Hyper-dimensional NMR spectroscopy for automated protein structure determination

点击翻译按钮获取中文摘要

Adequate spectral resolution is essential for the study of biomacromolecules by NMR, especially for proteins that are either large, unfolded, or membrane-bound. Using conventional methods, the analysis of spectra is greatly complicated by signal overlap. To address this problem we capitalize on our recent breakthroughs in NMR and computational methods: real-time (Jaravine and Orekhov 2006) fast spectroscopy of any dimensionality (Jaravine et al. 2006), and fully automated spectra analysis (Lopez-Mendez and Guntert 2006). Our R-MDD algorithm (Jaravine et al. 2006) can deliver highest spectral resolution affordable by spin magnetization lifetimes in the indirect dimensions. Recently, we have shown how this increased resolution facilitates and improves the accuracy of the assignment for a set of backbone experiments (Jaravine et al. 2008). Here we propose the extension of Hyper-Dimensional NMR methodology (Jaravine et al. 2008) to new applications in biomolecular NMR, in particular, for automated side-chain assignment and structure calculation. The possibility to enhance the resolution in NOESY experiments for the collection of conformational restraints (Luan et al. 2005) is especially promising in combination with fully automated structure calculation with the FLYA algorithm (Lopez-Mendez and Guntert 2006; Scott et al. 2006). A second direction of this project is the use of Hyper-Dimensional NMR methodology for the joint interpretation of information from the most sensitive experiments for resolution enhancement in the corresponding dimensions of less sensitive experiments. This research project will show how to achieve optimizations with respect to measurement time, resolution, sensitivity, data completeness, and accuracy. Applications of the fully automated approaches to the 23 kDa protein RcsD_HPt shall demonstrate the power of the method for answering questions in structural biology.