Expression of aromatase cytochrome P-450 in premenopausal and postmenopausal human ovaries: an immunocytochemical study.

The cellular distribution of the aromatase cytochrome P-450 enzyme in human ovaries has been investigated immunocytochemically, using an aromatase-specific monoclonal antibody. Ovaries of females ranging in age from prepubertal infant girl through to postmenopausal adulthood were obtained from immediate autopsy or after surgery. The results have revealed temporal and spatial changes in expression of aromatase at different stages of development. No immunoreactive aromatase was detected in the ovary of the 2.5 month infant. In premenopausal ovaries, aromatase was absent from the stromal compartment, but in follicles, a consistent pattern in expression of aromatase was observed, related to their size and developmental stage. Aromatase was not expressed in primordial, primary, or small secondary follicles less than 250 microns diameter. In slightly larger follicles (250-700 microns diameter) aromatase was first detected in a few thecal cells (TC). In more developed secondary through to large preovulatory follicles (greater than 1 cm) TC aromatase immunostain increased in intensity and number of positive cells, and the reaction was localized to a band of theca interna (TI) cells at the TI/theca externa interface. In granulosa cells (GC), aromatase was first detected in follicles in the initial stages of antrum formation (greater than 700 microns), and staining intensified as follicle diameter and antral cavity increased, being maximal in preovulatory follicles. GC aromatase was always found in the presence of TI immunostain. These two cell populations were separated by an unstained layer of TI cells giving the follicle walls a banded appearance. Immunostain was most intense in mural GC, was weaker in antral GC cells and was absent from the cumulus GC. Immunoreactive aromatase was also detected in functional corpora lutea (CL) but was absent from involuting CL's and corpora albicans. Our findings indicate that the immunostained cells of the CL are comprised of the former GC and possibly a subpopulation of former TI cells. In perimenopausal ovaries there was no evidence of any follicular or stromal aromatase immunostain. In postmenopausal ovaries no follicles were observed, but individual cells and clusters of cells in the stromal compartment of 3/7 specimens were found to have an aromatase immunostain reaction. In all cases, the aromatase immunostain reaction was cytoplasmic. The results provide the first direct evidence of the existence of TC aromatase, and of stromal cell aromatase in postmenopausal women.