2. Involvement of EWS-FLI1 in Transcription, Epigenetic Reprogramming, and Alternative Splicing in EwS
2.1. EWS-FLI1 in Transcription and Epigenetic Reprogramming
EWS-FLI1 is an aberrant transcription factor that drives cellular transformation by rewiring the epigenome to induce a large number of oncogenes. The N-terminus of EWSR1 contains a prion-like domain, characterized by an intrinsically disordered structure and low complexity. This domain has phase transition properties and manipulates multiple proteins involved in epigenome reprogramming and epigenetic alterations
[29,30,31,32,33,34,35,36][10][11][12][13][14][15][16][17]. In addition to the canonical ETS-binding sites, EWS-FLI1 binds to DNA sequences at the GGAA/T core motif
[37,38,39][18][19][20] via a conserved ETS domain. It regulates multiple proteins through its prion-like domain to tumor-specific enhancers and promotors, recruiting acetyltransferases and establishing de novo enhancers by generating H3K27ac, thus opening the chromosomal architecture, which contributes to the activation of target genes
[29,30,37,39][10][11][18][20]. The EWS-FLI1 protein complex includes RNA polymerase II
[23[21][22],
40], the core subunit hsRBP7 (human RNA polymerase II)
[41[23][24],
42], E2F3
[43[25][26],
44], EWSR1
[45][27], CBP/p300
[46][28], WDR5, ASH2, MLL
[30][11], and the BAF complex (mammalian SWI/SNF complex)
[29,47][10][29]. The threshold of GGAA motifs optimal for maximal expression is 20–26
[48][30], which differs from that in wild-type FLI1. Super-enhancer-associated MEIS1 and RING1B also contribute to the chromatin reprogramming through co-localization with EWS-FLI1 at the active enhancers to drive the malignancy of EwS
[49,50][31][32]. This specific coupling results in the activation of many genes (
Figure 1), such as NKX2.2
[51][33], NROB1
[52[34][35][36],
53,54], IGF1R
[55][37], BCL11B
[56][38], EZH2
[36][17], VRK1
[30][11], GLI1
[57][39], PTPL1
[58][40], PPPR1A
[59][41], ERG2
[60][42], GSTM4
[61][43], PAX7
[62][44], CHM1
[63][45], REST
[64][46], PHF19
[32][13], STEAP1
[65[47][48],
66], SLFN11(Schlafen 11)
[67][49], HDAC3
[68][50], TNC
[69][51], APCDD1
[49][31], IL1RAP
[70,71][52][53], MYC
[72][54], and PRC1 (protein regulator of cytokinesis 1)
[73][55].
Figure 1. The EWS-FLI1 protein complex drives the specific transcription profile of EwS. EWS-FLI1 recruits E2F3, hsRBP7, BAF, RING1B, RHA, P300, and MEIS1, among others, to GGAA repeats and further activates CHM1, EZH2, PAX7, NKX2.2, NROB1, and STEAP1, among others. EWS-FLI1 functions as a protein complex with ERG, BAF, RHA, DDX5, and U1C to drive alternative splicing. Among the genes, the purple ones, such as CHM1, could serve as TCR-based immunotherapy targets; the red ones, such as NKX2-2, serve as diagnostic markers in clinic diagnosis. EWS/FLI1 recruits LSD1 and unknown transcription factors (?) to repress TGFBR1 and IGFBP3, which still needs further research.
Among the direct targets of EWS-FLI1, NKX2-2 mediates oncogenic transformation via transcriptional repression and is necessary and sufficient for the oncogenic phenotype of EwS
[74][56]. Further work demonstrates that NKX2-2, KLF15, and TCF4 occupy similar super-enhancers and promoters, forming an inter-connected auto-regulatory loop and occupying 77.2% of promoters and 55.6% of enhancers shared with EWS-FLI1
[75][57], such as STEAP1
[76][58]; this kind of coordinate regulation drives the proliferation of EwS. NROB1 directly interacts with EWS-FLI1 to modulate multiple gene expressions and mediate the oncogenic phenotype of EwS
[77][59]. SLFN11 is a putative DNA/RNA helicase that recruits to the stressed replication fork and irreversibly triggers replication block and cell death. Overexpression of SLFN11 is associated with resistance to topoisomerase I inhibitors and poly (ADP-ribose) polymerase (PARP) inhibitor combinations
[67,78][49][60]. STEAP1 and IL1RAP are vital for the redox homeostasis of EwS
[65,70][47][52]. APCDD1, PHF19, GSTM4, and PTPL1 are genes that are involved in the proliferation of EwS.
EWS-FLI1 is also involved in transcriptional repression of tumor suppressors such as IGFBP3
[79][61] and PHLDA1
[53][35] to drive oncogenic transformation
[31,80][12][62]. The nucleosome remodeling and deacetylase (NuRD) complex is a typical ATP-dependent chromatin remodeling complex
[81][63] that plays a critical role in transcription and determines differentiation and development
[82][64]. EWS-FLI1 recruits the NuRD-LSD1 complex to repress LOX and TGFBR2
[80,83][62][65]. It also affects the transcriptional activation of AP-1
[33][14] and MRTFB
[84][66] and binds to the promotor of FOXO1 to repress its expression, thereby increasing tumor growth
[85][67]. EWS-FLI1 promotes the phosphorylation of cyclin-dependent kinase-2 and AKT to inhibit the activity of FOXO1, thus rewiring transcriptional repression
[85][67]. EWS-FLI1 is also involved in the regulation of microRNAs (miRNAs)
[86][68]. It downregulates miRNA-145 to initiate mesenchymal stem-cell reprogramming toward EwS stem cells
[87][69] and represses miR-708, which induces the overexpression of EYA3 and contributes to the chemoresistance to etoposide and doxorubicin
[88][70].
The histone methyltransferase EZH2 exhibits silencing activity via methylation of H3K27
[89][71]. In EwS, EWS-FLI1 upregulates EZH2 expression by interacting with the EZH2 promoter, thereby promoting tumor growth/metastasis and blocking endothelial/neuro-ectodermal differentiation
[36][17].
MiR-34a inhibits the proliferation and increases the sensitivity of EwS to doxorubicin and vincristine and is a strong predictor of a favorable prognosis in EwS
[90][72]. However, the exact mechanism underlying its downregulation remains elusive. Exportin 5 (XPO5), which mediates the nuclear export of pre-miRNAs and short hairpin RNAs
[91[73][74][75],
92,93], interacts with EWS-FLI1 based on mass spectrometry
[94][76]. XPO5 is highly expressed in various cancers including EwS. Furthermore, the phosphorylation of XPO5 alters the nucleus and cytoplasm shift
[95][77]. Investigating XPO5 and its relationship with EWS-FLI1 may offer new insights into the therapy of EwS. Post-translational modifications of EWS-FLI1 modulate its transcriptional activity. Phosphorylation and O-GlcNAcylation of the N-terminus of EWSR1
[96[78][79][80],
97,98], as well as acetylation of the C-terminal FLI1 domain by PCAF (KAT2B, lysine acetyltransferase 2B), enhance the transcriptional activity of EWS-FLI1
[99][81]. However, PCAF expression is lower in EwS tissues, which is a common feature of cancer.
2.2. EWS-FLI1 in Alternative Splicing
Pre-mRNA splicing is critical for gene expression, and most protein-encoding transcripts are alternatively spliced to provide diverse functions
[100,101][82][83]. The N-terminus of EWSR1 interacts with the hyperphosphorylated RNA polymerase II and recruits serine-arginine (SR) through its C-terminus. After chromosome translocation, the C-terminus of wild-type EWSR1 is replaced by FLI1, which hinders the recruitment of SR-splicing factors and interferes with mRNA splicing
[23][21], thus demonstrating the negative property of this chimeric protein
[102][84]. This leads to comprehensive alternative splicing of numerous genes. Meanwhile, EWS-FLI1 interacts with the splicing components (snRNP) U1C and SF1 to modulate pre-mRNA splicing
[103][85]. It also recruits the BAF complex to drive the alternative splicing of ARID1A and the preferential splicing of ARID1A-L, which is necessary for tumor growth
[104][86]. Work by Selvanathan
[94][76] demonstrates that EWS-FLI1 is involved in the alternative splicing of CLK1, CASP3, PPFIBP1, and TERT, which potentially regulate the oncogenesis of EwS.
3. The Regulation of EWS-FLI1
Transcription and post-transcriptional modifications are involved in the regulation of expression and activity of EWS-FLI1. Although the transcriptional regulation of EWS-FLI1 remains elusive, the BRD4 inhibitor JQ1 suppresses this activity
[32,34,105][13][15][87]. HDAC6 deacetylates specificity protein 1 (SP1), thereby inhibiting the recruitment of the SP1/P300 complex to the promoters of EWSR1 and EWS-FLI1 and downregulating EWS-FLI1
[106][88]. MiR-145 and let-7 repress EWS-FLI1 by targeting its mRNA
[87,107,108,109][69][89][90][91] and inhibit the proliferation of EwS. The RNA-binding protein LIN28B interacts with EWS-FLI1 transcripts to maintain the stability and ensure the expression of EWS-FLI1 to enhance the tumorigenicity of the self-renewal of EwS
[108][90]. At the post-transcriptional level, EWS-FLI1 degradation is proteasome dependent, and the protein has a half-life of 2–4 h
[110][92]. This process can be protected by the action of ubiquitin-specific protease 19 (USP19) at the N-terminus
[111][93] and accelerated by tripartite-motif-containing 8 (TRIM8) at K334
[112][94]; however, USP19 is expressed at low levels in EwS. Casein kinase 1 (CK1)-mediated phosphorylation of the VTSSS degron in the FLI1 domain activates speckle-type POZ protein (SPOP) activity, which degrades EWS-FLI1. In contrast, OTU-domain-containing protein 7A (OTUD7A) participates in the deubiquitination of the C-terminus and stabilizes EWS-FLI1
[113][95]. The inhibitor of chromosomal maintenance 1 (CRM1 and XPO1), KPT-330
[114][96], and IFN-γ
[115][97] suppress expression of EWS-FLI1 at the protein level. FOXM1, a downstream factor of EWS-FLI1, upregulates its expression
[116][98]. Cytosine arabinoside (ARA-C) downregulates EWS-FLI1 at the protein level and inhibits tumor growth
[117][99]; however, it shows hematologic toxicity and minimal activity in patients
[118][100].
STAG2 (stromal antigen 2) is a core subunit of the cohesion complex and is frequently mutated in multiple cancers
[119][101] including EwS
[10,120][102][103]. Mutation of STAG2 in EwS is associated with poor outcomes by improving metastasis
[10][102]. Mechanically, in addition to the disruption of PRC2-mediated regulation of gene expression in EwS
[121][104], the inactivation of STAG2 strongly altered CTCF-anchored loop extrusion and decreases promotor-enhancer interactions. As a result, the cis-mediated EWS-FLI1 activity at GGAA microsatellite neo-enhancers is downregulated and the cells are enhanced in their migration and invasion properties
[121,122][104][105].
Unlike STAG2, there was no evidence showing mutations of the ETS transcription factor ETV6 in EwS
[123][106]. ETV6 does not change the expression of EWS-FLI1 but co-occupy loci genome wide at the short consecutive GGAA repeats and constrains the transcriptional activity of EWS-FLI1
[124,125][107][108]. Upon inactivating ETV6, EWS-FLI1 overtakes and activates these cis-elements to promote mesenchymal differentiation by upregulating the expression of SOX11
[125][108].
4. CAR-T Therapy
Unlike TCR-based T-cell therapy, which is limited to specific HLA restriction and deficient HLA expression in EwS
[205][109] because of the presence of myeloid-derived suppressor cells, F2 fibrocytes, and M2-like macrophages in the microenvironment
[190][110], CAR-engineered T-cell therapy can target specific cell-surface antigens in tumors, independent of HLA. VEGFR2 is a potential target for CAR-T-cell therapy in EwS
[206][111]. In addition to TCR-T-cells, CAR-T-cells
[207][112] can target STEAP1, which is involved in the malignant phenotype of EwS
[65][47].
CAR-T targeting GPR64, ROR1, and IGF1R, which are highly expressed in EwS
[137[113][114],
208], leads to a selective killing of EwS in vivo
[209][115]. LINGO1, which is highly expressed in EwS
[210][116], is a direct target of EWS-FLI1
(Supplementary Figure S4B). EZH2 inhibition by GSK-126 induces GD2 surface expression in EwS
[211][117], and the combination of CAR-T therapy targeting GD2 and EZH2 inhibitors have synthetic cytotoxic in the treatment of EwS; this kind of combination provides new options for the clinical application. IL1RAP, a direct target of EWS-FLI1, is highly expressed in EwS, but minimally expressed in normal tissues, which makes it a promising surface target for EwS
[70][52] and a potential candidate for advanced CAR-T therapy. ICAM-1 can promote tumor cell/T-cell interaction and T-cell activation, and the knockdown of EWS-FLI1 upregulates ICAM-1 expression and leads to the upregulation of PD-L1 and PD-L2, both proteins that inhibit the activity of T-cells
[115][97]. Blocking PD-1 with a checkpoint inhibitor could increase the T-cell-mediated killing of EwS cells with low expression of EWS-FLI1.