Biotransformation of the double bond in allyl glycidyl ether to an epoxide ring. Evidence from hemoglobin adducts in mice.

ABSTRACT

Allyl glycidyl ether (AGE) is used industrially in the production of various epoxy resins. The compound is mutagenic and evidence for carcinogenicity in mice and rats has been reported. A previous study in mice showed that AGE reacts directly, without metabolic activation, with N-terminal valine in hemoglobin to form adducts (AGEVal). Metabolism of AGE may lead to formation of diglycidyl ether (I) through epoxidation of the double bond or 1-allyloxy-2,3-dihydroxypropane (II) through hydrolysis of the epoxide ring. 2,3-Dihydroxypropyl glycidyl ether (III) may be formed either by hydrolysis of I or epoxidation of II. The main aim of the present study was to investigate if AGE is metabolized to the reactive epoxides I or III by analysis of adducts with hemoglobin. Nine male mice (C3H/Hej) were administered AGE dissolved in tricaprylin, 4 mg/mouse, by intraperitoneal (i.p.) injection. Eleven male mice were administered 4 mg/mouse of AGE dissolved in acetone, by skin application. Adducts of I or III with N-terminal valine, N-(2-hydroxy-3-(2,3-dihydroxy)propyloxy)propylvaline (diOHPrGEVal), were demonstrated in mice administered AGE by i.p. injection. The levels were in the range 1600-5600 pmol/g globin. The level of diOHPrGEVal in mice administered AGE by skin application (n = 5) was below the detection limit of the analytical method, 20 pmol/g globin. The level of AGEVal, analyzed in mice administered AGE by skin application (n = 6), was about 20 pmol/g globin (median value), as compared with 1600 pmol/g globin previously found in mice administered AGE by i.p. injection. Neither AGEVal nor diOHPrGEVal were detected in control animals. Both adducts were analyzed using a modified Edman method for derivatization and using gas chromatography/tandem mass spectrometry for detection. The hydroxyl groups of the Edman derivative of diOHPrGEVal were protected by acetylation.