Production of activin A in hyperplasia and adenocarcinoma of the human endometrium

Objectives: To examine the possible localization and production of activin A in human normal endometrium, endometrial hyperplasia, and adenocarcinoma tissues.

Methods: Human endometrial tissues were collected from 45 patients who were undergoing abdominal hysterectomy. Tissue sections were stained with monoclonal antibodies against the inhibin/activin alpha- and beta A-subunits and activin A using an avidin-biotin-peroxidase complex technique. Concentrations of activin A and inhibin A in tissue extracts of the endometrial tissues were measured using enzyme-linked immunosorbent assays (ELISAs). The expressions of the inhibin alpha-subunit and activin beta A-subunit messenger RNA (mRNA) in the endometrial tissues were demonstrated by reverse transcription-polymerase chain reaction (RT-PCR) analysis.

Results: No immunostaining with an antibody against the inhibin alpha-subunit was observed in human normal endometrium, endometrial hyperplasia, and adenocarcinoma. By contrast, immunostaining for the activin beta A-subunit and activin A was observed in the cytoplasm of glandular cells in normal endometrium, endometrial hyperplasia, and tumor cells of endometrial adenocarcinoma. The percentages of stained cells in endometrial adenocarcinoma were higher than those in normal endometrium. Also, the percentages of stained tumor cells with poor differentiation were higher than those with good and moderate differentiation of the endometrium. The stromal cells in normal endometrium, endometrial hyperplasia, and adenocarcinoma were weakly immunoreactive with antibodies against the beta A-subunit and activin A. Immunoreactivity of activin A in tissue extracts from normal endometrium and endometrial adenocarcinoma was detected by the two-site ELISA. Immunoreactivity of activin A was significantly higher in adenocarcinoma than in normal endometrium. On the other hand, the immunoreactive inhibin A was not detected. The expression of the alpha-subunit mRNA in endometrial tissues was demonstrated as the RT-PCR products migrated at 905 bp and the PCR products of the beta A-subunit showed a band at 366 bp.

Conclusions: It is suggested that activin A, but not inhibins, are produced by endometrial tissues. The amounts of produced activin A were higher in adenocarcinoma tissues than in normal endometrium. Activin A might be involved in human endometrial tumorigenesis.