Activity of three selective estrogen receptor modulators on hormone-dependent responses in the mouse uterus and mammary gland

Abstract

Selective estrogen receptor modulators (SERMs) have the unique potential to provide estrogenic effects in the skeletal and cardiovascular system, while minimizing/eliminating side effects on reproductive organs. However, despite the unifying characteristic of mixed estrogen receptor (ER) agonist/antagonist activity, compounds within this class are not interchangeable. In order to define and compare the effects of SERMs on different hormone-responsive tissues, we evaluated effects of bazedoxifene acetate (BZA), lasofoxifene (LAS) and raloxifene (RAL) in the mammary gland and uterus of the ovariectomized mouse. Endpoints measured included those regulated by estradiol alone (uterine wet weight, uterine G protein-coupled receptor 105 (GPR105) mRNA expression and mammary gland indoleamine-pyrrole 2,3 dioxygenase (INDO) mRNA expression) as well as others that required the combination of estradiol and progesterone (uterine serine protease inhibitor Kazal type 3 (Spink3) mRNA expression, mammary gland morphology and mammary gland defensin β1 (Defβ1) mRNA expression). The three SERMs tested had variable agonist and antagonist activity on these endpoints. In the uterus, the SERMs were mixed agonists/antagonists on estradiol-induced wet weight increase, whereas all three SERMs were estrogen receptor antagonists on GPR105 mRNA expression. However, in the presence of progesterone, BZA and RAL were agonists on Spink3 expression, while LAS was primarily an antagonist. In the mammary gland, BZA and RAL were predominantly agonists on the endpoint of mammary morphology and all three SERMs were clear agonists on Defβ1 mRNA expression, an E + P-dependent marker. Finally, LAS and RAL had mixed agonist/antagonist activity on INDO mRNA expression, while BZA had only antagonist activity. These results demonstrate that compounds with small structural differences can elicit distinct biological responses, and that in general, SERMs tended to behave more as antagonists on endpoints requiring estrogen alone and agonists on endpoints requiring the combination of estrogen and progesterone.

Introduction

Estradiol (E2) is a hormone that plays a key role in reproduction and in the homeostasis of a variety of mammalian systems. When estrogen levels decline at menopause, many women experience vasomotor instability and vaginal atrophy, and are at increased risk for developing osteoporosis and heart disease. Hormone therapy can address many of these consequences of reduced ovarian productivity, but not all women are candidates for these medications.

Classically, estrogens act via binding to intracellular estrogen receptors (ER), two of which have been described (ERα and ERβ). Upon binding hormone, these receptors modulate transcription of target genes. Known ER ligands encompass a large class of steroidal and non-steroidal compounds (Anstead et al., 1997), and within this group is a subset of ligands called SERMs that can mimic the effects of E2 in some situations and block its effects in others. The mechanisms responsible for this activity are poorly understood but known to be cell type and gene promoter dependent (Miller, 2002). The cell-type dependent action of this class of compounds is clinically useful. For example, tamoxifen, a triphenylethylene, was the first SERM that was originally identified for use in breast cancer due to strong antiestrogenic activity in the mammary gland but was later discovered to prevent bone mineral density loss (Jordan et al., 1987). Tamoxifen also has significant uterotropic activity, and its clinical use is associated with an increased incidence of uterine cancer (Bernstein et al., 1999, Fisher et al., 1998, van Leeuwen et al., 1994). In rats, lasofoxifene (LAS), a tetrahydronaphthalene, has been shown to prevent bone loss with no effect on the uterine epithelium (Ke et al., 2004, Ke et al., 1998, Ke et al., 2000), and in an N-methylurea-induced model of rat mammary carcinoma, LAS demonstrated chemopreventive and therapeutic potential (Cohen et al., 2001). Raloxifene (RAL), a benzothiophene-based SERM, also has antiestrogenic activity in the mammary gland and maintains bone mineral density in the postmenopausal population without introducing the uterotrophic liability seen with tamoxifen (Delmas et al., 1997). Preclinically, RAL caused regression of mammary gland development in intact rats and antagonized estrogen-stimulated mammary gland proliferation in ovariectomized (OVX) rats (Buelke-Sam et al., 1998). A more recently developed SERM is bazedoxifene acetate (BZA). Like other SERMs, BZA preserves bone mineral density and biomechanical parameters in a rodent model of osteopenia, has no effect on uterine epithelium and antagonizes E2-dependent MCF-7 breast cancer cell proliferation in vitro (Komm et al., 2005).

In mammals, the development of the mammary gland and uterus is dependent on various hormonal contributions. Estrogens and progestins are well-established modulators of reproductive function in the uterus and play significant roles during pregnancy and parturition. In the mammary gland, these hormones support the development of a branched ductal system and secretory lobuloalveoli during pregnancy. Specifically, estrogens facilitate ductal elongation into the fat pad, whereas progestins enable ductal side branching and morphogenesis of the lobuloalveoli (Silberstein, 2001).

We previously described development of a short-term in vivo mouse model to measure uterine and mammary gland responses to estrogens and progestins using both gross tissue responses as well as molecular markers (Crabtree et al., 2006). In this report, we describe characterization of three SERMs (BZA, LAS and RAL) in this model and examine their effect on endpoints dependent on the co-administration of E2 and progesterone (P) in addition to those requiring E2 alone.