Insights on peptide topology in the computational design of protein ligands: the example of lysozyme binding peptides.

Herein, we compared the ability of linear and cyclic peptides generated in silico to target different protein sites: internal pockets and solvent-exposed sites. We selected human lysozyme (HuL) as a model target protein combined with the computational evolution of linear and cyclic peptides. The sequence evolution of these peptides was based on the PARCE algorithm. The generated peptides were screened based on their aqueous solubility and HuL binding affinity. The latter was evaluated by means of scoring functions and atomistic molecular dynamics (MD) trajectories in water, which allowed prediction of the structural features of the protein-peptide complexes. The computational results demonstrated that cyclic peptides constitute the optimal choice for solvent exposed sites, while both linear and cyclic peptides are capable of targeting the HuL pocket effectively. The most promising binders found in silico were investigated experimentally by surface plasmon resonance (SPR), nuclear magnetic resonance (NMR), and electrospray ionization mass spectrometry (ESI-MS) techniques. All tested peptides displayed dissociation constants in the micromolar range, as assessed by SPR; however, both NMR and ESI-MS suggested multiple binding modes, at least for the pocket binding peptides. A detailed NMR analysis confirmed that both linear and cyclic pocket peptides correctly target the binding site they were designed for.

In silico design of peptides as potential ligands to resistin.

Resistin is a hormone of biological interest due to its connection with several diseases of worldwide concern. This work aims to design a series of cyclic peptides as "lead compounds" to identify potential ligands to resistin. To this end, we propose an approach based on a peptide design algorithm plus a two-stage selection which accounts for selectivity, one of the most forgotten steps in the design of ligands. Following this approach, we have been able to identify several peptides as strong candidates for the design of elements of bio-recognition. Those peptides present low scoring binding energy to albumin, good water solubility, stability in water at 300 K, and high scoring binding energy to resistin. Among those peptides, two were chosen, to perform a more rigorous calculation of binding free energy based on the Alchemical Absolute Binding Free Energy method. We were able to establish a methodological route for the development of strong candidates for the design of ligands to resistin. Graphical Abstract Combined MD + MC + AABFE approach to design and screening of high-affinity binders to resistin.

X-ray accelerated photo-oxidation of As(III) in solution.

We performed near edge X-ray absorption spectroscopy (XANES) measurements on the arsenic K-edge of As(III) in solution under acidic and basic conditions, after exposure of the solutions to air. Spectra were recorded for increasing exposure times to the X-rays used to perform absorption spectroscopy measurements. We did not find changes for the solution under acidic conditions, whereas we observed significant changes in the case of solution under alkaline conditions. To interpret these changes, we compared the obtained spectra with XANES spectra of As(III) and As(V) solutions under alkaline conditions, not exposed to air, and used as standards. Principal component fits using these standards indicate an accelerated conversion of As(III) to As(V) due to the exposure to X-rays.

Interpretation of X-ray absorption spectra of As(III) in solution using Monte Carlo simulations.

We performed X-ray absorption spectroscopy measurements on the arsenic K-edge of As(III) in solution under acidic conditions. Extended X-ray absorption fine structure (EXAFS) and X-ray near edge structure (XANES) spectra were compared with theoretical calculations which use local atomic structure configurations, either derived from density functional theory (DFT) energy minimization (EM) calculations or based on classical Monte Carlo (MC) simulations, for a As(OH)3 cluster surrounded by water molecules. The nearest arsenic-oxygen distances obtained from the fit of the XAFS spectra are consistent with the distances present in configurations derived from Monte Carlo simulations but not with those obtained from DFT-EM calculations. Calculations of XANES using either DFT-EM or the average configuration obtained from MC simulations do not reproduce the XANES spectra in the vicinity of the absorption edge. However, specific local atomic structural configurations of the As(OH)3 and water molecules, obtained from MC simulations, which show some ordering of water molecules up to 5 Å from the arsenic, reproduce qualitatively the experimental spectra. These results highlight the capability of XANES to yield information about hydration of ions in solution.

Crystallization time scales for polydisperse hard-sphere fluids.

We study the evolution of crystallization in dense mono- and polydisperse hard sphere fluids, initially quenched to an amorphous configuration. We use as signatures of crystallization both the decay of the reduced pressure Z and the increase in the local and global orientational order parameters Q[over ¯](6). For a given realization of the crystallization process these parameters show sudden changes, both large and small, separated by long periods of quiescence. However, averaging over a large number of realizations, a well-defined scenario for their evolution appears. We find an initial fast relaxation to a disordered state, followed by a period of slow variation, associated to the presence of nucleation events, followed by a fast change, composed of the growth of a few crystals with different orientations, and a final and slow coarsening in a domain-growth process. No clear scaling for this whole process was found. We also find that the transition to an stable glassy fluid is quite sharp as the polydispersity is increased, showing a probable first-order phase transition behavior. A well-defined boundary between crystallizing and permanently amorphous fluids should exist, at least for a region in packing fractions. We looked for segregation at large values of polydispersity, but no effects of this type were found.

Aqueous solvation of As(OH)3: a Monte Carlo study with flexible polarizable classical interaction potentials.

A theoretical study of the hydration of arsenious acid is presented. This study included ab initio calculations and Monte Carlo simulations. The model potentials used for the simulations were ab initio derived and they include polarizability, nonadditivity, and molecular relaxation. It is shown that with these refined potentials it is possible to reproduce the available experimental evidence and therefore permit the study of clusters, as well as of the hydration process in solution. From the study of stepwise hydration and the Monte Carlo simulation of the condensed phase it is concluded that As(OH)(3) presents a hydration scheme similar to an amphipathic molecule. This phenomenon is explained as due to the existence of both a positive electrostatic potential and a localized lone pair in the vicinity of As. These results are used to rationalize the known passage of As(OH)(3) through aqua-glyceroporines.