Temporal expression patterns of individual genes in the uteri of immature, ovariectomized mice treated with ethynyl estradiol.

Fertuck KC, Zacharewski TR.

Department of Biochemistry & Molecular Biology, National Food Safety & Toxicology Center, and Institute for Environmental Toxicology, Michigan State University, East Lansing, MI.

Estrogens are classically defined as compounds that are capable of inducing a trophic effect on the uterus. This physiological change has been linked to alterations in the expression of estrogen-responsive genes, but the cascade of molecular events remains unknown. In the present study, the effects of ethynyl estradiol (EE), a synthetic estrogen, on global gene expression in the uterus were examined in immature, ovariectomized C57BL/6 mice. Mice were gavaged with 0.1mg/kg EE or vehicle, and uteri harvested at 0, 2, 8, 12, or 24 hr. An additional group received vehicle or EE daily for three consecutive days, followed by sacrifice 24hr following the final dose. Overall, while no differences in uterine weight were found between treatment groups (p=0.296) from 0 to 24 hrs, changes in global gene expression levels were observed during this period. Mu11KsubA Affymetrix GeneChips were used to assess relative expression levels of approximately 6500 genes in uterine tissue of two animals from each treatment and time-matched vehicle treated (control) group. Responses were compared using either (1) the raw values obtained from MicroArray Suite (MAS) 5.0, or after normalization by either (2) the PM-MM or (3) the PM-only model in the dChip 1.1 analysis program. In the 248 cases where an accession number was represented by two different probe sets, the PM-only model, and to a lesser extent the PM-MM model, was able to consistently provide similar results between each pair of probe sets, particularly when one probe set was expressed at an overall higher level than the other. However, the PM-only model was the least likely to provide results that were in agreement with published results for genes that are known to be induced by estrogen to only a moderate degree. The most complete view was obtained when the PM-only and PM-MM results were viewed concurrently; both models appeared to provide improvement over raw MAS output.