Organoid-based Discovery of Oncogenic Drivers and Treatment Resistance Mechanisms

项目摘要/摘要 癌症基因组图谱计划(TCGA)多尺度“组学”数据的泛滥 其他癌症分析项目已经揭示了显著的遗传和表观遗传学 复杂性和巨大的患者内部差异。因此，一个特别迫切的需求 存在精确的、可扩展的人类癌症模型，可以在功能上询问 这些广泛的数据集，识别良性乘客的司机致癌事件和 描述它们与治疗反应的相关性。 近四年来斯坦福大学癌症靶点的发现与发展 (CTD2)中心一直在探索用于癌症建模和研究的人类“有机类”培养方法 司机癌基因的发现。原生3D有机物培养提供了不同寻常的机会 在培养的原代人的表观/遗传“舌状突起”内启动肿瘤新生 野生型组织，与相应的、通常定义不清的长- 传代培养的2D癌细胞株。这为癌症创造了一个高度明确的基线 全程利用的建模和功能驱动癌基因验证。 我们的总体方法应用了最先进的系统生物学和健壮 大规模癌症概况数据集的计算资源，从而提名 在人类器官培养中接受直接功能评估的候选驱动程序。 该实验范围利用了Calvin Kuo(报告PI， 有机化合物)，季汉雷(多PI，癌症ITH，基因组学)，Christina Curtis(多PI，肿瘤 进化论，癌症系统生物学)，奥利维尔·格瓦尔特(癌症系统生物学， 表观遗传学)和迈克尔·巴西克(高通量功能基因组学)。因此， 目标1和2对TCGA拷贝数改变(CNA)的生物信息学优先顺序 以及用于通过基于有机物的条形码慢病毒筛选发现驱动器的甲基化数据 并采用了正交化的cDNA、shRNA和CRISPR方法。目标3利用了 纵向观察癌基因的新基因组和表观基因组进化 工程野生型有机化合物提名合作癌基因网络， 进行迭代的有机化合物功能验证。最后，目标4探讨了 有机类化合物模拟从头治疗耐药性，使用原型靶向和 化疗微扰剂作为原理证明，使用单细胞RNA- SEQ/瘤内异质性和外显子组测序终点。

PROJECT SUMMARY/ABSTRACT The deluge of multi-scale “omics” data from The Cancer Genome Atlas Project (TCGA) and other cancer profiling projects has revealed remarkable genetic and epigenetic complexity and tremendous intrapatient variation. Accordingly, a particularly acute need exists for accurate, scalable human cancer models that can functionally interrogate these extensive datasets, identify driver oncogenic events from benign passengers and characterize their relevance to treatment response. For the last four years the Stanford Cancer Target Discovery and Development (CTD2) Center has pursued human “organoid” culture methods for cancer modeling and driver oncogene discovery. Primary 3D organoid cultures afford the unusual opportunity to initiate cancer de novo within the epi/genetic “tabula rasa” of cultured primary human wild-type tissue, versus the corresponding and often poorly-defined complexity of long- passaged 2D cancer cell lines. This creates a highly defined baseline for cancer modeling and functional driver oncogene validation that is leveraged throughout. Our overall approach applies state-of-the-art systems biology and robust computational resources to large-scale cancer profiling datasets, thus nominating candidate drivers that undergo direct functional evaluation in human organoid culture. This experimental scope leverages a highly synergistic team of Calvin Kuo (reporting PI, organoids), Hanlee Ji (multi-PI, cancer ITH, genomics), Christina Curtis (multi-PI, tumor evolution, cancer systems biology), Olivier Gevaert (cancer systems biology, epigenetics) and Michael Bassik (high-throughput functional genomics). Accordingly, Aims 1 and 2 couple bioinformatic prioritization of TCGA copy number alteration (CNA) and methylation data for driver discovery via organoid-based barcoded lentiviral screens and orthogonal cDNA, shRNA and CRISPR approaches. Aim 3 exploits the ability to longitudinally observe de novo genomic and epigenomic evolution in oncogene- engineered wild-type organoids to nominate networks of cooperating oncogenes that undergo iterative organoid functional validation. Lastly, Aim 4 explores the utility of organoids to model de novo treatment resistance, using archetypal targeted and chemotherapy perturbagens as proof-of-principle and employing single cell RNA- seq/intratumoral heterogeneity and exome sequencing endpoints.