Mammary gland morphology and gene expression differ in female rats treated with 17β-estradiol or fed soy protein isolate.

Soy foods have been suggested to have both positive health benefits and potentially adverse effects as a result of their content of phytoestrogens. However, studies on the estrogenicity of soy foods are lacking. Here we directly compared the effects of soy protein isolate (SPI), the protein in soy infant formula, with those of 17β-estradiol (E2), on global gene expression profiles and morphology in the female rat mammary gland. Rats were fed AIN-93G diets containing casein or SPI beginning on postnatal d 30. Rats were ovariectomized on postnatal d 50 and treated with 5 μg/kg/d E2 or vehicle for 14 d. Microarray analysis revealed that E2 treatment altered expression of 780 genes more than or equal to 2-fold (P < 0.05), whereas SPI feeding altered expression of only 53 genes more than or equal to 2-fold. Moreover, the groups had only 10 genes in common to increase more than or equal to 2-fold. The combination of SPI feeding and E2 altered expression of 422 genes and reversed E2 effects on many mRNAs, including those involved in the c-myc signaling pathway, cyclin D1, and Ki67. ERα binding to its response element on the Tie-2/Tek and progesterone receptor promoters was increased by E2, but not SPI, and this promoter binding was suppressed by the combination of E2 + SPI for the Tie-2/Tek promoter but increased for the progesterone receptor promoter (P < 0.05). SPI reduced the ratio of epithelial to fat pad area and E2 + SPI reduced both epithelial and fat pad area (P < 0.05). These data suggest that SPI is only minimally estrogenic in the rat mammary gland even in the absence of endogenous estrogens.