Expression of androgen receptor and 5alpha-reductases in the human normal endometrium and its disorders.

Abstract

Androgen metabolism and actions are considered to play a very important role in the development and progression of the normal human endometrium and its disorders. Details regarding androgen actions in these tissues, however, have not been well studied. We first immunolocalized the androgen receptor (AR) and 5alpha-reductases, which catalyze the conversion of testosterone to the bioactive and potent androgen, 5alpha-dihydrotestesterone (DHT), in 18 normal cycling human endometria, 27 endometrial hyperplasia and 46 endometrioid endometrial adenocarcinomas. We also examined the mRNA expression of AR and 5alpha-reductases in 7 cases of endometrioid endometrial adenocarcinomas using reverse transcription polymerase chain reaction (RT-PCR). In the normal human endometrium, AR was immunolocalized predominantly in stromal cells of the proliferative phase of the menstrual cycle and in epithelial cells of the secretory phase, whereas 5alpha-reductase types 1 and 2 immunoreactivities were detected in the cytoplasm of epithelial cells but not in stromal cells throughout all phases of the menstrual cycle. In endometrial hyperplasia, the median labeling index (LI) for AR was 48.1%, whereas positive immunostaining for 5alpha-reductase Type 1 and Type 2 was detected in only 1 case. In endometrial carcinoma, AR immunoreactivity was detected in the nuclei of carcinoma cells and the number of positive cases was 39/44 (88.6%). Median LI for AR was 36.1%. Immunoreactivity for 5alpha-reductase Type 1 and Type 2 was detected in the cytoplasm of carcinoma cells and the number of positive cases was 37/44 cases (84.1%) and 34/44 (77.3%) for 5alpha-reductase Types 1 and 2, respectively. A significant positive correlation was detected between 5alpha-reductase Type 1 and Type 2 immunoreactivity (p < 0.001). AR LI was not correlated with the presence or absence of Type 1 and Type 2 5alpha-reductases. Results from our RT-PCR studies were consistent with those of immunohistochemistry. These results suggest that DHT may play more important roles than testosterone in the regulation of androgen action in endometrial cancer and normal human endometrium, especially in the secretory phase, in which both AR and 5alpha-reductase are increased. Androgenic actions may be also regulated predominantly by serum testosterone and not by DHT in endometrial hyperplasia because of the absence of 5alpha-reductases in the site of its actions.