Kinetics of formation of specific styrene oxide adducts in double-stranded DNA

Chemico-Biological Interactions 138 (2001) 111–124

Kinetics of formation of specific styrene oxideadducts in double-stranded DNA

Mikko Koskinen a,*, Ludmila Vodickova b, Pavel Vodicka c,SusanC. Warner a, Kari Hemminki a

a Department of Biosciences at No�um, Karolinska Institute, S-141 57 Huddinge, Swedenb National Institute of Public Health, S� robaro�a 48, 100 42 Prague 10, Czech Republic

c Institute of Experimental Medicine, Academy of Sciences of Czech Republic, Videcska 1083,14220 Prague 4, Czech Republic

Received 27 February 2001; received in revised form 23 April 2001; accepted 12 May 2001

Abstract

The possible carcinogenicity of styrene is believed to be related to the DNA-bindingproperties of styrene 7,8-oxide (SO). In order to compare the intrinsic reactivity of thedifferent nucleophilic sites in DNA towards SO and to evaluate the candidates for humanbiomonitoring we have determined the second-order rate constants and stabilities of severalSO-adducts in double-stranded DNA. These include �- and �-isomers of N7-substituted and�N2-substituted guanines, �- and �N3-substituted and �N6-substituted adenines as well as�N3- and �N4-substituted cytosines. The highest rate constants were found for the sponta-neously depurinating N7-guanines being ca. 3–15-fold higher than those for the stableadducts. When the relative proportions of different alkylation products were determined incourse of time, after a single addition of SO, the labile N7-guanines and N3-adenines werethe major products at early time points. After 144 h of incubation at 37 °C, �N6-SO-adenineand �N2-SO-guanine as well as �N3-SO-uracil were the major adducts. Regarding humanbiomonitoring, the N7-substituted guanines should be one of the main targets because of thehigh reactivity of the N7-atom of guanine. However, in the case of chronic styrene exposures

www.elsevier.com/locate/chembiont

Abbre�iations: Ade, adenine; Cyt, cytosine; GC, gas chromatography; Gua, guanine; HPLC, high-per-formance liquid chromatography; Hyp, Hypoxanthine; SO, styrene 7,8-oxide; Ura, Uracil.

* Corresponding author. Current address: Orion Corporation, Orion Pharma, Preclinical and ClinicalR&D, P.O. Box 65, FIN 02101 Espoo, Finland. Tel.: +358-10-429-3853; fax: +358-10-429-3441.

E-mail address: [email protected] (M. Koskinen).

0009-2797/01/$ - see front matter © 2001 Elsevier Science Ireland Ltd. All rights reserved.

PII: S0009 -2797 (01 )00254 -X

M. Koskinen et al. / Chemico-Biological Interactions 138 (2001) 111–124112

the chemically more stable DNA adducts may become important. © 2001 Elsevier ScienceIreland Ltd. All rights reserved.

Keywords: Biomonitoring; DNA adducts; Styrene; Styrene 7,8-oxide

1. Introduction

Styrene, an extensively used industrial chemical, is classified by the InternationalAgency for Research on Cancer (IARC) as a possible human carcinogen (group 2B)[1]. It is metabolised principally via formation of styrene 7,8-oxide (SO), a probablehuman carcinogen (Group 2A) [2]. The possible carcinogenicity of styrene isbelieved to be related to the genotoxic effects of SO, especially to its ability to formDNA adducts [3,4].

In DNA there is a wide variety of nucleophilic target sites for adduct formation.Under physiological conditions the main alkylation sites for aliphatic mono-substi-tuted epoxides are the ring-nitrogens at N7-guanine (Gua), N1- and N3-adenine(Ade), and N3-cytosine (Cyt). The reaction takes place through the �-carbon of theepoxides by the SN2 type of reaction mechanism [5]. The alkylation of exocyclicamino groups involves substrate ionisation [6,7]. Therefore, epoxides that have asubstituent capable of stabilising the positive charge, such as the vicinal aromaticgroup in the case of SO or vinyl group in the case of butadiene monoepoxide, canreact also at the exocyclic sites through the �-carbon [5]. SO-alkylation of aminogroups has been found at N2-position of Gua, N4-position of Cyt and N6-positionof Ade [6–9].

The different nucleophilic sites in DNA bases show differing extent of alkylation,resulting in adducts of differing stabilities as well as differing biological effects [5].All the ring-nitrogen DNA-alkylation products of SO and other alkyl epoxides areto some extent labile due to the positive charge in the purine or pyrimidine ringsystems: N7-SO-Gua and N3-SO-Ade adducts are prone to be depurinated fromDNA [10]; the N1-SO-Ade adduct undergoes a Dimroth rearrangement to N6-SO-Ade, or a deamination to N1-SO-hypoxanthine (Hyp) [11]; the N3-SO-Cyt adductdeaminates rapidly [8].

Recently we have investigated the formation of different SO induced DNAadducts in vitro and in vivo [8,10,12,13]. In order to better understand the chemicalbehaviour of specific SO-induced adducts in biological conditions, we have nowfollowed the kinetics of the formation and stabilities of these adducts as well astheir relative proportions in double-stranded DNA. The structural formulas of theadducts whose rate-constants were determined are collected in Fig. 1. The resultscan be used in predicting the suitable SO-DNA alkylation products to be used inhuman biomonitoring studies.

M. Koskinen et al. / Chemico-Biological Interactions 138 (2001) 111–124 113

2. Materials and methods

2.1. Chemicals

Chemicals were used as purchased from the manufacturers. Salmon testis DNA(sodium salt), alkaline phosphatase, prostatic acid phosphatase and snake venomphosphodiesterase were purchased from Sigma Chemical Co (St Louis, MO).Nuclease P1 was from Boehringer Mannheim Chemica (Mannheim, Germany) andDNase I from Fluka Chemie AG (Buchs, Switzerland). Racemic styrene oxide(�97% pure) was from Aldrich Chemie (Steinheim, Germany), methanol wasgradient grade from Merck (Darmstadt, Germany). All other chemicals were eitherfrom Sigma or Merck. The preparation of the adduct standards used as referencecompounds in identification and quantitation have been described earlier [8,10,12].

2.2. Stability of SO in aqueous solutions

100 �l of 16 mM SO solution in water or 10 mM Tris–HCl, 0.15 M NaCl wasplaced in 5 ml vials of the head-space gas-chromatograph (GC). The solution was

Fig. 1. Structural formulas of the adducts whose rate-constants are determined.


incubated at 37 °C 1, 3, 10, 25 and 50 h (three samples/interval), and SO wasanalysed by GC with FID detector. Analyses were carried out on glass packedcolumn OV 17 (injection and column temperature, 170 and 130 °C, respectively,carrier gas nitrogen, 44 ml/min, hydrogen and air 40 and 500 ml/min, respectively)1 ml of air phase was injected to the column by Hamilton gas tight syringe 37 °Cwarm. Similar experiments were also performed with saturated SO solution, withSO concentrations corresponding to the DNA alkylation studies below. These wereperformed in water, 10 mM Tris–HCl, 0.15 M NaCl and in DNA solution (2mg/ml buffer).

2.3. DNA alkylation and hydrolysis

In the experiments for the determination of the rate constants, 4 mg of salmontestis DNA (3.3 mg/ml) was incubated in 10 mM Tris buffer (pH 7.4), 0.15 M NaClwith 14 �l SO at 37 °C. The double-strandedness of DNA was tested by hyper-chromic effect [10]. The reactions were stopped at the intervals indicated byextracting the excess SO twice with ethyl acetate. The labile N7-SO-Gua andN3-SO-Ade adducts were immediately liberated from the DNA as alkylated basesby neutral thermal depurination keeping the solution in a boiling water bath for 30min, and DNA was precipitated at −20 °C with 4 M sodium acetate (0.1 Vol.)and ethanol (1.4 Vol.). After centrifugation for 20 min (14 000 rpm) the supernatantcontaining the N7-SO-Gua and N3-SO-Ade adducts was evaporated to dryness,and the alkylated bases were separated and quantitated by high-performance liquidchromatography (HPLC, Beckman, System Gold with 168 diode array detector),equipped with a C-18 column (Phenomenex, Kromasil, 250×4.5 mm, 5 �m), asdescribed [10]. The remaining DNA was washed with 70% ethanol and dissolved in1.5 ml of 50 mM Tris pH 7.2, 1 mM MgCl2. DNA was hydrolysed enzymaticallyto nucleosides using nuclease P1 (24 U/ml), alkaline phosphatase (3 U/ml), DNaseI (10 U/ml) and snake venom phosphodiesterase (0.3 U/ml), described previously[8]. After incubation for 26 h at 37 °C, the reaction was stopped by keeping themixture in a boiling water bath for 5 min, followed by cooling in an ice bath,centrifuging and filtering to another tube. After evaporation to dryness in afreeze-dryer, the obtained nucleoside adducts in the hydrolysate, including �- and�N1-SO-Ade, �N1-SO-Hyp, �N2-SO-Gua, �N6-SO-Ade, �N3-SO-Ura and �N4-SO-Cyt, were separated and quantitated by HPLC with UV-detection as described[8].

The 32P-postlabelling analysis with nuclease P1 enrichment for the DNA adducts,including �- and �N1-SO-Ade, �N2-SO-Gua, �N6-SO-Ade as well as �N3-SO-Ura,has been described in detail recently [13]. The hydrolysis to adducted dinucleotideswas carried out by prostatic acid phosphatase and nuclease P1. After T4-kinasemediated 32P-labelling, adducted 5�-nucleotides were obtained with snake venomphosphodiesterase and were separated by HPLC coupled with a radioisotopedetector (Packard Radiomatic 500TR, Flow Scintillation analyser).


2.4. Calculations

For the determination of the half-lives (t1/2) of SO and N1-SO-Ade adducts inDNA velocity constants were calculated from the slope of a semilogarithmic graphplotting concentration versus time, and t1/2 was derived from the velocity constant.

Second-order reaction kinetics is expected for the formation of most of theSO-induced DNA adducts. Since, in our experiments, in the beginning of theexposure the solution is saturated with SO the calculations were performed assum-ing a constant SO concentration. Further, we assume that there was no significantchange in the concentration of the alkylation sites in DNA. Therefore, we canexpect a linear increase in adduct concentration during the first 4 h (time points 1,2 and 4 h). The second-order reaction rate constants were calculated by deriving kfrom the equation:

[A]t

=k [SO][DNA]

[A]t

=k× [SO]× [DNA]

k=[A]

t [SO]=

�

[SO]

k=[A]

t [SO]=

�

[SO]

with k, second-order rate constant; [DNA], DNA concentration, that is constant;[A], adduct concentration (nmol adducts/mg DNA); [SO], SO concentration; �slope of the linear regression of the adduct concentration on time. SO concentrationin a saturated buffer solution was determined in this study to be 25 mM (notshown), in accordance with the value in the literature [2]. Because in the beginningof the reactions undissolved SO was present as droplets in the mixture, some errormight result from the interaction of these droplets with DNA. However, since themixture was not actively mixed during the reaction, the droplets sunk to the bottomof the tube and formed a small surface for interaction with DNA that wasconsidered negligible.

3. Results

3.1. Stability of SO in aqueous solutions

The stability of SO in aqueous solutions was studied by head-space GC. When 16mM SO in water was incubated at 37 °C, the half-life of SO was found to be 21.3h. In a solution of 10 mM Tris–HCl and 0.15 M NaCl, the hydrolysis of SO wassomewhat faster, t1/2 being 13.2 h. In a saturated solution, corresponding to theconditions used in the DNA alkylation studies, the t1/2 for SO were found to be


Table 1The second-order reaction rate constants of specific SO-induced adducts in double-stranded DNA

kDNA×10−4 (l/g DNA per h) raAdduct analysed

0.18 1.000�N7-SO-Guab

�N7-SO-Guab 1.0000.130.9820.06�N3-SO-Adeb

1.0000.01�N3-SO-Adeb

0.9100.06�N6-SO-Adec

0.9560.04�N2-SO-Guac

�N3-SO-Cytc,d 0.03 0.9260.9320.01�N4-SO-Cytd

a r gives the correlation coefficient of the linear regression analyses.b Analysed after neutral depurination as alkylated bases.c Analysed after enzymatic hydrolysis as alkylated nucleosides.d Analysed as �N3-SO-Ura.

42.3 and 32.7 h for water and buffer solutions, respectively. When SO wasincubated in a similar SO saturated buffer solution with DNA (2 mg/ml) the t1/2

was 36.3 h.

3.2. Kinetics of formation of SO-DNA adducts

The second-order reaction rates for the formation of different SO-induced DNAadducts were determined as summarised in Table 1. The DNA adducts involving alabile glycosidic bond, i.e. N7-SO-Gua and N3-SO-Ade, were analysed by neutralthermal hydrolysis following HPLC separation with UV detection. The curveplotting the adducts in the course of incubation showed a saturation at ca. 50 h forN7-SO-Gua (Fig. 2). This is mainly due to the hydrolysis and consumption of SOin aqueous solution. An additional factor contributing to the plateauing is thespontaneous depurination of these adducts. At 64 h a slight drop of the adduct levelwas detected. The reason for this remained obscure, most likely it is related to thevariations in the analyses. When N7-SO-Gua adducts were analysed separately inthe supernatant of the reaction solution and by the thermal hydrolysis in purifiedDNA, the amount of the adducts in the purified DNA started to decrease after 50h, while the spontaneously released adducts analysed from the reaction supernatantcontinued to increase (Fig. 3). The k-values for N7-SO-Gua and N3-SO-Ade inTable 1 reflect the total alkylation at these positions.

The stable adducts were enzymatically hydrolysed from DNA to nucleosides andseparated by HPLC with UV detection. The k-values were determined for theformation of �N6-SO-Ade, �N2-SO-Gua, �N4-SO-Cyt, as well as for �N3-SO-Cyt,which was determined as �N3-SO-uracil (Ura) (see below) (Table 1). The twodiastereomers of the adducted nucleosides were separated in the HPLC analyses.For the calculation of k, the concentrations of the two diastereomers were added tobe comparable with the adduct determinations by thermal hydrolysis (diastereomersnot separated). All of these adducts were stable up to 144 h of incubation (Fig. 4).


Fig. 2. The formation of N7-SO-Gua and N3-SO-Ade adducts in the course of time as determined byneutral depurination. The adduct levels include both spontaneously depurinated fractions as well as theadducts in DNA.

Other DNA adducts that were analysed by the enzymatic hydrolysis included�N1-SO-Ade, �N6-SO-Ade and �N1-SO-Hyp. In the case of �N6-SO-Ade a slightlag in the increase of adduct levels was observed at the early time points (Fig. 4),indicating the reaction showing a different rate-order from the above adducts. Thisis consistent with the fact that this adduct is not directly formed but requiresDimroth rearrangement of �N1-SO-Ade [11]. In the case of the �N3-SO-Ura asimilar lag would be expected, since these adducts are secondary products from�N3-SO-Cyt. However, the deamination of �N3-SO-Cyt appeared to be so fast that

Fig. 3. N7-SO-Gua adducts in double-stranded DNA determined by neutral depurination (ND) and thespontaneously released adducts determined in the supernatant of the reaction mixture (S.D.).


it was readily converted and no lag was observed. The analysis involved someincubation and boiling steps during which the conversion would have taken place.Obviously �N1-SO-Ade was also converted fast, since no original product wasobserved in the analyses. Instead, the corresponding deamination product, �N1-SO-Hyp, was identified. The estimation of the rate constant of �N1-SO-Ade from theformation of �N1-SO-Hyp was difficult since a considerable proportion of N1-sub-stituted Ade undergoes Dimroth rearrangement even at neutral conditions [9]. This,in turn, may introduce some overestimation to the k of �N6-SO-Ade.

In the case of �N1-SO-Ade, the determination of k was hampered due to theinstability of the adduct as a result of Dimroth rearrangement taking place at aneutral pH [9,11]. The two diastereomers of the adduct are expected to be formedat same concentrations. However, in the case of �N1-SO-Ade there was anunexpectedly large difference between the diastereomers indicating coeluting impu-rities. The instability of �N1-SO-Ade can be seen in Fig. 4. While the concentrationof the stable adducts continued to rise after 32 h, the concentration of the�N1-SO-Ade remained unchanged or even decreased.

The 32P-postlabelling technique was also used to follow the formation of nucleaseP1 resistant SO-DNA adducts. These included �N1-SO-Ade, �N1-SO-Ade (onlythe first eluting diastereomer), �N2-SO-Gua and �N3-SO-Ura (Fig. 5). The forma-tion of �N6-SO-Ade was also followed, even though it is partly sensitive towardsnuclease P1 [13]. The stability of the adducts was verified up to 120 h for�N2-SO-Gua, �N3-SO-Ura and �N6-SO-Ade. The instability of the �N1-SO-Ade inDNA was clearly observed in the 32P-postlabelling analyses and t1/2 of the adductwas estimated to be 94 h. The corresponding �-isomer was considerably more stablewith an estimated t1/2 of 450 h. In the milder hydrolysis conditions of the32P-postlabelling assay (as compared to the enzymatic hydrolysis for the UVdetection) N3-SO-Cyt was not readily deaminated. Thus, in the case of �N3-SO-Ura the lag in increase of adduct concentration was observed, as expected.

3.3. Relati�e proportions of different adducts in the course of time

When the relative proportions of the SO adducts in DNA (only adducts boundto DNA, excluding the spontaneously depurinated adducts) were plotted at 1, 16and 144 h (Fig. 6), a dramatic decrease in the proportions of �N7-SO-Gua as wellas �N3-SO-Ade was observed, as expected. However, somewhat surprising was theobservation that the proportion of �N7-SO-Gua remained essentially unchanged,even though marked depurination is taking place (as can be seen in Fig. 3). This isobviously due to the considerably faster depurination of the both isomers ofN3-SO-Ade and slightly faster depurination of the �N7-SO-Gua, leading to verylow proportions of these adducts with respect to �N7-SO-Gua.

Increased proportions of all the adducts determined by enzymatic hydrolysis wasobserved. This trend was clear even in the case of the unstable �N1-SO-Ade. Theincrease in the proportion was obviously due to the longer t1/2, as compared to thedepurinating adducts. The highest proportion of the stable adducts after 144 hincubation was detected for �N6-SO-Ade, �N2-SO-Gua and �N3-SO-Ura (Fig. 6).


Fig. 4. Formation of SO-induced DNA adducts in the course of time determined after enzymatichydrolysis as alkylated nucleosides. When resolved, the data points are given for each diastereomer ofthe nucleoside adduct. Otherwise, the levels given show the sum of the two diastereomers.

M.

Koskinen

etal./

Chem

ico-B

iologicalInteractions

138(2001)

111–

124120

Fig. 5. Formation of nuclease P1 resistant adducts determined by the 32P-postlabelling technique. The adducts were analysed as alkylated nucleotides. Inthe case of �N2-SO-Gua, data points for resolved diastereomers are given; in the case of �N1-SO-Ade only the first eluting diastereomer is shown; in theall other cases adducts the diastereomers are not resolved and the levels given show the sum of the two diastereomers.


Fig. 6. Relative proportions (percentage) of various adducts in the course of incubations, determined byneutral depurination (excluding spontaneously depurinated fractions) and enzymatic hydrolysis.

4. Discussion

In order to compare the intrinsic reactivity of the different nucleophilic sites indouble-stranded DNA towards SO, we determined rate constants for differentalkylation products. For the primary alkylation sites the increase in the alkylationin the beginning of the reaction was linear allowing the calculation of k. The highestk-values for the formation of �N7-SO-Gua and �N7-SO-Gua are in good agree-ment with the fact that the mono-substituted epoxides mainly alkylate the N7-posi-tion in Gua [5]. When the reactivity of SO is compared to aliphatic epoxides, itsk-value for �N7-SO-Gua of 0.18×10−4 is considerably lower than that forethylene oxide, i.e. 0.96×10−4 [14], and slightly lower than that of propyleneoxide, 0.25×10−4 [15]. Therefore, our finding on the reactivity of the �-carbon ofSO is in accordance with the observation that the reaction rate decreases with theincreasing length or lipophilicity of the substituent [16,17]. As compared toaliphatic alkyl epoxides, the total burden of alkylation due to SO is, however,increased somewhat because of the reactivity of the �-carbon.

The other nucleophilic sites in double-stranded DNA showed considerably lowerk as compared to N7-position of Gua, but many of the adducts at these sites mayhave a pronounced biological importance because of their chemical stability. As thelabile adducts reach a steady-state, the relative proportion of these stable adductscontinues to increase over time. Further, many of the products at these sites(�N2-SO-Ade, �N2-SO-Gua, �N3-SO-Ura, �N4-SO-Cyt) occupy Watson–Crickbase-pairing positions that may lead to increased mutagenicity by disrupting thenormal hydrogen bonding. The decreased accessibility, because of base-pairing,obviously also contributes to the lower k as compared to the N7-Gua. The highestk for the stable adducts was found for �N6-SO-Ade. This value does not, however,reflect only the reactivity at the N6-position of adenine, since it obviously contains


a portion due to the Dimroth rearrangement of the �N1-SO-Ade, taking placeduring the lengthy hydrolysis procedures. �N2-SO-Gua and �N3-SO-Cyt (deami-nating to �N3-SO-Ura) also showed marked reactivity. In the case of the secondaryalkylation products (i.e. rearrangement and deamination products), the shapes ofthe curve for adduct formation versus time were sigmoidal. This was seen for�N3-SO-Ura and �N6-SO-Ade that are formed by deamination and the Dimrothrearrangement from �N3-SO-Cyt and �N1-SO-Ade, respectively.

In the course of incubation, the proportion of unstable adducts decreased at thesame time as the proportion of the stable adducts increased. Somewhat contrastingto our earlier studies with nucleotides and DNA was the observation that �N7-SO-Gua was slightly less stable than the �-isomer [10,18]. In DNA the steric effects areobviously different to those in the nucleotides leading to the observed differencebetween DNA and nucleotides. The differing stabilities of isomers of N7-SO-Gualead to a large difference in the proportions of �N7-SO-Gua and �N7-SO-Guaremaining in DNA at 144 h. If the incubation had been extended beyond 144 h theproportion of �N7-SO-Gua would also have been decreased. SO substitutedN1-adenines in DNA are also known to be unstable [11]. The suggested half-life of450 h for the �-isomer of N1-substituted adenine is rather long as compared to thecorresponding value for propylene oxide, i.e. ca. 220 h [19], and somewhatunexpected since earlier these SO-adducts were found to be converted readily inDNA [7]. Because of the long t1/2, the proportion of �N1-SO-Ade was increased upto 144 h. However, with longer incubations, the proportion would have beenexpected to decrease. Regarding extended incubations, with no additional epoxidesupplied, SO-DNA adducts expected to reach the highest levels are �N6-SO-Ade,�N2-SO-Gua and �N3-SO-Ura. However, in the case of chronic in vivo exposure,as styrene or SO is continuously or repeatably available, the adducts are expectedto reach a steady-state level, in which the proportions reflect both the chemicalstability and DNA repair. The proportions of DNA adducts in a steady-state invivo would be an interesting subject to study.

SO-alkylation at O6-position of Gua can also be detected in the analysesinvolving enzymatic hydrolysis and UV-detection (data not shown). However, theywere found only in small amounts, and close eluting peaks (especially �N6-SO-Ade)and chemical instability of these adducts hampered their quantitation. The propor-tion of the O6-substituted products in the current analyses could be estimated to beca. 1%.

Based on the chemical properties presented in this study, with regard to biomon-itoring of workers exposed to styrene the following adducts should be considered astentative biomarkers: �N7-SO-Gua, �N1-SO-Ade, �N6-SO-Ade, �N2-SO-Gua and�N3-SO-Ura. �N7-SO-Gua seems to be a most suitable candidate due to therelatively high reactivity as well as its higher stability and better suitability to the32P-postlabelling assay as compared to the �-isomer [20]. However, because of itsrather short t1/2 the adduct may reflect a rather recent exposure. For �N1-SO-Adethe t1/2 should be long enough for some accumulation to take place, and this adductis also suitable for the 32P-postlabelling assay [13]. �N6-SO-Ade should also beconsidered because of its stability, which may lead to accumulation of adducts over


extended periods of exposure. However, sensitive analytical methods are stilllacking for this adduct. �N2-SO-Gua and �N3-SO-Ura are suitable substrates forthe 32P-postlabelling assay, and are also chemically stable enough to accumulateduring long exposures. Because DNA repair has an important role in addition tothe chemical stabilities for the formation of the steady-state, the relative DNAadduct levels in vivo remain to be established.

Acknowledgements

The study was supported by the European Communities, QLK4-1999-01368, byGACR-313/99/1460 and by the Swedish Council for Work Life Research.

References

[1] IARC, Styrene, IARC Monogr. Eval. Carcinog. Risks Hum. 60 (1994) 233–320.[2] IARC, Styrene-7,8-oxide, IARC Monogr. Eval. Carcinog. Risks Hum. 60 (1994) 321–346.[3] K. Hemminki, P. Vodicka, Styrene: from characterisation of DNA adducts to application in

styrene-exposed lamination workers, Toxicol. Lett. 77 (1995) 153–161.[4] K. Hemminki, M. Koskinen, H. Rajaniemi, C. Zhao, DNA adducts, mutations, and cancer, Regul.

Toxicol. Pharmacol. 32 (2000) 264–275.[5] M. Koskinen, K. Plna, Specific DNA adducts induced by some mono-substituted epoxides in vitro

and in vivo, Chem. Biol. Interact. 129 (2000) 209–229.[6] F. Latif, R.C. Moschel, K. Hemminki, A. Dipple, Styrene oxide as a stereochemical probe for the

mechanism of aralkylation at different sites on guanosine, Chem. Res. Toxicol. 1 (1988) 364–369.[7] T. Barlow, A. Dipple, Formation of deaminated products in styrene oxide reactions with deoxycy-

tidine, Chem. Res. Toxicol. 12 (1999) 883–886.[8] M. Koskinen, D. Calebiro, K. Hemminki, Styrene oxide-induced 2�-deoxycytidine adducts: implica-

tions for the mutagenicity of styrene oxide, Chem. Biol. Interact. 126 (2000) 201–213.[9] C. Qian, A. Dipple, Different mechanisms of aralkylation of adenosine at the 1- and N6-positions,

Chem. Res. Toxicol. 8 (1995) 389–395.[10] M. Koskinen, P. Vodicka, K. Hemminki, Adenine N3 is a main alkylation site of styrene oxide in

double-stranded DNA, Chem. Biol. Interact. 124 (2000) 13–27.[11] T. Barlow, J. Takeshita, A. Dipple, Deamination and Dimroth rearrangement of deoxyadenosine-

styrene oxide adducts in DNA, Chem. Res. Toxicol. 11 (1998) 838–845.[12] M. Koskinen, E.K.H. Schweda, K. Hemminki, Alkylation of anionic DNA bases by styrene

7,8-oxide, J. Chem. Soc. Perkin Trans. 2 (1999) 2441–2446.[13] M. Koskinen, P. Vodicka, L. Vodickova, K. Hemminki, 32P-postlabelling/HPLC analysis of various

styrene-induced DNA adducts in mice, Biomarkers 6 (2001) 175–189.[14] D. Segerback, Reaction products in hemoglobin and DNA after in vitro treatment with ethylene

oxide and N-(2-hydroxyethyl)-N-nitrosourea, Carcinogenesis 11 (1990) 307–312.[15] D. Segerback, S. Osterman-Golkar, B. Molholt, R. Nilsson, In vivo tissue dosimetry as a basis for

cross-species extrapolation in cancer risk assessment of propylene oxide, Regul. Toxicol. Pharmacol.20 (1994) 1–14.

[16] R. Kumar, J. Staffas, A. Forsti, K. Hemminki, 32P-postlabelling method for the detection of7-alkylguanine adducts formed by the reaction of different 1,2-alkyl epoxides with DNA, Carcino-genesis 16 (1995) 483–489.

[17] A. Hemminki, T. Vayrynen, K. Hemminki, Reaction kinetics of alkyl epoxides with DNA andother nucleophiles, Chem. Biol. Interact. 93 (1994) 51–58.


[18] K. Hemminki, A. Alhonen-Raatesalmi, P. Koivisto, P. Vodicka, Synthesis and stability of 2�-de-oxyguanosine 3�-monophosphate adducts of dimethyl sulfate, ethylene oxide and styrene oxide,Chem. Biol. Interact. 75 (1990) 281–292.

[19] K. Plna, R. Nilsson, M. Koskinen, D. Segerback, 32P-postlabelling of propylene oxide 1- andN6-substituted adenine and 3-substituted cytosine/uracil: formation and persistence in vitro and invivo, Carcinogenesis 20 (1999) 2025–2032.

[20] P. Vodicka, K. Hemminki, 32P-postlabeling of N-7, N2 and O6 2�-deoxyguanosine 3�-monophos-phate adducts of styrene oxide, Chem. Biol. Interact. 77 (1991) 39–50.

Documents

Kinetics of formation of specific styrene oxide adducts in double-stranded DNA