Single enantiomer of YK-4-279 demonstrates specificity in targeting the oncogene EWS-FLI1.

Oncogenic fusion proteins, such as EWS-FLI1, are excellent therapeutic targets as they are only located within the tumor. However, there are currently no agents targeted toward transcription factors, which are often considered to be 'undruggable.' A considerable body of evidence is accruing that refutes this claim based upon the intrinsic disorder of transcription factors. Our previous studies show that RNA Helicase A (RHA) enhances the oncogenesis of EWS-FLI1, a putative intrinsically disordered protein. Interruption of this protein-protein complex by small molecule inhibitors validates this interaction as a unique therapeutic target. Single enantiomer activity from a chiral compound has been recognized as strong evidence for specificity in a small molecule-protein interaction. Our compound, YK-4-279, has a chiral center and can be separated into two enantiomers by chiral HPLC. We show that there is a significant difference in activity between the two enantiomers. (S)-YK-4-279 is able to disrupt binding between EWS-FLI1 and RHA in an immunoprecipitation assay and blocks the transcriptional activity of EWS-FLI1, while (R)-YK-4-279 cannot. Enantiospecific effects are also established in cytotoxicity assays and caspase assays, where up to a log-fold difference is seen between (S)-YK-4-279 and the racemic YK-4-279. Our findings indicate that only one enantiomer of our small molecule is able to specifically target a protein-protein interaction. This work is significant for its identification of a single enantiomer effect upon a protein interaction suggesting that small molecule targeting of intrinsically disordered proteins can be specific. Furthermore, proving YK-4-279 has only one functional enantiomer will be helpful in moving this compound towards clinical trials.


INTRODUCTION
Transcription factors are often considered to be 'undruggable,' despite being promising therapeutic targets. Transcription factors are the driving force in many cancers and are the "bottleneck" of many signaling cascades [1]. Many oncogenic transcription factors result from chromosomal translocations that are unique anticancer targets. ETS family translocations occur in several cancers, including prostate, malignant melanoma of soft parts, myxoid liposarcomas, desmoplastic small round cell tumors, and Ewing's sarcoma family of tumors (ESFT) www.impactjournals.com/oncotarget [2]. Ninety-five percent of ESFT cases contain a balanced t (11;22) or t(21;22) rearrangement, combining the aminoterminus of EWS to the carboxyl-terminus of FLI1 or ERG, both of which contain the highly conserved ets DNA binding domain [3]. Currently, there are no clinically available targeted agents that inhibit these unique tumorspecific proteins.
Unlike targeting an enzyme at the ATP binding site, development of a therapeutic target for a transcription factor requires very specific disruption of a DNA-protein or protein-protein interaction [4]. EWS-FLI1 is predicted to be an intrinsically disordered protein (IDP), which is a protein lacking stable secondary or tertiary structures under physiological conditions [5]. IDPs often have a great potential for binding to small molecules due to higher induced-fit sampling properties and have the potential for multiple binding sites to small molecules [6]. IDPs have already been targeted for drug discovery, such as the kinase and phosphorylation sites located within areas of intrinsic disorder [7]. The c-Myc oncoprotein can be inhibited by small molecules that bind to the disordered region of c-Myc [8,9]. EWS-FLI1 requires disorder for maximal transactivation of transcription [10] and the disordered nature of the transcription factor facilitates the protein-protein complexes that lead to oncogenesis [11].
Oncogenesis of EWS-FLI1 requires protein partnering with RNA Helicase A (RHA), which is necessary to enhance the transformation of EWS-FLI1 [12]. The purification of recombinant EWS-FLI1 [13] allowed for the screening of a library of small molecules with surface plasmon resonance to identify compounds with direct binding [14]. The small molecule lead compound and its derivative, YK-4-279, bind to EWS-FLI1 and are able to disrupt the EWS-FLI1/RHA interaction. Treatment with YK-4-279 specifically inhibits EWS-FLI1 function both in vivo and in vitro. Our small molecule, YK-4-279, is the first molecule to directly target EWS-FLI1.
YK-4-279 contains a chiral center. Chiral discrimination between enantiomers is extremely important, as stereopure drugs can often reduce the total dose of drug given and minimize any toxicity resulting from the inactive enantiomer [15]. Examples of specificity of enantiomers ranges from harmless, such as estrone with an inactive (-) form and active (+) form, to penicillamine, which is extremely toxic when dosed in the L-form [16]. When dosed in the (R)-form, thalidomide acts as a sedative, but treatment with the (S)-enantiomer is highly teratogenic [17]. However, chiral inversion of thalidomide occurs within the body, and both the (S)-and (R)-forms are interconverted with both oral and intravenous dosing. In 1992, the FDA revised its policy on the registration of new drugs, and now requires separate pharmacological, pharmacokinetic, and toxicological profiles for each enantiomer in a racemic mixture [18]. As a result of these changes requiring additional toxicology studies of racemic compounds, single enantiomer drugs have since dominated the pharmaceuticals approved in the United States [19].
In order to evaluate enantiospecific effects of YK-4-279, we separated both enantiomers to purity and tested each in comparison to the racemate. Our results indicate a significant difference in activity between the two enantiomers. These experiments clarify the activity of YK-4-279 based upon its single enantiomer activity and contribute to the future clinical development of YK-4-279 in an era where novel approaches to cancer therapy are critical to improving patient care.

YK-4-279 can be separated into enantiomers
The chiral molecule YK-4-279 was resolved into its individual enantiomers using preparative high pressure liquid chromatography (HPLC). The analytical trace indicates the presence of two enantiomers ( Figure 1A). The retention time of the first enantiomer is 8.2 minutes, while the retention of the second enantiomer is 20.0 minutes. Conditions from the batch analysis were scaled linearly to a 250mm x 77mm preparative column. Aliquots of each sample were analyzed by chemical and chiral HPLC and mass spectrometry (MS). The HPLC results show chemical and chiral purity of both fractions, with chemical purity and enantiomeric excess values of >99%. MS analyses of the enantiomers are consistent with the structure of YK-4-279.
The enantiomers were analyzed for optical rotation to determine the polarization angle. The first enantiomer to elute from the column has an optical rotation of -191.9°, while the second enantiomer rotates the polarized light clockwise +192.6°. To further determine the enantiomer  (1 -30µM) was added to EWS-FLI1, followed by the addition of RHA and detection with antibody (**, p < 0.05 compared to vehicle control, using a two-tailed Student's t-test). (B) TC32 cells were treated with 10µM small molecule for 15 hours. Immunoprecipitation of EWS-FLI1 and RHA was detected via western blot. (C) TC32 cells were treated with 0.3, 1, 3, and 10µM of small molecule for 15 hours followed by immunoprecipitation. (D) Immunoblot of 10% input for (C). (E) Densitometry was calculated for each band and the ratio RHA to EWS-FLI1 was plotted. conformation, crystals were grown in hexane/ethyl acetate. Crystals were obtained from both (+) and (-) enantiomers. The crystals of both the (+) and (-) enantiomers were analyzed by x-ray crystallography. The (+) enantiomer has (R)-configuration while the (-) enantiomer has (S)configuration ( Figure 1B). Small molecule circular dichroism demonstrated the expected mirror image spectra of the individual enantiomers, while the racemate had zero net rotation of light, indicating a mixture of two isomers in equal proportions ( Figure 1C).

(S)-YK-4-279 disrupts binding of RHA with EWS-FLI1
Previous studies from our group indicate the ability of YK-4-279 to block the binding between EWS-FLI1 and RHA [14]. To determine if enantiospecific blocking of this protein-protein interaction occurs, binding between EWS-FLI1 and RHA was measured using ELISA. Wells coated with EWS-FLI1 were treated with a range from 1µM to 30µM YK-4-279, (S)-YK-4-279, (R)-YK-4-279, or vehicle. Both the racemate and (S)-YK-4-279 are able to inhibit the protein-protein interaction with as little as 1µM small molecule, while (R)-YK-4-279 does not inhibit the binding at 30µM (Figure 2A). Multiple ELISA runs suggests that (S)-YK-4-279 is only slightly more potent at disrupting the complex based upon increased dissociation at 1µM.
Next, ESFT cells were treated with YK-4-279, (S)-YK-4-279, and (R)-YK-4-279, followed by immunoprecipitation of EWS-FLI1. We treated TC32 cells with 10µM of the racemic or enantiomeric small molecule for 15 hours, consistent with the K D value of YK-4-279 [14]. Immunoblotting showed coimmunoprecipitation of EWS-FLI1 with RHA in the presence of vehicle or (R)-YK-4-279, but there is a significant reduction in complexed RHA in the lysates from cells treated with either YK-4-279 or (S)-YK-4-279 ( Figure 2B). The control IgG lanes do not indicate the pull-down of either EWS-FLI1 or RHA.
To determine the relative potency of YK-4-279 and (S)-YK-4-279 to disrupt the binding between EWS-FLI1 and RHA in ESFT cells, we titrated down the amount of small molecule (range = 0.3 to 10µM). TC32 cells were treated for 15 hours before immunopreciptation followed by immunoblotting for RHA complexes ( Figure 2C). Cells treated with vehicle or (R)-YK-4-279 were again able to pull down RHA, as were cells treated with lower doses of YK-4-279 and (S)-YK-4-279, but the binding between EWS-FLI1 and RHA was disrupted in a dosespecific fashion. Treatment did not affect the EWS-FLI1 or RHA levels ( Figure 2D). Densitometry of RHA relative to FLI1 was calculated and shows an IC 50 of 4.9 µM for the racemic and 1.8 µM for the (S)-YK-4-279, with a significant difference between vehicle and cells treated with 0.3µM (S)-YK-4-279, but not for cells treated with the racemic molecule ( Figure 2E). The experiment was repeated three times and a representative blot is shown.

EWS-FLI1 functional activity is reduced by only one enantiomer
Since (S)-YK-4-279 is able to block the binding between EWS-FLI1 and RHA, we next investigated the ability of the enantiomers to reduce transcriptional activity. We transfected COS7 cells with EWS-FLI1 and the NR0B1-luciferase reporter plasmid, which contains 25 EWS-FLI1 DNA-binding sites. YK-4-279 and (S)-YK-4-279 were able to inhibit EWS-FLI1 transcriptional activity in a dose-dependent manner compared to vehicletreated cells ( Figure 3A), with an IC 50 of 0.96µM for racemic and 0.75µM for the (S)-enantiomer. At a dose of 0.3µM, luciferase activity was significantly reduced from cells treated with (S)-YK-4-279 compared to vehicle treated (p < 0.05), while treatment with racemic compared to control was not significant. The (R)-enantiomer did not show significant reduction from control. Data is averaged from three experiments, each performed in triplicate. Expression of EWS-FLI1 into COS7 was similar in each experiment and a representative Western blot is shown ( Figure 3B). EWS-FLI1 has been shown to increase Cyclin D1 levels by altering the D1b/D1a mRNA level in ESFT cells through effects on transcript elongation [20]. Since racemic YK-4-279 significantly decreases Cyclin D1 levels in TC32 cells [14], we evaluated whether this effect was enantiospecific. We treated A673i cells with 1µM of YK-4-279 or enantiomer for 8 hours and then analyzed mRNA for levels of Cyclin D1a and Cyclin D1b using quantitative RT-PCR. As a positive control, we reduced EWS-FLI1 protein levels in A673i with an induced shRNA. With 1µM racemic YK-4-279 treatment, there is a reduction in the D1b/D1a levels ( Figure 3C). The (S)-YK-4-279 is approximately 35% more potent at the D1b/D1a reduction than racemic and reduced this ratio to a level equivalent to the EWS-FLI1 reduction. Consistent with other experiments, the (R)-YK-4-279 cells maintained a D1b/D1a ratio equivalent to the vehicle treated cells. There was no decrease in EWS-FLI1 protein expression in cells treated with small molecule ( Figure  3D). Therefore, an enantiospecific effect was seen in a second EWS-FLI1 activity assay.

YK-4-279 demonstrates enantiospecific cellular effects in EWS-FLI1 containing cells
After determining that (S)-YK-4-279 is able to block both the interaction between EWS-FLI1 and RHA as well as reduce EWS-FLI1 function while (R)-YK-4-279 does not, we tested the small molecules for cytotoxicity in a panel of ESFT cell lines compared to cell lines that lack ets rearrangements. TC32, along with six other cell lines expressing EWS-FLI1, were treated with either a vehicle or dose of small molecule ranging from 0.1 to 30µM of compound for three days ( Figure 4A). Six of these cell lines demonstrated significant cytotoxicity to (S)-YK-4-279 compared to racemic (p < 0.05, two-tailed Student's t-test) while the (R)-YK-4-279 enantiomer demonstrated no specific toxicity. Experiments were repeated three times in triplicate and mean IC 50 values ranged from 0.33µM to 1.83µM for racemic YK-4-279, 0.16µM to 0.87µM for (S)-YK-4-279, and 11.69µM to 25.98µM for (R)-YK-4-279 ( Figure 4B, Table 1 Figure 4E). The slope, representing the rate of caspase activity increase, is greater between 0.3 and 1µM for the (S)-YK-4-279 (m = 7.23 ± 0.32) than the racemic (m = 4.43 ± 0.66), but the caspase activity begins to reach saturation by 3µM.

(S)-YK-4-279 and (R)-YK-4-279 lack racemization in xenograft tumors
Since there was such a clear difference in activity between the two enantiomers, we evaluated the potential for racemization in whole organisms, including ESFT xenografts. YK-4-279 in racemic or purified enantiomeric form was administered by intravenous (IV) injection or oral gavage to Sprague-Dawley rats ( Table 2). As expected, since the racemic contains a 1:1 mixture of each enantiomer, rats treated with each enantiomer individually had 1.8-to 2.3-fold higher concentrations of the specific enantiomer in plasma than rats treated with the racemic (data not shown). There was no racemization indicated in plasma after IV or gavage administration since (R)-YK-4-279 was not present in the plasma when (S)-YK-4-279 was administered and vice versa ( Table 2). However, maximum urine concentrations of YK-4-279 showed 3.1% of (R)-YK-4-279 when (S)-YK-4-279 was administered and 6.9 % of (S)-enantiomer when (R)-enantiomer was given by IV injection. After gavage administration, there were more pronounced racemizations with 36.2% of (R)-enantiomer present in urine when (S)-enantiomer was administered and 35.7% of (S)-enantiomer when (R)-enantiomer was given. The A4573 ESFT xenograft model was used to determine if any racemization occurred in tumors. (S)-YK-4-279 was administered by intravenous (IV) injections to A4573 tumor-bearing SCID mice. Tumors were resected two hours following dosing. At that time, there was no (R)-YK-4-279 present in tumor tissues after dosing with (S)-YK-4-279. This lack of racemization within the tumors supports development of the (S)-YK-4-279 for clinical trials rather than the racemic mixture.

DISCUSSION
We recognized that the small molecule YK-4-279 has a chiral center and can be separated into its enantiomers. In binding assays and transcriptional assays, (S)-YK-4-279 is active and titrations indicate more potency when compared to the racemic, while (R)-YK-4-279 is inactive. The cellular response to (S)-YK-4-279 demonstrates that it is the active enantiomer in both apoptosis and growth assays in EWS-FLI1 containing cells. These data characterize the enantiospecificty of YK-4-279.
We show that YK-4-279 and (S)-YK-4-279 are able to block the EWS-FLI1/RHA interaction, but (R)-YK-4-279 cannot. This clearly demonstrates the importance of stereochemistry for the targeted disruption of small molecule modulators of protein-protein interactions. It further supports the hypothesis that small molecule interactions with intrinsically disordered proteins can have significant specificity, despite relatively low affinity [21]. While the structure of the full-length EWS-FLI1 protein still eludes us due to its intrinsic disorder, the functional difference of the two enantiomers may help us  In order to evaluate the potency of the enantiomers towards cells containing EWS-FLI1, we tested a panel of cells in cytotoxicity and caspase assays. Our racemic small molecule consists of equal parts of the (S)-and (R)-enantiomers; thus one would expect a two-fold difference in potency between the active enantiomer and the racemate. Apoptosis was measured by the activation of caspase-3. Through a dose titration, we do see the expected two-fold difference in caspase activation between ESFT cells treated with YK-4-279 and (S)-YK-4-279 for 18 hours, at lower concentrations while this saturates at the highest concentrations. In each of the ESFT cell lines studied for cytotoxicity, (S)-YK-4-279 had a lower IC 50 value than YK-4-279. The actual fold-differences between the racemic and the active enantiomer vary from 1.18-fold in RDES cells to more than a log-fold difference in TC71 cells. We tested cells containing all three major variant translocation types of Ewing's sarcoma and this difference cannot be attributed to the specific translocation [3]. One explanation for this is based on the steep nature of the dose response curve, which causes large changes in IC 50 based on small shifts in viability. Differences in cellular uptake or metabolism of YK-4-279, independent of translocation type, may also account for the fold-difference between racemic and active enantiomer.
The cytotoxicity and caspase results indicate that (S)-YK-4-279 is the active enantiomer, and that

A 4 5 7 3 T C 3 2 R D E S M M H -E S -1 S T A -E T 7 . 2 S K E S M C F 7 M D A -M B -2 3 1 P A N C 1
ESFT Non-ESFT www.impactjournals.com/oncotarget it retains specificity to ESFT cells containing an ETS family transcription factor in comparison to non-Ewing's cells. Previously, non-transformed cell lines were tested with racemic YK-4-279 and found to have IC 50 values of >30µM [14]. It is important to note that cell growth of both MDA-MB-231 and PANC1 cells, breast and pancreas respectively, were reduced by YK-4-279 and (S)-YK-4-279. These cell lines were previously reported to have IC 50 values of >20µM; however, we received these cell lines from a different source than previously used and the current clones are more sensitive. Despite the increased sensitivity in cytotoxicity studies, the IC 50 values for MDA-MB-231 and PANC1 cells were lower for the racemic than the active enantiomer and there was no statistical significance between treatments with racemic and (S)-YK-4-279. When tested for caspase-3 activation, neither of these cell lines exhibited more than 2-fold caspase-activation, in comparison to the ESFT cells, which averaged 8.  [22], including apoptosis-inducing factor (AIF) [23], endonuclease G [24], and HtrA2 [25], which may explain the increase in cytotoxicity without caspase activation in these two non-ESFT cell lines. Further studies focusing on caspase-independent cell death may explain why these cells exhibit cytotoxicity toward YK-4-279 but do not undergo apoptosis and may inform other potential mechanisms of activity.
In preparation for clinical trials, we evaluated the potential racemization of YK-4-279. Pharmacokinetic studies in rats show no conversion between (S)-YK-4-279 and (R)-YK-4-279 in the plasma following IV dosing and very minimal conversion in urine. However, there is increased racemization in urine, but no sign of conversion in plasma, after the gavage administration, which may be attributed to low oral bioavailability (2-6% for (S)-YK-4-279 and 8-15% for (R)-YK-4-279) and much slower elimination of YK-4-279 compared to the IV administration. Importantly, no enantioconversion was seen in the tumors. Taken together, these results suggest that the chirality of YK-4-279 does not change in the plasma or tumor tissues. The conversion in the urine indicates either a tissue-specific presence of an enzyme that isomerizes YK-4-279 or a gastric effect that alters compound structure. Additional experiments testing biospecificity are necessary to further study the pharmacokinetics of (S)-YK-4-279 in vivo to advance the small molecule to clinical trials. Although xenograft mice treated with YK-4-279 exhibited no toxicity when treated with 75 mg per kg body weight [14], separating out the inactive enantiomer may allow for a reduction in dosage or an increased effect.
ETS rearrangements are present in 40-70% of all prostate cancers, including those that are the most clinically aggressive [26][27][28]. These aggressive prostate cancers fuse the promoter of TMPRSS2, an androgen responsive gene, to an ETS family transcription factor, such as ETV1 or ERG [29], both of which have also been identified in joining to EWS in ESFT to form an oncogenic fusion protein [30][31][32]. Recent experiments from our group have also indicated that YK-4-279 is able to inhibit ERG and ETV1 fusion-positive prostate cancer cell lines [33], and the activity and function of the enantiomers should be further investigated in prostate cancer. Our data support the ability to disrupt a critical oncogenic proteinprotein interaction with YK-4-279 and the importance of enantiospecificity in disrupting EWS-FLI1 from RHA. A small molecule targeted to the ETS family of transcription factors with minimal side effects would be an important development in treatment for patients with ESFT and other ETS-family oncogenic fusion proteins.

Separation of YK-4-279
The chemical and chiral HPLC analyses were performed using Waters XBridge C18 (250mm x 4.6mm) and Chiral Technologies Chiralpak AD (250mm x 4.6mm) columns, respectively. The enantiomers were resolved by preparative HPLC using a Chiralpak AD column (250mm x 77mm) packed in-house using a Varian Dynamax TM Rampak Column Packing Station model 41.4/77. Mobile phase was 60% 2-propanol in heptane and flow rate was maintained at 250ml/min. All mobile phase batches were premixed by volume. Sample solution was prepared for purification by dissolving the sample in 10:40:50 dichloromethane/reagent alcohol/heptane. The fractions collected during the purification were transferred to round bottom flasks and evaporated using mild temperature conditions (30 -35°C) until all solvent was removed. The analytical HPLC system used for method development and sample analyses was a Waters 2695 Alliance Systems coupled to a Waters 996 Photo Diode (PDA) detector. Preparative HPLC separations were performed using a Waters Delta Prep 2000/4000 equipped with #7 pump heads coupled to a Waters 484 UV-Vis detector. Optical rotation analyses were performed using a Perkin-Elmer Model 343 Polarimeter with c = 10 mg/ml in methanol and T = 25°C.

Resolution and Crystallography of YK-4-279
Crystals suitable for x-ray diffraction experiments were obtained through slow evaporation of a solution containing ethyl acetate. The diffraction experiments were performed at 100K on a Rigaku R-axis Rapid www.impactjournals.com/oncotarget diffactometer, equipped with a Mo Kα radiation source (60kV, 40mA). HKL-2000 was used for control of the data collection as well as data reduction [34]. The structure was solved and refined by the HKL-3000SM system, which is integrated with SHELXS, SHELXL [35], and 0 [36,37]. Absolute configurations of both compounds were determined using anomalous dispersion. Both compounds crystallized in space group P1 with five molecules of the YK-4-279 and disordered solvent molecules in the asymmetric unit.

Small Molecule Circular Dichroism
Circular dichroism was studied using 100µM YK-4-279 and 50µM of each enantiomer in a buffer containing 1% ethanol, 5mM sodium phosphate, and 50mM potassium fluoride, pH 7.4. Data is averaged from 5 or more independent scans. Scans were performed on a Jasco J-715 spectropolarimeter from 400 -190 nm with a bandwidth of 5.0nm, a response of 8 seconds, and a scan speed of 50 nm/minute.

ELISA
EWS-FLI1 was used to coat the surface of a 96 well plate (MaxiSorb) at a concentration of 150ng/ well in 100µL buffer containing 20mM Tris, 500mM sodium chloride, and 1M imidazole. Following overnight incubation, the plate was blocked with 4% BSA in PBS and washed with PBS + 0.1% Tween-20. Small molecules were added at concentrations of 1µM, 3µM, 10µM, and 30µM, followed by RHA (300ng/well) and allowed to incubate overnight. Following washing, incubation with RHA antibody (Everest) and anti-goat secondary, protein binding was detected using a TMB Peroxidase EIA Substrate Kit (Bio-Rad) per the manufacturer's instructions.

Quantitative RT-PCR
A673i cells [38] were treated with 1µM small molecule for 8 hours. Positive control cells were treated with tetracycline for 72 hours to knock down expression of EWS-FLI1. Total RNA was extracted using an RNeasy Micro Kit (Qiagen) and reverse transcribed using the Transcriptor First Strand cDNA Synthesis Kit (Roche) per the manufacturer's instructions. qPCR was performed using an Eppendorf Mastercycler Realplex with FastStart Universal SYBR Green Master with ROX (Roche) with primers for 18S, CyclinD1a, and CyclinD1b. Data were analyzed for expression relative to 18S using the comparative Ct method. Data from four separate experiments performed in duplicate were averaged.

Immunoprecipitation and Western Blotting
TC32 were grown to ~70% confluency and cells were treated for 15 hours with small molecule. Nuclear lysate was collected using the Active Motif Magnetic Co-IP Kit (Active Motif). Protein concentration was determined using bicinchoninic acid protein assay for each lysate (Pierce) and lysate was bound to 1µg of antibody overnight at 4C on a rotating axis with addition of same concentration of small molecule as used for 15 hour treatment. Magnetic beads (Active Motif) were added to the lysates and tumbled for 2 hours at 4C. Immunoprecipitated proteins were resolved using 10% PAGE and transferred to a polyvinylidene difluoride membrane (Millipore). Membranes were blocked in 5% nonfat dry milk in TTBS (20mM Tris-HCl, 150mM NaCl, 0.5% Tween 20) for 2 hours. FLI1 (Santa Cruz Biotechnologies) and RHA (Abcam) antibodies were used at 1:1000 dilutions in 5% nonfat dry milk for 2 hours. Horseradish peroxidase-linked anti-rabbit antibody (GE Healthcare) was added for 1 hour. Detection was carried out using Millipore Immobilon Western Chemiluminescent HRP Substrate per the manufacturer's instructions (Millipore Corp.). Chemiluminescence was detected using a Fujifilm LAS-3000 imaging system. Densitometry values were obtained using ImageJ software.

Reporter Assay
Cells were transiently transfected with the NR0B1 luciferase reporter [20] and full length EWS-FLI1 into COS7 cells with Fugene-6 (Roche) according to the manufacturer's protocol. Two hours after transformation, cells were treated with 0.3, 1, and 3µM of YK-4-279, (S)-YK-4-279, and (R)-YK-4-279. Luciferase activity was measured 16 hours after treatment. Protein concentration was normalized by absorbance at 280nm. All luciferase assays were performed using a luciferase assay kit according to the manufacturer's protocol (Promega).

Cellular Proliferation Assays
TC32, TC71, RDES, MMH-ES-1, STA-ET 7.2, A4573, and PC3 cells were grown in RPMI with 10% FBS and 1% HEPES. SKES cells were grown in McCoy's 5A medium with 15% FBS. MCF7, MDA-MB-231, PANC1, and ASPC1 cells were grown in colorless DMEM with 10% FBS. Cells were grown at a plating density of 5,000 -15,000 cells/well, depending on cell line, in a 96-well plate. Small molecule or vehicle alone (DMSO) were added to cells in appropriate growth media the day after plating. After three days, viable cells were quantified using WST (Roche) according to the manufacturer's protocol. IC 50 values were calculated by sigmoidal dose-response