A recent shift towards proteomics has seen many structural genomics initiatives set up for high-throughput structure determination using traditional methods of x-ray crystallography and NMR. The next step in the proteomic revolution focuses on the interplay of multi-protein complexes and transient protein-protein interactions, which are involved in many cellular functions. Greater understanding of protein-protein interactions will inevitably lead to better comprehension of the regulation of cellular process, which has implications in biomedical sciences and biotechnology. Even though many high-resolution initiatives focus on proteins and protein complexes, their structure-determination success rates are still low. An emerging approach uses chemical cross-linking and mass spectrometry to derive a set of sparse distance constrains, which can be used for building models of proteins and to map out residues in protein interaction interface based on partial structural information. This technique allows low-resolution identification of protein structures and their interactions in cases where traditional structure determination techniques did not produce results. Chemical cross-linkers have been successfully used for many years in identifying interacting proteins. However, recent advances in mass spectrometry have allowed the identification of exact insertion points of low-abundance cross-links and hence has opened up a new perspective on the use of cross-linkers in combination with computational structure prediction. For protein interaction studies, the approach uses chemical cross-linking information with molecular docking, so that the cross-links are treated as explicit constraints in the calculations. This study focuses on a low-cost and rapid approach to structure prediction, where partial structural information and distance constraints can be used to obtain the relative orientation of interacting proteins and domains, specifically as a rescue strategy where traditional high resolution structure determination methods were unsuccessful. This hybrid biochemical/bioinformatics approach was applied in the determination of structure of the latexin:carboxypeptidase A complex, and succeeded in achieving 4 Å rmsd compared to the crystal structure determined subsequently (Mouradov et al., 2006). Application of the bioinformatics/biochemical approach to multi-domain proteins was carried out on murine acyl-CoA thioesterase 7 (Acot7). X-ray crystallography provided structures of the two separate domains of Acot7, however the full length protein did not crystalise. Combining chemical cross-linking, mass spectrometry, molecular docking and homology modeling we were able to reconstruct how the two domains are arranged in the full length protein (Forwood et al., 2007). Limitations of this technique caused by the enormous complexity of the cross-linking reaction mixtures were identified and emphasized by analysing a large (four protein) complex of DNA polymerase III, where only one inter-protein cross-link was identified. A rapid and cost-effective method for identification of cross-linked peptides using a commercially available cross-linker was developed as part of the overall aim of streamlining the hybrid biochemical/bioinformatics in order for it to become a generally applicable technique for rapid protein structure characterisation (King et al., 2008). Finally an in-house software package was developed for assignment of cross-linked peptides based on m/z values.
| Identifer | oai:union.ndltd.org:ADTP/286000 |
| Creators | Dmitri Mouradov |
| Source Sets | Australiasian Digital Theses Program |
| Detected Language | English |
Page generated in 0.0014 seconds