- Email: [email protected]
Haemoglobin adducts of epoxybutanediol from exposure to 1,3-butadiene or butadiene epoxides
ELSEVIER
Chemico-Biological
Interactions
105 (1997) 181- 198
Haemoglobin adducts of epoxybutanediol from exposure to 1,3-butadiene or butadiene epoxides Hermes Licea PCrez a, Jaana Lghdetie b, Helena Hinds0 Landin Ilkka KilpelZnen ‘, Pertti Koivisto d, Kimmo Peltonen d, Siv Osterman-Golkar a,*
a,
a Department
of Radiobiology, Stockholm Universit_y, S-106 91 Stockholm, Sweden of Medical Genetics, Institute of Biomedicine, University of Turku, FIN-205 20 Turku, Finland ’ Helsinki University, Institute of Biotechnology, FIN-00560 Helsinki, Finland d Molecular Dosimetry Group, Finnish Institute of Occupational Health (FIOH), FIN-00250 Helsinki, Finland
b Department
Received
16 April
1997; received
in revised form 6 June 1997; accepted 12 June 1997
Abstract Epoxybutanediol is one of the reactive metabolites of butadiene. It is formed via hydrolysis followed by oxidation of the primary metabolite of butadiene, epoxybutene, or via hydrolysis of diepoxybutane, a secondary metabolite of butadiene. Groups of male Sprague Dawley rats were treated by intraperitoneal injection of epoxybutene, epoxybutanediol or diepoxybutane. N-(2,3,4-Trihydroxybutyl)valine adducts in haemoglobin, formed from epoxybutanediol in its reaction with N-terminal valine, were measured using the N-alkyl Edman method followed by aeetylation of the Edman derivatives and analysis by gas chromatography mass speetrometry. The same adducts were also measured in male Wistar
Abbreviations: DQFCOSY, tathione; HBI, haemoglobin
double
quantum
filtered
COSY;
GC,
gas chromatography;
GSH,
glu-
binding index; HSQC, heteronuclear single quantum coherence; Ml, 1,2-dihydroxy+(N-acetylcysteinyl)butane; M2, I-hydroxy-2-(N-acetylcysteinyl)-3-butene or 2-hydroxyI-(N-acetylcysteinyl)-3-butene; MS, mass spectrometry; MS-MS, tandem mass spectrometry; NCI, negative ion chemical ionisation; NMR, nuclear magnetic resonance; PFPITC, pentafluorophenyl isothiocyanate; PFPTH, pentafluorophenylthiohydantoin; * Corresponding author. 0009-2797/97/$17.00
0 1997 Elsevier
PII SOOOS-2797(97)00049-5
Science
Ireland
ValGlyGly,
Ltd. All rights
valylglycylglycine.
reserved
182
H.L. Perez et al. I Chrmico-Biological Interactions 105 (1997) 181-198
rats exposed to butadiene by inhalation and in a few workers with occupational exposure to butadiene. Haemoglobin binding indexes, HBI, (pmol adduct/g per pmol of alkylating agent, or, for butadiene, per ppm x h), were calculated. The HBI for epoxybutanediol (about IO) is comparable to that of ethylene oxide in the rat demonstrating a similar capacity of the two compounds to alkylate nucleophilic sites in vivo. The HBI of diepoxybutane (about 8) for epoxybutanediol adduct formation is approximately the same as that of epoxybutanediol itself. Epoxybutanediol adduct formation was nonlinearly related to exposure in butadiene exposed rats. The epoxybutanediol-haemoglobin adduct levels were substantially higher than those of epoxybutene in both butadiene-exposed rats and humans suggesting an important role of epoxybutanediol in the toxicity of butadiene. Adducts of epoxybutanediol are probably useful for biomonitoring of human exposure to butadiene. 0 1997 Elsevier Science Ireland Ltd. Keywords: Haemoglobin adducts; Rats; Humans; Epoxybutene; Diepoxybutane; Biomarker
Butadiene
metabolites;
Epoxybutanediol;
1. Introduction 1,3-Butadiene is a high priority compound for risk assessment. It is used in the production of various polymers and copolymers. Sources of exposure to butadiene include cigarette smoke and emissions from butadiene production, storage, transport, and end use. In vitro and in vivo metabolism studies show that butadiene undergoes oxidation to form several reactive, potentially carcinogenic, metabolites and that there are significant species differences in metabolism. The carcinogenicity of butadiene in rodents is well established, with a high sensitivity of mice compared with rdtS. There is a corresponding striking difference in the clastogenic response to butadiene in these species (see the review by Melnick and Huff [l]). It is likely that quantitative differences in the metabolism are a key factor for the biological activity of butadiene. Measurements of specific adducts of reactive butadiene metaboiites may aid in the interpretation of species differences and of dose response curves for various genotoxic effects and in this way provide a basis for risk estimation for humans. Haemoglobin adducts of the primary metabolite of butadiene, epoxybutene have been measured previously in experimental animals [2-51 and in humans [6]. The haemoglobin binding index, HBI, defined as pmol adduct/g globin per ppm x h of butadiene, for epoxybutene adduct formation is low for humans (about 0.001) when compared with rats (about 0.03) and is considerably lower than, for example, the HBI for ethylene in humans (of the order of 0.2 [7]; about 0.07 [8]). The low alkylating efficiency of epoxybutene is related to a rapid clearance of the compound. However, this clearance may lead to formation of diepoxybutane or epoxybutanediol. The latter epoxide may be formed via hydrolysis of epoxybutene followed by oxidation, or via hydrolysis of diepoxybutane (Fig. 1).
H.L. PPrez et al. /C%emico-Biological lnreractions 105 (1997) 181-198
183
This study is a part of the project ‘Multi end-point analysis of the genetic damage induced by 1,3-butadiene and its major metabolites in somatic and germ cells of mice, rats and man. Genetic risk estimation by the parallelogram method’ sponsored by the European Commission, and concerns the dosimetry of epoxybutanediol. Groups of male Sprague Dawley rats were treated by intraperitoneal injection of the three reactive metabolites, epoxybutene, epoxybutanediol and diepoxybutane. N-2,3,4_Trihydroxybutyl adducts in haemoglobin, formed from epoxybutane-
Butadiene
Epoxybutene
J
\
HyoH \
OH Butenediol
Diepoxybutane
Epoxybutanediol
OH OH OH + OH Erythritol Fig. 1. Metabolic pathways for formation and elimination of epoxybutanediol from butadiene or from metabolites of butadiene. Possible pathways, other than oxidation and hydrolysis, are not included in the figure.
184
H.L.
P&z
et al. / Chemico-Biological
Interactions
105 (1997) 181p198
diol in its reaction with N-terminal valine, were measured using the N-alkyl Edman method, modified to allow quantification of these hydrophilic adducts, and analysis by gas chromatography mass spectrometry. The same adducts were also measured in male Wistar rats exposed to butadiene by inhalation and in a few workers with occupational exposure to butadiene. The study underscores the importance of dosimetry of all the reactive metabolites of butadiene and suggests that epoxybutanediol plays an important role in the toxicity of butadiene.
2. Materials
and methods
2.1. Chemicals 3,4-Epoxybutane-1,2-diol (butadiene diolepoxide, epoxybutanediol; purity > 96%) was synthesised at the Department of Chemistry, University of Helsinki, Finland, by oxidation of butenediol [9]. This synthetic product was used for the animal experiments. Butadiene diepoxide (diepoxybutane (97%) was obtained from Aldrich-Chemie, Steinheim, Germany (used in the animal experiments) and from Lancaster Synthesis, Lancashire, UK (used for syntheses of reference compounds). 3,4-Epoxy-1 -butene (98%) was obtained from Aldrich-Chemie. [14C]Valine was from Amersham, Buckinghamshire, UK. Triethylamine ( > 99%) was obtained from Sigma, St. Louis, MO. Pentafluorophenyl isothiocyanate (PFPITC, > 97%) was obtained from Fluka, Buchs, Switzerland, and was purified on a Sep-Pak silica mini column [lo]. All other chemicals used were of analytical grade. All glassware was silanised with dichlorodimethylsilane [ 111. The internal standard [globin containing 0.2 nmol (2H,)dihydroxypropylvaline per mg] was prepared by treatment of human erythrocytes with (*H,)glycidol [l I]. 2.2. Synthesis
of N-(2,3,4-trihydroxybutyl)valine
Diepoxybutane, 400 ~1, was hydrolysed in 3 ml of water for three days at 37°C (the half life is approximately 100 h [12]). Remaining diepoxybutane was extracted with toluene. N-(2,3,4_Trihydroxybutyl)valine was prepared by incubation of L-valine (1 mmol; 1 pCi/mmol 14C) with part of the hydrolysate, containing approximately 1.5 mmol epoxybutanediol, and 2 ml 0.5 M NaOH at 45°C for 3 h. The product was isolated by chromatography on a Dowex 50W x 4 column (2.5 x 39 cm). The column was eluted 1 M HCl and 18-20 ml fractions were collected. Samples of 50 ~1 of each fraction were used for radioactivity determination. Fraction 19, containing radioactivity associated to 14C-labelled and alkylated L-valine, was evaporated to give an oily product. The product was dissolved in 2 ml of water and filtered through a Cl8 Sep-Pak mini column (Millipore, Waters, Milford, MA). The yield was 60 pmol product according to determination of radioactivity. The product was dried under vacuum and part of it was submitted to NMR analyses. The radiochemical purity was determined to be over 96% by HPLC with radioactivity detection (see below, NMR studies).
H.L.
Perez et al. /Chemico-Biological
Interactions
2.3. Synthesis of N-(2,3,4-trihydroxybutyl)ValGlyGly
105 (1997) 181-198
185
ethyl ester
ValGlyGly ethyl ester (1 PCi 14C/mmol) was prepared according to OstermanGolkar et al. [6]. The tripeptide (0.1 mmol) was incubated with approximately 0.5 mmol hydrolysate of diepoxybutane (see above) at room temperature for 48 h. Excess epoxide was hydrolysed by addition of 1 ml 1 M HCl. The alkylated tripeptide was used without further purification for calibration of the analytical method. The concentration of trihydroxybutylvaline in the tripeptide was determined as follows: (a) Samples of the tripeptide were incubated with PFPITC and pentafluorophenylthiohydantoin (PFPTH) derivatives were extracted as described below (see Quantitative determination of haemoglobin adducts). The adduct content was calculated on the basis of extracted radioactivity. (b) Samples of internal standard globin were mixed with alkylated tripeptide or trihydroxybutylvaline, respectively, and hydrolysed in 6 M HCl (16 h, 120°C). The samples were then analysed using the N-alkyl Edman method for free amino acids [ 10,l l] and processing of the samples as described below. ValGlyGly ethyl ester was also reacted with the synthetic epoxybutanediol used in the animal experiments. 2.4. NMR studies Prior to NMR analysis the N-(2,3,4-trihydroxybutyl)valine was further purified with reverse phase HPLC. The conditions applied in the HPLC purification comprised a linear gradient from 0 to 30% methanol during 25 min. Formic acid, 50 mM, was used as a buffer. An Inertsil ODS-2 column (150 x 4.0 mm) was used in chromatography and the flow rate was 0.5 ml/min. The radiochemical purity was determined with a Radiomatic on-line radioactivity detector. The size of the loop used was 0.5 ml and the flow rate of the scintillation liquid was adjusted to 2.0 ml/min. The range of energy monitored during the analysis was from zero to 153 MeV. NMR spectra were obtained in 99.9% D,O at 500 MHz, using a Varian Unity 500 spectrometer. The spectra were recorded at 27°C and referenced to the water resonance (4.70 ppm). The ‘H spectrum was assigned using 2D DQFCOSY and 2D (iH-i3C) HSQC methods. 2.5. Animal treatments Adult male Sprague Dawley rats, age 2-3 months, body weight 270-550 g, from the Laboratory Animal Centre of the University of Turku were used. The rats were given a single intraperitoneal injection of the chemicals at the following doses: epoxybutene, 78.3 mg (1117 pmol)/kg body weight; diepoxybutane, 16.7 and 33.4 mg (194 and 387 pmol)/kg body weight; epoxybutanediol, 30.0 and 60.0 mg (289 and 578 pmol)/kg body weight. Epoxybutene was dissolved in corn oil and diepoxybutane and epoxybutanediol were dissolved in physiological saline. Control animals were given saline or corn oil only. The animals were killed by CO, asphyxiation 24 or 48 h after the injection. Blood was collected in heparinised tubes by cardiac puncture.
186
H.L.
P&z
et ul. iCl~emico-Bic~l~~~ical Intemctions
105 (1997) 181-198
The same rats were used for assays of micronuclei induction in somatic cells and germ cells (bone marrow polychromatic erythrocytes and spermatids), respectively [9,13]. The adduct study was added to the protocol at a late stage. Therefore, blood samples were not available from all animals. Blood samples were also obtained from male Wistar rats exposed to butadiene in inhalation chambers controlled for constant concentration of the gas with a Miran spectrophotometer. The concentrations used were 0, 50, 200, or 500 ppm, 6 h per day for 5 consecutive days [14]. The animals were killed 1 day after the last exposure and blood was taken by heart puncture. Three animals per exposure group were assessed for adducts of epoxybutene and of epoxybutanediol. 2.6. Humun
samples
Samples from two workers exposed to an estimated median air concentration of about 1 ppm butadiene at a Portuguese petrochemical plant and from two referents from the same plant, but with no known exposure to butadiene, were selected for this pilot study. These samples were previously assessed for adducts of epoxybutene
161. 2.7. Radioactivity
measurements
Radioactivity assays were made using a 1217 Rackbeta liquid scintillation counter (LKB, Wallac), OptiPhase ‘HiSafe’ scintillation cocktail and an internal standard kit (Wallac) for determination of counting efficiency. 2.8. Quantitative
determination
of huemoglobin
adducts
2.8. I. Sample prepurution Erythrocytes were separated from the plasma by centrifugation at 1000 x g for 10 min, and were then washed twice with saline. The cells were lysed by adding one volume of distilled water for each volume of packed cells. Globin was precipitated from the hemolysates according to Mowrer et al. [15]. Haemoglobin adducts were analysed using the N-alkyl Edman method [10,16] with a few modifications: samples of 50 mg of globin were dissolved in 1 ml of formamide. Internal standard (50 ~1 from a solution containing 0.385 mg (*H,)dihydroxypropyl-globin/ml), 40 ~1 1 M NaOH and 10 ~1 PFPITC were added and the samples were left on a rocking mixer at room temperature over night and then at 45°C for 1.5 h. Two millliters of water was added to the formamide phase and the PFPTH derivatives were extracted with diethyl ether (3 x 3 ml). The ether was evaporated under nitrogen. The samples were dissolved in 2 ml toluene and washed once with 1 ml 0.1 M sodium carbonate and once with 1 ml water. The sodium carbonate and water phases were pooled and extracted with ethyl ether (2 x 3 ml). The ether extract was evaporated to dryness, dissolved in 2 ml toluene and washed with 0.1 ml water. The toluene extracts were pooled in a new test-tube and evaporated under nitrogen at 50°C. The samples were acetylated with triethylamine, 25% (v/v), and acetic
H.L.
PPrez et al. / Chemieo-Biological
Interactions
105 (1997) 181p198
187
anhydride, 25% (v/v), in acetonitrile for 15 min essentially as described by Paulson and Lindberg [17] and evaporated under nitrogen at 30°C. The dry samples were dissolved in 3 ml of pentane and washed with 2 ml 60% (v/v) methanol/water. The pentane phase was transferred to a new test tube and was evaporated under nitrogen at 30°C. Before GC/MS-MS analysis the samples were dissolved in 50 ~1 of toluene. Haemoglobin adducts produced by the direct reaction of terminal carbon of epoxybutene with N-terminal valine [N-(2-hydroxy-3-butenyl)valine] were measured in the butadiene-exposed rats using the method described by Osterman-Golkar et al. [6]. 2.9. Muss-spectrometric
analysis
The samples and standards were analysed by GC/MS-MS on a Finnigan TSQ 700 mass spectrometer, using negative ion chemical ionisation (NCI). The operating procedures for the gas chromatograph were: helium as a carrier gas at constant gas pressure 8 psi (55 kPa); temperature programming 1 min at lOO”C, lO”C/min to 320°C and then 320°C for 7 min. The column used was a 30 m SE-54 (0.32 mm i.d., 1.0 pm phase thickness) fused silica capillary column from Alltech, Deerfeld, IL. The operating procedures for the mass spectrometer were: methane reagent gas at an ion source pressure of 4.8 torr (640 Pa), ion source temperature 150°C offset 30.0 V, ionisation energy 70 eV. Argon was used as collision gas at a pressure of 1.06 mtorr (0.14 Pa). Two ~1 of the samples, in toluene, were injected on column. The retention times were about 22 min and 45 s for the analyte and 20 min 35 s and 20 min and 42 s (double peak) for the internal standard. The daughter ions, 409, 367 and 59, respectively, formed from m/z 451 ([M-103]-) of the analyte and the daughter ions, 342, 300 and 284, respectively, formed from m/z 384 ([M-103] -) of the internal standard, were measured. 2.10. Calibration Control globin was dissolved in formamide to give a concentration of 10 mg/ml. Defined amounts of trihydroxybutylValGlyGly ethyl ester in the range 0, l-80 pmol were mixed with 1 ml control globin solution. Internal standard globin containing 3.85 pmol (2H,)dihydroxypropylvaline was added and the derivatisations were carried out as described above. The calibration curve was linear over the range of concentrations studied.
3. Results and discussion 3.1. NMR The product diepoxybutane)
prepared by reaction of epoxybutanediol (from hydrolysate with L-valine was studied by NMR. ‘H NMR of the product,
of 6
188
N.L.
PPrez et al. /Chew&o-Biological
Interactions
105 (1997) 181-198
Fig. 2. N-(2,3,4-Trihydroxybutyl)valine
0.96-1.03 (6h, m, H3), 2.18-2.28 (lh, m, H2), 3.14-3.24 (2h, m, H4), 3.48-3.54 (lh, m, Hl), 3.58-3.68 (3h, m, H6 and H7) and 3.664.2 (lh, m, H5). Because of the diastereomers the signals are split to multiples and a range of chemical shifts are reported. However, there was no indication of the positional isomer (i.e. adduct formed by reaction at carbon 3 of 3,4-epoxybutane-1,2-diol) present in the sample. This is in accordance with preferential nucleophilic attack at the terminal carbon of several related epoxides, such as propylene oxide [18], epichlorohydrin [1 I] and glycidyl ethers [19]. The structure of the epoxybutanediol product and the numbering of the protons are presented in Fig. 2. (ppm):
3.2. Preparation of the N-(trihydroxybutyljvaline
PFPTH derivative
As described previously for N-(dihydroxypropyl)valine [1 11,several modifications of the original N-alkyl Edman method [10,16] were required to make analysis of the N-(trihydroxybutyl)valine adduct possible. These modifications were: silanisation of all glass ware used in the procedure, addition of water to the reaction mixture (in formamide) before extraction of the alkylvaline-PFPTH and finally acetylation of the PFPTH derivative. Model experiments with radiolabelled N-(trihydroxybutyl)valine, and measurement of the yield in each step of the procedure described by Hindso Landin et al. [11], showed considerable losses of the N-(trihydroxybutyl)valine-PFPTH during the rather extensive washing of the toluene extract with water and sodium carbonate. The partition coefficient of N-(trihydroxybutyl)valinePFPTH for toluene/water was estimated as approximately 6. The number of washings and the volumes of water were therefore reduced to a minimum (see Section 2). 3.3. GCIMS The synthetic N-(triacetoxybutyl)valine-PFPTH was analysed by negative ion chemical ionisation GC/MS. The derivative gave one dominating peak in the chromatogram with a fragmentation pattern similar to those of several other alkylvaline-PFPTHs [20-221 (Fig. 3). The dominating fragments, HI/Z 534 ([M-20] -) and m/z 451 ([M-103] -), correspond to loss of HF and F + (CH,),(CH),CO from the valyl residue, respectively. The fragment m/z 323 corresponds to loss of the whole triacetoxybutyl substituent and the minor fragment MT/Z 225 corresponds
H.L. Pgrez et al. /Chemico-Biological
189
Interactions 105 (1997) 181-198
I
I
534
loo-
4
. E+07 1.66
I
80-
5J-
40-
20323 100
Fig. 3. Negative ion chemical tafluorophenyl-Z-thiohydantoin.
414
554
J&J\....,6co--+7
,..'.."“I. l..;.!..,L_
200
ionisation
mass spectrum
of I-(2,3,4-triacetoxybutyl)-5-isopropyl-3-pen-
to loss of C,F,NCS. Daughter ions of the fragment m/z 451 (Fig. 4) were used for analyses by GC/MS-MS. The GC/MS-MS ion chromatograms of samples from butadiene, epoxybutene or epoxybutanediol-exposed animals showed an additional peak with a slightly shorter GC retention time (22 min 35 s and 22 min 45 s for the
4
I 100
60
E+06 1.62
59
60
451
367
40
20
222
263 28p307
100
m
Fig. 4. Collision induced daughter spectrum propyl-3-pentafluorophenyl-2-thiohydantoin.
300
349 1
I-'"
400
of m/z 451 ([M-103] -) from
‘--'I
!xnJ
l-(2,3,4-triacetoxybutyl)-5-iso-
190
H.L.
Ptrez
et al. / Chemico-Biological
Interactions
105 (1997) 181~ 198
E+O5 2.312 Intensity
lnkxnal standard r.,..:
AA ., . . . . . . . . . . . . . . . .._ .,~ ., . . .& .., 20:oo
2050
21:40
22:30
24:lO
23:20
Time
-E+05 1 .949
Intensity
lntemal standard
20:oo
2050
21:40
22
23:20
24:lO
Time
Fig. 5. MS-MS ion chromatogram of a sample from (a) a diepoxybutane-exposed tanediol-exposed
rat. Retention
rat (b) an epoxybutimes 22 min 35 s for peak I and 22 min 45 s for peak II, respectively.
first and second peak, respectively) (Fig. 5). The two products gave the same daughter ions and were named I and II based on their GC retention times. The analyses of peak I were not calibrated but the quantities of this adduct were estimated directly from the peak areas of daughter ions of m/z 451 assuming equal response of type I and type II products.
H.L.
P&e= et al. /Chemico-Biological
3.4. The stereochemistry
of epoxybutanediol
Interactions
191
105 (1997) 181-198
adduct formation
(Fig. 6)
Presumably the peaks I and II (Fig. 5) correspond to different stereoisomeric forms of the trihydroxybutylvaline-PFPTH. Peak II was dominating, with only traces of peak I, in samples from diepoxybutane-exposed animals (Fig. 5a) and in samples of reference compounds prepared via hydrolysis of diepoxybutane. Both peaks, I and II, were seen in samples from animals exposed to epoxybutanediol (Fig. 5b), epoxybutene or butadiene. The formation of different stereoisomeric forms of epoxybutanediol and of its adducts, respectively, is explained in Fig. 6. The commercial diepoxybutane [( + )-butadiene diepoxide] is a racemic mixture of
-3-l OH
OH 1
0
1
iH
==JJOH OH
0:
V 2
&
OH
z OH
3
2 ‘:. OH +
4
OH
k +R
i)H
OH la
OH
OH 2a
Fig. 6. Formation of different stereoisomeric ( + )-diepoxybutane or from 3-butene-1,2-dial, haemoglobin (R = N-terminal valine).
3a
4a
forms of adducts of epoxybutanediol, formed from respectively, in reaction with N-terminal valine in
192
H.L.
PPrez et al. /Chemico-Biological
interactions
105 (1997) 181-198
(SS) and (RR) stereoisomers. Apparently hydrolysis of this mixture, in vivo as well as in vitro, mainly gives the diolepoxides assigned 1 and 2, whereas epoxybutanediol, prepared from 2-butene-1,4-diol [9] or formed metabolically from epoxybutene, contains the diolepoxide structures l-4. Using our analytical conditions the acetylated PFPTH derivatives of structures l(a) and 2(a) (adducts of 1 and 2) and 3(a) and 4(a) (adducts of 3 and 4) are resolved as two peaks. Analysis of ValGlyGly ethyl ester, treated in vitro with epoxybutanediol prepared from butenediol, gave, as expected, both peaks I and II. Using a different type of derivatives (trimethylsilylated PFPTHs), Rydberg et al. [23] observed two peaks of trihydroxybutylvaline in the chromatogram of a sample from haemoglobin treated in human erythrocytes with rat-diepoxybutane. 3.5. Adduct
levels in rats and humans
Table 1 presents N-(2,3,4-trihydroxybutyl)valine adduct levels in haemoglobin from male Sprague Dawley rats administered the three reactive metabolites, epoxybutene, epoxybutanediol or diepoxybutane by intraperitoneal injection. The HBI, given as pmol adduct/g per pmol of alkylating agent, has been calculated in order to facilitate an intercomparison of the various agents in this study and a comparison with other compounds with respect to their capacity to alkylate macromolecules in vivo. The HBI values are related to formation of epoxybutanediol adducts of type II; the HBI for formation of epoxybutanediol adducts of both types, I + II, may then be estimated using the adduct ratio (I + II)/II. Treatment of rats with diepoxybutane gave mainly adducts of type I (Fig. 5a), and treatment with epoxybutanediol or epoxybutene or exposure to butadiene gave adducts of type I and II (Fig. 5b). This observation suggests that the in vivo oxidation of butenediol is not stereospecific but gives a mixture of the diolepoxides l-4 (Fig. 6). The single peak formed by diepoxybutane may have been formed by direct formation of adducts of diepoxybutane followed by hydrolysis of the epoxide bond or by hydrolysis to epoxybutanediol followed by formation of adducts. The contribution of each pathway is not known, however, the latter is presumably dominating. Studies of the reactions of diepoxybutane with valinamide in vitro show that the primary reaction product of diepoxybutane with the valine-NH, gives a ring structured product in a reaction which is fast compared with hydrolysis of the second epoxide ring [23]. Thus, trihydroxybutylvaline is formed only in small amounts after direct reaction of diepoxybutane with N-terminal valine in haemoglobin. Epoxybutanediol and diepoxybutane were administered to the animals at two doses for each compound. The resulting adduct data (although limited to analysis of a few animals) were compatible with linear dose-response relationships for epoxybutanediol adduct formation. The HBI for epoxybutanediol, about 3.7 for adduct II and accordingly about 10 (2.7 x 3.7 pmol/g per pmol/kg) for adducts I + II, is comparable to the HBI, about 16, for ethylene oxide in the rat as estimated from data given by Walker et al. [24], demonstrating a similar capacity of the two compounds (epoxybutanediol and ethylene oxide) to alkylate nucleophilic
Diepoxybutane
Epoxybutene Epoxybutanedial
Agent
1117 289 578 194 387
60 16.7 33.4
0
Dose (pmol/kg)
reactive metabohtes
1900cd, 2700’ 1800 k 283 2800 & 100
83+13 8604, 1100’
2 k 2.8
Adduct level + S.D.; adduct II (pmol/g)”
of rats administered
78.3 30
0
Dose (mg/kg)
levels in haemoglobin
1
4+0.5 2.7 k 0.1
epoxybutanediol,
or
4.0 9.3 7.2
0.07 3.4
HBI; adduct II (pmol/ g per PmoVkg)
epoxybutene,
Adduct ratio f S.D.; [(I+ II)/IIl”
of butadiene:
a See Figs. 5 and 6. I and II are different stereoisomeric forms of the adduct. b Groups of two to three rats were treated by i.p. injection and blood samples were collected after 48 h unless otherwise stated. ’ Adduct level of an individual rat. d The animals were sacrificed 24 h after treatment,
Rats, Sprague Dawley, i.p.b
Species, strain, route of exposure
Table I N-(2,3,4-Trihydroxybutyl)valine diepoxybutane
194
H.L.
P&r
et al. /Chemico-Biological
Interactions
105 (1997) 181-198
Table 2 N-(2,3,4-Trihydroxybutyl)valine tion and of human subjects
adduct levels in haemoglobin of rats exposed with occupational exposure to butadiene
Species,
level
Rats,
strain
Wistarh
(ppm)
Butadiene
Adduct level & S.D.; adduct 11 (pmol/g)’
0 50 200 500
912 680 i 230 119Ok275’ 585 * 230
Humans’ Controls Occupationally exposed
Around
1”
Adduct ratio [(I + lI):II]”
3.2 f 0.2d
to butadiene
by inhala-
HBI; adduct (pmol/g h)
II
per ppm x
0.5 0.2 0.04
1.8 (0.05)’ 3.3 (0.02)’ 10 (0.15)’ 14 (0.20)’
* See Figs. 5 and 6. 1 and II are different stereoisomeric forms of the adduct. ‘Three animals per group. The rats were exposed to butadiene by inhalation 6 h/day for 5 days and killed 1 day after the last exposure. ‘Two animals only. One high value excluded (outside the calibration curve). ’ Average ( f SD.) calculated from adduct data in animals exposed to butadiene at 50, 200 or 500 ppm. ’ Adduct levels of individual subjects are indicated. r Values for adducts of epoxybutene (N-(2-hydroxy-3-butenyl)valine) are given within parenthesis. g Median concentration of butadiene estimated as 1 ppm [6].
sites in vivo. The efficiency of diepoxybutane with respect to formation of epoxybutanediol-hemoglobin adducts (HBI around 8) is approximately the same as that of the epoxybutanediol itself. Table 2 presents N-(2,3,4_trihydroxybutyl)valine adduct levels in haemoglobin from male Wistar rats exposed to butadiene and from a few human subjects with occupational exposure to butadiene, respectively. Studies on tissue concentrations of epoxides (epoxybutene and diepoxybutane) in mice and rats during exposure to butadiene [25] demonstrate very low blood levels of diepoxybutane in the rats. Therefore, epoxybutanediol, formed by oxidation of butenediol, is probably the main precursor of epoxybutanediol adducts in butadiene-exposed rats. The HBI, expressed as pmol trihydroxybutylvaline/g globin per ppm x h, decreases with butadiene concentration from 0.5 at 50 ppm to 0.04 at 500 ppm demonstrating an effect of dose rate. Since the formation of epoxybutene adducts is essentially linearly related to exposure level ~ the epoxybutene adduct levels are 0.6, 21, 88 and 180 pmol N-(2-hydroxy-3_butenyl)valine/g’ globin in the rats exposed to 0, 50, 200 and 500 ppm, respectively ~ the non-linearity with concentration is probably related to the metabolism of butenediol. A dose rate effect may in part explain the low HBI-value observed for the animals administered an acute dose of epoxybutene (Table 1). ’ This adduct
constitutes
approximately
60’%) of the total adducts
with N-terminal
valine [5]
H.L.
Ptrez
et al. /Chemico-Biological
Interactions
105 (1997) 181-198
195
The HBI for epoxybutanediol adduct formation from butadiene (3.2 x 0.4 pmol/ g per ppm x h) at the lowest air concentration, 50 ppm, is comparable to the HBI for ethylene oxide adducts in ethylene exposed rats (estimated as about 1 [16]). The analyses of epoxybutanediol adduct levels in humans were limited to a few samples from butadiene-exposed workers and controls investigated previously with respect to levels of adducts of the primary butadiene metabolite epoxybutene [7] (Table 2). The epoxybutene adducts measured were those formed by reaction of the The adduct terminal carbon in epoxybutene, i.e., N-(2-hydroxy-3-butenyl)valine. levels of epoxybutanediol (type II) were much higher (about 70-fold) than the adduct levels of epoxybutene measured in the same subjects. The presence of adducts of both epoxybutene and epoxybutanediol in the two control subjects supports the notion that these subjects are exposed to butadiene, possibly due to occasional visits to contaminated areas of the plant [6]. Since epoxybutanediol is downstream compared to epoxybutene in the metabolic pathway of butadiene (Fig. 1) the high adduct ratio (epoxybutanediol/epoxybutene) indicates an efficient formation of epoxybutanediol and a considerably slower elimination of this epoxide than of epoxybutene. Sabourin et al. [26] examined the urinary excretion of mercapturic acids in mice, rats, hamsters, and monkeys. All four species produced two metabolites, Ml [ 1,2-dihydroxy-4-(N-acetylcysteinyl)butane] and M2 [ 1-hydroxy-2-(N-acetylcysteinyl)-3-butene or 2-hydroxy-1-(N-acetylcysteinyl)-3-butene]. M2 is formed by conjugation of glutathione (GSH) with epoxybutene, and Ml appears to be formed by GSH conjugation with butenediol, the product of hydrolysis of epoxybutene. The butenediol conjugation is thought to occur via the intermediary formation of a reactive a$-unsaturated carbonyl compound. When comparing the four species, the ratio of Ml to the sum of Ml + M2 was linearly related to hepatic epoxide hydrolase activities for each species. Bechtold et al. [27] showed that Ml is the predominant metabolite in urine from butadiene-exposed workers. The ratio of epoxybutanediol adducts (type II) to epoxybutene adducts [N-(2-hydroxy-3-butenyl)valine, representing attack on the terminal carbon of epoxybutene] in the butadiene-exposed rats is 4-26 in the individual rats of this study (to be completed and published in another context) as compared to about 70 in the human subjects. The large amounts of trihydroxybutyl adducts as compared to epoxybutene adducts in haemoglobin and the high urinary concentrations of Ml as compared to M2 suggest that the hydrolysis pathway of epoxybutene is important in both species. The strong focus on epoxybutene and diepoxybutane in previous discussions on the toxicity of butadiene may, at least in part, be related to the techniques which have been used to study the formation of reactive metabolites in vitro and in vivo. These techniques generally involve extraction of the microsomal samples, tissue homogenates, or blood with an organic solvent such as o-xylene [28] or methylene chloride [29,30]. The hydrophilic epoxybutanediol is not recovered by these extractions but stays in the water phase. Cheng and Ruth [31] solved the problem by evaporation of the media to dryness (epoxybutanediol is not volatile) followed by extraction of the residue with ethyl acetate overnight. These authors incubated rat
196
H.L.
PPrez et al. /Chemico-Biological
Inteructions
105 (1997) 181-198
liver microsomes in the presence of NADPH and found that butenediol and epoxybutanediol were the major products. In conclusion, this study shows that measurements of epoxybutanediol adducts with N-terminal valine in haemoglobin can be used for dosimetry studies in experimental animals and to determine individual exposures in human epidemiological studies. Epoxybutanediol gives substantially higher haemoglobin adduct levels than epoxybutene in both butadiene-exposed rats and humans suggesting an important role of epoxybutanediol in the toxicity of butadiene.
Acknowledgements We thank Mr Vlado Zorcec and MS Annika Gustavsson for mass spectrometric analyses and MS Aila Suutari for technical assistance. This study was supported by the European Commission, project EV5V-CT94-0543. The NMR spectra are available on request.
References [l] R.L. Melnick, J. Huff, 1,3-Butadiene: toxicity and carcinogenicity in laboratory animals and in humans, Rev. Environ. Contam. Toxicol. 124 (1992) I1 1~ 144. [2] S. Osterman-Golkar, A. Kautiainen, E. Bergmark, K. Hakansson, J. Maki-Paakkanen, Hemoglobin adducts and urinary mercapturic acids in rats as biological indicators of butadiene exposure, Chem.-Biol. Interact. 80 (1991) 291-302. [3] O.E. Albrecht, J.G. Filser, H.D. Neumann, Biological monitoring of 1,3-butadiene: species differences in haemoglobin binding in rat and mouse, in: M. Sorsa, K. Peltonen, H. Vainio, K. Hemminki (Eds.), Butadiene and Styrene: Assessment of Health Hazards. IARC Scientific Publications, vol. 127, International Agency for Research on Cancer, Lyon, 1993, pp. 1355142. [4] S.M. Osterman-Golkar, J.A. Bond, J.B. Ward Jr., M.S. Legator, Use of haemoglobin adducts for biomonitoring exposure to 1,3-butadiene, in: M. Sorsa, K. Peltonen, H. Vainio, K. Hemminki (Eds.), Butadiene and Styrene: Assessment of Health Hazards. IARC Scientific Publications, vol 127. International Agency for Research on Cancer, Lyon, 1993, pp. 1277134. [5] N.Yu. Tretyakova, Y. Lin, P.B. Upton, R. Sangaiah, J.A. Swenberg, Macromolecular adducts of butadiene, Toxicology 113 (1996) 70-76. [6] S. Osterman-Golkar, K. Peltonen, T. Anttinen-Klemetti, H. Hindso Landin, V. Zorcec, M. Sorsa, Haemoglobin adducts as biomarkers of occupational exposure to 1.3-butadiene, Mutagenesis 10 (1996) 145-149. [7] S. Osterman-Golkar, J.A. Bond, Biomonitoring of 1,3-butadiene and related compounds, Environ. Health Perspect. 104 (Suppl. 5) (1996) 907-915. [8] F. Granath, 0. Rohlen, G. Goransson, L. Hansson, A.L. Magnusson. M. Tornqvist, Relationship between dose in vivo of ethylene oxide and exposure to ethene studied in exposed workers, Hum. Exp. Toxicol. 15 (1996) 8266833. [9] J. Llhdetie, K. Peltonen, T. Sjoblom, Germ cell mutagenicity of three metabolites of 1,3-butadiene in the rat: induction of spermatid micronuclei by butddiene mono-, di- and diolepoxides in vivo, Environ. Mol. Mutagen., in press. [IO] M. Tornqvist, J. Mowrer, S. Jensen, L. Ehrenberg, Monitoring of environmental cancer initiators through hemoglobin adducts by a modified Edman degradation method, Anal. Biochem. 154 (1986) 2555266.
H.L. [1 1] H. Hindso
Perez et al. / Chemico-Biological
Landin,
S. Osterman-Golkar,
drin by hemoglobin adducts, Anal. [12] L. Ehrenberg, S. Hussain, Genetic l&113. [13] J. Lahdetie,
J. Grawe,
polychromatic
Flow
erythrocytes
V. Zorcec,
M. Tiirnqvist,
analysis
Biomonitoring epoxides,
of micronucleus
1,2;3,4_diepoxybutane,
estimation
of petrochemical M. Tornqvist, for monitoring
11 (1986) 215-231. [16] M. Tornqvist, A. Kautiainen, to gasoline [17] J. Paulson, matogr.
emissions, S. Jensen, occupational R.N.
of epichlorohy-
Mutat.
induction
Res. 86 (1981)
in rat bone
3,4-epoxy-I-butene
and
3,4-diol, Cytometry, in press. [14] M. Sorsa, K. Peltonen, D. Anderson, N.A. Demopoulos, H.G. Neumann, Assessment of environmental and occupational exposures to butadiene [15] J. Mowrer, hemoglobin
191
10.5 (1997) 181-198
Biochem. 240 (1996) l-6. toxicity of some important
cytometric by
Interactions
marrow
1,2-epoxybutdne-
S. Osterman-Golkar, as a model for risk
Mutagenesis 11 (1996) 9- 17. L. Ehrenberg, Modified Edman degradation applied to exposure to alkylating agents, Toxicol. Environ. Chem.
Gatz,
L. Ehrenberg,
Hemoglobin
adducts
in animals
and diesel exhausts, 1 Alkenes, J. Appl. Toxicol. 8 (1988) 1599170. C. Lindberg, Increasing thermospray response for cortisol by derivatization,
exposed J. Chro-
554 (1991) 1499154.
[18] K. Svensson, S. Osterman-Golkar, Kinetics of metabolism of propene and covalent macromolecules in the mouse, Toxicol. Appl. Pharmacol. 73 (1984) 363-372. [19] H. Licea Perez, K. Plna, S. Osterman-Golkar, Dosimetry of glycidyl tion of haemoglobin adducts, Chem.-Biol. Interact. 103 (1997) l-16. [20] P. Rydberg, N-substituted
ethers
Golkar, Monitoring occupational blood, Stand. J. Work Environ.
V. Zorcec,
J. MCki-Paakanen,
H.
exposure to styrene by hemoglobin Health 19 (1993) 255-263.
to
in mice by quantifica-
B. Liining, C.A. Wachtmeister, M. Tornqvist, Synthesis and valines and their phenyland pentafluorophenyl-thiohydantoins,
Stand. 47 (1993) 813-817. [21] A. Christakopoulos, E. Bergmark,
binding
characterization of Acta Chem.
Norppa,
adducts
S. Osterman-
and metabolites
in
[22] S.M. Osterman-Golkar, J.P. MacNeela, M.J. Turner, V.E. Walker, J.A. Swenberg, S.J. Sumner, N. Youtsey, T.R. Fennell, Monitoring exposure to acrylonitrile using adducts with N-terminal valine in hemoglobin, Carcinogenesis 15 (1994) 2701-2707. [23] P. Rydberg, A.L. Magnusson, V. Zorcec, F. Granath, M. Tornqvist, valines in hemoglobin from butadiene metabolites, Chem.-Biol. Interact. [24] V.E. Walker, J.P. MacNeela, J. Swenberg, ethylene oxide: formation and persistence repeated
exposure
Adducts to N-terminal 101 (1996) 193-205.
M.J. Turner Jr., T.R. Fennell, Molecular dosimetry of of N-(2-hydroxyethyl)valine in hemoglobin following
of rats and mice, Cancer
Res. 52 (1992) 4320-4327.
[25] J.R. Thornton-Manning, A.R. Dahl, W.E. Bechtold, W.C. Griffith Jr., R.F. Henderson, Disposition of butadiene monoepoxide and butadiene diepoxide in various tissues of rats and mice following a low-level inhalation 1731. [26] P.J. Sabourin, L.T. Burka, W.E.
exposure Bechtold,
to A.R.
1,3-butadiene, Dahl,
M.D.
Carcinogenesis Hoover,
16 (1995)
I.Y. Chang,
R.F.
1723Hender-
son, Species differences in urinary butadiene metabolites; identification of 1,2-dihydroxy-4-(Nacetylcysteinyl)butane, a novel metabolite of butadiene, Carcinogenesis 13 (1992) 1633-1638. [27] W.E. Bechtold, M.R. Strunk, I.Y. Chang, J.B. Ward Jr., R.F. Henderson, Species differences urinary
butadiene
metabolites:
comparisons
of metabolite
ratios
between
in
mice, rats, and humans,
Toxicol. Appl. Pharmacol. 127 (1994) 4449. [28] R.J. Duescher, A.A. Elfarra, Human liver microsomes are efficient catalysts for 1,3-butadiene oxidation: Evidence for major roles by cytochrome P450 2A6 and 2E1, Arch. Biochem. Biophys. 311 (1994) 342-349. [29] M.W.
Himmelstein,
of l$butadiene Carcinogenesis
M.J. Turner,
B. Asgharian,
and butadiene epoxides 15 (1994) 1479- 1486.
J.A. Bond,
Comparison
in mice and rats exposed
of blood
to 1,3-butadiene
concentrations by inhalation,
198
H.L. PCrez et al. /Chemico-Biological
Interactions 105 (1997) 181-198
[30] M.W. Himmelstein, B. Asgharian, J.A. Bond, High concentrations of butadiene epoxides in livers and lungs of mice compared to rats exposed to 1,3-butadiene, Toxicol. Appl. Pharmacol. 132 (1995) 281-288. [31] X. Cheng, J.A. Ruth, A simplified methodology for quantitation of butadiene metabolites: Application to the study of 1,3-butadiene metabolism by rat liver microsomes, Drug Metab. Dispos. 21 (1993) 12lLl24.
Chemico-Biological
Interactions
105 (1997) 181- 198
Haemoglobin adducts of epoxybutanediol from exposure to 1,3-butadiene or butadiene epoxides Hermes Licea PCrez a, Jaana Lghdetie b, Helena Hinds0 Landin Ilkka KilpelZnen ‘, Pertti Koivisto d, Kimmo Peltonen d, Siv Osterman-Golkar a,*
a,
a Department
of Radiobiology, Stockholm Universit_y, S-106 91 Stockholm, Sweden of Medical Genetics, Institute of Biomedicine, University of Turku, FIN-205 20 Turku, Finland ’ Helsinki University, Institute of Biotechnology, FIN-00560 Helsinki, Finland d Molecular Dosimetry Group, Finnish Institute of Occupational Health (FIOH), FIN-00250 Helsinki, Finland
b Department
Received
16 April
1997; received
in revised form 6 June 1997; accepted 12 June 1997
Abstract Epoxybutanediol is one of the reactive metabolites of butadiene. It is formed via hydrolysis followed by oxidation of the primary metabolite of butadiene, epoxybutene, or via hydrolysis of diepoxybutane, a secondary metabolite of butadiene. Groups of male Sprague Dawley rats were treated by intraperitoneal injection of epoxybutene, epoxybutanediol or diepoxybutane. N-(2,3,4-Trihydroxybutyl)valine adducts in haemoglobin, formed from epoxybutanediol in its reaction with N-terminal valine, were measured using the N-alkyl Edman method followed by aeetylation of the Edman derivatives and analysis by gas chromatography mass speetrometry. The same adducts were also measured in male Wistar
Abbreviations: DQFCOSY, tathione; HBI, haemoglobin
double
quantum
filtered
COSY;
GC,
gas chromatography;
GSH,
glu-
binding index; HSQC, heteronuclear single quantum coherence; Ml, 1,2-dihydroxy+(N-acetylcysteinyl)butane; M2, I-hydroxy-2-(N-acetylcysteinyl)-3-butene or 2-hydroxyI-(N-acetylcysteinyl)-3-butene; MS, mass spectrometry; MS-MS, tandem mass spectrometry; NCI, negative ion chemical ionisation; NMR, nuclear magnetic resonance; PFPITC, pentafluorophenyl isothiocyanate; PFPTH, pentafluorophenylthiohydantoin; * Corresponding author. 0009-2797/97/$17.00
0 1997 Elsevier
PII SOOOS-2797(97)00049-5
Science
Ireland
ValGlyGly,
Ltd. All rights
valylglycylglycine.
reserved
182
H.L. Perez et al. I Chrmico-Biological Interactions 105 (1997) 181-198
rats exposed to butadiene by inhalation and in a few workers with occupational exposure to butadiene. Haemoglobin binding indexes, HBI, (pmol adduct/g per pmol of alkylating agent, or, for butadiene, per ppm x h), were calculated. The HBI for epoxybutanediol (about IO) is comparable to that of ethylene oxide in the rat demonstrating a similar capacity of the two compounds to alkylate nucleophilic sites in vivo. The HBI of diepoxybutane (about 8) for epoxybutanediol adduct formation is approximately the same as that of epoxybutanediol itself. Epoxybutanediol adduct formation was nonlinearly related to exposure in butadiene exposed rats. The epoxybutanediol-haemoglobin adduct levels were substantially higher than those of epoxybutene in both butadiene-exposed rats and humans suggesting an important role of epoxybutanediol in the toxicity of butadiene. Adducts of epoxybutanediol are probably useful for biomonitoring of human exposure to butadiene. 0 1997 Elsevier Science Ireland Ltd. Keywords: Haemoglobin adducts; Rats; Humans; Epoxybutene; Diepoxybutane; Biomarker
Butadiene
metabolites;
Epoxybutanediol;
1. Introduction 1,3-Butadiene is a high priority compound for risk assessment. It is used in the production of various polymers and copolymers. Sources of exposure to butadiene include cigarette smoke and emissions from butadiene production, storage, transport, and end use. In vitro and in vivo metabolism studies show that butadiene undergoes oxidation to form several reactive, potentially carcinogenic, metabolites and that there are significant species differences in metabolism. The carcinogenicity of butadiene in rodents is well established, with a high sensitivity of mice compared with rdtS. There is a corresponding striking difference in the clastogenic response to butadiene in these species (see the review by Melnick and Huff [l]). It is likely that quantitative differences in the metabolism are a key factor for the biological activity of butadiene. Measurements of specific adducts of reactive butadiene metaboiites may aid in the interpretation of species differences and of dose response curves for various genotoxic effects and in this way provide a basis for risk estimation for humans. Haemoglobin adducts of the primary metabolite of butadiene, epoxybutene have been measured previously in experimental animals [2-51 and in humans [6]. The haemoglobin binding index, HBI, defined as pmol adduct/g globin per ppm x h of butadiene, for epoxybutene adduct formation is low for humans (about 0.001) when compared with rats (about 0.03) and is considerably lower than, for example, the HBI for ethylene in humans (of the order of 0.2 [7]; about 0.07 [8]). The low alkylating efficiency of epoxybutene is related to a rapid clearance of the compound. However, this clearance may lead to formation of diepoxybutane or epoxybutanediol. The latter epoxide may be formed via hydrolysis of epoxybutene followed by oxidation, or via hydrolysis of diepoxybutane (Fig. 1).
H.L. PPrez et al. /C%emico-Biological lnreractions 105 (1997) 181-198
183
This study is a part of the project ‘Multi end-point analysis of the genetic damage induced by 1,3-butadiene and its major metabolites in somatic and germ cells of mice, rats and man. Genetic risk estimation by the parallelogram method’ sponsored by the European Commission, and concerns the dosimetry of epoxybutanediol. Groups of male Sprague Dawley rats were treated by intraperitoneal injection of the three reactive metabolites, epoxybutene, epoxybutanediol and diepoxybutane. N-2,3,4_Trihydroxybutyl adducts in haemoglobin, formed from epoxybutane-
Butadiene
Epoxybutene
J
\
HyoH \
OH Butenediol
Diepoxybutane
Epoxybutanediol
OH OH OH + OH Erythritol Fig. 1. Metabolic pathways for formation and elimination of epoxybutanediol from butadiene or from metabolites of butadiene. Possible pathways, other than oxidation and hydrolysis, are not included in the figure.
184
H.L.
P&z
et al. / Chemico-Biological
Interactions
105 (1997) 181p198
diol in its reaction with N-terminal valine, were measured using the N-alkyl Edman method, modified to allow quantification of these hydrophilic adducts, and analysis by gas chromatography mass spectrometry. The same adducts were also measured in male Wistar rats exposed to butadiene by inhalation and in a few workers with occupational exposure to butadiene. The study underscores the importance of dosimetry of all the reactive metabolites of butadiene and suggests that epoxybutanediol plays an important role in the toxicity of butadiene.
2. Materials
and methods
2.1. Chemicals 3,4-Epoxybutane-1,2-diol (butadiene diolepoxide, epoxybutanediol; purity > 96%) was synthesised at the Department of Chemistry, University of Helsinki, Finland, by oxidation of butenediol [9]. This synthetic product was used for the animal experiments. Butadiene diepoxide (diepoxybutane (97%) was obtained from Aldrich-Chemie, Steinheim, Germany (used in the animal experiments) and from Lancaster Synthesis, Lancashire, UK (used for syntheses of reference compounds). 3,4-Epoxy-1 -butene (98%) was obtained from Aldrich-Chemie. [14C]Valine was from Amersham, Buckinghamshire, UK. Triethylamine ( > 99%) was obtained from Sigma, St. Louis, MO. Pentafluorophenyl isothiocyanate (PFPITC, > 97%) was obtained from Fluka, Buchs, Switzerland, and was purified on a Sep-Pak silica mini column [lo]. All other chemicals used were of analytical grade. All glassware was silanised with dichlorodimethylsilane [ 111. The internal standard [globin containing 0.2 nmol (2H,)dihydroxypropylvaline per mg] was prepared by treatment of human erythrocytes with (*H,)glycidol [l I]. 2.2. Synthesis
of N-(2,3,4-trihydroxybutyl)valine
Diepoxybutane, 400 ~1, was hydrolysed in 3 ml of water for three days at 37°C (the half life is approximately 100 h [12]). Remaining diepoxybutane was extracted with toluene. N-(2,3,4_Trihydroxybutyl)valine was prepared by incubation of L-valine (1 mmol; 1 pCi/mmol 14C) with part of the hydrolysate, containing approximately 1.5 mmol epoxybutanediol, and 2 ml 0.5 M NaOH at 45°C for 3 h. The product was isolated by chromatography on a Dowex 50W x 4 column (2.5 x 39 cm). The column was eluted 1 M HCl and 18-20 ml fractions were collected. Samples of 50 ~1 of each fraction were used for radioactivity determination. Fraction 19, containing radioactivity associated to 14C-labelled and alkylated L-valine, was evaporated to give an oily product. The product was dissolved in 2 ml of water and filtered through a Cl8 Sep-Pak mini column (Millipore, Waters, Milford, MA). The yield was 60 pmol product according to determination of radioactivity. The product was dried under vacuum and part of it was submitted to NMR analyses. The radiochemical purity was determined to be over 96% by HPLC with radioactivity detection (see below, NMR studies).
H.L.
Perez et al. /Chemico-Biological
Interactions
2.3. Synthesis of N-(2,3,4-trihydroxybutyl)ValGlyGly
105 (1997) 181-198
185
ethyl ester
ValGlyGly ethyl ester (1 PCi 14C/mmol) was prepared according to OstermanGolkar et al. [6]. The tripeptide (0.1 mmol) was incubated with approximately 0.5 mmol hydrolysate of diepoxybutane (see above) at room temperature for 48 h. Excess epoxide was hydrolysed by addition of 1 ml 1 M HCl. The alkylated tripeptide was used without further purification for calibration of the analytical method. The concentration of trihydroxybutylvaline in the tripeptide was determined as follows: (a) Samples of the tripeptide were incubated with PFPITC and pentafluorophenylthiohydantoin (PFPTH) derivatives were extracted as described below (see Quantitative determination of haemoglobin adducts). The adduct content was calculated on the basis of extracted radioactivity. (b) Samples of internal standard globin were mixed with alkylated tripeptide or trihydroxybutylvaline, respectively, and hydrolysed in 6 M HCl (16 h, 120°C). The samples were then analysed using the N-alkyl Edman method for free amino acids [ 10,l l] and processing of the samples as described below. ValGlyGly ethyl ester was also reacted with the synthetic epoxybutanediol used in the animal experiments. 2.4. NMR studies Prior to NMR analysis the N-(2,3,4-trihydroxybutyl)valine was further purified with reverse phase HPLC. The conditions applied in the HPLC purification comprised a linear gradient from 0 to 30% methanol during 25 min. Formic acid, 50 mM, was used as a buffer. An Inertsil ODS-2 column (150 x 4.0 mm) was used in chromatography and the flow rate was 0.5 ml/min. The radiochemical purity was determined with a Radiomatic on-line radioactivity detector. The size of the loop used was 0.5 ml and the flow rate of the scintillation liquid was adjusted to 2.0 ml/min. The range of energy monitored during the analysis was from zero to 153 MeV. NMR spectra were obtained in 99.9% D,O at 500 MHz, using a Varian Unity 500 spectrometer. The spectra were recorded at 27°C and referenced to the water resonance (4.70 ppm). The ‘H spectrum was assigned using 2D DQFCOSY and 2D (iH-i3C) HSQC methods. 2.5. Animal treatments Adult male Sprague Dawley rats, age 2-3 months, body weight 270-550 g, from the Laboratory Animal Centre of the University of Turku were used. The rats were given a single intraperitoneal injection of the chemicals at the following doses: epoxybutene, 78.3 mg (1117 pmol)/kg body weight; diepoxybutane, 16.7 and 33.4 mg (194 and 387 pmol)/kg body weight; epoxybutanediol, 30.0 and 60.0 mg (289 and 578 pmol)/kg body weight. Epoxybutene was dissolved in corn oil and diepoxybutane and epoxybutanediol were dissolved in physiological saline. Control animals were given saline or corn oil only. The animals were killed by CO, asphyxiation 24 or 48 h after the injection. Blood was collected in heparinised tubes by cardiac puncture.
186
H.L.
P&z
et ul. iCl~emico-Bic~l~~~ical Intemctions
105 (1997) 181-198
The same rats were used for assays of micronuclei induction in somatic cells and germ cells (bone marrow polychromatic erythrocytes and spermatids), respectively [9,13]. The adduct study was added to the protocol at a late stage. Therefore, blood samples were not available from all animals. Blood samples were also obtained from male Wistar rats exposed to butadiene in inhalation chambers controlled for constant concentration of the gas with a Miran spectrophotometer. The concentrations used were 0, 50, 200, or 500 ppm, 6 h per day for 5 consecutive days [14]. The animals were killed 1 day after the last exposure and blood was taken by heart puncture. Three animals per exposure group were assessed for adducts of epoxybutene and of epoxybutanediol. 2.6. Humun
samples
Samples from two workers exposed to an estimated median air concentration of about 1 ppm butadiene at a Portuguese petrochemical plant and from two referents from the same plant, but with no known exposure to butadiene, were selected for this pilot study. These samples were previously assessed for adducts of epoxybutene
161. 2.7. Radioactivity
measurements
Radioactivity assays were made using a 1217 Rackbeta liquid scintillation counter (LKB, Wallac), OptiPhase ‘HiSafe’ scintillation cocktail and an internal standard kit (Wallac) for determination of counting efficiency. 2.8. Quantitative
determination
of huemoglobin
adducts
2.8. I. Sample prepurution Erythrocytes were separated from the plasma by centrifugation at 1000 x g for 10 min, and were then washed twice with saline. The cells were lysed by adding one volume of distilled water for each volume of packed cells. Globin was precipitated from the hemolysates according to Mowrer et al. [15]. Haemoglobin adducts were analysed using the N-alkyl Edman method [10,16] with a few modifications: samples of 50 mg of globin were dissolved in 1 ml of formamide. Internal standard (50 ~1 from a solution containing 0.385 mg (*H,)dihydroxypropyl-globin/ml), 40 ~1 1 M NaOH and 10 ~1 PFPITC were added and the samples were left on a rocking mixer at room temperature over night and then at 45°C for 1.5 h. Two millliters of water was added to the formamide phase and the PFPTH derivatives were extracted with diethyl ether (3 x 3 ml). The ether was evaporated under nitrogen. The samples were dissolved in 2 ml toluene and washed once with 1 ml 0.1 M sodium carbonate and once with 1 ml water. The sodium carbonate and water phases were pooled and extracted with ethyl ether (2 x 3 ml). The ether extract was evaporated to dryness, dissolved in 2 ml toluene and washed with 0.1 ml water. The toluene extracts were pooled in a new test-tube and evaporated under nitrogen at 50°C. The samples were acetylated with triethylamine, 25% (v/v), and acetic
H.L.
PPrez et al. / Chemieo-Biological
Interactions
105 (1997) 181p198
187
anhydride, 25% (v/v), in acetonitrile for 15 min essentially as described by Paulson and Lindberg [17] and evaporated under nitrogen at 30°C. The dry samples were dissolved in 3 ml of pentane and washed with 2 ml 60% (v/v) methanol/water. The pentane phase was transferred to a new test tube and was evaporated under nitrogen at 30°C. Before GC/MS-MS analysis the samples were dissolved in 50 ~1 of toluene. Haemoglobin adducts produced by the direct reaction of terminal carbon of epoxybutene with N-terminal valine [N-(2-hydroxy-3-butenyl)valine] were measured in the butadiene-exposed rats using the method described by Osterman-Golkar et al. [6]. 2.9. Muss-spectrometric
analysis
The samples and standards were analysed by GC/MS-MS on a Finnigan TSQ 700 mass spectrometer, using negative ion chemical ionisation (NCI). The operating procedures for the gas chromatograph were: helium as a carrier gas at constant gas pressure 8 psi (55 kPa); temperature programming 1 min at lOO”C, lO”C/min to 320°C and then 320°C for 7 min. The column used was a 30 m SE-54 (0.32 mm i.d., 1.0 pm phase thickness) fused silica capillary column from Alltech, Deerfeld, IL. The operating procedures for the mass spectrometer were: methane reagent gas at an ion source pressure of 4.8 torr (640 Pa), ion source temperature 150°C offset 30.0 V, ionisation energy 70 eV. Argon was used as collision gas at a pressure of 1.06 mtorr (0.14 Pa). Two ~1 of the samples, in toluene, were injected on column. The retention times were about 22 min and 45 s for the analyte and 20 min 35 s and 20 min and 42 s (double peak) for the internal standard. The daughter ions, 409, 367 and 59, respectively, formed from m/z 451 ([M-103]-) of the analyte and the daughter ions, 342, 300 and 284, respectively, formed from m/z 384 ([M-103] -) of the internal standard, were measured. 2.10. Calibration Control globin was dissolved in formamide to give a concentration of 10 mg/ml. Defined amounts of trihydroxybutylValGlyGly ethyl ester in the range 0, l-80 pmol were mixed with 1 ml control globin solution. Internal standard globin containing 3.85 pmol (2H,)dihydroxypropylvaline was added and the derivatisations were carried out as described above. The calibration curve was linear over the range of concentrations studied.
3. Results and discussion 3.1. NMR The product diepoxybutane)
prepared by reaction of epoxybutanediol (from hydrolysate with L-valine was studied by NMR. ‘H NMR of the product,
of 6
188
N.L.
PPrez et al. /Chew&o-Biological
Interactions
105 (1997) 181-198
Fig. 2. N-(2,3,4-Trihydroxybutyl)valine
0.96-1.03 (6h, m, H3), 2.18-2.28 (lh, m, H2), 3.14-3.24 (2h, m, H4), 3.48-3.54 (lh, m, Hl), 3.58-3.68 (3h, m, H6 and H7) and 3.664.2 (lh, m, H5). Because of the diastereomers the signals are split to multiples and a range of chemical shifts are reported. However, there was no indication of the positional isomer (i.e. adduct formed by reaction at carbon 3 of 3,4-epoxybutane-1,2-diol) present in the sample. This is in accordance with preferential nucleophilic attack at the terminal carbon of several related epoxides, such as propylene oxide [18], epichlorohydrin [1 I] and glycidyl ethers [19]. The structure of the epoxybutanediol product and the numbering of the protons are presented in Fig. 2. (ppm):
3.2. Preparation of the N-(trihydroxybutyljvaline
PFPTH derivative
As described previously for N-(dihydroxypropyl)valine [1 11,several modifications of the original N-alkyl Edman method [10,16] were required to make analysis of the N-(trihydroxybutyl)valine adduct possible. These modifications were: silanisation of all glass ware used in the procedure, addition of water to the reaction mixture (in formamide) before extraction of the alkylvaline-PFPTH and finally acetylation of the PFPTH derivative. Model experiments with radiolabelled N-(trihydroxybutyl)valine, and measurement of the yield in each step of the procedure described by Hindso Landin et al. [11], showed considerable losses of the N-(trihydroxybutyl)valine-PFPTH during the rather extensive washing of the toluene extract with water and sodium carbonate. The partition coefficient of N-(trihydroxybutyl)valinePFPTH for toluene/water was estimated as approximately 6. The number of washings and the volumes of water were therefore reduced to a minimum (see Section 2). 3.3. GCIMS The synthetic N-(triacetoxybutyl)valine-PFPTH was analysed by negative ion chemical ionisation GC/MS. The derivative gave one dominating peak in the chromatogram with a fragmentation pattern similar to those of several other alkylvaline-PFPTHs [20-221 (Fig. 3). The dominating fragments, HI/Z 534 ([M-20] -) and m/z 451 ([M-103] -), correspond to loss of HF and F + (CH,),(CH),CO from the valyl residue, respectively. The fragment m/z 323 corresponds to loss of the whole triacetoxybutyl substituent and the minor fragment MT/Z 225 corresponds
H.L. Pgrez et al. /Chemico-Biological
189
Interactions 105 (1997) 181-198
I
I
534
loo-
4
. E+07 1.66
I
80-
5J-
40-
20323 100
Fig. 3. Negative ion chemical tafluorophenyl-Z-thiohydantoin.
414
554
J&J\....,6co--+7
,..'.."“I. l..;.!..,L_
200
ionisation
mass spectrum
of I-(2,3,4-triacetoxybutyl)-5-isopropyl-3-pen-
to loss of C,F,NCS. Daughter ions of the fragment m/z 451 (Fig. 4) were used for analyses by GC/MS-MS. The GC/MS-MS ion chromatograms of samples from butadiene, epoxybutene or epoxybutanediol-exposed animals showed an additional peak with a slightly shorter GC retention time (22 min 35 s and 22 min 45 s for the
4
I 100
60
E+06 1.62
59
60
451
367
40
20
222
263 28p307
100
m
Fig. 4. Collision induced daughter spectrum propyl-3-pentafluorophenyl-2-thiohydantoin.
300
349 1
I-'"
400
of m/z 451 ([M-103] -) from
‘--'I
!xnJ
l-(2,3,4-triacetoxybutyl)-5-iso-
190
H.L.
Ptrez
et al. / Chemico-Biological
Interactions
105 (1997) 181~ 198
E+O5 2.312 Intensity
lnkxnal standard r.,..:
AA ., . . . . . . . . . . . . . . . .._ .,~ ., . . .& .., 20:oo
2050
21:40
22:30
24:lO
23:20
Time
-E+05 1 .949
Intensity
lntemal standard
20:oo
2050
21:40
22
23:20
24:lO
Time
Fig. 5. MS-MS ion chromatogram of a sample from (a) a diepoxybutane-exposed tanediol-exposed
rat. Retention
rat (b) an epoxybutimes 22 min 35 s for peak I and 22 min 45 s for peak II, respectively.
first and second peak, respectively) (Fig. 5). The two products gave the same daughter ions and were named I and II based on their GC retention times. The analyses of peak I were not calibrated but the quantities of this adduct were estimated directly from the peak areas of daughter ions of m/z 451 assuming equal response of type I and type II products.
H.L.
P&e= et al. /Chemico-Biological
3.4. The stereochemistry
of epoxybutanediol
Interactions
191
105 (1997) 181-198
adduct formation
(Fig. 6)
Presumably the peaks I and II (Fig. 5) correspond to different stereoisomeric forms of the trihydroxybutylvaline-PFPTH. Peak II was dominating, with only traces of peak I, in samples from diepoxybutane-exposed animals (Fig. 5a) and in samples of reference compounds prepared via hydrolysis of diepoxybutane. Both peaks, I and II, were seen in samples from animals exposed to epoxybutanediol (Fig. 5b), epoxybutene or butadiene. The formation of different stereoisomeric forms of epoxybutanediol and of its adducts, respectively, is explained in Fig. 6. The commercial diepoxybutane [( + )-butadiene diepoxide] is a racemic mixture of
-3-l OH
OH 1
0
1
iH
==JJOH OH
0:
V 2
&
OH
z OH
3
2 ‘:. OH +
4
OH
k +R
i)H
OH la
OH
OH 2a
Fig. 6. Formation of different stereoisomeric ( + )-diepoxybutane or from 3-butene-1,2-dial, haemoglobin (R = N-terminal valine).
3a
4a
forms of adducts of epoxybutanediol, formed from respectively, in reaction with N-terminal valine in
192
H.L.
PPrez et al. /Chemico-Biological
interactions
105 (1997) 181-198
(SS) and (RR) stereoisomers. Apparently hydrolysis of this mixture, in vivo as well as in vitro, mainly gives the diolepoxides assigned 1 and 2, whereas epoxybutanediol, prepared from 2-butene-1,4-diol [9] or formed metabolically from epoxybutene, contains the diolepoxide structures l-4. Using our analytical conditions the acetylated PFPTH derivatives of structures l(a) and 2(a) (adducts of 1 and 2) and 3(a) and 4(a) (adducts of 3 and 4) are resolved as two peaks. Analysis of ValGlyGly ethyl ester, treated in vitro with epoxybutanediol prepared from butenediol, gave, as expected, both peaks I and II. Using a different type of derivatives (trimethylsilylated PFPTHs), Rydberg et al. [23] observed two peaks of trihydroxybutylvaline in the chromatogram of a sample from haemoglobin treated in human erythrocytes with rat-diepoxybutane. 3.5. Adduct
levels in rats and humans
Table 1 presents N-(2,3,4-trihydroxybutyl)valine adduct levels in haemoglobin from male Sprague Dawley rats administered the three reactive metabolites, epoxybutene, epoxybutanediol or diepoxybutane by intraperitoneal injection. The HBI, given as pmol adduct/g per pmol of alkylating agent, has been calculated in order to facilitate an intercomparison of the various agents in this study and a comparison with other compounds with respect to their capacity to alkylate macromolecules in vivo. The HBI values are related to formation of epoxybutanediol adducts of type II; the HBI for formation of epoxybutanediol adducts of both types, I + II, may then be estimated using the adduct ratio (I + II)/II. Treatment of rats with diepoxybutane gave mainly adducts of type I (Fig. 5a), and treatment with epoxybutanediol or epoxybutene or exposure to butadiene gave adducts of type I and II (Fig. 5b). This observation suggests that the in vivo oxidation of butenediol is not stereospecific but gives a mixture of the diolepoxides l-4 (Fig. 6). The single peak formed by diepoxybutane may have been formed by direct formation of adducts of diepoxybutane followed by hydrolysis of the epoxide bond or by hydrolysis to epoxybutanediol followed by formation of adducts. The contribution of each pathway is not known, however, the latter is presumably dominating. Studies of the reactions of diepoxybutane with valinamide in vitro show that the primary reaction product of diepoxybutane with the valine-NH, gives a ring structured product in a reaction which is fast compared with hydrolysis of the second epoxide ring [23]. Thus, trihydroxybutylvaline is formed only in small amounts after direct reaction of diepoxybutane with N-terminal valine in haemoglobin. Epoxybutanediol and diepoxybutane were administered to the animals at two doses for each compound. The resulting adduct data (although limited to analysis of a few animals) were compatible with linear dose-response relationships for epoxybutanediol adduct formation. The HBI for epoxybutanediol, about 3.7 for adduct II and accordingly about 10 (2.7 x 3.7 pmol/g per pmol/kg) for adducts I + II, is comparable to the HBI, about 16, for ethylene oxide in the rat as estimated from data given by Walker et al. [24], demonstrating a similar capacity of the two compounds (epoxybutanediol and ethylene oxide) to alkylate nucleophilic
Diepoxybutane
Epoxybutene Epoxybutanedial
Agent
1117 289 578 194 387
60 16.7 33.4
0
Dose (pmol/kg)
reactive metabohtes
1900cd, 2700’ 1800 k 283 2800 & 100
83+13 8604, 1100’
2 k 2.8
Adduct level + S.D.; adduct II (pmol/g)”
of rats administered
78.3 30
0
Dose (mg/kg)
levels in haemoglobin
1
4+0.5 2.7 k 0.1
epoxybutanediol,
or
4.0 9.3 7.2
0.07 3.4
HBI; adduct II (pmol/ g per PmoVkg)
epoxybutene,
Adduct ratio f S.D.; [(I+ II)/IIl”
of butadiene:
a See Figs. 5 and 6. I and II are different stereoisomeric forms of the adduct. b Groups of two to three rats were treated by i.p. injection and blood samples were collected after 48 h unless otherwise stated. ’ Adduct level of an individual rat. d The animals were sacrificed 24 h after treatment,
Rats, Sprague Dawley, i.p.b
Species, strain, route of exposure
Table I N-(2,3,4-Trihydroxybutyl)valine diepoxybutane
194
H.L.
P&r
et al. /Chemico-Biological
Interactions
105 (1997) 181-198
Table 2 N-(2,3,4-Trihydroxybutyl)valine tion and of human subjects
adduct levels in haemoglobin of rats exposed with occupational exposure to butadiene
Species,
level
Rats,
strain
Wistarh
(ppm)
Butadiene
Adduct level & S.D.; adduct 11 (pmol/g)’
0 50 200 500
912 680 i 230 119Ok275’ 585 * 230
Humans’ Controls Occupationally exposed
Around
1”
Adduct ratio [(I + lI):II]”
3.2 f 0.2d
to butadiene
by inhala-
HBI; adduct (pmol/g h)
II
per ppm x
0.5 0.2 0.04
1.8 (0.05)’ 3.3 (0.02)’ 10 (0.15)’ 14 (0.20)’
* See Figs. 5 and 6. 1 and II are different stereoisomeric forms of the adduct. ‘Three animals per group. The rats were exposed to butadiene by inhalation 6 h/day for 5 days and killed 1 day after the last exposure. ‘Two animals only. One high value excluded (outside the calibration curve). ’ Average ( f SD.) calculated from adduct data in animals exposed to butadiene at 50, 200 or 500 ppm. ’ Adduct levels of individual subjects are indicated. r Values for adducts of epoxybutene (N-(2-hydroxy-3-butenyl)valine) are given within parenthesis. g Median concentration of butadiene estimated as 1 ppm [6].
sites in vivo. The efficiency of diepoxybutane with respect to formation of epoxybutanediol-hemoglobin adducts (HBI around 8) is approximately the same as that of the epoxybutanediol itself. Table 2 presents N-(2,3,4_trihydroxybutyl)valine adduct levels in haemoglobin from male Wistar rats exposed to butadiene and from a few human subjects with occupational exposure to butadiene, respectively. Studies on tissue concentrations of epoxides (epoxybutene and diepoxybutane) in mice and rats during exposure to butadiene [25] demonstrate very low blood levels of diepoxybutane in the rats. Therefore, epoxybutanediol, formed by oxidation of butenediol, is probably the main precursor of epoxybutanediol adducts in butadiene-exposed rats. The HBI, expressed as pmol trihydroxybutylvaline/g globin per ppm x h, decreases with butadiene concentration from 0.5 at 50 ppm to 0.04 at 500 ppm demonstrating an effect of dose rate. Since the formation of epoxybutene adducts is essentially linearly related to exposure level ~ the epoxybutene adduct levels are 0.6, 21, 88 and 180 pmol N-(2-hydroxy-3_butenyl)valine/g’ globin in the rats exposed to 0, 50, 200 and 500 ppm, respectively ~ the non-linearity with concentration is probably related to the metabolism of butenediol. A dose rate effect may in part explain the low HBI-value observed for the animals administered an acute dose of epoxybutene (Table 1). ’ This adduct
constitutes
approximately
60’%) of the total adducts
with N-terminal
valine [5]
H.L.
Ptrez
et al. /Chemico-Biological
Interactions
105 (1997) 181-198
195
The HBI for epoxybutanediol adduct formation from butadiene (3.2 x 0.4 pmol/ g per ppm x h) at the lowest air concentration, 50 ppm, is comparable to the HBI for ethylene oxide adducts in ethylene exposed rats (estimated as about 1 [16]). The analyses of epoxybutanediol adduct levels in humans were limited to a few samples from butadiene-exposed workers and controls investigated previously with respect to levels of adducts of the primary butadiene metabolite epoxybutene [7] (Table 2). The epoxybutene adducts measured were those formed by reaction of the The adduct terminal carbon in epoxybutene, i.e., N-(2-hydroxy-3-butenyl)valine. levels of epoxybutanediol (type II) were much higher (about 70-fold) than the adduct levels of epoxybutene measured in the same subjects. The presence of adducts of both epoxybutene and epoxybutanediol in the two control subjects supports the notion that these subjects are exposed to butadiene, possibly due to occasional visits to contaminated areas of the plant [6]. Since epoxybutanediol is downstream compared to epoxybutene in the metabolic pathway of butadiene (Fig. 1) the high adduct ratio (epoxybutanediol/epoxybutene) indicates an efficient formation of epoxybutanediol and a considerably slower elimination of this epoxide than of epoxybutene. Sabourin et al. [26] examined the urinary excretion of mercapturic acids in mice, rats, hamsters, and monkeys. All four species produced two metabolites, Ml [ 1,2-dihydroxy-4-(N-acetylcysteinyl)butane] and M2 [ 1-hydroxy-2-(N-acetylcysteinyl)-3-butene or 2-hydroxy-1-(N-acetylcysteinyl)-3-butene]. M2 is formed by conjugation of glutathione (GSH) with epoxybutene, and Ml appears to be formed by GSH conjugation with butenediol, the product of hydrolysis of epoxybutene. The butenediol conjugation is thought to occur via the intermediary formation of a reactive a$-unsaturated carbonyl compound. When comparing the four species, the ratio of Ml to the sum of Ml + M2 was linearly related to hepatic epoxide hydrolase activities for each species. Bechtold et al. [27] showed that Ml is the predominant metabolite in urine from butadiene-exposed workers. The ratio of epoxybutanediol adducts (type II) to epoxybutene adducts [N-(2-hydroxy-3-butenyl)valine, representing attack on the terminal carbon of epoxybutene] in the butadiene-exposed rats is 4-26 in the individual rats of this study (to be completed and published in another context) as compared to about 70 in the human subjects. The large amounts of trihydroxybutyl adducts as compared to epoxybutene adducts in haemoglobin and the high urinary concentrations of Ml as compared to M2 suggest that the hydrolysis pathway of epoxybutene is important in both species. The strong focus on epoxybutene and diepoxybutane in previous discussions on the toxicity of butadiene may, at least in part, be related to the techniques which have been used to study the formation of reactive metabolites in vitro and in vivo. These techniques generally involve extraction of the microsomal samples, tissue homogenates, or blood with an organic solvent such as o-xylene [28] or methylene chloride [29,30]. The hydrophilic epoxybutanediol is not recovered by these extractions but stays in the water phase. Cheng and Ruth [31] solved the problem by evaporation of the media to dryness (epoxybutanediol is not volatile) followed by extraction of the residue with ethyl acetate overnight. These authors incubated rat
196
H.L.
PPrez et al. /Chemico-Biological
Inteructions
105 (1997) 181-198
liver microsomes in the presence of NADPH and found that butenediol and epoxybutanediol were the major products. In conclusion, this study shows that measurements of epoxybutanediol adducts with N-terminal valine in haemoglobin can be used for dosimetry studies in experimental animals and to determine individual exposures in human epidemiological studies. Epoxybutanediol gives substantially higher haemoglobin adduct levels than epoxybutene in both butadiene-exposed rats and humans suggesting an important role of epoxybutanediol in the toxicity of butadiene.
Acknowledgements We thank Mr Vlado Zorcec and MS Annika Gustavsson for mass spectrometric analyses and MS Aila Suutari for technical assistance. This study was supported by the European Commission, project EV5V-CT94-0543. The NMR spectra are available on request.
References [l] R.L. Melnick, J. Huff, 1,3-Butadiene: toxicity and carcinogenicity in laboratory animals and in humans, Rev. Environ. Contam. Toxicol. 124 (1992) I1 1~ 144. [2] S. Osterman-Golkar, A. Kautiainen, E. Bergmark, K. Hakansson, J. Maki-Paakkanen, Hemoglobin adducts and urinary mercapturic acids in rats as biological indicators of butadiene exposure, Chem.-Biol. Interact. 80 (1991) 291-302. [3] O.E. Albrecht, J.G. Filser, H.D. Neumann, Biological monitoring of 1,3-butadiene: species differences in haemoglobin binding in rat and mouse, in: M. Sorsa, K. Peltonen, H. Vainio, K. Hemminki (Eds.), Butadiene and Styrene: Assessment of Health Hazards. IARC Scientific Publications, vol. 127, International Agency for Research on Cancer, Lyon, 1993, pp. 1355142. [4] S.M. Osterman-Golkar, J.A. Bond, J.B. Ward Jr., M.S. Legator, Use of haemoglobin adducts for biomonitoring exposure to 1,3-butadiene, in: M. Sorsa, K. Peltonen, H. Vainio, K. Hemminki (Eds.), Butadiene and Styrene: Assessment of Health Hazards. IARC Scientific Publications, vol 127. International Agency for Research on Cancer, Lyon, 1993, pp. 1277134. [5] N.Yu. Tretyakova, Y. Lin, P.B. Upton, R. Sangaiah, J.A. Swenberg, Macromolecular adducts of butadiene, Toxicology 113 (1996) 70-76. [6] S. Osterman-Golkar, K. Peltonen, T. Anttinen-Klemetti, H. Hindso Landin, V. Zorcec, M. Sorsa, Haemoglobin adducts as biomarkers of occupational exposure to 1.3-butadiene, Mutagenesis 10 (1996) 145-149. [7] S. Osterman-Golkar, J.A. Bond, Biomonitoring of 1,3-butadiene and related compounds, Environ. Health Perspect. 104 (Suppl. 5) (1996) 907-915. [8] F. Granath, 0. Rohlen, G. Goransson, L. Hansson, A.L. Magnusson. M. Tornqvist, Relationship between dose in vivo of ethylene oxide and exposure to ethene studied in exposed workers, Hum. Exp. Toxicol. 15 (1996) 8266833. [9] J. Llhdetie, K. Peltonen, T. Sjoblom, Germ cell mutagenicity of three metabolites of 1,3-butadiene in the rat: induction of spermatid micronuclei by butddiene mono-, di- and diolepoxides in vivo, Environ. Mol. Mutagen., in press. [IO] M. Tornqvist, J. Mowrer, S. Jensen, L. Ehrenberg, Monitoring of environmental cancer initiators through hemoglobin adducts by a modified Edman degradation method, Anal. Biochem. 154 (1986) 2555266.
H.L. [1 1] H. Hindso
Perez et al. / Chemico-Biological
Landin,
S. Osterman-Golkar,
drin by hemoglobin adducts, Anal. [12] L. Ehrenberg, S. Hussain, Genetic l&113. [13] J. Lahdetie,
J. Grawe,
polychromatic
Flow
erythrocytes
V. Zorcec,
M. Tiirnqvist,
analysis
Biomonitoring epoxides,
of micronucleus
1,2;3,4_diepoxybutane,
estimation
of petrochemical M. Tornqvist, for monitoring
11 (1986) 215-231. [16] M. Tornqvist, A. Kautiainen, to gasoline [17] J. Paulson, matogr.
emissions, S. Jensen, occupational R.N.
of epichlorohy-
Mutat.
induction
Res. 86 (1981)
in rat bone
3,4-epoxy-I-butene
and
3,4-diol, Cytometry, in press. [14] M. Sorsa, K. Peltonen, D. Anderson, N.A. Demopoulos, H.G. Neumann, Assessment of environmental and occupational exposures to butadiene [15] J. Mowrer, hemoglobin
191
10.5 (1997) 181-198
Biochem. 240 (1996) l-6. toxicity of some important
cytometric by
Interactions
marrow
1,2-epoxybutdne-
S. Osterman-Golkar, as a model for risk
Mutagenesis 11 (1996) 9- 17. L. Ehrenberg, Modified Edman degradation applied to exposure to alkylating agents, Toxicol. Environ. Chem.
Gatz,
L. Ehrenberg,
Hemoglobin
adducts
in animals
and diesel exhausts, 1 Alkenes, J. Appl. Toxicol. 8 (1988) 1599170. C. Lindberg, Increasing thermospray response for cortisol by derivatization,
exposed J. Chro-
554 (1991) 1499154.
[18] K. Svensson, S. Osterman-Golkar, Kinetics of metabolism of propene and covalent macromolecules in the mouse, Toxicol. Appl. Pharmacol. 73 (1984) 363-372. [19] H. Licea Perez, K. Plna, S. Osterman-Golkar, Dosimetry of glycidyl tion of haemoglobin adducts, Chem.-Biol. Interact. 103 (1997) l-16. [20] P. Rydberg, N-substituted
ethers
Golkar, Monitoring occupational blood, Stand. J. Work Environ.
V. Zorcec,
J. MCki-Paakanen,
H.
exposure to styrene by hemoglobin Health 19 (1993) 255-263.
to
in mice by quantifica-
B. Liining, C.A. Wachtmeister, M. Tornqvist, Synthesis and valines and their phenyland pentafluorophenyl-thiohydantoins,
Stand. 47 (1993) 813-817. [21] A. Christakopoulos, E. Bergmark,
binding
characterization of Acta Chem.
Norppa,
adducts
S. Osterman-
and metabolites
in
[22] S.M. Osterman-Golkar, J.P. MacNeela, M.J. Turner, V.E. Walker, J.A. Swenberg, S.J. Sumner, N. Youtsey, T.R. Fennell, Monitoring exposure to acrylonitrile using adducts with N-terminal valine in hemoglobin, Carcinogenesis 15 (1994) 2701-2707. [23] P. Rydberg, A.L. Magnusson, V. Zorcec, F. Granath, M. Tornqvist, valines in hemoglobin from butadiene metabolites, Chem.-Biol. Interact. [24] V.E. Walker, J.P. MacNeela, J. Swenberg, ethylene oxide: formation and persistence repeated
exposure
Adducts to N-terminal 101 (1996) 193-205.
M.J. Turner Jr., T.R. Fennell, Molecular dosimetry of of N-(2-hydroxyethyl)valine in hemoglobin following
of rats and mice, Cancer
Res. 52 (1992) 4320-4327.
[25] J.R. Thornton-Manning, A.R. Dahl, W.E. Bechtold, W.C. Griffith Jr., R.F. Henderson, Disposition of butadiene monoepoxide and butadiene diepoxide in various tissues of rats and mice following a low-level inhalation 1731. [26] P.J. Sabourin, L.T. Burka, W.E.
exposure Bechtold,
to A.R.
1,3-butadiene, Dahl,
M.D.
Carcinogenesis Hoover,
16 (1995)
I.Y. Chang,
R.F.
1723Hender-
son, Species differences in urinary butadiene metabolites; identification of 1,2-dihydroxy-4-(Nacetylcysteinyl)butane, a novel metabolite of butadiene, Carcinogenesis 13 (1992) 1633-1638. [27] W.E. Bechtold, M.R. Strunk, I.Y. Chang, J.B. Ward Jr., R.F. Henderson, Species differences urinary
butadiene
metabolites:
comparisons
of metabolite
ratios
between
in
mice, rats, and humans,
Toxicol. Appl. Pharmacol. 127 (1994) 4449. [28] R.J. Duescher, A.A. Elfarra, Human liver microsomes are efficient catalysts for 1,3-butadiene oxidation: Evidence for major roles by cytochrome P450 2A6 and 2E1, Arch. Biochem. Biophys. 311 (1994) 342-349. [29] M.W.
Himmelstein,
of l$butadiene Carcinogenesis
M.J. Turner,
B. Asgharian,
and butadiene epoxides 15 (1994) 1479- 1486.
J.A. Bond,
Comparison
in mice and rats exposed
of blood
to 1,3-butadiene
concentrations by inhalation,
198
H.L. PCrez et al. /Chemico-Biological
Interactions 105 (1997) 181-198
[30] M.W. Himmelstein, B. Asgharian, J.A. Bond, High concentrations of butadiene epoxides in livers and lungs of mice compared to rats exposed to 1,3-butadiene, Toxicol. Appl. Pharmacol. 132 (1995) 281-288. [31] X. Cheng, J.A. Ruth, A simplified methodology for quantitation of butadiene metabolites: Application to the study of 1,3-butadiene metabolism by rat liver microsomes, Drug Metab. Dispos. 21 (1993) 12lLl24.