Variability in human sensitivity to 1,3-butadiene: Influence of polymorphisms in the 5′-flanking region of the microsomal epoxide hydrolase gene (EPHX1)

The carcinogenic effects of 1,3-butadiene (BD), a mutagenic chemical widely used in the manufacture of synthetic rubber, are likely initiated through its epoxide metabolites. In humans, these epoxides are detoxified predominantly by hydrolysis, a reaction mediated by the microsomal epoxide hydrolase (mEH; EPHX1) enzyme. It appears reasonable to hypothesize that BD-exposed individuals possessing lower mEH detoxification capacity may have elevated risk of adverse health effects. The interindividual levels of mEH enzymatic activity vary considerably, and polymorphisms in the mEH gene may contribute to this variability. In addition to the well-studied coding region polymorphisms encoding Tyr113His and His139Arg substitutions, seven other polymorphic sites in the 5′-flanking region of the mEH gene have been reported. These polymorphisms appear to differentially affect mEH gene transcriptional activities. The 5′-flanking region polymorphisms exist in two linkages, the -200 linkage (-20OC/T, -259C/T, -290T/G) and the -600 linkage (-362A/G, -613T/C, -699T/C), whereas the -399T/C polymorphism exists as an independent site. Because these polymorphisms may affect total mEH enzymatic activity, we hypothesized that they influence the mutagenic response associated with occupational exposure to BD. We genotyped the 5′-region of the mEH gene in 49 non-smoking workers from two styrene-butadiene rubber facilities in southeast Texas and evaluated the linkage patterns against results obtained from an autoradiographic HPRT mutant lymphocyte assay, used as a biomarker of genotoxic effect. In the study population, 67% were exposed to low BD levels, <150 parts per billion, and 33% were exposed to >150 ppb. We used the observed HPRT mutant (variant) frequency (VF) in the studied population and a 4-way first-order interaction statistical model to estimate parameters that describe the influence of exposure, genotypes and the interaction between the two on the HPRT VF in the target population. The background (baseline) VF, defined as the VF (× 10_-6) ± S.E.M. at low levels of BD exposure (<150 ppb) where all the genotypes under study are homozygous wild-type, was estimated to be 4.02 ± 1.32. Exposure to >150 ppb of BD alone resulted in an estimated increase in VF of 3.42 ± 2.47 above the baseline level. Inheritance of the variant ATT allele in the -600 linkages resulted in an estimated increase in VF of 3.39 ± 1.67 above the baseline level. When the interaction between BD exposure and the ATT allele in the -600 linkage group was considered, a statistically significant positive interaction was observed, with an estimated increase in the VF of 10.89 ± 2.16 (95% CI = 6.56-15.20; p = 0.0027) above baseline. These new data confirm and extend our previous findings that sensitivity to the genotoxic effects of BD is inversely correlated with predicted mEH activity.