| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
AstraZeneca Pharmaceuticals, Wilmington, Delaware 19850 (J.L., K.S.K., C.W.S.); 431 83 Mölndal, Sweden (H.K., G.A., E.N.); and 151 85 Södertälje, Sweden (C.D.)
Address all correspondence and requests for reprints to: Jianwei Liu, AstraZeneca Pharmaceuticals, 1800 Concord Pike, LW207D, Wilmington, Delaware 19850. E-mail: jianwei.liu{at}astrazeneca.com.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTS AND DISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
or ERß ligand binding domain (LBD), a biotinylated coactivator peptide, europium-labeled anti-FLAG antibody, and streptavidin-conjugated allophycocyanin. Upon agonist binding, the biotinylated coactivator peptide is recruited to FLAG-tagged ER LBD to form a complex and thus allow fluorescence resonance energy transfer (FRET) to occur between europium and allophycocyanin. Compounds with estrogen antagonism block the agonist-mediated recruitment of a coactivator and prevent FRET. The assay was used to evaluate the preference of ERs for various coactivators and ligands. Both ER and ERß exhibited strong preferences for coactivator peptides corresponding to steroid receptor coactivator-1 and PPAR
coactivor-1 vs. peroxisome proliferator-activated receptor-interacting protein and cAMP response element binding protein-binding protein. 17ß-Estradiol acted as a nonselective agonist for ER and ERß. Genistein showed full agonism for ER and only partial agonism for ERß, but with higher potency for ERß than ER. Both raloxifene and tamoxifen behaved as full antagonists in this functional assay. The results obtained using compounds with a wide range of potency correlated well with those from a cell-based reporter gene assay. Therefore, this simple in vitro functional assay is predictive of ligand-dependent transactivation function of the receptor and, as such, is useful in nuclear receptor applications including mechanistic studies. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTS AND DISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
The ER domain structure is typical of NRs. The amino-terminal region is involved in transactivation of gene expression and has been termed the activation function domain, or AF1 domain. The middle region contains a two-zinc finger structure, which plays an important role in binding to specific DNA response elements and in receptor dimerization. The carboxyl-terminal region also contains an activation function domain, termed AF2. Within AF2 lies a ligand-binding domain (LBD) that is crucial for binding of receptor specific ligands and also corepressors and coactivators. These binding interactions affect receptor dimerization, nuclear translocation, and modulation of target gene expression by AF2 (2, 3).
The LBD is structurally conserved among the NR family and consists of between 10 and 12 -helices folded in a globular domain (4). A central hydrophobic pocket accommodates the cognate ligand, which upon binding induces a conformational change in the LBD, exposing a coactivator-docking site on the LBD surface. The subsequent recruitment of coactivators to the NR promotes transactivation. However, the binding of raloxifene or tamoxifen to ERs results in a different conformation change in which the coactivator-binding site is blocked from interaction with a coactivator, thereby blocking AF2-mediated transcriptional activity.
Several classes of NR coactivators (NCoA) interact with ERs (5). One class consists of the cAMP response element binding protein-binding protein (CBP) and the related factor, p300. These molecules promote the transcription of NR-responsive genes as well as cAMP response element binding protein-binding protein and activator protein-1-responsive genes, and therefore may integrate multiple signal transduction pathways (6, 7). A second class of coactivator proteins is called steroid receptor coactivators (SRC) and includes SRC1/NCoA1, SRC2/transcriptional intermediary factor 2 (TIF2)/glucocorticoid receptor interacting protein (GRIP)/NCoA2, and SRC3/receptor-associated coactivator 3 (RAC3)/p300/CBP-cointegrator-associated protein (pCIP)/activator for TR (ACTR)/amplified in breast cancer 1 (AIB1)/TR-associated molecule-1 (TRAM-1)/NCoA3 (8). These three proteins also interact with CBP/p300 in ligand-dependent transactivation by NRs. SRC1’s role appears to be restricted to NRs. Two isoforms of SRC1 have been identified: SRC1a and SRC1e. These isoforms have divergent sequence at their C- termini and differ in their ability to enhance ER-driven transactivation (9). In vitro studies indicate that SRC3 enhances the transcriptional activity of ER, but not ERß (10). Other coactivators that have been shown to interact with NRs include PRIP [peroxisome proliferator-activated receptor (PPAR) interacting protein]/RAP250 (11, 12) and PGC-1 (PPAR coactivor-1; Ref. 13).
All nuclear coactivators contain one or more copies of the highly conserved LXXLL amphipathic -helix motif, also called the NR box or NRB. The NRB resides within a larger domain termed the nuclear receptor interaction domain, or NRID, that directly binds to hormone-liganded NRs. For example, Zhou et al. (14) used NRIDs from CBP and SRC-1 in a coactivator FRET assay and showed that both interacted with PPAR and ER but with differing affinities: the CBP NRID had a higher affinity for PPAR, whereas SRC-1 NRID preferred ER. Using similar time-resolved fluorescence resonance energy transfer (FRET) technology, we used coactivator NRBs and developed a homogeneous in vitro functional assay to quantitate the ligand-induced recruitment of nuclear coactivator peptides to ER and ERß LBDs. Using such a FRET assay, we successfully investigated receptor-coactivator interactions and pharmacological profiles of some agonists, partial agonists and SERMs.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTS AND DISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
). An agonist will induce the recruitment of the NRB peptide to ER LBD, resulting in FRET. Compounds with antagonist activity will block the agonist-induced recruitment of the NRB peptide, resulting in decreased FRET. Thus, this homogeneous in vitro assay can be used to quantitate effects of binding of both agonists and antagonists to ER LBDs.
|
and ERß. The biotinylated NRB peptides used in the experiments were chosen to reflect the diversity of coactivators known to bind ERs (Fig. 1B
). The biotinylated peptides SRC1-II and SRC1-IV represent two different NRBs, both of which are expressed in SRC1a, whereas SRC1-IV is missing in SRC1e. The similar length of NRBs was designed to minimize effect of position or orientation differences of two fluorophores on FRET.
Initial studies were designed to optimize the concentrations of reagents used to achieve FRET. Various concentrations of SRC1-II peptide and streptavidin (SA)-APC were incubated with 1 nM Eu-anti-FLAG antibody, 10 nM FLAG-ER LBD, and 100 nM 17ß-estradiol (E2). The nonspecific binding was measured by a parallel incubation in the absence of the agonist. No ligand-dependent signal was observed when the biotinylated NRB peptide was replaced with a nonbiotinylated peptide (data not shown). As shown in Fig. 2, varying the concentrations of SRC1-II peptide and SA-APC affected FRET intensity. The signal decreased as the ratio of SRC1-II peptide to SA-APC exceeded 4. At ratios above 4, one has exceeded the relative binding capacity of the streptavidin for biotinylated peptide. Under these conditions some SRC1-II binds to the ER but does not complex with SA-APC, resulting in lower FRET values. The optimal ratio was about 3. Therefore, 100 nM SRC1-II and 30 nM SA-APC were chosen as standard concentrations.
|
and ERß (data not shown). The assay tolerated dimethylsulfoxide (DMSO) up to 5% without a significant reduction in signal.
The effect of agonist concentration on FRET was evaluated, using different concentrations of SRC1-II peptide and ERß LBD. E2 induced a concentration-dependent increase in FRET, with an EC50 value that was independent of the amount of SRC1-II peptide used in the assay (Fig. 3A). The EC50 values for E2 in the presence of 50, 100, and 200 nM SRC1-II were 3.4, 2.9, and 3.5 nM, respectively. A reduction in total FRET was observed at high SRC1-II concentration (200 nM), for reasons stated above. As shown in Fig. 3B, increasing the concentration of ERß LBD resulted in an increase in the maximal response to E2, whereas the EC50 for E2 remained unchanged (2.0–3.4 nM, with Hill coefficients of 0.9–1.2). These results also demonstrated that the assay has high precision with coefficients of variation less than 5%. In dose response experiments, a Z' value of 0.7–0.9 was achieved even when the signal to background ratio was less than 2-fold as observed at low agonist concentrations. The Z' value is used to assess both specific signal and precision for accurate assay performance (17). A value of above 0.5 indicates that the assay is capable of providing robust, high precision data. The high precision resulted, in part, from the simple liquid handling steps for FRET and also from the ability to normalize the FRET intensity. Another important feature of this assay is that the signal is stable for more than 48 h, which would allow batch processing for high throughput screening applications.
|
, these peptides differed significantly in their ability to bind ligand-activated ER LBDs. These experiments used a saturating concentration of agonist; therefore, the EC50 values reflect the affinities of NRB peptides for the ERs. The affinity rank order of NRB peptides for ER LBD was similar to ERß LBD: SRC1-II 
PGC1 SRC1-IV PRIP, CBP. ER LBD showed slightly higher affinity for CBP than PRIP, whereas ERß LBD exhibited the opposite. The NRB peptides exhibited different efficacies (i.e. maximum FRET signals). However, the efficacy rank order was the same for E2-bound ER LBD and ERß LBD and paralleled potency rank order. Such efficacy difference should reflect conformation differences in the ER-coactivator peptide complexes so that efficiency of FRET varied. Even though the coactivator peptides have similar length of amino acids at both ends of LXXLL, binding of different coactivator peptides to ER may induce different confirmation changes. In turn, the distance between the fluorescent donor in ER and acceptor in the coactivator peptide may vary among the ER-coactivator peptide complexes.
|
and ERß LBDs as observed with E2. However, the magnitude of FRET seen with ERß LBD was reduced compared with that seen with E2, indicating that genistein is a partial agonist for ERß LBD in this assay system.
The CBP NRB showed a much weaker interaction with ER LBDs than SRC1, which is consistent with literature data (6, 15). The results in Fig. 4 are also consistent with those of Torchia et al. (18), who reported that SRC1-II had higher binding affinity than SRC1-IV for ER. Bramlett et al. (19) reported conflicting results; however, the SRC1-II and SRC1-IV peptides used in their study were 7 and 10 amino acids shorter, respectively, at the N terminus than those used in our assays. The difference in length of the NRB peptides may account for the discrepancy between their results and ours. Adjacent sequences around the LXXLL motif can increase or decrease the affinity of a NRB for NRs (20), hence our decision to use relatively longer NRB peptides.
In the presence of high concentrations of NRB peptides (i.e. above 300 nM), SRC1-II and PGC1 were observed to interact with ER LBDs in the absence of hormone (data not shown). Under these conditions, agonist increased the coactivator peptide affinity for ERs by 3-fold. Similar observations were reported with NRB binding using ELISA and scintillation proximity assays (21). Low affinity ligand-independent interaction of ERs with coactivators may account for basal constitutive transactivation activity of ERs observed in our cell-based functional assay (not shown).
Although the NRB peptides had different affinities for agonist-ER complex, the potency of E2 for ER LBD was similar when high concentrations of the different NRBs were used in the FRET coactivator recruitment assay (Fig. 5). The efficacy rank order of the NRB peptides was similar to their affinity rank order for ERs. The same EC50 (3.0 nM) was reported in SRC1-II recruitment assay using a time-resolved fluorescence coactivator peptide-binding assay (22). Under these assay conditions, the ability to induce FRET is dependent on agonist concentration. Thus, the EC50 for E2 is independent of the NRB peptide used in the assay, but the efficacy is dependent on which NRB peptide is used.
|
). E2 showed the same affinity and efficacy for ER and ERß LBDs, which is consistent with the literature data that E2 is a nonselective ER agonist. Genistein showed a more than 10-fold higher selectivity for ERß LBD vs. ER LBD, but also exhibited partial agonism on ERß LBD and full agonism on ER LBD (Fig. 6A). This is in agreement with the results obtained with genistein and multiple NRB peptides (Fig. 4). Furthermore, in a cell-based functional assay, genistein was reported to induce 50–60% of the ERß response seen with E2 (23).
|
Raloxifene and tamoxifen were tested for both agonist and antagonist activities in the recruitment assay. During a 5-h incubation, neither compound showed agonist activity for ER or ERß LBD. E2 was then added to the mixture and incubated overnight. Raloxifene completely antagonized E2-induced coactivator recruitment, with 30-fold selectivity for ER LBD vs. ERß LBD (Fig. 6B). Tamoxifen also exhibited full antagonism in this assay (data not shown). Similar results have been reported in cell-based assays (23, 25, 26) and in binding assays (27).
Therefore, the FRET coactivator recruitment assay is able to reflect both the binding affinities and cell-based functions of antagonists and to discriminate agonist, partial agonist, and antagonist activities. Moreover, the in vitro functional FRET assay can be used to screen for both agonism and antagonism of compounds in a single run of compound screening.
Correlation with a Cell-Based Functional Assay
The recruitment assay described here provides a simple and robust in vitro functional assay for agonists, partial agonists, and antagonists and is consistent with literature data. To further evaluate its correlation with the function of transcriptional activation, several agonists covering a broad range of potencies from benzoxazole series were tested in both a cell-based reporter gene assay and the FRET in vitro functional assay. The cell-based assay used full-length ERß and an ER-response element (ERE)-reporter system. As shown in Fig. 7, the rank order of compound potency was similar in the in vitro functional assay and the cell-based functional assay. However, all compounds tested showed higher potency in the cell-based assay than in the in vitro coactivator recruitment assay. Because the FRET assay measures the binding of a single NRB to ligand-bound ERß LBD, the recruitment potencies of ligands in this assay may underestimate the ability of coactivators to induce transcription in a cell-based reporter gene assay. For example, coactivators contain multiple NRBs and can exhibit higher affinity for ER than a single NRB, presumably reflecting the potential for cooperative interactions for enhanced binding and signal amplification (28). Indeed, cocrystallization studies of PPAR and an 88-amino-acid-long SCR-1 fragment containing two NRBs have shown that one coactivator can bridge across a receptor dimer (29). Most importantly, the rank order of compound potency in the coactivator FRET assay accurately reflected the results of the more complex whole cell transcriptional assay.
|
In conclusion, the coactivator FRET assay described here was developed in a 384-well high throughput format and successfully used to explore the preference of ERs for different NRBs and to profile some compounds for their agonist and antagonist properties against ERs. The results obtained using this method were comparable to those obtained from a cell-based functional assay and reported in literature, and demonstrate the value of this homogeneous, high throughput method to dissect the mechanisms of interaction between NR and coactivators or ligands and to screen compounds for NR regulation in drug discovery applications.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTS AND DISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
competent cells were purchased from Life Technologies, Inc. (Rockville, MD). LANCE Eu-W1024-labeled anti-FLAG antibody, SA-APC, and low-volume 384-well white plates were purchased from Perkin-Elmer Corp. Wallac, Inc. (Gaithersburg, MD). Ninety-six-well cell culture plates (Biocoat Collagen I Cellware) were purchased from BD Biosciences (Bedford, MA). DMEM was from Mediatech (Herndon, VA), and charcoal-stripped FBS was from Biocell (Rancho Dominguez, CA). Chlorophenol red-ß-D-galactopyranoside was purchased from Roche Molecular Biochemicals (Indianapolis, IN). Other reagents were purchased from Sigma.
ER and ERß Constructs
The human ER LBD fragment (amino acids 301–553) was obtained by PCR amplication. An NdeI site followed by a nucleotide sequence coding for a FLAG tag (underlined) was designed in the forward primer, 5'-GCCATATGGACTACAAGGACGACGATGACTCTAAGAAGAACAGCC-3'. The reverse primer contained a BamHI site after a stop codon (underlined), 5'-GGCGGGATCCTCAAGTGGGCGCATGTAGGCGG-3'. The 0.8-kb amplified DNA product was digested with the appropriate enzymes and subcloned into corresponding sites of a modified pET-28 vector (Novagen, Madison, WI). The vector contained maltose binding protein (MBP) and His tag coding sequences followed by a thrombin cleavage site (LVPRGS) coding sequence before the multiple cloning sites.
The expression plasmid was transformed into BL21(DE3), which were grown at 37 C in the presence 50 µg/ml kanamycin, until induction at OD600 = 1.7 with 250 µM isopropyl-ß-D-thiogalactoside. After induction, the culture was transferred to 20 C and left overnight. Cells were harvested by centrifugation and resuspended in buffer A (25 mM bis-Tris-propane, 10% glycerol, 300 mM NaCl, 1 mM triscarboxyethylphosphine (TCEP), Complete EDTA-free proteinase inhibitors Roche Diagnostics and 0.5 mg/ml lysozyme) followed by sonication and centrifugation at 40,000 x g for 45 min. The supernatant was applied to a nickel-charged nitrilotriacetic acid column and the 6xHis-MBP-FLAG-ERLBD was eluted in buffer A containing 50 mM imidazole. Fractions containing the fusion protein 6xHis-MBP-FLAG-ERLBD were combined and desalted on a Sephadex G25 column preequilibrated with buffer B [25 mM bis-Tris propane, 10% vol/vol glycerol, 50 mM NaCl, 1 mM TCEP (pH 9.5) at 4 C]. The fusion protein was then cleavaged with thrombin (5 U thrombin/mg protein) and the resulting MBP and FLAG-ERßLBD were separated by a Source 15Q column preequilibrated with buffer B. FLAG-ERLBD was eluted with a NaCl gradient of 1–100% in buffer C [25 mM bis-Tris propane, 10% vol/vol glycerol, 1 M NaCl, 1 mM TCEP (pH 9.5) at 4 C]. The Source 15Q fractions containing FLAG-ERLBD were desalted on a Sephadex G25 preequilibrated with buffer B. The purity of FLAG-ERLBD was estimated to about 95% judged by Commassie stained SDS-PAGE. The mass (calculated, 30,967 Da) of FLAG-ERLBD was verified (found, 30,965 Da) using an HP 1100 series HPLC connected to an electrospray MSD detector (Hewlett-Packard Co.). Protein concentration was determined by a Bradford-like dye-binding assay (Bio-Rad Laboratories, Inc.). Human FLAG-ERß LBD (amino acids 254–510) was obtained in a similar way with higher than 95% purity.
Recruitment Assay
The optimized reaction mixture contained 5 nM of purified FLAG-ER LBD, 30 nM SA-APC, 1.0 nM 77 Eu-FLAG antibody, and 100 nM biotin-SRC1-II peptide in the assay buffer [20 mM HEPES, 50 mM KCl, 1 mM EDTA, 0.05% Nonidet P-40, 1 mM dithiothreitol, and 1 mg/ml BSA (pH 7.2)]. Agonist experiments were performed in a 384-well plate by adding 10 µl of the reaction mixture to 10 µl of agonists (diluted with the assay buffer) using a CyBi-Well liquid dispenser. For determination of antagonist activities, 10 µl of the reaction mixture were added to 5 µl of compounds, after by 5 µl of 16 nM E2 (4 nM final concentration). The reactions were routinely incubated overnight at 4 C because E2 binding to ER reached equilibrium after 9 h (30). The samples were counted using a Wallac, Inc. Victor V 1420 Multilabel Counter with a setting of 50-µsec time delay, excitation at 340 nm, and emission at both 615 nm and 665 nm. Fluorescence intensity ratio at 665 nm/615 nm (multiplied by 104) was used for agonist EC50 plotting directly. For antagonist activities, FRET in the absence of E2 was calculated as 100% inhibition; DMSO control, 0% inhibition. All of the compounds were dissolved in DMSO.
Cell-Based Functional Assay
HEK293 cells were routinely maintained in DMEM supplemented with 10% HFBS, 2 mM L-glutamine, and G418 (0.4 mg/ml). Cells were transfected with a vector containing full-length human ERß and a second vector containing the ß-galactosidase (ß-gal) gene with the vitellogenin estrogen response element (ERE) in its promoter. These expression vectors were constructed with a similar way as described by Ernst et al. (31). The transfection protocol used the calcium phosphate method (Profection Kit, Promega Corp.) according to the supplier’s recommendations. The day after transfection cells were trypsinized and seeded onto collagen I coated 96-well culture plates in DMEM (phenol red free) containing 2% FBS (charcoal stripped) and 2 mM L-glutamine at 25,000 cells per well. After approximately 2–4 h, compounds were added and incubated overnight at 37 C for determination of agonist activities. The level of ß-gal expression was determined using chlorophenol red-ß-D-galactopyranoside as substrate in a lysis buffer (Roche Molecular Biochemicals, Indianapolis, IN) as per the manufacturer’s instructions.
|
ACKNOWLEDGMENTS |
|---|
LBD, Linda Hagman for purifying FLAG-tagged ER LBD, Dr. Philippe Cronet for supplying a modified pET28 plasmid, and Dr. Deborah S. Hartman for reviewing the manuscript critically. |
FOOTNOTES |
|---|
; PPAR, peroxisome proliferator-activated receptor; PRIP/RAP250, PPAR-interacting protein; SA, streptavidin; SERMs, selective ER modulators; SRC, steroid receptor coactivator; TCEP, triscarboxyethylphosphine. Received for publication September 23, 2002. Accepted for publication November 25, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTS AND DISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
transcriptional activity by the coactivator PGC-1. J Biol Chem 275:16302–16308 and estrogen receptor ß to partial estrogen agonists/antagonists. Mol Pharmacol 54:105–112. Nature 395:137–143[CrossRef][Medline]
NURSA Molecule Pages Link:
This article has been cited by other articles:
 |
|
 |
|
  J. R. Gunther, Yuhong Du, E. Rhoden, I. Lewis, B. Revennaugh, T. W. Moore, Sung Hoon Kim, R. Dingledine, Haian Fu, and J. A. Katzenellenbogen A Set of Time-Resolved Fluorescence Resonance Energy Transfer Assays for the Discovery of Inhibitors of Estrogen Receptor-Coactivator Binding J Biomol Screen, February 1, 2009; 14(2): 181 - 193. [Abstract] [PDF]
|
|
|
|
 |
|
 S. Wang, C. Zhang, S. K. Nordeen, and D. J. Shapiro 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 J. Biol. Chem., February 2, 2007; 282(5): 2765 - 2775. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Wang and S. S. Simons Jr. Corepressor Binding to Progesterone and Glucocorticoid Receptors Involves the Activation Function-1 Domain and Is Inhibited by Molybdate Mol. Endocrinol., June 1, 2005; 19(6): 1483 - 1500. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. S. Ozers, K. M. Ervin, C. L. Steffen, J. A. Fronczak, C. S. Lebakken, K. A. Carnahan, R. G. Lowery, and T. J. Burke Analysis of Ligand-Dependent Recruitment of Coactivator Peptides to Estrogen Receptor Using Fluorescence Polarization Mol. Endocrinol., January 1, 2005; 19(1): 25 - 34. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
