Comprehensive Glycoproteomic Tool Development for Cancer Biomarkers

描述(由申请人提供)：该项目致力于寻找更好的方法来破译临床样本中蛋白质的糖基化。糖基化是蛋白质结构和功能的重要修饰物，在疾病过程中起重要作用。但目前我们对大多数蛋白质的糖基化知之甚少。目前用于探测蛋白质上的多糖的方法不适合于满足这一需求，因为它们需要大量的材料和许多加工步骤。在这里，我们提出了一种实用的方法来探测蛋白质的糖基化作用，它将提供：1)用有限的样本获得结构和组成信息的能力 用途；2)能够精确地比较样本之间的葡聚糖水平；以及3)准备好转化为临床检测。我们将通过新的信息学技术来实现这一目标，这些技术有助于结合使用质谱学(MS)和凝集素结合来研究多糖。第二阶段将专注于胰腺癌的糖蛋白生物标记物。MS提供了多糖的单糖组成和一些序列信息，但它留下了关于序列或连锁变体的模棱两可的信息。同样，凝集素可以用少量的样品精确测量特定的结构，但它们不能提供每个糖链的完整图像。我们预测，定量地整合这两种类型的信息将比单独使用任何一种方法提供更准确的信息。我们将使用糖的基序-亚结构的共同语言将凝集素实验与MS实验定量地联系起来。在目标1中，我们将开发一种算法，根据凝集素结合确定样品中最有可能存在的糖链基序。在目标2中，我们将开发集成工具 凝集素和MS数据，并将使用该方法来表征和比较三种不同纯化糖蛋白的葡聚糖。我们将确定MS和凝集素数据的连接是否提供了比这两种方法单独提供的更完整的信息，而且样本消耗有限，能够在样本之间进行精确的比较。

DESCRIPTION (provided by applicant): This project addresses the need for better methods for deciphering the glycosylation of proteins in clinical samples. Glycosylation is an important modifier of protein structure and function and contributes to disease processes. But we currently know little about the glycosylation of most proteins. The current methods for probing glycans on proteins are not suitable for meeting this need, as they require much material and many processing steps. Here we propose and practical approach to probing protein glycosylation that will provide: 1) the ability to obtain structural and compositional information with limited sample usage; 2) the ability to precisely compare glycan levels between samples; and 3) ready translation into a clinical assay. We will achieve this goal through novel informatics techniques that facilitate the combined use of mass spectrometry (MS) and lectin binding for studying glycans. Phase II will focus on glycoprotein biomarkers of pancreatic cancer. MS provides the monosaccharide compositions of glycans and some sequence information, but it leaves ambiguities about sequence or linkage variants. Likewise, lectins can give precise measurements of specific structures using small amounts of sample, but they do not provide a complete picture of each glycan. We predict that quantitatively integrating the two types of information will give more accurate information than either method alone. We will quantitatively link lectin experiments to MS experiments using the common language of motifs - substructures of glycans. In Aim 1, we will develop an algorithm for identifying what glycan motifs are most likely present in a sample based on lectin binding. In Aim 2, we will develop tools for integrating lectin and MS data and will use the method to characterize and compare the glycans of three different purified glycoproteins. We will determine whether the linking of MS and lectin data provides more complete information than either method alone, with limited sample consumption and the ability to make precise comparisons between samples.