Styrene is widely used in the production of various plastics, synthetic rubber and resins. Occupational exposure occurs mainly via inhalation and relatively high exposure occurs due to its use in manual application techniques. The aim of this study was to evaluate if genetic polymorphisms of metabolic enzymes modulate the level of urinary styrene metabolites and styrene oxide adducts with N- terminal valine of human globin (SO-Hb) in 75 workers occupationally exposed to styrene and 77 unexposed controls. The mean air concentration of styrene in the breathing zone of workers (30.4 ppm) was higher than the threshold limit value of 20 ppm recommended by the American Conference of Governmental Industrial Hygienists (ACGIH), and the biological exposure index adopted by the ACGIH for exposure to styrene prior to the next shift

Styrene is one of the most important organic chemicals used worldwide. In humans, styrene metabolism involves oxidation by cytochrome P450 monooxygenases (CYPs) to styrene-7,8-oxide, an epoxide thought to be responsible for the genotoxic effects of styrene exposure, and detoxification by means of epoxide hydrolase (mEH) and glutathione S-transferases (GSTs). The objective of this study was to investigate if genetic polymorphisms of metabolic enzymes modulate the level of urinary styrene metabolites and styrene oxide adducts with N-terminal valine of human globin (SO-Hb) in 75 workers occupationally exposed to styrene and 77 unexposed controls. The mean air concentration of styrene in the breathing zone of workers (30.4 ppm) was higher than the threshold limit value of 20 ppm recommended by the American Conference of Governmental Industrial Hygienists (ACGIH), and the biological exposure index adopted by the ACGIH for exposure to styrene prior to the next shift (MA+PGA= 400mg/g creatinine) was exceeded, indicating that styrene exposure for this group of workers was higher than recommended. A highly significant correlation was observed between styrene concentration in the breathing zone and the MA+PGA in urine of workers (r=0.85, P<0.001). The levels of SO-Hb adducts in exposed workers were significantly increased as compared with controls, although no difference was observed between subjects stratified as high and medium exposure categories based on MA+PGA excretion. Regarding the effect of the genetic polymorphisms we found that the level of SO-Hb adducts might be modulated by the predicted mEH enzymatic activity in the exposed workers. From our data we conclude that SO-Hb adduct measurement is a complementary method to MA+PG measurement for assessing exposure to styrene at occupational and environmental levels, which reflects a more extensive exposure period. Styrene is widely used in the production of various plastics, synthetic rubber and resins. Occupational exposure occurs mainly via inhalation and relatively high exposure occurs due to its use in manual application techniques. The aim of this study was to evaluate if genetic polymorphisms of metabolic enzymes modulate the level of urinary styrene metabolites and styrene oxide adducts with Nterminal valine of human globin (SO-Hb) in 75 workers occupationally exposed to styrene and 77 unexposed controls. The mean air concentration of styrene in the breathing zone of workers (30.4 ppm) was higher than the threshold limit value of 20 ppm recommended by the American Conference of Governmental Industrial Hygienists (ACGIH), and the biological exposure index adopted by the ACGIH for exposure to styrene prior to the next shift (MA+PGA= 400mg/g creatinine) was exceeded, indicating that styrene exposure for this group of workers was higher than recommended The levels of SO-Hb adducts were significantly higher (P<0.01) in the exposed subjects (5.98 ± 0.41 pmol/g globin) when compared with controls (2.59 ± 0.25 pmol/g globin) and a significant difference (P<0.02) was found in levels of SO-Hb adducts between non-smokers (2.19 ± 0.27 pmol/g globin) and smokers (3.55 ± 0.51 pmol/g globin) among the control group. From our data we conclude that SO-Hb adduct measurement is a sensitive and specific means of assessing exposure to styrene at occupational and environmental level. Regarding the effect of the genetic polymorphisms, we found that the level of SOHb adducts might be modulated by the predicted mEH enzymatic activity in the exposed workers. In exposed individuals, EPHX1 genotypes associated with low hydroxylation reaction yielded an increase of SO-Hb adducts levels compared to high activity EPHX1 genotypes. Since we know that mEH rapidly converts SO to phenyl Cover Letter

Styrene is widely used in the production of various plastics, synthetic rubber and resins.
Occupational exposure occurs mainly via inhalation and relatively high exposure occurs due to its use in manual application techniques.
The aim of this study was to evaluate if genetic polymorphisms of metabolic enzymes modulate the level of urinary styrene metabolites and styrene oxide adducts with Nterminal valine of human globin (SO-Hb) in 75 workers occupationally exposed to styrene and 77 unexposed controls.
The mean air concentration of styrene in the breathing zone of workers (30.4 ppm) was higher than the threshold limit value of 20 ppm recommended by the American Conference of Governmental Industrial Hygienists (ACGIH), and the biological exposure index adopted by the ACGIH for exposure to styrene prior to the next shift (MA+PGA= 400mg/g creatinine) was exceeded, indicating that styrene exposure for this group of workers was higher than recommended The levels of SO-Hb adducts were significantly higher (P<0.01) in the exposed subjects (5.98 ± 0.41 pmol/g globin) when compared with controls (2.59 ± 0.25 pmol/g globin) and a significant difference (P<0.02) was found in levels of SO-Hb adducts between non-smokers (2.19 ± 0.27 pmol/g globin) and smokers (3.55 ± 0.51 pmol/g globin) among the control group. From our data we conclude that SO-Hb adduct measurement is a sensitive and specific means of assessing exposure to styrene at occupational and environmental level.
Regarding the effect of the genetic polymorphisms, we found that the level of SO-Hb adducts might be modulated by the predicted mEH enzymatic activity in the exposed workers. In exposed individuals, EPHX1 genotypes associated with low hydroxylation reaction yielded an increase of SO-Hb adducts levels compared to high activity EPHX1 genotypes. Since we know that mEH rapidly converts SO to phenyl Cover Letter ethylene glycol (which is considered as rather non-toxic), in the case of excessive styrene exposure we can hypothesise that individuals with low enzymatic activity have higher levels of free SO that can rapidly react with the N-terminal valine in haemoglobin. However, this result should be regarded as preliminary, since the number of individuals in each category of mEH expected enzymatic activity is small.

"Response to Reviews"
-In order to respond to the first comment of the referee, I agree that the last sentence of the abstract is a little bit strong, as we show that MA+PGA correlated with styrene in air but Hb adducts did not. (However the Hb adducts are measuring dose over a much longer period that MA+PGA). I have suggested an alternative end to the abstract and on page 15.
-The reviewer's second comment relates to background levels and the effect of smoking. I have added a bit more about background levels on page 14 (add new reference Toxicology Letters, 1999, vol 108, 117-126). I presume that any effects of smoking in the exposed population may be overwhelmed by the variation in occupational exposure over the lifetime of the adduct formation. It is also conceivable that there is saturation in the metabolite production or adduct formation! This would explain why high doses do not increase the adduct and why smoking does not increase the adduct in the exposed population. I have modified the text a little on page 14 .
-Concerning to normality of the variables only the styrene in air departed significantly from the normality. MA+PGA (exposed): Median=381and Mean=443mg/g cr; Hb Adducts (control): Median=1.95 and Mean=2.59pmol/g globin and exposed median=5.78 and mean=5.98pmol/g globin -As said the reviewer the limit of detection is expressed in mg/L instead of mg/g creatinine, in the paper we put in mg/g creatinine by distraction! -Finally, as suggested by the reviewer we changed the Table 5 and the text in page 11 concerning to this table.

LIST OF CHANGES
-Abstract, line 19"…, although no difference was observed between subjects stratified as high and medium exposure categories based on MA+PG excretion." -Abstract, line 22 "From our data we conclude that SO-Hb adduct measurement is a complementary method to MA+PG measurement for assessing exposure to styrene at occupational and environmental levels, which reflects a more extensive exposure period." -Statistical analysis, page 8, line 16, "The level of styrene in air was the only parameter that departed significantly from normality (median=19.0 ppm vs mean=30.4 ppm in exposed group)." -Results, page 9, line 8 "…limit of detection (14 mg/L)." -Since all the individuals in control group had values of MA and PGA below the LOD we did not use in the statistical analysis.
-Discussion, page 14, line 3 "…and exposure data a significant difference…" -Discussion, page 14, line 7 "An effect in the exposed group may be undetectable because it is overwhelmed by the variability in the adduct levels induced by occupational exposure, or conceivably because of saturation in the metabolic activation process of styrene and/or adduct formation." -Discussion, page 14, line 15 "Rappaport and Yeowell O'Connell calculate that the background levels of the SO-albumin adduct are too great to have arisen from non-occupational exposures to styrene or from cigarette smoking (Rappaport and Yeowell O'Connell, 1999). Possibilities were suggested that these adducts represent unknown exposures or endogenous production of reactive species which react with proteins to produce the same analyte in the assay." -Discussion, page 14, line 22 "but despite the fact that a significant correlation was found between MA+PGA and SO-Hb adducts, this difference was not statistically significant. It should be appreciated also however that the exposure classification was based on urinary metabolites which reflect a different exposure period than Hb adducts.

Introduction 1 2
Styrene is a monomer widely used in the production of plastics, synthetic rubber 3 and polyester resins. The highest human exposure to styrene occurs during the 4 production of fiberglass-reinforced polyester products (Miller et al., 1994), where 5 unsaturated resins containing about 40% styrene as a reactive diluent are commonly 6 used, and as much as 10% of it can evaporate into the workroom air during lamination 7 by hand procedures. 8 In humans, styrene metabolism has been well characterised (Sumner and Fennell, 9 1994). The first step of styrene metabolism is oxidation by cytochrome P450 enzymes 10 (CYP) including CYP2E1, CYP2B6, CYP1A2 and other isozymes (Nakajima et al., 11 1994) to styrene-7,8-oxide (SO), which is a highly reactive epoxide and has been 12 classified as a probable human carcinogen ( The study population consisted of 75 individuals, who have been occupationally 5 exposed to styrene for 9 years (SD 10 years) in factories manufacturing glass-fibre 6 reinforced plastics in Portugal. The workers were mainly involved in hand-spraying 7 lamination processes using unsaturated polyester. The control group included 77 healthy 8 volunteers of comparable sex and age distribution, with no history of styrene exposure 9 and living in the same areas as the exposed individuals. Demographic characteristics of 10 the study population are shown in Table 1. 11 Prior to the study, a questionnaire was supplied concerning health status, smoking 12 habits, alcohol consumption, medication and occupational history. This study was 13 The mean values, standard errors (SE) and ranges recorded for styrene in air, 3 styrene metabolites in urine and for SO-Hb adducts are reported in Table 2. During 4 sample preparation for the SO-Hb adducts, 6 control and 18 exposed samples were lost, 5 due to technical reasons. 6 Our results showed that MA+PGA levels in the control group were below the limit 7 of detection (14 mg/L). The study of workplace air revealed the presence of toluene and 8 acetone, although air levels of these two chemicals were less than 1% of the styrene 9 concentration. The styrene exposures, expressed as TWA for each subject, were 10 between 0.5 and 114 ppm, mean 30.4 ppm ( Table 2). Considering the current threshold 11 Levels of SO-Hb adducts were significantly higher (P<0.01) in the exposed 19 subjects (5.98 ± 0.41 pmol/g globin) when compared with controls (2.59 ± 0.25 pmol/g 20 globin) ( Table 2). 21 The effect of cigarette smoking on the levels of SO-Hb adducts was further 22 investigated (Table 3). Among the control subjects the smokers had a significant 23 increase of SO-Hb adducts when compared with the non-smokers (P<0.02). In the 24 exposed group, no significant difference was observed between smokers and non-1 smokers. 2 To further investigate the effect of styrene exposure on the levels of SO-Hb 3 adducts, the subjects were classified into three exposure categories on the basis of 4 urinary levels of MA+PGA [zero (controls); MA+PGA ≤ 400 mg/g Cr (medium-level 5 exposure, mean 151 ± 16 mg/g Cr, range 23-390 mg/g Cr); MA+PGA > 400 mg/g Cr 6 (high-level exposure, mean 813 ± 63 mg/g Cr, range 423-1770 mg/g Cr)]. The cut-off 7 value, 400 mg/g Cr, was based on the BEI value for MA+PGA currently proposed by 8 the ACGIH (2003) for exposure to styrene prior to the next shift. Figure 1 summarizes 9 the mean levels of SO-Hb adducts recorded for the subjects stratified by styrene 10 exposure. A significant difference was found between medium-level exposure group 11 (5.71 ± 0.45 pmol/g globin) and control group (2.59 ± 0.25 pmol/g globin; P<0.01), and 12 the high-level exposure group (6.25 ± 0.68 pmol/g globin) when compared with the 13 control group (P<0.01). Levels of SO-Hb adducts in the high-level exposure group were 14 higher than in the medium-level exposure group; however, the increase was not 15 statistically significant. 16 A strong and significant correlation was found between the concentration of 17 styrene in air and concentration of MA+PGA in the urine of the workers (r=0.85, 18 P<0.001, N=75). A significant correlation was also found between the individual levels 19 of MA+PGA and the SO-Hb adducts among exposed individuals (r=0.30, P<0.02, 20 N=57). However, no correlation was obtained between SO-Hb adducts and 21 concentration of styrene in the air, age and years of exposure. 22 The results of genotype analysis for sets of exposed workers, controls, and the 23 whole set are presented in Table 4. The distribution of genotypes in each group was in 24 Hardy-Weinberg equilibrium. 25 Table 5 shows results obtained for the effect of EPHX1 genetic polymorphisms 1 studied on the levels of protein adducts. Concerning the EPHX1 genotypes in codons 2 113 and 139, individuals were classified according to the expected enzymatic activity 3 (Sarmanová et al., 2000). We could not find any difference in the urinary metabolites 4 for each genotype of styrene metabolizing enzymes. However, the levels of SO-N-5 terminal valine adducts in haemoglobin suggest that EPHX1 genotypes might modulate 6 the levels of this endpoint (Table5). A decrease in the levels of this biomarker is 7 observed with an increase in EPHX1 activity in the exposed groups (altogether and by 8 the level of exposure). No effect was observed for the other polymorphisms evaluated in 9 this study on SO-Hb adducts. 10 12 4. Discussion 1 2 Our group of workers had highly variable exposure to styrene (0.5-114 ppm) and in 3 48% of them, the recommended TLV-TWA (20 ppm) by the ACGIH (2003) was 4 exceeded, indicating that styrene exposure for this group of workers was higher than 5 recommended. This was also confirmed by analysis of the urinary metabolites of 6 styrene (MA+PGA), which in 48% of the cases was higher than the biological exposure 7 index (BEI) proposed by the ACGIH (2003). The present data on styrene exposure 8 clearly showed a good correlation between styrene concentrations in inhaled air and its 9 urinary metabolites. The two major urinary metabolites, MA and PGA, are routinely 10 used as biomarkers of internal dose of styrene, and a clear correlation between their 11 levels and styrene in the workplace air has been reported in a number of studies 12 (Vodicka et al., 1999;Migliore et al., 2006). 13 However, SO is metabolised by two distinct metabolic pathways: microsomal 14 epoxide hydrolase (mEH) and glutathione-S-transferases (GST). In the urine of styrene 15 exposed factory workers, the impact of these metabolic pathways has been measured by Christakopoulos et al., 1993), we observed that SO-Hb adducts were significantly 12 higher in exposed subjects as compared to controls and correlated with internal dose 13 parameter (MA+PGA). 14 In a previous study by Christakopoulos et al. (1993) an average haemoglobin 15 adduct level of 18 pmol/g globin was observed in reinforced plastic workers who had 16 been exposed to an average styrene concentration of 74 ppm (0.24 adducts per ppm of 17 styrene) . The adduct level observed in this study is about a 3-fold less. However, the 18 styrene air concentration in Christakopoulos´ study was about 2.5 fold higher than that 19 of this study. The adduct levels observed in these two studies are therefore similar if we 20 expressed the amount of adducts per unit of exposure (0.24 versus 0.20 adducts/ppm 21 styrene). Significant differences have also been observed by Fustinoni et al. (1988), 22 who found significant differences in Hb adduct levels between workers (average 23 exposure estimated at 24 ppm) and controls. 24 Subjects were divided according to smoking status to determine any role that 1 cigarette smoking may have on SO-Hb adduct formation. By comparing the smoking 2 status and exposure data a significant difference was observed in levels of SO-Hb 3 adducts between non-smokers and smokers within the control group. In contrast, no 4 effect was observed within the exposed group when smoking status was taken into 5 account. From these results we conclude that the only significant effect of smoking 6 status on SO-Hb adducts is within the control group. An effect in the exposed group 7 may be undetectable because it is overwhelmed by the variability in the adduct levels 8 induced by occupational exposure, or conceivably because of saturation in the metabolic 9 activation process of styrene and/or adduct formation. 10 This supports the hypothesis of Rappaport et al. (1996) that cigarette smoking is a 11 source of SO-background adducts among people not occupationally exposed to styrene. 12 Because this adduct is also observed in the blood of non-smoking individuals this 13 indicates that other factors such as diet or environmental pollution can also contribute to 14 the formation of SO-Hb adducts. The presence of background levels of alkylated 15 proteins among control subjects as well as in unexposed experimental animals has been 16 reported for several chemicals (Welie et al., 1992;Farmer et al., 1993) including 17 styrene (Brenner et al., 1991;Christakopoulos et al., 1993;1996;18 Yeowell-O'Connell et al., 1996). Rappaport and Yeowell O'Connell calculate that the 19 background levels of the SO-albumin adduct are too great to have arisen from non-20 occupational exposures to styrene or from cigarette smoking (Rappaport and Yeowell 21 O'Connell, 1999). Possibilities were suggested that these adducts represent unknown 22 exposures or endogenous production of reactive species which react with proteins to 23 produce the same analyte in the assay. 24 To better evaluate the effect of styrene exposure on the levels of SO-Hb adducts 1 we stratified the subjects by styrene exposure on the basis of urinary levels of 2 metabolites (zero, medium-level, and high-level exposure). After stratification by these 3 three exposure categories, differences were found in levels of SO-Hb adducts between 4 the two exposed groups (medium-level and high-level exposure) and the controls. The 5 high-level exposure group had more SO-Hb adducts than the medium-level exposure 6 group, but despite the fact that a significant correlation was found between MA+PGA 7 and SO-Hb adducts, this difference was not statistically significant. It should be 8 appreciated also however that the exposure classification was based on urinary 9 metabolites which reflect a different exposure period than Hb adducts. 10 Age and exposure time did not influence the SO-Hb adducts, which is not 11 surprising because SO-Hb adducts persist with a half span of erythrocytes (120 days), 12 and their level is closely related to the concentration of SO in circulation. 13 Regarding the effect of the genetic polymorphisms, we found that the level of SO-14 Hb adducts might be modulated by the predicted mEH enzymatic activity in the 15 exposed workers. In exposed individuals, EPHX1 genotypes associated with low 16 hydroxylation reaction yielded an increase of SO-Hb adducts levels compared to high 17 activity EPHX1 genotypes. Since we know that mEH rapidly converts SO to phenyl 18 ethylene glycol (which is considered as rather non-toxic), in the case of excessive 19 styrene exposure we can hypothesise that individuals with low enzymatic activity have 20 higher levels of free SO that can rapidly react with the N-terminal valine in 21 haemoglobin. However, this result should be regarded as preliminary, since the number 22 of individuals in each category of mEH expected enzymatic activity is small. 23 In conclusion, the results of this study based on a reasonable number of 24 observations show a strong correlation between styrene concentration in the workplace 25 and styrene urinary metabolites and that SO-Hb adduct measurement is a sensitive and 1 specific means of assessing exposure to styrene at occupational and environmental 2 level. The present results suggest the importance of individual susceptibility factors in 3 modulating genotoxicity, although cautious interpretations are required since the size of 4 the studied population limits the power of many of the analysis. Because the effects of 5 these polymorphisms are relatively subtle, and some important alleles are relatively 6 rare, a much larger study population will be necessary to evaluate their effects on 7 biomarkers, especially when gene-gene interactions are considered. Conducting 8 adequate studies for addressing these questions represents a major challenge.  Cigarettes/day, mean ± SD 5 ± 7 6 ± 6  SO-Hb adducts (pmol/g globin) 2.59 ± 0.25 (N=71) 5.98 ± 0.41 (N=57)* *P<0.01, when compared the SO-Hb adducts between the controls and the exposed group Table 3. Effect of smoking status on the SO-Hb adducts (pmol/g globin)