Bisphenol A diglycidyl ether as a potential metabolic source of bisphenol A.

Epoxy resins find widespread use as adhesives and coatings in the construction and food industries. The majority of the products are based on resins that use bisphenol A diglycidyl ether (BADGE) as a precursor. This raises the question if BADGE could be a potential source of bisphenol A (BPA). The recent paper by Perez et al. (1) on the estrogenicity of BPA and some of its analogs and derivatives states that little is known about the disposition and metabolism of this compound in mammals. In fact, preliminary reports of its fate in rats have been published in abstract form (2,3). In searching for information on the metabolism of BPA and its derivatives, the authors identified our papers on the fate of BADGE in mice (4,5). These papers were based on a thorough study of the metabolism of [14C]BADGE in which most of the metabolites were identified. The objectives of the study were twofold: to describe the overall fate of BADGE and to assess the amount of O-dealkylation to either glycidaldehyde (a bacterial mutagen) or to glyceraldehyde (relatively innocuous). Because the 14C-glycidyl label was not available to us, we hoped to use the yield of BPA (and its proposed glucuronide and sulfate conjugates) as indicators of the generation of glycidaldehyde. To our surprise, BPA could not be confirmed as a metabolite of BADGE. BADGE afforded six metabolites (and three conjugates), mostly novel, which accounted for 1-24% of the dose. These were formed by hydrolytic and glutathione-dependent opening of the epoxide rings. Further metabolism of the products occurred by oxidation and decarboxylation. Loss of one glycidyl group could be confirmed (1-3%) of the dose. The only evidence for the presence of BPA was in a neutral solvent extract of feces. This fraction (0.1% of the dose) was subjected to further chromatographic analysis and a small proportion, well under 0.1% of the dose, may have been BPA. This could not be confirmed. When BADGE bis-diol was dosed to mice, no BPA could be detected. Further attempts were made to demonstrate deglycidylation (and BPA production) using liver microsomes from mouse, rat, and rabbit, operating in an oxidative mode. Even inhibiting epoxide hydrase with trichloromethyloxirane, no BPA could be detected. Citing our papers (4,5), Perez et al. (1) state that "a small amount ofBADGE (-5%) underwent oxidative dealkylation to give glycidaldehyde (which has alkylating properties) and BPA." Our studies showed that only traces (<<0.1%) of a metabolite which may have been BPA were formed from BADGE. Thus, the available evidence shows that BADGE is not a significant metabolic source of bisphenol A.