Differential effects of xenoestrogens on coactivator recruitment by estrogen receptor (ER) alpha and ERbeta

It has been proposed that tissue-specific estrogenic and/or antiestrogenic actions of certain xenoestrogens may be associated with alterations in the tertiary structure of estrogen receptor (ER) alpha and/or ERbeta following ligand binding; changes which are sensed by cellular factors (coactivators) required for normal gene expression. However, it is still unclear whether xenoestrogens affect the normal behavior of ERalpha and/or ERbeta subsequent to receptor binding. In view of the wide range of structural forms now recognized to mimic the actions of the natural estrogens, we have assessed the ability of ERalpha and ERbeta to recruit TIF2 and SRC-1a in the presence of 17beta-estradiol, genistein, diethylstilbestrol, 4-tert-octylphenol, 2',3',4', 5'-tetrachlorobiphenyl-ol, and bisphenol A. We show that ligand-dependent differences exist in the ability of ERalpha and ERbeta to bind coactivator proteins in vitro, despite the similarity in binding affinity of the various ligands for both ER subtypes. The enhanced ability of ERbeta (over ERalpha) to recruit coactivators in the presence of xenoestrogens was consistent with a greater ability of ERbeta to potentiate reporter gene activity in transiently transfected HeLa cells expressing SRC-1e and TIF2. We conclude that ligand-dependent differences in the ability of ERalpha and ERbeta to recruit coactivator proteins may contribute to the complex tissue-dependent agonistic/antagonistic responses observed with certain xenoestrogens.     

Colocalization and ligand-dependent discrete distribution of the estrogen receptor (ER)alpha and ERbeta

To investigate the relationships between the loci expressing functions of estrogen receptor (ER)alpha and that of ERbeta, we analyzed the subnuclear distribution of ERalpha and ERbeta in response to ligand in single living cells using fusion proteins labeled with different spectral variants of green fluorescent protein. Upon activation with ligand treatment, fluorescent protein-tagged (FP)-ERbeta redistributed from a diffuse to discrete pattern within the nucleus, showing a similar time course as FP-ERalpha, and colocalized with FP-ERalpha in the same discrete cluster. Analysis using deletion mutants of ERalpha suggested that the ligand-dependent redistribution of ERalpha might occur through a large part of the receptor including at least the latter part of activation function (AF)-1, the DNA binding domain, nuclear matrix binding domain, and AF-2/ligand binding domain. In addition, a single AF-1 region within ERalpha homodimer, or a single DNA binding domain as well as AF-1 region within the ERalpha/ERbeta heterodimer, could be sufficient for the cluster formation. More than half of the discrete clusters of FP-ERalpha and FP-ERbeta were colocalized with hyperacetylated histone H4 and a component of the chromatin remodeling complex, Brg-1, indicating that ERs clusters might be involved in structural changes of chromatin.     

Selectivity of antibodies to estrogen receptors alpha and beta (ERalpha and ERbeta) for detecting DNA-bound ERalpha and ERbeta in vitro

Antibodies are widely used to detect estrogen receptor (ER) in ER-DNA complexes in electrophoretic mobility shift assays (EMSA). We compared the specificity of antibodies raised to different regions of ERalpha or ERbeta for detecting recombinant human ERalpha (rhERalpha) and recombinant rat ERbeta (rrERbeta) when bound to a consensus estrogen response element (ERE). ERalpha-specific antibodies specifically slowed the migration of the ER-ERE complex by 32 to 84% and inhibited rhERalpha-ERE binding by 17 to 75%. None of antibodies to ERbeta supershifted rhERalpha-ERE complex. Some ERalpha-specific antibodies increased whereas some decreased rrERbeta-ERE binding. Anti-ERbeta antibodies supershifted different amounts of the rrERbeta-ERE complex. Our results indicate that supershift and inhibition of ER-ERE interaction with a specific antibody are equally reliable in the detection of rhERalpha and rrERbeta. ERalpha antibody Ab10, antisera G20 and AT3B, and ERbeta-antiserum Y19 offered the best discrimination between ERalpha and ERbeta. Comparison of the peptide sequences against which various antibodies were raised indicate directions for new ERalpha and ERbeta- specific antibody development. We conclude that a cognate ER antibody that retards the migration of the ER-ERE complex by at least 40% or inhibits ER-ERE interaction by at least 8% provides a reliable detection of a specific ER isoform in EMSA.     

Interaction of tetrahydrocrysene ketone with estrogen receptors alpha and beta indicates conformational differences in the receptor subtypes

Estrogen receptors (ER) alpha and beta bind estradiol (E2) and other estrogenic ligands with different affinities. To measure the rate of E2 association with ERa and ERbeta, we employed tetrahydrocrysene ketone (THCK), a fluorescent ligand that is an agonist with ERalpha and an antagonist with ERbeta. We report that THCK binds E2-liganded and unliganded ERalpha and ERbeta, indicating a THCK binding site(s) other than the E2 binding pocket. THCK fluorescence was greater for ligand-occupied ERbeta than ERalpha, suggesting differences in the microenvironment of the THCK binding site(s). THCK fluorescence was also significantly greater for E2-, 4-hydroxytamoxifen-, and tamoxifen aziridine-liganded versus unliganded ER, allowing calculations of E2 association rate constants (ka) of 7.60 +/- 0.75 and 5.12 +/- 0.30 x 10(5) M(-1) s(-1) for E2-ERalpha and E2-ERbeta, respectively. THCK did not affect ERalpha binding to estrogen response element (ERE) DNA, but decreased ERbeta-ERE binding. We conclude that THCK binding site(s) on ERalpha versus ERbeta are different and important for ER function.     

Activities of a non-classical estrogen,Z-bis-dehydrodoisynolic acid,with ERalpha and ERbeta

(+/-)-Z-bis-Dehydrodoisynolic acid [(+/-)-Z-BDDA] is highly estrogenic in vivo, yet binds to estrogen receptor (ER) poorly. This paradox has raised the possibility of alternative ERs and/or molecular mechanisms. To address the possibility of high activities of Z-BDDA with ERbeta, we determined the activities of (+)-Z-BDDA and (-)-Z-BDDA, in cell culture and in vitro, comparing ERbeta to ERalpha. Transfectional analysis in Hela cells showed (-)-Z-BDDA is an agonist for gene activation with both ERalpha (EC(50) congruent with 0.3nM) and ERbeta (EC(50) congruent with 5nM), while little to no activity was observed with (+)-Z-BDDA. Similarly, in gene repression assays, (-)-Z-BDDA was active (EC(50) congruent with 0.2nM), but again minimal activity was exhibited by (+)-Z-BDDA. Binding to ERalpha and ERbeta in vitro used both competition and a direct binding assay. For ERalpha, the relative affinity of (-)-Z-BDDA was approximately 6% by competition and 1.7% by direct binding versus 17beta-estradiol (E2; 100%), while (+)-Z-BDDA also demonstrated binding, but with relative affinities of only 0.08% by competition and 0.3% by the direct assay. For ERbeta, the affinity of (-)-Z-BDDA was approximately 7% by competition and 1.5% by the direct assay relative to E2 (100%), while (+)-Z-BDDA had lower affinity, approximately 0.2% that of E2 by both assays.The paradox of potent in vivo activity but lower activity in receptor binding and in cell culture reporter gene assays, previously seen with ERalpha is now also associated with ERbeta. The failure of ERbeta to explain the activity-binding paradox indicates the need for additional in vivo metabolic and pharmacokinetic studies and continued consideration of alternative mechanisms.     

Dissecting physiological roles of estrogen receptor alpha and beta with potent selective ligands from structure-based design

The distinct roles of the two estrogen receptor (ER) isotypes, ERalpha and ERbeta, in mediating the physiological responses to estrogens are not completely understood. Although knockout animal experiments have been aiding to gain insight into estrogen signaling, additional information on the function of ERalpha and ERbeta will be provided by the application of isotype-selective ER agonists. Based on the crystal structure of the ERalpha ligand binding domain and a homology model of the ERbeta-ligand binding domain, we have designed steroidal ligands that exploit the differences in size and flexibility of the two ligand binding cavities. Compounds predicted to bind preferentially to either ERalpha or ERbeta were synthesized and tested in vitro using radio-ligand competition and transactivation assays. This approach directly led to highly ER isotype-selective (approximately 200-fold) and potent ligands. To unravel physiological roles of the two receptors, in vivo experiments with rats were conducted using the ERalpha- and ERbeta-selective agonists in comparison to 17beta-estradiol. The ERalpha agonist induced uterine growth, caused bone-protective effects, reduced LH and FSH plasma levels, and increased angiotensin I, whereas the ERbeta agonist did not at all or only at high doses lead to such effects, despite high plasma levels. It can thus be concluded that estrogen effects on the uterus, pituitary, bone, and liver are primarily mediated via ERalpha. Simultaneous administration of the ERalpha and ERbeta ligand did not lead to an attenuation of ERalpha-mediated effects on the uterus, pituitary, and liver parameters.     

In vitro fluorescence anisotropy analysis of the interaction of full-length SRC1a with estrogen receptors alpha and beta supports an active displacement model for coregulator utilization

Binding of full-length P160 coactivators to hormone response element-steroid receptor complexes has been difficult to investigate in vitro. Here, we report a new application of our recently described fluorescence anisotropy microplate assay to investigate binding and dissociation of full-length steroid receptor coactivator-1a (SRC1a) from full-length estrogen receptor alpha (ERalpha) or estrogen receptor beta (ERbeta) bound to a fluorescein-labeled (fl) estrogen response element (ERE). SRC1a exhibited slightly higher affinity binding to flERE.ERbeta than to flERE.ERalpha. Binding of SRC1a to flERE.ERalpha and to flERE.ERbeta was 17beta-estradiol (E2)-dependent and was nearly absent when ICI 182,780, raloxifene, or 4-hydroxytamoxifen were bound to the ERs. SRC1a binds to flERE.E2-ERalpha and flERE.E2-ERbeta complexes with a t1/2 of 15-20 s. Short LXXLL-containing nuclear receptor (NR) box peptides from P160 coactivators competed much better for SRC1a binding to flERE.E2-ER than an NR box peptide from TRAP220. However, approximately 40-250-fold molar excess of the P160 NR box peptides was required to inhibit SRC1a binding by 50%. This suggests that whereas the NR box region is a primary site of interaction between SRC1a and ERE.E2-ER, additional contacts between the coactivator and the ligand-receptor-DNA complex make substantial contributions to overall affinity. Increasing amounts of NR box peptides greatly enhanced the rate of dissociation of SRC1a from preformed flERE.E2-ER complexes. The data support a model in which coactivator exchange is facilitated by active displacement and is not simply the result of passive dissociation and replacement. It also shows that an isolated coactivator exhibits an inherent capacity for rapid exchange.