In vivo relevance of in vitro detected estrogenic effects of food associated compounds

The aim of the present thesis was to study the in vivo relevance of in vitro detected estrogenic effects of food associated compounds, with emphasis on prenatal exposure. The estrogenic potency of 21 food packaging-associated compounds was studied in ERα or ERβ transfected U2-OS (human osteoblasts devoid of endogenous estrogen receptors) cell lines. Six plasticizers and three anti-oxidants were slightly estrogenic in the ERα cells. BPA, NP, tris (2-ethylhexyl) trimellitate (TEHTM), propyl gallate and butylated hydroxy anisole (BHA) were estrogenic both in ERα and ERβ cells. These compounds appeared to be more estrogenic relative to estradiol (E2) in ERβ than in ERα cells. To study the in vivo relevance of these effects, in vivo biomarker-responses of BPA, NP, DEHA, DEHP, DIHP, p,p’-DDE and quercetin were studied in ER-Luc male mice. Of these seven compounds, BPA, DEHP and quercetin induced estrogenic effects after a single oral dosage at exposure levels 10-104 times higher than the established Tolerable Daily Intakes (TDI’s). It remains to be seen whether these margins are sufficiently high to allow the conclusion that these compounds are unlikely to represent a human health risk. As exposure to estrogenic compounds during developmental stages could be an important risk factor in developing hormone dependent cancers later in life, these seven compounds were studied in pregnant ER-Luc mice as well. After oral exposure of the mother animal the compounds were unable to significantly induce luc-activity in any of the tissues including fetuses. Unexpectedly however, NP, BPA and DIHP significantly lowered the placental luc-activity. The results indicate that at the current levels of exposure to food associated estrogenic compounds, direct estrogenic effects in the fetus are not expected. The mechanism and consequences of the significant luc-reduction in the placenta should be investigated to a further extent to elucidate its possible significance. Because of the absence of significant luc-induction in the fetuses, the fate and distribution of radioactively labeled E2, NP and p,p’-DDE were studied in pregnant C57black mice in order to investigate whether the compounds can actually reach the fetuses. E2 did not reach the fetus at a level above the detection limit, NP and p,p’-DDE levels were above the detection limit in fetuses exposed via the mother from GD8-16. Levels of E2 and NP detected in the placenta were significantly higher than those in the fetuses, up to five fold for E2 and three fold for NP, possibly due to prevention of transport of the compounds to the fetuses by placental binding proteins. Based on the findings of the present thesis it was concluded that in spite of the in vitro estrogenicity of various food-born estrogens, in vivo estrogenic effects are not likely expected as estrogenicity in vivo is shown at levels 10-104 fold higher than the TDI’s for humans and placental binding proteins may reduce fetal exposure. However, the estrogenic compounds were given as a single acute dose, whereas TDI values and also risks associated with dietary exposure can be expected to result from chronic combined exposure. Therefore, it remains to be seen whether the estimated margins are sufficiently high. Future experiments are therefore needed that focus on long term and combined exposure. Another factor that remains to be solved relates to the fact that only in in vitro assays the relative ability of compounds to differentially activate ERα or ERβ can be quantified. All together the results of the present thesis reveal that the extent at which in vitro estrogenic effects may result in in vivo estrogenicity may vary with the compound under study and should be evaluated on a case by case basis taking differences in absorption, distribution, metabolism and excretion, the intrinsic estrogenic potency of chemicals for either the ERα or ERβ and the differential levels of expression of these receptors in different tissues in vivo into account.