Prediction of ligand binding affinity using a multiple-conformations-multiple-protonation scheme: application to estrogen receptor α.

A fast method that can predict the binding affinities of chemicals to a target protein with a high degree of accuracy will be very useful in drug design and regulatory science. We have been developing a scoring function for affinity prediction, which can be applied to extensive protein systems, and also trying to generate a prediction scheme that specializes in each target protein, with as high a predictive power as possible. In this study, we have constructed a prediction scheme with target-specific scores for estimating ligand-binding affinities to human estrogen receptor α (ERα), considering the major conformational change between agonist- and antagonist-bound forms and the change in protonation states of histidine at the ligand-binding site. The generated scheme calibrated with fewer training compounds (23 for the agonist-bound form, 17 for the antagonist-bound form) demonstrated good predictive power (a predictive r(2) of 0.83 for 154 validation compounds); this was also true for compounds with frameworks that were quite different from those of the training compounds. Our prediction scheme will be useful in drug development targeting ERα and in primary screening of endocrine disruptors, and provides a successful method of affinity prediction considering the major conformational changes in a protein.

Binding characteristics of estrogen receptor (ER) in Atlantic croaker (Micropogonias undulatus) testis: different affinity for estrogens and xenobiotics from that of hepatic ER.

An estrogen receptor (ER) was identified in cytosolic and nuclear fractions of the testis in a marine teleost, Atlantic croaker (Micropogonias undulatus). A single class of high affinity, low capacity, and displaceable binding sites was identified by saturation analysis, with a Kd of 0.40 nM in cytosolic extracts and a Kd of 0.33 nM in nuclear extracts. Competition studies demonstrated that the receptor was highly specific for estrogens (diethylstilbestrol > estradiol >> estriol = estrone) and also bound several antiestrogens. Testosterone and 5alpha-dihydrotestosterone had much lower affinities for the receptor, whereas no displacement of specific binding occurred with 11-ketotestosterone or any of the C21 maturation-inducing steroids. A variety of xenoestrogens, including o,p'-dichlorodiphenyltrichloroethane (DDT), chlordecone (Kepone), nonylphenol, hydroxylated polychlorinated biphenyls (PCBs), and the mycotoxin zearalenone, bound to the receptor with relatively low binding affinities, 10(-3) to 10(-5) that of estradiol. A comparison of the binding affinities of various ligands for the testicular ER and the hepatic ER in this species revealed that the testicular ER was saturated at a lower [3H]estradiol concentration (1 nM vs. 4 nM). The binding affinities of several compounds, including testosterone and nafoxidine, exhibited marked differences for the two ERs; and most of the estrogens and xenoestrogens tested had higher binding affinities for the testicular receptor. Minor amounts of estradiol (0.12 ng/g tissue/h) were produced by testicular tissue fragments incubated in vitro, and estradiol was detected in male Atlantic croaker plasma. The identification of a testicular ER and evidence that estradiol is produced by the testes in croaker suggest that estrogens participate in the hormonal control of testicular function in teleosts.

Evaluation of clinical and environmental anti-estrogens with human estrogen receptor expressed in Saccharomyces cerevisiae: a novel role for ABC-cassette transporters in mediating anti-estrogenic activity.

The effectiveness of anti-estrogens in treating estrogen-dependent diseases is limited by the acquired resistance of some diseases to anti-estrogens. This effect could occur by the export of anti-estrogens by cell membrane transport proteins. To study this phenomenon we have expressed human estrogen receptor (hER) and an estrogen-sensitive reporter in wild-type yeast and two transport-defective strains. In the wild-type strain, the most effective anti-estrogen was nafoxidine. 4-Hydroxy tamoxifen and clomiphene were inactive whereas tamoxifen had significant inhibitory activity in the wild-type strain. Using a strain missing the ABC-cassette transporter Snq2, clomiphene had anti-estrogenic activity. 4-Hydroxy tamoxifen had anti-estrogenic activity only in yeast lacking the transporter Pdr5. Whole cell binding assays indicated that 4-hydroxy tamoxifen is exported by Pdr5. Environmental chemicals such as polychlorinated biphenyls function as partial estrogens and anti-estrogens in yeast. In the absence of Pdr5 or Snq2, the estrogenic activity of 4-hydroxy, 2',4',6'-trichloro biphenyl (3-PCB) was substantially reduced in comparison to its activity in the wild-type strain. Interestingly, the antiestrogenic activity of 3-PCB was equivalent in the wild-type and transporter-defective strains. Our results suggest a novel role for ABC-cassette transporters in regulating the activity of clinical and environmental anti-estrogens.

FLP recombinase/estrogen receptor fusion proteins require the receptor D domain for responsiveness to antagonists, but not agonists.

The ligand-binding domains of steroid receptors convey ligand-dependent regulation to certain proteins to which they are fused. Here we characterize fusion proteins between a site-specific recombinase, FLP, and steroid receptor ligand-binding domains. These proteins convert ligand binding into DNA recombination. Thus, ligand binding is directly coupled to an enzyme activity that is easily measured by DNA rearrangements or heritable genetic changes in marker gene expression, as opposed to the multiple events leading to transcription. Recombination by a FLP-estrogen receptor (FLP-EBD) fusion is activated by all tested estrogens, whether agonists or antagonists, indicating that all induce EBD release from the 90-kDa heat shock protein complex. Altering the distance between FLP and the EBD domain in the fusion proteins, by reducing the included length of the estrogen receptor D domain, affects ligand efficacy. A FLP-EBD with no D domain shows reduced inducibility by agonists and, unexpectedly, complete insensitivity to induction by all antagonists tested. A FLP-EBD including some D domain shows a ligand-inducible phenotype intermediate to those displayed by FLP-EBDs containing all or none of the D domain. Thus, we observed a tethered interference between FLP and the EBD domains that differs depending on the distance between the two domains, the conformations induced by agonists or antagonists, and which presents a previously undetectable distinction between estrogen agonists and antagonists in yeast.

Characterization of estrogen and antiestrogen binding to the cytosol and microsomes of breast tumors.

The binding of [3H]estradiol and [3H]hydroxytamoxifen to the cytosol and microsomal fractions of several human breast tumors was investigated. By washing microsomal membranes with a KCl-free or a KCl-containing medium we could distinguish between intrinsic, extrinsic and contaminant estradiol binding sites in these membranes. We observed that treatment of the microsomes with low salt medium removes about 80% of the total estradiol binding sites, whereas 20% are not extractable. The concentration of unextractable [3H]estradiol binding sites in the microsomes varies in proportion to the level of cytosolic estrogen receptors (ER). About 10% of the total extranuclear specific estrogen binding sites was consistently found tightly associated to the microsomal fraction, which displays an affinity for estradiol (Kd = 0.1-0.6 nM) similar to that of the cytosolic ER. The displacement of [3H]estradiol with unlabeled hormone or with the antiestrogens, nafoxidine, enclomiphene and tamoxifen (TAM) exhibits identical IC50 values either in the cytosol or in the microsomal membranes. On the other hand, the microsomal fraction of breast tumors also binds [3H]hydroxyTAM, but with higher capacity and lower affinity than those of the cytosolic fraction. Furthermore, we did not observe correlation between the concentrations of ER and of antiestrogen binding sites (AEBS) in the tumors. These results indicate that microsomal membranes of human breast tumors contain estrogen binding sites which may be related to the cytosol ER recycling and that specific AEBS are predominantly localized in this membrane system. Furthermore, it is shown that the magnitude of estradiol binding to microsomes depends on the ER positive degree of the tumors, whereas the magnitude of the antiestrogen binding to the microsomes is independent of the ER status of the tumors.

Unique molecular properties of a urea- and salt-stable DNA-binding estrogen receptor dimer covalently labeled with the antiestrogen [3H]desmethylnafoxidine aziridine. A comparison with the estrogen-receptor complex.

A new antiestrogen affinity ligand for the covalent labeling of estrogen receptors, [3H]desmethylnafoxidine aziridine, has been used to investigate the salt- and temperature-independent formation of DNA-binding estrogen receptor forms from untransformed (300 kilodaltons) receptor. Calf uterine estrogen receptor proteins labeled with [3H]estradiol or [3H]desmethylnafoxidine aziridine were quantitatively transformed (greater than 90%) to their DNA-binding configuration in low ionic strength buffers by brief exposure to 3 M urea at 0 C. The urea effect was hormone-dependent and partially reversible. The transformed receptors were purified (ca 250-fold) by affinity chromatography on single-stranded DNA-agarose in the continued presence of 3 M urea to prevent transformation reversal. Scatchard analyses revealed a single class of high affinity radioligand binding sites (Kd = 0.34 nM) unchanged by urea-induced transformation and purification. The DNA-binding receptor form labeled with [3H]desmethylnafoxidine aziridine was stable as a probable dimer in 3 M urea with 0.4 M KCl and displayed no evidence of size (Stokes radius 7.3 to 7.5 nm; 4.2 to 4.3 S; Mr = 136,800) heterogeneity. Sodium dodecyl sulfate-polyacrylamide gradient gel electrophoresis indicated the presence of an intact 67 kDa steroid-binding receptor subunit. Reverse-phase chromatography of the covalently labeled receptor on C4 and phenyl stationary phases revealed no evidence of structural heterogeneity. The surface charge of the estrogen- and antiestrogen-receptor complexes, however, was distinctly different in both the presence and absence of 3 M urea. Thus, exposure to urea was an effective salt- and temperature-independent means for achieving the complete transformation of receptor to its stable DNA-binding dimer configuration. The ligand-induced differences in receptor surface charge and the urea effects on DNA-binding (but not hormone-binding) suggest that both electrostatic and hydrophobic or hydrogen bonding receptor domains are influenced by ligand binding.

Receptor binding of NBD-labeled fluorescent estrogens and progestins in whole cells and cell-free preparations.

We have studied the interactions of four fluorescent steroid conjugates with either the estrogen or progesterone receptor, both in whole cells and cell-free receptor preparations. The fluorophore, nitrobenzoxadiazole (NBD), was conjugated with a synthetic progestin, with a steroidal estrogen, a non-steroidal estrogen, and with an antiestrogen. With all compounds, receptor-specific binding could be detected by fluorescence measurements following extraction from the protein into an organic solvent. In the native state, however, the NBD-ligand-receptor complex is essentially non-emissive, although these ligands fluoresce strongly when associated with non-specific binders such as albumin. The binding site concentrations and relative affinities determined by fluorescence (after extraction) correspond well with those determined by [3H]estradiol or [3H]R5020 binding to their respective receptors. In T47D breast cancer cells, the NBD-progestin showed receptor-mediated uptake and nuclear localization. These compounds have provided valuable information about the interactions of low and medium affinity ligands with their receptors; however, the successful use of fluorescent ligands for detecting steroid receptors under native-bound conditions, by "imaging" modalities (fluorescence microscopy and flow cytometry) will require the development of fluorophores that are emissive while receptor bound or assay protocols that enable the environment of ligands associated with the receptor to be controlled.

Accumulation of oestrone by pig blastocysts and its potential physiological significance for blastocyst development in vitro.

Oestrone accumulation of Day-5 pig blastocysts and the potential physiological significance of oestrone and oestradiol-17 beta for blastocyst development were investigated in vitro. After 6 h of in-vitro culture in medium supplemented with 10 nM-[3H]oestrone, the accumulation amounted to 550 +/- 49 d.p.m. (s.e.m.) per 10 blastocysts. The accumulation of [3H]oestrone (or its metabolite(s] was reduced (P less than 0.001) in the presence of a 100-fold excess of unlabelled oestrone or oestradiol-17 beta to 135 +/- 14 d.p.m. or 148 +/- 28 d.p.m. per 10 blastocysts, respectively. The accumulation of [3H]oestrone was not affected in the presence of a 100-fold excess of unlabelled progesterone, testosterone or oestrone sulphate. When blastocysts were post-incubated for 30 or 60 min in [3H]oestrone-free medium, blastocysts retained 74.1 +/- 16.8% and 66.0 +/- 10.4%, respectively of their initial radioactivity. In parallel experiments with [3H]progesterone the respective values were 23.8 +/- 3.0% and 21.7 +/- 2.1%. The presence of the antioestrogen nafoxidine (15 micrograms/ml) in basic culture medium impaired (P less than 0.001) the transformation of morulae to blastocysts (21.5 +/- 8.9%) compared to controls (98.3 +/- 1.7%). The inhibitory effects could be overcome (P less than 0.001) by a supplementation with 1 nM- or 100 nM-oestradiol-17 beta (62.5 +/- 12.8% and 80.0 +/- 6.2% development to blastocysts) but not with 1 nM- or 100 nM-oestrone (30.3 +/- 9.6% and 45.2 +/- 10.5%). Blastocyst expansion was also decreased P less than 0.01) to 61.0 +/- 11.4% of control values in the presence of 15 micrograms nafoxidine ml.(ABSTRACT TRUNCATED AT 250 WORDS)

Studies on mechanisms of estrogen and antiestrogen action in human breast cancer.

PIP: Studies have shown that patients whose tumors are ER (estrogen receptor) positive have at least tenfold chance of responding successfully to endocrine therapy. However, not all ER-containing tumors necessarily respond, and ERs are thus considered necessary but not sufficient markers of hormone dependence. It has been proposed that progesterone receptors (PgR) found in human breast tumors may serve as a marker of estrogen action in breast cancer, since PgR synthesis is known to be controlled by estrogens in the uterus. The authors have recently worked on a modal system which shows that PgR are under estrogen control and that PgR synthesis involves the ER. Their study concerning the effects of antiestrogens show that tamoxifen (TAM) is a potent inducer of PgR on the cells and that TAM's estrogenic property is masked at very high doses (1 mcg/M) which also inhibit cell growth. Nafoxidine (NAF), another antiestrogen, had little effect on PgR when tested over a wide dose range. It has been reported that growth-inhibitory effects of both antiestrogens can be reversed by estradiol; this indicates that the effects of these compounds are mediated through the ER system. A complex system in cells in which ER binding and translocation and turnover of nuclear receptors or their processing mediate the induction of PgR by estradiol is described.^ieng

In vitro and in vivo interactions of [3H]dimethylstilbestrol with the estrogen receptor.

PIP: The binding of an antiestrogen with the uterine estrogen receptor (R) has been studied directly in vitro using tritiated dimethylstilbestrol (DMS). The affinity of DMS for R as determined at equilibrium was similar to that of estradiol-17beta (E2) (K(D) approximately .3 nN) when taking into account the higher nonspecific binding of DMS. The number of DMS binding sites was constantly found to be inferior to that of E2. The fact that the DMS binding entity specifically bound estrogen and antiestrogen, was destroyed by pronase, displayed an 8S sedimentation constant, and interacted in vitro with DNA, strongly suggested that DMS interacted directly with R. The association of DMS to R was a simple 2nd-order reaction while its dissociation was a 1st-order reaction with 2 slopes. The association and dissociation rate constants of the R-DMS complex were, respectively, slower and higher than those of the R-E2 complex. The rapid dissociation rate of DMS could be responsible for its inability to protect the receptor binding sites against thermo-inactivation. Tritiated DMS was able in vivo to induce the nuclear translocation of the receptor. However, as with other short-acting antiestrogens and contrary to Nafoxidine, the time of nuclear retention of R was short. These results are in agreement with the assumption that the length of the nuclear retention of R is determinant in explaining the weak agonist activity of this compound.^ieng

Antiestrogen action: differential nuclear retention and extractability of the estrogen receptor.

Since there is a much longer uterine nuclear retention of the U-11, 100A (antiestrogen) receptor complex (UARC) than of the estradiol receptor complex (ERC) at 4-12 hrs after injection, experiments were designed to determine if there is a difference between the relative nuclear affinities for the two RCs as determined by extraction with various ionic strength mediums. Although the UARC was retained longer in the nuclear fraction in vivo, the UARC was completely extractable with 0.3M KCl or 50mM spermine, whereas the ERC demonstrates a salt-resistant form. This suggests that the ERC is more tightly bound to nuclear components through this salt-resistant form of the receptor. In addition, various intercalating agents were used to distinguish the different nuclear chromatin DNA sites where the UARC and ERC may be binding. With actinomycin D (50 uM) more ERC than UARC was retained in the nuclear fraction. However, with ethidium bromide (100uM) less ERC than UARC was retained. Also, the ERC selectively released by ethidium bromide is precisely that fraction not released by salt. These results indicate that the UARC and ERC bind to different chromatin loci.

Interaction of delta-5-androstene-3beta, 17beta-diol with estradiol and dihydrotestosterone receptors in human myometrial and mammary cancer tissue.

Specific receptor binding of estradiol (E-2) and dihydrotestosterone (DHT) was studied in human myometrial tissue and in human mammary cancer tissue. The inhibition of binding for E-2 and DHT by E-2, testosterone (T), DHT, dehydroepiandosterone (DHEA), dehydroepiandrosterone-sulfate (DHEA-S), androstendione (A) and 5-androstene-3beta, 17beta-diol (Adiol) was tested with the use of dextran-coated charcoal separation of bound and free E-2, respectively, and DHT. The percentage of binding inhibition was calculated with reference to the inhibition obtained with nafoxidine in a molar concentration ratio of 1,000 for E-2 binding, respectively, with cyproterone acetate in a molar concentration ratio of 10,000 for DHT binding. In 15 samples of myometrium tested, receptors were found for both E-2 and DHT. From 19 samples of mammary carcinoma tissue one showed no binding activity, three samples did bind E-2 only, five samples DHT only, and ten samples showed binding of both steroids. A 50% inhibition of E-2 binding, in myometrial as well as in tumor tissue, required a molar concentration ratio of 40 for Adiol, of more than 2,000 for DHEA. No significant inhibiting activity could be found for A up to a molar concentration ratio of 10,000 and for DHEA-S up to 40,000. With regard to DHT binding, Adiol is more active than E-2 and less active than T. Of the substances tested Adiol is therefore the only one which exerts a significant inhibiting influence at a molar ratio not far beyond the physiological range. This signifies that Adiol might interfere at the receptor level in the estrogenic stimulation of mammary cancer cells.

PIP: Specific receptor binding of estradiol-17beta (E2) and dihydrotestosterone (DHT) was studied in human myometrial tissue and in human mammary cancer tissue. The inhibition of binding for E2 and DHT by E2, testosterone (T), DHT, dehydroepiandrosterone (DHEA), dehydroepiandrosterone-sulfate (DHEA-S), androstenedione (A) and 5-androstene-3beta, 17beta-diol (Adiol) was tested with the use of dextran-coated charcoal separation of bound and free E2, respectively, and DHT. The percentage of binding inhibition was calculated with reference to the inhibition obtained with nafoxidine in a molar concentration of 1000 for E2 binding, respectively, with cyproterone acetate in a molar concentration ratio of 10,000 for DHT binding. In 15 samples of myometrium tested, receptors were found for both E2 and DHT. From 19 samples of mammary carcinoma tissue 1 was without binding activity, 3 samples bound E2 only, 5 samples DHT only, and 10 showed binding of both. A 50% inhibition of E2 binding in myometrial as well as in tumor tissue, required a molar concentration ratio of 40 for Adiol, of more than 2000 for T and for DHT, and of about 20,000 for DHEA. Significant inhibiting activity for A up to a molar concentration ratio of 10,000 was absent. This was also true for DHEA-S up to 40,000. With regard to DHT binding, Adiol is more active than E2 and less active than T. Thus Adiol was the only substance which exerted a significant inhibiting influence at a molar ratio near physiological range. Adiol might interfere at the receptor level in the estrogenic stimulation of mammary cancer cells.

Estradiol receptor activity in tissues of human breast cancer.

Using the method for receptor determination recommended by the EORTC breast, cancer group we found 33.8% of our tissue specimen experimentally estradiol sensitive. No relation could be seen between receptor binding ability of the tumor tissues and lymph node involvement, tumor histology, clinical stages of the disease and menopausal state of the patients, respectively. In most cases primary tumor tissues and tissues of invaded lymph nodes show a corresponding behaviour in relation to estradiol uptake. The differences observed in the effects of various antiestrogenic substances on estradiol binding ability can not be explained at present.