[en] Complete text of publication follows: Characterization of heavy metals interactions with proteins is fundamental for understanding the molecular factors and mechanisms governing ions toxicity and speciation in cells. This line of research will also help in developing new molecules able to selectively and efficiently bind toxic metal ions, which could find application for bio-detection or bioremediation purposes. We have used the regulatory calcium-binding protein Calmodulin (CaM) from A. thaliana as a structural model and, starting from it, we have designed various mutants by site-directed mutagenesis. We have analysed thermodynamics of uranyl ion binding to both sites I and II of CaM N-terminal domain and we have identified structural factors governing this interaction. Selectivity for uranyl ion has been tested by studying reactions of the investigated peptides with Ca"2"+, in the same conditions used for UO_2"2"+. Spectro-fluorimetric titrations and FTIR analysis have shown that the affinity for uranyl increases by phosphorylation of a threonine in site I, especially approaching the physiological pH, where the phospho-threonine side chain is deprotonated. Based on structural models obtained by Molecular Dynamics, we tested the effect of a two residues deletion on site I properties. We obtained an almost two orders of magnitude increase in affinity for uranyl, with a sub-nanomolar dissociation constant for the uranyl complex with the non phosphorylated peptide, and an improved uranyl/calcium selectivity. Allosteric effects depending on Ca"2"+ and UO_2"2"+ binding have been investigated by comparing thermodynamic parameters obtained for mutants having both sites I and II able to chelate metal ions with those of mutants consisting of just one active site