Human endometrium is infiltrated by natural killer (NK) cells throughout the menstrual cycle. The number of endometrial NK cells is low in the proliferative phase, but acutely increases after ovulation, and reaches a peak in the late secretory phase, suggesting that endometrium recruits these leukocytes selectively from circulating peripheral blood. We investigated the expression of macrophage inflammatory protein (MIP)-1beta, a potential chemoattractant for NK cells, in the endometrium. RT-PCR and ELISA revealed that MIP-1beta is expressed in the endometrium throughout the menstrual cycle at both the message and protein levels. MIP-1beta expression is stronger in the secretory phase endometrium than in the proliferative phase endometrium. Immunohistochemistry revealed that MIP-1beta is localized in the surface epithelial cells, glandular epithelial cells, and perivascular stromal cells throughout the menstrual cycle. Stromal cells in a wider perivascular area became immunoreactive in the secretory phase. There was a strong correlation between the endometrial MIP-1beta concentration and the number of endometrial NK cells. Progesterone significantly induced MIP-1beta secretion from cultured endometrial stromal cells, whereas 17beta-estradiol had a weak effect. These results suggest that endometrial MIP-1beta may be involved in the recruitment of NK cells from circulating peripheral blood.

No abstract

Aromatase has been shown to be expressed at a higher level in human breast cancer tissue than in normal breast tissue, by means of enzyme activity measurement, immunocytochemistry, and RT-PCR analysis. Cell culture including MCF-7 breast cancer cells, animal experiments using aromatasetransfected breast cancer cells, and transgenic mouse studies have demonstrated that estrogen production in situ plays a more important role than circulating estrogens in breast tumor promotion. In addition, tumor aromatase is believed to be able to stimulate breast cancer growth through both autocrine and paracrine pathways, as demonstrated by a three-dimensional cell culture study. RT-PCR and gene transcriptional studies have revealed that the aromatase promoter is switched from a glucocorticoid-stimulated promoter, I.4, in normal tissue to cAMP-stimulated promoters, I.3 and II, in cancerous tissue. Recently, we identified and characterized a cAMP-responsive element (CREaro) upstream from promoter I.3 by DNA deletion and mutational analyses. Our results from promoter functional analysis also demonstrated an interaction between the CREaro and the silencer element (S1) that was identified previously in our laboratory. In the presence of cAMP, the positive regulatory CREaro can overcome the action of the silencer on the function of promoter I.3. On the basis of results generated from our own and other laboratories, we propose that, in normal breast adipose stromal cells and fibroblasts, aromatase expression is driven by promoter I.4 (glucocorticoid dependent), and that the action of promoters I.3 and II is suppressed by the silencer negative regulatory element. However, in cancer cells and surrounding adipose stromal cells, the cAMP level increases, and aromatase promoters are switched to cAMP-dependent promoters -I.3 and II. Furthermore, we applied the yeast one-hybrid screening method to search for proteins interacting with the silencer element, S1. The major protein identified was ERRα-1; however, SF-1, which is present in the ovary, is not detected in breast cancer tissue. Using a reporter plasmid with the aromatase genomic fragment containing promoter I.3 and S1, in breast cancer SK-BR-3 cells, ERRα-1 was found to have a positive regulatory function. It is believed that the silencer element in the human aromatase gene may function Endocrine-Related Cancer (1999) 6 149-156 differently in different tissues, as a result of distinct expression patterns of transcription factors.

No abstract

Two isozymes of porcine aromatase, the placental and the blastocyst forms, were expressed in CHO cells using the mammalian cell transfection method. Using an`in-cell' assay (a 3 H-water release method), catalytic parameters of the porcine placental aromatase were found to be very similar to those of the human enzyme; however, the activity of the blastocyst isozyme was found to be one-thirtieth that of the placental isozyme. Product isolation assay (using testosterone as the substrate) revealed that the major steroid products were 17b-estradiol and 19-nortestosterone. The product ratio of estradiol/19-nortestosterone was found to be 94 : 6 for the porcine placental form, 6 : 94 for the porcine blastocyst form, and 92 : 8 for the human wild-type aromatase. Therefore, the porcine blastocyst aromatase isozyme catalyzes mainly androgen 19-desmethylation rather than aromatization. In addition, inhibition profile analyses on the placental and blastocyst isozymes were performed using three steroidal inhibitors [4-hydroxyandro-stenedione (4-OHA), 7a-(4 H -amino)phenylthio-1,4-androstandiene-3,17-dione (7a-APTADD), and bridge (2,19-methyleneoxy) androstene-3,17-dione (MDL 101,003)], and four nonsteroidal inhibitors [aminoglutethimide (AG), CGS 20267, ICI D1033, and vorozole (R83842)]. While the two isozymes of porcine aromatase share 93% amino-acid sequence identity, our results indicate that the two porcine aromatase isozymes have distinct responses to various aromatase inhibitors.