Permissive Specificity in Peptide Binding to HLA-DR

描述（由申请人提供）：当CD4+ T细胞识别主要组织相容性复合体（MHCII） II类分子呈递的肽抗原时，适应性免疫反应开始。MHCII分子的两个突出特征是它们的多态性和每个等位基因结合大量肽的能力。我们的长期目标是正式描述肽识别过程的三个阶段：1)肽- mhcii复合物的形成，2)该复合物将被选择用于呈现，3)该复合物将被T细胞识别。预测表位和T细胞对它们的反应在MHCII发挥重要作用的两个实际领域是重要的，即疫苗接种和移植。本应用程序的目标是详细描述这三个阶段中的第一个阶段，即HA-DR （DR）结合肽的预测。这里提出的研究将产生一个结合过程的热力学描述，足以为生成预测算法提供基础。预测绑定的能力将是取得总体进展的重要一步。MHCII和多肽是一种灵活的结合系统，具有特异性和容许性。这种灵活性被热力学现象所证明，特别是协同性的存在和等温熵焓补偿（iEEC）的发生。协同性在一个相互作用受到其他相互作用影响的系统中被观察到，并与蛋白质折叠有关。iEEC描述了熵（增加对结构空间的搜索）补偿较差的结合能（焓）的能力。这是容许性背后的驱动机制，因为存在一个亲和范围，在这个亲和范围内，具有较差结合特性的肽有可测量的概率找到允许结合的构象。为了在结合预测模型中包含灵活性，需要在分子水平上更深入地了解协同性和iEEC之间的相互作用，以确定肽结合的允许特异性。为了将我们的研究应用于所有DR等位基因，我们还必须测量定义不同等位基因的多态磁带的贡献。因此，在Aim 1中，我们将研究肽/DR复合物形成过程中涉及的能量学，将其反卷积为焓和熵组分，并将iEEC的程度与协同性联系起来。在目标2中，我们将定义DR残基对亲和力、折叠和能量学的生物物理贡献，以概述一个适用于所有HLA等位基因的结合模型。肽结合通常发生在辅助分子HLA-DM （DM）存在的情况下。我们有证据表明，DM通过设置容许阈值来影响肽结合。因此，在Aim 3中，我们将从热力学角度研究dr -肽-DM相互作用，将DM活性与各种pDR复合物的热力学特征联系起来。在目标4中，我们概述了一个初始的肽结合预测模型，该模型包含了前三个目标中导出的结构和热力学数据，我们建议通过预测随机生成的抗原序列中的结合肽来验证该方法。

DESCRIPTION (provided by applicant): The adaptive immune response starts when CD4+ T cells recognize peptide antigens presented by class II molecules of the Major Histocompatibility Complex (MHCII). Two outstanding features of MHCII molecules are their polymorphism and the ability of each allele to bind a large panoply of peptides. Our long-term goal is to formally describe the three stages of the peptide recognition process: 1) formation of a peptide-MHCII complex, 2) will the complex be selected for presentation, and 3) will the complex be recognized by a T cell. Predicting epitopes and T cell responses to them is important in two practical areas where MHCII plays a significant role, vaccination and trasnplantation. The goal of the present application is to characterize in detail the first of these three stages, i.e., the prediction of HA-DR (DR) binding peptides. The research proposed here will generate a thermodynamic description of the binding process sufficient to provide a basis for generating predictive algorithms. The ability to predict binding will be an important step in making progress overall. The MHCII and peptides represent a flexible binding system, featuring specificity and permissiveness. This flexibility is evidenced by thermodynamic phenomena, in particular the presence of cooperativity and the occurrence of isothermal entropy-enthalpy compensation (iEEC). Cooperativity is observed in systems where one interaction is affected by other interactions and is associated with protein folding. iEEC describes the ability of entropy (increased search of structural space) to compensate for poorer binding energy (enthalpy). This is the driving mechanism behind permissiveness, in that there is an affinity range in which a peptide with poor binding characteristics has a measurable probability of finding a conformation that allows binding. To include flexibility in a binding prediction model, a deeper understanding of the interaction between cooperativity and iEEC in determining the permissive specificity of peptide binding at the molecular level is needed. To apply our studies to all DR alleles we have to also measure the contribution of the polymorphic cassettes that define the different alleles. Therefore, in Aim 1 we will investigate the energetics involved in peptide/DR complex formation, deconvoluted into enthalpic and entropic components and we will correlate the extent of iEEC to cooperativity. In Aim 2 we will define the biophysical contribution of DR residues to affinity, folding, and energetics to outline a binding model valid for all HLA alleles. Peptide binding usually takes place in the presence of an adjunct molecule, HLA-DM (DM). We have evidence that DM affects peptide binding by setting a permissiveness threshold. Therefore in Aim 3 we will investigate the DR-peptide-DM interaction in thermodynamic terms, correlating DM activity to the thermodynamic signatures of various pDR complexes. In Aim 4 we outline an initial peptide-binding prediction model encompassing the structural and thermodynamic data derived in the first three aims and we propose to validate this approach by predicting binding peptides within randomly generated antigenic sequences.