Abstract Sequential changes in transcriptional cell state are essential for normal development and are coopted in cancer. Controlling these changes for therapeutic benefit has been limited due to a lack of tools that reflect different states in live cells. Here, we describe a new reporter method termed “TRECS” that integrates epigenomic and transcriptomic measurements to define endogenous genomic elements that label individual cells in different cell states. We use TRECS in the neural crest-derived, high-risk pediatric solid tumor neuroblastoma, where we demonstrate broad presence of cells with distinct transcriptomes, associated with functional chemoresistance and sensitivity. Experiments using neural crest stem cells and TRECS mouse knock-in model identified that TRECS-labelled cells reflect early developmental stages in the neural crest. These cells display real-time plasticity of transcriptional state and phenotype, in a manner unlinked to cell cycle control. Investigation of nominated loci demonstrates state-specific enrichment of elements marked by H3K27ac, H3K4me1 and open chromatin by ATAC-seq, which flexibly change as cells transition between these phenotypically divergent states. To investigate whether the primary nominated locus is a driver or reporter of cell state, we integrate micro-C, transcriptomics, truncation experiments and functional CRISPRi to identify that this region functions as a pure endogenous reporter of cell state. This, therefore, provides a mechanism to identify new, state-controlling transcription factors. Motif analysis demonstrated enrichment of AP1 transcription factor motifs in the chemoresistant state, and knockout of these AP1 transcription factors results in rewiring of cell state and enhanced chemosensitivity without effects on cell growth. To capitalize on the endogenous flexibility in this system and identify mechanisms to enforce cell state changes independent of cell growth and death, we performed high-content image-based small molecule screening to identify targets suitable to enhance chemosensitivity. These experiments identified EP300/CBP, master histone acetyltransferases, as crucial controllers of a primitive, chemoresistant cell state. Transient acetyltransferase and bromodomain-based inhibition of EP300/CBP results in transcriptional and epigenetic reprogramming in vitro and in vivo, leading to enhanced chemosensitivity and prolonged survival in murine models. These results demonstrate an unbiased method to identify non-coding genomic loci enriched in specific cell states, which can be harnessed to identify master transcription factors driving these cell states and similarly, mechanisms to enforce changes in cell state. Citation Format: Noha Shendy, Yang Zhang, Stephanie Nance, Ha Won Lee, Shivendra Singh, Yousef Khashana, Vernon Ebegboni, Estevez Prado Daniel, Mohammad Ali Mohammad Nezhady, K. Elaine Ritter, Anoop Kavirayani, Bensheng Ju, Grace McKay-Corkum, Qi Liu, Yiping Fan, Gang Wu, Jun Qi, John B. Easton, Anand G. Patel, JUN YANG, Taosheng Chen, Brian Abraham, Adam D. Durbin. A genome derived non coding reporter of dynamic cancer cell state [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 385.

Despite comprehensive and multi-modal therapy, outcomes for children and adolescents with rhabdomyosarcoma (RMS) have plateaued over the past four decades. This is not for a lack of progress in the basic and translational studies of RMS. Indeed, advances in animal models and/or patient tissue sample acquisition and analysis have improved our understanding of RMS biology. Large-scale sequencing efforts have generated transcriptomic, genomic, and epigenomic datasets that highlight the heterogeneity of RMS and have the potential to improve prognostication and the application of precision medicine in patients with RMS. However, few of these discoveries have been clinically translated, and limitations to the accessibility, uniformity, and application of these new models and datasets hinder their utility. Here, we discuss how advances in understanding RMS biology, optimization of preclinical models, and strategies for translating basic science discoveries to the clinic can potentially improve outcomes for patients with RMS.

Abstract Cancer cells are dependent on the control of transcription for maintenance of the malignant cell state. EP300 and CBP are paralogous, commonly expressed, master epigenetic enzymes, whose activity controls normal and malignant transcription. Here, using a cancer-wide integrative chemical-genetic analysis, we find enhanced dependency on EP300 compared to CBP in most cancer lineages, including osteosarcoma (OS). OS is a highly lethal malignancy of bone and has enhanced dependency on EP300, compared with CBP. To take advantage of selective pharmacology targeting EP300 alone that spares CBP in untransformed cells and thereby reduces toxicity, we used the EP300-targeted degrader JQAD1. We identify a specific genetic subgroup of OS, marked by dysregulation of the PI3K-AKT-mTOR pathway, that is both genetically dependent on EP300 and uniquely sensitive to EP300 degradation. Expression of constitutively active AKT in insensitive OS cells induces sensitivity to JQAD1, driven by physical relocalization of EP300, CBP and H3K27ac to genetic subtype-enriched dependency loci. Mechanistically, combinations of AKT inhibitors and JQAD1 suppress the growth of AKT-dysregulated OS. These observations extend across >850 cancer cell lines, where multiple AKT inhibitors, including the FDA-approved capivasertib, positively combine with EP300 degraders in a manner mechanistically dependent on control of protein synthesis. Finally, combination treatment is synergistic in AKT-dysregulated, orthotopic OS xenografts. These findings reveal genetic and transcriptional determinants of EP300 degrader function in high-risk OS and provide a foundation for biomarker-directed investigation of co-targeting of EP300 and AKT for therapeutic gain across cancers marked by enhanced protein translation. Citation Format: Ian Delahunty, Stephanie Nance, Yang Zhang, Francois Lamoureux, Qi Liu, W. Charlie Wright, K. Elaine Ritter, Noha A. Shendy, Benedicte Brounais, Sandra Kietlinska, Barbara De Kegel, Matthew G. Rees, Mustafa Kocak, Ashish B. George, Anoop M. Kavirayani, Paris P. Prinsen, Yousef Khashana, Melissa M. Ronan, Jennifer Roth, Alejandro Sweet-Cordero, Xiaotu Ma, Lillian M. Guenther, Paul Geeleher, Benjamin Ory, Jun Qi, Brian Abraham, Adam D. Durbin. Dysregulated AKT signaling reprograms osteosarcoma to drive selective reliance on EP300 [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 4024.

The third in a series of Paediatric Therapeutic Development Workshops focused on rhabdomyosarcoma. Rhabdomyosarcoma is the most common soft tissue sarcoma in children, with 90% survival for those with the lowest risk disease, but just 20-30% in children with metastatic disease. An urgent unmet need exists to develop targeted therapeutics for high-risk disease and to reduce the toxicity of treatment. The results of trials of CAR T-cells and ADCs against FGFR4 are awaited with great interest, and developing a FGFR4 degrader is a priority. Directly targeting PAX3::FOXO1 and PAX7::FOXO1 fusion proteins is a high priority. An in vivo study of a MYOD1 degrader approach is required prior to clinical development. Degraders of P300/CBP should be evaluated preclinically with a view to clinical investigation. ROR2 is an interesting target for the L122R mutant MYOD1 rhabdomyosarcoma. Development of a TEAD degrader is a high priority, and this should be evaluated in combination with a Notch inhibitor. Considering targets with existing clinical agents, antibody-drug conjugates targeting cell-surface antigen B7-H3/CD276 are showing preclinical promise in other paediatric cancers and are also deemed a high priority for evaluation in rhabdomyosarcoma. Based on currently available evidence, MEK inhibitors should be evaluated, potentially with BRAF or PI3K inhibitors, in combination with chemotherapy in the maintenance setting. Understanding the mechanism of action underpinning drug combinations, gaining access to therapeutics and optimising clinical trial design will be essential to enable combinatorial testing in patients.

The second Paediatric Therapeutic Development Workshop focused on medulloblastoma. Between 60-70% of patients with medulloblastoma survive, but survivors have significant long-term side effects, and the highest-risk groups have a probability of survival <10%. Thus, the unmet need is to develop therapeutics targeting specific vulnerabilities in medulloblastoma including poor prognosis disease groups (SHH-medulloblastoma, MYCN amplified or TP53 mutated; and Group 3 medulloblastoma, c-MYC amplified) and developing less-toxic therapies for good prognosis disease (WNT-medulloblastoma). The Workshop concluded that (i) targeting SRC by a degrader is a high priority, (ii) inhibition of c-MYC and MYCN tumour-relevant functions for poor prognosis groups is a priority, (iii) targeting WNT-medulloblastoma via a radiolabelled theranostic antibody is an innovative approach for good prognosis tumours to further reduce toxicity, and (iv) B7-H3 has many advantages for CAR T-cell and ADC-based approaches. Based on currently available evidence, combinations of central nervous system penetrant selective PARP-1, CHK1/2 or CDK9 inhibitors with an ATR inhibitor could potentially be evaluated in early-phase trials for high-risk patients; however, these combinations require robust evaluation in pre-clinical models first. Early-phase clinical studies should be international, have novel designs to address small patient numbers and based on an understanding of biology with correlative biological studies. Both developing therapeutics targeting specific vulnerabilities in medulloblastoma and evaluating combinations of existing medicinal products are required to improve outcome and reduce long term sequalae.

Abstract Rhabdomyosarcoma (RMS) is a high-risk and lethal pediatric sarcoma that resembles developing skeletal muscle. RMS tumors have low mutation burdens, but these scant mutations often alter genes involved in transcriptional control. Transcriptional dysregulation is critical to RMS pathogenesis, supported by studies in both RMS tumors carrying mutationally derived chimeric transcription factors (“fusion positive (FP)”), or those without (“fusion negative” (FN)). However, mechanisms to selectively target dysregulated transcription in RMS remain outstanding. Here, we develop a novel approach targeting RMS transcription comprising simultaneous targeting of two distinctly acting transcriptional co-activators. We discover a common cell identity-controlling pan-RMS core regulatory circuit (CRC) composed of oncogenic and lineage-specific myogenic master transcription factors (mTFs). These mTFs are regulated by super-enhancers, and they co-bind genome-wide to control the malignant transcriptome of both FP- and FN-RMS. Using a super-enhancer-based reporter screen, we identify the EP300/CBP inhibitor A485 as a potent inhibitor of the pan-RMS CRC, though efficacy of this compound was limited by toxicity. To enhance on-target specificity, we identify the protein EYA2 as a co-factor that binds directly to SIX1, a member of the pan-RMS CRC and exploit a recently developed second-generation EYA1/2 tyrosine phosphatase inhibitor, LG1-34, to inactivate its function. While A485 and LG1-34 independently reduce mTF transcription and drive RMS cell death, in combination, these agents function synergistically to reduce RMS growth in vitro and in vivo. These results demonstrate that combined targeting of enhancer maintenance and CRC cofactors is a powerful strategy to suppress the RMS transcriptome and enforce RMS cell death. Citation Format: Annika Gustafson, Stephanie Nance, Berkley Gryder, Noha A. Shendy, Lars Wick, Grace McKay-Corkum, K. Elaine Ritter, Stephen Connor Purdy, Arthur R. Wolin, Sheera R. Rosenbaum, Sabateeshan Mathavarajah, Nickerson A. Demelfi, Yueyang Wang, Yang Zhang, Mark W. W. Zimmerman, Anoop M. Kavirayani, John Hardin, Alexander LaVeck, Xiang Wang, Neekesh V. Dharia, Andrew Hong, Guillaume Kugener, Jesse S. Boehm, Jennifer Roth, Javed Khan, Francisca Vasquez, Kristin B. Artinger, Rui Zhao, David M. Langenau, Jun Qi, Kimberly Stegmaier, Heide L. Ford, Adam D. Durbin, Brian J. Abraham. Synergistic targeting of EP300/CBP and EYA co-activators collapses the rhabdomyosarcoma core regulatory circuit [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 3500.

Abstract High-risk Neuroblastomas (NB) carrying in-frame fusion mutations in the ATRX chromatin remodeler(“ATRX-IFF”) are chemoresistant, and children with these mutations display poor overall survival. There are no specific agents to target these tumors, and the mechanisms driving chemoresistance remain unknown. To gain insight into the mechanisms of chemoresistance, we used scRNAseq in combination with cisplatin to study if cellular plasticity drives chemoresistance. In contrast to non-ATRX-IFF-bearing NBs, ATRX-IFF NBs did not appear to display transcriptionally plastic subpopulations, suggesting that the ATRX-IFF promotes an inherent chemoresistance program. To this end, CRISPR-mediated engineering of ATRX-IFF into the endogenous locus in wildtype cells, results in enhanced chemoresistance to multiple clinically-used agents. Next, to identify potential targetable proteins directly regulated by the ATRX-IFF, we combined chromatin binding assays with cell surface proteomics. Using cell lines and orthotopic PDXs, we identified that the ATRX-IFF binds to, and nucleates a super-enhancer at the PTK7 locus, thereby driving extremely high level cell surface expression of PTK7. PTK7 is a pseudokinase receptor that is targeted by experimental CAR-T cell therapies in active preclinical development. Ongoing work is aimed at understanding the dependency of PTK7 on the ATRX-IFF, and using preclinical CAR-T cell models to target PTK7 in ATRX-IFF NB. Citation Format: Mohammad Ali Mohammad Nezhady, Sheetal Bhatara, Siarhei Hladyshau, Estevez Prado Daniel, Vernon Ebegboni, Alja Kozulic-Pirher, Arnav Barpujari, Bo Wang, Yuan Feng, Qiqi Jin, Xueying Liu, Paul Geeleher, Xiaotu Ma, Jiyang Yu, Emily Bernstein, Kelly Goldsmith, Hunter Jonus, Adam D. Durbin. ATRX in-frame fusions promote endogenous chemoresistance programs but yield immunotherapeutic vulnerabilities in neuroblastoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 2995.

Abstract A major limitation in cancer treatment is chemoresistance, particularly at relapse. Previously, relapse was thought to be driven by chemoresistance arising from new acquired genetic mutations; however, emerging evidence indicates that cancer cells exploit non-genetically-driven processes such as epigenetically regulated transcriptional plasticity to drive chemoresistance, invasion and proliferation. High-risk neuroblastoma (NB) demonstrates transcriptional plasticity with defined cell states and low mutational burden. NB cells exist primarily in two distinct states: a chemosensitive adrenergic state (ADRN) and a less common drug-tolerant persister cell state, the mesenchymal (MES) state. MES cells are enriched at relapse, suggesting that cells switch to this state under therapeutic pressure. Despite a myriad of approaches to treating high risk NB, relapsed patients have exceptionally poor survival. Unfortunately, mechanisms maintaining cell state and permitting state switching in NB are poorly understood. We hypothesize that dissection of the pathways promoting NB cell state plasticity will reveal new approaches to drive cell state interconversions that facilitate chemosensitivity. Thus, here, we took advantage of a recently developed novel fluorescent reporter system of the NB MES cell state developed by the Durbin Lab, to perform whole exome-CRISPR-cas12 knockout screening and identify master controllers of cell state maintenance in ADRN and MES-dominant cell lines. Integrated pathway analysis demonstrated conserved pathway modules involved in maintenance of cells in distinct cell states. To validate these findings, we performed targeted CRISPR knockouts in reporter carrying cells and identified changes in chemoresistance by chemosensitivity assays. To profile the effects of chemotherapy directly on cell state maintenance, we performed fluorescence reporter assays, followed by RNAseq to confirm changes in target loss nominated from CRISPR screens. Cell state switching is an intriguing paradigm by which NB cells may evade conventional therapies. We have identified pathways playing a role in maintaining either the ADRN or MES cell state in NB. Continued interrogation of these fundamental mechanisms represents a new approach that may potentially be leveraged for therapeutic gain. Citation Format: Grace McKay-Corkum, Stephanie Nance, Noha Shendy AM, Shilpa Narina, Shondra Miller, Alex Carisey, Qiqi Jin, Jiyang Yu, Adam D. Durbin. Exome-scale CRISPR screening reveals master controllers of cell state maintenance in high-risk neuroblastoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 1782.

Selective degradation of a disease-associated protein of interest (POI) is a powerful therapeutic strategy. FDA-approved and investigational glue and degrader drugs function by recruiting a POI to an E3 ubiquitin ligase that mediates POI polyubiquitination and triggers proteasomal degradation. However, using E3 ligases as an intermediary and requiring POI polyubiquitination makes this mechanism of action complex and difficult to rationalize and optimize. These issues have led to interest in evaluating whether direct recruitment of non-ubiquitinated POIs to the proteasome might achieve the same pharmacological outcome. Here, we examined the potential of direct-to-proteasome non-ubiquitinated POI recruitment. Using a tag strategy, we first demonstrated that the proteasomal 19S cap region proteins, RPN13 and RPN1, can recruit non-ubiquitinated POI model proteins, such as BRD4, to the proteasome and induce degradation. Subsequently, we developed small molecule-based bifunctional recruiter molecules (Proteasome Cap Targeting Chimeras, CAP-TACs) and showed that they recruit several distinct POIs, including BRD4, PRMT5, and FKBP12, to specific subunits in the 19S cap region and induce their ubiquitination-independent, proteasome-dependent degradation. This study provides further evidence that bifunctional small molecules can re-localize POIs to the proteasome and induce their degradation in the absence of ubiquitination, which broadens the capabilities of targeted protein degradation.

ZFTA-RELA is the most recurrent genetic alteration seen in paediatric supratentorial ependymoma (EPN) and is sufficient to initiate tumours in mice1. Despite its oncogenic potential, ZFTA-RELA (ZR) is observed nearly exclusively in childhood EPN, with tumours located distinctly in the supratentorial brain of the central nervous system1. We proposed that specific chromatin modules accessible during brain development would render distinct cell lineage programs at direct risk of transformation by ZR. To test this hypothesis, we performed combined single-nucleus assay for transposase-accessible chromatin and RNA (snMultiome) sequencing of the developing mouse forebrain compared with ZR-driven mouse and human EPN. We demonstrated that specific developmental lineage programs present in transient progenitor cells and regulated by PLAG/L family transcription factors were at risk of neoplastic transformation. Binding of this chromatin network by ZR or other PLAG/L family motifs targeting fusion oncoproteins led to persistent chromatin accessibility at oncogenic loci and oncogene expression. Cross-species analysis of mouse and human ZR EPN revealed significant cell type heterogeneity indicating incomplete neurogenic and gliogenic differentiation, with a small percentage of cycling progenitor-like or radial glial-like cells that established a putative tumour cell hierarchy. In vivo lineage tracing studies identified neoplastic clones that aggressively dominated tumour growth and established the entire EPN cellular hierarchy. These findings identify developmental epigenomic states that are critical for fusion-oncoprotein-driven transformation and show how these states continue to shape tumour progression.

Neuroblastoma is a malignancy of the peripheral sympathetic nervous system, accounting for 15% of childhood cancer deaths. Despite multi-modal treatment with systemic chemotherapy, surgery, radiation, two cycles of high-dose chemotherapy requiring stem cell rescue, and immunotherapy with anti-GD2 antibody treatment combined with retinoic acid, overall survival for patients is only about 50% and treatment -associated toxicity is immense. Thus, less toxic and more effective treatment options are desperately needed. Retinoic acid is clinically used to induce growth arrest of neuroblastoma cells and epigenetic rewiring of the enhancer landscape, leading to suppression of oncogenic MYCN expression. However, the epigenetic and growth suppressing effects of retinoic acid are completely reversible, leading to tumor re-growth. Here, we sought to identify epigenetic modifiers that enhance the antiproliferative effects of retinoids in neuroblastoma. In a drug screen with 452 compounds targeting epigenetic modifiers, we identified PF-9363, an inhibitor of the histone H3K23 acetyltransferases KAT6A/B, as synergistically enhancing neuroblastoma growth arrest in combination with retinoic acid, with a BLISS synergy score of 17 when both compounds are administered at 1µM each. Importantly, the combination treatment with retinoic acid plus PF-9363 resulted in durable growth arrest of neuroblastoma, which persisted beyond retinoid withdrawal in in vitro models, in contrast to single retinoid treatment. Durable growth arrest was due to sustained downregulation of MYCN and the essential transcription factors PHOX2B and GATA3, which was caused by gene expression silencing through deposition of H3K27me3 across their regulatory elements. In xenograft models of neuroblastoma in NSG mice, the KAT6A/B inhibitor PF-9363 alone demonstrated significant anti-tumor activity when given at 5 mg/kg daily by oral gavage and tumor volumes were significantly different compared to mice treated with vehicle starting at day 4 (p<0.01) through the end of treatment on day 28 (p<0.01). Moreover, the combination treatment with retinoic acid and PF-9363 also induced significant growth suppression starting from day 4 onward, compared to the vehicle control (p<0.01), and exhibited the most sustained growth suppressing effect among all treatment conditions even after retinoic acid was withdrawn on day 14, which was significantly different from retinoid treatment alone (p<0.05), where drug withdrawal led to rapid tumor proliferation. Furthermore, expression of GD2, a cell surface target for antibody therapy, was induced on GD2low neuroblastoma through the treatment with retinoic acid plus PF-9363 after 14 days of treatment determined by flow cytometry in vitro and immunofluorescence in vivo, rendering neuroblastoma cells more susceptible for anti-GD2 immunotherapy. Our studies nominate KAT6A/B inhibition as a novel approach to enhance the effectiveness of differentiation and immunotherapies in neuroblastoma and may be relevant for other tumors of neuro-ectodermal origin. Citation Format: Mark Zimmerman, Adam Durbin, Brian Abraham, Thomas Look, Francesca Alvarez-Calderon, Ulrike Gerdemann, Nina Weichert-Leahey. Treatment of neuroblastoma with retinoic acid plus a KAT6A/B inhibitor induces sustained growth arrest and increases GD2 expression as a target for immunotherapy [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Functional and Genomic Precision Medicine in Cancer: Different Perspectives, Common Goals; 2025 Mar 11-13; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2025;85(5 Suppl):Abstract nr A020.

Neuroblastoma is a malignancy of the peripheral sympathetic nervous system, accounting for 15% of childhood cancer deaths. Despite multi-modal treatment regimens and post-consolidation therapy, including anti-GD2 antibody treatment combined with retinoic acid, overall survival for patients with high-risk disease is only ∼50%, and treatment-associated toxicity is immense. Thus, less toxic and more effective treatment options are desperately needed. Retinoic acid is clinically used to induce growth arrest of neuroblastoma cells and epigenetic rewiring of the enhancer landscape, leading to suppression of oncogenic MYCN expression. However, the epigenetic and growth-suppressing effects of retinoic acid are completely reversible, leading to tumor re-growth. Here, we sought to identify epigenetic modifiers that enhance the antiproliferative effects of retinoids in neuroblastoma. In a drug screen with 452 compounds targeting epigenetic modifiers, we identified PF-9363, an inhibitor of the histone H3K23 acetyltransferases KAT6A and B, as synergistically inhibiting neuroblastoma growth in combination with retinoic acid. Importantly, this growth suppression was durable and persisted beyond retinoid withdrawal in in vitro models, in contrast to retinoid treatment alone. Durable growth suppression was due to sustained downregulation of MYCN and the essential transcription factors PHOX2B and GATA3, caused by gene expression silencing through PRC2-dependent deposition of H3K27me3 across their regulatory enhancers. In NSG xenograft models of neuroblastoma, the KAT6A/B inhibitor PF-9363 alone demonstrated significant anti-tumor activity when given at 5 mg/kg daily by oral gavage and tumor volumes were significantly different compared to vehicle treated mice starting at day 4 through the end of treatment on day 28. Moreover, the combination treatment with retinoic acid and PF-9363 exhibited profound and sustained growth suppression compared to vehicle treated mice and exhibited the most sustained growth suppressing effect among all treatment conditions, persisting even after retinoic acid was withdrawn. This outcome was significantly different from retinoid treatment alone, where drug withdrawal resulted in rapid tumor proliferation in NSG mice. Reflecting the epigenetic reprogramming of these tumor cells, we observed that expression of GD2, a cell surface glycosphingolipid that is a target of antibody-based and CAR T cell therapies in neuroblastoma, was induced on GD2low neuroblastoma cells by treatment with retinoic acid plus PF-9363 both in vitro and in vivo, leading to increased effectiveness of anti-GD2 CAR T cell therapy in neuroblastoma. These studies nominate KAT6A/B inhibition as a novel approach to enhance the effectiveness of differentiation and GD2-immunotherapy in neuroblastoma and may be relevant for other tumors of neuro-ectodermal origin. Nina Weichert-Leahey, Alla Berezovskaya, Mark W. Zimmerman, Francesca Alvarez-Calderon, Ulrike Gerdemann, Silvi Salhotra, Nathaniel Mabe, Adam D. Durbin, Kimberly Stegmaier, Brian J. Abraham, A. Thomas Look. KAT6A and B inhibition increases efficacy of differentiation and GD2 targeting immunotherapy in neuroblastoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 2601.

Despite multi-modal treatment intensification, nearly half of children with high-risk neuroblastoma (NBL) still succumb to their disease, underscoring the urgent need for novel therapeutic strategies with potent and durable anti-tumor activity. As in many pediatric cancers, gene expression signatures defining NBL cell state are under epigenetic control. Through epigenetic rewiring of the enhancer landscape, treatment with retinoic acid leads to suppression of oncogenic MYCN expression and subsequently to growth arrest and differentiation. However, the epigenetic and antiproliferative effects of retinoic acid are completely reversible upon drug withdrawal, leading to rapid tumor cell proliferation. Here, we sought to identify epigenetic modifiers that enhance the antiproliferative effects of retinoids in NBL. We conducted a screen with 452 compounds targeting epigenetic modifiers to identify synergistic growth inhibitors. We identified PF-9363, a first-in-class-inhibitor of the histone acetyltransferases KAT6A/B, as synergistically inhibiting NBL growth in combination with retinoic acid in NBL. Importantly, this growth suppression persisted beyond retinoid withdrawal, in stark contrast to retinoid monotherapy. Moreover, the KAT6A/B inhibitor PF-9363 synergizes with RA through sustained downregulation of MYCN and the transcription factors PHOX2B and GATA3, caused by gene expression silencing through PRC2-dependent deposition of H3K27me3 across their regulatory enhancers, extending the antiproliferative and differentiating effects of retinoic acid. In NSG xenograft models of NBL, PF-9363 alone demonstrated significant anti-tumor activity and tumor volumes were significantly reduced compared to vehicle treated mice. Moreover, combination treatment with retinoic acid and PF-9363 led to the most profound and sustained tumor growth suppression, lasting beyond retinoid withdrawal, an effect not observed with either agent alone, where tumors resumed rapid growth following drug removal. Reflecting the epigenetic reprogramming of these tumor cells, expression of GD2 was significantly upregulated in GD2low NBL cells by treatment with retinoic acid plus PF-9363 both in vitro and in vivo, thereby enhancing the efficacy of dinutuximab and anti-GD2 CAR T cell therapy. In conclusion, as KAT6A/B have recently emerged as druggable epigenetic targets in adult cancers—with early-phase clinical trials demonstrating favorable safety profiles and durable responses—our studies highlight KAT6A and KAT6B as promising novel therapeutic targets to enhance the efficacy of differentiation therapy and GD2 immunotherapy in NBL. Nina Weichert-Leahey, Alla Berezovskaya Berezovskaya, Mark W. Zimmerman, Francesca Alvarez-Calderon, Marlana Winschel, Silvi Salhotra Salhotra, Nathaniel Mabe Mabe, Ulrike Gerdemann, Kimberly Stegmaier, Adam D. Durbin, Derek A. Oldridge, Brian J. Abraham, A. Thomas Look. KAT6A and KAT6B are therapeutic targets to enhance the efficacy of differentiation therapy and GD2 immunotherapy in neuroblastoma [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Discovery and Innovation in Pediatric Cancer— From Biology to Breakthrough Therapies; 2025 Sep 25-28; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2025;85(18_Suppl_2):Abstract nr A020-PR004.

The quest for effective cancer therapeutics has traditionally centered on targeting mutated or overexpressed oncogenic proteins. However, challenges arise in cancers with low mutational burden or when the mutated oncogene is not conventionally targetable, which are common situations in childhood cancers. This obstacle has sparked large-scale unbiased screens to identify collateral genetic dependencies crucial for cancer cell growth. These screens have revealed promising targets for therapeutic intervention in the form of lineage-selective dependency genes, which may have an expanded therapeutic window compared to pan-lethal dependencies. Many lineage-selective dependencies regulate gene expression and are closely tied to the developmental origins of pediatric tumors. Placing lineage-selective dependencies in a transcriptional network model is helpful for understanding their roles in driving malignant cell behaviors. Here, we discuss the identification of lineage-selective dependencies and how two transcriptional models, core regulatory circuits and gene regulatory networks, can serve as frameworks for understanding their individual and collective actions, particularly in cancers affecting children and young adults.

Rhabdomyosarcoma is a soft-tissue sarcoma that occurs most frequently in pediatric patients and has poor survival rates in patients with recurrent or metastatic disease. There are two major sub-types of RMS: fusion-positive (FP-RMS) and fusion-negative (FN-RMS); with FP-RMS typically containing chromosomal translocations between the PAX3/7-FOXO1 loci. Regardless of subtype, RMS resembles embryonic skeletal muscle as it expresses the myogenic regulatory factors (MRFs), MYOD1 and MYOG. During normal myogenesis, these developmental transcription factors (TFs) orchestrate the formation of terminally differentiated, striated, and multinucleated skeletal muscle. However, in RMS these TFs become dysregulated such that they enable the sustained properties of malignancy. In FP-RMS, the PAX3/7-FOXO1 chromosomal translocation results in restructured chromatin, altering the binding of many MRFs and driving an oncogenic state. In FN-RMS, re-expression of MRFs, as well as other myogenic TFs, blocks terminal differentiation and holds cells in a proliferative, stem-cell-like state. In this review, we delve into the myogenic transcriptional networks that are dysregulated in and contribute to RMS progression. Advances in understanding the mechanisms through which myogenesis becomes stalled in RMS will lead to new tumor-specific therapies that target these aberrantly expressed developmental transcriptional pathways.

Rhabdomyosarcoma (RMS) is the most frequent soft tissue sarcoma of childhood. The fusion positive (FP) subtype expressing the chimeric transcriptional factor PAX3-FOXO1, the driver of malignancy, has limited treatment options due to resistance to treatments. As a pediatric cancer, RMS has a low mutational burden with epigenetic transcriptional deregulations playing key roles. Thus, targeting the basic transcriptional machinery including catalytic effectors of transcription is an emerging anti-cancer approach. The transcriptional cyclin-dependent kinase CDK12 regulates the transcription elongation phase, particularly of the DNA damage response (DDR) genes, by phosphorylating Ser2 (pSer) on RNA PolII. In Ewing sarcoma, CDK12 inhibition leads to a DNA repair deficient phenotype and high sensitivity to DNA damaging agents. PAX3-FOXO1 RMS cells display elevated levels of replication stress highly relying to DDR for survival. Therefore, in this study we evaluate the impact of CDK12 inhibition on PAX3-FOXO1 RMS tumorigenic features. We found that CDK12 mRNA levels were highly up-regulated in RMS primary samples compared to healthy control muscles, with one of the highest expression among the adult and pediatric tumors analyzed. Then, we treated PAX3-FOXO1 cell lines with THZ531, an inhibitor of CDK12 and its paralog CDK13. Six hours of THZ531 treatment at nanomolar concentrations decreased pSer2 of RNA PolII and reduced PAX3-FOXO1 transcripts and protein levels. The drug also reduced cell survival in 2D and 3D culture conditions. THZ531-treated cells accumulated in G2/M phase and underwent PARP- and Casp 3/7-dependent apoptosis, increased markers of DNA damage associated to DNA double-strand breaks (neutral comet assay), and reduced the levels of the DDR players ATM, ATR, RAD51, BRCA1 and BRCA2. Similar results were obtained after treatment with another CDK12/13i with Cyclin K degrader ability, SR-4835, and with a CDK12-specific degrader, BSJ-4-116, suggesting on-target effects. The three drugs showed similar low nanomolar concentrations effective on PAX3-FOXO1 RMS organoids. Preliminary RNAseq data in two PAX3-FOXO1 RMS cell lines showed that THZ531 downregulated only a low number of genes, common to the two cell lines. The most downregulated gene families were those of DDR pathways. Moreover, a negative correlation between gene length and gene expression levels was found. CDK12 KO in two RMS cell lines affected cell survival in 2D and 3D conditions compared to control cells suggesting on-target activity of the drugs. Phosphoproteomics analysis after THZ531 treatment showed a significant decrease of a number of serines and threonine phosphopeptides also involving a subset on DDR proteins. THZ531 treatment resulted synergic in vitro with Olaparib (PARPi) and with JQ1 (BRD4i). Interestingly, BRD4 is needed for PAX3-FOXO1 activity on oncogenic super-enhancers. Analysis of nascent RNAs and mapping of RNA PolII to identify direct targets of CDK12 is ongoing. Supported by Associazione Italiana per la Ricerca sul Cancro (AIRC) IG #27794 grant to RR Marika Attili, Lucrezia D'Archivio, Riccardo Mazzocchi, Francesca Antonella Aiello, Matteo Cassandri, Silvia Pomella, Adam D. Durbin, Francesco Marampon, Kim PJ Schellekens, Michael T Meister, Jarno Drost, Concetta Quintarelli, Simone Sidoli, Michele Ceribelli, Craig J Thomas, Franco Locatelli, Biagio De Angelis, Kristy R Stengel, Rossella Rota. CDK12 pharmacological targeting promotes cell death by abrogating DNA damage response, impairs PAX3-FOXO1 expression and synergizes with BRD4 and PARP inhibitors in fusion positive rhabdomyosarcoma [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Discovery and Innovation in Pediatric Cancer— From Biology to Breakthrough Therapies; 2025 Sep 25-28; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2025;85(18_Suppl_2):Abstract nr B015.