Comparison of ethylene, propylene and styrene 7,8-oxide in vitro adduct formation on N-terminal valine in human haemoglobin and on N-7-guanine in human DNA

Article abstract of DOI:10.1016/S1383-5718(98)00106-5

Epoxides react at various nucleophilic sites in macromolecules such as haemoglobin and DNA. To study the reaction rate constants of ethylene oxide (EO), propylene oxide (PO) and styrene 7,8-oxide (SO) towards two of these positions, i.e., the N-terminal v

Full text of DOI:10.1016/S1383-5718(98)00106-5

Ž
.
Mutation Research 418 1998 21–33
Comparison of ethylene, propylene and styrene 7,8-oxide in vitro
adduct formation on N-terminal valine in human haemoglobin
and on N-7-guanine in human DNA
)
W. Pauwels, H. Veulemans
Laboratory for Occupational Hygiene and Toxicology, Katholieke UniÕersiteit LeuÕen, KapucijnenÕoer 35r6, 3000 LeuÕen, Belgium
Received 17 March 1998; revised 11 June 1998; accepted 21 July 1998
Abstract
Epoxides react at various nucleophilic sites in macromolecules such as haemoglobin and DNA. To study the reaction rate
Ž
.
Ž
.
Ž
.
constants of ethylene oxide EO , propylene oxide PO and styrene 7,8-oxide SO towards two of these positions, i.e., the
N-terminal valine in haemoglobin and N-7-guanine in DNA was the central aim of this investigation. These two reactive
sites are the most studied haemoglobin and DNA adducts, respectively. Further attention, therefore, was also paid to the
applicability in vivo of the in vitro determined reaction constants. The determination of the second-order rate constants
y
1
y1
w
Ž
.
Ž . x
between EO and PO and N-terminal valine in Hb 2.7 l mol Hb h
and 1.0 l mol Hb h , respectively were consistent
y
3
y1
w
Ž
.
with the literature values. The constants for the reaction with N-7-guanine 16=10
l mol DNA nucleotide h
Ž
l mol DNA nucleotide h , respectively were lower than previously published values, probably due to
and
3
y1
7.7=10y
.
x
differences in the methodology used. The use of the in vitro obtained values to model the in vivo situation lead to a
consistent picture for EO and PO. In contrast, for SO the in vitro ratio between the adduct formation on N-terminal valine
y
1
y
3
y1
w
Ž
.
x
w
Ž
.
x
was about two orders of magnitude
1.5 l mol Hb h
and N-7-guanine 0.71=10
l mol DNA nucleotide h
greater than for the in vivo situation. This was probably due to a lower than expected reactivity of SO towards N-terminal
valine in vivo. Further research is needed to elucidate whether the use of SO in vitro, contrasting with the in vivo
experiments in which SO was metabolically formed from styrene, could entail an explanation for this discrepancy.
Concerning the methodological part, the use of dipeptide standards to replace the alkylated globins as standard lead to an
improvement of the method. Especially the commercial availability of the standards, their stability and accurately known
adduct content will make them to the standards of choice in the future. q 1998 Elsevier Science B.V. All rights reserved.
Keywords: Epoxides; Haemoglobin adducts; DNA adducts; Edman degradation
1. Introduction
The epoxides studied are the derivatives oxides
which have a widespread use as monomers in the
chemical industry. Moreover, ethylene oxide and to a
certain extent propylene oxide, also have a number
of specific applications. Especially, their use in the
gaseous form as a disinfectant, sterilising agent,
fumigant or insecticide leads to relatively high expo-
sures. The highest exposures for styrene are observed
Ž
.
of ethylene, propylene and styrene, respectively,
)
Corresponding author. Fax: q32-16-33-69-97; E-mail:
hendrik.veulemans@med.kuleuven.ac.be
1383-5718r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved.
Ž
.
PII: S1383-5718 98 00106-5

(
)
22
W. Pauwels, H. VeulemansrMutation Research 418 1998 21–33
in the production of fiberglass-reinforced plastics,
where manual application techniques are used.
The overall evaluation of the carcinogenicity of
these three chemicals and their epoxides is reconsid-
mainly as metabolite of ethylene arising from en-
dogenous production, from environmental sources
Ž
.
plant hormone and combustion product or from
w x
cigarette smoking 6 .
w x
ered by the IARC 1 and is summarized in Table 1.
The reaction rate constants of the three epoxides
towards N-terminal valine in haemoglobin has been
determined in different in vitro experiments using
The metabolism of the alkenes could lead to the
formation of the corresponding reactive epoxides.
These epoxides are able to bind covalently to nucleo-
w
x
human blood 8–10 . The reaction rate constants of
EO and PO towards N-7-guanine in DNA, so far,
has been determined in a solution of calf thymus
Ž
philic sites in proteins and DNA reviewed in Refs.
x.
w
1,2 . The measurement of DNA and haemoglobin
w
x
adducts is a useful tool for monitoring exposure to
electrophilic agents. Binding with DNA gives some
information about the potential risk since DNA
adduct formation is in many cases the initial step in
DNA 11,9 . The comparison of the adduct formation
by the three epoxides on N-terminal valine in
haemoglobin and N-7-guanine in DNA in whole
blood was the primary objective of this study. The
second-order reaction rate constants between the three
epoxides and either N-terminal valine in haemoglobin
or N-7-guanine in DNA were calculated. Further
attention was also paid to whether the in vitro deter-
mined reaction rate constants can be used in vivo by
comparison of the ratio of the in vitro reaction rate
constants towards N-terminal valine and N-7-guanine
with the ratio between the adduct levels observed in
a number of animal experiments.
w x
chemical carcinogenesis 3 . The determination of
DNA adducts has focused mainly on the N-7-ad-
ducts on guanine, since these are by far the main
Ž
DNA adducts formed with the three epoxides re-
w
x.
viewed in Refs. 1,2 . Regarding haemoglobin
adducts, up to now, only the modified Edman degra-
dation technique has found to be a practical method
for the determination of N-terminal valine adducts
formed by electrophilic agents. The method devel-
w x
oped by Tornqvist et al. 4 provides a rapid, sensi-
Ring tests for the determination of N-terminal
valine adducts of EO and SO indicated that the use
of different standards for the Edman degradation
method leads to differences in absolute adduct levels
¨
tive and specific technique for the determination of
adducts on the N-terminal valine in haemoglobin.
For EO, a large number of studies are available.
This compound, therefore, has been used as a model
w
x
measured 12,13 . A secondary objective of this
study, therefore, was to determine if the synthetic
N-alkylated-dipeptide standards, which became re-
cently available, could be used to replace the former
w
x
for dose monitoring and risk estimation 5–7 . More-
over, EO may be an important contributor to the
collective genotoxic dose to human populations,
Table 1
Ž
w x.
Evaluations of the carcinogenicity of ethylene, propylene, styrene and their corresponding epoxides IARC 1
Agent
Degree of evidence of carcinogenicity
Overall evaluation of
carcinogenicity for humans
Human
Animal
Ethylene
Ethylene oxide EO
Propylene
Propylene oxide PO
Styrene
Styrene 7,8-oxide SO
inadequate
limited
inadequate
inadequate
inadequate
inadequate
inadequate
sufficient
inadequate
sufficient
limited
3
1
3
2B
2B
2A
Ž
.
Ž
.
Ž
.
sufficient
Group 1: human carcinogen.
Group 2A: probable human carcinogen.
Group 2B: possible human carcinogen.
Group 3: not classifiable as to its carcinogenicity to humans.

(
)
W. Pauwels, H. VeulemansrMutation Research 418 1998 21–33
23
standards, prepared through reaction between
haemoglobin and the epoxides. In this way problems
with absolute calibration of standards could be
avoided in the future.
ml PO or 12 ml SO. After 5 h stirring at 378C the
reaction mixture was diluted with 100 ml methanol.
Adducts and unreacted guanosine were precipitated
by the addition of 900 ml diethyl ether. After filtra-
tion the precipitate was resuspended in methanol.
This allowed the removal of the unreacted guanosine
by filtration. After evaporation of the solvent under
N2 the residue was dissolved in 0.1 N HCl and
heated to 1108C for 30 min. After cooling to room
temperature, neutralization of the solution with KOH
resulted in precipitation of the adduct. After recrys-
2. Materials and methods
2.1. Chemicals
Ž
.
Pentafluorophenyl isothiocyanate PFPITC was
Ž
.
Ž
.
obtained from Fluka Buchs, Switzerland and was
used without further purification. Formamide ana-
lytical grade was extracted with pentane before use.
Ethylene oxide 50 mgrml in methanol was ob-
tained from Supelco Bellefonte, USA , 1,2-pro-
pylene oxide )98% was delivered by Merck
Overijse, Belgium and styrene 7,8-oxide 97% by
tallization only the N-7- 2-hydroxyethyl adduct
turned out to be pure by GC–MS. The propylene and
styrene oxide adducts contained still some guanine
and needed to undergo an additional purification step
on a preparative Sep-Pakw Vac 12 cm3 C18-cartridge
Ž
.
Ž
.
.
Ž
Ž
.
Ž
.
Waters Chromatography Division . The cartridges
Ž
.
Ž
.
were conditioned with MeOH, followed by a flush-
ing with bidistilled water. The samples were dis-
solved in slightly acidified water and brought on the
columns. After washing with water and 10% MeOH,
the adduct was eluted from the cartridges with 40%
MeOH. The procedure was repeated until the product
turned out to be pure on GC–MS.
Ž
.
Aldrich Bornem, Belgium . The dipeptide standards
Ž
.
Ž
.
N- 2-hydroxyethyl Val-Leu-anilide ) 98% and
Ž
.
Ž
.
N-DL- 2-hydroxypropyl Val-Leu-anilide )95% for
the determination of the N-terminal valine adducts of
EO and PO, respectively, were purchased from
Ž
.
Bachem Biochemica Heidelberg, Germany .
Haemoglobin containing 192 pmol N-hydroxyethyl-
valinermg Hb, EO-Hb, was a kind gift from N. Van
Sittert, Shell, The Netherlands. The standard for the
determination of SO adducts, SO-Hb, and the inter-
(
)
2.1.2. Synthesis of N-7- 2-hydroxy-2-phenyl ethyl-
) [
(
]
guanine N-7-b-SO-adduct 15
Ž
.
Guanosine 3 g was suspended in ethanol: H2O
. Ž
Ž
.
w2
x
Ž
.
nal standard I.S. used for all analyses, H8 SO-Hb,
was prepared as described in Severi et al. 14 . The
amount of SO-Hb was found to be 2.03 nmolrmg
Hb, whereas the H8 SO-Hb contained 1.73 nmol of
N- 2-hydroxy-2- H8 phenylethyl valinermg Hb.
The level of N-terminal valine SO-adduct in the
reference globin and the I.S. was determined by acid
hydrolysis of the globin, followed by the determina-
tion of the amount of alkylated valine as described in
1:1 400 ml . The mixture was heated until the
w
x
guanosine was dissolved. After addition of 4 ml
styrene oxide, the solution was stirred during 3 days
at 378C. After evaporation of the solvent, the residue
was redissolved in methanol. The unreacted guano-
sine was only partially soluble and the major part
could be removed by filtration. The solvent was
evaporated and the residue redissolved in 0.1 N HCl.
The hydrolysis of the N-7-guanine adduct from the
ribose was carried out at 1108C for 30 min. After
cooling to room temperature the mixture was brought
to pH 7 with KOH, resulting in the precipitation of
the adduct together with some unreacted guanine.
The adduct was again separated from the unre-
acted guanine through the same C18-purification
technique as described above. In contrast with the
reaction in acetic acid, which allowed to obtain pure
N-7-a-SO-adduct, this reaction resulted in the forma-
tion of a mixture of N-7-a-SO en N-7-b-SO-adduct
in a 3:2 ratio. The amount of N-7-b-SO-adduct
w2
x
x
Ž
w2
.
w
x
Severi et al. 14 .
All other chemicals and solvents were of analyti-
cal grade and used without further purification.
(
)
2.1.1. Synthesis of N-7- 2-hydroxyethyl guanine, N-7-
(
)
(
2-hydroxypropyl guanine and N-7- 2-hydroxy-1-
)
(
)
phenylethyl guanine N-7-a-SO-adduct
The synthesis of these N-7-adducts was based on
w
x
the method described by Latif et al. 15 . Guanosine
Ž
.
2.8 g was suspended portion-wise in glacial acetic
Ž
.
acid 75 ml and was treated with either 5 ml EO, 10

(
)
24
W. Pauwels, H. VeulemansrMutation Research 418 1998 21–33
Ž
.
Ž
.
could be calculated by subtraction of the amount of
N-7-a-SO-adduct from the total amount of adduct.
injector 1 ml with an inlet splitter 1:10 . Helium
was used as the carrier gas. The oven was pro-
grammed at 208C from 130 to 2008C, which was
held for an additional 7.5 min. The EO concentration
was determined from established calibration curves.
The relative concentration at 0 min was set to 100%.
2.2. Elimination of the three epoxides from blood in
Õitro
Ž
.
Ž
The initial concentration of EO 328 mM was
Ten microliters of a stock solution of PO 105 ml
Ž
.
different from the initial concentration of PO 30
in 100 ml saline was added to 10 ml freshly drawn
.
Ž
.
mM and SO 40 mM , due to reasons of sensitivity
of detection. All initial concentrations, however, were
well within the range of the linear dose–N-terminal
valine adduct relationship 0–1000 mM . For two
reasons, to determine the second-order rate reaction
constant between SO and N-7-guanine an elimina-
human blood resulting in a 30 mM initial concentra-
tion. The blood was incubated in an oil bath at 378C
and at 0, 5, 15, 30, 45, 60, 90 and 120 min intervals
aliquots of 1 ml blood were removed and immedi-
ately extracted with 1 ml ethyl acetate. The amount
of PO in the ethyl acetate layer was determined in
the same way as described for EO.
Ž
.
Ž
tion constant at a higher initial concentration 50
mM was calculated. Firstly, the determination of
.
The first-order elimination rate constant for SO
N-7-SO-guanine adducts required higher incubation
concentrations of the epoxide compared to EO and
PO. Secondly, the elimination constant of SO was
relatively high compared to the other epoxides. Only
for SO, therefore, and not for EO and PO, substantial
effects on the second-order rate reaction constants
were expected from a slower elimination at higher
human blood was estimated essentially as described
w
x
in Rappaport et al. 16 and Yeowell-O’Connell et al.
w
x
10 . Briefly, freshly drawn human blood was incu-
bated with SO at an initial concentration of 40 mM.
One milliliter aliquots were taken at 0, 5, 10, 15, 20,
30, 60, 90 and 120 min intervals and immediately
w2
x
extracted with hexane containing H8 SO as the I.S.
After drying of the hexane layer the SO concentra-
tion was determined by GC–MS. The elimination
experiment was repeated with an initial concentra-
tion of 50 mM, since for the modification of DNA in
vitro, higher incubation concentrations were re-
quired.
Ž
.
incubation concentrations see Section 3 .
The apparent first-order elimination constant ke
Ž
.
for EO was estimated as follows. All materials used
for the handling of the solutions containing EO were
kept at y208C before use because of the low boiling
Ž
.
point of EO 118C . The incubation of blood with
EO, was carried out in completely filled and tightly
closed vials. The solution of 50 mg EOrml methanol
was diluted in order to obtain a stock solution of 6.5
mgrml. To six different vials containing 1.5 ml
blood EO solution was added to a final concentration
of 328 mM EO. The blood was spiked with EO by
inserting the needle of a cooled Hamilton syringe
through the membrane, avoiding an increase of pres-
sure in the vial. After mixing, the vials were placed
2.3. Modification of haemoglobin and DNA by in
Õitro incubation of whole blood
2.3.1. Modification of haemoglobin
Again, the necessary precautions were taken to
minimalize the influence of the high volatility of EO
on the results. The solution of 50 mg EOrml
methanol was diluted in order to obtain a stock
solution of 6.5 mgrml. Sufficient amounts of this
solution were added to 1.5 ml blood, resulting in
final concentrations of 9.8, 29.5, 98.4, 295 and 984
mM.
Ž
in an oil bath at 378C. At convenient intervals 0, 15,
.
30, 60, 90 and 120 min the content of one of the
vials was extracted in a standardized way with 750
ml ethyl acetate. The ethyl acetate layer was injected,
without further purification on a HP 5890 series II
gas chromatograph with a PORAPLOT Q fused sil-
Ž
.
PO 639 mlr10 ml acetone was added to 3 ml
human blood to obtain the final concentrations of 0,
10, 30, 100, 300 and 1000 mM. The same final
concentrations were obtained by addition of suffi-
Ž
ica capillary column 25 m=0.32 mm, phase thick-
.
Ž
ness 10 mm from Chrompack Bergen op Zoom,
The Netherlands . The chromatograph was equipped
.
Ž
cient amounts of SO 144 ml in 10 ml isotonic
.
w x
with a flame ionization detector and an automatic
saline to 4 ml aliquots of human blood 10 .

(
)
W. Pauwels, H. VeulemansrMutation Research 418 1998 21–33
25
All blood samples were occasionally stirred dur-
ing a 2 h incubation at 378C.
2.3.2. Modification of DNA
Similar procedures were used for the incubation
of blood to determine the formation of N-7-guanine
adducts, except that the initial concentrations had to
be higher, because of the lower sensitivity. For EO,
Ž
sufficient amounts of a 1 M aqueous solution stored
.
at y808C were added to 10 ml blood to obtain
initial concentrations of 1, 2, 5 and 10 mM. The
blood samples were incubated for 3 h at 378C.
Five milliliters freshly drawn blood was spiked
with 2, 4, 18 and 35 ml PO, resulting in initial
concentrations of 5, 11, 51 and 100 mM, respec-
tively. Amounts of 5.5, 29, 57, 143 and 285 ml SO
were added to 5 ml blood in order to obtain initial
concentrations of 9.4, 49, 96, 240 and 460 mM,
respectively. The samples were incubated for 2 h at
378C, with occasional stirring.
2.4. Determination of N-terminal Õaline adducts
2.4.1. Isolation of Hb
Red blood cells were separated from plasma by
centrifugation at 800 g for 10 min and were washed
with 3 volume of saline prior to lysis with 1 volume
of water. Haemoglobin was purified by the method
w
x
of Mowrer et al. 17 . Briefly, acidic propanol was
added to the red blood cells and, following centrifu-
gation, the globin was precipitated by the addition of
ethyl acetate. The globin was filtered and washed
with ethyl acetate and pentane and then dried by
passing a gentle stream of air over the globin.
2.4.2. DeriÕatization
N-terminal valine adducts were analysed using the
w x
method of Tornqvist et al. 4 as described in Severi
¨
w
x
w x
et al. 14 and Pauwels et al. 13 . Fifty milligrams
Ž
globin and 10 ml I.S. containing 1.2 mg globinrml
and 1.73 nmol N- 2-hydroxy-2- H8 phenylethyl -
valinermg Hb were reacted with pentafluorophenyl
Ž
w2
x
.
.
isothiocyanate under basic conditions. The resulting
thiohydantoin derivatives were extracted into diethyl
ether, washed with NaHCO3 and dissolved in toluene
for analysis by GC–MS.

(
)
26
W. Pauwels, H. VeulemansrMutation Research 418 1998 21–33
Ž
.
Ž
2.4.3. Determination by means of GC–MS
ficients for EO-Hb r s 0.993 , N- 2-hydroxy-
.
Ž
.
Ž
The most important differences for the determina-
tion of the three adducts are given in Table 2.
Adduct formation of SO on the N-terminal valine in
haemoglobin occurs mainly through the b-carbon of
SO resulting in the formation of two diastereoiso-
ethyl Val-Leu-anilide r s 0.997 and N- 2-hy-
droxypropyl Val-Leu-anilide rs0.9986 were only
.
Ž
.
slightly lower.
2.5. Determination of N-7-guanine adducts
Ž
.
w x
mers of N- 2-hydroxy-2-phenyl ethylvaline 18 .
Ž
.
Also N- 2-hydroxypropyl valine, but not the EO-ad-
duct, contains two chiral carbons leading to the
formation of two diastereoisomers. The analyses were
carried out using a HP 5890 Series II gas chromato-
graph coupled to a HP 5970 quadrupole mass spec-
2.5.1. Isolation of DNA
The incubation experiments were carried out in
Vacutainerw tubes, with EDTA or sodium citrate as
anticoagulant. The isolation procedure for EO was
slightly different, since the samples were initially
prepared for HPLCrElectrospray Mass Spectrome-
try, requiring a higher purity. After the incubation
with EO, blood lymphocytes were isolated using
Leucoprep tubes Becton Dickinson, Aalst, Belgium ,
following the manufacturer’s protocol. For the other
epoxides, after centrifugation at 1200=g for 10
min, the lymphocytes were transferred to a clean
tube. After washing with phosphate buffered saline
Ž
.
trometer EI mode and a standard chemstation
G1701AA version 03.00. The chromatographic sepa-
ration of the different components in the samples
was made on a DB-5-ms fused silica capillary col-
Ž
.
Ž
.
umn 30 m=0.32 mm, 0.12 mm phase thickness .
Helium was used as the carrier gas.
The interface temperature was 2708C and the
source pressure was 0.005 Pa. Analyses were carried
out using selected ion monitoring. Due to the manual
injection of the samples the observed retention times
could deviate sometimes from the values given in
Table 2. The identification of the peaks was assured
by the determination of the retention time relative to
the I.S.
Ž
.
2= and 0.9% NaCl, the cells were lysed by
Ž
.
Ž
addition of a cold 48C hypotonic solution 20 mM
.
Tris HCL, pH 8.0, 10 mM EDTA . After pelleting
the cells at 4000 rpm the supernatant was discarded.
To the pellet 5 ml DNAzole reagent Gibco BRL,
Life Technologies, Merelbeke, Belgium was added.
Ž
.
Ž
.
To reduce the RNA amount -3% the solution was
centrifuged for 10 min at 10,000=g. DNA was
precipitated from the supernatant, which was trans-
ferred to a clean tube, by the addition of 2.5 ml
ethanol. The DNA pellet, obtained after centrifuga-
tion, was washed twice with 95% ethanol and twice
with 70% ethanol. For the incubation with EO, the
DNA was air dried and weighted with an analytical
balance. The determination of the amount of DNA
extracted from the blood for the PO and the SO-ex-
periment, was done by solving the DNA in 1.5 ml
water followed by spectrophotometrical determina-
tion, assuming that one A260 unit equals 50 mg
double stranded DNA per ml. The A260rA280 ratio,
which was between 1.6 and 1.9 for all extractions,
was routinely determined to see if the DNA was free
of proteins. The DNA was dried by lyophilization.
2.4.4. Calibration
Calibration curves were prepared through the ad-
dition of different amounts of the calibration refer-
ence globin or alkylated dipeptide to a constant
amount of I.S. The two calibration curves for EO
were established between 0 and 449 pmol adductrg
Ž
Hb and 0 and 575 pmol adductrg Hb for N- 2-hy-
.
droxyethyl Val-Leu-anilide and EO-Hb, respectively.
The calibration curve for PO was established from
Ž
.
calibration with N- 2-hydroxypropyl Val-Leu-anilide
as standard between 0 and 523 pmol PO-adductrg
Hb and for SO with SO-Hb as standard between 0
and 406 pmol SO-adductrg Hb. All established cali-
bration curves were linear. Although it should be
better to use different deuterated analogues as I.S.
for each N-terminal valine adduct separately, the
obtained correlations for the calibration curves were
considered sufficient for the purposes of this study.
The correlation coefficient for the calibration curve
for SO was rs0.9995, whereas the correlation coef-
2.5.2. Determination by means of HPLC-UV
The DNA samples were dissolved in 100 ml
water and submitted to a neutral thermal hydrolysis

(
)
W. Pauwels, H. VeulemansrMutation Research 418 1998 21–33
27
Ž
.
1008C-30 min in order to release the N-7-adducts.
determined as the sum of both isomers. Calibration
curves for all adducts were prepared through the
injection of different solutions containing 10y7 –10y6
M adduct. All absolute calibration curves were linear
in the concentration range studied.
Ž
.
Samples 25 ml were injected without any further
derivatization on a HP 1090 Series II Liquid Chro-
matograph coupled with a HP 1040 Diode Array
Detector. The products were separated on a 4=25
mm reverse phase C18-column, filled with Nucleosil-
Ž
.
100 with 5 mm particles Machery–Nagel, Germany .
The eluent used for HPLC was a 0.1 M ammonium
3. Results
Ž
.
acetate buffer pHs5.2 . The best separation was
obtained with a methanol gradient going from 6% to
10% in 12 min for EO, 6% to 10% in 14 min for PO
and from 6% to 30% in 18 min for SO. The flow
was 1 mlrmin. For quantitative determinations the
UV-signal was measured at two wavelengths: ls
254 nm and 282 nm, with a band width of 4 nm.
3.1. Comparison of the calibration curÕes obtained
with EO-Hb and N- 2-hydroxyethyl Val-Leu-anilide
as standard
(
)
In Fig. 1, calibration curves for the determination
Ž
.
of N- 2-hydroxyethyl valine are presented. The cali-
bration curve obtained from haemoglobin alkylated
with EO shows a less steep slope than the curve
established from the alkylated dipeptide as standard.
Although the difference between both curves is small,
a possible explanation may be the difference in
efficacy of the Edman degradation reaction between
the dipeptide standard and the globin standard.
2.5.3. Calibration
In contrast with the situation for EO, for PO
theoretically the formation of two regioisomers of
the N-7-guanine adduct is possible. The reaction
with PO, however, resulted only in the formation of
Ž
.
the b-adduct N-7- 2-hydroxypropyl guanine in de-
tectable quantities. In contrast, the reaction between
N-7-guanine and SO results in the formation of two
3.2. Elimination of the epoxides from whole blood
Ž
.
regioisomers, i.e., N-7- 2-hydroxy-1-phenyl ethyl-
Ž
.
guanine and N-7- 2-hydroxy-2-phenyl ethylguanine
19 . The SO-adduct formation on N-7-guanine was
If the epoxides are uniformly mixed and the
concentration is low relative to the pool of available
w
x
Ž
.
Ž
.
Fig. 1. Calibration curves area standardrarea I.S. in function of concentration standard for the determination of N- 2-hydroxyethyl valine
with either EO-Hb – – – – or N-2-hydroxyethyl-Val-Leu-anilide –'– as standard and 2 H8 SO as I.S.
Ž
.
Ž
.
w
x

(
)
28
W. Pauwels, H. VeulemansrMutation Research 418 1998 21–33
Table 3
Ž
lated using the following relationship derived from
x.
Elimination constants and half-lives for the three epoxides in
human blood
w
Ref. 21 :
y1
Ž
.
Ž .
t1r2 h
w
x
Ž
EpoxyY
k
e
Agent
Initial
concentration
ke
h
.
kEpoxyY
s
s
e
w xw
x
Y
Epox 1yeyk t
0Ž
.
Ž
.
mM
EO
PO
SO
328
30
40
0.20
0.05
0.78
0.11
3.5
13.6
0.90
6.4
bY
k
e
1yeyk
Ž
.
e t
w x
Y
Ž
.
50,000
Ž
with kEpoxyY : second-order reaction constant
l
moly1 hy1 ; EpoxyY : adduct concentration in
. w
x
y1
Ž
.
;
mM; ke: elimination constant of the epoxide h
w x
Y : haemoglobin or DNA nucleotide concentration
in human blood M ; Epox : initial epoxide con-
centration in mM; bY : slope of the linear regression
Ž
. w
x
nucleophiles in DNA and haemoglobin, then the
overall rate of reaction is pseudo-first-order, with the
rate constant ke representing the sum of all first-order
0
of the adduct concentration on the initial epoxide
w
x
Ž
rate constants of the individual reactions 20 . This
fits with the observation that in all elimination exper-
iments, even for the incubation with a high concen-
tration of SO, first-order kinetics turned out to be the
most appropriate model to describe the results. Ac-
concentration mMrmM and forced through the ori-
.
Ž .
gin ; t: incubation time h ;
The value used for the blood concentration of
Ž
.
haemoglobin 2.3 mM was taken from the literature
w
x
22 . The DNA Isolation kit allowed to isolate an
w
x
cordingly, the concentration of the epoxide, Epox ,
average yield of 350 mg DNAr10 ml blood col-
Ž
as a function of the incubation time can be described
lected from healthy human subjects s113 mM
as Epox s Epox e t, with Epox being the
initial concentration. Values of the elimination con-
stant, ke and the corresponding half-life for the
different epoxides are given in Table 3.
DNA nucleotide , ranging from 200–700 mgr10 ml.
yk e
w
x
w
x
w
x
.
0
0
With these data, the equation was worked out and
the results are presented in Table 4. The haemoglobin
adduct formation was linear for the whole concentra-
tion range studied. For the determination of the DNA
binding, the formation of N-7-EO and N-7-PO-
guanine was linear up to epoxide concentrations of
about 10 mM. For SO, the least reactive epoxide
towards DNA, the dose–adduct relationship was lin-
ear up to 240 mM. At this concentration, however,
3.3. Reaction of N-terminal Õaline and N-7-guanine
with the epoxides in Õitro
The second-order rate constants for the reaction of
the two site-specific adducts studied can be calcu-
Table 4
y1
w Ž
.
x
for the three epoxides in human blood
Second-order rate constants l mol Hb or DNA nucleotide h
Agent
bHab
kEpox-Hb
bDbNA
kEpox-DNA
Ratio kEpox-Hbr
kEpox-DNA
EO
PO
SO
12=10y3
4.5=10y3
3.5=10y3
2.7
1.0
1.5c
2.6=10y6
0.60=10y6
0.13=10y6
16=10y3
5.1=10y3
169
196
2113
0.71=10y3d
a
Ž
.
Ž
bHb is the slope mM adductrmM epoxide of the relationship between dose and N-terminal valine adduct in haemoglobin linear between
.
0 and 1000 mM epoxide and forced through the origin .
b
Ž
.
Ž
bDNA is the slope mM adductrmM epoxide of the relationship between dose and N-7-guanine adduct in DNA linear between 0–10 mM
.
for EO, between 0–11 mM for PO and between 0–240 mM for SO; forced through the origin .
cAs elimination constant ke s0.78 hy1, obtained with an initial SO concentration of 40 mM, was used.
dAs elimination constant ke s0.11 hy1, obtained with an initial SO concentration of 50 mM, was used.

(
)
W. Pauwels, H. VeulemansrMutation Research 418 1998 21–33
29
w x
lysis of the blood was observed, which was even
more pronounced at higher concentrations.
in salt water 1 . In contrast, a much faster non-en-
zymatic elimination of 3–4% SO per hour, resulting
in a half-life between 12.5 and 16.7 h, was observed
The reaction rate constant of N-terminal valine in
haemoglobin as well as of N-7-guanine was highest
for EO. N-terminal valine in haemoglobin showed
the slowest reaction rate constant towards PO,
whereas for N-7-guanine the slowest reaction rate
constant was observed for SO. In Table 4 also the
ratios between the second-order reaction rate con-
stant for haemoglobin and DNA are presented. For
EO and PO comparable ratios of 169 and 196,
respectively, were obtained. In contrast, for SO the
ratio between kSO-Hb and kSO-DNA was 2113, which
reveals an increased preference for N-terminal valine
relative to N-7-guanine of about one order of magni-
tude compared to the other epoxides studied.
w
x
in a buffer solution 24 . Moreover, Dent and Schnell
25 concluded that EO and PO are poor substrates
for epoxide hydrolase in contrast with SO. Also,
Mendrala et al. 26 observed a high glutathione
S-transferase activity towards SO in vitro. These
arguments lead to the conclusion that the fast elimi-
nation for SO from whole blood at low concentration
is according to expectations with literature data.
w
x
w
x
4.3. Second-order rate constants for the reaction
with N-terminal Õaline and N-7-guanine
The second-order rate reaction constants for the
reaction with N-terminal valine and N-7-guanine
have previously been determined for both ethylene
and propylene oxide. The constant for the reaction
4. Discussion
between human N-terminal valine and ethylene ox-
4.1. Use of dipeptides as standard
y1
Ž
.
w x
ide was estimated to be 2.97 l mol Hb h
8 ,
.
y1
Ž
.
which is close to our estimate of 2.7 l mol Hb h
Peptide standards have already been used for the
determination of N-terminal valine adducts of EO
w x
For propylene oxide Segerback et al. 9 came to a
¨
kPO-Hb of 1.01 l mol Hb h y1, virtually equal to our
estimate.
Ž
.
w
x
and SO 23,13 . This study clearly demonstrates that
the use of an alkylated dipeptide standard is as
suitable as an alkylated globin standard for the deter-
mination of N-terminal valine adducts by means of
the modified Edman degradation technique. More-
over, advantages such as commercial availability,
stability and accuracy allow a better comparability
between labs.
The way of determination of the second-order rate
reaction constant in DNA in this study is different
from the method used in previous studies. The incu-
bations in this study were done with whole blood,
whereas the available literature values are based on
in vitro expositions of calf thymus DNA to the
epoxides. The lower accessibility of DNA in whole
blood could explain our estimate of the second-order
rate reaction constant for ethylene oxide being about
4.2. Elimination of the epoxides from whole blood
half of the literature value of 30=10y3 l mol DNA
Ž
y1
.
w
x
A slower elimination at a higher initial concentra-
tion for SO can probably be attributed to the lowered
contribution at this concentration of enzymatic pro-
cesses and reactions with nucleophiles to the elimi-
nation of SO. When the elimination of the three
epoxides at comparable concentrations is considered,
the fastest elimination was observed for SO. Al-
though ethylene oxide is clearly the most reactive of
the three epoxides studied, this is not reflected in its
half-life in human blood. This is in accordance with
the observation of a relative long half-life of 4 days
for ethylene oxide and 4.1 days for propylene oxide
nucleotide h
11 . For PO, a smaller difference
y1
between our 5.1=10y3 l mol DNA nucleotide h
Ž
.
and the value of 7.7=10y3 l mol DNA nucleotide
Ž
y1
.
w x
h
obtained by Segerback et al. 9 was seen.
¨
4.4. Comparison of literature Õalues of the in Õitro
and in ÕiÕo ratio between N-terminal Õaline and
N-7-guanine adduct formation
w
x
Osterman-Golkar et al. 27 suggest that adduct
formation in short term experiments could be related

Table 5
Comparison of literature values of the ratio between adduct formation on N-terminal valine and N-7-guanine in vivo for the three epoxides studied
Compound
Target tissue
for DNA adducts
Species
Type of dosing, regimen and amounts
Ratio N-terminal valiner
N-7-guanine
Reference
Ž
.
experiment duration
Ž
.
w
w
w
w
w
w
w
w
w
x
x
x
x
x
x
x
x
x
Ethylene
EO
liver
liver
liver
lung
lung
liver
liver
liver
lung
liver
lung
liver
lung
liver
lung
liver
lung
rat 3 days
inhalation 300 ppm, 12 hrday
243
28
29
30
31
31
32
28
32
32
Ž
.
.
Ž
.
rat 6 h
i.p. injection 2.8–20.4 mmolrkg BW
53–65
267–342
993–1254
295–367
71
174
66
285
246–361
171–245
318–731
171–464
797–872
222–259
6–55
Ž
rat 6 h
inhalation 3–33 ppm
Ž
.
Ž
.
.
mice 4 weeks
inhalation 33–100 ppm 6 hrday, 5 daysrweek
Ž
.
Ž
rat 4 weeks
inhalation 33–300 ppm 6 hrday, 5 daysrweek
Ž
.
Propylene
PO
mice 7 h
an estimated inhalation of 0.88 mmolrkg BW
Ž
.
rat 3 days
inhalation 300 ppm, 12 hrday
Ž
Ž
.
.
mice 3 h
mice 3 h
mice 6–7 h
mice 7 h
i.p. injection 0.1 mmolrkg BW
i.p. injection 0.1 mmolrkg BW
Ž
Ž
.
w x
9
inhalation or i.p. injection 0.05–0.32 mmolrkg BW
inhalation 0.11–0.32 mmolrkg BW
.
w x
9
Ž
.
.
w x
9
rat 6–7 h
inhalation or i.p. injection 0.05–0.19 mmolrkg BW
inhalation or i.p. injection 0.05–0.19 mmolrkg BW
i.v. injection 0.07–0.35 mmolrkg BW
Ž
w x
9
rat 6–7 h
Ž
Ž
.
.
w x
9
dog 4 h
dog 4 h
mice 3 h
mice 3 h
w x
i.v. injection 0.07–0.35 mmolrkg BW
9
Ž
Ž
.
.
w
x
x
Styrene
i.p. injection 0.28–4.35 mmolrkg BW
33
33
w
i.p. injection 0.28–4.35 mmolrkg BW
9–32

(
)
W. Pauwels, H. VeulemansrMutation Research 418 1998 21–33
31
w
x
w x
to the integrated dose of the epoxide according to the
following relationship:
et al. 29 and Potter et al. 30 resulting from short
term experiments.
The ratio for propylene oxide adducts with Hb
and DNA varies between 66 and 872, without any
clear relationship across the species, the target tissue,
the way of administration, the exposure regimen or
the amount of epoxide administered. However, the
average value for all ratios was 317, which is, given
the high variation observed, not considered to be
substantially different from our in vitro obtained 196.
In contrast with EO and PO-binding, for SO a
deviation of about two orders of magnitude is seen
between the in vitro and the in vivo observed ratio.
w
x
w x
Ds EpoxyY rkEpoxyY
Y
where D has units mM h. Assuming that the inte-
grated dose of the epoxide in blood is the same for
N-terminal valine in haemoglobin and N-7-guanine
adduct formation in leucocytes, from this equation it
can be deduced that:
w
x w
x
Epox-Hb r Hb
.
kEpox -HbrkEpox -DNA
s
Ž
w
x w
x
Epox-DNA r DNA
.
r
Ž
In other words, the ratio of the second-order rate
constants should agree with the ratio between the
amount of N-terminal valine and N-7-guanine adduct
in leucocytes expressed in mol adduct per mol
macromolecule. Furthermore, this in vitro ratio is
compared with a number of animal studies wherein
the ratio of the adduct formation on haemoglobin
and on DNA in different tissues has been studied
4.5. Discussion on the extrapolation from the in Õitro
obtained results to the in ÕiÕo situation
The applicability of the in vitro obtained reaction
constant to the in vivo situation was already demon-
w x
strated in a study of Ehrenberg and Tornqvist 6 in
¨
which the adduct formation by EO on N-terminal
valine in vivo is modelled. They observed that based
on the kEO-Hb and other metabolic parameters for
rodents at low dose a reasonably consistent picture
was obtained concerning the estimates of the adduct
level increments per unit of administered dose in
vivo. Moreover, the authors came to the conclusion
that for humans, although the value for the in vivo
dose per unit of exposure was uncertain because of
unreliable data, the most likely value for this rela-
tionship agrees with the rodent data. Due to the
lower sensitivity of the DNA adduct determination
technique, the experimental conditions for
haemoglobin and DNA adduct determination in this
study differed substantially. Nevertheless, our in vitro
obtained ratio between N-terminal valine and N-7-
guanine adduct formation by EO and PO, seems to
agree very well with the in vivo obtained ratio in
short term animal experiments.
Ž
.
Table 5 . At present it remains, however, unknown
whether this ratio remains constant across species
and tissues. Especially, the amount of N-7-guanine
adducts could be different due to variations in the
tertiary structure of the DNA.
Nevertheless, the data obtained by Walker et al.
w
x
31 for ethylene oxide show a good agreement
between the levels of N-7-guanine adducts in leuco-
cytes and in the lung. In the liver slightly lower
N-7-guanine adduct levels were found. Literature
values for the ratio between N-terminal valine and
N-7-guanine adduct formation in liver and lung, the
two most studied tissues, are presented in Table 5.
w
x
The different ratio in the study of Walker et al. 31
with respect to our in vitro obtained value, can
possibly be explained by RNA interferences, due to
insufficient purification, with a different degree of
w
x
alkylation than the DNA 28 . Another possible ex-
planation could be the relatively long duration of the
experiment. The life-time of haemoglobin adducts
Only for the N-terminal valinerN-7-guanine ratio
for SO a major discrepancy was observed between
the in vitro and in vivo results. This is probably due
to lower than expected values for N-terminal valine
adducts by SO in vivo, which was already concluded
Ž
.
120 days compared to a half-life for N-7-guanine
w
x
adducts between 1.0 and 5.8 days 31 , results in an
increased HbrDNA adduct ratio for long term exper-
iments. Our in vitro obtained value of 169 for EO-bi-
w
x
by Christakopoulos et al. 18 , extrapolating from
EO-data. Apparently, the ratio of the in vitro ob-
tained second-order rate reaction constants for the
Ž
.
nding Table 4 , therefore, has preferentially to be
compared to the values obtained by Osterman-Golkar

(
)
32
W. Pauwels, H. VeulemansrMutation Research 418 1998 21–33
reaction of SO with N-terminal valine and N-7-
guanine shows a 100-fold difference as to the in vivo
adduct levels. Inaccuracy of the techniques and prob-
lems with the detection limit of the methods can only
partly account for this discrepancy. Several studies
indicate that our results showed no major deviations
from the results of other laboratories using indepen-
dent variants of the same technique or even com-
the Prime Minister; Scientific, Technical and Cul-
tural Affairs.
References
w x
1
IARC, Monographs on the evaluation of carcinogenic risks to
humans, Vol. 60, Some industrial chemicals. Lyon 1994, pp.
45–213 and pp. 233–346.
w
x
pletely different techniques 10,13,33,34 . Probably,
one or more of the three assumptions that were made
for the extrapolation does not hold in the specific
case of SO. First of all, a linear extrapolation was
assumed from the higher incubation concentrations
of SO for DNA adduct formation to the concentra-
tion used for Hb adduct formation. Secondly, the
species difference was not assumed to affect the ratio
between both adducts. The in vitro results were
determined in human blood, whereas the in vivo
results for SO were obtained in mice. In this context,
w x
2
M. Uziel, N.B. Munro, D.S. Katz, T. Vo-Dinh, E. Zeighami,
M.D. Waters, J.D. Griffith, DNA adduct formation by 12
chemicals with populations potentially suitable for molecular
Ž
.
epidimiological studies, Mutation Res. 277 1992 35–90.
w x
3
K. Hemminki, DNA adducts, mutations and cancer, Carcino-
Ž
.
genesis 14 1993 2007–2012.
w x
4
M. Tornqvist, J. Mowrer, S. Jensen, L. Ehrenberg, Monitor-
¨
ing of environmental cancer initiators through haemoglobin
adducts by a modified Edman degradation method, Anal.
Ž
.
Biochem. 154 1986 255–266.
w x
5
M. Tornqvist, L. Ehrenberg, On cancer risk estimation of
urban air pollution, Environ. Health Perspect. 104 1994
¨
Ž
.
173–182, Suppl. 4.
Ž
the potential protective role of the different species
w x
6
L. Ehrenberg, M. Tornqvist, The research background for
¨
risk assessment of ethylene oxide: aspects of dose, Mutation
.
specific GST-types remains to be elucidated. Espe-
cially the erythrocytic GSTT1-1 enzyme could pre-
Ž
.
Res. 330 1995 41–54.
w x
vent SO and the other epoxides from reacting with
7
M. Tornqvist, H. Hindsø Landin, Hemoglobin adducts for in
¨
vivo dose monitoring and cancer risk estimation, J. Occup.
w
x
N-terminal valine 35 . A last assumption was that
the formation of N-7-guanine adduct levels in leuco-
cytes is comparable to the levels in other tissues.
Data on this issue for styrene 7,8-oxide are lacking,
but could differ from the other epoxides due to its
deviating chemical properties. In contrast with the
other epoxides, styrene and its metabolites contain an
aromatic ring, resulting in increased lipophilic prop-
erties. This certainly leads to higher concentrations
Ž . Ž
.
Environ. Med. 37 9 1995 1077–1085.
w x
8
D. Segerback, Reaction products in haemoglobin and DNA
¨
Ž
after in vitro treatment with ethylene oxide and N- 2-hy-
.
Ž
.
droxyethyl -N-nitrosourea, Carcinogenesis 11 1990 307–
312.
w x
9
D. Segerback, S. Osterman-Golkar, B. Molholt, R. Nilsson,
¨
In vivo tissue dosimetry as a basis for cross-species extrapo-
lation in cancer risk assessment of propylene oxide, Regul.
Ž
.
Toxicol. Pharmacol. 20 1994 1–14.
w
x
10 K. Yeowell-O’Connell, W. Pauwels, M. Severi, Z. Jin, M.R.
Walker, S.M. Rappaport, H. Veulemans, Comparison of
styrene-7,8-oxide adducts formed via reaction with cysteine,
N-terminal valine and carboxylic acid residues in human,
Ž
w x.
in fat reviewed in Ref. 1 , but could also work at
the subcellular level with a higher affinity for, e.g.,
membranes. In this context, it should be emphasized
that no data are available of both haemoglobin and
DNA adduct formation in SO treated animals. Al-
though treatment with the alkene compared to the
epoxide did not lead to major differences in the
adduct ratio for ethylene and propylene, due to its
higher affinity for membranes the situation for styrene
could be different.
Ž
.
mouse and rat Hb, Chem.–Biol. Interact. 106 1997 67–85.
w
x
11 D. Segerback, Alkylation of DNA and haemoglobin in the
¨
mouse following exposure to ethene and ethene oxide,
Ž
.
Chem.–Biol. Interac. 45 1983 139–151.
w
x
12 M. Tornqvist, A.-L. Magnusson, P.B. Farmer, Y.-S. Tang,
¨
A.M. Jeffrey, L. Wazneh, G.D.T. Beulink, H. Van der Waal,
Ž
N.J. Van Sittert, Ring test for low levels of N- 2-hydroxy-
.
ethyl valine in human haemoglobin, Anal. Biochem. 203
Ž
.
1992 357–360.
w
x
13 W. Pauwels, P.B. Farmer, S. Osterman-Golkar, M. Severi, R.
Cordero, E. Bailey, H. Veulemans, Ring test for the determi-
nation of N-terminal valine adducts of styrene 7,8-oxide on
haemoglobin by the modified Edman degradation technique,
Acknowledgements
Ž
.
J. Chromatogr. B 702 1997 77–83.
This project was supported by the Belgian Pro-
gram for Health Hazards funded by the Services of
w
x
14 M. Severi, W. Pauwels, P. Van Hummelen, D. Roosels, M.
Kirsch-Volders, H. Veulemans, Urinary mandelic acid and

(
)
W. Pauwels, H. VeulemansrMutation Research 418 1998 21–33
33
w
w
w
x
haemoglobin adducts in fiberglass-reinforced workers ex-
26 A.L. Mendrala, P.W. Langvardt, K.D. Nitschke, J.F. Quast,
R.J. Nolan, In vitro kinetics of styrene and styrene oxide
metabolism in rat, mouse, and human, Arch. Toxicol. 67
Ž
.
posed to styrene, Scand. J. Work Environ. Health 20 1994
451–458.
w
w
w
w
x
Ž
.
15 F. Latif, R.C. Moschel, K. Hemminki, A. Dipple, Styrene
oxide as a stereochemical probe for the mechanism of aralky-
lation at different sites an guanosine, Chem. Res. Toxicol. 1
1993 18–27.
x
27 S. Osterman-Golkar, L. Ehrenberg, D. Segerback, I.
¨
Hallstrom, Evaluation of genetic risks of alkylating agents:
¨
¨
Ž
.
Ž
.
1988 364–369.
II. Haemoglobin as dose monitor, Mutation Res. 34 1976
1–10.
x
16 S.M. Rappaport, D. Ting, Z. Jin, K. Yeowell-O’Connell, S.
Waidyanatha, T. McDonald, Application of Raney nickel to
measure adducts of styrene oxide with haemoglobin and
x
28 I. Eide, R. Hagemann, K. Zahlsen, E. Tareke, M. Tornqvist,
¨
R. Kumar, P. Vodicka, K. Hemminki, Uptake, distribution,
and formation of hemoglobin and DNA adducts after inhala-
Ž .
tion of C2 –C8 1-alkenes olefins in the rat, Carcinogenesis
Ž
.
albumin, Chem. Res. Toxicol. 6 1993 238–244.
x
17 J. Mowrer, M. Tornqvist, S. Jensen, L. Ehrenberg, Modified
¨
Ž .
16 1995 1603–1609.
Edman degradation applied to haemoglobin for monitoring
w
x
occupational exposure to alkylating agents, Toxicol. Environ.
29 S. Osterman-Golkar, P.B. Farmer, D. Segerback, E. Bailey,
¨
Ž
.
Chem. 11 1986 215–231.
C.J. Calleman, K. Svensson, L. Ehrenberg, Dosimetry of
ethylene oxide in the rat by quantitation of alkylated histidine
x
18 A. Christakopoulos, E. Bergmark, V. Zorcec, H. Norppa, J.
Maki-Paakkanen, S. Osterman-Golkar, Monitoring occupa-
tional exposure to styrene from hemoglobin adducts and
metabolites in blood, Scand. J. Work Environ. Health 19
Ž
.
in haemoglobin, Teratogen. Carcinogen. Mutagen. 3 1983
395–405.
w
w
x
30 D. Potter, D. Blair, W.P. Watson, S. Wright, The relationship
between alkylation of haemoglobin and DNA in Fisher 344
Ž
.
1993 255–263.
w
x
w14
x
19 P. Vodicka, K. Hemminki, Identification of alkylation prod-
rats exposed to C ethylene oxide, Arch. Toxicol. Suppl.
Ž
.
ucts of styrene oxide in single- and double-stranded DNA,
13 1989 254–257.
Ž
.
x
Carcinogenesis 9 1988 1657–1660.
31 V.E. Walker, T.R. Fennell, P.B. Upton, T.R. Skopek, V.
w
x
20 L. Ehrenberg, E. Moustacchi, S. Osterman-Golkar, Dosime-
Prevost, D.E.G. Shuker, J.A. Swenberg, Molecular dosimetry
Ž
try of genotoxic agents and dose–response relationships of
of ethylene oxide: formation and persistence of 7- 2-
Ž
.
.
their effects, Mutation Res. 123 1983 121–182.
hydroxyethyl guanine in DNA following repeated exposures
w
w
w
x
Ž .
of rats and mice, Cancer Res. 52 1993 4328–4334.
21 L. Ehrenberg, S. Hussain, Genetic toxicity of some important
epoxides, Mutation Res. 86 1981 1–113.
22 ICRP, Report of the Task Group on Reference Man, ICRP
publication number 23, Pergamon, Oxford, 1975, pp. 37–40.
23 R.M. Lawrence, G.M. Sweetman, R. Tavares, P.B. Farmer,
Ž
.
w
w
x
32 K. Svensson, K. Olofsson, S. Osterman-Golkar, Alkylation
of DNA and haemoglobin in the mouse following exposure
to propene and propylene oxide, Chem.–Biol. Interact. 78
x
x
Ž
.
1991 55–66.
x
Synthesis and characterization of peptide adducts for use in
monitoring human exposure to acrylonitrile and ethylene
oxide, Teratogen. Carcinogen. Mutagenen. 16 1996 139–
148.
33 W. Pauwels, P. Vodicka, M. Severi, K. Plna, H. Veulemans,
K. Hemminki, Adduct formation on DNA and haemoglobin
in mice intraperitoneally administered with styrene, Carcino-
Ž
.
Ž
.
genesis 17 1996 2673–2680.
w
x
w
x
24 M.K. Viaene, Acute en chronische neurotoxische effecten
van styrene en hun reversibiliteit. Resultaten van een cross-
sectional onderzoek bij werknemers tewerkgesteld in de
glasvezelversterkte polyesterindustrie. Ph.D. Thesis, KU-
34 S. Cantoreggi, W.K. Lutz, Covalent binding of styrene to
Ž
.
DNA in rat and mouse, Carcinogenesis 14 1993 335–350.
w
x
35 U. Fost, E. Hallier, H. Ottenwalder, H.M. Bolt, H. Peter,
¨
¨
Distribution of ethylene oxide in human blood and its impli-
Ž
.
Ž
.
Leuven, 1998 pp. 175.
cations for biomonitoring, Hum. Exp. Toxicol. 10 1991
25–31.
w
x
25 J.G. Dent, S.R. Schnell, Inhibition of microsomal-membrane
bound and purified epoxide hydrolase by C2 –C8 1,2-alkene
Ž
.
oxides, Biochem. Pharmacol. 30 1981 1712–1714.