JANUARYIFEBRUARY 1994 VOLUME 7, NUMBER 1 0 Copyright 1994 by the American Chemical Society
Articles Metabolic Pathways of 1-Butyl [3-13C]Acrylate. Identification of Urinary Metabolites in Rat Using Nuclear Magnetic Resonance and Mass Spectroscopy Igor Linhart,*lt Richard Hraba1,J Jaroslav Smejkal,? and Jii3 Miteras National Institute of Public Health, Centre of Industrial Hygiene and Occupational Diseases, Srob6roua 48, 100 42 Prague, Czech Republic, Institute of Chemical Technology, Laboratory of Nuclear Magnetic Resonance Spectroscopy, Technick6 1905,166 28 Prague, Czech Republic, and National Institute of Public Health, Centre of Health and Living Conditions, Srob6rova 48, 100 42 Prague, Czech Republic Received August 2, 1993" 1-Butyl acrylate, an industrial monomer, is rapidly metabolized by carboxylesterase-catalyzed hydrolysis to acrylic acid and 1-butanol. Acrylic acid enters the intermediary metabolism and is efficiently degraded to carbon dioxide as the metabolic end product. To obtain a virtually complete metabolic pattern, rats were dosed by a single intraperitoneal dose of 1 mmol/kg 1-butyl [3-13C] acrylate. The urine was then analyzed by a one-dimensional 'H-detected and two-dimensional lH-13C shift-correlated heteronuclear multiple-quantum NMR experiment. In this experiment, three urinary metabolites, namely, 3-hydroxypropanoic acid, N-acetyl-S(2-~arboxyethyl)cysteine, and N-acetyl-S-(2-carboxyethyl)cysteinesulfoxide, were identified by comparing their 'H and 13C chemical shifts with those of authentic standards. In another experiment, to enhance minor metabolic pathways, rats were dosed with 0.25 mmol/kg of a carboxylesterase inhibitor, tri-o-tolyl phoshpate, prior to 0.5 mmol/kg butyl [3-l3C1acrylate. Under these conditions,N-acetyl-S-(2-carboxyethyl)cysteine,N-acetyl-S-[2-(butoxycarbonyl)ethyllcysteine, and N-acetyl-S-(2-carboxyethy1)cysteinesulfoxide were found in urine. No metabolites which would arise from a possible metabolic activation of 1-butyl acrylate to 1-butyl oxiranecarboxylate and its subsequent hydrolysis or glutathione conjugation were found. It is estimated that any metabolite amounting to more than 1% of the dose should be detected under these conditions. To study the routes by which BA enters the intermediary metabolism, incorporation of the label into urinary carboxylic acids was followed by GUMS. Significant enrichment was found in 3-hydroxypropanoic acid and citric and isocitric acid but not in lactic acid. These results confirm that the main metabolic pathway of butyl acrylate is its carboxylesterase-catalyzed hydrolysis. The resulting acrylic acid enters the intermediary metabolism via a minor pathway of propanoic acid catabolism and tricarboxylate cycle. Glutathione conjugation leading to mercapturic acids is a minor pathway of butyl acrylate metabolism. t
t
Centre of Industrial Hygiene and Occupational Diseases. Laboratory of Nuclear Magnetic Resonance Spectroscopy.
I Centre of Health and Living Conditions.
Abstract published in Advance ACS Abstracts, November 15,1993.
0893-228x/94/2707-0001$04.50/00 1994 American Chemical Society
2 Chem. Res. Toxicol., Vol. 7, No. 1, 1994
Introduction l-Butyl acrylate (BA)l is an industrial monomer used along with other acrylic acid (AA) derivatives in the production of polymers for surface coatings, adhesive and sealant emulsions, emulsion polishes, and acrylic elastomers. BA bears an electrophilic olefinic group in its molecule which may react with nucleophilic sites of biological molecules such as glutathione (1-4) and protein thiol groups (4,5). Moreover, it might be metabolically activated by monooxygenases to the corresponding epoxide, oxiranecarboxylicacid butylester (butyl glycidate, BG). Formation of epoxides during biotransformation was demonstrated for analogous compounds, namely, acrylonitrile (6-8) and acrylamide (9), and is considered to be closely linked with toxic effects such as mutagenity and carcinogenity (10). However, this type of metabolic activation has not been reported for AA esters as yet. Unlike acrylonitrile and acrylamide, AA esters are rapidly enzymatically hydrolyzed in biological systems (4,5,1113). Hydrolytic products, AA and the corresponding alcohol, are further efficiently metabolized to carbon dioxide as the metabolic end product. At physiological pH, AA exists in its anionic form. The negatively charged carboxylic group significantly decreases the reactivity of the olefinic double bond toward nucleophiles, so that reactivity of acrylic acid with nucleophiles is very limited under physiological conditions (1, 12). Therefore, hydrolysis of acrylate esters is considered to be a detoxication pathway. Nevertheless, after high oral doses of AA a small amount of mercapturic acid was formed in rats (14). It was postulated that AA enters the intermediary metabolism via a minor pathway of propanoate catabolism and subsequent metabolization of resulting acetyl S-coenzyme A in the tricarboxylate cycle (14, 15). However, a very recent study comparing the metabolism of carbon-13labeled acrylic acid and propanoic acid showed that these compounds differ in their metabolism significantly (14). Prior to the carboxylesterase-catalyzedhydrolysis, acrylic acid ester may undergo various metabolic reactions. Among them, metabolic activation to electrophilic oxirane carboxylates is of major concern. In experiments with radiolabeled BA three urinary metabolites were detected (3). Two of them were identified as N-acetyl-S-(2carboxyethy1)cysteine (ACEC) and N-acetyW(2-carboxyethy1)cysteine sulfoxide (ACECO); the third one remained unknown. In experiments with AA (14) and ethyl acrylate (15), 3-hydroxypropanoic acid (HPA) was identified as the only metabolic intermediate on the route from AA to carbon dioxide. The aim of present study is to obtain more detailed information on biotransformation of BA including the routes by which BA enters the intermediary metabolism. Investigations on biotransformation pathways of small molecules which enter the intermediary metabolism like acrylate esters present a 1 Abbreviations: BA, l-butyl acrylate; AA, acrylic acid; BG, l-butyl glycidate; HPA, 3-hydroxypropanoic acid; ACEC, N-acetyl-S-(2-carboxyethy1)cysteine;ABEC, N-acetyl-S-[2-(butoxycarbonyl)ethyllcysteine; ACECO, N-acetyl-S-(2-carboxyethyl)cysteineS-oxide; DHPA, 2,3-dihydroxypropanoicacid; BSTFA,N,O-bis(trimethylsily1)trifluoroacetic acid; TMCS, trimethylchlorosilane; 1-BHC, N-acetyl-S-[l-(butoxycarbony1)2-hydroxyethyllcysteine;2-BHC, N-acetyl-S-[2-(butoxycarbonyl)-2-hydroxyethyllcyeteine; 1-CHC, N-acetyl-S-(l-carboxy-2-hydroxyethyl)cysteine; 2-CHC, N-acetyl&( 2-carboxy-2-hydroxyethyl)cysteine;TOTP, tri-o-tolylphosphate; HMQC,heteronuclear multiple-quantum coherence; CIA, citric acid; ICA, isocitric acid; ABZC, N-acetyl-S-benzylcysteine; CEC, S-(2-carboxyethyl)cysteine; BEC, S-[2-(butoxycarbonyl)ethy1lcysteine.
Linhart et al.
considerable methodological challenge. Therefore, powerful tools such as one- and two-dimensional 'H-'3C-NMR techniques following the administration of 13C-enriched (8),[1,2,3-W31substrates such as [1,2,3-13C31acrylonitrile acrylamide (91,and [1,2,3-13C31acrylic acid (14)were used to obtain a virtually complete pattern of biotransformation pathways. Alternatively, specifically l3C-enriched substrates are used. In the latter case, lH-13C shift-correlated NMR spectroscopy (16, 17) or techniques of selective observation of 'H resonances from hydrogens directly bound to 13Catoms were applied (18,19). Among these techniques, proton-detected heteronuclear multiple-quantum coherence (HMQC) NMR seems to be the method of choice, if high sensitivity is required (19). Therefore, in the present study we apply the HMQC NMR technique in tandem with the administration of butyl [3J3C1acrylate. Incorporation of the label into physiological metabolites is followed by GUMS.
Experimental Section Caution. Diazomethane is explosive and carcinogenic. Sharp edges and ground glass must be avoided during its generation. All operations should be done in an efficient fume hood. Chemicals. l-Butyl [3-13C]acrylate (98% pure according to GLC) was synthesized in our laboratory (20) from [Wlmethyl iodide 99 % 13C-enriched (Aldrich, Steinheim, Germany) and diluted with unlabeled BA to 92 atom % lSC. Authentic samples of 3-hydroxypropanoic acid (21), l-butyl oxiranecarboxylate (butyl glycidate, BG) (22),ACEC (23),N-acetyl-S-[2-(butoxycarbony1)ethyllcysteine (ABEC) (23), ACECO ( 3 ) , 2,3-dihydroxypropanoic acid (DHPA) (24), and N-acetyl-S-benzylcysteine (ABZC) (25)were also prepared in our laboratory using known synthetic procedures. Dioxane was dried by refluxing with sodiumand distilled. Silylationreagents,bis(trimethylsily1)trifluoroacetamide (BSTFA) and its mixture with 1% of trimethylchlorosilane (TMCS), were purchased from Pierce (The Netherlands). Tri-o-tolyl phosphate (TOTP) was from Eastman Organic Chemicals (Rochester, NY). Solid-phase extraction plastic cartridges 20 mm by 8 mm i.d. packed with HEMA Q, a strongly basic anion exchanger with a hydrophilic methacrylate copolymer matrix, were from Tessek (Prague, Czechoslovakia). All other chemicals were of analytical or reagent grade and were obtained from commercial sources. Synthesis of Authentic Standards. Mercapturic acids derived from BG were prepared by the addition of N-acetylcysteine to BG. To a stirred solution of 160 mg (1 mmol) of N-acetylcysteine and 170mg (2 mmol) of sodium bicarbonate in 10 mL of water was added 144mg (1mmol) of BG. The reaction mixture was stirred under nitrogen at room temperature for 7 h. Hydrochloric acid (diluted 1:4) was then added to pH 2, the solution was saturated with ammonium sulfate, and products were extracted into ethyl acetate. Drying with magnesiumsulfate followed by evaporation of ethyl acetate in vacuo yielded 250 mg (88%)ofthe crude product, whichwasthen fractionated by HPLC on a 8 X 250 mm steel column packed with Separon SGX C18, particle size 7 pm (Tessek, Prague, Czechoslovakia), using methanol-0.5% acetic acid (54:48) as a mobile phase at a flow rate of 2.5 mL/min. Two peaks absorbing at 215 nm were separated. Fractions were evaporated to dryness and characterized by NMR and mass spectra: N-Acetyl-S-[ l-(butoxycarbonyl)-2-hydroxyethyl]cysteine (1-BHC): 1H-NMR (acetone-ds, 6 in ppm): n-CdHe, 0.93,1.41,1.64, and 4.15; CHsCO, 2.00 (8); CHaS, 3.20 (2 dd, 13.8 and 4.7 Hz), 3.04 (dd, 13.8 and 7.2 Hz),and 2.99 (dd, 13.8 and 8 Hz); SCHzCH,4.41 (m);SCH, 3.50 (dd, 8.9 and 5.7 Hz) and 3.55 (dd, 8.7 and 5.6 Hz); C&OH, 3.85 (m), 3.74 (dd, 11.0 and
Chem. Res. Toxicol., Vol. 7, No.1, 1994 3
Metabolic Pathways of Butyl Acrylate Table 1. 1H and
1 F NMR
Chemical Shifts of the Methylene or Methyl Group in the Potential Metabolites of 1-Butyl [3-1F]Acrylate (DaO,pD = 7) chemical shifts (6 in ppm) compound 'H '3C 3.75 (t) 61.5 3-hydroxypropanoic acid (HPA) 31.0 2.74 (t) N-acetyl-S-(2-~arboxyethyl)cysteine (ACEC) 32.4 2.84 (t) N-acetyl-S-[2-(butoxycarbonyl)ethy1lcysteine (ABEC) 49.7 3.08 (m) N-acetyl-S-(2-carboxyethy1)cysteine S-oxide (ACECO) 65.3 3.54 and 3.65 (d) 2,3-dihydroxypropanoic acid (DHPA) 62.9 3.82 (m) N-acetyl-S-[l-(butoxycarbonyl)-2-hydroxyethyl]cysteine(1-BHC) 63.8 and 64.0 3.72 and 3.83 (m) N-acetyl-S-(l-carboxy-2-hydroxyethyl)cysteine (1-CHC) 37.2 and 37.4 N-acetyl-S-[2-(butoxycarbonyl)-2-hydroxyethyllcysteine(2-BHC) 2.90 and 3.14 (m) 38.5 and 38.7 2.92 and 3.05 (m) N-acetyl-S-(2-carboxy-2-hydroxyethyl)cysteine (2-CHC) . . 2.80 (t) 30.9 S-(2-carboxyethyl)cysteine(CEC) 2.86 (t) 32.2 S-[2-(butoxycarbonyl)ethyllcysteine (BEC) 1.24 (d) 20.9 lactic acid 2.36 and 1.47 (8) 29.2 pyruvic acid 3.17 37.6 @-alanine
5.6 Hz), and 3.77 (dd, 11.1and 5.7 Hz). W-NMR (acetone-de, 6 inppm): n-C,Hs, 14.0,22.6,31.3, and65.4;CH&O, 29.7; CHzS, 33.9 and 34.1; CHS, 50.1 and 49.8; CHNH, 53.3 and 52.8; CH2OH, 62.9; CO, 170.9, 171.0, 172.0, and 172.1. N-Acetyl-S-[2-(butoxycarbonyl)-2-hydroxyethyl]cysteine (2-BHC): 1H-NMR (acetone-de, 6 in ppm): n-C4Hs, 0.93, 1.40, 1.64, and 4.13; CHsCO, 2.00 ( 8 ) ; CH2S, 2.85-3.04 (m, 3 H), 3.14 (dd, 13.9 and 4.9 Hz), and 3.17 (dd, 13.9 and 4.9 Hz); CgNH, 4.41 (m); CHOH, 4.69 (m). W-NMR (acetone-& 6 in ppm): n-C4H9,13.9,19.7,31.4, and 65.4; CHaCO, 29.5; cysteineCHz, 35.4 and 35.5;CHzS, 37.2 and 37.4;CHNH,53.3 and 53.5; CHOH, 72.1; CO, 170.9, 172.3, 172.4, and 173.3. FAB MS spectra of both 1-BHC and 2-BHC showed intense (M H)+ and (M Na)+ ions at m/z 308 and 330, respectively. The crude product containing 1-BHC and 2-BHC was used for preparation of corresponding free mercapturic acids, i.e., N-acetyl-S-(l-carboxy-2-hydroxyethyl)cysteine(1-CHC) and N-acetyl-S-(2-carboxy-2-hydroxyethyl)cysteine (2-CHC). The hydrolysis was accomplished by stirring the mixture of 70 mg (0.23 mmol) of the adduct with 26 mg (0.46 mmol) of potassium hydroxide in 3 mL of ethanol-water (51) for 1.5 h at room temperature. After acidification by diluted hydrochloric acid to pH 2, the solution was evaporated to dryness. The residue was redissolved in acetone, filtered, and evaporated to dryness in vacuo. The oily residue was dissolved in DzO to measure NMR spectra. Relevant chemical shifts are shown in Table 1. S-(2-Carboxyethyl)cysteine(CEC). To the solution of 1.75 g (10 mmol) of cysteine hydrochloride monohydrate in 5 mL of water purged with nitrogen was added 10 mL of 20 % ammonium hydroxide, followed by 0.8 g of acrylic acid (11mmol) in 2.5 mL of water. The reaction mixture was stirred for 7 h under nitrogen at room temperature and allowed to stand overnight. Excess ammonia was evaporated in vacuo, the residue was diluted to a volume of 18mL, and the pH was adjusted to 3 by hydrochloric acid. White crystals decomposing at 224-226 "C were filtered off and dried at 35 OC in vacuo. 'H-NMR (DzO, pD 7,6 in ppm): CH&O,2.51 (m); SC&CH2,2.80 (t, J = 7.0 Hz); SC&CH, 3.01 ( d d , J = 14.8and8.1Hz)and3.15(dd,J= 14.8and4.1Hz);CH, 3.93 (dd, J = 8.1 and 4.1 Hz). W-NMR (D20, pD 7,6 in ppm): CH~CHZS, 39.7; CHZCH~S, 30.9; Cys-CHz, 35.1; Cys-CH, 56.5; CO, 171.6. FAB MS (hydrochloride): (M + H)+ 194. S-[2-(Butoxycarbonyl)ethyl]cysteine.To the solution of 1.75 g (10 mmol) of cysteine hydrochloride monohydrate in 5 mL of water purged with nitrogen was added 10 mL of 20% ammonium hydroxide, followed by 1.4 g (11mmol) of BA. The reaction mixture was stirred 3 h at room temperature, excess ammonia evaporated in vacuo, and the pH adjusted to 6 by hydrochloric acid. Crystals were filtered off and purified by column chromatographyon silicagel using chloroform-methanolacetic acid (7:2:1) as eluent: mp 173-174 OC. 'H-NMR (DzO, pD 7, 6 in ppm): n-CdH9, 0.90 (t),1.36 (hex), 1.63 (p), and 4.15 (t); SCH2CH2, 2.73 (t, J = 6.7 Hz); SC&CH2, 2.86 (t, J = 6.9 Hz); Cys-CHz, 3.03 (dd, J = 14.7 and 7.4 Hz) and 3.11 (dd, J = 14.7 and 3.9 Hz); Cys-CH, 3.89 (m). 13C-NMR (D20, pD 7 6 in ppm):
+
+
n-C& 15.1,20.8,28.8,and 68.0; SCH2CH2,36.5; SCHzCHz,32.2; Cys-CH2, 34.8; Cys-CH, 56.0; CO, 163 and 175.5. FAB MS (hydrochloride): (M + H)+ 250. Animals. Adult female Wistarrats (Charles River, Germany) weighing 225-265 g were placed in individual glass metabolic cages with free access to pelleted food (Velaz, Czechoslovakia) and drinking water. Animals were divided into three groups, three rats in each. The first group was dosed by a single intraperitoneal dose of 1 mmol/kg [WIBA in sunflower oil (2 mL/kg). The second group was pretreated with an intraperitoneal dose of 0.25 mmol/kg TOTP followed in 18 h by 0.5 mmol/kg [13C]BAin sunflower oil. The control group received sunflower oil only. Urine samples were collected after 20 h. Aliquots of 5 mL were taken from each individual sample, and the remaining urine was pooled within a group. Samples were frozen and stored at -20 OC until worked up. NMR Studies. NMR spectra were measured on a Bruker AM 400 spectrometer (400.13 MHz for lH and 100.62 MHz for 13C spectra, respectively) at 27 O C . For the measurement of proton-decoupled W-NMR spectra 0.5-mL aliquots of pooled urine were filtered through 0.45-pm membrane filters, the pH was adjusted to 7, and 0.1 mL of DzO was added. For other measurements, aliquots (6 mL) of pooled urine were freeze-dried and redissolved in 6 mL of D2O to exchangehydroxyl and amino protons for D. Solutions were filtered through 0.45pm membrane filters and freeze-dried again. Lyophilizates were dissolved in 1mL of deuterium oxide (100.0 atom % , Aldrich, Steinheim, Germany) and filtered, an internal standard, sodium 4,4-dimethyl-4silapentanesulfonate, was added, and the samples were sealed in NMR sample tubes. A 5-mm dual (IH, W) probehead was used for both lH and 13C experiments while for the detection of protons bound to 13Catoms an inverse broadband probehead with the same inner diameter was utilized. Resonances of protons coupled to 13Care approximately 2 orders less intensive than those of protons coupled to 12Catoms, which presents a considerable dynamic range problem. The onedimensional 'H-detected heteronuclear multiple-quantum experiment (HMQC) preceded by the bilinear (BIRD) pulse sequence alleviates this hindrance (26-28). The pulse sequence used is as follows: 90,(1H)-1/2J-180.(1H, 13C)-1/zJ-90.(1H)-r-
90~('3C)~90~('H)~'/zJ~90e('3C)~t~/~-18O~('H)-t~~~9O~('3C)-'/zJACQ('H). Phase cycling: 4 - (4x, 4 - x ) , B - 4(x, - x ) , ACQ - 2 ( x , -2, -x, x ) .
'H and 13C resonances were correlated to each other by means of a two-dimensionalversion of the above-mentioned experiment (27). Sixty-four experiments of 2048 data points each were multiplied with an unshifted sine-bell window function and zero filled to the 2048 X 2048 data matrix. To obtain lH and 13Cchemical shift values directly comparable with those of BA metabolites, as measured in the HMQC-BIRD experiments, authentic standards were dissolved in DzO. The sample solutions were then neutralized by adding sodium carbonate in D2O.
4 Chem. Res. Toxicol., Vol. 7, No. 1, 1994 GC/MS Studies. Mass spectra were measured in electron impact mode at 70 eV on a JEOL DX 303 instrument with DA 5000 data station coupled with Hewlett-Packard 5890 gas chromatograph. Carboxylic acids were isolated from urine using strong anion exchanger columns by a modified procedure of Verhaege et al. (29). In brief, 0.5-mL aliquots of pooled urine were mixed with 0.1 mL of 0.01 M barium hydroxide solution to precipitate sulfates and a part of phosphates. Samples were centrifuged at 200g, 0.2 mL of the supernatant was diluted with 0.4 mL of water, and the pH was adjusted to 7.5-8. The resulting solutions were applied to HEMAQ columns activated as described (29). Columns were then washed with 5 mL of doubly distilled water, and excess water was removed by centrifugation (5 min at 200 g), followed by washing with 1 mL of methanol and 1 mL of ethyl acetate. Acidic compounds were then eluted with 1-butanol-formic acid-concentrated hydrochloric acid (80/20/ 0.5 v/v) followed by 1mL of methanol. Combined eluates were evaporated to dryness in vacuo, and the residues were dissolved in 100 pL of dry dioxane. Aliquots of dioxane solution (30 pL) were transferred into the silylation vials, mixed with 60 pL of BSTFA, and injected into the GC. Alternatively, to enhance elution of citric acid (CIA) as well as the other polar acids which could not be eluted with the above-mentioned nonaqueous solvent mixture, water-2-propanol-formic acid-concentrated hydrochloric acid (60/20/20/0.5 v/v) was used as an eluent. In the latter case, carboxylic acids were derivatized by heating for 1 h at 60 "C with a solution of hydroxylamine hydrochloride in pyridine (4 mg/mL) prior to silylation with BSTFA-TMCS. Derivatized samples were analyzed on a 30 m by 0.32 mm i.d. fused silica capillary column coated with cross-linked methylsilicone, film thickness 0.33 pm. The flow rate of helium carrier gas was 2 mL/min. The injector port temperature was 280 "C, and the column oven temperature was held at 40 "C for 1 min with the stream splitter closed and thereafter programmed to 250 "C at 10 "C/min with the splitter open. FAB MS. Fast atom bombardment mass spectra were measured on a JEOL DX 303 instrument with a DA 5000 data station. Thioglycerol was used as a matrix and xenon as a bombardment gas. GC Analyses. Urine was worked up and analyzed for ACEC and ABEC as reported previously (2). In brief, acidified urine was extracted with ethyl acetate and derivatized with diazomethane. The extracts were analyzed on a 2 m long by 3 mm i.d. glass column packed with 3% OV 225 on Gas Chrom Q 100/120 mesh using a flame ionization detector. The flow rate of nitrogen carrier gas was 35 mL/min. Column, injector, and detector temperatures were 215, 250, and 200 "C, respectively. ABZC was used as an internal standard. TLC Analyses. Urine samples were deproteinized by addition of 2 volumes of ethanol followed by centrifugation (5 min at 3000g). Supernatants and authentic standard solutions were spotted on TLC plates coated with silicagel 60 WF 254 (E. Merck, Prague, Czechoslovakia),eluted with 1-butanol-acetic acid-water (4/1/5 v/v). Mercapturic acids and cysteine conjugates were detected by spraying with iodoplatinate reagent (30).
Results Comparing proton-decoupled 13C-NMRspectra of the urine from the exposed and control groups, tentative assignment of the metabolite signals arising from 3-1% of the acrylate molecule could be made. In rats dosed with 1mmol/kg [l3C1BA,ACEC and HPA were major metabolites, ACECO was a minor one, and ABEC could not be detected at all (Figure 1). In contrast, in TOTP-pretreated rats dosed with 0.5 mmol/kg ['WIBA, signals of S-CH2 carbons from ACEC, ABEC, and ACECO but not from CHzOH of HPA were detected (Table 1). Carbon-13 signals of specificallylabeled metabolites are not sufficient for unequivocal identification. To obtain additional characteristics, namely, the proton resonances, their
Linhart et al. "PA
90
80
70
60
50
40
30
10
P PM
Figure 1. J-modulated spin-echoexperiment for l3C coupled to 'H. Urine of rata dosed with 1mmol/kg [l3C1BA(top trace) and of control rats (lower trace). Signals of carbons CH2 and quaternary carbons are positive while signals of CHs and CH are negative.
multiplicity, and lH-13C correlation, one- and two-dimensional HMQC experiments were measured. In these experiments, signals derived from protons bound directly to the labeled carbon are measured and correlated with corresponding 13C resonances. So, selective observation of protons bound to 13Cin one-dimensional NMR showed marked differences between the spectra of BA-treated rat urine and those of control urine (Figure 2). All 13C-bound proton signals are split into doublets by a one-bond l3ClH coupling constant. They were assigned by comparing chemical shifts and multiplicity with the corresponding signals of authentic standard measured in DzO at pD 7. Since 1-CHC and 2-CHC were not obtained as individual standards, the assignment of the corresponding chemical shift values in the mixture of 1-CHC + 2-CHC was based on the similarity of their spectra with those of 1-BHC and 2-BHC. Neither the 1-butyl ester group nor the N-acetyl group markedly influences the resonances of 13C-labelderived methylenes. While slight differences in chemical shifts between free carboxylic acid and the butyl ester enabled us to distinguish between ACEC and ABEC, it was not possible to distinguish between cysteine (CEC, BEC) and N-acetylcysteine derivatives (ACEC, ABEC) solely on the basis of 'H and 13C chemical shifts (Table 1). Nevertheless, mercapturic acids ACEC and ABEC, unlike cysteine adducts CEC and BEC, are known metabolites of BA. In addition, ACEC and ABEC unlike CEC and BEC were detected in urine from exposed rats by TLC (Table 2). Major proton signals derived from [l3C1BAin the urine of the BA-treated group were assigned
Chem. Res. Toxicol., Vol. 7, No. 1, 1994 5
Metabolic Pathways of Butyl Acrylate
I
I
n
I
I
I
I
I
I
_
I
.
,
8.8 I
I
I
5!0
I
u!o
I
I
I
I
I
PPM
3!0
,
.
, , . . , . , , .
.
.lo
4.6
, . . .
, . . , A!"
sln
a;"
, . . . , . . ~ . , 2.8
Figure 3. Two-dimensional W-lH shift-correlatedNMR spectrum of a D20solution of the lyophilized rat urine after a single ip dose of [WIBA obtained using the BIRD pulse sequence to suppress signals of protons bound to other than l3C nuclei.
ACEC
I
.
70
Figure 2. 'H-NMR spectraof the urine of controlrats (top) and rata dosed with 1 mmol/kg [13C]BA (middle) and 0.5 mmol/kg [WIBA 18h after pretreatmentwith TOW (bottom). Toptraces are all lH resonances; lower traces are resonances of protons coupled to l3C atoms (BIRD-HMQC experiment). Table 2. TLC Mobilities (Rf Values) of Mercapturic Acids ACEC and ABEC and Cysteine Adducts CEC and BEC. ACEC ABEC CEC BEC 0.58 0.72 0.20 0.49 The analyses were performed on silica gel WF254 plates (Merck, Czechoslovakia) with 1-butanol-acetic acid-water (4/1/5v/v) as mobile phase.
to the 3-CHzOH protons of HPA and SCH2CH2CO of ACEC. Weak signals which could be assigned to SO-CH2CH2 protons of ACECO were also detected. To potentiate minor metabolic pathways of BA, the main pathway, i.e., carboxylesterase-catalyzed hydrolysis, was inhibited by pretreatment with TOTP. In TOTP-pretreated rats, additional signals assigned to ABEC could be detected. On the other hand, no signals of HPA were observed. To obtain additional evidence for the structural assignment and to get a better resolution, the two-dimensional WlH shift-correlated (HMQC-BIRD)NMR spectrum of the urine from BA-treated rats (Figure 3) was measured. A correlation was found for all signals detected in onedimensional experiments. These signals agreed well with those of authentic standards, confirming the structures of HPA, ACEC, and ACECO. ACECO consists of two
diastereomers with chiral centers on Cys-CH and sulfoxide grouping with distinct 'H and 13C resonances. Proton signals of CHPCH~SO are split into 2 multiplets at 3.04 and 3.16 ppm and may interfere with those of 2-BHC and/ or 2-CHC, complicating both detection and unequivocal structural assignment. Correlation with the carbon resonance at approximately 51 ppm provided important evidence to confirm the structure of ACECO. Monooxygenase-catalyzed metabolic activation of BA would lead to the formation of BG as a reactive intermediate and, subsequently, to the excretion of urinary metabolites arising from its hydrolysis (DHPA) and/or conjugation with glutathione (1-BHC, 2-BHC, 1-CHC, 2-CHC) as shown in Figure 4. Therefore, HMQC NMR spectra were searched for 13C-label-derivedsignals of these potential metabolites (Table 1). No BG-derived metabolite was detected. Carboxylic acid profiles as followed by GC/MS of trimethylsilylated urine extracts are shown in Figure 5. Urinary carboxylic acids were identified by comparing found and reported spectra (31). Isotopic 13Cenrichment was calculated from the difference in the ratio of the ion intensities in urine from treated animals as compared to controls (for physiological metabolites) or to unlabeled authentic standards (for ACEC). For each peak, an average spectrum was taken and corrected for background. In rats treated with 1 mmol/kg [l3C1BA, significant enrichment was found in major metabolites detected, i.e., HPA and ACEC. A slight 13C enrichment was detected also for citric acid (CIA) and isocitric acid (ICA). In contrast, in TOTP-pretreated rats dosed with 0.5 mmol/ kg PCIBA, less enrichment was found for HPA and no significant enrichment was detected for CIA and ICA (Table 3). No other urinary carboxylicacids were detected to be 13C-enriched. To enable identification of metabolites bearing an enolizable carbonyl group, derivatization with hydroxylamine hydrochloride prior to silylation was applied. No additional metabolites were detected by this procedure. Although the main metabolic pathways of BA, namely, carboxylesterase-catalzyedhydrolysis and mercapturic acid formation, were reported previously (2,3), HPA has not been identified among metabolites of BA as yet. It should be noted that, in a previous experiment
6 Chem. Res. Toxicol., Vol. 7, No. 1, 1994
Linhart et al.
Table 3. IsotoDic 1F Enrichment of Urinary Metabolites of 1-Butyl I3-WIAcrylate 13C enrichment in %" (mean f SD)at a dose of compound 1mmol/kg BA 0.5 mmol/kg BA t TOTP 79 f 3 18 f 2 ( R = 2) 3-hydroxypropanoicacid (HPA) 92 f 4 ( R = 3) N-acetyl-S-(2-carboxyethyl)cysteine(ACEC) 90 f 2 citric acid (CIA) 7.5 f 1.8 nsb ( R = 5) isocitric acid (ICA) 5.2 f 1.4 nsb (n = 5) a The enrichment was calculated using characteristic ions at mlz 219 and 177 (HPA);364, 321, and 230 (ACEC);465,375,363,347, and 273 (CIA and ICA). Not significant. _
_
_
_
~
~
~
~
~
~
I100
la
15
LE
\COOR
El 9
30
R.T. 411.
1 4
BG COOR
/
\
\
'2
NHCOCH
OH
COOR(H)
OH
OH
COOH S
DHPA
+
2-BHC
(2-CHC)
NHCOCH, R = I-Butyl
5
cbS
