Proteolysis TArgeting Chimeras (PROTACs) can be used to target both the catalytic and noncatalytic functions of a protein, which can be particularly beneficial for proteins with important scaffolding functions like Aurora A. However, instability, poor selectivity profiles, and the hook effect often limit the applicability of PROTACs as chemical probes. In this study, we report the development of CCT400028, a second-generation alisertib-derived Aurora A PROTAC. The hook effect was removed through rational optimization of the CRBN-targeting warhead to decrease affinity for cereblon, which, combined with improved stability to hydrolysis, expands the range of concentrations and duration at which maximal degradation can be achieved. Potent Aurora A degradation was shown in three pediatric tumor cell lines, as well as excellent selectivity and on-target mechanism of action. CCT400028 and a matched inactive control analogue fulfill the criteria for a degrader chemical probe for studying Aurora A degradation in vitro.

Aurora A kinase (AURKA) is an oncogene frequently overexpressed in adult solid tumors, hematologic malignancies, and pediatric cancers. AURKA plays an important role in mitosis, and cancer cells sensitive to loss of AURKA include lines derived from MYCN amplified tumors such as neuroblastoma, as well as those with RB1 loss such as neuroendocrine small cell cancers and CDK4/6-resistant breast cancer (Mou, et al., 2021). Several AURKA inhibitors are effective in preclinical tumor models, but this activity has failed to translate into clinical efficacy. Recent studies have found that AURKA has kinase-independent scaffolding functions that are not effectively blocked through enzymatic inhibition. For example, AURKA limits the proteasomal degradation of MYCN through direct protein-protein interactions and prevents transcription/replication conflicts during S phase (Otto, et al, Cancer Cell, 2009; Buchel, et al., Cell Reports, 2017). To address the limitations of inhibitors, we have designed bifunctional targeted protein degraders of AURKA that enable removal of both enzymatic and scaffolding functions. Combinatorial chemical libraries of AURKA binders, CRBN binders, and linkers were generated by high-throughput parallel synthesis and screened for optimal properties in vitro and in vivo. These degraders form stable ternary complexes between AURKA and the E3 ligase CRBN, leading to efficient ubiquitylation and proteosome-dependent elimination of AURKA. Optimized AURKA degraders potently and selectively degraded AURKA in the MYCN amplified neuroblastoma cell line IMR32. Compounds with oral bioavailability and CNS exposure in mice were prioritized for further development. Our lead AURKA degrader, NRX-4972, has 58% oral bioavailability in C57BL/6 mice and moderate clearance (17.5 mL/min/kg). Three days of daily oral administration at 30 mg/kg to athymic nude mice bearing subcutaneous IMR32 tumors resulted in strong AURKA degradation (84% degraded relative to vehicle controls 6 hours after the final administration), high exposure in plasma (6.94 µM) and tumors (6,939 ng/g) and brain exposure of 178 ng/g. The closely related kinase AURKB was not degraded. AURKA degradation also promoted rapid regression of IMR32 neuroblastoma tumors, while inhibitors only achieved stasis. Analysis of tumors isolated from treated mice revealed that AURKA degraders, but not inhibitors, rapidly induced DNA damage and apoptosis. The differentiated activity of AURKA degraders compared to enzymatic inhibitors in these preclinical studies suggests the potential for improved safety and efficacy in cancer patients. The ability of AURKA degraders to penetrate the CNS not only broadens their therapeutic potential to include pediatric brain cancers such as medulloblastoma but also offers a promising avenue for treating adult cancers with brain metastases, addressing an urgent unmet medical need. Hua Tian, Eric Wegrzyniak, Ya-Wen Lu, Jeffrey T. Mihalic, Ryan Rountree, Tina Acholla, Bikash Adhikari, Karthik Arumugam, Paul Auger, Graham Carlson, Robert Cass, Coleen E. Casey, Abhinav Chaterjee, Lorenz Eing, James Iuliano, Adrienne Le, Yifan Li, Victoria Louie, Filippo Marchioni, Daniel Medina-Cleghorn, Isabel Morgado, Michael Mormino, Madeleine Nemchek, Erick Palomares, Rusha Sardhara, Julie Sheung, Sangita Sridharan, Gintvile Valinciute, Dipna Venkatachalam, Simon Vezina-Dawod, Derek Wodka, Stephanie Yung, Martin Eilers, John M. Maris, Yael P. Mossé, Martine F. Roussel, Elmar Wolf, Gwenn M. Hansen. Identification of selective, orally bioavailable aurora A degraders for treatment of pediatric and adult cancers [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 6379.

Deregulated MYCN is a driver of aggressive pediatric and adult neuroendocrine tumors, but critical oncogenic processes downstream of MYCN remain poorly defined. In neuroblastoma, MYCN interacts with and activates the Aurora-A kinase. Here we show that Aurora-A is a CDK-activating kinase for CDK12 by phosphorylating T893 in the T-loop, thereby enhancing its kinase activity. Aurora-A-dependent activation of CDK12 controls phosphorylation of T4 of RNA polymerase and recruits transcription termination complexes, thereby preventing transcription-replication conflicts. Enhanced crosslinking and immunoprecipitation sequencing reveals that Aurora-A associates with splice sites on nascent RNA. RNA-bound Aurora-A is catalytically inactive. MYCN competes with RNA for binding to Aurora-A and displaces Aurora-A from RNA in cells, promoting its CDK12 kinase activity. Combining Aurora-A and CDK12 inhibition potently suppresses the growth of MYCN-amplified neuroblastoma cells and patient-derived xenografts. Our data demonstrate that an Aurora-A/CDK12-dependent transcription termination pathway is a critical and targetable dependency of MYCN-driven tumors. ### Competing Interest Statement M.B. is a shareholder and employee of Proxygen. V.N.-P. is an employee of Proxygen.

Abstract MYC family proteins are primary drivers of oncogenic processes in a variety of cancer histologies. N-myc overexpression and amplification induce aggressive pediatric cancers, the most common of which are solid extracranial tumors in children (neuroblastoma, NB) and malignant pediatric CNS tumors (medulloblastoma, MB). Due to the oncogenic addiction observed in these tumor types, N-myc is considered an attractive therapeutic target. However, direct small molecule targeting of N-myc remains technically challenging. Alternative approaches to target this pathway include inhibiting proteins thought to stabilize N-myc, such as Aurora A kinase (AURKA). While inhibition of AURKA was found to be effective in preclinical tumor studies, no AURKA inhibitors have been approved for clinical use due to a lack of efficacy. Here, we aim to develop chimeric degrader molecules suitable for preclinical and clinical use that degrade AURKA and concomitantly reduce N-myc levels. We used automated solid-phase synthesis to generate a library of >1000 chimeric degraders using derivatives of six known AURKA ligands (Series 1-6). Due to the limitations of using N-myc-driven MB cells in vitro, we employed NB cells as the proof-of-concept model. While the degraders based on five of the ligands showed efficient AURKA degradation, only degraders based on series 3 and 6 could also diminish N-myc. Selected series were further characterized in pharmacokinetic and pharmacodynamic experiments using NB tumor xenografts implanted in the flank in CD1-nude mice. In summary, chimeric degraders based on two individual AURKA ligands efficiently degraded AURKA and N-myc in vitro and in vivo, providing the foundation for further development of novel therapies for patients with N-myc-driven cancers. Further studies will focus on assessing antitumor properties of lead candidates in NB and patient-derived orthotopic MB xenograft models.

Cellular growth is a fundamental process of life and must be precisely controlled in multicellular organisms. Growth is crucially controlled by the number of functional ribosomes available in cells. The production of new ribosomes depends critically on the activity of RNA polymerase (RNAP) II in addition to the activity of RNAP I and III, which produce ribosomal RNAs. Indeed, the expression of both, ribosomal proteins and proteins required for ribosome assembly (ribosomal biogenesis factors), is considered rate-limiting for ribosome synthesis. Here, we used genetic screening to identify novel transcriptional regulators of cell growth genes by fusing promoters from a ribosomal protein gene (Rpl18) and from a ribosomal biogenesis factor (Fbl) with fluorescent protein genes (RFP, GFP) as reporters. Subsequently, both reporters were stably integrated into immortalized mouse fibroblasts, which were then transduced with a genome-wide sgRNA-CRISPR knockout library. Subsequently, cells with altered reporter activity were isolated by FACS and the causative sgRNAs were identified. Interestingly, we identified two novel regulators of growth genes. Firstly, the exon junction complex protein RBM8A controls transcript levels of the intronless reporters used here. By acute depletion of RBM8A protein using the auxin degron system combined with the genome-wide analysis of nascent transcription, we showed that RBM8A is an important global regulator of ribosomal protein transcripts. Secondly, we unexpectedly observed that the glycolytic enzyme aldolase A (ALDOA) regulates the expression of ribosomal biogenesis factors. Consistent with published observations that a fraction of this protein is located in the nucleus, this may be a mechanism linking transcription of growth genes to metabolic processes and possibly to metabolite availability.

The family of AURORA kinases is essential for cell cycle progression and dysregulation of AURORA-A in cancer led to a large number of clinical and pre-clinical inhibitors. However, ATP competitive AURORA-A inhibitors usually do not target non-catalytic functions that have also been identified as mechanisms promoting tumorigenesis. To target non-catalytic as well as catalytic functions, we developed a series of PROTACs (PROteolysis TArgeting Chimeras) based on the selective AURORA-A kinase inhibitor MK-5108 (VX-689) and the CEREBLON E3-ligase ligand thalidomide. The most potent PROTAC, JB301, had good physicochemical properties and cell penetration resulting in degradation of AURORA-A in leukemic cells at single digit nM concentration.

The mitotic kinase AURORA-A is essential for cell cycle progression and is considered a priority cancer target. Although the catalytic activity of AURORA-A is essential for its mitotic function, recent reports indicate an additional non-catalytic function, which is difficult to target by conventional small molecules. We therefore developed a series of chemical degraders (PROTACs) by connecting a clinical kinase inhibitor of AURORA-A to E3 ligase-binding molecules (for example, thalidomide). One degrader induced rapid, durable and highly specific degradation of AURORA-A. In addition, we found that the degrader complex was stabilized by cooperative binding between AURORA-A and CEREBLON. Degrader-mediated AURORA-A depletion caused an S-phase defect, which is not the cell cycle effect observed upon kinase inhibition, supporting an important non-catalytic function of AURORA-A during DNA replication. AURORA-A degradation induced rampant apoptosis in cancer cell lines and thus represents a versatile starting point for developing new therapeutics to counter AURORA-A function in cancer.

High invasiveness and resistance to chemo- and radiotherapy of glioblastoma multiforme (GBM) make it the most lethal brain tumor. Therefore, new treatment strategies for preventing migration and invasion of GBM cells are needed. Using two different migration assays, Western blotting, conventional and super-resolution (dSTORM) fluorescence microscopy we examine the effects of the dual PI3K/mTOR-inhibitor PI-103 alone and in combination with the Hsp90 inhibitor NVP-AUY922 and/or irradiation on the migration, expression of marker proteins, focal adhesions and F-actin cytoskeleton in two GBM cell lines (DK-MG and SNB19) markedly differing in their invasive capacity. Both lines were found to be strikingly different in morphology and migration behavior. The less invasive DK-MG cells maintained a polarized morphology and migrated in a directionally persistent manner, whereas the highly invasive SNB19 cells showed a multipolar morphology and migrated randomly. Interestingly, a single dose of 2 Gy accelerated wound closure in both cell lines without affecting their migration measured by single-cell tracking. PI-103 inhibited migration of DK-MG (p53 wt, PTEN wt) but not of SNB19 (p53 mut, PTEN mut) cells probably due to aberrant reactivation of the PI3K pathway in SNB19 cells treated with PI-103. In contrast, NVP-AUY922 exerted strong anti-migratory effects in both cell lines. Inhibition of cell migration was associated with massive morphological changes and reorganization of the actin cytoskeleton. Our results showed a cell line-specific response to PI3K/mTOR inhibition in terms of GBM cell motility. We conclude that anti-migratory agents warrant further preclinical investigation as potential therapeutics for treatment of GBM.