Redirecting T Cell Responses to Self/Tumor Antigens Using Enforced Costimulation

描述(申请人提供)：人们对开发治疗前列腺癌的疫苗很感兴趣，前列腺癌是美国男性最常见的恶性肿瘤。然而，将可移植小鼠肿瘤模型中开发的疫苗转化为治疗人类前列腺癌的努力只取得了有限的成功。在这个项目的初始资金阶段，我们利用一个新的转基因小鼠模型来证明，无法通过接种疫苗获得有效的抗前列腺癌免疫可能至少部分是由于肿瘤反应性T细胞通过外周耐受而失活所致。为了克服T细胞的耐受性，我们和其他人针对不同的共刺激途径来重定向T细胞对耐受性抗原的反应。特别是，我们发现CD134(OX40)和CD137(4-1BB)的强化刺激可以有效地重定向暴露于同源可溶性多肽的CD8T细胞，使其经历大规模的克隆性扩张和效应器分化，而不是无能和缺失。重要的是，含有增强的OX40和/或4-1BB共刺激作用的疫苗可以抑制多种免疫原性小鼠肿瘤的生长，是治疗人类癌症的临床试验的基础。我们进一步测试了增强的OX40/4-1BB双重共刺激在高度严格的条件下重定向T细胞反应的潜力，其中耐受表位来自广泛和结构性表达的实质自身抗原。与可溶性多肽相比，当CD8T细胞遇到同源实质来源的自身抗原时，强制双重共刺激促进了克隆性扩张，但令人惊讶的是，没有促进效应器的分化。具有讽刺意味的是，CD4T细胞具有较弱的效应器分化倾向，在自身抗原和强制双重共刺激的反应下，它被扩增并分化为Th1效应器。因此，CD8T细胞对自身抗原解偶联效应器的特异性分化仅限于CD8T细胞池中，并表明CD8T细胞对强制双重共刺激的反应强烈地受到递呈耐受抗原的APC的影响。然而，当CD4和CD8T细胞在强制双重共刺激过程中同时遇到同源实质自身抗原时，CD8T细胞被推动进行效应器分化。因此，在缺乏CD4T细胞的情况下，强制双重共刺激扩大了“无害”的自我反应性CD8T细胞。值得注意的是，这种“无害的”CD8 T细胞的扩张类似于十年前但无法解释的临床观察，即携带肿瘤的个体通常含有克隆扩增的无能CD8 T细胞，这些细胞是组织/肿瘤相关分化抗原的特异性。因此，我们的模型为理解和克服肿瘤免疫的这一顽固障碍提供了一个独特的机会。这一修订的竞争性更新应用的实验目标将阐明“无害的”CD8 T细胞被扩增的机制，以及重要的是如何在临床相关的前列腺癌模型中推动它们表达治疗有用的效应功能。

DESCRIPTION (provided by applicant): There has been considerable interest in developing vaccines to treat prostate cancer, the most common malignancy in American men. Nevertheless, efforts to translate vaccines developed in transplantable murine tumor models to treat human prostate cancer have only achieved limited success. During the initial funding period of this project we utilized a novel transgenic mouse model to demonstrate that the inability to achieve effective anti-prostate tumor immunity through vaccination is likely at least partially caused by the inactivation of tumor-reactive T cells through peripheral tolerization. To overcome T cell tolerance, we and others have targeted various costimulatory pathways to redirect the response of T cells encountering tolerogenic antigens. In particular, we found that enforced stimulation of CD134 (OX40) and CD137 (4-1BB) potently redirects CD8 T cells exposed to cognate soluble peptide to undergo massive clonal expansion and effector differentiation rather than anergy and deletion. Importantly, vaccines incorporating enforced OX40 and/or 4-1BB costimulation inhibit the growth of a variety of immunogenic murine tumors and are the basis for clinical trials to treat human cancers. We further tested the potential of enforced OX40/4-1BB dual costimulation to redirect T cell responses under highly stringent conditions where tolerizing epitopes derive from a widely and constitutively expressed parenchymal self-antigen. In contrast to soluble peptide, enforced dual costimulation boosted clonal expansion but surprisingly not effector differentiation when CD8 T cells encountered cognate parenchymally-derived self-antigen. CD4 T cells, which ironically possess a weaker tendency to undergo effector differentiation, expanded and differentiated into Th1 effectors in response to self-antigen plus enforced dual costimulation. Thus, specificity to self-antigen uncoupled effector differentiation from expansion exclusively in the CD8 T cell pool, and indicated that the response of CD8 T cells to enforced dual costimulation is strongly influenced by the APC presenting the tolerizing antigen. Nevertheless, when CD4 and CD8 T cells simultaneously encountered cognate parenchymal self-antigen during enforced dual costimulation, CD8 T cells were pushed to undergo effector differentiation. Thus, in the absence of CD4 T cell help enforced dual costimulation expands "harmless" self-reactive CD8 T cells. Notably, this expansion of "harmless" CD8 T cells is analogous to the decade-old but unexplained clinical observation that tumor-bearing individuals often harbor clonally expanded populations of anergic CD8 T cells specific for tissue/tumor- associated differentiation antigens. Our model thus represents a unique opportunity to understand and overcome this stubborn barrier to tumor immunity. The experimental aims of this revised competitive renewal application will elucidate the mechanism by which "harmless" CD8 T cells are expanded, and importantly how they can be pushed to express therapeutically useful effector functions in a clinically relevant prostate cancer model.