Certain types of estrogenic compounds have been shown to have tissue-specific actions. In addition, some tissues may exhibit differential gene regulation by agonists and antagonists. Our previous studies using structurally modified estrogenic molecules had indicated differential effects on specific estrogen responses, indicating that the activity of the estrogen receptor protein can be altered depending not only upon the structure of the bound ligand but also the regulated gene itself.
The mechanism of differential induction, however, was not determined, and might involve altered binding to the estrogen response element (ERE), altered transcription, or post-transcriptional modification of gene products. Our previous studies indicated that differential induction by modified diethylstilbestrol (DES) agonists could not be accounted for by differences in ligand affinity for the estrogen receptor (ER) or differential binding of the ER to a consensus vitellogenin A2 (vit A2) ERE. To determine if this differential hormonal responsiveness was reflected at the level of transcription, we analyzed mouse uterine mRNA of several estrogen-responsive genes, including glucose-6-phosphate dehydrogenase (G6PD), ornithine decarboxylase (ODC) and lactoferrin, by Northern blot following injection with the modified agonists DES, indenestrol A (IA), indenestrol B (IB) and Z-pseudo DES (ZPD). All compounds induced the G6PD message, although IB and ZPD induced expression only transiently, while DES and IA maintained the message for 24 h. No difference in induction was seen for ODC message, which was induced equally by all the compounds. In contrast, lactoferrin, a highly estrogen-responsive gene, was induced only by DES and IA and not by the other agonists IB or ZPD, showing that the lactoferrin gene was differentially regulated by these compounds. To determine whether this difference was due to altered transcriptional activity, the mouse lactoferrin estrogen-responsive module (mERM) linked to a chloramphenicol acetyl transferase (CAT) reporter gene was tested in transfected cells. Using the mouse estrogen receptor in RL95 cells, DES and IA induced expression of CAT, but IB did not, confirming the differential response seen in vivo. To show whether this difference in transcription occurred because of altered binding to the lactoferrin ERE, which is not a perfect consensus ERE, a gel shift assay was used to examine DNA binding of ER bound to the agonists. All ligands produced equivalent binding to the lactoferrin ERE suggesting that differential regulation was not a result of altered DNA binding. Taken together, these observations indicate that the differential induction of lactoferrin by these compounds occurs via altered activation of the transcriptional components unique to lactoferrin and is likely to involve altered interaction with co-activators. Surprisingly, unlike the mouse ER, the human estrogen receptor activated and induced expression of lactoferrin estrogen-responsive module-CAT with all the compounds. Mouse ER is also known to vary from the human ER in its activity with the triphenylethylene estrogen tamoxifen, which has agonist activity with the mouse ER but mixed antagonist/agonist activity with the human ER. The data show that human and mouse estrogen receptors are activated differently by this group of stilbestrol estrogen ligands when assayed on the lactoferrin response element, which is the first description of this type of gene and species specific difference. Lactoferrin gene regulation by estrogen receptor can be used as a model to study the mechanism of differential gene activation by different estrogen agonists and antagonists using a more physiological situation than commonly used with in vitro gene reporter systems.