From in vivo gene targeting of oestrogen receptors to optimization of their modulation in menopause

Abstract

The ancestral status of oestrogen receptor (ER) in the family of the steroid receptors has probably contributed to the pleiotropic actions of oestrogens, and in particular, that of 17β-oestradiol (E2). Indeed, in addition to their well-described role in sexual development and reproduction, they influence most of the physiological processes. The pathophysiological counterpart of these actions includes prevention of osteoporosis, atheroma and type 2 diabetes, and also the promotion of uterus and breast cancer growth. Thus, the major challenge consists in uncoupling some beneficial actions from other deleterious ones, that is, selective ER modulation. Tamoxifen and raloxifene are already used, as they prevent the recurrence of breast cancer and mimic oestrogen action mainly on bone. Both E2 and tamoxifen exhibit a proliferative and, thus, a protumoural action on the endometrium. Activation of ERα and ERβ regulates target gene transcription (genomic action) through two independent activation functions, AF-1 and AF-2, but can also elicit rapid membrane-initiated steroid signals. In the present review, we attempted to summarize recent advances provided by the in vivo molecular 'dissection' of ERα, allowing the uncoupling of some of its actions and potentially paving the way to optimized selective ER modulators.

Figure 1

Schematic representation of the mouse ERα gene structure and the strategies of Esr1 gene inactivation. (A) The Esr1 gene encompasses eight coding exons and at least six non-coding 5′ exons. The full length 66 kDa ERα protein is composed of six domains (A to F), comprising a DNA-binding domain (DBD), a ligand-binding domain (LBD) and two activation functions (AF-1 and AF-2). The translation of the physiologically expressed 46 kDa isoform (not shown) is initiated at AUG2. This isoform lacks the entire A and B domains and hence, AF-1. (B) The first strategy of ERα gene targeting consisted in inserting a neomycin cassette in the first exon of the ERα gene (referred to as αER-NeoKO) (Lubahn et al., 1993). The αER-NeoKO expresses at least two truncated ERα proteins, due to natural and non-natural splicing events, devoid of AF-1 function but with a functional AF-2. The splicing involving the neomycin cassette generates a chimeric 55 kDa isoform (Kos et al., 2002; Pendaries et al., 2002). The level of expression of these truncated isoforms is enough to mediate several actions of E2 in the vessel wall. (C) The second knockout approach (referred to as ERα^−/− mice) consisted in introducing LoxP sites and then excising the second coding exon of Esr1 gene coding for parts of the DBD (Dupont et al., 2000). This strain does not allow the expression of any functional ERα. Accordingly, the vascular effects of E2 in that persisted αER-NeoKO were abolished in ERα^-/-, demonstrating that ERα appears to mediate most of the vascular effects of E2 (Pare et al., 2002; Pendaries et al., 2002). (D) ERαAF-1° mice consist in targeting the first exon of ERα gene coding for the A and B domains and thereby AF-1 (deletion corresponding to amino acids 2–148) (Billon-Gales et al., 2009). The phenotype of ERαAF-1° mice is reminiscent to that αER-NeoKO mice, although the leakage and, thereby, the expression of chimeric 55 kDa isoform is highly variable (Kos et al., 2002). (E) ERαAF-2° mice consist in deleting seven amino acids in the helix 12 and, thereby, AF-2 (deletion corresponding to amino acids 543–549) (ms submitted to publication).

Figure 2

Structure of oestradiol, full antagonist, ER-specific agonists and SERMs.