Search

The physical adsorption (physisorption) of proteins to surfaces is an important but incompletely understood factor in many biological processes and is of increasing significance in bionanotechnology as well. Avidin is an important protein because of strong avidin-biotin binding, which has been exploited in numerous applications. We have undertaken thorough experimentation on the physisorption of avidin, to chemically different flat surfaces of Si and graphite and also to the curved version of the latter, on multiwalled carbon nanotubes (MWNTs) of different diameters. The difference in the behavior of avidin on Si versus graphite is drastic; on Si, avidin deposits as single globular tetrameric units and maintains functionality, whereas on graphite, it forms irregular networks of two-layer thick filaments, where the first layer has lost its biological activity. On MWNTs, avidin also deposits as one-dimensional formations, or stripes, but these appear to order in a perpendicular arrangement to the MWNT axis. A better understanding of protein-surface interactions is essential for the development of robust and reliable methods for biofunctionalization of materials. This work also provides insights into the importance of the nanoscale surface architecture.

It is well known that peptide and protein fibrillation is strongly affected by the solution conditions, but a fundamental understanding of how amyloid fibril nucleation depends on solution pH, salt concentration, and solvent is absent. Here, we use expressions from Debye-Hückel theory to describe the interactions between charged amino acids in combination with our recently developed nonstandard nucleation theory to predict the concentration dependence of the fibril nucleation rate under different solvent conditions. The general rule that emerges from these considerations is that changes in solution pH, salt concentration, and solvent that increase the bonding energy between the fibril building blocks decrease the fibril solubility and promote fibril nucleation, in line with experimental observations. The simple analytical relations among the nucleation rate, fibril solubility, and binding energies provide a tool to controlling and understanding amyloid fibril formation by changing the solution conditions.

It is well established that amyloid fibril solubility is protein specific, but how solubility depends on the interactions between the fibril building blocks is not clear. Here we use a simple protein model and perform Monte Carlo simulations to directly measure the solubility of amyloid fibrils as a function of the interaction between the fibril building blocks. Our simulations confirms that the fibril solubility depends on the fibril thickness and that the relationship between the interactions and the solubility can be described by a simple analytical formula. The results presented in this study reveal general rules how side-chain-side-chain interactions, backbone hydrogen bonding, and temperature affect amyloid fibril solubility, which might prove to be a powerful tool to design protein fibrils with desired solubility and aggregation properties in general.

We consider the nucleation of amyloid fibrils when the process occurs by direct polymerization of fully extended peptides (i.e., β-strands) into fibrils composed of successively layered β-sheets with alternating weak and strong hydrophobic surfaces. We extend our recently developed nucleation model (Kashchiev, D.; Cabriolu, R.; Auer, S. J. Am. Chem. Soc. 2013, 135, 1531-1539) to derive general expressions for the work to form such fibrils, the fibril solubility, the nucleation work, the equilibrium concentration of nuclei, and the fibril nucleation rate as explicit functions of the supersaturation of the protein solution. Analysis of these expressions illustrates the effect of increased asymmetry between the weak and strong hydrophobic β-sheet surfaces on the thermodynamics and kinetics of the polymerization process. In particular, the application of our theoretical framework to a simple model peptide system shows that lowering the hydrophobicity of one β-sheet surface can hamper protein fibrillation because the threshold concentration below which the fibril nucleation is practically arrested, and above which the process occurs vigorously--because then each monomer in the solution acts as a fibril nucleus--is shifted to higher concentrations. This effect is entirely due to the effect of asymmetry of the two hydrophobic β-sheet surfaces on the fibril solubility. In addition, with increasing asymmetry, the nucleation rate of one fibril polymorph becomes increasingly dominant, illustrating that there is a morphological selection between the two possible polymorphs.

Fibrils of amyloid proteins are currently of great interest because of their involvement in various amyloid-related diseases and nanotechnological products. In a recent kinetic Monte Carlo simulation study (Cabriolu, R.; Kashchiev, D.; Auer, S. J. Chem. Phys.2012, 137, 204903), we found that our simulation data for the rate of amyloid fibril nucleation occurring by direct polymerization of monomeric protein could not be described adequately by nucleation theory. It turned out that the process occurred in a peculiar way, thus confounding the nucleation paradigm and demanding a new theoretical treatment. In the present study, we reconsider the theoretical approach to nucleation of amyloid fibrils and derive new expressions for the stationary rate of the process. As these expressions provide a remarkably good description of the simulation data, by using them we propose a theoretical dependence of the amyloid-β(40) fibril nucleation rate on the concentration of monomeric protein in the solution. This dependence reveals the existence of a threshold concentration below which the fibril nucleation in small enough solution volumes is practically arrested, and above which the process occurs vigorously, because then each monomeric protein in the solution acts as fibril nucleus. The presented expressions for the threshold concentration and for the dependence of the fibril nucleation rate on the concentration of monomeric protein can be a valuable guide in designing new therapeutic and/or technological strategies for prevention or stimulation of amyloid fibril formation.

Characterization of the folding transition in polypeptides and assessing the thermodynamic stability of their structured folds are of primary importance for approaching the problem of protein folding. We use molecular dynamics simulations for a coarse grained polypeptide model in order to (1) obtain the equilibrium conformation diagram of homopolypeptides in a broad range of the chain lengths, N = 10, ..., 100, and temperatures, T (in a multicanonical ensemble), and (2) determine free energy profiles (FEPs) projected onto an optimal, so-called "natural", reaction coordinate that preserves the height of barriers and the diffusion coefficients on the underlying free energy hyper-surface. We then address the following fundamental questions. (i) How well does a kinetically determined free energy landscape of a single chain represent the polypeptide equilibrium (ensemble) behavior? In particular, under which conditions might the correspondence be lost, and what are the possible implications for the folding processes? (ii) How does the free energy landscape depend on the chain length (homopolypeptides) and the monomer interaction sequence (heteropolypeptides)? Our data reveal that at low T values equilibrium structures adopted by relatively short homopolypeptides (N < 60) are dominated by α-helical folds which correspond to the primary and secondary minima of the FEP. In contrast, longer homopolypeptides (N > 70), upon quasi-equilibrium cooling, fold preferentially in β-bundles with small helical portions, while the FEPs exhibit no distinct global minima. Moreover, subject to the choice of the initial configuration, at sufficiently low T, essentially metastable structures can be found and prevail far from the true thermodynamic equilibrium. We also show that, by sequence-enabling the polypeptide model, it is possible to restrict the chain to a very specific part of the configuration space, which results in substantial simplification and smoothing of the free energy landscape as compared to the case of the corresponding homopolypeptide.

In this study, we address the questions of how important is the kinetics in protein aggregation, and what are the intrinsic properties of proteins that cause this behavior. On the basis of our recent quantitative calculation of the equilibrium phase diagram of natively folded α-helical and β-sheet forming peptides, we perform molecular dynamics simulations to demonstrate how the aggregation mechanism and end product depend on the temperature, concentration, and starting point in the phase diagram. The results obtained show that there are severe differences between the thermodynamically predicted and the kinetically obtained aggregate structures. The observed differences help to rationalize the suggestion that monomeric proteins in their native functional structure can be metastable with respect to the amyloid state, and that the native fold is a special property that protects them from aggregation.