融合蛋白(FP)经常被用作使用X射线方法确定大分子结构的生物技术工具。这里,我们探索了不同蛋白质标签在各种FP中的使用,通过在部分分子置换(MR)中使用它们并使用ARP/wARP构建剩余的FP结构来获得初始相。通常,在结晶之前去除标签,然而,留下标签可能会促进晶体形成,以及通过将复合物的已知阶段扩展到未知阶段来确定结构。在这项研究中,蛋白质数据库被挖掘为最新的FPs列表,其中包含最常用的蛋白质标签,麦芽糖结合蛋白(MBP),绿色荧光蛋白(GFP),硫氧还蛋白(TRX),谷胱甘肽转移酶(GST)和小泛素样修饰蛋白(SUMO)。使用蛋白质标签的部分MR,其次是自动建模,在116FP的子集上进行了测试。分析了该方法的效率,并鉴定了影响融合蛋白大部分坐标构建的因素。使用MBP,GFP,和SUMO作为相发生器,可以在116例测试的36例中构建至少75%的感兴趣的蛋白质。我们的结果表明,标签选择有显著的影响;标签具有更大的结构稳定性,比如GFP,提高成功率。进一步的统计分析确定了该分辨率,威尔逊B因子,溶剂百分比,完整性,多重性,FP中的蛋白质标签百分比(考虑氨基酸),使用我们的方法,链接器长度起着关键作用。在缺乏结构同源的情况下,这种方法值得包含在蛋白质晶体学的工具包中。
Fusion proteins (FPs) are frequently utilized as a biotechnological tool in the determination of macromolecular structures using X-ray methods. Here, we explore the use of different protein tags in various FP, to obtain initial phases by using them in a partial molecular replacement (MR) and constructing the remaining FP structure with ARP/wARP. Usually, the tag is removed prior to crystallization, however leaving the tag on may facilitate crystal formation, and structural determination by expanding phases from known to unknown segments of the complex. In this study, the Protein Data Bank was mined for an up-to-date list of FPs with the most used protein tags, Maltose Binding Protein (MBP), Green Fluorescent Protein (GFP), Thioredoxin (TRX), Glutathione transferase (GST) and the Small Ubiquitin-like Modifier Protein (SUMO). Partial MR using the protein tag, followed by automatic model building, was tested on a subset of 116 FP. The efficiency of this method was analyzed and factors that influence the coordinate construction of a substantial portions of the fused protein were identified. Using MBP, GFP, and SUMO as phase generators it was possible to build at least 75 % of the protein of interest in 36 of the 116 cases tested. Our results reveal that tag selection has a significant impact; tags with greater structural stability, such as GFP, increase the success rate. Further statistical analysis identifies that resolution, Wilson B factor, solvent percentage, completeness, multiplicity, protein tag percentage in the FP (considering amino acids), and the linker length play pivotal roles using our approach. In cases where a structural homologous is absent, this method merits inclusion in the toolkit of protein crystallographers.