Molza A-E, Férey N, Czjzek M, Le Rumeur E, Hubert J-F, Tek A, Laurent B, Baaden M, Delalande O
Institut de Génétique et Développement de Rennes, CNRS, UMR6290, Univ. Rennes 1, Campus Santé, 2 av du Pr Léon Bernard, 35043 Rennes Cedex, France.
Faraday Discuss. 2014;169:45-62. doi: 10.1039/c3fd00134b. Epub 2014 Jun 20.
At present, our molecular knowledge of dystrophin, the protein encoded by the DMD gene and mutated in myopathy patients, remains limited. To get around the absence of its atomic structure, we have developed an innovative interactive docking method based on the BioSpring software in combination with Small-angle X-ray Scattering (SAXS) data. BioSpring allows interactive handling of biological macromolecules thanks to an augmented Elastic Network Model (aENM) that combines the spring network with non-bonded terms between atoms or pseudo-atoms. This approach can be used for building molecular assemblies even on a desktop or a laptop computer thanks to code optimizations including parallel computing and GPU programming. By combining atomistic and coarse-grained models, the approach significantly simplifies the set-up of multi-scale scenarios. BioSpring is remarkably efficient for the preparation of numeric simulations or for the design of biomolecular models integrating qualitative experimental data restraints. The combination of this program and SAXS allowed us to propose the first high-resolution models of the filamentous central domain of dystrophin, covering repeats 11 to 17. Low-resolution interactive docking experiments driven by a potential grid enabled us to propose how dystrophin may associate with F-actin and nNOS. This information provides an insight into medically relevant discoveries to come.
目前，我们对肌营养不良蛋白（由DMD基因编码且在肌病患者中发生突变的蛋白质）的分子了解仍然有限。为了克服其原子结构缺失的问题，我们基于BioSpring软件并结合小角X射线散射（SAXS）数据，开发了一种创新的交互式对接方法。BioSpring借助增强弹性网络模型（aENM）实现对生物大分子的交互式处理，该模型将弹簧网络与原子或伪原子之间的非键项相结合。由于包括并行计算和GPU编程在内的代码优化，这种方法甚至可以在台式机或笔记本电脑上用于构建分子组装体。通过结合原子模型和粗粒度模型，该方法显著简化了多尺度场景的设置。BioSpring在准备数值模拟或设计整合定性实验数据约束的生物分子模型方面非常高效。该程序与SAXS的结合使我们能够提出肌营养不良蛋白丝状中央结构域的首个高分辨率模型，涵盖重复序列11至17。由势场网格驱动的低分辨率交互式对接实验使我们能够提出肌营养不良蛋白可能与F-肌动蛋白和神经元型一氧化氮合酶（nNOS）结合的方式。这些信息为未来医学相关的发现提供了见解。
