Huang Wei, Ravikumar Krishnakumar M, Parisien Marc, Yang Sichun
Center for Proteomics and Department of Nutrition, Case Western Reserve University, Cleveland, OH, USA.
Alan Edwards Centre for Research on Pain, McGill University, Montreal, Quebec, Canada.
J Struct Biol. 2016 Dec;196(3):340-349. doi: 10.1016/j.jsb.2016.08.001. Epub 2016 Aug 2.
Structural determination of protein-protein complexes such as multidomain nuclear receptors has been challenging for high-resolution structural techniques. Here, we present a combined use of multiple biophysical methods, termed iSPOT, an integration of shape information from small-angle X-ray scattering (SAXS), protection factors probed by hydroxyl radical footprinting, and a large series of computationally docked conformations from rigid-body or molecular dynamics (MD) simulations. Specifically tested on two model systems, the power of iSPOT is demonstrated to accurately predict the structures of a large protein-protein complex (TGFβ-FKBP12) and a multidomain nuclear receptor homodimer (HNF-4α), based on the structures of individual components of the complexes. Although neither SAXS nor footprinting alone can yield an unambiguous picture for each complex, the combination of both, seamlessly integrated in iSPOT, narrows down the best-fit structures that are about 3.2Å and 4.2Å in RMSD from their corresponding crystal structures, respectively. Furthermore, this proof-of-principle study based on the data synthetically derived from available crystal structures shows that the iSPOT-using either rigid-body or MD-based flexible docking-is capable of overcoming the shortcomings of standalone computational methods, especially for HNF-4α. By taking advantage of the integration of SAXS-based shape information and footprinting-based protection/accessibility as well as computational docking, this iSPOT platform is set to be a powerful approach towards accurate integrated modeling of many challenging multiprotein complexes.
对于多结构域核受体等蛋白质-蛋白质复合物的结构测定,高分辨率结构技术一直颇具挑战。在此,我们展示了多种生物物理方法的联合应用,即iSPOT,它整合了小角X射线散射(SAXS)的形状信息、羟自由基足迹法探测的保护因子,以及来自刚体或分子动力学(MD)模拟的大量计算对接构象。在两个模型系统上进行了具体测试,结果表明iSPOT能够基于复合物各个组分的结构,准确预测大型蛋白质-蛋白质复合物(TGFβ-FKBP12)和多结构域核受体同二聚体(HNF-4α)的结构。尽管单独的SAXS或足迹法都无法为每个复合物提供明确的结构信息,但在iSPOT中无缝整合这两种方法后,能将最佳拟合结构范围缩小,其与相应晶体结构的均方根偏差(RMSD)分别约为3.2Å和4.2Å。此外,这项基于从现有晶体结构综合推导数据的原理验证研究表明,使用刚体或基于MD的柔性对接的iSPOT能够克服独立计算方法的缺点,尤其是对于HNF-4α。通过利用基于SAXS的形状信息与基于足迹法的保护/可及性以及计算对接的整合,这个iSPOT平台有望成为一种强大的方法,用于对许多具有挑战性的多蛋白复合物进行精确的整合建模。