetic plots. These data are presented in Table I. In order to determine whether muconaldehyde might react with two molecules of glutathione, initial concentrations of 8 and GSH were increased 1000-fold to 1.0 X 1k2and 2.5 X 1k2M, respectively. Under these conditions, addition of a second glutathione molecule might be expected to compete with the unimolecular rearrangement of the initialmono adduct (see below). Nevertheless, the ratio of moles thiol consumed to moles of muconaldehyde consumed (determined at various times over 20 min) was 1.05 f 0.11 (data not shown). Reaction of Muconaldehyde with Glutathione. In the reaction of muconaldehyde (8) with glutathione, disappearance of the reactants was followed by a slower appearance of a new UV absorbance at ,A, = 340 nm
Chem. Res. Toxicol., Vol. 6, No. 4,1993 581
GSH Reactiuity of Muconaldehyde and Metabolites 1.51
i
A
A
E
0
Y
8C m
e
k.,
8 n
k d r 5 . 2 X 1O"S"
a
O Y 0
200
300
250
350
400
450
nm Figure 1. Time course of the reaction of muconaldehyde and glutathione at pH 7.4. Muconaldehyde (4.0 X 1od M) and glutathione (1.2 X 1o-L M) in 0.1 M phosphate buffer (pH 7.4) were incubated at 37 O C . UV spectra were recorded at various times over 3 h. (A) 0 min; (B) 5 min; (C) 10 min; (D) 30 min; (E) 1 h; (F)3 h.
2
4
6
r
1.2 x 1 0 ' 4 ~ "
,
8
I
1 0 1 2 1 4
Time ( 5 ) x 10 Figure 3. Comparison of data and best fit numerical solutions of kinetic Scheme 111. Muconaldehyde (1.0 X leaM) and glutathione (1.2 X le2M) in 0.1 M phosphate buffer (pH 7.4) were incubated at 25 O C for 1 min at room temperature after which the reaction was diluted 1:lOO in 5 X 10-9M DTNB in 0.1 M phosphate buffer (pH, 7.4). Optical density was recorded at 420nm. A numerical solutionto the kinetic equations of Scheme I11 was computed and optimized to the kinetic data. (Solid line)
Experimentaldata; (Opencircles) Optimized numerical solution. Scheme 111. Decomposition of the Initial Muconaldehyde-Glutathione Adduct in the Presence of DTNB
Time (min)
Figure 2. (218-Reverse phase chromatogram of the reaction of 14C-muconaldehydeand glutathione at pH 7.4. 14C-Muconaldehyde (1.0 X 10-8M) and glutathione (1.2 X 10-9 M) in 0.1 M phosphate buffer (pH 7.4) were incubated at 25 O C for 3 h. Eighty pL of the reaction mixture were chromatographed on a reverse phase column and 0.5 mL fractions were collected and counted. UV absorption at 340 nm (solid line); DPM (open circles). Scheme 11. Pathway of the Reaction of (E,E)-Muconaldehyde with Glutathione at pH 7.4 8
+
GSH
7 [PI1 kl
k t
P, + P, + ... + Pn
(Figure 1). In order to determine the proportion of the reaction mixture represented by this product, excess glutathione was reacted with [14C]muconaldehyde. Reverse-phase chromatography revealed a single compound which absorbed at 340nm. Radioactivity recovered under the UV absorbing HPLC peak represented only 5% of recovered counts, with the majority of counts eluting in the void volume (Figure 2). Reinjection and recovery of a portion of the UV-absorbing peak fractions indicated that less than 20% of the compound had decomposed during the chromatography. Kinetics of Reaction of Muconaldehyde with Glutathione. The reaction was assumed to follow the pathway shown in Scheme 11. In this scheme, k d is the sum of all rate constants for decomposition of PIto products ( k d = kz + k3 + ...k,). In order to determine the rate constants, k-1 and k d , the initially formed adduct, PI,was rapidly generated (
