Pre-Clinical Core

核心2项目总结 临床前模型和病理学核心 临床前核心（核心#2）利用了几位Moffitt研究者的知识和专业知识， 核心设施，并将追求三个目标，系统地支持计划项目在三个关键 阶段; 1）模型开发，2）结果的测量和分析，以及3）在人类队列中的验证 和类器官模型。在目标1中，核心#2负责小鼠模型生成和表征。的 核心将使用胚胎干细胞基因工程小鼠模型（ESC-GEMM）的方法， 肺癌小鼠模型的快速生成。为此，将从现有的肺中建立ESC， 癌GEMM也含有调节等位基因，用于有效靶向ESC和控制基因 表情将用诱导型表达盒靶向ESC，然后注射到胚泡中， 产生嵌合动物。此类嵌合体中的胚胎干细胞来源的肺细胞携带诱导肺形成所需的等位基因 肿瘤发生，嵌合体将被用作实验动物而无需进一步繁殖。这种方法 允许非常快速地产生具有调节的代谢酶的表达或活性的肺癌模型， 感兴趣的酶在目标2中，核心#2将简化这些小鼠中肿瘤负荷的定量。 通过提供鼠肺癌模型的数字病理学（DP）分析来建立模型。作为输入，组织学切片 使用自动载玻片扫描仪在多个放大倍数下成像。由此产生的高分辨率图像 使用我们开发的创新机器学习算法进行处理和分析。平台 允许肺肿瘤和整体内的单个细胞的鉴定、表征和定量。 肺切片将应用深度学习方法来提供肺的全面表征 肿瘤，包括肿瘤分级和免疫细胞浸润的评估。组织学样本和 由该计划项目创建的数据集将用于创建新算法的训练库 发展DP平台允许高度准确和快速的图像分析， 肺肿瘤模型的表征。在目标3中，核心2将提供资源和专业知识， 前两个目标的小鼠研究结果。核心将提供大的，注释良好的人类 由组织微阵列代表的队列和用于生成类器官模型的集中式工作流程。 核心将协助项目规划和方法开发，诊断咨询，生物标志物评分， 结果解释和手稿撰写。其他病理学资源将包括显微镜评价 动物和人体组织的充分性和诊断，抗体选择指导，优化 免疫组化（IHC）和多重免疫荧光成像。

CORE 2 PROJECT SUMMARY PRECLINICAL MODELS AND PATHOLOGY CORE The Preclinical Core (Core #2) leverages the knowledge and expertise from several Moffitt investigators and core facilities and will pursue three aims that systematically support the Program Projects at three critical stages; 1) model development, 2) measurement and analysis of outcome, and 3) validation in human cohorts and organoid models. In the Aim 1, Core #2 is tasked with mouse model generation and characterization. The Core will use the embryonic stem cell-genetically engineered mouse model (ESC-GEMM) approach to enable the rapid generation of lung cancer mouse models. To this end, ESCs will be established from existing lung cancer GEMMs that also harbor regulatory alleles for efficient targeting of ESCs and control of gene expression. ESCs will be targeted with inducible expression cassettes, followed by injection into blastocysts to generate chimeric animals. ESC-derived lung cells in such chimeras carry the alleles necessary to induce lung tumorigenesis, and chimeras will be used as experimental animals without further breeding. This approach allows for the very rapid generation of lung cancer models having modulated expression or activity of metabolic enzymes of interest. In Aim 2, Core #2 will streamline the quantification of tumor burden in these mouse models by providing digital pathological (DP) analysis of murine lung cancer models. As input, histology slides are imaged at multiple magnifications using automated slide scanners. The resulting high-resolution images are processed and analyzed using innovative machine learning algorithms we have developed. The platform allows for the identification, characterization, and quantification of lung tumors and individual cells within whole lung sections. Deep learning methods will be applied to provide a comprehensive characterization of lung tumors, including tumor grading and the assessment of immune cell infiltration. Histology samples and datasets created by the program projects will be used to create training libraries for new algorithm development. The DP platform allows for highly accurate and rapid image analysis, accelerating the characterization of lung tumor models. In Aim 3, Core #2 will provide resources and expertise to validate outcomes from the mouse studies of the first two Aims. The Core will provide large, well-annotated human cohorts represented by tissue microarrays and a centralized workflow for the generation of organoid models. The Core will assist with project planning and method development, diagnostic consultation, biomarker scoring, result interpretation, and manuscript writing. Additional pathology resources will include microscopic evaluation of animal and human tissues for adequacy and diagnosis, antibody selection guidance, optimization for immunohistochemistry (IHC), and multiplexed immunofluorescence imaging.