Proteasomal recruiters of PAX3-FOXO1 Designed via Sequence-Based Generative Models

摘要 融合阳性的肺泡型横纹肌肉瘤(FP-ARMS)是儿童最致命的癌症之一，主要是 依赖于PAX3-FOXO1融合癌蛋白，一种劫持正常基因的嵌合转录因子 表达和染色质状态。患有PAX3-FOXO1阳性手臂的儿童五年存活率约为30% 转移时为10%。Pax3-FOXO1在很大程度上被认为是一种不可药物的蛋白质，没有小分子 被开发来结合和抑制其活性。最近，我们开发了新的算法来设计特定的 选择性结合和降解致病蛋白的多肽，包括经典的“无法下药”的转录 因子和融合癌蛋白。在这项提案中，我们将整合我们最先进的发电模式，以 Novo设计对PAX3-FOXO1(而不是PAX3或FOXO1)具有选择性的高亲和力多肽引导降解物 展示FP-ARM在体外模型中的降解。这项工作的成果将激励 下游体内研究和加速FP-ARM的蛋白质靶向药物。

Abstract Fusion-positive alveolar rhabdomyosarcoma (FP-ARMS), one of the most fatal childhood cancers, is primarily dependent on the PAX3-FOXO1 fusion oncoprotein, a chimeric transcription factor that hijacks normal gene expression and chromatin state. The five-year survival for children with PAX3-FOXO1-positive ARMS is ~30% and <10% when metastatic. PAX3-FOXO1 is largely considered an undruggable protein, with no small molecule developed to bind and inhibit its activity. Recently, we have developed novel algorithms to design specific peptides that selectively bind and degrade pathogenic proteins, including classically “undruggable” transcription factors and fusion oncoproteins. In this proposal, we will ensemble our state-of-the-art generative models to de novo design high-affinity peptide-guided degraders selective to PAX3-FOXO1 (and not PAX3 or FOXO1) and demonstrate degradation within in vitro models of FP-ARMS. The outcomes of this work will motivate downstream in vivo studies and accelerate protein-targeting medicines for FP-ARMS.