Development of a Computational Framework for TCR Engineering

描述(申请人提供)：T细胞受体(TCR)是疾病免疫反应的关键介质，具有显著的治疗潜力。例如，临床试验表明，过继转移表达特定抗原特异性TCRs的基因工程T细胞可以导致体内肿瘤的消退。然而，TCR的治疗能力是有限的，因为它们的结合亲和力相对较弱，而基因工程T细胞的疗效受到转导的TCR链与内源性TCR链错配的限制。通过体外分子进化提高TCR亲和力的努力可能会影响抗原特异性，而改善链配对的尝试尚未产生有效的解决方案。在这里，我们建议开发一个计算模型框架来设计具有高亲和力、高抗原特异性和有偏链配对的TCR。结构导向的计算设计将用于全面扫描界面突变，模拟它们对特异性和亲和力的影响，并针对关键抗原决定因素量身定做设计。这项工作将通过两个具有互补专业知识的实验室的密切互动来完成：TCR的蛋白质相互作用建模和基于结构的设计(Pi Zhiping Weng)和实验TCR生物物理(Pi Brian Baker)。实验评估，包括亲和力的测量以及功能和特异性的免疫学分析，将用于反复测试和改进建模框架。该框架和相关算法将通过一个网络可访问的数据库公开发布，该数据库作为测量野生型和突变TCR结合亲和力和结构的储存库。该建议有三个具体目的：目的1.设计TCR的点突变以提高对抗原的亲和力和特异性；目的2.对具有多个替换的TCR环进行灵活的重新设计，以进一步改善和重定向结合；目的3.通过计算设计的恒定区对工程TCR链进行偏向特异性配对。

DESCRIPTION (provided by applicant): T cell receptors (TCRs) are key mediators of the immune response to disease, with significant therapeutic potential. For example, clinical trials have shown that adoptive transfer of T cells genetically engineered to express TCRs of defined antigen specificity can lead to tumor regression in vivo. However, the therapeutic capabilities of TCRs are limited due to their relatively weak binding affinities, and the efficacy of genetically engineered T cells is limited by the mispairing of transduced TCR chains with those endogenously produced. Efforts to improve TCR affinity via in vitro molecular evolution can impact antigen specificity, and attempts to improve chain pairing have yet to yield an effective solution. Here we propose to develop a computational modeling framework to engineer TCRs with high affinity, high antigen specificity, and biased chain pairing. Structure-guided computational design will be used to comprehensively scan interface mutations, model their impact on specificity and affinity, and tailor designs to target key antigenic determinants. This work will be accomplished through the close interaction of two laboratories with complementary expertise: protein interaction modeling and structure-based design of TCRs (PI Zhiping Weng) and experimental TCR biophysics (PI Brian Baker). Experimental assessments, including measurements of affinities and immunological assays of function and specificity, will be used to iteratively test and improve the modeling framework. The framework and associated algorithms will be publically released, facilitated via a web-accessible database serving as a repository for measurements of wild type and mutant TCR binding affinities and structures. The proposal has three specific aims: Aim 1. Design point mutations in TCRs to improve affinity and specificity towards antigen; Aim 2. Perform flexible redesign of TCR loops with multiple substitutions to further improve and redirect binding; Aim 3. Bias specific pairing of engineered TCR chains via computationally designed constant regions.