The gene fusion KIAA1549::BRAF is a common oncogenic driver in pediatric pilocytic astrocytoma. The chimeric fusion protein that comprises most of the type 1 transmembrane KIAA1549 protein fused in its C-terminal cytosolic domain to the truncated BRAF oncogene without its N-terminal autoinhibitory domains. KIAA1549 has a large N-terminal ectodomain that undergoes extensive O-mannosylation directed by the POMT1/2 protein O-mannosyltransferase complex localized in the ER. Here, we investigated how posttranslational modifications contribute to KIAA1549 protein maturation in the secretory pathway. Using a panel of KIAA1549 expression constructs and glycoengineered HEK293 cells, we provide evidence that O-mannosylation of the KIAA1549 ectodomain is required for ER exit, acquisition of complex N-glycans, and subsequent furin-like cleavage in the Golgi. In the absence of POMT1/2 O-mannosylation KIAA1549 is retained in ER and fails to undergo proprotein processing. Our findings confirm and extend our previous studies with the KIAA1549::BRAF fusion protein, where oncogenic transformation was dependent on POMT1/2 O-mannosylation and affected on furin processing. Thus, these intrinsic posttranslational modifications of the KIAA1549 ectodomain are critical determinants of the cellular trafficking that lead to activation of the BRAF oncogene, potentially involving a cascade of cleavage events in the Golgi that ultimately release a truncated but fully active BRAF kinase domain into the cytosol.

The most common fusion oncoprotein in the most common pediatric solid tumor, pediatric low-grade glioma (pLGG), is KIAA1549::BRAF. Although MAPK inhibitors show early promise, they often fail to achieve durable cures, highlighting the need to develop new therapeutics for KIAA1549::BRAF-driven tumors. In this study we leverage genome-scale anti-transformation CRISPR screens to uncover KIAA1549::BRAF-specific genetic vulnerabilities. Across all genes, POMT1 and POMT2, the two members of the heterodimeric Protein O-mannosyltransferase (POMT) complex emerge as the strongest KIAA1549::BRAF-specific dependencies. We demonstrate that KIAA1549 is a direct substrate of this complex, and that glycosylation of KIAA1549::BRAF by POMT is necessary for its maturation through the secretory pathway and subsequent oncogenic signaling. These data highlight for the first time a new route to directly target KIAA1549::BRAF independent of the MAPK pathway. This work also uncovered an unexpected and previously unrecognized role for KIAA1549 in driving KIAA1549::BRAF oncogenic signaling. The prevailing dogma has been that BRAF fusions exert their transforming activity through truncation and removal of N-terminal regulatory domains, thereby rendering the BRAF kinase domain constitutively active. We demonstrate that BRAF fusion partners, including KIAA1549, contain domains that facilitate membrane localization, thereby enhancing BRAF signaling. These findings challenge the traditional view of BRAF fusion biology and establish a role for fusion partners in shaping oncogenic potential. Sean Misek, Anna Borgenvik, Daniel Christen, Gloria Kyrila, Alexander Zhang, Sarah Reel, Michelle Boisvert, Kelly Cai, Kevin Zhou, Elizabeth Gonzales, Lorena Lazo de la Vega, Timothy Ragnoni, Nicole Persky, Tanaz Abid, Esteban Miglietta, Sergey Vakhrushev, Michael Stumpe, Jacquelyn Jones, Seth Malinowski, Lobna Elsadek, Aaron Fultineer, Kira Tang, Antonio Maldera, Hyesung Jeon, Sangita Pal, Todd Golub, William Hahn, Eric Fischer, Jesse Boehm, Jörn Dengjel, Henrik Clausen, Nada Jabado, Till Milde, Anne Carpenter, Beth Cimini, Keith Ligon, Katherine Janeway, Michael Eck, David Root, David Jones, Timothy Phoenix, Rameen Beroukhim, Hiren Jitendra Joshi, Tilman Tilman, Adnan Halim, Pratiti Bandopadhayay. O-mannosylation and protein maturation checkpoints represent therapeutic opportunities in BRAF fusion protein oncogenesis [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Fusion-Positive Cancer: From Discovery to Therapy; 2026 Jan 13-15; Philadelphia PA. Philadelphia (PA): AACR; Cancer Res 2026;86(1_Suppl):Abstract nr B026.

Results of differential gene expression analysis between GTML tumours and GTS tumours.

Results of differential gene expression analysis between GTML3 tumor cells treated for 48 hours with DMSO or Dox.

Pilocytic astrocytomas (PAs) are low-grade gliomas (LGGs) and predominantly affect children under 14 years of age. PAs are driven by aberrant activation of the RAS/RAF/MEK/ERK axis, primarily by BRAF alterations. As a key activator of this pathway, the Ser/Thr-kinase BRAF plays a crucial role in promoting cellular survival and proliferation and is one of the most frequently mutated kinases in cancer. While the most prevalent mutation, BRAFV600E, is commonly associated with hairy cell leukemia, melanoma, and thyroid carcinomas, the KIAA1549::BRAF fusion is the hallmark oncogene of LGGs, particularly PAs. Further, due to advancements in precision oncology diagnostics, this fusion is also increasingly identified in other tumor types. Various breakpoint variants of the KIAA1549::BRAF fusion gene were reported, with the most common involving KIAA1549 exons 1-16 and BRAF exons 9-18. The oncogenic potential of KIAA1549::BRAF is attributed to the loss of exons encoding the autoinhibitory domains of BRAF, although there is increasing evidence that the fusion partners of RAF oncoproteins can influence their functionality and signalling. Despite the high frequency of KIAA1549::BRAF fusions, the function, topology, and localization of KIAA1549, an enigmatic putative transmembrane protein, remains unclear. In contrast to the cytoplasmic localisation of wild-type BRAF, point mutants and insertion or deletion mutants, KIAA1549::BRAF fusions retain the single transmembrane domain of the KIAA1549 portion, localizing it to the plasma membrane. This localization requires trafficking through the ER/Golgi system during maturation. Additionally, using TAILS mass spectrometry and genomic approaches, we demonstrate that KIAA1549 and KIAA1549::BRAF fusions are cleaved at a specific cleavage motif in the extracellular portion of the protein. Mutation of the cleavage motif slows down cell growth compared to non-cleavage impaired fusions. Known protease inhibitors can target this cleavage event, offering a promising strategy for drug development. Apart from a very recently approved RAF inhibitor (RAFi), there are very little options available for directly targeting KIAA1549::BRAF fusion proteins. This is due to the fact that the RAFi developed for BRAFV600E are ineffective against BRAF fusions and probably also due to our limited understanding of the influence of KIAA1549. Consequently, our data could support the development of urgently needed targeted therapies for this BRAF fusion protein. Daniel Christen, Sean Misek, Anna Borgenvik, Gloria Kyrila, Alexander Zhang, Sarah Reel, Michelle Boisvert, Kelly Cai, Kevin Zhou, Elizabeth M Gonzales, Amy Goodale, Esteban Miglietta, Jacquelyn Jones, Seth Malinowski, Lobna Elsadek, Merve Ozdemir, Zachary Eisenbies, Joohee Lee, April Apfelbaum, Jenna Robinson, Antonio Maldera, Daniel Bondeson, Jason Kwon, Mounica Vallurupalli, Sangita Pal, Todd Golub, William Hahn, Eric Fischer, Jesse Boehm, Jörn Dengjel, Henrik Clausen, Nada Jabado, Till Milde, Beth Cimini, Keith Ligon, Kathrine Janeway, Michael Eck, David Root, David Jones, Timothy N Phoenix, Rameen Beroukhim, Hiren Jitendra Joshi, Tilman Brummer, Adnan Halim, Pratiti Bandopadhayay. Molecular insights into KIAA1549::BRAF fusion proteins: Implications for targeted therapy in pediatric low-grade gliomas [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 A032.

Pediatric low-grade gliomas (pLGGs) are the most common type of brain tumor in children. We (and others) have found that the second most common alteration in pLGGs is in Fibroblast Growth Factor Receptor 1 (FGFR1). Current treatment options for patients with pLGGs are surgical resection, chemotherapy, and targeted inhibitors. Despite the prevalence of FGFR1 alterations, the only approved targeted therapies in pLGGs are MEK and BRAF inhibitors, not FGFR inhibitors. Therefore, identifying effective therapeutic strategies for FGFR1-altered gliomas remains a large gap in the field. To investigate this question, we generated novel human isogenic neural stem cell lines that are driven by FGFR1-alterations found in pLGGs. Using these models, we have found that single agent treatment with currently available FGFR-inhibitors is unlikely to result in curative responses, and that combination approaches are likely. The goal of this project is to identify combination approaches that induce cytotoxic responses rather than those that are primarily cytostatic. To do so, we will leverage CRISPR-cas screening approaches to identify genes, which when ablated in the presence of FGFR-inhibitors, induce apoptosis. We have designed CRISPR-cas12a based lentiviral libraries that specifically target genes that are clinically druggable with current agents or are predicted to be druggable. We will transduce our isogenic FGFR-driven cell lines with the “druggable genome library” and then will isolate apoptotic cells using Annexin V labeling. This will allow us to evaluate which genes were ablated in cells in which apoptosis was induced, thereby nominating potential genes to target in combination approaches. Overall, this project aims to identify effective potential combination therapies for FGFR1-altered pLGGs. Sarah W Lamson, April A Apfelbaum, Anna Borgenvik, Noelia Ares Bovida, Verónica Rendo, Allison Uebele, John Doench, Pratiti Bandopadhayay. Identifying effective targeted inhibitors as single-agents or in combination for FGFR1-altered pediatric low-grade gliomas [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 B023.

Fusions between protein-coding genes are common oncogenic drivers across cancers, typically pairing a proto-oncogene with partner that does not independently drive cancer. In all therapeutically actionable fusions, the proto-oncogene is the drug target, the contributions to oncogenicity of the fusion partner have largely been ignored. We studied the role of BRAF fusion partners and found that they are necessary for transformation. In the setting of KIAA1549::BRAF, the most common fusion protein across brain tumors, we found that KIAA1549 is necessary for the oncogenicity of KIAA1549::BRAF and engenders a striking and specific dependency on the protein O-mannosyltransferase complex (POMT1/2). Specifically, we show that genetic silencing or pharmacologic inhibition of the protein O-mannosyltransferase complex (POMT1/2) reverses fusion-induced transformation, thereby representing a novel and MAPK independent therapeutic target. Furthermore, POMT1/2 is required to glycosylate and enable maturation of the K::B fusion protein. These findings represent a proof-of-concept for targeting the partners in oncogenic fusions as a potential cancer therapeutic strategy. ### Competing Interest Statement PB has received grant funding from Novartis Institute of Bio-medical Research, and has served on paid advisory boards for QED Therapeutics, and Day One Biopharmaceuticals. RB consults for and owns equity in Scorpion Therapeutics, Kar-yoverse Therapeutics, and LOH Therapeutics. WCH is a consultant for Thermo Fischer, Solasta Ventures, KSQ Ther-apeutics, Frontier Medicines, Jubilant Therapeutics, RAPPTA Therapeutics, Serinus Biosciences, Kestral Therapeutics, Crane Biotherapeutics, Function Oncology, Recursion Pharma, and Weaver Biotherapetics and Perceptive. DER receives research funding from members of the Functional Genomics Consortium (Abbvie, BMS, Jannsen, Merck, Vir), and is a director of Addgene, Inc. TM is advisory board member for Ipsen Pharma GmbH. SAM is a consultant for Karyoverse Therapeutics and is a consultant for and owns equity in LOH Therapeutics. ESF is a founder, scientific advisory board (SAB) member, and equity holder of Civetta Therapeutics, Proximity Therapeutics, Neomorph, Inc. (also board of direc-tors), StelexisBiosciences, Inc., Anvia Therapeutics, Inc. (also board of directors) and CPD4, Inc. (also board of directors). He is an equity holder and SAB member for Avilar Therapeu-tics, PhotysTherapeutics, and Ajax Therapeutics and an equity holder in Lighthorse Therapeutics. E.S.F. is a consultant to Novartis, EcoR1 capital, Odyssey and Deerfield. The Fischer lab receives or has received research funding from Deerfield, Novartis, Ajax, Interline, Bayer and Astellas. TB received speaker honoraria from Pierre Fabre and the European Society for Medical Oncology (ESMO)

Results of differential gene expression analysis between pre-GTS1 tumor cells treated for 48 hours with DMSO or Dox.

List of genes that are part of a quiescent cell signature and display regions of differentially opened chromatin.

Fusions between protein-coding genes are common oncogenic drivers across cancers, typically pairing a proto-oncogene with a partner that does not independently drive cancer. In all therapeutically actionable fusions, the proto-oncogene is the drug target; the contributions to oncogenicity of the fusion partner have largely been ignored. KIAA1549::BRAF (K::B) is the most common fusion protein across brain tumors, and particularly in pediatric gliomas. We found that KIAA1549 is necessary for oncogenicity of K::B, in contrast to the notion that the BRAF kinase domain alone is enough for transformation. We therefore performed genome-wide CRISPR/Cas9 anti-transformation screens in isogenic neural stem cell models with K::B overexpression to identify genetic dependencies specific to this oncogenic driver. We found that the BRAF fusion partner KIAA1549 engenders a striking and specific dependency on the protein O-mannosyltransferase complex (POMT1/2). Specifically, we show that genetic silencing or pharmacologic inhibition of the POMT1/2 complex reverses K::B transformation, thereby representing a novel therapeutic target beyond the MAPK pathway. Furthermore, POMT1/2 is required to glycosylate and enable maturation of K::B, reflecting the native biology of KIAA1549 but representing a non-canonical process for BRAF. These findings represent a proof-of-concept for targeting the partners in oncogenic fusions as a potential cancer therapeutic strategy. Anna Borgenvik, Sean Misek, Daniel Christen, Gloria Kyrila, Alexander Zhang, Sarah Reel, Michelle Boisvert, Kelly Cai, Kevin Zhou, Elizabeth M Gonzales, Lorena Lazo de la Vega, Amy Goodale, Esteban Miglietta, Jacquelyn Jones, Seth Malinowski, Lobna Elsadek, Merve Ozdemir, Zachary Eisenbies, Joohee Lee, April Apfelbaum, Jenna Robinson, Antonio Maldera, Daniel Bondeson, Jason Kwon, Mounica Vallurupalli, Sangita Pal, Todd Golub, William Hahn, Eric Fischer, Jesse Boehm, Jörn Dengjel, Henrik Clausen, Nada Jabado, Till Milde, Beth A Cimini, Keith Ligon, Kathrine Janeway, Michael Eck, David Root, David Jones, Timothy N Phoenix, Rameen Beroukhim, Hiren Jitendra Joshi, Tilman Brummer, Adnan Halim, Pratiti Bandopadhayay. O-mannosylation represents a therapeutic opportunity in BRAF fusion protein oncogenesis [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 B019.

Medulloblastomas comprise a molecularly diverse set of malignant pediatric brain tumors in which patients are stratified according to different prognostic risk groups that span from very good to very poor. Metastasis at diagnosis is most often a marker of poor prognosis and the relapse incidence is higher in these children. Medulloblastoma relapse is almost always fatal and recurring cells have, apart from resistance to standard of care, acquired genetic and epigenetic changes that correlate with an increased dormancy state, cell state reprogramming and immune escape. Here, we review means to carefully study metastasis and relapse in preclinical models, in light of recently described molecular subgroups. We will exemplify how therapy resistance develops at the cellular level, in a specific niche or from therapy-induced secondary mutations. We further describe underlying molecular mechanisms on how tumors acquire the ability to promote leptomeningeal dissemination and discuss how they can establish therapy-resistant cell clones. Finally, we describe some of the ongoing clinical trials of high-risk medulloblastoma and suggest or discuss more individualized treatments that could be of benefit to specific subgroups.

Abstract Since the KIAA1549::BRAF fusion was discovered as the most common driver of pediatric low-grade glioma (pLGG), it has been hypothesized that the fusion induces oncogenicity through BRAF activation as a result of KIAA1549 replacing the BRAF negative regulatory N-terminus. This led to the rapid translation of MAPK pathway inhibitors into the clinical setting. Despite most tumors exhibiting promising initial responses, some tumors are not sensitive and about half of responsive tumors grow back after treatment cessation. Therefore, strategies that result in sustained tumor responses are desperately needed. We have recently performed genome-scale CRISPR/Cas9 screens across isogenic neural stem cell models transduced to express pLGG-associated oncogenes (KIAA1549::BRAF, BRAFV600E, and multiple FGFR1 and MYB family alterations) to generate a dependency map of genetic vulnerabilities associated with expression of these oncogenes. We also included normal neural stem cells to allow identification of genetic dependencies specifically induced by the expression of each oncogene. Surprisingly, through these efforts, we have discovered KIAA1549::BRAF expressing cells to harbor striking and specific dependency on multiple members of an enzymatic complex that exerts its activity outside of the MAPK signaling axis*. Interestingly, this enzymatic complex has been described to modify only a few substrates, including KIAA1549, suggesting specificity. We have now validated this dependency across other isogenic models of KIAA1549::BRAF-expressing cells. Finally, our KIAA1549::BRAF-expressing models exhibit preferential sensitivity to a tool compound that targets our novel enzyme, suggesting novel therapeutic potential for KIAA1549::BRAF. These findings highlight a MAPK pathway-independent avenue for therapeutically targeting the most frequent genomic alterations in pediatric brain tumors. Consequently, we also propose that the fusion partner KIAA1549 is instrumental for the aberrant BRAF signaling driving pLGGs. (* we are currently in the process of working with our IP offices to ensure that we can disclose the names of the genes and proteins at the ISPNO meeting)

Abstract t was originally hypothesized that loss of BRAF N-terminal negative regulatory domains, which are deleted in the KIAA1549:BRAF rearrangement, results in constitutive activation of the BRAF kinase. This model suggests that BRAF fusion partners are dispensable for BRAF oncogenic signaling. Paradoxically, our data suggest expression of truncated BRAF is insufficient for transformation and identify specific domains in KIAA1549 that are necessary for transformation. These domains are critical for regulating subcellular localization of the fusion protein and highlight an aberrant pattern of cellular localization that is not observed with wildtype BRAF. Altered subcellular localization results in proteolytic cleavage of KIAA1549:BRAF, and we identify the protease responsible for this cleavage event. This protease is therapeutically tractable and clinically relevant inhibitors have been developed, presenting a MAPK-independent mechanism to target the fusion. We also highlight an unexpected role for rare fusion partners, including FAM131B, in BRAF activation. In total, these data suggest that BRAF fusion partners are not indispensable for transformation as was previously thought, presenting unexplored opportunities to therapeutically target pLGG tumors with BRAF rearrangements. We are currently working on the IP measures required to disclose the specific enzyme(s) at ISPNO.

Medulloblastoma, the most common malignant pediatric brain tumor, often harbors MYC amplifications. Compared to high-grade gliomas, MYC -amplified medulloblastomas often show increased photoreceptor activity and arise in the presence of a functional ARF/p53 suppressor pathway. Here, we generate an immunocompetent transgenic mouse model with regulatable MYC that develop clonal tumors that molecularly resemble photoreceptor-positive Group 3 medulloblastoma. Compared to MYCN -expressing brain tumors driven from the same promoter, pronounced ARF silencing is present in our MYC -expressing model and in human medulloblastoma. While partial Arf suppression causes increased malignancy in MYCN -expressing tumors, complete Arf depletion promotes photoreceptor-negative high-grade glioma formation. Computational models and clinical data further identify drugs targeting MYC-driven tumors with a suppressed but functional ARF pathway. We show that the HSP90 inhibitor, Onalespib, significantly targets MYC-driven but not MYCN-driven tumors in an ARF-dependent manner. The treatment increases cell death in synergy with cisplatin and demonstrates potential for targeting MYC-driven medulloblastoma.