Lucia Banci (Chemistry and CERM, University of Florence, Italy)
Abstract:The description and understanding of the interaction of metal ions with biological molecules requires the combination of several approaches involving bioinformatics, computation and experimental investigations. Here two facets of our research are highlighted which have heavily exploited, and may in the future benefit from, mathematical/computational advances. Knowledge of the genomes of a large number of organisms has allowed the analysis and correlation of different properties and features ranging from evolution to structure prediction, to mechanistic details of processes. In contrast, the metal binding abilities of the products of a gene sequence somehow are `hidden' and cannot be so readily extracted from sequence information. Therefore specific tools are needed in order to identify those sequences `able' to bind a specific metal ion and moreover `likely' to bind a specific metal in vivo. We have developed bioinformatic approaches for this, and screened the genomes of several organisms, from the most ancestral to humans and plants, with respect to ability to bind specific metal ions. From this analysis we have inferred how the number and types of proteins, binding a specific metal ion, have changed along evolution. Secondly, the interaction between metal ions and protein scaffolds can be experimentally characterized with atomic resolution through structural studies, in the crystal with X-ray diffraction and/or in solution with NMR spectroscopy. Mass spectrometry is also suited to determine affinity constants of metal ions for proteins, which range over several orders of magnitude. Working in solution, using NMR, is particularly well suited for the characterization of the cell biology of metals at atomic resolution. This system typically involves transfer of metal ions from a transporter to a final recipient metallo-protein. The thermodynamic and kinetic properties of the metal-protein interactions and of the metal transfer processes need also to be known for a complete description of these systems and processes. Here there may be rich opportunities with mathematical approaches having barely begun to be deployed. We have applied the combination of these various approaches particularly to the case of copper. These studies allowed us to draw a comprehensive picture of the distribution of copper proteins in the cell and of the processes where copper is involved. Computational modeling of these processes would greatly contribute to the understanding of the factors affecting them.