Recognizing the tumor ecosystem: Integrating stromal and cancer antigen signals to achieve precision recognition of solid tumors by CAR T cells

识别肿瘤生态系统：整合基质信号和癌抗原信号，实现CAR T细胞对实体瘤的精准识别

• 批准号: 10559489
• 负责人: WENDELL A LIM
• 金额: $ 55.2万
• 依托单位: UNIVERSITY OF CALIFORNIA, SAN FRANCISCO
• 依托单位国家: 美国
• 财政年份: 2020
• 资助国家: 美国
• 起止时间: 2020-12-01 至 2026-11-30
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10559489
• 关键词: Antigen Presentation; Antigens; B lymphoid malignancy; Cell Therapy; Cells; Complex; Cytotoxic T-Lymphocytes; Ecosystem; Endothelial Cells; Engineering; Epithelium; Fibroblasts; Goals; Growth; Immune; Immunotherapeutic agent; In Vitro; Individual; Malignant - descriptor; Malignant Neoplasms; Malignant neoplasm of lung; Malignant neoplasm of ovary; Molecular; Normal tissue morphology; Organ; Play; Pre-Clinical Model; Role; Signal Transduction; Solid; Solid Neoplasm; Source; Specificity; Stromal Cells; Stromal Neoplasm; Supporting Cell; Surface Antigens; T-Lymphocyte; Testing; Therapeutic; Therapeutic Agents; Tissues; Toxic effect; Tumor Antigens; VTCN1 gene; Work; Xenograft Model; cancer cell; candidate identification; cell killing; cell type; chimeric antigen receptor; chimeric antigen receptor T cells; combinatorial; cross reactivity; design; efficacy evaluation; engineered T cells; improved; mesothelin; mouse model; neoplastic cell; pancreatic cancer model; receptor; spatial integration; success; tool; tumor; tumor microenvironment; tumor xenograft

Project Summary/Abstract Despite the remarkable success of engineered chimeric antigen receptor (CAR) T cells in the treatment of B cell malignancies, their application to solid cancers has been far less successful. One of the major challenges limiting their utility is the difficulty in identifying ideal surface antigens that can be used to discriminate between cancer and normal tissues – many potential targets that are highly expressed in solid tumors are also found at lower levels in normal epithelial organs, leading to off-tumor toxicity. Nonetheless, we know that solid tumors comprise a complex and sophisticated tissue with a distinct ecosystem of malignant, immune and stromal cells. From first principles, one would predict that there should be ample discriminatory information in the tumor, if one could design therapeutic T cells that could integrate information from across different cells in the tumor ecosystem. We have recently developed new CAR T cell recognition circuits that can sense and respond to combinations of antigens, even if they are present on distinct cells within the same tissue microenvironment. These circuits utilize a synNotch receptor to detect a priming antigen, which in turn induces the expression of a CAR that kills cells based on a killing antigen. In preliminary results, we have shown that T cells with this kind of prime-and-kill circuit can recognize unique combinations of neighboring cells to induce killing. These types of engineered T cells are one of the first known therapeutic agents that can integrate molecular information from across different cells within the same tissue. In this proposal, we hypothesize that this prime-and-kill T cell recognition circuit could be used to recognize solid tumors based on information distributed across the tumor ecosystem. Specifically, we will target combinatorial integration of signals that are present in cancer cells and cancer-associated stromal cells, which play a central supportive role in a number of solid cancers. As a test case, we propose to investigate whether antigens from cancer associated fibroblasts can be used to locally prime CAR T cells to then kill based on a cancer associated antigen. Even if this cancer associated antigen in not perfectly specific (i.e., it is expressed in other normal tissues), the combination of stromal and cancer cell signals should be far more specific for the tumor. Prior efforts have unsuccessfully explored using single antigen CARs to target stromal or cancer cells individually, but here we test whether using integrated combinatorial recognition of the cancer cell/stromal cell ecosystem can result in significantly improved recognition specificity. If so, then this kind of integrated tumor ecosystem recognition could be applied to a large number of solid cancers.

项目总结/摘要 尽管工程化嵌合抗原受体（CAR）T细胞在B细胞治疗中取得了显著的成功， 虽然它们在恶性肿瘤中的应用远没有在实体癌中成功。其中一个主要的挑战是限制 它们的用途是难以鉴定可用于区分癌症、非癌症和非癌症的理想表面抗原 和正常组织-许多在实体瘤中高度表达的潜在靶点也在较低的表达水平上被发现。 正常上皮器官中的水平，导致肿瘤外毒性。尽管如此，我们知道实体瘤包括 一种复杂而精密的组织，具有独特的恶性、免疫和基质细胞生态系统。从 根据第一原理，如果可以的话，可以预测在肿瘤中应该有足够的区分信息。 设计治疗性T细胞，可以整合肿瘤生态系统中不同细胞的信息。 我们最近开发了新的CAR T细胞识别电路，可以感知和响应以下组合： 抗原，即使它们存在于相同组织微环境内的不同细胞上。这些电路 利用synNotch受体检测引发抗原，引发抗原反过来诱导CAR的表达， 基于杀伤抗原的细胞。在初步结果中，我们已经表明，具有这种引发和杀伤作用的T细胞， 电路可以识别相邻细胞的独特组合以诱导杀伤。这些类型的工程T 细胞是第一个已知的治疗剂，可以整合来自不同的分子信息， 同一组织中的细胞。 在这个提议中，我们假设这种引发和杀死T细胞的识别电路可以用来识别固体 基于分布在肿瘤生态系统中的信息来评估肿瘤。具体来说，我们将针对组合 整合存在于癌细胞和癌症相关基质细胞中的信号，这些信号在肿瘤细胞中起着中心作用。 在许多实体癌中起支持作用。作为一个测试案例，我们建议调查抗原是否来自 癌症相关的成纤维细胞可用于局部引发CAR T细胞，然后基于癌症相关的成纤维细胞杀伤CAR T细胞。 抗原的即使这种癌症相关抗原不是完全特异性的（即，它以其他正常的方式表达， 组织），基质和癌细胞信号的组合应该对肿瘤更具特异性。之前 已经进行的努力没有成功地探索使用单抗原汽车单独靶向基质细胞或癌细胞， 在此，我们测试使用癌细胞/基质细胞生态系统的整合组合识别是否可以 导致识别特异性显著提高。如果是这样，那么这种整合的肿瘤生态系统 识别可以应用于大量的实体癌。