Location: Arkansas Children's Nutrition CenterTitle: Feeding soy protein isolate and treatment with estradiol have different effects on mammary gland morphology and gene expression in weanling male and female rats) Author
Submitted to: Physiological Genomics
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/16/2013
Publication Date: 12/1/2013
Citation: Miousse, I., Sharma, N., Blackburn, M., Vantrease, J., Gomez-Acevedo, H., Hennings, L., Shankar, K., Cleves, M., Badger, T.M., Ronis, M.J. 2013. Feeding soy protein isolate and treatment with estradiol have different effects on mammary gland morphology and gene expression in weanling male and female rats. Physiological Genomics. 45(22):1072-1083. Interpretive Summary: In recent years, concerns have emerged that a family of compounds naturally present in soy, isoflavones, may have a feminizing effect in children due to its structural similarity to estrogen. To test this hypothesis, we fed newly weaned male and female rats a diet where the protein source was soy protein isolate, the same compound used in soy infant formula. We compared these with animals fed a diet composed of milk protein, as a negative control, and with animals fed a milk protein diet supplemented with estrogen, as a positive control. At puberty, we studied an estrogen-sensitive tissue: the mammary gland. Our results showed that the diet supplemented with estrogen caused an increase in the number of a structure called "terminal end bud" in male rats, but not in females. Such increases have been associated with increased breast cancer risk in previous studies. However, no such changes were observed in either male or female rats fed soy protein isolate. The soy protein isolate diet also did not increase markers of growth, contrarily to the estrogen-supplemented diet. We further looked at the expression of all genes in the mammary gland of these animals by microarray. There were over 15 times fewer genes in which expression was increased or decreased at least two-folds in the soy protein isolate diet than in the estrogen-supplemented diet. In males, there was a small overlap between genes in which expression was changed by estrogen and soy protein isolate. Some of these genes were associated with a decrease in the development of fat cells. We also found that serum insulin was decreased by soy protein isolate in males. In conclusion, we found that young male rats were more sensitive to estrogen and soy protein isolate in the diet than young female rats. Soy protein isolate had a very small effect on the mammary gland compared to estrogen. In young male animals, diets containing estrogen and soy protein isolate shared some similarities in their ability to repress the expression of a small subset of genes, notably genes that stimulate the development of fat cells.
Technical Abstract: Isoflavones are phytochemical components of soy diets that bind weakly to estrogen receptors (ERs). To study potential estrogen-like actions of soy in the mammary gland during early development, we fed weanling male and female Sprague-Dawley rats a semi-purified diet with casein as the sole protein source from PND21 to PND33, the casein diet supplemented with estradiol (E2) at 10 ug/kg/day or the same diet substituting soy protein isolate (SPI) for casein. In contrast to E2, the SPI diet induced no significant change in mammary morphology. In males, there were 34 genes for which expression was changed, =2-fold in the SPI group versus 509 changed significantly by E2, and 8 versus 174 genes in females. Nearly half of SPI-responsive genes in males were also E2-responsive, including adipogenic genes. Serum insulin was found to be decreased by the SPI diet in males. SPI and E2 both down-regulated the expression of ER (Esr1) in males and females, and ER beta (Esr2) only in males. Chromatin immunoprecipitation (ChIP) revealed an increased binding of ER alpha to the promoter of the progesterone receptor (Pgr) and Esr1 in both SPI and E2 treated males compared to the casein group but differential recruitment of ER. ER promoter binding did not correlate with differences in Pgr mRNA expression. This suggests that SPI fails to recruit appropriate co-activators at E2-inducible genes. Our results are unsupportive of an E2-like proliferative effect of SPI feeding and indicate that SPI behaves like a SERM rather than a weak estrogen in the developing mammary gland.