Development of an HTS Compatible Assay for Discovery of RORα Modulators using AlphaScreen® Technology

Abstract

The retinoid acid receptor-related orphan receptors (RORs) represent important targets for treatment of metabolic and immune disorders. Here we describe the application of AlphaScreen® technology to develop an HTS compatible assay to facilitate the discovery of RORα modulators. Using the ligand binding domain (LBD) of RORα and a peptide derived from the NR1 box of the nuclear receptor coactivator PGC-1α, a 384-well format assay was developed exhibiting high sensitivity, requiring only low nanomolar concentration of reagents. Recently it was shown that oxysterols such as 7α-hydroxycholesterol (7α-OHC) function as modulators of the RORs. In this assay 7α-OHC produced a dose-dependent response with an EC₅₀ of 162 nM, Z’ factor of 0.6 and a S/B ratio of 4.2 demonstrating that the assay is HTS compatible. Validation of the assay was afforded by screening against the Sigma LOPAC1280™ library in 384-well format. In summary, the results presented here demonstrate that this assay can be used to screen large chemical libraries to discover novel modulators of RORα.

INTRODUCTION

The retinoid-related orphan receptors (RORs) are members of the nuclear receptor (NR) superfamily. Three ROR genes have been identified and each generates several isoforms as a result of alternative promoter usage and exon splicing. In humans, four distinct RORα (NR1F1) isoforms have been identified (RORα 1–4) and they are predominantly expressed in blood, brain, skeletal muscle, and fat cells ^1–2. The RORβ (NR1F2) and RORγ (NR1F3) genes each produce two different isoforms ^3–4. RORβ isoforms are expressed specifically in the brain, while the RORγ isoforms are highly expressed in the thymus, although significant expression has been also reported in the liver, skeletal muscle, adipose tissue and kidney⁵.

RORs have been suggested to function as constitutive transcriptional regulators in the absence of ligands by interacting with corepressors as well as coactivators^6–7. Reports have suggested that cholesterol, 7-hydroxycholesterol (7α-OHC), and cholesterol sulfate might be ligands for RORα ^8–9. More recently, it was shown that 7α-OHC binds to and modulates the transcriptional output of RORα¹⁰. In this study, 7α-OHC was able to modulate the presence of the p160 family coactivator SRC2 at a RORα target gene promoter (G6Pase). Thus, it is clear that ROR ligands can modulate recruitment of coregulators. The interaction between NRs and their coactivators typically occurs between a region of the ligand-binding domain (LBD) referred to as activation function 2 (AF-2) and regions on the coactivator/corepressor termed nuclear receptor boxes (NR boxes) which contain a hydrophobic LXXLL motif important for mediating this interaction¹¹.

RORs have emerged as important drug targets for the treatment of a variety of medical conditions, such as metabolic disorders, atherosclerosis, and certain cancers. The goal of this study was to develop a robust and sensitive high-throughput screening (HTS) biochemical assay for discovery of RORα modulators. It has been suggested that the interaction between RORα and PGC-1α (peroxisome proliferator-activated receptor coactivator-1α) serves as an important link between the circadian clock and various metabolic pathways and may play a pivotal role in lipid homeostasis in muscle ^12–13. We took advantage of this apparent constitutive interaction to design an assay that detects conformational change induced by ligand binding to the LBD of RORα that would perturb interaction with the LKKLL motif present in NR box 1 of PGC-1α. It is important to note that this type of NR-LBD/NR-box peptide interaction assay with purified proteins, while very useful to detect compound binding, is artificial and not necessarily directly reflective of the compounds pharmacology in the context of a native transcription complex. NR co-activators (CoA) and co-repressors (CoR) are large multi-domain proteins that interact with nuclear receptors via a minimal domain referred to as the receptor interaction domain or RID. Within the RID are several nuclear receptor boxes (NR boxes) each of which contains a classical hydrophobic LXXLL motif. These NR boxes facilitate cofactor interaction with NRs but by themselves have no function and do not mediate transcription. Peptides that contain one NR box will interact with hydrophobic clefts on the NR’s surface often in a ligand-dependent fashion for NRs that are not constitutively active, whereas in case of constitutively active RORα, the NR box peptide derived from the PGC-1α RID interacts with the receptor even in the absence of ligand. The ability to alter the level of interaction of the apo receptor with the PGC-1α NR box peptide requires ligand to bind in a fashion such to alter the conformation of the LBD. Thus, this the assay is useful for identifying compounds that bind to RORα but does not directly provide insight into the cellular pharmacology of the ligand. This is also true for a radioligand displacement binding assay so there is no one single assay for nuclear receptors that effectively captures both binding potency and ligand pharmacology.

Discovery of novel modulators of NRs has been enhanced by using the AlphaScreen® (Amplified Luminescent Proximity Homogeneous Assay) technology. This has become one of the methods of choice for HTS, having the best sensitivity and dynamic range over alternative technologies^14–15.

Here we demonstrate an HTS compatible AlphaScreen® assay to facilitate the identification of small molecule modulators of RORα.

MATERIALS AND METHODS

Reagents

7α-Hydroxycholesterol was purchased from Steraloids. Sodium chloride (NaCl) was obtained from Fisher Chemicals. Hepes and BSA were purchased from Sigma Aldrich. The AlphaScreen® Histidine (Nickel chelate) detection kit containing Ni-chelate acceptor beads and streptavidin donor beads was purchased from PerkinElmer. The AlphaScreen® TruHits kit containing streptavidin-donor beads and biotinylated acceptor beads coated with BSA was from PerkinElmer. N-terminal biotinylated peptides PGC-1α NR1 and NR2 (SMPDGTPPPQE-AEEPSLLKKLLLAPANTQL and AKSICQQQKPQRRPCSELLKYLTTNDDPPH) and the recombinant His-RORα-LBD protein were provided by Eli Lilly and Company.

The sequence of the His-RORα-LBD (262–523) protein is: MGHHHHHHGETSPTVSMAELEHLAQNISKSHLETCQYLREELQQITWQTFLQEEIENYQNKQREVMWQLCAIKITEAIQYVVEFAKRIDGFMELCQNDQIVLLKAGSLEVVFIRMCRAFDSQNNTVYFDGKYASPDVFKSLGCEDFISFVFEFGKSLCSMHLTEDEIALFSAFVLMSADRSWLQEKVKIEKLQQKIQLALQHVLQKNHREDGILTKLICKVSTLRALCGRHTEKLMAFKAIYPDIVRLHFPPLYKELFTSEFEPAMQIDG. Baculovirus expression system was used to express the His-RORα-LBD protein. This was purified using Ni column and by size exclusion chromatography method to a 95% purity.

The Library of Pharmacologically Active Compounds (Sigma LOPAC1280™) was obtained from Sigma Aldrich and was reformatted at 1 mM concentration in DMSO into 384 well format sources plates obtained from Greiner Bio-One.

Cell culture and transcriptional assays

Luciferase reporter assays were conducted using a pBind Gal4-tagged RORα LBD construct along with a UAS luciferase reporter in the presence or absence of full length PGC-1α cotransfected into HEK293T cells. Reverse transfections were performed in bulk using 1x10⁶ cells in 6 cm plates, 3 µg of total DNA in a 1:1:1 ratio of receptor, reporter and coactivator respectively, and FuGene6 (Roche) in a 1:3 DNA:lipid ratio. Following 24h incubation, cells were counted and plated in 384 well plates at a density of 10,000 cells/well. Cells were treated with vehicle or 10 µM 7α-hydroxycholesterol 4h after plating. Following additional 20h incubation, luciferase levels were assayed by one-step addition of 30 µL BriteLite (Perkin Elmer) and read using an Envision multilabel plate reader (Perkin Elmer).

Radioligand Receptor Binding Assay

Ninety nanograms of purified GST-RORα-LBD were incubated with 3nM of [³H]-25-hydroxycholesterol in assay buffer (50 mM HEPES, pH 7.4, 0.01% bovine serum albumin, 150 mM NaCl and 5 mM MgCl₂) as described previously¹⁰. In brief, for the competition assay, various concentration of compounds were incubated with receptor in the presence of [³H]-25-hydroxycholesterol for 1 hr. The assays were terminated by rapid filtration through pre-soaked Whatman GF/B filters (0.5% PEI in PBS) in Multiscreen plates (Millipore) and were washed with ice-cold assay buffer (3×0.1 ml). The data is presented as specific counts in presence of various concentrations of the above compounds.

AlphaScreen® assay development and characterization in 384-well format

All experiments were performed in white opaque 384-well plates (Greiner Bio-One) under green light conditions (<100 lux) at room temperature. All dilutions were made in 1X assay buffer containing 25 mM Hepes (pH 7.4), 100 mM NaCl, and 0.1% bovine serum albumin (BSA) was added to minimize non-specific interactions. The final DMSO concentration was 0.25% for compound dose-response curves. The luminescence was read on PerkinElmer Envision 2104.

The final assay volume was 15 µL for the peptide-protein cross-titration experiments and 20 µL for kinetics experiments. For the optimization of the constitutive interaction between His-RORα-LBD and Biotin-PGC-1α NR1, cross-titrations of protein and peptide, respectively were performed. For titrating the Biotin-PGC-1α NR1 at 25, 50, 75, and 100 nM, respectively of His-RORα-LBD, a mix of 12 µL of receptor and beads (12.5 µg/mL each) was added to each well followed by addition of 3 µL of peptide (8.4 pM-500 nM). For titrating the His-RORα-LBD (8.4 pM-166 nM) against 70 nM of Biotin-PGC-1α NR1 or Biotin-PGC-1α NR2, a mix of 12 µL of protein corresponding to each dilution and beads at 12.5 µg/mL was added to each well followed by addition of 3 µL of peptide. For kinetics experiments 75 nM His-RORα-LBD and 30 nM Biotin-PGC-1α NR1 were incubated for 0 to 6h at room temperature with 30 minutes interval and 13 time points were generated. Two independent experiments were performed with each time point representing a separate experiment to avoid photo-bleaching. Plates were sealed and incubated for 1h.

For compound dose-response curves (425 pM-25 µM), a mix of 12 µL of 75 nM of His-RORα-LBD and beads (range of 15–40 µg/mL for each bead depending on the experiment as indicated) was added to the wells and 4 µL of increasing concentration of compound and incubated for 1h. After this pre-incubation step, 4 µL of 30 nM of Biotin-PGC-1α NR1 were added and further incubated for 2h. The final assay volume was 20 µL. To investigate the effect of solvent DMSO on the compound EC50 value, four different DMSO final concentrations were tested (0.25%, 0.5%, 1%, and 5%).

AlphaScreen® TruHits assay

Biotin and steptavidin bind strongly which brings the donor and acceptor beads into close proximity generating a signal upon excitation at 680 nm. Any compound which causes a decrease in the signal represents an interfering compound which is not specific for the target of interest, whereas those which have no effect on the signal are potential true hits. The final assay volume was 20.5 µL. The assay buffer, plate format, light conditions and luminescence reading instrument are described in the previous section. A mix of streptavidin donor (SA-D) beads and biotinylated acceptor beads (b-A) each at a final concentration of 1.5 µg/mL in a volume of 20 uL were added to each well. For generating the compound dose-response curves, 0.5 µL of increasing concentrations of 7α-hydroxycholesterol compound (0.3 nM - 25 µM) or LOPAC hits (20 nM – 50 µM) were added, the plates were sealed and incubated for 1h at room temperature. The final DMSO concentration was 0.25%.

HTS LOPAC assay

Assay plates were filled with 12 µL per well of 1X assay buffer (described above) using a BioRAPTR FRD (Beckman Coulter, Inc.). First, the compound or controls (100 nL) in DMSO were added via the pintool instrument (Kalypsys, GNF) to have 0.5% nominal percentage of DMSO at 5 µM screening compound concentration when using a 1 mM source plate. Each compound was transferred as a single instance per well but was pinned across three separate assays plates such that they were analyzed in triplicate. All reagent preparation and additions, as well as the experiment, were kept under green light conditions (<100 lux) from this point onward. Next, a mix of 4 µL of 5X His-RORα-LBD protein (375 nM) and 5X beads (50 µg/mL each of Ni-chelate acceptor beads and streptavidin donor beads) in assay buffer was dispensed into wells followed by incubation at room temperature for 1h. To start the reaction, 4 µL per well of Biotin-PGC-1α NR1 in assay buffer was added at 150 nM (5X) and the reaction was allowed to proceed for 2h at room temperature. The final assay volume was 20 µL. Plates were then read using the Envision plate reader from PerkinElmer Life Sciences using the HTS AlphaScreen modality reading at a height of 0 mm, exciting for 0.04 seconds with an emission time of 0.09 seconds per well with cross talk corrections applied.

The compounds tested in the assay validation stage were from Sigma LOPAC1280™ library containing 1280 proven pharmacologically-active compounds. Detailed information about the library composition is available at: http://www.sigmaaldrich.com/chemistry/drug-discovery/validation-libraries/lopac1280-navigator.html.

All data were normalized on a per plate basis to their individual controls. Since each compound was represented three times we also calculated its average percent activity. The percent activation for each compound was calculated as follows in the Eq. below, where Test Compound is defined as wells containing His-RORα-LBD, Ni-chelate acceptor beads and streptavidin donor beads, and Biotin-PGC-1α NR1 in the presence of test compound. Positive Control is defined as wells containing the same except 7α-OHC is used at 25 µM in place of compounds. Low Control wells again are the same but contain DMSO:

• Eq. % Activation = 100 * (1-((Test Compound – Median Positive Control) / (Median Low Control – Median Positive Control)).

The AlphaScreen® compound response curves (CRC) assay was run under identical conditions to the HTS version in triplicate using 10 mM source starting concentration and 3-fold serial dilutions for 10 points. Final assay compound concentrations are thus 50 µM at the highest dilution progressing downward accordingly.

Assay performance

To monitor assay sensitivity and evaluate the robustness of this assay, signal to background (S/B) ratios and the Z’ factor were calculated using the equations previously described¹⁶.

Data analyses

All EC50 values were calculated by non-linear regression analysis (sigmoidal dose-response-variable slope) using Prism® software (Graphpad Software, Inc., San Diego, CA). Data analysis for the LOPAC1280™ compound library was performed using internal informatics tools and using Symyx Assay Explorer version 3.2 (Symyx Technologies, Inc.). For CRC analysis, for each test compound, percent activation was plotted against compound concentration. A four parameter equation describing a sigmoidal dose-response curve was then fitted with adjustable baseline using Assay Explorer software (Symyx Technologies Inc). The reported EC50 values were generated from fitted curves by solving for the X-intercept value at the 50% inhibition level of the Y-intercept value. In cases where the highest concentration tested (i.e. 50 µM) did not result in greater than 50% inhibition, the EC50 was determined manually as greater than 50 µM.

RESULTS AND DISCUSSION

Using a cell-based co-transfection assay we found that PGC-1α was able to modulate RORα transactivation by 2.5-fold (Figure 1). This result is in agreement with published data suggesting that PGC-1α may function as a coactivator for RORα^{6–7, 12–13}. Moreover, 7α-OHC was able to decrease the transactivation signal by RORα in the absence and presence of PGC-1α by 2- and 1.5-fold, respectively. These data confirm the recent finding that 7α-OHC is able to modulate conformational change of RORα as an inverse agonist¹⁰. PGC-1α contains two distinct NR boxes, NR1 and NR2, which are both essential for the interaction with RORα. In order to identify the optimal binding partner for RORα in the context of an AlphaScreen® assay, we investigated the affinity of RORα to biotinylated peptides representative of NR1 and NR2 of PGC-1α . Initial experiments demonstrated that the PGC-1α NR1 peptide was recruited 5-fold more efficiently to RORα as compared to PGC-1α NR2 peptide (data not shown). Therefore, the Biotin-PGC-1α NR1 peptide was chosen for further assay development as described below.

Figure 1. RORα is activated by PGC-1α and 7α-hydroxycholesterol modulates RORα interaction with PGC-1α.

293T cells were cotransfected with UAS-luciferase and Gal4-RORα in the absence or presence of PGC-1α as mentioned in Material and Methods. The cells were treated with 10µM of 7α-hydroxycholesterol or DMSO (vehicle) as control for 20 hr followed by luciferase activity measurement. Relative change was determined by normalizing to the cells treated with vehicle only. Each data point was performed in 6 replicates and represented as mean ±SEM.

Development of the 384-well format AlphaScreen® assay

Both binding partners were titrated in order to find the best reagent concentration and to determine the sensitivity of the assay. First, the Biotin-PGC-1α NR1 peptide was titrated against four concentrations of His-RORα-LBD. As illustrated in Figure 2A, the highest S/B was achieved with 75 nM His-RORα-LBD protein. The S/B decreased with 100 nM receptor concentration suggesting perhaps that the available binding sites on the acceptor beads had become saturated with the protein. The EC₅₀ values for Biotin-PGC-1α were similar in all cases while the S/B and Z’ factor varied depending on the His-RORα-LBD concentration as shown in the table attached to Figure 2A. While holding the peptide at the EC₉₀ concentration (70 nM) corresponding to the EC50 value using 75 nM His-RORα-LBD, the receptor was titrated and the results are shown in Figure 2B. The EC₅₀ value for receptor was determined to be 36 nM and the S/B and Z’ factor were 162.61 and 0.73, respectively. These results demonstrated that it was possible to build an assay that was robust and sensitive. Based on these results, further assay optimization was performed using final concentrations of receptor and peptide of 75 nM and 30 nM, respectively.

Figure 2. RORα-PGC-1α cross-titrations and validation of His-RORα-LBD protein conformational change.

A) Increasing concentrations of Biotin-PGC-1α NR1 peptide (8.4 pM-500 nM) were incubated with a constant amount of 25, 50, 75, and 100 nM, respectively of His-RORα-LBD as described in Material and Methods and table shows Hill Slope, EC50 (M), S/B, and Z’ for the indicated His-RORα-LBD concentrations. B) Increasing concentrations of His-RORα-LBD (8.4 pM-166 nM) at a constant amount of 70 nM of Biotin-PGC-1α NR1 (EC90 value) in a 15 µL assay volume. The final concentration of each bead was 12.5 µg/mL. Plates were incubated for 1h at room temperature. Each data point is represented as the mean ± S.D. of 3 independent experiments (N=3). C) Increasing concentrations of compound 7α-hydroxycholesterol (425 pM-25 µM) were pre-incubated with His-RORα-LBD (75 nM) and both acceptor and donor beads for 1h at room temperature. The peptide Biotin-PGC-1α (30 nM) was added to each well and further incubated for 2h at room temperature. The final concentration of each bead was 30 µg/mL in 20 µL final volume. Data are represented as the mean ± S.D. of 2 independent experiments (N=7). The EC50 values were calculated by non-linear regression analysis (sigmoidal dose-response-variable slope) using GraphPad Prism.

Next the kinetics of the interaction was assessed to determine the optimal timeframe for the assay (Supplemental Figure 1). Using protein and peptide concentration described above in Materials and Methods section, a linear increase in assay signal was detected over the first 2 h of incubation. This optimal reaction timeframe is similar with results obtained for an AlphaScreen®assay for the NR retinoic acid receptor gamma (RARγ)¹⁴. The assay signal decreased over the remaining four hours indicating that the optimal timeframe for the assay is around two hours post mixing. Such a signal decrease is similar to that observed in a recent report describing an AlphaScreeen® assay for a TNF receptor, where the signal decreased after 60 min of incubation¹⁷. Subsequent assay validation was performed reading AlphaScreen® signal 2h post mixing of all reagents.

Compound modulation of RORα/PGC-1α interaction

Following optimization of assay conditions, experiments were performed to determine if a small molecule ligand can modulate RORα/PGC-1α interaction. To this end, 7α-OHC was titrated and the effect of different bead concentrations (15, 20, 30, 40 µg/mL) on EC₅₀ and S/B was also investigated to identify the optimal bead concentration (data not shown). In all cases the EC₅₀ of the test compound was similar. However, not surprisingly, the S/B decreased with decreasing bead concentration and a similar trend was previously reported¹⁸. Optimal S/B was obtained using either 30 or 40 µg/mL of AlphaScreen® beads, respectively. Results for the assay using 30 µg/mL of each bead are shown in Figure 2C. Under these assay conditions the EC₅₀ was determined to be 162 nM and the S/B and Z’ factor were 4.2 and 0.6, respectively. TruHits assay performed with various concentrations of 7α-OHC as indicated in Materials and Methods section confirmed that the effect of the compound is due to the specific binding to RORα-LBD (data not shown).

HTS compatibility

The experiments described in this section were performed to characterize the assay under various conditions to support transfer to the HTS laboratory. First, the effect of DMSO concentration on assay performance was determined. The potency of 7α-OHC was determined at increasing DMSO concentrations ranging from 0.25% to 5% (data not shown). No significant difference in the EC₅₀ value or the S/B ratio was observed indicating that the assay can tolerate up to 5% DMSO. The performance of the assay was further monitored over the course of three days and the data are shown in Figure 3A. Again, there were no significant differences in the performance of the assay run on different days. The S/B ratios for day 1, 2, and 3 were 6.6, 6.2, and 6.9, respectively. The Z’ factors for day 1, 2, and 3 were 0.72, 0.73, and 0.58, respectively. Finally, reagent stability experiments were conducted in order to generate a convenient protocol and define the tolerance of the assay to potential delays encountered during the HTS process. Figure 3B shows that the storage of the master mix consisting of the protein and both bead types over night at 4 °C was not favorable. This behavior was expected based on the decreased signal observed already after 6h of incubation of assay reagents (Supplemental Figure 1). Therefore, delays were avoided during the HTS experiments and the total incubation time of the experiments described below was 3h.

Figure 3. Assays for HTS transfer.

Increasing concentrations of compound 7α-hydroxycholesterol (425 pM-25 µM) were pre-incubated with His-RORα-LBD (75 nM) and both beads for 1h at room temperature. The peptide Biotin-PGC-1α (30 nM) was added to each well and further incubated for 2h at room temperature. A) The experiment was performed on 3 separate days. B) The dose-response curve labeled t0 represents the results after initial pre-incubation of the protein with both beads, compound, and peptide. The dose-response curve labeled ∆16h represents the results of the assay performed after the master mix containing protein and beads was stored in the dark at 4 °C for 16 h. The compound and peptide dilutions dilutions were also stored in the dark at 4 °C for 16 h. The EC50 values were calculated by non-linear regression analysis (sigmoidal dose-response-variable slope) using GraphPad Prism. Data are represented as the mean ± S.D. of 2 independent experiments (N=7).

LOPAC screen using AlphaScreen® assay

To determine the performance of the assay in terms of robustness (Z’) under HTS conditions, the assay was screened against the LOPAC library using modified conditions specific to the HTS platform used as described in the Material and Methods section. Under these modified conditions the EC₅₀ for 7α-OHC was consistent with that determined during assay optimization (data not shown). Briefly, compounds were analyzed in triplicate at nominal concentration of 5 µM using 7α-OHC as the positive control (25µM). Each plate contained positive control wells and low control wells (DMSO only), both of which were used in Z’ factor calculations¹⁶. By utilizing the sample field as the low control in conjunction with the positive control we calculated the Z-factor¹⁶ which when compared to Z’ allows for determination of the inter- and intra-plate variability (Figure 4). The assay Z’ factor remained greater the 0.5 for the entire LOPAC run indicating it to be robust and worthy of further analysis of active compounds. The Z-factor was generally reproducible within the triplicate test plates and indicates limited inter plate variability.

Figure 4. Z’ and Z-factor results from the HTS LOPAC screen.

One data point represents one plate of data. Calculations are described in the text. Both Z’ and Z are shown for each assay plate and are displayed in exact order as the run sequence. Z’, shown in red is consistently >0.5 whilst the Z-factor as depicted in blue indicates interplate variability as expected when numerous hits are identified within the plate. Intraplate variability is consistent as shown by looking at the pairing of plates 1–3, 4–6, 7–9 and 10–12 which is expected. The entire run statistics are also tabulated and include % CV analysis for the 320 sample wells, S/B as calculated by dividing the positive control by the low control for each plate, and average and standard deviation.

Hits were identified in the LOPAC HTS effort by applying a mathematical algorithm in which two values are calculated: (1) the average activity of the low control, and (2) three times their standard deviation. The sum of these two values was used as a cutoff parameter, i.e. any compound that exhibited greater average % activity than the cutoff parameter was declared active. In this case, the cutoff value equated to 19.61% which yielded 30 unique compounds with an average activity above this threshold (Figure 5). To further assess the reproducibility of the hits, we looked at the number of agreements or disagreements of the triplicates among the hits in relation to the hit cutoff. The results were exceptional in that only 4 of 30 compounds had a point in disagreement with activity falling below the cutoff. Of these 4 compounds, the highest average activity was 23.54 ±10.52%. The lowest single point activity scored for a hit compound was 15.54% indicating that, of the 4 compounds with a point in disagreement, all were very close to the hit cutoff to begin with. Since this result was not unexpected, these hits were then “cherry picked” and used for subsequent analysis.

Figure 5. Scatter plot analysis results from the HTS LOPAC screen.

All data from 12 assay plates including controls is displayed. The nominal cutoff for activators of this assay is shown by the dashed line. The positive control is shown as red circles, compounds in black and low control in green. Active compounds were identified based on their average result being greater than the nominal cutoff threshold value of 19.61%. In this display, more than 30 test wells have activity exceeding the threshold value which reflects the reproducibility of the compounds being tested in triplicate. Note this assays identifies negative modulators of RORα and PGC-1α interaction which are displayed as negative % activation.

In order to identify RORα target-independent positives, we employed the AlphaScreen® TruHits assay for all 30 hit compounds. No interfering compounds were identified in this assay demonstrating that these are specific RORα modulators (data not shown). Concentration response curves (CRCs) analysis showed that 23 of these compounds displayed dose-dependent and sigmoidal response (Figure 6A and Supplemental Figure 2). Only 4 of all hit compounds have EC₅₀ values in the nanomolar range, whereas the rest of them displayed values in the micromolar range (Table 1 and Supplemental Table 1). We then focused on further characterization of the 4 most potent compounds. Thus, to confirm whether the activity of these compounds in the AlphaScreen® is due to direct binding to RORα we carried out competitive radioligand receptor binding assays. We demonstrate that all these 4 compounds dose-dependently competed with [³H]-25-hydroxycholesterol for RORα, thus confirming the true binding activity of these compounds to the protein (Figure 6B).

Figure 6. LOPAC hits concentration response curves (CRCs) and radioligand receptor binding assay.

A) The compound hits were titrated starting from 50 µM, with 3-fold serial dilution steps for 10 points. Each point on a curve is displaying the average of three results along with error bars. B) Radioligand competition assays were done using purified GST-RORα-LBD as protein and 3nM of [³H]-25-hydroxycholesterol as radioligand in the assay buffer as mentioned in Materials and Methods. The competition assay was performed for these compounds at 8 different concentrations starting from 10µM and 3-fold subsequent dilutions to 3nM. Data shown are representative results from the experiment performed in triplicates.

Table 1. LOPAC hits CRC.

The most active compounds are displayed including unique identifiers (Sigma ID) which are directly purchasable from Sigma Aldrich. All EC50 values shown are in molar (M) concentration. Averaged Max % response and SD max % response are derived from triplicate data for a concentration in which the values are achieving the greatest agonist activity in this assay. The Hill Slope and the description of the biological function of each compound are shown.

Sigma ID	Final EC50 (M)	Averaged Max % Response	SD Max % Response	Hill Slope	Description
C 1112	223.5E-9	107.68	8.12	1.73	Cannabinoid receptor agonist
P 8887	330.7E-9	148.76	4.26	0.28	Positive allosteric modulator of GABA-A chloride channels
U 5882	393.4E-9	109.46	1.84	1.79	Free radical lipid peroxidation inhibitor
P 0130	832.9E-9	121.57	6.39	1.07	Suppresses ovulation; induces maturation and secretory activity of the uterine endothelium

CONCLUSION

We have established an HTS compatible assay for identification of small molecule compounds that modulate the interaction between RORα and PGC-1α using AlphaScreen® technology. We demonstrate that the assay is robust and sensitive with Z’ factor values over 0.5. By screening this assay against the 1280 pharmacologically active compounds in the Sigma LOPAC1280™ library, we demonstrate the HTS readiness of the assay. In summary, the results presented highlight the suitability of this assay for larger HTS campaigns.