Modeling the Cu + Binding in the 1-16 Region of the Amyloid- Peptide Involved in Alzheimer’s Disease Sara Furlan, †,‡ Christelle Hureau, †,‡ Peter Faller, †,‡ and Giovanni La Penna* ,§ LCC (Laboratoire de Chimie de Coordination), CNRS, 205 route de Narbonne, F-31077 Toulouse, France; UPS, INPT, LCC, UniVersite ´ de Toulouse, F-31077 Toulouse, France; and ICCOM (Institute for Chemistry of Organo-metallic Compounds), CNR (National Research Council), Via Madonna del Piano 10, I-50019 Sesto Fiorentino, Firenze, Italy ReceiVed: April 1, 2010; ReVised Manuscript ReceiVed: August 26, 2010 The coordination of copper to the amyloid- (1-16) (A) peptide has been investigated because of its relevance for understanding Cu redox activity when the ion is embedded in peptides involved in neurodegenerative diseases. In this work, several reasonable models of Cu + coordination were built on the basis of experimental information and investigated by first-principles molecular dynamics simulations in the Car-Parrinello scheme. The propensity of a linear Nδ (His)-Cu-Nδ (His) coordination for Cu + is shown by all the models investigated here, with distortions due to weak interactions with the carbonyl O of His 6 and His 13 and with the amide N of His 14. Though the His 6-Cu-His 14 linear coordination is favored in truncated models, the His 13- Cu-His 14 linear coordination is favored by interactions present in the complete solvated and in vacuo models of Cu-A (1-16). These interactions include steric hindrance for the expulsion of His 13, hydrogen bonds between Asp and His side chains and a network of electrostatic interactions stabilizing two separated 1-10 and 11-16 peptide regions. The role of linear His 13-Cu-His 14 coordination in stabilizing Cu(I) and in increasing the Cu(II)/Cu(I) reorganization energy can be therefore modulated by boundary conditions acting on the A (1-16) ligand. Introduction Complexes of amyloid- (A) peptides with metal ions are linked to the development of Alzheimer’s disease. 1,2 Among the metal ions, copper is particularly investigated because it has been linked to aggregation of A peptides and to generation of reactive oxygen species (ROS), two events intimately related to the development of the disease. 3-10 The nature of the Cu(I) and Cu(II) environments affects the Cu(I)-A/Cu(II)-A redox properties and hence the A-mediated ROS production by Cu ions. Cu(II) coordination has recently been discussed in details (ref 4 and references therein) and deeper insights have been very recently obtained by NMR 11 and pulsed-EPR studies. 12-15 The molecular mechanism of the different metal-binding impact on A aggregation is not understood. Metal ions bind to the N-terminal portion (region 1-16), which is unstructured in the amyloid fibrils. 16 Therefore, the metal ion is likely to affect intermediate states (oligomers which are difficult to study) rather than aggregated states. The difference of Cu(I) from the better known Cu(II) and Zn(II) is that the divalent metals are bound to more ligands (4 or more) and include, in their most stable form at pH 7.4, His 6 and the N-terminal Asp 1, which lead to a higher structuration of the peptide around the metal ion. Zn(II) induces rapid A aggregation, which has been assigned to binding of Zn(II) to ligands of two peptides. 17,18 This latter arrangement seems less likely for the stable Cu(I)- A complex and, hence, Cu(I) might have a less important promoting effect than Zn(II) and a possible indirect protective role. Another factor is the overall charge, as aggregation is higher when the peptide is neutral. A (1-42) is negatively charged (-3) around physiological pH, and binding of positively charged metal ions will simply lower its charge. However, protons released upon metal binding have to be considered as well. Even on the basis of this simple argument, Cu(I) is expected to have a less promoting effect compared to Zn(II) and Cu(II). Since Cu, in either redox state, binds to the N-terminal portion of A, the truncated peptide A (1-16) has been established as a model for metal binding to the entire A in the soluble, nonaggregated state. Indeed, no difference in spectroscopic signatures or in binding affinities has been observed between the truncated Cu-A (1-16) and the native Cu-A (1-40/42) peptides. 17,19-21 Two distinct Cu(II)-A complexes, which differ by the protonation state of the peptide, coexist at physiological pH. 22 These two complexes have mainly been evidenced by EPR because they are more easily distinguishable by their magnetic parameters. 23,24 In both forms the Cu(II) ion lays on a distorted square-planar geometry and in the form predominant at physi- ological pH it is surrounded by the -NH 2 terminus, a carbonyl function and two imidazole rings from the His residues, while in the minor form, one of the two imidazole rings is replaced by the Asp1-Ala2 deprotonated amide function. Cu(I) environ- ment has more recently been revealed by XAS spectroscopy 19,25,26 associated to DFT calculations. 19,27 These showed that the His- His motif present in A provides to Cu(I) the framework for a linear Nδ-Cu-Nδ geometry, where the two Nδ atoms belong to His 13 and His 14 imidazole side chains (Im, hereafter), respectively. The DFT calculations showed that this linear geometry is also open to other ligands to forms 3-coordinated T-shaped geometries. Coordination number 2, with Im-Cu-Im adopting a linear geometry, is indeed found for many nonpeptidic Cu(I) * Corresponding author. E-mail: glapenna@iccom.cnr.it. † CNRS-LCC. ‡ UPS-LCC. § CNR-ICCOM. J. Phys. Chem. B 2010, 114, 15119–15133 15119 10.1021/jp102928h  2010 American Chemical Society Published on Web 11/01/2010