Department of Biosciences, Bioanalytical Research Laboratories , University of Salzburg , Hellbrunner Straße 34 , 5020 Salzburg , Austria.
Christian Doppler Laboratory for Innovative Tools for Biosimilar Characterization , University of Salzburg , Hellbrunner Straße 34 , 5020 Salzburg , Austria.
Anal Chem. 2018 May 1;90(9):5728-5736. doi: 10.1021/acs.analchem.8b00019. Epub 2018 Apr 18.
Hybrid mass spectrometry (MS) is an emerging technique for characterizing glycoproteins, which typically display pronounced microheterogeneity. Since hybrid MS combines information from different experimental levels, it crucially depends on computational methods. Here, we describe a novel software tool, MoFi, which integrates hybrid MS data to assign glycans and other post-translational modifications (PTMs) in deconvoluted mass spectra of intact proteins. Its two-stage search algorithm first assigns monosaccharide/PTM compositions to each peak and then compiles a hierarchical list of glycan combinations compatible with these compositions. Importantly, the program only includes those combinations which are supported by a glycan library as derived from glycopeptide or released glycan analysis. By applying MoFi to mass spectra of rituximab, ado-trastuzumab emtansine, and recombinant human erythropoietin, we demonstrate how integration of bottom-up data may be used to refine information collected at the intact protein level. Accordingly, our software reveals that a single mass frequently can be explained by a considerable number of glycoforms. Yet, it simultaneously ranks proteoforms according to their probability, based on a score which is calculated from relative glycan abundances. Notably, glycoforms that comprise identical glycans may nevertheless differ in score if those glycans occupy different sites. Hence, MoFi exposes different layers of complexity that are present in the annotation of a glycoprotein mass spectrum.
混合质谱（MS）是一种新兴的糖蛋白分析技术，糖蛋白通常表现出明显的微观异质性。由于混合 MS 结合了来自不同实验水平的信息，因此它严重依赖于计算方法。在这里，我们描述了一种新的软件工具 MoFi，它可以将混合 MS 数据集成到完整蛋白质去卷积质谱中，以分配聚糖和其他翻译后修饰（PTM）。它的两阶段搜索算法首先将单糖/PTM 组成分配给每个峰，然后编译与这些组成相容的聚糖组合的层次列表。重要的是，该程序仅包括那些由糖肽或释放聚糖分析得出的聚糖文库支持的组合。通过将 MoFi 应用于利妥昔单抗、ado-曲妥珠单抗emtansine 和重组人促红细胞生成素的质谱，我们展示了如何整合自上而下的数据以细化在完整蛋白质水平上收集的信息。因此，我们的软件表明，单个质量通常可以用相当数量的糖型来解释。然而，它同时根据得分对蛋白质形式进行排名，得分是根据相对聚糖丰度计算得出的。值得注意的是，如果这些糖占据不同的位置，即使包含相同聚糖的糖型在得分上也可能不同。因此，MoFi 揭示了糖蛋白质谱注释中存在的不同复杂层次。
