Immunolocytochemical Localization of Estrogen Receptors a and b in the Human Reproductive Organs

Abstract & Commentary

Synopsis: This is the first study to determine the distribution of estrogen receptor subtypes in human reproductive tissues. The findings aid in understanding the actions of various estrogens and estrogen analogs.

Source: Pelletier G, El-Alfy M. J Clin Endocrinol Metab 2000;85:4835-4840.

The notion that we can synthesize or find a compound with estrogen action only in those tissues thought to benefit from estrogen exposure, and not in those tissues where its actions are unwanted or deleterious, has become the modern day holy grail of reproductive endocrinology. To date, there have been few reasons to suggest that this was not an achievable goal. However, the present study does much to reveal why the hope of finding an adequate selective estrogen receptor (ER) devoid of estrogen action in key tissues such as the breast is false. Pelletier and El-Alfy performed a key study to establish what types of ERs are expressed in which human reproductive tissues. Much of our previous information on tissue localization of receptor subtypes was garnered from investigations done in other animals, mostly the rat. As background, Pelletier and El-Alfy point out that the ER belongs to a receptor superfamily that also includes receptors for androgens, progestins, glucocorticoids, mineralocorticoids, thyroid hormone, retinoic acid, and vitamin D. There are two main subtypes of ERs, referred to as ERa and ERb. ERb is 95% homologous with ERa in the DNA-binding domain and 60% homologous in the ligand-binding domain. Using specific antibodies and human tissues obtained at the time of surgery, Pelletier and El-Alfy observed the following. In the ovary, ERb was found in granulosa cells of the growing follicle of all stages. Corpora lutea were unlabeled with either ER. ERa immunoreactivity was detected predominantly in theca interna cells. In the uterus, there was intense staining for ERa in luminal and glandular epithelial cells, the stroma, and muscle cells. There was only weak staining for ERb in epithelial cells. In the vagina, ERa staining was observed in the deep layers of the stratified epithelium, cells of the lamina propria, and smooth muscle cells. No ERb immunoreactivity could be detected. In breast tissue, ERa immunoreactivity was present in epithelial cells of intralobular acini and interlobular ducts and ERb showed a similar distribution.

COMMENT by Sarah L. Berga, MD

The concept of a selective ER modulator that could surpass the risk: benefit profile of endogenous or exogenous estrogens sounded attractive and seemed theoretically feasible. Sadly, the premise is based on the notion that there would be enough differential ER tissue localization that one, therefore, could find an agent devoid of estrogen action in the breast and endometrium with preservation of estrogen action in the brain, bone, bladder, cardiovascular tree, pelvic floor, and vagina. As the present study reveals, the tissue expression of ERa and ERb is not sufficiently distinct to make this likely. In particular, mammary tissue contains about the same amount of ERa and ERb. We want estrogen action in the pelvic floor and vagina to maintain support and sexual function, which means we want a ligand that binds to ERa, but that ligand would also have unwanted estrogen action in the breast. We want estrogen action in the brain. In particular, the cortical centers mediating memory are enriched with ERb. That means we want a ligand that binds to both ERa and ERb. Thus, we are stuck because there is not enough tissue specificity of ER subtype distribution to avoid estrogen action in the breast and endometrium. If we cannot achieve tissue specificity with ligands that preferentially bind to a particular ER subtype, does this mean that there is no hope? Certainly not. But it does mean that it will be much more difficult to get rid of estrogen action only in the breast and endometrium. There may be enough tissue specificity conferred by the various proteins that modulate ER translocation to the nucleus for binding to the DNA at estrogen response elements (EREs), from proteins that modulate the configuration of the DNA and therefore the availability of the ERE in the nucleus, or from the cytosolic chaperones that mediate ligand binding to ER. However, since these proteins, often referred to as co-regulators and co-repressors, are not well understood, we cannot be sure that this new hope for tissue specificity will stand the test of further investigation. Given that there are many more tissues that clearly benefit from estrogen exposure, it seems that it would be easier to develop a compound to subtract estrogen’s action from those tissues in which its actions are unwanted (endometrium) or possibly not of benefit (breast). Until we develop this type of pharmacological "scalpel," it seems that we will have to rely on surgical modalities to remove these tissues when deemed desirable to do so.