Structural and Computational Biology Structural characterization of macromolecular assemblies
Inter- and intra-cellular communication is fundamental for the survival of multi-cellular organisms, and defects in this process are often key features of many diseases. At the molecular level, the basis for information transfer is the formation of complex networks of interacting components.
Protein domains are discrete structural and functional units that have been shuffled and reorganized throughout the course of evolution of multicellular organisms. Indeed, many domains function in signalling cascades through the formation of complexes. Given their relatively small size and compact fold, protein domains are highly appropriate for structural studies, in particular in solution by Nuclear Magnetic Resonance (NMR). Undeniably, the last decade has witnessed a dramatic increase in the popularity and success of NMR for structural studies of proteins in solution. Protein expression using 15N and 13C labelling is now technically and economically affordable for most labs and has opened up a new avenue in NMR since it allows the use of triple resonance experiments in a routine manner, thereby increasing the quality of NMR data and thus that of the calculated structures. Furthermore, recent advances in NMR techniques and instrumentation allow the study of large complexes, thereby extending the molecular weight limit of systems that can be studied up to 100 kDa.
- Structure determination of protein domains involved in signal transduction pathways by NMR
- Protein folding and stability using NMR
- Software development to facilitate the assignment of NMR data
We seek to determine the rules that govern specificity and selectivity of protein domains. These domains are normally quite variable in sequence. Indeed, most of the best conserved residues are devoted to defining the elements of secondary structure and the three-dimensional fold. However, small variations at loops or even at elements of the secondary structure are responsible for ligand recognition specificity. To obtain a complete picture of a domain family, efforts must focus not only on a single structure but a small collection of representative sequences. For this purpose, we use sequence alignments and phylogenetic tree reconstructions to select sets of sequences that may fully represent the properties of a given type of domain. We then determine their structures in solution, in both free and bound forms.
The main bottle-neck of the NMR process continues to lie in the resonance assignment step, which is often performed manually. Given that an efficient use of NMR in structural or functional genomics relies on the availability of the automation of data acquisition and analysis, we are currently developing a software package that allows protein resonance assignments with the minimum of human intervention. This software will combine peak-picking and assignment processes, thereby improving the performance of both tasks when compared with other programmes that deal with these processes as two separate entities. We also plan to assign not only protein backbone and carbon side-chains but also aliphatic side-chain protons in an automated manner.
This group receives financial support from the following sources:
- Ministerio de Educación y Ciencia (Spanish Ministry of Science & Education)
- Ministerio de Economía y Competitividad (MINECO)
- European Commission (EC), Fondo Europeo de Desarrollo Regional (FEDER), "Una manera de hacer Europa"