Building of a functional single cell proteomics resource for the study the DNA damage response

DNA损伤反应（DDR）在确保基因组稳定性和细胞存活方面至关重要。基于抗体的免疫荧光技术为识别DNA损伤细胞和研究DDR机制提供了独特的机会。我以前已经开发了一种空间定向的单细胞蛋白质组学方法的多重循环免疫荧光（Cyc-IF），它允许可视化和分析60蛋白在单细胞水平在一个单一的组织载玻片。在一次检测中分析多种蛋白质，可以以前所未有的规模研究细胞状态。在这里，我建议使用Cyc-IF平台来开发DDR检测，并创建DDR细胞表型图谱，以表征与DNA损伤相关的可能细胞状态。AIM 1. DDR细胞模型的开发。我们将通过用不同种类的DNA损伤诱导剂处理基因工程小鼠细胞系来产生DDR的细胞模型。这组细胞系包含DDR相关基因中的遗传畸变，包括参与DNA修复、细胞周期和促存活信号传导途径的基因。细胞系将用诱导各种类型DNA损伤的试剂（例如辐照、烷化剂、DNA损伤检查点抑制剂）处理。将使用经处理和未经处理的细胞构建组织微阵列（TMA）。我们之前已经在Cyc-IF平台上获得并验证了许多抗体。我们将用DDR特异性抗体完成我们的小组，以研究DNA损伤，DNA修复，表观遗传改变，细胞周期，应激反应，代谢和主要信号通路。TMA将通过Cyc-IF与我们的抗体组进行分析，每个样本将通过我们预先建立的计算管道在单细胞水平上进行分析。这将允许鉴定DNA损伤的细胞和相关的表型。AIM 2. DDR细胞表型图谱的构建。机器学习将用于对与DDR相关联的单元状态进行分类。为了提高我们方法的鲁棒性，数据将被分成训练集和测试集。我们将首先确定哪些标记是高度相关的或反相关的。无监督分层聚类（例如PCA/tSNE/k均值）将用于鉴定与DDR相关的细胞群，与细胞基因型无关，逻辑回归将用于鉴定可预测不同类型DNA损伤和修复机制存在的蛋白质。最后，我们将使用一组标记和表达分布作为背景，以开发将在未来实验中使用的归一化方法。总之，该计划将有助于确定与DDR相关的特定细胞状态，并将用于建立一个新的资源，这将是特别有用的研究细胞畸变，导致DNA损伤的积累。

The DNA damage response (DDR) is essential in ensuring the preservation of genomic stability and cell survival. Antibody-based Immunofluorescence represent a unique opportunity to identify cells with DNA damage and to study mechanisms of DDR. I have previously developed a spatial oriented single-cell proteomics approach of multiplex cyclic-immunofluorescence (Cyc-IF), which allows the visualization and analysis of 60 proteins at the single cell level in a single tissue slide. Analysis of multiple proteins in a single assay allows to study cell states with unprecedented dimensions. Here, I propose to use the Cyc-IF platform to develop a DDR assay and to create a DDR cell phenotype atlas to characterize the possible cell states associated with DNA damage. AIM 1. Development of cellular models of DDR. We will produce cellular models of DDR through treatment of genetically engineered murine cell lines with different classes of DNA-damage inducing agents. This panel of cell lines contains genetic aberrations in DDR related genes including genes involved in the DNA repair, cell cycle and pro-survival signaling pathways. The cell lines will be treated with agents that induce various type of DNA damage (e.g. irradiation, alkylating agents, DNA damage checkpoint inhibitors). A tissue microarray (TMA) will be constructed using the treated and untreated cells. We have previously acquired and validated many antibodies on the Cyc-IF platform. We will complete our panel with antibodies specific for the DDR in order to study DNA damage, DNA repair, epigenetic alterations, cell cycle, stress response, metabolism and major signaling pathways. The TMA will be analyzed through Cyc-IF with our antibody panel and each sample will be analyzed at the single cell level through our pre-established computational pipeline. This will allow the identification of DNA damaged cells and associated phenotypes. AIM 2. Building of a DDR cell phenotype atlas. Machine learning will be used to classify the cell states associated to the DDR. To increase the robustness of our approach, the data will be separated into a training set and a test set. We will first determine which markers are highly correlated or anti-correlated. Unsupervised hierarchical clustering (e.g. PCA/tSNE/k-mean) will be used to identify cell populations that are associated to the DDR, independently of the cell genotype, and a logistic regression will be used to identify proteins that can predict the presence of different type of DNA damage and repair mechanisms. Finally, we will use the panel of markers and expression distribution as a background to develop a normalization method that will be used across future experiments. In conclusion, this program will help identify specific cell states associated with the DDR and will be used to build a new resource that will be particularly useful in studying cell aberrations that result from the accumulation of DNA damage.