Abstract

   Ewing sarcoma (ES) poses a significant therapeutic challenge due to the
   difficulty in targeting its main oncodriver, EWS::FLI1. We show that
   pharmacological targeting of the EWS::FLI1 transcriptional complex via
   inhibition of P300/CBP drives a global transcriptional outcome similar
   to direct knockdown of EWS::FLI1, and furthermore yields prognostic
   risk factors for ES patient outcome. We find that EWS::FLI1 upregulates
   LMNB1 via repetitive GGAA motif recognition and acetylation codes in ES
   cells and EWS::FLI1-permissive mesenchymal stem cells, which when
   reversed by P300 inhibition leads to senescence of ES cells.
   P300-inhibited senescent ES cells can then be eliminated by senolytics
   targeting the PI3K signaling pathway. The vulnerability of ES cells to
   this combination therapy suggests an appealing synergistic strategy for
   future therapeutic exploration.

Supplementary Information

   The online version contains supplementary material available at
   10.1186/s12943-024-02115-7.

   Keywords: Ewing sarcoma, EWS::FLI1, Lamin B1, P300/CBP, Senescence,
   Senolytics, Pharmacological targeting

One sentence summary

   Pharmacological inhibition of P300/CBP disrupts the EWS::FLI1
   transcriptional complex in Ewing sarcoma, leading to senescence and
   vulnerability to senolytics, providing a promising therapeutic strategy
   for this challenging cancer.

Introduction

   Ewing sarcoma (ES), the second most common form of pediatric bone
   cancer, primarily affects individuals at a median age of 15 years and
   occurs with an incidence of approximately three cases per million
   annually [[42]1, [43]2]. Over the last fifty years, improvements in
   diagnosis, surgery, chemotherapy, and radiation have increased survival
   rates to nearly 70% for localized ES, yet the prognosis for patients
   with metastatic or recurrent forms remains challenging, with a
   five-year survival rate still below 25% [[44]3, [45]4]. This disparity
   underscores an urgent need for new therapeutic targets and strategies.

   ES is defined by chromosomal translocations that form fusion genes
   encoding aberrant transcription factors crucial for pathogenesis. The
   predominant translocation, t(11;22)(q24;q12), present in about 85% of
   cases, leads to the formation of EWS::FLI1 [[46]5]. Acting as a
   transcription factor, the EWS::FLI1 fusion protein alters cellular
   functions and promotes the expression of key oncogenes such as CMYC,
   ID2, and CCND1, as well as genes associated with super enhancer
   activities like MEIS1 and APCDD1 [[47]6, [48]7]. The preferential
   binding of EWS::FLI1 to GGAA microsatellites creates enhancers that
   amplify expression of the oncogene, promoting tumor aggression and
   treatment resistance [[49]8].

   However, the structural complexity of EWS::FLI1 and the absence of a
   defined active site make direct targeting a challenge [[50]9, [51]10].
   Research efforts are thus increasingly focused on understanding and
   disrupting the regulatory network of EWS::FLI1 as a strategy to
   counteract its oncogenic effects [[52]2, [53]6, [54]11].

   The oncogenic potential of EWS::FLI1 is highly dependent on the
   cellular context, as evidenced by its varying effects across different
   cell types [[55]12]. In permissive environments like mesenchymal stem
   (MSC) and neuroectodermal cells, the EWS::FLI1 fusion protein not only
   drives ES-like transformation and tumorigenesis by promoting cellular
   proliferation [[56]12], but also significantly contributes to immune
   evasion [[57]13]. This oncogenic protein modulates immune-related gene
   expression, alters cytokine profiles to favor immunosuppression,
   interferes with effective antigen presentation by downregulating MHC
   class I molecules, and manipulates the tumor microenvironment to
   encourage the recruitment of immunosuppressive cells like M2
   macrophages [[58]13]. EWS::FLI1 toxicity is apparent when expressed in
   non-permissive cells, as it triggers cellular stress responses that
   result in cell cycle arrest, apoptosis, or senescence [[59]14].

   P300 (E1A Binding Protein P300) and CBP (CREB Binding Protein) are
   critical histone acetyltransferases that mark promoters and enhancers
   for gene activation via acetylation of histone H3 at lysine 27
   (acH3K27) [[60]15]. This modification is essential for cancer
   progression in various types such as liver cancer, prostate cancer,
   melanoma, renal carcinoma, leukemia, lymphoma, and lung cancer
   [[61]15]. Acetylation by P300/CBP enhances cell proliferation,
   survival, and metastasis, contributing to therapy resistance and immune
   evasion mechanisms [[62]15]. There is a crucial partnership in ES,
   where P300/CBP is dynamically localized to interact with the activation
   domain of the EWS::FLI1 fusion protein. This results in enhanced
   oncogenic capacity through chromatin relaxation and remodeling, and
   subsequent tumor progression [[63]6, [64]8, [65]16, [66]17]. Given the
   challenges of directly targeting EWS::FLI1, our study focuses on
   P300/CBP as alternative therapeutic targets, aiming to disrupt this
   crucial interaction and thereby reduce oncogenic activity.

   Despite their crucial role in cellular processes, the specific function
   of P300/CBP in ES remains uncharted [[67]15]. This study underscores
   the dependence of the EWS::FLI1 fusion protein on P300/CBP to govern
   vital downstream targets crucial for ES cell growth and malignancy. We
   unravel the intriguing regulatory mechanism by which EWS::FLI1 prevents
   ES cancer cells and permissive cells from entering senescence, acting
   concurrently as a transcriptional activator of LMNB1 and repressor of
   P15. Moreover, we demonstrate pharmacological targeting of the
   EWS::FLI1/P300/CBP axis holds promise as a therapeutic approach for ES.

Results

siRNA-mediated knockdown of P300/CBP affects ES cell viability

   While P300/CBP is recognized as a promoter of cancer cell proliferation
   [[68]15, [69]18], its impact on ES viability remains uncertain. The
   influence of P300/CBP on the survival of three ES cell lines, SKES1,
   A4573, and TC71, was evaluated through siRNA-mediated knockdown.
   P300/CBP-targeted siRNA decreased cell viability across all examined ES
   cell lines, underscoring the vital role these histone
   acetyltransferases play in cell survival (Fig. [70]1A). The efficiency
   of siRNA-mediated knockdown was confirmed by assessing P300 and CBP
   protein levels via Western blot (Fig. [71]1B). Knockdown of P300/CBP
   led to a reduction in acetylated H3K18 and H3K27, indicating a shift
   toward a transcriptionally repressive chromatin state. Knockdown of
   either P300 or CBP reduced the expression of well-known targets of
   EWS::FLI1, such as NR0B1, MEIS1, and C-MYC levels in SKES1 and A4573 ES
   cell lines. However, silencing both P300 and CBP together had a more
   pronounced effect, resulting in heightened downregulation across a
   plethora of targets (Fig. [72]1C). Thus, knockdown of P300/CBP by RNAi
   emphasizes its indispensable role within the EWS::FLI1 transcriptional
   axis and in maintaining ES cell viability.

Fig. 1.

   [73]Fig. 1
   [74]Open in a new tab

   siRNA-mediated knockdown of P300/CBP unveils a pivotal role in ES.
   A Bar graphs show the significant reduction in cell viability across
   SKES1, A4573, and TC71 ES cell lines after siRNA-mediated knockdown of
   P300/CBP at 72 hours. The data represent mean ± SEM; **p<0.001,
   ****p<0.0001 by two-way ANOVA, n=4. Results are presented as percentage
   viability, normalized to the control sample. B Western blot analysis
   confirms the knockdown of P300/CBP at the protein level and reveals
   decreased levels of acetylated histone markers (acH3K18 and acH3K27)
   after 48 hours of siRNA treatment. “Cont” refers to untreated cells,
   “siCont” indicates cells treated with scramble siRNA, and “siCBP/P300”
   refers to cells treated with siRNA targeting both CBP and P300.
   C RT-qPCR analysis at 48 hours following P300/CBP knockdown shows
   decreased expression of EWS::FLI1 regulated genes (NR0B1, MEIS1, c-MYC,
   ID2) in SKES1 and A4573 cell lines. Silencing P300 or CBP individually
   led to a reduction in the target genes, with a more pronounced effect
   observed upon concurrent silencing of both. Data are presented as mean
   ± SEM; *p<0.05, by two-way ANOVA. Results are presented as relative
   expression to B2M (n=6)

Pharmacological inhibition P300/CBP inactivates EWS::FLI1 oncogenic activity

   To pharmacologically emulate the effects of P300/CBP siRNA-mediated
   knockdown, we treated ES cell lines with our recently developed
   P300/CBP inhibitor, iP300w [[75]18]. The treatment reduced cell
   viability in a dose-dependent manner across all tested ES cell lines.
   All tested ES cell lines exhibited significant sensitivity to iP300w at
   a concentration of 0.1 μM after 48 h of treatment (Fig. [76]2A and
   Supplementary Fig. 1A). In contrast, the viability of the non-sarcoma
   human myoblast LHCN cell line and the osteosarcoma cell line remained
   unaffected at this concentration (Fig. [77]2A and Supplementary Fig.
   1A). However, a slight but statistically significant decrease in cell
   viability was observed in the LHCN and SJSA-1 cell lines at 1.0 μM. The
   impact of P300/CBP inhibition on cell proliferation and morphology was
   readily observable under bright-field microscopy. iP300w-treated
   samples exhibited a marked decrease in cell confluence, along with a
   noticeable increase in cell size and a more flattened, spread-out
   morphology. (Fig. [78]2B). Ki-67 staining demonstrated almost complete
   suppression of cell proliferation at 48 h of the iP300w treatment (Fig.
   [79]2C and Supplementary Fig. 1B).

Fig. 2.

   [80]Fig. 2
   [81]Open in a new tab

   Successful pharmacological targeting of ES using iP300w. A Bar graphs
   show the dose-dependent inhibition of viability in ES cell lines
   (SKES1, A673, A4573, and TC71) compared to normal human myoblast cells
   (LHCN) following iP300w treatment for 48 hours. The ATP assay data are
   represented as mean ± SEM; p<0.05, by two-way ANOVA. Results are
   presented as fold difference compared to control (untreated group at 48
   hours) (n=3). B SKES1 cell morphology after 48 hours of treatment with
   1µM iP300w. Scale bar 100µm. C Immunofluorescence for Ki67 (red)
   expression in SKES1 cells indicates cell-cycle arrest after 48 hours of
   iP300w treatment (1 µM). DAPI (blue) was used to stain the nuclei.
   D Western blot analysis reveals for EWS::FLI1, P300, CBP, acH3K9,
   acH3K18 and acH3K27, in SKES1 and TC71 cell lines treated with 1 µM
   iP300w for 4 and 12 hours. E Bar graphs show RT-qPCR results
   illustrating unchanged EWS::FLI1 expression following P300/CBP
   inhibition, with a significant reduction in EWS::FLI1 target gene
   expression (NKX2.2 and CMYC) at 24 hours post-treatment. Data were
   normalized to B2M. The data represent the mean ± SEM, p< 0.05, by
   t-test. Results are presented as fold differences compared to the
   control (untreated group at 24 hours) (n = 6). F Gross morphology of
   tumors from treated (iP300w, 5.6 mg/kg daily) and control mice at the
   terminal point of the experiment (day 14). G The graphs display tumor
   volume and weight measurements, demonstrating significant reductions in
   tumor size and mass following treatment. Data are presented as mean ±
   SEM; ***p< 0.001, ****p< 0.0001, by t-test (control, n=6; treatment,
   n=7). H Western blot for acetylated H3K18 and H3K27 in tumor samples
   presented in “F”. I RT-qPCR analysis of EWS::FLI1, its target genes,
   senescence-related genes, and cell cycle genes in the xenografts from
   experiment “F”. Data were normalized to B2M and compared to the control
   group using a t-test. The data represent mean ± SEM; *p< 0.05, **p<
   0.01, ***p< 0.001, ****p< 0.0001, (control, n=6; treatment, n=7)

   Next, we investigated the transcriptional and epigenetic effects of
   iP300w treatment in both SKES1 and TC71 cell lines. Within 4 h,
   acetylation at H3K18 and H3K27 notably decreased, and was entirely
   diminished by the 12-h mark (Fig. [82]2D). As anticipated, there was no
   change in acetylation levels on H3K9, affirming the iP300w specific
   inhibition of P300/CBP-mediated H3 acetylation [[83]19]. Concurrently,
   protein levels of P300 and CBP remained stable, suggesting iP300w
   modulates the activity of these acetyltransferases without altering
   their expression (Fig. [84]2D). Moreover, RT-qPCR analysis at 24 h
   post-iP300w treatment revealed a significant downregulation of
   EWS::FLI1 target genes, notably NKX2.2, NR0B1 and C-MYC (Fig. [85]2E,
   Supplementary Fig. 1C). Remarkably, the sensitivity of EWS::FLI1
   transcriptional activity to treatment was evident within just 4 h
   (Supplementary Fig. 1D). Notably, EWS::FLI1 protein levels decreased
   with iP300w treatment while transcription remained unaffected. This
   suggests that the stability of EWS::FLI1 may be dependent on
   P300/CBP-mediated acetylation (Fig. [86]2D, E), similar to the
   relationship between P300/CBP and other transcription factors like
   TP53, SOX10, and MCL-1 [[87]20–[88]22].

   Finally, we assessed the efficacy of iP300w in suppressing ES tumor
   formation in a SKES1 xenograft tumor model (Fig. [89]2F, G and
   Supplementary Fig. 2A). Mice treated with iP300w exhibited significant
   reductions in both tumor size and mass compared to control groups. By
   the end of the experiment (day 14), average tumor volumes were
   significantly smaller in the treated group (300 mm^3, 0.3 g) compared
   to the untreated control group of mice (2200 mm^3, 2 g) (Fig. [90]2H,
   G). Consistent with the in vitro observations, treated tumors showed a
   notable decrease in H3K18 and H3K27 acetylation, as well as in the
   expression of EWS::FLI1 target genes (Fig. [91]2H and I, and
   Supplementary Fig. 1E). To evaluate the potential side effects of
   iP300w treatment, we monitored body weight changes (Supplementary Fig.
   2B) and assessed serum levels of BUN, creatinine, sodium, potassium,
   osmolality, ALP, ALT, and glucose (Supplementary Fig. 2C). Among all
   the parameters, only glucose showed a slight upregulation in the
   iP300w-treated group. BUN was actually modestly decreased with
   statistical significance in the treated group. Furthermore,
   histopathological analysis revealed no noticeable tissue damage or
   pathological changes in the liver and kidneys (Supplementary Fig. 2D).
   In summary, iP300w efficiently suppresses EWS::FLI1/P300/CBP oncogenic
   axis both in vitro and in vivo.

EWS::FLI1 knockdown and P300/CBP inhibition induce similar transcriptional
alterations in ES

   Global transcriptional analysis was employed in iP300w-treated ES cells
   to identify molecular mechanisms underlying EWS:FLI1/P300/CBP-driven
   malignancy. An advantage of chemical inhibition over siRNA knockdown is
   its rapid effect. To capture immediate and highly sensitive
   transcriptional responses while avoiding secondary targets, RNA-Seq was
   performed at 4 h post-treatment. Comparison revealed distinct
   clustering based on differentially expressed genes (DEGs), effectively
   segregating treated cells from controls (Fig. [92]3A). Principal
   Component Analysis (PCA) revealed a clear trend: while different ES
   cell lines clustered together in the control set, they were distinctly
   separated in the iP300w-treated groups, highlighting the substantial
   impact of iP300w treatment on cell transcriptomes (Supplementary Fig.
   3A, B). Differential gene expression (DGE) analyses revealed a
   pronounced predominance of downregulated genes over upregulated ones,
   in a ratio of 4:1, indicating widespread transcriptional repression
   induced by the treatment (Fig. [93]3B). This trend remained consistent
   across all four ES cell lines, demonstrating a notable overlap,
   especially among the downregulated genes.

Fig. 3.

   [94]Fig. 3
   [95]Open in a new tab

   P300/CBP inhibition and EWS::FLI1 knockdown yield comparable
   transcriptional changes in ES. A Heatmap illustrates the gene
   expression profiles of ES cells treated with iP300w (1 µM) for 4 hours,
   with hierarchical clustering indicating significant concordance among
   different ES samples. B Venn diagram shows the overlap of DEGs in ES
   cells post P300/CBP inhibition, with a predominant downregulation of
   gene expression. C The Venn diagram of DEGs in three ES cell lines
   following EWS::FLI1 knockdown reveals both common and unique DEGs
   across the cell lines [[96]23]. D The comparative Venn diagram
   illustrates the overlap between DEGs in iP300w-treated cells and
   EWS::FLI1 knockdown cells. E The UpsetR comparative analysis shows the
   overlap of DEGs between ES cell lines and non-ES cell lines treated
   with iP300w. F GSEA enrichment analyses revealed significant enrichment
   (p=8.54e-52) of EWS-FLI1 related pathways (Reactome Riggi) in
   iP300w-treated ES cell lines compared to other cancer cell lines. The
   EWS-FLI1 related genes in the “Reactome Riggi” were identified through
   the overexpression of EWS-FLI1 in human mesenchymal stem cells.
   G Volcano plot showing the distribution of DEGs influenced by the
   treatment, with EWS::FLI1 binding sites within 50KB of the promoter
   region, Canonical GGAA microsatellite EWS-ETS activation sites were
   identified through ChIP-Seq and selected based on their presence in at
   least 15 out of 18 ES cell lines [[97]23]. H GSEA enrichment analyses
   reveal significant enrichment (p=2.18e-4) of cell cycle checkpoints in
   iP300w-treated ES cell lines (EWS) compared to other cancer cell lines
   (Other). I KEGG pathway enrichment analyses on DEG in iP300w treated ES
   cells (4 hours). J KEGG pathway analyses were conducted on DEGs
   following EWS::FLI1 knockdown in ES cell lines. Note that iP300w and
   EWS::FLI1 knockdown both affect the cell cycle and induce cell
   senescence

   To assess the dependency of EWS::FLI1 transcriptional activity on
   P300/CBP, we compared transcriptional alterations following EWS::FLI1
   knockdown (48 h) and P300/CBP inhibition (4 h). Transcriptional
   profiles of A673, TC71, and SKES1 after EWS::FLI1 knockdown were
   accessed through the Gene Expression Omnibus (GEO) data repository
   [[98]23]. Although the knockdown approach requires a much longer
   timeframe, meaning much greater involvement of indirect targets, Venn
   diagram analysis indicated a substantial number of common genes
   affected by EWS::FLI1 knockdown across different ES cell lines, similar
   to what was observed previously with P300/CBP inhibition (Fig. [99]3B
   and C). Subsequent comparison between the two approaches revealed
   significant overlaps of approximately one-fourth of the differentially
   expressed genes (DEGs), indicating a shared transcriptional response to
   both EWS-FLI1 knockdown and iP300w treatment (Fig. [100]3D).

   To underscore the significance of P300/CBP in EWS::FLI1 activity, we
   utilized UpsetR for a comparative analysis of RNA-Seq data between ES
   cell lines and non-ES cell lines treated with iP300w (Fig. [101]3E). In
   the non-ES cell line pool, we included diverse cancer cell lines such
   as breast carcinoma (T-47D), rhabdomyosarcoma (A204), synovial sarcoma
   (Fuji), and osteosarcoma (HOS, U2-OS). The comparison of DEGs revealed
   that ES cells exhibited heightened sensitivity to iP300w as compared to
   other cancer cell lines. For instance, 330 DEGs were uniquely affected
   in ES cells compared to 124 affected genes in other lines (Fig.
   [102]3E). Furthermore, Gene Set Enrichment Analysis (GSEA) demonstrated
   a robust transcriptional correlation between iP300w-treated ES cells
   and mesenchymal stem cells (MSC) overexpressing EWS::FLI1
   (p = 8.54e-52; Reactome EWS::FLI1) (Fig. [103]3F). Notably, 74 of DEGs
   influenced by the treatment contained EWS::FLI1 binding sites within 50
   KB of the promoter region (Fig. [104]3G). Additionally, GSEA revealed a
   more pronounced impact on the cell cycle in iP300w-treated ES cell
   lines compared to other cancer cell lines (p = 2.18e-4; Reactome: cell
   cycle; Fig. [105]3H). This is further supported by the ATP assay
   results, which showed that iP300w did not significantly affect the
   viability of non-ES cell lines (Supplementary Fig. 1A, 3C).

   Finally, KEGG pathway analysis revealed modulation of pathways
   associated with cancer, stem cell pluripotency, and cellular signaling
   in iP300w-treated cells. Particularly notable changes were observed in
   the Hippo, Wnt, and Hedgehog pathways, indicating the widespread
   effects of P300/CBP inhibition on regulatory networks crucial for
   tumorigenesis and cell differentiation (Fig. [106]3I). Conversely, in
   EWS::FLI1 knockdown cells, DNA replication and cell cycle-related
   pathways were predominantly affected (Fig. [107]3J). Among the common
   most affected pathways between iP300w treatment and EWS::FLI1 knockdown
   were “Transcriptional misregulation in cancer” and “Cellular
   senescence”. Collectively, these findings highlight that at the
   transcriptional level, inhibition of P300/CBP with iP300w can yield
   effects comparable to EWS::FLI1 knockdown in ES cells.

Differential survival outcomes in ES patients linked to P300/CBP related
genes

   Expanding upon our gene expression findings, we explored the potential
   clinical relevance of EWS::FLI1/P300/CBP axis-related genes. We
   correlated the DEGs from iP300w-treated ES cell lines with the overall
   gene expression profiles of biopsy samples from 142 Ewing sarcoma
   patients archived in the International Cancer Genome Consortium (ICGC)
   database and [108]GSE63157 [[109]24]. Employing Non-negative Matrix
   Factorization (NMF), we categorized these patients into two distinct
   groups based on the expression patterns of the iP300w affected DEGs.
   This stratification is visually represented in an NMF clustering
   consensus map (Fig. [110]4A), while a heatmap (Fig. [111]4B) further
   elaborates on the expression profiles of the top five genes from each
   group, highlighting the potential impact of these DEGs on patient
   outcomes.

Fig. 4.

   [112]Fig. 4
   [113]Open in a new tab

   P300/CBP-related genes are linked to poor prognosis in patients with
   ES. A NMF clustering consensus map categorizing 142 ES patients into
   two distinct groups based on the expression patterns of iP300w-induced
   DEGs. B Heatmap illustrates the expression profiles of the top five
   genes from each group identified by NMF clustering. C Kaplan-Meier
   survival curves reveal significant differences in survival outcomes
   between the two patient groups identified through NMF clustering.
   Additional analyses include survival stratification by median age, as
   well as metastatic and relapse states. D Cox regression analysis
   presenting the top five and bottom five hazard ratio genes
   significantly associated with overall survival among the 74 identified
   genes. E The Lasso Cox regression model correlates gene expression with
   overall survival, assigning risk scores based on gene expression
   levels. Higher risk scores are identified as significant risk factors

   To explore the clinical implications of our findings, we conducted
   survival analysis of the two distinct patient groups identified via NMF
   clustering. There were significant differences (p = 0.0237) in survival
   outcomes among these groups (Fig. [114]4C), underscoring the potential
   prognostic relevance of P300/CBP related genes in ES. We stratified the
   patient samples by median age; however, this categorization did not
   reveal a significant impact on survival (p = 0.0939). Additionally, we
   assessed the role of metastatic and relapse state in patient prognosis.
   The analysis revealed that event-fee patients had significantly better
   survival compared to those with metastasis (Hazard Ratio = 0.4861,
   p = 0.0014) and those with relapse (Hazard Ratio = 0.4608, p = 0.0124).
   However, there was no significant difference in survival between the
   groups with metastasis and relapse (p = 0.913) (Fig. [115]4C). The
   overall survival analysis across the three groups was highly
   significant (p < 0.001).

   Given the classification of samples into two categories through NMF, a
   direct understanding of the relationship between DEGs and patient
   overall survival was not immediately clear. To bridge this gap, we
   employed Cox regression analysis to examine the link between DEG
   expression and overall survival, identifying 74 genes significantly
   associated with patient survival (Wald p < 0.05), with the top 5 and
   bottom 5 hazard ratio genes presented for clarity (Fig. [116]4D).
   Recognizing patient survival is unlikely to be regulated by a single
   gene, we conducted a Lasso Cox analysis of these genes to construct a
   model correlating gene expression with overall survival (Supplementary
   Fig. 4). The model assigned scores based on the expression of different
   genes, and higher scores were identified as a significant risk factor,
   evidenced by a Hazard ratio of 5.582 and p < 0.0001, indicating a
   strong association with patient survival, comparable to the impact of
   metastasis or relapse (Fig. [117]4E).

   Our clinical analysis, based on survival data, illuminates the pivotal
   prognostic role of P300/CBP related genes in ES, revealing a robust
   correlation between patient stratification according to distinct gene
   expression patterns and diverse overall survival outcomes. These
   findings underscore the compelling prognostic relevance of targeting
   P300/CBP in ES.

P300/CBP inhibition triggers senescence in ES

   Our findings, indicating iP300w treatment triggers rapid cell cycle
   arrest and morphological changes characteristic of senescent cells,
   coupled with KEGG pathway enrichment in Cellular Senescence, motivated
   us to directly evaluate the relationship between P300/CBP and
   senescence in ES. We performed β-galactosidase (βGal) staining, a
   widely recognized marker for cellular senescence, to assess senescence
   in ES cell lines. A notable increase in β-Gal-positive cells was
   detected in SKES1 and TC71 24 h after iP300w treatment, even in the
   absence of evident morphological changes at this stage (Fig. [118]5A;
   Supplementary Fig. 5A). Both cell lines exhibited a pronounced increase
   in β-Gal-positive staining 72 h after treatment. Concurrent Hematoxylin
   and Eosin (H&E) staining in SKES1 cells revealed significant
   morphological changes characteristic of senescence, such as enlarged
   and flattened cell structures, irregular shapes, and increased
   granularity (Fig. [119]5A; Supplementary Fig. 5A). To further validate
   our findings, we confirmed senescence induction by iP300w through β-Gal
   staining in additional EWS::FLI1-driven Ewing sarcoma cell lines,
   including CHLA25, A4573, and CHLA9 (Supplementary Fig. 5B). We then
   expanded our analysis to track early and late senescence-associated
   gene expression over a seven-day period in these treated cell lines.
   Although cell cycle gene alterations highlighted a significant impact
   on proliferation, it was surprising to observe that the early
   senescence markers P16 and P21 in most cases were not upregulated (Fig.
   [120]5B; Supplementary Fig. 5C). Upon broadening our panel of
   senescence-related markers, we noted a considerable decrease in LMNB1
   expression in all ES cell lines or xenograft tumors treated with iP300w
   (Figs. [121]2I and [122]5B). This effect was further confirmed at the
   protein level by immunofluorescence (Fig. [123]5C and D; Supplementary
   Fig. 5D). Interestingly, we observed that only the ES cell lines
   harboring functional P53 mutations (SKES1, TC71, CHLA25) responded with
   P15 induction upon iP300w treatment, whereas the cell lines with intact
   P53 (A4573 and CHLA9) did not show this response (Fig. [124]5B;
   Supplementary Fig. 5C). Remarkably, alterations in the expression of
   P15 was observed as early as 4 h into the treatment (Supplementary Fig.
   7A).

Fig. 5.

   [125]Fig. 5
   [126]Open in a new tab

   P300/CBP Inhibition triggers senescence in ES cancer cell lines.
   A β-Gal and H&E staining in SKES1 cells at 24 and 72 hours of
   treatments with iP300w. B Gene expression analysis shows changes in
   proliferation and senescence markers in SKES1 cells following 1, 3, and
   7 days of treatment with 1 μM iP300w. Data were normalized to B2M and
   are presented as fold change compared to the control group. The data
   represent mean ± SEM; p< 0.05, by one-way ANOVA (n=6). C Immunostaining
   shows a complete absence of LMNB1 (red) expression in iP300w-treated
   cells after 72 hours. Phalloidin (green) highlights cell morphology,
   and DAPI (blue) stains the nuclei. Scale bar: 50 µm. D Immunostaining
   for P15 in cells treated with iP300w for 72 hours. E RT-qPCR analysis
   of senolytics target gene expression following 24 and 72 hours of
   iP300w (1µM) treatment. Data were normalized to B2M and are presented
   as fold change compared to the control group. The data represent mean ±
   SEM; *p<0.05, by one-way ANOVA (n=6). F ATP assays show cell viability
   of SKES1 with different senolytics (1 μM) used alone or combined with
   iP300w (1 μM) following 72 hours of treatment. The data represent mean
   ± SEM; **p<0.01, ****p<0.0001 by one-way ANOVA. Results are presented
   as fold change compared to the control group (n=4). G Synergy scores
   between iP300w and various senolytics (Dasatinib, Fisetin, Navitoclax,
   and Quercetin) were calculated using the Highest Single Agent (HSA)
   model via SynergyFinder Plus. The 3D plots display synergy scores
   across different concentrations of senolytics (X-axis: 0 μM, 0.01 μM,
   0.1 μM, 1 μM, 10 μM, 50 μM) combined with iP300w (Y-axis: 0 μM, 1 μM).
   The Z-axis represents the synergy score, with positive values
   indicating synergy and negative values indicating antagonism. The
   strongest synergistic effects were observed with Dasatinib, while other
   senolytics demonstrated varying levels of synergy or antagonism. H FACS
   analysis of Annexin V staining (histogram) in SKES1 cells following 72
   hours of treatment with iP300w (1 μM), Dasatinib (1 μM), Dasatinib +
   iP300w, and Doxorubicin (1 μM). I Quantification of Annexin V staining.
   Data are presented as log10 fluorescence intensity. The data represent
   mean ± SEM; *p<0.05, ****p<0.0001 by one-way ANOVA (n=3)

   To determine whether P300/CBP inhibition also induces senescence in ES
   cell lines driven by EWS::ERG translocation, we extended our analysis
   to the CADO-ES1 and COG-E-352 cell lines. Initial results showed that
   iP300w treatment reduced cell viability in EWS::ERG cells, though this
   effect was less pronounced compared to EWS::FLI1 cell lines at 48 h, as
   measured by the ATP assay (Supplementary Fig. 6A). Notably, after
   extending the treatment to 72 h, we observed a complete loss of Ki67
   staining, indicating a profound effect on cell proliferation
   (Supplementary Fig. 6B). Similarly to the EWS::FLI1 cell lines, both
   CADO-ES1 and COGE352 lines underwent senescence following iP300w
   treatment. This was evidenced by changes in cell morphology, expression
   of senescence-related genes, and β-Gal positivity (Supplementary Fig.
   6C). Downregulation of LMNB1 was consistently observed as a hallmark of
   the senescence process (Supplementary Fig. 6C, E, F, H).

   We then reasoned that if senescence is indeed one of the initial
   responses to P300/CBP inhibition, senolytics might facilitate the
   clearance of iP300w-treated ES cells. Senolytics comprise a group of
   drugs that target various senescence-related pathways [[127]25]. To
   identify the most appropriate senolytic, we assessed the commonly
   affected PI3K and BCL-2 pathways by RT-qPCR. Gene expression analyses
   revealed a significant upregulation of PI3K, mTOR, MCL-1, and ERBB2,
   and a reduction of BCL-2 and SRC2 (Fig. [128]5E). RNA-Seq data
   suggested signs of these changes were apparent as early as 4 h
   following P300/CBP inhibition (Supplementary Fig. 7A). Hence, for our
   testing, we opted for Dasatinib and Fisetin, established senolytics
   targeting the PI3K pathways, while Navitoclax and Quercetin, known
   BCL-2 targets, served as negative controls [[129]25–[130]27]. We
   proceeded to assess the potential synergistic effects of these
   senolytics when combined with iP300w, testing each senolytic across a
   concentration range from 0 to 50 μM. Dasatinib displayed an IC50
   between 1–10 μM, while Fisetin and Navitoclax fell within the 20–50 μM
   range (Supplementary Fig. 7B, C). Quercetin showed no effectiveness
   even at the highest tested concentrations. Combining iP300w with these
   senolytics significantly reduced cell viability, as evidenced by the
   percentage fold change compared to the control group in the ATP assay
   (Supplementary Fig. 7B, C). Particularly noteworthy was the marked
   improvement observed with the combination of Dasatinib with iP300w that
   resulted in a 70% inhibition of cell viability at 1 μM concentrations
   for both agents (Fig. [131]5F; Supplementary Fig. 7D). This underscores
   a potent synergistic effect when iP300w is used alongside selected
   senolytics. To quantitatively assess the synergistic effects of
   combining iP300w with senolytics, we utilized SynergyFinder Plus. Our
   analysis revealed Dasatinib at 1 μM, when combined with 1 μM of iP300w,
   achieved the highest synergy score (Fig. [132]5G; Supplementary Fig.
   7E). After confirming that the combination of Dasatinib with iP300w
   exhibited the strongest synergistic effect, we proceeded to evaluate
   the ability of this combined treatment to induce apoptosis. While
   Dasatinib alone caused only minimal apoptosis in SKES1 cells, the
   addition of iP300w significantly enhanced this effect, achieving levels
   comparable to those induced by Doxorubicin (Fig. [133]5H, I;
   Supplementary Fig. 7F, G). Bringing our findings together, we concluded
   that ES undergoes senescence as an early response to P300/CBP
   inhibition, characterized by alterations in LMNB1 and PI3K expression.

Regulation of LMNB1 by the EWS::FLI1

   The noncanonical senescence pathway induced by iP300w in ES prompted
   further investigation into the interplay between LMNB1/P15 and
   EWS::FLI1/P300/CBP transcriptional axis. To explore this, we leveraged
   previously published high-resolution ChIP-Seq data obtained from
   studies examining the epigenetic effects of gain and loss of EWS::FLI1
   function in ES and MCS [[134]8, [135]16].

   Analysis of ChIP-Seq data revealed EWS::FLI1 binding in the regulatory
   elements of both LMNB1 and P15. The binding region in LMNB1 comprises
   consecutive GGAA repeats (more than 14), which are associated with
   EWS::FLI1 transcriptional activation, while the binding site in P15
   features a single GGAA motif that facilitates oncogene transcriptional
   repression (Fig. [136]6A). Notably, the EWS:FLI1 peak in LMNB1 is also
   associated with acetylated H3K27 and P300, indicating active chromatin.
   Upon EWS::FLI1 knockdown, the EWS:FLI1 binding is lost as expected,
   along with simultaneous loss of P300 binding and reduction of H3K27
   acetylation (Fig. [137]6A). The epigenetic alterations in the LMNB1
   enhancer are reflected in the transcriptional levels, as the gene is
   downregulated by EWS:FLI1 knockdown (Fig. [138]6B). Contrary to LMNB1,
   EWS::FLI1 binding of P15 was not accompanied by P300 and maintained low
   H3K27 acetylation. Upon EWS::FLI1 knockdown, the repression of the
   locus was lost, as indicated by increased acetylation and gene
   induction (Fig. [139]6A and B). To confirm EWS::FLI1 transcriptionally
   regulates LMNB1 through P300/CBP, we reassessed our data from
   experiments in which we knocked down both proteins (Fig. [140]1).
   Indeed, the transcription of LMNB1 upon P300/CBP knockdown mirrored
   those observed with EWS::FLI1 knockdown (Fig. [141]6C).

Fig. 6.

   [142]Fig. 6
   [143]Open in a new tab

   EWS::FLI1/ regulates LMNB1 and P15 and cellular senescence in ES.
   A ChIP-Seq analysis reveals that EWS::FLI1 binds to regulatory elements
   of LMNB1 and P15 in the SKNMC cell line [[144]8]. Specifically,
   EWS::FLI1 targets the repetitive GGAA motif in the regulatory elements
   of LMNB1 and a single motif in P15. EWS::FLI1 binding sites in LMNB1
   are associated with P300 and acetylation of H3K27. B Transcriptional
   levels (RNA-Seq) of LMNB1 and P15 following EWS::FLI1 knockdown (96
   hours) reveals reduction of LMNB1 and induction of P15. C RT-qPCR for
   LMNB1 and P15 expression 48 hours post P300/CBP knockdown. Data were
   normalized to B2M and are presented as fold change compared to the
   control group. The data represent mean ± SEM; p<0.05 by t-test (n=4).
   D ChIP-Seq data shows EWS::FLI1 binding in LMNB1 and P15 regulatory
   elements in MSC overexpressing EWS::FLI1 [[145]16]. E LMNB1 and P15
   expression (RNA-Seq) in MSC overexpressing EWS::FLI1 and GFP control
   cells. Data are presented as mean expression levels. The data represent
   mean ± SEM; p<0.05 by t-test (n=4)

   Overexpression of EWS::FLI1 in non-permissive cells leads to induction
   of cell cycle arrest and cell death [[146]14]. Therefore, we
   hypothesized the opposite mechanism might apply when EWS::FLI1 is
   expressed in a permissive environment. To investigate this, we
   retrieved ChIP-Seq data in which EWS::FLI1 was overexpressed in MSC to
   promote malignant transformation [[147]16, [148]28]. As predicted,
   EWS::FLI1 overexpression resulted in EWS::FLI1 binding to the
   repetitive GGAA motif, facilitating H3K27 acetylation and LMNB1
   expression (Fig. [149]6D and E). Additionally, EWS-FLI1 binds to the
   repressive single GGAA motif, leading to reduced H3K27 acetylation at
   the P15 locus (Fig. [150]6D). In summary, our findings demonstrate the
   EWS::FLI1/P300/CBP axis directly regulates the expression of key
   senescence-related genes LMNB1, promoting cell viability and preventing
   senescence in Ewing sarcoma and permissive cells.

   To further investigate whether EWS::FLI1 and EWS::ERG-mediated
   senescence evasion mechanisms operate independently of the canonical
   TP53 and P16/P21 pathways, we examined the binding sites of these
   oncoproteins at their regulatory regions in SKNMC, SKES1, and TC32
   cells. Our analysis revealed that both EWS::FLI1 and EWS::ERG
   consistently bind only to LMNB1, with no direct interactions observed
   with TP53, P16, or P21. This suggests that any transcriptional activity
   of P16, P21, and TP53 might be independent of these oncoproteins
   (Supplementary Fig. 8A-E). Additionally, the previously observed
   binding of EWS::FLI1 to the repressive motif of P15 in SKNMC was also
   identified in SKES1 cells, both of which have nonfunctional P53.
   Notably, EWS::FLI1 binding to P15 is not associated with P300 peaks,
   indicating a P300/CBP-independent regulatory mechanism (Supplementary
   Fig. 8A). In summary, we demonstrate that both EWS::FLI1 and EWS::ERG
   cooperate with P300 to maintain LMNB1 expression and evade
   senescence-related processes in Ewing sarcoma.

LMNB1 and P15 suppress apoptosis and sustain iP300w induced senescence

   To assess the functional relevance of iP300w induced senescence
   mediated by LMNB1/P15 in ES cells with non-functional P53, we conducted
   gene rescue experiments. Our assumption was that simultaneous
   overexpression of LMNB1 and knockdown of P15 would delay the
   senescence. To achieve constitutive LMNB1 overexpression, SKES1 cells
   were transduced with a lentiviral vector carrying LMNB1, while P15 was
   knocked down using siRNA (Supplementary Fig. 9A). β-Gal staining
   revealed a significant induction in 21% of the cells by 24 h of iP300w
   treatment, a process notably reduced to 4% upon overexpression of LMNB1
   and knockdown of P15 (Fig. [151]7A, B).

Fig. 7.

   [152]Fig. 7
   [153]Open in a new tab

   The restoration of LMNB1 and P15 expression in iP300w-treated cells
   delays senescence and induces apoptosis. A β-Gal staining was performed
   on SKES1 cells treated with iP300w (1 µM) for 48 hours, as well as on
   control untreated cells. LMNB1 refers to SKES1 cells that
   constitutively overexpress LMNB1 from a viral construct, while siP15
   denotes SKES1 cells with P15 knocked down. LMNB1+siP15 refers to the
   condition where LMNB1 is expressed and P15 is knocked down for 48
   hours. B Quantification of β-Gal positive cells presented in “A”. The
   data represent mean ± SEM; *p<0.05, ***p<0.001 by one-way ANOVA (n=3).
   C Annexin V staining (histogram) of SKES1 cells in the condition
   described above. D Quantification of intensity of Annexin V staining.
   The data represent mean ± SEM; *p<0.05, ****p<0.0001 by one-way ANOVA
   (n=3)

   Senescent cells evade apoptosis through a variety of mechanisms, such
   as enhancing anti-apoptotic pathways, modifying pro-survival signaling,
   and regulating apoptotic regulators, enabling them to persist in a
   senescent state [[154]29]. We conducted Annexin V staining to evaluate
   the apoptotic status in LMNB1/P15 rescued cells. As anticipated, SKES1
   cells treated with iP300w exhibited a significant reduction in
   fluorescence signal and a marked decrease in Annexin V-positive cells
   (Fig. [155]7C, D; Supplementary Fig. 9B, C). However, following LMNB1
   overexpression and P15 RNA interference, both the fluorescence
   intensity and the proportion of Annexin V-positive cells were partially
   reversed (Fig. [156]7C, D; Supplementary Fig. 9B, C). These functional
   experiments provide direct evidence that LMNB1 and P15 play a critical
   role in iP300w-induced senescence.

Discussion

   This study unveils the critical mechanism of the EWS::FLI1/P300/CBP
   axis in maintaining ES viability and exposes its vulnerability for
   pharmacological targeting by the P300/CBP inhibitor, iP300w.
   Additionally, we elucidate a novel mechanism by which EWS::FLI1
   prevents ES or permissive MSC from undergoing senescence by controlling
   the expression of LMNB1. EWS::FLI1, together with P300/CBP induce LMNB1
   transcription by binding to multiple GGAA repeats in its regulatory
   elements. P300/CBP inhibition rapidly inactivated EWS::FLI1
   transcriptional activity, leading to the downregulation of LMNB1,
   consequently inducing cell senescence (Fig. [157]8). Finally, we
   demonstrate that PI3K-targeting senolytics effectively eliminate
   iP300w-induced senescent ES cells, suggesting a novel, combined
   therapeutic approach.

Fig. 8.

   [158]Fig. 8
   [159]Open in a new tab

   The proposed mechanism by which EWS::FLI1 regulates LMNB1 in ES cancer
   cell lines. In proliferative ES cells, the EWS::FLI1/P300 complex
   induces LMNB1 by binding to its enhancer motif containing 14 GGAA
   repeats. Inhibition of P300/CBP by iP300w treatment suppresses
   EWS::FLI1 transcriptional activity, leading to reduced LMNB1
   transcription. Consequently, SA-β-Ga and other early senescence markers
   are induced. In a subset of ES cancer cell lines with non-functional
   P53, EWS::FLI1 binds to a single GGAA motif in the regulatory elements
   of P15, independent of P300, and represses its transcription. Following
   treatment with iP300w and subsequent LMNB1 suppression, P15 is rapidly
   induced, contributing to the induction of senescence in ES cells

   While prior studies have shown EWS::FLI1 can recruit P300 to
   differentially regulate super enhancers [[160]8], the role of P300/CBP
   in ES viability remains undefined [[161]15]. To explore this, we
   initially knocked down P300/CBP using siRNA and demonstrated that the
   oncogenic activity of EWS::FLI1 largely depends on these
   acetyltransferases. Prompted by our initial results and recognizing the
   translation limitations of the siRNA approach [[162]30], we
   investigated pharmacological targeting of P300/CBP, capitalizing on the
   new generation of more specific spirocyclic P300/CBP inhibitors,
   including iP300w [[163]19, [164]31]. We initially described iP300w as
   an effective inactivator of DUX4, a pioneer transcriptional factor that
   recruits P300/CBP through its C-terminal domain and plays a key role in
   early embryonic development and facioscapulohumeral muscular dystrophy
   (FSHD) [[165]31, [166]32]. The translocation of the DUX4 C-terminus
   transforms CIC from a transcriptional repressor to an activator,
   thereby driving the development of CIC::DUX4 sarcoma. Notably, iP300w
   displayed superior efficacy against CIC::DUX4 sarcoma compared to A-485
   [[167]18], another P300/CBP inhibitor from the same spirocyclic
   chemical series found to be effective against certain hematological
   malignancies, androgen receptor-positive prostate cancer, and NUT
   midline carcinoma [[168]19, [169]33]. Likewise, iP300w demonstrated
   high efficacy against ES, evidenced by rapid induction of cell arrest
   and inactivation of EWS::FLI1 transcriptional activity, comparable to
   the effects seen with knockdown of EWS::FLI1 [[170]23, [171]34]. The
   observed effect was specific to ES cells, with minimal reduction in
   cell viability noted in human myoblasts and osteosarcoma cell lines at
   the tested concentrations. Furthermore, the extent of global
   transcriptional alterations in ES cell lines treated with iP300w
   surpassed that observed in synovial sarcoma, rhabdomyosarcoma, or
   osteosarcomas. Our findings underscore the pivotal role of P300/CBP in
   ES malignancy, as evidenced by a robust correlation between patient
   stratification based on specific EWS::FLI1/P300/CBP-related gene
   expression patterns and overall survival rates. Notably, this
   correlation demonstrates greater predictive power for survival outcomes
   than traditional clinical features [[172]35]. Similarly, elevated
   levels of P300 have been identified as indicators of high malignancy
   and poor prognostic survival factors in liver, nasopharyngeal, and
   prostatic cancers [[173]36–[174]38].

   Despite extensive research over the decades, treatment options for
   Ewing sarcoma) remain limited. FDA-approved drugs such as vincristine,
   doxorubicin, and cyclophosphamide show some effectiveness but lack
   specificity for the EWS::FLI1-driven malignancy [[175]39–[176]41].
   These are drugs that target DNA replication and cell proliferation,
   leading to unintended effects on healthy cells. Ifosfamide, another
   commonly used chemotherapeutic agent for ES, is an alkylating agent
   that cross-links DNA strands [[177]42]. It is associated with
   significant side effects, including renal toxicity, cardiotoxicity, and
   neurotoxicity. Etoposide and topotecan, both topoisomerase inhibitors,
   are used in the treatment of various cancers, including ES, but their
   use can also lead to serious side effects such as secondary leukemias
   and myelosuppression [[178]43, [179]44]. Other drugs that have shown
   some effectiveness for ES, when used in combination with other
   treatments, include irinotecan, actinomycin D, regorafenib,
   clofarabine, cladribine, and trabectedin [[180]40, [181]45–[182]48].
   Recently, efforts to target the EWSR1::FLI1 fusion or its downstream
   pathways have been largely unsuccessful. For instance, TK-216, a
   pharmacological agent designed to disrupt the interaction between
   EWSR1::FLI1 and RNA helicase A, demonstrated responses in 3 of 85
   patients in a phase I/II clinical trial, with an overall disease
   control rate of 25% and a median duration of response of 25.5 months
   [[183]49]. Similarly, trabectedin, a DNA minor groove-binding agent,
   has demonstrated minimal clinical activity as a single agent despite
   its effects on EWSR1::FLI1 transcriptional targets [[184]50].
   Consequently, there is a critical need for therapies capable of
   specifically targeting EWS::FLI1 activity. The current study
   demonstrates that inhibiting P300/CBP can inactivate the EWS::FLI1
   oncogenic axis, making it a rationally targeted pharmacological
   approach for treating ES.

   Pathway analyses of differentially expressed genes alerted by iP300w
   revealed activation of the senescence pathway early in the treatment.
   The senescence phenotype was further evidenced by altered cell
   morphology, reduced cell proliferation, β-Gal positive staining, and
   changes in gene expression related to senescence. Senescence induced by
   P300/CBP inhibition, whether via small molecules or knockdown, has been
   previously observed in various cancers, such as melanoma, non-small
   cell lung cancer, and breast cancer [[185]51–[186]53]. This effect is
   primarily attributed to global changes of chromatin acetylation, with
   subsequent impact on the regulatory elements of senescence-related
   genes, among others [[187]54]. Conversely, recent reports indicate P300
   inhibition prevents age-related cellular senescence, primarily driven
   by P300, inducing a dynamic, hyper-acetylated chromatin state and
   promoting the formation of active enhancer elements in the non-coding
   genome [[188]55]. The rapid induction of senescence in ES by iP300w
   suggests a more specific interplay. This prompted further mechanistic
   exploration, uncovering a novel mechanism by which EWS::FLI1 directly
   and concurrently regulates the expression of LMNB1. Surprisingly, the
   classical senescence pathways typically mediated by P16 or P53/P21 were
   not affected by iP300w in ES [[189]56, [190]57].

   EWS::FLI1 exhibits distinct binding preferences depending on the