Residual Dipolar Couplings in de-novo Protein Fold Determination

Activity: Talks and presentationsPoster presentationOther

Description

Residual dipolar couplings are global long-range restraints and have been shown improve the accuracy and precision of solution structures determined by NMR. De-novo determination of protein backbone folds, though, is less straight forward and more difficult. Using a procedure termed molecular fragment replacement (MFR) it has been possible to solve protein structures with residual dipolar couplings (RDCs) and backbone chemical shifts. In a first step the query protein is broken into small overlapping peptide fragments (typically 7-10 residues). For each query range a representative database of protein structures is mined for peptide fragments with homologous backbone chemical shifts and residual dipolar couplings. From this collection of fragments consensus values for the backbone angles can be derived or those fragments can be directly assembled into a protein backbone structure by aligning their local alignment tensors and translating the pieces for best chain overlap keeping this orientation fixed. This simple implementation of the MFR assembly is quite adequate when complete sets of precise RDCs are available but still suffers from a number of limitations. The sequential chain building process is a purely geometric algorithm, which may result in non-physical structures and it is prone to errors, as accidental incorporation of unsuitable fragments can disrupt the assembly process. A newer implementation uses a ‘rigid body docking’ algorithm for chain extension. A new peptide fragment is placed at the end of the growing protein chain and its orientation with respect to the rest of the protein is refined against dipolar coupling by simulated annealing. Fragments that fail to converge to an orientation compatible with the rest of the protein are discarded. The accepted fragments for a particular residue range are averaged ‘on the fly’ with the previously built protein backbone and the process is continued until the end of the chain. NOE information that may be available at this stage (e.g. easily obtainable 1HN-1HN NOEs or well resolved Methyl-Methyl NOEs), can thus be easily incorporated into the fragment assembly and ad extra precision and accuracy to the model building.
Period2011
Event titleNicht angegeben
Event typeOther
LocationUnknownShow on map