【作者】 舒驰；

【导师】 吴玉章；

【作者基本信息】 第三军医大学 ， 免疫学， 2011， 硕士

【摘要】 I型糖尿病（Type 1 diabetes,T1D）是由自身反应性T细胞介导的以胰腺β细胞损伤为主要特征的一种器官特异性自身免疫性疾病。无论在非肥胖糖尿病（nonobese diabetic,NOD）小鼠还是T1D患者,自身反应性CD8⁺细胞毒性T淋巴细胞（cytotoxic T lymphocyte）是T1D发病不可或缺的直接效应细胞。针对β细胞自身反应性T细胞进行特异性的免疫干预与阻断,重塑机体对β细胞自身抗原的免疫耐受,是一种可能从根本上预防T1D发生或发展的理想策略。抗原特异性治疗与其他形式的治疗方案相比,其优势在于无潜在的代谢活性并且能限制目的致病表位肽的反应范围。愈来愈多证据表明:负向调节疫苗是针对自身免疫性疾病治疗的一种非常有效的抗原特异性免疫调节剂。它通常是根据天然表位肽人工修饰设计合成,能特异性下调致病自身反应性T细胞应答效应的拮抗型改造肽（altered peptide ligands, APL）。大量研究结果提示,反复给予天然肽或拮抗型表位改造肽可通过诱导自身反应性CD8⁺T细胞克隆无能或活化后凋亡（AICD）从而实现特异性的免疫耐受。改造肽是在天然表位肽的基础上对表位进行氨基酸改造形成的具有更高稳定性和免疫调节活性的短肽。目前常用的APL设计方法包括:1.TCR结合位点的替换;2.MHC-I类分子锚着残基替换;3.残基的特殊修饰,已成功应用于其他自身免疫疾病如多发性硬化（multiple sclerosis,MS）负调疫苗的治疗研究中。胰岛素是启动T1D唯一所必需的关键靶抗原,胰岛素特异性CD8⁺T细胞被发现是在T1D患者和NOD小鼠中最早浸润至胰岛的CD8⁺T细胞,它通过启动β细胞损伤,致使β细胞释放大量危险信号并暴露更多自身抗原,进而通过表位扩展,形成针对更广泛自身抗原的“瀑布式”自身免疫反应。因此,针对胰岛素特异性CD8⁺T细胞的免疫干预,对于阻断T1D的自身免疫启动显得尤为关键。基于能够模拟T1D患者疾病过程的人源化HLA（human leucocyte antigen,人白细胞抗原）转基因NOD小鼠模型（NOD.β2mnullHHD）,已陆续鉴定了胰岛素来源HLA-A*0201限制性CTL表位,其中胰岛素A链2-10 (mInsA_2-10)和胰岛素B链5-14(mInsB_5–14)与人胰岛素序列高度同源,且具有交叉反应性,为T1D的免疫干预以及将来的临床转化研究提供了良好的靶标。本研究根据自身抗原表位肽mInsA_2-10和mInsB_5–14设计计算得到若干APL,在多克隆T细胞动物模型NOD.β2mnullHHD中筛选得到能下调肽特异性自身免疫性CD8⁺T细胞应答效应的拮抗型APL,从而得到具有能预防、保护效应的I型糖尿病负调疫苗。本文第一部分,我们通过三方面实验验证上述两条已知自身抗原表位的自身反应免疫原性。通过A2分子相对亲和力实验证实该表位是HLA-A*0201限制性的,在人源化NOD小鼠T细胞功能水平通过CTL效应细胞因子分泌检测和CFSE增殖检测证实:其具备包含特异性分泌IFN-γ以及特异性增殖反应等表位特异性自身反应性。即mInsA_2-10、mInsB_5–14是具有优势免疫原性的HLA-A*0201限制性自身反应性CD8⁺T细胞表位。第二部分,我们通过反复系统性给予可溶性抗原的方式分别将两种自身抗原肽经腹腔注射给予幼龄小鼠,观察是否产生保护性预防T1D发生的作用。结果,给予可溶性mInsA_2-10并不能对人源化NOD小鼠产生保护作用,反而加速了其T1D的疾病进程以及增加发病率。mInsB_5–14处理组虽然能够降低人源化NOD小鼠胰腺的病理损伤以并缓解淋巴细胞浸润程度,以及降低分泌IFN-γ的CTL频率,但对T1D发病率并没有显著的干预作用。第三部分,我们首先利用InsightII工作站分别模拟构建表位mInsA_2-10与HLA-A2分子的peptide-MHC三维复合体(mInsA_2-10-HLA-A2)。分别通过非天然氨基酸逐个替换的方法和TCR位点单个天然氨基酸的方法对mInsA_2-10与mInsB_5–14进行改造,通过计算各候选APL与HLA分子结合能量变化、RMSD、以及氢键数量,初步筛选得到基于mInsA_2-10的4条APL与基于mInsB_5–14P6位点改造的6条APL作为候选APL用于后续生物功能的筛选。通过与HLA-A2分子亲和力检测、CFSE染色增殖抑制能力分析以及IFN-γ分泌检测等实验鉴定候选APL中具有拮抗功能的APL。结果显示,基于mInsA_2-10第4位点改为D型谷氨酸的mInsA_2-10D-Q4和第6位点改为D型半胱氨酸的mInsA_2-10D-C6为拮抗型APL,可以明显下调靶器官胰岛浸润自身反应性CD8⁺T细胞特异性分泌IFN-γ并抑制CD8⁺T细胞在mInsA_2-10刺激下的增殖反应。根据mInsB_5–14第6位由组氨酸改为苯丙氨酸的改造肽mInsB_5–14H6F也能显著下调肽特异性外周自身反应性CD8⁺T淋巴细胞分泌IFN-γ的应答水平。综上,本研究在HLA-A*0201转基因NOD小鼠模型中验证了来自鼠胰岛素的自身抗原表位mInsA_2-10和mInsB_5–14是HLA-A*0201限制性免疫优势自身反应性CD8⁺T细胞表位;发现通过腹腔注射方式反复给予T1D相关自身抗原CD8⁺T细胞表位mInsA_2-10或mInsB_5–14并不能够在该人源化NOD小鼠体内产生预防T1D发生的保护效应;通过对mInsA_2-10与mInsB_5–14进行计算机辅助模拟氨基酸替换改造,并通过体外细胞功能实验证实:mInsA_2-10D-C6、mInsA_2-10D-Q4与mInsB_5–14H6F均为具有负调天然肽特异性CTL应答功能的拮抗型APL,今后我们将在人源化NOD小鼠模型体内实验中进一步验证这三条拮抗型APL的治疗效应,从而将这些肽用于抗T1D治疗性疫苗研发的设计应用当中。本课题所建立的基于分子模拟、分子动力学并结合自由能计算的计算机辅助APL设计平台以及利用人源化疾病动物模型的生物学功能筛选体系,可以极大的提高负调疫苗的研究效率。更多还原

【Abstract】 Type 1 Diabetes （T1D） is an autoimmune disease characterized by T-cell–mediated elimination of insulin-producing pancreatic isletβ-cells in both humans and nonobese diabetic （NOD） mice. Cytotoxic CD8⁺ T lymphocyte （CTL）, reactive toβcell antigens, are required for T1D development in the NOD mouse model of the disease, and CD8⁺ T cells specific forβcell antigens can be detected in the peripheral blood of T1D patients. Antigen-specific immunotherapy is expected to be an ideal strategy for treating T1D. Recently, multiple lines of investigation have demonstrated downregulation of CTL response associated with diabetogenic auto-reactive T cell by antagonist peptide vaccination is an efficacy approach for T1D prophylaxis and therapy. Administration of self-peptides or altered peptide ligands （APLs） for the induction of diabetogenic CD8⁺ T cell tolerance provides a new strategy for diabetes immunotherapy. The advantage of peptide-specific therapy over other forms of therapy is that it lacks potential metabolic activity and can limit the range of the response to the desired pathogenic peptide epitopes without increasing the possibility for hyperactivation of self-reactive T cells. Recently, modulation of CTL response by manipulating T cell epitopes has been considered as a particularly attractive approach for autoimmune disease immunotherapy.Insulin was the firstβ-cell protein to which an autoimmune response was documented in type 1 diabetic patients, and accumulating evidence suggests it is a key target of pathogenic T-cells in both NOD mice and human. mInsB_5–14 and mInsA2–10, derived from insulin were confirmed to be the immune-dominant HLA-A*0201 restricted CTL epitopes in the humanized HLA transgenic NOD model （NOD.β2mnullHHD） which expressed chimeric HLA-A*0201 molecule.There are few research about modulation of CTL response by epitope or APL for T1D immunotherapy in polyclone animal models. APLs with increased pMHC complex affinity for the TCR molecule, which are designed by modifications at the TCR contact sites rather than the MHC anchor residues, have an expected potency to induce weaker T cell responses and may even alter cross-reactive T cells from a hyperactivate state. There have been many successful studies on evaluating the design of APLs for similar or increased MHC binding affinity through the calculation of pMHC interaction energies using in silico strategy.In this study, we design several altered peptide ligands （APL） based on these two peptides to downregulate the peptide specific auto-reactive CD8⁺ T cell responses in order to obtain an antagonist peptide vaccine for T1D therapy.In the first part of this study, we identified the established autoreactive epitopes again through HLA-A*0201 binding assays, ELISpot IFN-γsecretion assays and CFSE proliferation assays ensure that they are HLA-A*0201 restricted immunodominant self-antigen epitopes in humanized NOD mice T1D development. Then, we found that administration of these two HLA-A*0201 restricted auto-antigen epitopes to NOD.β2mnullHHD mice by intraperitoneal injection cannot prevent the development of T1D. Administration of the soluble mInsA_2-10 in 4 week-old cannot protect them from T1D, however would even accelerate the occurrence and increase the incidence of T1D. There are no significant protective effects from T1D in soluble mInsB_5–14 administration group either.In the third part, we generated several altered peptide ligands of mInsA_2-10 and mInsB_5–14 with D-animo mutation or single TCR contact residue substitution by InsightII studio。The candidate antagonist APLs were identified by HLA-A2 binding, CFSE proliferation suppressing assay and IFN-γsecretion inhibition assay. It is demonstrated that mInsA_2-10D-C6 and mInsA_2-10D-Q4, APLs with D-cysteine substitution at site 6 and D-glutamic substitution at site 4 were found as potential antagonist APLs for that they could obviously reduce the IFN-γsecretion of islet-infiltrating CTL and suppress the specific proliferation of CTLs by mInsA_2-10 stimulation. mInsB_5–14H6F, an APL with phenylalanine substitution at residues 6 was verified as a potential antagonist APL which could also significantly downregulate the mInsB_5–14-specific CTL responses.We concluded that mInsA_2-10 and mInsB_5–14 derived from murine inuslin are both immunodominant HLA-A*0201 restricted CTL epitope in NOD.β2mnullHHD mice; There were no protective effects in administration of the soluble wild peptide mInsA_2-10 or mInsB_5–14 in humanized NOD mice; mInsA_2-10D-C6 and mInsA_2-10D-Q4 derived from mInsA_2-10 and mInsB_5–14H6F derived from mInsB_5–14 were all identified to be antagonist APLs in vitro assays.These three APLs could be used for treating a CD8⁺ T cell-mediated diabetes model,and it will be of interest to further examine the immunological effects of the antagonist APLs in the humanized mice model in order to make this peptide to be applicable for anti-T1D therapeutic vaccine design.更多还原