Suppressing the Aberrant Transcriptional Functionality of EWS::FLI1 Oncoprotein by Designer polyQ Fusions with Its Homologous Peptides

TL;DR

This paper explores a new method to fight Ewing sarcoma by trapping a harmful protein, EWS::FLI1, using specially designed polyQ fusion proteins.

Contribution

The novel use of polyQ fusion proteins to sequester and reduce the activity of the EWS::FLI1 oncoprotein in Ewing sarcoma cells.

Findings

PolyQ fusion proteins co-precipitate and co-localize with EWS::FLI1, forming insoluble aggregates.

The polyQ fusions reduce the cellular availability of EWS::FLI1 and alter downstream gene expression.

Atx793Q-N172-LCD increases P21 and decreases c-Myc, suggesting therapeutic potential.

Abstract

Background/Objectives: The oncoprotein EWS::FLI1 is a chimeric transcription factor that aberrantly brings transcriptional deregulation relevant to Ewing sarcoma. It is also regarded as a therapeutic target for suppressing oncogenic progression, but the inhibition and clearance of the EWS::FLI1 oncoprotein remain a challenge. Methods: We apply a polyglutamine (polyQ) fusion strategy to directly target EWS::FLI1 in suppression of its transcriptional malfunction in A673 cells derived from Ewing sarcoma. Based on the template of the N-terminal fragment of polyQ-expanded ataxin-7 (Atx793Q-N172) and the homologous peptides of EWS::FLI1, we have designed and constructed three polyQ fusion proteins, namely Atx793Q-N172-SYGQ1, Atx793Q-N172-SYGQ2, and Atx793Q-N172-LCD. Results: Supernatant/pellet fractionation and immunofluorescence imaging reveal that the polyQ fusion proteins co-precipitate and co-localize with EWS::FLI1 in A673 cells, indicating that the polyQ fusions we have designed can sequester endogenous EWS::FLI1 into insoluble aggregates and reduce its cellular availability. Moreover, these polyQ fusions, especially Atx793Q-N172-LCD, alter the expression of EWS::FLI1 downstream genes, with an increase in P21 (CDKN1A) and a decrease in c-Myc. Conclusions: These results demonstrate that the engineered polyQ fusions entrap endogenous EWS::FLI1 protein into aggregates and reduce its soluble fraction in Ewing sarcoma cells. This study provides an alternative potential for treating Ewing sarcoma and other tumors by directly targeting the oncogenic proteins in the future.

Figures

Full text

1. Introduction

Ewing sarcoma is a malignant tumor that often occurs in bone and soft tissues of children and adolescents [1,2]. Research has reported that these tumors are generally caused by various chromosomal translocations, which can generate a series of oncogenic proteins fused by the E-twenty-six (ETS) transcription factor family and FET (FUS, EWSR1, and TAF15) family [3]. These genetically mutated chimeric oncoproteins bring aberrant transcriptional regulation, either reducing the crucial functional genes or promoting oncogenic transformation [4,5,6]. Among these chimeras, EWS::ETS is an extensively studied oncoprotein essential to Ewing sarcoma [7]. The EWS::ETS chimeric proteins contribute to oncogenic progression by the ETS domain binding to DNA sequences with a central GGAA/T core motif driving aberrant enhancer remodeling [8,9,10,11], and the chimeric oncoprotein functions in cell-cycle regulation and differentiation [12,13].

The chimeric protein EWS::FLI1 is widely regarded as the oncogenic protein for Ewing sarcoma caused by a chromosomal translocation (t11;22) (q12;q24) [14,15,16]. As an aberrant transcription factor, EWS::FLI1 consists of the N-terminus of EWSR1 from the FET family and the C-terminus of FLI1 from the ETS transcription factor family [17,18]. The N-terminus of EWSR1 mainly comprises a low complexity domain (LCD) [19] including two SYGQ-rich (rich in Ser, Tyr, Gly, and Gln residues) regions [20], while the C-terminus of FLI1 carrying a DNA-binding domain (DBD) not only contributes simply to the DNA-binding function, but also is implicated separately in the transcriptional regulation [16,21].

Previous studies have shown that protein interactions occur between the FET family proteins through their LCD sequences [22,23,24,25], and the FET proteins exhibit liquid–liquid phase separation (LLPS) properties [26]. Since the LCD portion of EWSR1 can self-interact and accumulate, it implies that the chimeric protein EWS::FLI1 can also undergo LLPS by the LCD-LCD self-interactions [27,28], which may modulate oncogenic transcription [10,29]. Moreover, both the EWS::FLI1 chimera and the EWSR1 protein purified in vitro can aggregate rapidly, while FLI1 remains soluble [13,30].

EWS::FLI1 is only expressed in Ewing sarcoma-derived cells and contributes to the progression of the cancer, which is regarded as an ideal candidate for targeted therapy for the disease. However, its structural composition contains only LCD regions and lacks a well-defined pocket for ligand interaction, leading to a series of difficulties and challenges in designing specific blocking targets [31,32]. So, exploring new therapeutic opportunities for treating Ewing sarcoma is becoming important and urgent.

To suppress the aberrant functionality of EWS::FLI1 in oncogenic progression for therapeutic purposes, several strategies have been developed for targeting EWS::FLI1 [17,33], such as small-molecule inhibitors [34,35,36] and more recently, gene editing [37,38]. In this study, we applied a series of polyglutamine (polyQ) fusion proteins combined with the homologous peptides derived from EWS::FLI1. As previously indicated [39], the polyQ fusion proteins can self-aggregate in cells, sequester cellular essential components into aggregates, deplete the soluble availability of the sequestered proteins, and cause their dysfunctions. Our previous study has shown that the polyQ fusions sequester the USP7/HDM2 proteins into aggregates effectively, leading to a reduction in the availability of these key enzymes and thereby the modulation of the P53 signaling pathway [40]. Thus, we proposed a novel strategy for modulating the cellular processes and therapeutic potential against various diseases by using the aggregation and sequestration properties of the engineered polyQ fusion proteins. On the basis of this polyQ-fusion technology and the SYGQ-rich regions that recapitulate the LCD-LCD interaction of the EWS::FLI1 molecules, we designed and engineered three polyQ fusion proteins combining the SYGQ-based homogenous peptides with Atx7_93Q_-N172. These polyQ fusion proteins can sequester EWS::FLI1 into aggregates in A673 cells and suppress its aberrant functionality in oncogenic progression, which may provide a potential therapeutic application for treating Ewing sarcoma and even other tumors.

2. Materials and Methods

2.1. Plasmids, Reagents, and Antibodies

The SYGQ1, SYGQ2, and LCD fusions were cloned into a pcDNA3.0 (NLS-Atx7_10Q_-N172 or NLS-Atx7_93Q_-N172) vector via EcoR I/Xba I sites, respectively [41], which encode the HA-tagged fusion proteins. The plasmids encoding FLAG-tagged EWSR1 and FLI1, and EWS::FLI1 were generated by cloning their DNA fragments into a pcDNA3.1 (FLAG) vector via BamH I/Xho I sites. The c-Myc Luciferase Reporter plasmid (pGL3-c-Myc-Fluc) was cloned from a commercial pGL3-NFκB-FLuc plasmid, and the Renilla Luciferase plasmid (pGMLR-Renilla) was purchased from Yeason and Promega. Mutagenesis technology was applied to replace the c-Myc response element with EBS1 (ETS transcriptional factor-binding site 1 from CDKN1A promoter) [42], namely pGL3-EBS1-FLuc. EBS1 is the direct binding site of EWS::FLI1 at the CDKN1A promoter, which implies that EBS1 can suppress the expression of P21. The DNA sequences of the constructed plasmids in this study were confirmed by DNA sequencing (Table S1), and the oligonucleotides, including PCR primers, are listed in Table S2.

The anti-HA and anti-FLAG antibodies were purchased from Sigma-Aldrich (St. Louis, MO, USA), while the anti-GAPDH, anti-P21, and anti-Myc antibodies were purchased from Proteintech (Chicago, IL, USA). The anti-FLI1 antibody (Abcam, Waltham, MA, USA) against the C-terminal epitope of FLI1 was applied to detect EWS::FLI1 in A673 cells, while the rabbit and mouse anti-EWS antibodies (Absin (Shanghai, China) or Santa Cruz (Dallas, TX, USA)) were used to detect EWSR1 in HEK 293T cells. The secondary antibodies were commercially obtained from Jackson ImmunoResearch Laboratories (West Grove, PA, USA). All the primary and secondary antibodies are listed in Table S3.

2.2. Cell Culture and Plasmid Transfection

The medium for the cell culture was DMEM (HyClone, Marlborough, MA, USA) supplemented with 10% FBS (Gibco, Waltham, MA, USA) and penicillin–streptomycin (Invitrogen, Carlsbad, CA, USA). The culture conditions were maintained in a humidified atmosphere (5% CO_2_) at a temperature of 37 °C. Plasmid transfections were performed by using PolyJet reagent (SignaGen, Rockville, MD, USA) or Lipofectamine 3000 (Invitrogen). The HEK 293T and A673 cells were purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China).

2.3. Western Blotting and Protein Analysis

Protein analysis was carried out by Western blotting on the cell lysates or separated fractions. The proteins on the SDS-PAGE gels were transferred onto PVDF membranes (Millipore, Burlington, MA, USA), and then the blots were sectioned as necessary and stained with the indicated antibodies. Appropriate primary and secondary antibodies were applied for detecting the specific proteins, followed by developing the protein bands with an ECL kit (ThermoFisher Scientific, Waltham, MA, USA). The Sage Capture V3.2.2 software (http://www.sagecreation.com.cn/en/ (accessed on 11 January 2022)) was used for collecting integral values and quantitative protein analysis.

2.4. Supernatant/Pellet (S/P) Fractionation Experiment

S/P fractionation was performed following a previously described protocol in our laboratory [43,44]. The cultured A673 or HEK 293T cells were collected about 48 h post-transfection and lysed on ice in 100 μL of RIPA buffer (50 mM Tris (pH 7.5), 150 mM NaCl, 1 mM EDTA, 1% NP-40, and Roche protease inhibitors) for a 30 min incubation. Subsequently, the cell lysates were centrifuged at 13,000× g for 15 min at 4 °C. Afterwards, the separated supernatant (100 μL) was mixed with the 2× loading buffer (100 μL, 4% SDS). The pellet was washed three times with the RIPA buffer at 4 °C and re-suspended in 50 μL of the 4× loading buffer (8% SDS). After boiling for 5 min, equal volumes of both the supernatant and pellet were analyzed by SDS-PAGE and Western blotting. Generally, the volume of supernatant was larger than that of the pellet, then the respective fraction was estimated according to the following formula: F_s_ = nS/(nS + P); F_p_ = P/(nS + P), where “n” denotes the ratio of the supernatant volume over the pellet (V_S_/V_P_).

2.5. Total Protein Extraction

About 48 h post-transfection, the cultured cells were harvested and lysed in 100 μL of the RIPA buffer containing 8 M urea to solubilize protein aggregates. Then, the lysates were mixed with an equal volume (~100 μL) of 2× loading buffer (4% SDS, 8 M urea), denatured by boiling, and analyzed via SDS-PAGE followed by Western blotting.

2.6. Immunoprecipitation (IP)

An IP experiment was carried out as described previously [45]. Briefly, the HEK 293T or A673 cells were harvested approximately 48 h post-transfection and lysed with the RIPA buffer. The cell lysates were centrifuged (4 °C) at 13,000× g for 20 min, and then the supernatant was incubated with anti-FLAG beads (Abmart, Shanghai, China) for 4 h, and the beads were washed three times with the RIPA buffer. Subsequently, 40 μL of the 2× loading buffer (4% SDS) was mixed with the beads, and the mixtures were boiled for denaturation before immunoblotting analysis.

2.7. Dual Luciferase Reporter Assay

A673 cells were co-transfected with pGL3-c-Myc-FLuc or pGL3-EBS1-FLuc and pGMLR-Renilla plasmids along with Atx7_93Q_-N172, Atx7_93Q_-N172-SYGQ1, Atx7_93Q_-N172-SYGQ2, or Atx7_93Q_-N172-LCD. At about 48 h post-transfection, the cells were lysed following the instructions of the Dual Luciferase Reporter Gene Assay Kit (Yeason Biotech, Shanghai, China). GloMax^®^ 20/20 Luminometer (Promega, Madison, WI, USA) was used to record the dual luminescence data, and the ratio of Firefly luciferase over Renilla luciferase (FLuc/RLuc) was applied to represent the relative activity.

2.8. Immunofluorescence (IF) Imaging

For IF imaging, the HEK 293T or A673 cells cultured on glass coverslips for ~48 h after transfection were washed with a PBS buffer (10 mM Na_2_HPO_4_ and 1.8 mM KH_2_PO_4_ (pH 7.3), 140 mM NaCl, and 2.7 mM KCl), and fixed with 4% paraformaldehyde (PFA) for 15 min. After washing three times with the PBS buffer, cell permeabilization was conducted with the PBS buffer containing 0.1% Triton X-100 and 5% BSA for 1 h at room temperature. Afterwards, the cells were incubated overnight at 4 °C with the specific primary antibodies diluted in PBS. After washing with PBS, the cells were incubated with either FITC or TRITC conjugated secondary antibodies (Jackson ImmunoResearch Laboratories; FITC—fluorescein isothiocyanate; TRITC—tetramethyl rhodamine isothiocyanate). Nuclei were stained with Hoechst33342 (ThermoFisher). Confocal imaging was performed on a Leica TCS SP8 WLL system (Leica Microsystems, Wetzlar, Germany), and the images were captured using LAS X acquisition software (https://www.leica-microsystems.com/ (accessed on 16 March 2022)).

2.9. Quantitative Real-Time PCR (qPCR) Assay

About 48 h post-transfection, the cultured A673 cells were harvested for RNA extraction using TRIzol reagent (Life Technologies, Waltham, MA, USA). cDNA synthesis was completed by using 4× Reverse Transcription Master Mix (EZBioscience, Suzhou, China) with 2 μg of total RNA per sample. The qPCR reactions (a total volume of 20 μL) were run in triplicate on a LightCycler96 system (Roche, Basel, Switzerland) using Hieff qPCR SYBR Green Master Mix (Yeasen Biotech, Shanghai, China), which includes 2 μL of cDNA template, 0.5 mM of each primer (Table S2), and 10 μL of Master Mix. The relative mRNA levels were calculated via the 2^−ΔΔCt^ algorithm, normalized to the GAPDH expression. The qPCR assay was performed in triplicate to ensure data reliability.

2.10. Statistical Analysis

The band intensities on Western blots were quantified by using ImageJ V1.53 software (https://imagej.net/ (accessed on 11 January 2022)). Integrated grayscale values of proteins were normalized to respective controls. Data were obtained in triplicate and presented as Mean ± SD. Statistical analysis was conducted using GraphPad Prism V7.0 (https://www.graphpad.com/ (accessed on 11 January 2022)) with Student’s t-test or one-way ANOVA. Statistical significance was defined as p < 0.05, where different p-values were indicated with asterisks, showing * (p < 0.05), ** (p < 0.01), *** (p < 0.001), or N.S. (no significance) in the graphs.

3. Results

3.1. Endogenous EWS::FLI1 Distributes Mainly in the Soluble Fraction

The EWS::FLI1 protein consists of the N-terminus of EWSR1 (residues 1–264) and the C-terminus of FLI1 (residues 218–452) (Figure 1A). Its N-terminal SYGQ-rich regions are considered to be a factor that determines the solubility of the protein [26], while the SYGQ-rich segments, designated as SYGQ1 (26–72) and SYGQ2 (201–264) (Figure 1A), have been reported to be sufficient to recover the aberrant function of EWS::FLI1 [13,46]. At first, the supernatant/pellet (S/P) fractionation experiment was carried out, and the result showed that the EWS::FLI1 protein dominantly distributed in the soluble fraction, while only a subtle fraction of the protein remained in the pellet (Figure 1B). As a comparison, EWSR1 also appeared mainly in the soluble fraction with a small fraction of insolubility (Figure 1C). Immunofluorescence (IF) imaging exhibited that the endogenous EWS::FLI1 protein formed nuclear puncta in A673 cells as liquid-like droplets (Figure 1D). It is consistent with the previous observation that the exogenous EWS::FLI1 forms droplet-like biomolecular condensates in A673 cells [46].

Next, we over-expressed the three proteins (EWSR1, FLI1, and EWS::FLI1) in HEK 293T cells, respectively (Figure S1). S/P fractionation showed that nearly half amount of EWSR1 (Figure S1A) and a major fraction of FLI1 (Figure S1B) were in the soluble fraction, whereas, unexpectedly, the exogenous EWS::FLI1 protein distributed mainly in the insoluble fraction (Figure S1C). Then, IF imaging exhibited that the exogenous EWSR1 and FLI1 still diffused mainly in the nucleus, but EWS::FLI1 formed some nuclear puncta (Figure S1D), indicating that EWS::FLI1 could assemble into biomolecular condensates or aggregates [47] in an over-expression manner. However, the morphology and dynamics of EWS::FLI1 in cells remain largely unknown. Our future work will define whether these biomolecular assemblies are droplet-like condensates or insoluble aggregates by fluorescence recovery after photobleaching (FRAP) experiments. This implies that the chimeric oncoprotein EWS::FLI1 is more prone to biomolecular condensation and further aggregation than its constructing partners EWSR1 and FLI1.

3.2. Design and Construction of the polyQ Fusion Proteins Targeting EWS::FLI1 Directly

In the EWS::FLI1 chimera, two SYGQ-rich regions located in its N-terminus have been considered to play a role in protein interaction that contributes to protein condensation, aggregation, and sequestration [30,46]. On the other hand, the polyQ tracts harbored in proteins possess the aggregation propensities that even lead to proteinopathies. Based on the Atx7_93Q_-N172 template [40,41], we designed and engineered three polyQ fusion proteins, each combining with a homologous SYGQ-rich peptide derived from EWS::FLI1, namely Atx7_93Q_-N172-SYGQ1, Atx7_93Q_-N172-SYGQ2, and Atx7_93Q_-N172-LCD (Figure 2A). Because EWS::FLI1, as well as EWSR1, is a nucleus-addressed protein, an NLS sequence is also included in the polyQ fusions for directing the designer proteins into the nucleus.

We firstly characterized the interaction between the polyQ fusions and EWS::FLI1 or EWSR1 by using the Atx7_10Q_-N172 template that does not have aggregation propensity [41]. Immunoprecipitation (IP) showed that Atx7_10Q_-N172-LCD could obviously precipitate endogenous EWS::FLI1 in A673 cells (Figure 2B) and endogenous EWSR1 in HEK 293T cells (Figure S2), but Atx7_10Q_-N172-SYGQ1/SYGQ2 could not. It indicates that Atx7_10Q_-N172-LCD can interact with EWS::FLI1 or EWSR1 in cells, while the other two polyQ fusions were undetectable under the experimental condition. The full-length LCD (consisting of SYGQ1 and SYGQ2) may have multivalent interactions, conferring the LCD fusion on strong interaction with EWS::FLI1 or EWSR1, although its individual portions, SYGQ1 and SYGQ2, interact with EWS::FLI1 or EWSR1 relatively weakly. As an inference, our designed polyQ fusions at least Atx7_93Q_-N172-LCD can interact with endogenous EWS::FLI1 as well as EWSR1 in their aggregated forms through their fused SYGQ-rich peptides.

3.3. Sequestration of Endogenous EWS::FLI1 by the polyQ Fusion Proteins

First, we examined whether over-expression of the polyQ fusions affects the total protein level of endogenous EWS::FLI1. After transfection of the Atx7_93Q_-N172-SYGQ1, Atx7_93Q_-N172-SYGQ2, or Atx7_93Q_-N172-LCD plasmid, the protein level of EWS::FLI1 was detected by Western blotting. The results exhibited that the total protein levels of EWS::FLI1 in A673 cells had no observable changes caused by over-expression of the three polyQ fusions (Figure 3A), while that of EWSR1 in HEK 293T cells had a slight increase by the over-expression of Atx7_93Q_-N172-LCD, due to a yet unknown reason (Figure S3A).

We then detected the protein levels of EWS::FLI1 in supernatant and pellet fractions upon over-expression of the three polyQ fusions in A673 cells by using S/P fractionation. Compared with free Atx7_93Q_-N172, the three polyQ fusions, especially Atx7_93Q_-N172-LCD, caused a remarkable increase in the protein level of EWS::FLI1 in the pellet fraction, whereas a considerable decrease in the supernatant (Figure 3B). It suggests that the designer polyQ fusions can sequester endogenous EWS::FLI1 into aggregates, resulting in a reduction in its soluble availability.

We also investigated the co-localization and sequestration of EWS::FLI1 in A673 cells by IF imaging (Figure 4). As expected, all the polyQ fusions could form punctum-like condensates or aggregates in the nucleus. As our previous observation that Atx7_93Q_-N172 could self-aggregate in cells [40,41], the nuclear puncta formed by these polyQ fusions were thought to be insoluble aggregates. Compared with Atx7_93Q_-N172, which did not co-localize with endogenous EWS::FLI1, the three peptide-fused species could co-localize with the endogenous EWS::FLI1 in the nucleus, suggesting that they could sequester endogenous EWS::FLI1 into the nuclear aggregates. Similarly, these fusions also formed nuclear puncta in HEK 293T cells, but only Atx7_93Q_-N172-LCD could co-localize with the endogenous EWSR1 (Figure S4) and sequester it into the nuclear aggregates (Figure S3B). Together, these data suggest that the polyQ fusions at least Atx7_93Q_-N172-LCD can effectively sequester endogenous EWS::FLI1 into nuclear aggregates.

3.4. The polyQ Fusions Influence Expression of the Downstream Genes by Sequestering EWS::FLI1

EWS::FLI1, as the chimeric oncoprotein, has aberrant transcriptional functions in frequent occurrence and crucial mechanism in Ewing sarcoma. Some target genes of EWS::FLI1 have been identified, showing that induction of the EWS::FLI1 target genes is implicated in transformation and/or tumor progression, while some relevant genes are reported to be repressed by EWS::FLI1 [48]. Among these targets, CDKN1A/P21 and c-Myc are thought to be the direct targets of EWS::FLI1 [21,42,49]. So, we next investigated whether the polyQ fusions influence the expression of these target genes.

When the polyQ fusions were over-expressed in A673 cells, the mRNA levels of CDKN1A and c-Myc were determined by quantitative PCR (qPCR). Our data showed that the mRNA level of CDKN1A was increased significantly upon over-expression of the respective polyQ fusions as compared with Atx7_93Q_-N172 (Figure 5A), whereas that of c-Myc was decreased significantly (Figure 5B). Similar results were also obtained from the protein assay by Western blotting. It showed that the polyQ fusions, especially Atx7_93Q_-N172-LCD, significantly enhanced the protein level of P21 (Figure 5C) but reduced that of c-Myc (Figure 5D). Together, the designer polyQ fusions, especially the LCD-fused form, can modulate expression of the target genes by sequestering EWS::FLI1, which has the potential to be applied in therapeutic suppression of the EWS::FLI1 oncogenicity.

To further validate the effects of polyQ fusion proteins on the downstream target gene expression, we constructed two dual-luciferase reporter plasmids [42], namely CDKN1A-FLuc (firefly luciferase expression plasmid containing the EWS::FLI1 response element EBS1 in the CDKN1A promoter) (Figure 6A) and c-Myc-FLuc (firefly luciferase expression plasmid with the c-Myc promoter response element) (Figure 6B). The reporter system also included a Renilla luciferase plasmid for an evaluation standard, and the relative activity of dual luciferase (FLuc/RLuc) was then recorded and quantified. In A673 cells, when Atx7_93Q_-N172-fused SYGQ1, SYGQ2, or LCD was co-transfected with the reporter plasmids, the transcriptional activity of the firefly luciferase downstream of the CDKN1A promoter response element was significantly increased, as compared with transfection of Atx7_93Q_-N172 (Figure 6C). Among the polyQ fusions, Atx7_93Q_-N172-LCD was most effective in promoting the activity. On the other hand, the transcriptional activity of the c-Myc promoter significantly decreased, and as with Atx7_93Q_-N172-LCD, it was the most suppressive (Figure 6D). Collectively, these results suggest that the polyQ fusions may deplete the availability and functionality of EWS::FLI1 in cells by sequestering the oncoprotein into inclusions and thereby influence the gene expression of its downstream targets.

4. Discussion

The chimeric oncoprotein EWS::FLI1 is the key driving factor of Ewing sarcoma and thereby becomes a therapeutic target for treating the disease [31]. This aberrant transcription factor plays a critical role in regulating the expression of many downstream genes and orchestrating the oncogenic progression responsible for malignant transformation. However, the aberrant transcription factor EWS::FLI1 has neither obvious molecular pockets nor specific interaction targets due to the intrinsically disordered structure of LCD in the EWSR1 portion and the lack of ligand-binding sites in the FLI1 portion [50]; directly targeting the oncoprotein remains challenging [51]. So, there is a great demand for investigating the novel therapeutic strategies that may provide improved specificity and efficacy.

To date, great endeavors have been made to target the EWS::FLI1 oncoprotein for therapeutic purposes in Ewing sarcoma [17,33], such as modulating its interactome and splicing [24,52,53], silencing the oncoprotein by oligonucleotides [54], modulating the turnover of EWS::FLI1 [55], and more importantly, gene editing by CRISPR/Cas9 [37,38]. Previous studies focused on targeting the key factors in the signaling pathways relevant to Ewing sarcoma [56]. These include inhibition of the transcriptional programs of EWS::FLI1 by targeting LSD1 [57] or BET bromodomain [58], ablation of EWS::FLI1 through deregulating the deubiquitinase USP9X [59] or USP19 [60] and the E3 ligase TRIM8 [61], and modulation of the HDAC3/HSP90 signaling for EWS::FLI1 [62]. Small-molecule inhibitors TK216 and YK-4-279 directly target on the EWS::FLI1/RHA complex [34,35,36], but they may have a high probability of causing off-target effects.

Our previous studies established a novel strategy based on polyQ fusion proteins to modulate the functionalities of transcription factors by sequestering the targeted proteins (enzymes) [40,63]. We have designed and constructed several polyQ fusion proteins by combining the polyQ template (such as Atx7_93Q_-N172) with the target-binding peptides. The designer polyQ-IRF and polyQ-PMI fusions can specifically bind to and effectively sequester USP7/HDM2 into aggregates, leading to the depletion of USP7/HDM2, and consequently enhancement of the P53 functionality [40]. Based on this polyQ-fusion technology, three polyQ-fusion proteins have been developed to sequester the aberrant transcription factor EWS::FLI1 and suppress its functionality in oncogenic progression, and thereby possibly reverse the transformation of the sarcoma. These polyQ fusion proteins are proposed to suppress the aberrant functionality of the aberrant transcription factor by directly targeting the EWS::FLI1 oncoprotein, reminiscent of the effects of small-molecule inhibitors and interference RNAs.

EWS::FLI1 shares the same LCD region with EWSR1, implying that they have a similar mechanism underlying biomolecular interaction and sequestration. The LCD region of EWSR1 plays important roles in normal cellular functions and disease pathologies [64,65]. It is also proposed that sequestration of the EWS::FLI1 oncoprotein into the nucleolus may be mediated by LCD-LCD interactions [13,46]. Moreover, two fragments derived from the LCD region of EWSR1 combined with the C-terminus of FLI1 are sufficient to recover the EWS::FLI1 functionality [13].

Our data suggest that EWS::FLI1 is different from its partner EWSR1 in cellular distribution, condensate formation, and the functioning niches. The designed polyQ fusions exert considerable sequestration effects on the EWS::FLI1 oncoprotein in A673 cells, but they may have less effect on the EWSR1 protein in HEK 293T cells. On the other hand, the SYGQ1 and SYGQ2 fusions have little sequestration effect on endogenous EWSR1 in HEK 293T cells, while the LCD fusion may sequester EWSR1 relatively strongly owing to its high expression in these cells. These designer polyQ fusions may have a priority and specificity in sequestering the EWS::FLI1 chimera in Ewing sarcoma cells. Among these three fusions, it seems that the LCD-fused form sequesters both EWS::FLI1 and EWSR1 more strongly than the fusions with short SYGQ peptides. Intriguingly, the SYGQ2 fusion sequesters EWS::FLI1 moderately but does not sequester EWSR1. We think that the SYGQ2 fusion is probably an optimal candidate for suppressing the aberrant functionality of EWS::FLI1 for therapeutic purposes. So, the polyQ fusion proteins with the homogeneous peptide derived from EWS::FLI1 may be feasible to treat Ewing sarcoma by sequestering the oncogenic protein into insoluble aggregates or inclusions and depleting its cellular availability and functionality. However, the targeting specificity and therapeutic application opportunity remain to be further explored and evaluated in the future.

Both CDKN1A and c-Myc are the EWS::FLI1 downstream genes associated with the pathogenesis of Ewing sarcoma [21,42,49]. CDKN1A is the direct target of EWS::FLI1, with its promoter containing two ETS binding sites, EBS1 and EBS2 [42], while c-Myc, as the target gene of EWS::FLI1, contributes to tumorigenesis after the transcriptional activation by EWS::FLI1 [21,49]. EWS::FLI1 mainly suppresses transcription of P21 while upregulating the transcriptional activation of c-Myc, both contributing to tumorigenesis [21,42,49]. Our designed polyQ fusion proteins self-aggregate in cells and sequester endogenous EWS::FLI1 into aggregates, leading to depletion of the availability of EWS::FLI1. This sequestration effect may result in restoration of the transcriptional inhibition of CDKN1A and the aberrant transcription activation of c-Myc caused by EWS::FLI1, and thereby suppress tumor progression effectively. Further work will in-depth study the impact of these polyQ fusions on Ewing sarcoma, such as cell proliferation, viability, and apoptosis. Therefore, these designer polyQ fusions confer the ability to remedy the oncogenic function of EWS::FLI1 and likely hold therapeutic potential for Ewing sarcoma in the future.

5. Conclusions

In summary, we have designed and engineered three polyQ fusion proteins combined with the homologous peptides of EWS::FLI1 and applied them to reverse the aberrant transcriptional functionality in oncogenic progression. The designed polyQ fusions, especially the SYGQ2 fusion, sequester EWS::FLI1 into aggregates or inclusions with relative specificities, effectively reduce the aberrant transcriptional function of EWS::FLI1, directly improve the expression of P21 and reduce that of c-Myc, and potentially suppress tumor progression. Thus, directly targeting EWS::FLI1 by the polyQ fusion proteins is a promising strategy to treat Ewing sarcoma with high therapeutic feasibility and opportunity. Further research should be implemented for validating the polyQ-fusion strategy in therapeutic potential, focusing on sequestration mechanism, delivery method, targeting specificity, and functional analysis in vivo.