1891

Previous studies have shown that neonatal exposure to Diethylstilbestrol (DES) or genistein induces uterine adenocarcinoma by 18 months of age in a rodent model; however, low dose exposure to DES has not been associated with endometrial pathology. In the present study, we tried to determine whether neonatal low-to-high-dose DES or an environmentally relevant high dose of Genistein might alter the sensitivity of the uterus to hormonal carcinogenesis induced by prolonged exposure to diethylstilbestrol in adulthood. Mice were divided into 2 groups and were treated with 0.01 μg/kg, 1 μg/kg, 1 mg/kg diethylstilbestrol, 50 mg/kg genistein or vehicle at days 1-5. First group were sacrificed on day 19 and the second group (4-6 months) were having the ovariectomy 7 days before sacrifice. Total RNA were extracted from uterus and were amplified by the FairPlay Microarray Labeling Kit following manufacturer’s protocol. Uterus cDNA labeled and cDNA amplified from Universal Mouse Reference RNA were labeled with Cy3 and Cy5 respectively. Florescent dye labeled cDNAs was hybridized on MWG Mouse oligo array. Use hierarchical clustering, a non-supervised approach as an exploring tool with various combinations of linkage methods and distance measuring algorithms, we discovered that the transcriptomes of various treatment and time point groups are very reproducible and consistent regardless which algorithm is using. As we predicted, the clustering results showing that the expression profiles of the DES 1 μg/kg group is similar to the genistein 50 mg/kg group. These two groups always clustered together under one cluster. Also, the low dose DES 0.01 μg/kg has a similar expression profiles as the control group was confirmed by the clustering analysis. Use Significant Analysis for Microarray (SAM), 83 genes was identified were differentially expressed and significantly changed between the day 19 and 4∼6 month. Those genes included transaldolase, calpain 5, erythroid ankyrin, cathepsin k and caspase-8. The preliminary statistical and bioinformatics analyses have already met our predictions of this study. With further data analysis and selected genes for confirmation, future findings should be able to help us to understand the mechanism of estrogen-like compounds on the development of endometrial cancer.

[Proc Amer Assoc Cancer Res, Volume 47, 2006]