Discovering Network-Based Drivers of Single-Cell Transcriptional State in Tumor Immune Microenvironment to Reveal Immuno-Therapeutic Targets and Treatment Synergies

项目摘要/摘要： 实体瘤不仅包括肿瘤细胞，还包括多种免疫细胞类型，它们渗入到肿瘤的微血管中。 环境。传统的癌症治疗方法侧重于直接杀死肿瘤细胞，但最近 免疫检查点抑制疗法的目标是激活组织中的抗肿瘤免疫细胞。 在过去的几年里，免疫疗法在临床肿瘤学中起着革命性的作用，但生物标记物 治疗反应有限，治疗对肿瘤微环境的影响尚不完全清楚。这有 德雷克实验室和其他人积极努力，在不同治疗下更好地描述肿瘤中的免疫细胞类型 旨在揭示新的治疗目标并确定改进的治疗反应预测因素。我们的 该集团在应用高通量单细胞RNA测序(scRNA-Seq)方面拥有丰富的经验 在单个细胞水平上具有完全转录分辨率的肿瘤微环境。我们假设 在单细胞水平描绘肿瘤微环境并应用先进的基于网络的分析 治疗流水线-幼稚和免疫疗法治疗的肿瘤将改善 肿瘤浸润性免疫细胞类型中的转录程序、它们与预后的关系及其临床意义 与肿瘤细胞的相关相互作用。目标1)尽管分辨率很高，但scRNA-Seq数据通常是稀疏的， 在任何给定细胞中检测到的少数基因。我们的目标是开发一套源自卡里法诺的强大工具 基于网络的单细胞数据调节蛋白活性推断实验室，缓解基因表达 并提供了一个可扩展的管道，用于推断细胞群、肿瘤-免疫相互作用和 调节蛋白在不同的细胞状态下被不同的激活。我们通过比较以下内容来验证此管道 在一组肾透明细胞癌(CcRCC)患者的数据集中，同时用流式细胞术分析标记物。目标 2)我们将专门利用我们的新分析管道来询问肿瘤浸润性监管的驱动因素 T细胞，一种由多种常规治疗方式诱导的免疫抑制群体，包括 前列腺癌的雄激素剥夺疗法。我们将通过CRISPR验证预测的肿瘤渗透驱动因素 基因敲除筛查和应用先进的转录干扰筛查来识别哪些药物逆转了 肿瘤特异性Treg签名。这些都有望成为未来合并的主要候选者。 免疫疗法研究。目的3)确定免疫治疗引起的微环境改变。 前列腺癌免疫治疗加雄激素剥夺的两个临床试验中的应答者和无应答者 癌症和一项抗PD1加抗IL1b治疗ccRCC的试验，确定了潜在的新治疗靶点。在……里面 此外，我们将把我们新开发的分析管道应用于已发表的scRNA-Seq数据集，以识别 黑色素瘤治疗反应的预测因子。在有经验的免疫治疗导师的共同指导下 和计算系统生物学在哥伦比亚大学医学中心的背景下，这个项目将准备 作为一名具有翻译生物信息学研究独特背景的内科科学家职业生涯的实习生。

Project Summary/Abstract: Solid tumors consist not only of tumor cells, but also of immune cell types infiltrating the tumor micro- environment. Traditional approaches to cancer therapy have focused on killing tumor cells directly, but recent immune checkpoint inhibitor therapies have instead aimed to activate anti-tumor immune cells in the tissue. Immunotherapy has been transformative in clinical oncology over the past several years, but biomarkers of response are limited and effect of treatment on tumor micro-environment is incompletely understood. This has motivated efforts by Drake lab and others to better profile immune cell types in tumors under various treatment conditions, aiming to reveal novel therapy targets and identify improved predictors of treatment response. Our group has considerable experience applying high-throughput single-cell RNA sequencing (scRNA-Seq) to profile tumor micro-environment with full transcriptional resolution at the level of individual cells. We hypothesize that profiling the tumor microenvironment at single-cell level and applying an advanced network-based analysis pipeline to treatment-naïve and immunotherapy-treated tumors will improve characterization of the transcriptional program in tumor-infiltrating immune cell types, their association with outcome, and their clinically relevant interactions with tumor cells. Aim 1) Despite high resolution, scRNA-Seq data are typically sparse, with a minority of genes detected in any given cell. We aim to develop a powerful set of tools originating in the Califano Lab for network-based inference of regulatory protein activity in single-cell data, mitigating gene expression dropout and providing a scalable pipeline for inference of cell populations, tumor-immune interactions, and regulatory proteins differentially activated in distinct cell states. We validate this pipeline by comparison to markers concurrently profiled by flow cytometry in a dataset of clear cell renal carcinoma (ccRCC) patients. Aim 2) We will specifically leverage our novel analysis pipeline to interrogate drivers of tumor-infiltrating regulatory T-cells, an immunosuppressive population induced by multiple conventional treatment modalities, including androgen deprivation therapy in prostate cancer. We will validate predicted tumor-infiltration drivers by CRISPR knockout screen and apply an advanced transcriptional perturbation screen to identify drugs which invert the tumor-specific Treg signature. These are expected to serve as prime candidates for future combination immunotherapy studies. Aim 3) We will identify changes in micro-environment induced by immunotherapy in responders and non-responders across two clinical trials of immunotherapy plus androgen deprivation in prostate cancer and one trial of anti-PD1 plus anti-IL1b in ccRCC, identifying potentially novel therapeutic targets. In addition, we will apply our newly developed analytic pipeline to published scRNA-Seq datasets to identify predictors of treatment response in melanoma. With joint guidance from experienced mentors in Immunotherapy and Computational Systems Biology in the setting of Columbia University Medical Center, this project will prepare the trainee for a career as a physician-scientist with a unique background in translational bioinformatics research.