| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Departments of Pharmaceutical Chemistry and Cellular & Molecular Pharmacology (R.V.W., N.J.C., D.C.C., T.S.S.), University of California, San Francisco, California 94143-0446; Department of Pharmacology and Cancer Biology (C.-Y.C., D.P.M.), Duke University Medical Center, Durham, North Carolina 27710; Graduate Program in Chemistry and Chemical Biology (N.J.C.), University of California, San Francisco, California 94143-0446; Departments of Medicine and Cell Biology (R.N.D.), NSF Center for Biological Timing, University of Virginia Health Sciences Center, Charlottesville, Virginia, 22908; Metabolic Research Unit, Diabetes Center and Department of Medicine (J.D.B., F.S.), University of California, San Francisco, California 94143-0540
Address all correspondence and requests for reprints to: Fred Schaufele, HSW-1119, 513 Parnassus, University of California, San Francisco, California 94143-0540
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
helix, containing variants of the sequence LXXLL, found in many coactivators. In cells, it is unknown whether ER interactions with the different LXXLL-containing helices are uniformly similar or whether they vary with LXXLL sequence or activating ligand. Using fluorescence resonance energy transfer (FRET), we confirm in the physiological environment a direct interaction between the estradiol (E2)-bound ER and LXXLL peptides expressed in living cells as fusions with spectral variants of the green fluorescent protein. This interaction was blocked by a single amino acid mutation in the hydrophobic cleft. No FRET was detected when cells were incubated with the antiestrogenic ligands tamoxifen and ICI 182,780. E2, diethylstilbestrol, ethyl indenestrol A, and 6,4'-dihydroxyflavone all promoted FRET and activated ER-dependent transcription. Measurement of the level of FRET of ER with different LXXLL-containing peptides suggested that the orientations or affinities of the LXXLL interactions with the hydrophobic cleft were globally similar but slightly different for some activating ligands. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
AND ERß ARE members of a large class of nuclear receptors that regulate the transcription of genes in response to binding small molecule ligands (1, 2, 3). The regulatory roles of ER in disorders like breast cancer and osteoporosis make it an important therapeutic target (4, 5, 6, 7, 8, 9). One of the signature features of the ER-targeting compounds is that they may have different stimulatory or repressive effects depending on the cellular context. For instance, the breast cancer drug tamoxifen is an antiestrogen in breast tissue but, in the uterus, it mimics the estrogenic activity of the physiological hormone, E2 (7, 10). Other compounds, such as the osteoporosis drug raloxifene, show a different clinical profile (11). Improved designer estrogens with higher selectivity for specific tissues would permit tissue-specific, estrogen-regulated disorders to be treated with minimal side effects (12, 13).
The mechanisms by which E2 and the selective ER modulators (SERMs) show tissue-specific activities remain unresolved but are at least partially related to ligand-regulated alterations in ER structure and function. In both its unliganded and liganded state, the ER is part of larger complexes with other accessory proteins (14). These accessory proteins, some of which are still unidentified, can stabilize ER structure and regulate transcription at different DNA effector sites (15). Ligand binding causes a conformation change in the ER (16, 17), which alters the affinities of the receptor for these accessory proteins (18, 19). One possible mechanism for SERM activity is that different ligands recruit different sets of accessory proteins and thereby differentially regulate gene transcription (20, 21, 22). Differential cofactor interactions, together with tissue-dependent expression of ER, ERß, and each cofactor, could explain tissue-selective SERM activity.
To elucidate, and ultimately predict, differential SERM action, it is therefore essential to measure the ligand-induced, direct interactions between the ER and different accessory proteins in the cellular environment (20, 21, 22). Many coactivators that interact with the E2-activated ER contain one or more copies of a consensus sequence, LXXLL (L, leucine; X, any amino acid) (23, 24). Structural studies have shown that an isolated LXXLL peptide will interact with a hydrophobic cleft that forms on one surface of the E2-bound ER (16). This hydrophobic cleft constitutes the activation function AF-2, which is conserved amongst nuclear receptors and participates in ligand-regulated gene transcription (25). Two-hybrid interaction assays have proved very useful for identifying and characterizing the ligand-regulated interactions of LXXLL-containing factors and peptides with ER expressed in cells (20, 21, 22). However, two-hybrid assays measure only whether proteins interact, and not whether they interact with differing structural characteristics or affinities.
We applied a microscope-based assay using fluorescence energy resonance transfer (FRET) to measure in living cells the ability of a ligand to modulate LXXLL interactions with ER. FRET measures the proximity of two molecules as a consequence of the degree to which the fluorescence energy excited in a donor fluorophore, linked to one factor, is not emitted and instead is nonradiatively transferred to an acceptor fluorophore, linked to another factor (26, 27, 28, 29, 30). We observed in the cellular environment that ER, fused to the red fluorescent protein (RFP) interacted directly with LXXLL peptides, fused to the cyan (CFP) or green (GFP) fluorescent proteins. These interactions were promoted by E2 but blocked by tamoxifen and another SERM, ICI 182,780, which confirmed prior studies in two-hybrid (21) and fluorescence colocalization (31) assays. Like E2, the synthetic ligands diethylstilbestrol (DES), ethyl indenestrol A (EIA), and 6,4'-dihydroxyflavone (DHF) promoted FRET between ER-RFP and two different LXXLL peptides fused to GFP. All these interactions were dependent upon the integrity of AF-2 within the ligand binding domain of ER. E2, DES, EIA, and DHF yielded similar levels of FRET for the interaction of ER with one LXXLL peptide. However, small, ligand-selective differences in the level of FRET were measured for interaction with the other LXXLL target sequence. This indicated that there were subtle, ligand-specific, and LXXLL-specific differences in the orientation or affinity of LXXLL interaction with ER. The accurate measurement of such nuances in the interactions of ER in the cellular environment will help distinguish the similarities and cell-type dependent differences in ligand-selective ER activities.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
and LXXLL for FRET Measurements (21, 31), was fused to the carboxy terminus of CFP. X-ray crystallographic structures of LXXLL bound to ER (16) predict that, if LXXLL binds directly to ER, the CFP fluorophore should project toward RFP fused to the carboxy terminus of ER. This positioning would be optimal for FRET from the CFP donor to the RFP acceptor.
The ER-RFP fusion was transcriptionally active (Fig. 1
). ER-RFP or control expression vectors were transfected into ER-deficient HepG2 cells together with either of two different, E2-sensitive promoters controlling the expression of a luciferase reporter. One promoter consisted of three copies of an estrogen response element linked to a minimal TATA box (32). This reporter defines the "classical" activities of ER-RFP, in which estrogen response is mediated by direct ER interaction with a single DNA binding site in the promoter. The second promoter, from the complement 3 gene (C3), contains three suboptimal ER binding elements, which together allow ER to bind and regulate transcription in response to E2 (33).
|
, black bars) and C3 (Fig. 1, white bars) promoters were activated upon E2 addition. In contrast, tamoxifen and ICI 182,780 did not activate ER-RFP at either promoter, even though wild-type ER weakly activated the C3 promoter in the presence of tamoxifen, but not ICI 182,780 (22, 33). Thus, ER-RFP was defective in tamoxifen activation. Because the estrogenic activities of E2 were not disrupted by the fusion of RFP to the carboxy terminus of ER, ER-RFP remained viable for studying agonist activation via AF-2.
Controls for the Accurate Measurement of FRET between CFP-LXXLL and ER-RFP
The measurement of FRET relies on the accurate quantification of the amount of fluorescence emitted by the donor and acceptor fluorophores upon donor excitation (29). The donor CFP is excited optimally by blue light to emit light of energy in the blue-green (cyan) wavelength, whereas the RFP acceptor emits red light upon excitation by light of mid-visible wavelengths, including cyan. If the cyan fluorescent CFP is in close proximity to RFP (<100 Å apart), some of the fluorescence energy from CFP will be absorbed by, and excite, RFP. Thus, when excited by blue light, energy transfer from CFP to RFP would decrease the emission of cyan light and increase the emission of red light.
For controls, we first quantified the amount of fluorescence in cells that independently expressed ER-RFP or CFP-LXXLL. Expression vectors encoding ER-RFP or CFP-LXXLL were transfected into GHFT1–5 pituitary progenitor cells grown in estrogen-free media and plated onto microscope coverslips. GHFT1–5 cells contain endogenous ER, but promoter responses to E2 in GHFT1–5 cells are not significantly altered upon ER overexpression (34, 35). Because overexpression of ER in GHFT1–5 cells does not reduce ER response as it does in many other cell types (36), the actions of expressed ER measured in GHFT1–5 cells likely mimic those of endogenous receptors. GHFT1–5 cells also have a flat morphology, which facilitates data collection by fluorescence microscopy (31, 37). The transfected cells were treated with E2, or other ligands as discussed later, or with the control vehicle (ethanol). After allowing 24 h for expression, the amounts of fluorescence emitted from the control cells separately expressing ER-RFP and CFP-LXXLL were measured in the cyan, red, and FRET channels by quantitative fluorescence microscopy. Digital images from cells expressing CFP-LXXLL were collected by specifically exciting CFP with light of wavelengths between 431 and 434 nm and collecting emissions between 455 and 480 nm (Fig. 2A, cyan). Digital images from the ER-RFP control cells were collected by 550–560 nm excitation and 580–630 nm emission (Fig. 2B, red).
|
-RFP and CFP-LXXLL (Fig. 2B, cyan; Fig. 2A, red). This bleedthrough was accurately quantified by marking each nucleus containing CFP-LXXLL or ER-RFP as a contiguous assembly of pixels containing more fluorescence than the background. The total amount of cyan and red fluorescence above the background fluorescence was then measured within the nucleus of each control cell. Red fluorescence from 27 different E2-treated cells expressing only CFP-LXXLL was negligible: on average, 0.0009 ± 0.0026 the amount of cyan fluorescence. This means that the amount of energy emitted by CFP in the red channel was statistically insignificant. Similarly, the bleedthrough of ER-RFP into the cyan channel was 0.0016 ± 0.0019 the amount of emission in the red’ channel. These ratios were similar regardless of the ligand treatment for each cell (not shown). Although the amount of CFP-LXXLL or ER-RFP expressed in each transiently transfected cell varied, plotting the amount of bleedthrough as a function of the amount of CFP-LXXLL or ER-RFP fluorescence in each cell (Fig. 3, A and B, open boxes) showed that these ratios were consistently measured regardless of the amount of CFP-LXXLL or ER-RFP expressed.
|
-RFP caused the coexpressed LXXLL peptide to colocalize with the E2-bound ER, whereas the LXXLL peptide, by itself, distributed throughout the cell (Fig. 2). To determine if there was a hormone-regulated, direct interaction of LXXLL with colocalized ER, we measured FRET between coexpressed CFP-LXXLL and ER-RFP. The low level of CFP and RFP bleedthrough enabled us to selectively and accurately measure the amounts of CFP-LXXLL and ER-RFP coexpressed in the same cell. These values are then used to correct for the contributions of the known amounts of CFP and RFP to the FRET channel, as discussed below.
In cells coexpressing CFP-LXXLL and ER-RFP, FRET was detected, upon blue light excitation, as an increase in acceptor fluorescence transferred from the donor (FRET channel: 431–434 nm excitation/580–630 emission) relative to a decrease in donor fluorescence (cyan channel: 431–434/455–480). Therefore, FRET was measured as an increase in the ratio of FRET/cyan fluorescence from a cell expressing both CFP-LXXLL and ER-RFP relative to the FRET/cyan ratios emitted from independently expressed CFP-LXXLL and ER-RFP. In the control CFP-LXXLL-expressing cells, the amount of bleedthrough fluorescence into the FRET channel was 0.2455 ± 0.0094 that emitted in cyan channel (Fig. 2A), which did not vary with the overall amount of CFP-LXXLL in the cell (Fig. 3A, Xs). The bleedthrough of ER-RFP fluorescence to the FRET channel was 0.0288 ± 0.0066 that emitted in the red channel (Fig. 2B), which also did not vary with the amount of ER-RFP present in each cell (Fig. 3B, black boxes).
To calculate the relevant FRET/cyan from donor ratio, we first calculated the amounts of CFP-LXXLL and ER-RFP present in the coexpressing cells. This was achieved by subtracting the minor contributions of CFP to the red channel (0.09% the value of cyan fluorescence using the matched control data in the prior section) and of RFP to the cyan channel (0.16% the value of the corrected red channel). We then subtracted the contribution of RFP to the FRET channel (2.88% the value of the corrected red channel). This remaining signal in the FRET channel contained the CFP bleedthrough to the FRET channel plus any sensitized emissions that resulted from the transfer of energy from CFP to RFP. If there was no FRET, the FRET/donor (remaining FRET/corrected cyan, hereafter FRET/cyan) ratio remained that of the donor CFP alone (0.2455). However, if there was transfer of energy from CFP to RFP, the amount of CFP fluorescence decreased and the amount of FRET increased, so that the FRET/cyan ratio increased.
The FRET/cyan ratio averaged from 32 E2-treated cells coexpressing CFP-LXXLL and ER-RFP increased to 0.5412 ± 0.2018. Because cells with low amounts of CFP-LXXLL relative to ER-RFP have fewer CFP donors in close proximity to the RFP acceptor, the FRET/cyan ratio varied with the relative amounts of RFP and CFP fluorescence measured in each cell. To account for this variation, the FRET/cyan ratio was graphed against the relative amounts of bleedthrough-corrected cyan and red fluorescence for each E2-treated cell (Fig. 4, black boxes). The slope of this graph was linear and consistent between experiments, indicating that the acceptor (RFP)-driven level of FRET within each cell was a constant. If the LXXLL peptide were not attached to CFP (Fig. 4, gray triangles) or if CFP-LXXLL and ER-RFP coexpressing cells were treated with the antiestrogen ICI 182,780 instead of E2 (Fig. 4, white boxes), the FRET/cyan ratio remained identical to the 0.2455 FRET/cyan ratio of CFP-LXXLL alone (Fig. 4, Xs at acceptor/donor = 0) regardless of the relative amounts of CFP-LXXLL and ER-RFP measured in the cell. This validated our calculations and demonstrated the accuracy with which we measure the energy transfer. Thus, we observed a ligand-regulated direct interaction of an LXXLL peptide with ER in living cells.
|
-RFP-RFP with the same F6 LXXLL peptide, but labeled with GFP instead of CFP. Control measurements similar to those described above for the CFP-LXXLL construct were conducted to determine the bleedthrough of GFP-LXXLL fluorescence into the red and FRET channels. These GFP-LXXLL bleedthrough constants, and bleedthrough constants determined from parallel ER-RFP control cells, were used to calculate the amount of ligand-regulated FRET in cells coexpressing GFP–LXXLL and ER-RFP. Coexpressing cells treated with tamoxifen showed a FRET/donor (FRET/green) ratio of 0.0530 ± 0.0043 (n = 18) (Fig. 5A, white boxes), that was not significantly different from the 0.0541 ± 0.0039 ratio measured in the control cells expressing GFP-LXXLL alone (Fig. 5A, Xs).
|
-RFP and GFP-LXXLL showed a FRET to green ratio of 0.0971 ± 0.0291 (n = 44) that varied proportionally to the RFP/GFP ratio (Fig. 5A, black boxes). This confirmed that E2 promotes an interaction between ER and LXXLL in living cells, whereas tamoxifen does not. When E2-treated cells coexpressing ER-RFP and GFP not containing the 19-amino acid LXXLL peptide were analyzed, the FRET/GFP ratio remained at 0.0549 ± 0.0098 (n = 14). In addition, mutation of a single lysine in the hydrophobic cleft of ER to alanine (K362A) abolished FRET of GFP-LXXLL with ER-RFP in E2-treated cells (Fig. 5A, white circles) as the FRET/GFP ratio remained as 0.0557 ± 0.0063 (n = 27). This demonstrated that the cleft, which is essential for E2-dependent transcription via AF-2 (25), is required for direct interaction of LXXLL with ER in the physiological environment of the living cell.
In the presence of E2, the FRET/green ratio from coexpressed wild-type ER and LXXLL increased with the RFP/GFP ratio in the cells (Fig. 5A), confirming that bona fide FRET was detected. The slopes of these graphs, summarized in Table 1 as the range of slopes encompassing the 95% confidence intervals, emphasized that interaction of the F6 LXXLL peptide with ER was promoted by E2 (slope = 0.017 to 0.021), but not by the SERMs tamoxifen (slope = -0.001 to 0.001) or ICI 182,780 (slope = -0.002 to 0.000). Thus, FRET precisely measured a ligand-specific, AF-2-dependent direct interaction between ER and an LXXLL peptide in living cells.
|
-RFP and another 19-amino acid-long, LXXLL-containing peptide, "D2" (21), fused to GFP. D2 contains sequences flanking the LXXLL motif that differ from those in the F6 peptide.
Control measurements established that the FRET/GFP and RFP/GFP ratios for GFP-LXXLL-D2 alone were no different from those measured for GFP-LXXLL-F6 (not shown). When coexpressed with ER-RFP, GFP-LXXLL-D2, like GFP-LXXLL-F6, showed E2- and AF-2-dependent FRET that was not promoted by tamoxifen (Fig. 5B). In the presence of saturating (10-6 M) E2, the slopes of the FRET/GFP vs. RFP/GFP graphs (reported as 95% confidence intervals) were similar, but slightly different, for GFP-LXXLL-F6 (Table 1, 0.017–0.021) and GFP-LXXLL-D2 (0.022–0.026). This suggested that the D2 and F6 peptides bound to the hydrophobic cleft of E2-bound ER with marginal differences in orientation or with slightly different affinities.
Using FRET to Distinguish SERM-Regulated Interactions
The ability of FRET to measure subtle differences in the direct interactions of the LXXLL motif and ER in living cells could be used as a sensitive new assay for detecting specific activities of new SERMs in vivo. All ligands that trigger LXXLL motif binding to ER are known to strongly activate transcription at promoters containing the classical ERE promoter element. Therefore, a compound that elicits a strong level of FRET between ER-RFP and GFP-LXXLL in this assay might also activate transcription of an ERE-driven gene in a reporter assay.
To test this hypothesis, we synthesized and tested two synthetic ligands of uncharacterized estrogenic or antiestrogenic activities that were reported previously to bind ER with high affinity (38, 39, 40): ethyl indenestrol A and 6,4'-dihydroxyflavone (Fig. 6). These compounds and a known ER agonist diethylstilbestrol were compared with E2 for their abilities to elicit FRET in cells coexpressing ER-RFP and GFP-LXXLL-F6 or GFP-LXXLL-D2. E2, DES, EIA, and DHF all were able to trigger significant levels of FRET between ER-RFP and both LXXLL peptides (Table 1, slopes, italic). All of these ligand-regulated interactions were blocked upon mutation of lysine 362 in ER to alanine (Table 1, ER-RFP K362A), indicating that LXXLL was interacting directly with the hydrophobic pocket of ER in each case.
|
-RFP with GFP-LXXLL-F6, the levels of FRET activated by E2, DES, EIA, and DHF were not statistically different (P > 0.05) (Table 1, GFP-LXXLL-F6 slopes). Similarly, the levels of FRET determined for GFP-LXXLL-D2 interaction with ER-RFP in the presence of DES and EIA were not significantly different than those observed with GFP-LXXLL-F6 (Table 1). In contrast, DHF activated a significantly greater level of FRET with the D2 LXXLL peptide than with the F6 LXXLL peptide (Table 1, boldface). Thus, all compounds promoted the direct interactions of two different LXXLL peptides with ER in the cellular environment, but precise FRET measurements allowed subtle variations in those interactions to be observed.
The similar levels of FRET with the F6 peptide suggested that the ER-binding compounds EIA and DHF both caused ER to adopt a conformation that permitted LXXLL-F6 to bind into the hydrophobic AF-2 cleft in the same orientation as occurs when E2 or DES binds to ER. Because reporter gene assays show that DES and E2 activate transcription from a classical ERE in an AF-2-dependent fashion, the similar AF-2/LXXLL-F6 interactions adopted by the EIA and DHF-bound ERs suggested that these compounds might activate transcription at an ERE site. Reporter gene assays were performed in HeLa cells using transiently transfected wild-type ER and a luciferase gene driven by the classical ERE from the vitellogenin promoter (Fig. 7). The promoter was activated upon expression of unliganded ER. This activation was blocked by the SERM raloxifene, which acts as an antiestrogen for AF-2-dependent transcription (41). Incubation with 10-5 M EIA and 10-5 M DHF both activated transcription from an ERE site as effectively as 10-5 M E2. Thus, the ability of two different LXXLL peptides to productively interact with ER AF-2 in living cells was associated with agonist activity of four separate ligands, each with a distinct chemical structure.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Only recently has FRET been used to measure protein-protein interactions in the environment of living eukaryotic cells (28, 49, 50). FRET has been used to detect nuclear receptor interactions with cofactor fragments, labeled with spectral variants of GFP (49). We show here that LXXLL peptides by themselves are sufficient to interact, in an agonist-regulated fashion, directly with ER in the cellular environment (Table 1). Moreover, these interactions are wholly dependent upon AF-2 in ER. A weak interaction of LXXLL with ER was also detected in the absence of ligand. This ligand-independent interaction was blocked by the same K362A mutation in AF-2 that abrogated agonist-dependent FRET. Thus, in living cells, both the ligand-independent and agonist-dependent interactions of LXXLL with ER are dependent upon AF-2.
Our approach allowed us to precisely quantify the level of FRET between the interacting factors. Because the amount of FRET falls very rapidly, to the sixth power, with the separation of the fluorophores (28, 29, 30), differences in the relative spatial orientations of ER and LXXLL affect the amount of FRET measured in each complex. In our studies, the efficiency of FRET between ER and a given LXXLL peptide was similar for each ligand. However, there were subtle differences, particularly with 6,4'-dihydroxyflavone for which the level of FRET was higher for ER interaction with LXXLL-D2 than with LXXLL-F6. The slight variations in the levels of FRET indicate subtly different ligand-specific interactions. Different levels of FRET may suggest that the LXXLL peptide is bound to AF-2 in a different position such that the GFP and RFP fluorophores are different distances apart. Alternatively, the on- or off-rates for the interaction of LXXLL with ER may be different, leading to quantitatively different levels of FRET. For instance, the higher level of LXXLL FRET with all liganded ERs than with the unliganded ER is consistent with prior observations that agonist binding dramatically stabilizes the LXXLL interaction with ER (31, 51).
The FRET studies described here, by themselves, showed the similarities and differences in LXXLL interactions with ER bound by four different activating ligands and two different AF-2-blocking ligands. Other techniques may complement the FRET studies of ligand-selective nuclear receptor action. Determining the kinetics of fluorescence recovery after photobleaching fluorophore-linked nuclear receptors (46, 48) and cofactors (48) at isolated regions within the nucleus may help to determine if the ligand-selective changes in FRET efficiency arise from altered kinetics in the LXXLL interaction with ER. Measuring the recovery of FRET after selectively photobleaching the interacting cofactor would more precisely correlate fluorescence recovery with direct interactions with ER in the cellular environment.
The benefits of drugs that regulate nuclear receptor activities in some tissues are often counteracted by unwanted receptor actions in other tissues. It is therefore important to identify compounds with desirable selective modulatory properties (12, 13, 20, 21, 22, 41, 52). However, most current assays for interaction are insufficient to distinguish the tissue-selective actions of new compounds from previously existing SERMs (22, 41). The precise measurement of ER/cofactor interactions afforded by FRET will allow the detection, in different cellular environments, of more subtle differences in the interactions of ER, or any other nuclear receptor, bound to different ligands. This will aid the development of clinically effective compounds that regulate specific interactions in specific cell types. Indeed, we found that FRET between ER and two LXXLL-containing peptides was useful in predicting the E2-mimicking activity of two previously untested ER-binding compounds, EIA and DHF. However, different levels of FRET for the LXXLL interactions with AF-2 suggest that these compounds possess somewhat distinct properties. This precise quantification of FRET between nuclear receptors and conformation-specific peptide probes developed by us (20, 21, 22) and others (53) will greatly contribute to a better mechanistic understanding of estrogen action and may be potentially useful for discovering SERMs with improved tissue-selective actions.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
was fused, in frame with an eight-amino acid linker, to the amino terminus of RFP by inserting a PCR-generated ER cDNA into the NheI and BamHI sites of pDsRed1-N1, an RFP expression vector (CLONTECH Laboratories, Inc., Palo Alto, CA). The ER-RFP expression vector was cotransfected with either the 3xERE-TATA-Luc or C3-Luc reporters into HepG2 cells and the transfected cells were treated with E2, tamoxifen, ICI 182,780 or ethanol control vehicle as previously described (21). Cells were then lysed and the amount of luciferase activity in the extracts was measured as previously described (21).
The synthesis, ER binding properties and transcriptional activation profiles of EIA and DHF will be reported elsewhere. Transfection conditions and assay protocols used for the testing of EIA and DHF with wild-type human ER and the ERE reporter gene assay in HeLa cells (Fig. 7) were identical to those reported previously (41, 54).
Cellular Imaging
GHFT1–5 cells were transfected with the ER-RFP expression vector and doxycycline-inducible GFP-LXXLL or CFP-LXXLL expression vectors as previously described (31). Transfected cells were grown for 24 h in estrogen-free media containing 3–5 µg/ml doxycycline. A total of 10-6 M of each ligand was then added and the cells grown for a further 24 h before data collection. Quantitative fluorescence images were collected with a Hamamatsu ORCA cooled interline camera attached to an Olympus Corp. IX-70 microscope controlled by Universal Imaging Corp. (Downingtown, PA) Metamorph software. Filter combinations, described in the Results section, were obtained from Chroma Technology Corp. (Brattelboro, VT). Fluorescence quantification of marked nuclei and background was performed using Metamorph software. Background subtractions and bleedthrough corrections were applied using Microsoft Excel (Microsoft Corp., Redmond, WA).
FRET Analysis
FRET measurements can be accomplished using relatively affordable fluorescence microscopic equipment and image collection software. Control measurements of the cells separately expressing the donor or the acceptor fluorophore (Fig. 2) indicate the point at which the user’s equipment no longer accurately quantifies the fluorescence ratios critical for FRET determination. Only cells containing donor and acceptor fluorescence amounts greater than those that are accurately measured are included for calculating FRET. Because the ratios measured are physical parameters of the fluorophores, FRET measurements are highly consistent between separate experiments provided that all parameters affecting the relative ratios of fluorescence quantification in the donor, acceptor and FRET channels are kept constant. This includes using the same 1) objective lens, 2) relative integration times for the different channels, 3) dichroic mirror, 4) excitation/emission filters, and 5) camera.
Average fluorescence ratios ± SD were calculated from data collected on multiple days using Microsoft Excel. Ninety-five percent confidence intervals in the slopes and Y-intercepts of the FRET/donor vs. acceptor/donor graphs were calculated using GraphPad Software, Inc. Prism software (San Diego, CA). Only data up to an acceptor/donor ratio of 4 were included in the calculation of the slopes for GFP-LXXLL FRET with ER-RFP. The linearity of the graphs tended to decrease beyond this point, as the amount of acceptor became more saturating. This acceptor/donor ratio of 4 should not be used by others as a defined parameter as it depends on the nature of the molecular interaction (our unpublished data) and the ability of the user’s equipment used to quantify the specific acceptor and donor fluorescence.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Current address: Department of Medicinal Chemistry and Molecular Pharmacology, Purdue University, West Lafayette, Indiana 47907. 
Abbreviations: CFP, Cyan fluorescent protein; DES, diethylstilbestrol; DHF, 6,4'-dihydroxyflavone; EIA, indenestrol A; FRET, fluorescence resonance energy transfer; GFP, green fluorescent protein; L, leucine; RFP, red fluorescent protein; SERM, selective ER modulator; X, any amino acid.
Received for publication October 8, 2001. Accepted for publication December 21, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
and ER ß. Proc Natl Acad Sci USA 96:3999–4004 and ß. Mol Cell Biol 19:8226–8239 assembles essential cooperating factors in common subnuclear domains. Mol Endocrinol 1665–1676
and ERß at AP1 sites. Science 277:1508–1510 reveals a role for ligand in the nuclear distribution of the receptor. Mol Biol Cell 10:471–486-coactivator complexes in living cells. Mol Cell Biol 21:4404–4412 and ERß) at AP-1 sites. Chem Biol 8:427–436[CrossRef][Medline]This article has been cited by other articles:
 |
|
 |
|
  E. M. Kofoed, M. Guerbadot, and F. Schaufele Structure, Affinity, and Availability of Estrogen Receptor Complexes in the Cellular Environment J. Biol. Chem., January 22, 2010; 285(4): 2428 - 2437. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Eeckhoute, R. Metivier, and G. Salbert Defining specificity of transcription factor regulatory activities J. Cell Sci., November 15, 2009; 122(22): 4027 - 4034. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Nakamura, V. M. Nemani, E. K. Wallender, K. Kaehlcke, M. Ott, and R. H. Edwards Optical Reporters for the Conformation of {alpha}-Synuclein Reveal a Specific Interaction with Mitochondria J. Neurosci., November 19, 2008; 28(47): 12305 - 12317. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. M. Fernandez, M. C. Lewis, A. S. Pechenino, L. L. Harburger, P. T. Orr, J. E. Gresack, G. E. Schafe, and K. M. Frick Estradiol-Induced Enhancement of Object Memory Consolidation Involves Hippocampal Extracellular Signal-Regulated Kinase Activation and Membrane-Bound Estrogen Receptors J. Neurosci., August 27, 2008; 28(35): 8660 - 8667. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Sanchez, K. Sauve, N. Picard, and A. Tremblay The Hormonal Response of Estrogen Receptor beta Is Decreased by the Phosphatidylinositol 3-Kinase/Akt Pathway via a Phosphorylation-dependent Release of CREB-binding Protein J. Biol. Chem., February 16, 2007; 282(7): 4830 - 4840. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Padron, L. Li, E. M. Kofoed, and F. Schaufele Ligand-Selective Interdomain Conformations of Estrogen Receptor-{alpha} Mol. Endocrinol., January 1, 2007; 21(1): 49 - 61. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Paulmurugan and S. S. Gambhir An intramolecular folding sensor for imaging estrogen receptor-ligand interactions PNAS, October 24, 2006; 103(43): 15883 - 15888. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Khan, R. Barhoumi, R. Burghardt, S. Liu, K. Kim, and S. Safe Molecular Mechanism of Inhibitory Aryl Hydrocarbon Receptor--Estrogen Receptor/Sp1 Cross Talk in Breast Cancer Cells Mol. Endocrinol., September 1, 2006; 20(9): 2199 - 2214. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Schaufele, X. Carbonell, M. Guerbadot, S. Borngraeber, M. S. Chapman, A. A. K. Ma, J. N. Miner, and M. I. Diamond The structural basis of androgen receptor activation: Intramolecular and intermolecular amino-carboxy interactions PNAS, July 12, 2005; 102(28): 9802 - 9807. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. N. Day and F. Schaufele Imaging Molecular Interactions in Living Cells Mol. Endocrinol., July 1, 2005; 19(7): 1675 - 1686. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Kim, R. Barhoumi, R. Burghardt, and S. Safe Analysis of Estrogen Receptor {alpha}-Sp1 Interactions in Breast Cancer Cells by Fluorescence Resonance Energy Transfer Mol. Endocrinol., April 1, 2005; 19(4): 843 - 854. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. S Lewis, T J Thomas, R. G Pestell, C. Albanese, M. A Gallo, and T. Thomas Differential effects of 16{alpha}-hydroxyestrone and 2-methoxyestradiol on cyclin D1 involving the transcription factor ATF-2 in MCF-7 breast cancer cells J. Mol. Endocrinol., February 1, 2005; 34(1): 91 - 105. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. E. Ross, H. S. Radomska, B. Wu, P. Zhang, J. N. Winnay, L. Bajnok, W. S. Wright, F. Schaufele, D. G. Tenen, and O. A. MacDougald Phosphorylation of C/EBP{alpha} Inhibits Granulopoiesis Mol. Cell. Biol., January 15, 2004; 24(2): 675 - 686. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Tanaka, M. Nishi, M. Morimoto, T. Sugimoto, and M. Kawata Yellow Fluorescent Protein-Tagged and Cyan Fluorescent Protein-Tagged Imaging Analysis of Glucocorticoid Receptor and Importins in Single Living Cells Endocrinology, September 1, 2003; 144(9): 4070 - 4079. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Bai and V. Giguere Isoform-Selective Interactions between Estrogen Receptors and Steroid Receptor Coactivators Promoted by Estradiol and ErbB-2 Signaling in Living Cells Mol. Endocrinol., April 1, 2003; 17(4): 589 - 599. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Schaufele, X. Wang, X. Liu, and R. N. Day Conformation of CCAAT/Enhancer-binding Protein alpha Dimers Varies with Intranuclear Location in Living Cells J. Biol. Chem., March 14, 2003; 278(12): 10578 - 10587. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. N. Day, T. C. Voss, J. F. Enwright III, C. F. Booker, A. Periasamy, and F. Schaufele Imaging the Localized Protein Interactions Between Pit-1 and the CCAAT/Enhancer Binding Protein {alpha} in the Living Pituitary Cell Nucleus Mol. Endocrinol., March 1, 2003; 17(3): 333 - 345. [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 |
