Production of low-estrogen goldfish diet for in vivo endocrine disrupter test.

A low-estrogenic diet for goldfish Carassius auatus was produced for an in vivo estrogen activity test, because commercial fish feed has estrogenic activity and may affect the results of estrogen assays. The newly produced diet (FD5) was formulated with defatted rice bran and casein, and did not contain any soybean meal or fish meal. Phytoestrogen contents (genistein, daidzein, equol, and coumestrol) of FD5 were measured by liquid chromatography-mass spectroscopy/mass spectroscopy (LC-MS/MS) and compared with those of the commercial trout diet (TD) and carp diet (CD). The genistein, daidzein, and coumestrol contents of TD and CD were much higher (5-2000 times) than those of FD5, but equol was detected only in FD5. Estrogenic activity of the fish diets was estimated in vitro by the yeast estrogen-screen assay (YES assay). The estrogenic activity was detected in TD and CD, but not in FD5. The in vivo estrogenic activity of the diets was examined by determining the production of vitellogenin in male goldfish. When male goldfish were fed TD or CD, plasma vitellogenin levels increased, but fish that were fed FD5 maintained low vitellogenin levels. These results indicate that FD5 produced in the present study has a low estrogenic activity, and FD5 would be suitable for the in vivo estrogen activity test using goldfish.

Survey of estrogenic activity in fish feed by yeast estrogen-screen assay.

Fishes have been used as laboratory animal for research of estrogenic endocrine disrupters by many researchers. However, much less attention was paid to the possibility that compounds with estrogenic activity are present in fish diets. In order to examine this possibility, we measured the estrogenic activity in commercial fish feed by in vitro yeast estrogen-screen (YES) assay based on the binding ability of tested compounds to estrogen receptors. Estrogenic activity was detected in all the commercial fish feed examined (0.2-6.2 ng estradiol equivalent/g fish feed), some phytoestrogens (genistein, formononetin, equol and coumestrol; relative activity to estradiol, 8.6 x 10(-6)-1.1 x 10(-4) by giving a value of 1.0 to estradiol) and some androgens (testosterone, 11-ketotestosterone and 5 alpha-dihydrotestosterone; relative activity to estradiol, 3.0 x 10(-6)-1.2 x 10(-4)). Therefore, it is possible that these compounds could affect the results of in vivo estrogen assay, such as vitellogenin production in male fish, especially when fish are fed commercial feed.