We introduce Replica-Exchange-with-Tunneling (RET) simulations as a tool for studies of the conversion between polymorphic amyloids. For the 11-residue amyloid-forming cylindrin peptide we show that this technique allows for a more efficient sampling of the formation and interconversion between fibril-like and barrel-like assemblies. We describe a protocol for optimized analysis of RET simulations that allows us to propose a mechanism for formation and interconversion between various cylindrin assemblies. Especially, we show that an interchain salt bridge between residues K3 and D7 is crucial for formation of the barrel structure.

Recent experiments suggest that an amino acid sequence encodes not only the native fold of a protein but also other forms that are essential for its function or are important during folding or association. These various forms populate a multifunnel folding and association landscape where mutations, changes in environment, or interaction with other molecules switch between the encoded folds. We introduce replica exchange with tunneling as a way to efficiently simulate switching between distinct folds of proteins and protein aggregates. The correctness and efficiency of our approach are demonstrated in a series of simulations covering a wide range of proteins, from a small 11-residue large designed peptide to two 56-residue large mutants of the A and B domains of protein G.

Amyloid-β peptides form polymorphous amyloid fibrils are correlated with the pathogenesis of Alzheimer’s disease. Recently, a new ssNMR high-resolution structure has been reported for wild-type Aβ1–42 fibrils that is characterized by a strand-turn-strand-turn-strand motif instead of the U-shape form seen in previously known wild-type Aβ-fibril structures. Analyzing molecular dynamics simulations we comment on the relative weight of the new fibril structure and present evidence that its stability depends on hydrophobic contacts involving the C-terminal residues I41 and A42, but not on the salt bridge K28–A42. We further argue that Aβ1–42 peptides with this structure may assemble in fibrils with a 2-fold packing symmetry and discuss two possible arrangements.

Using molecular dynamics, we study the self-assembly of phenylalanine with charged end-groups at various temperatures and concentrations. As in the case of diphenylalanine, we observe the formation of nanotubes; however, phenylalanine aggregates in layers of four, not six, molecules. The observed aggregates are consistent with recent experimental measurements of fibrils obtained from mice with phenylketonuria. We investigate the stability and the mechanism by which these tubular structures form and discuss potential toxicity mechanisms.

Single amino acid mutations in amyloid-beta (Aβ) peptides can lead to early onset and increased severity of Alzheimer’s disease. An example is the Osaka mutation (Aβ1–40E22D), which is more toxic than wild-type Aβ1–40. This mutant quickly forms early stage fibrils, one of the hallmarks of the disease, and these fibrils can even seed fibrilization of wild-type monomers. Using molecular dynamic simulations, we show that because of formation of various intra- and intermolecular salt bridges the Osaka mutant fibrils are more stable than wild-type fibrils. The mutant fibril also has a wider water channel with increased water flow than the wild type. These two observations can explain the higher toxicity and aggregation rate of the Osaka mutant over the wild type.

We compare the efficiency of multicanonical and replica exchange molecular dynamics for the sampling of folding/unfolding events in simulations of proteins with end-to-end β-sheet. In Go-model simulations of the 75-residue MNK6, we observe improvement factors of 30 in the number of folding/unfolding events of multicanonical molecular dynamics over replica exchange molecular dynamics. As an application, we use this enhanced sampling to study the folding landscape of the 36-residue DS119 with an all-atom physical force field and implicit solvent. Here, we find that the rate-limiting step is the formation of the central helix that then provides a scaffold for the parallel β-sheet formed by the two chain ends.

Recent epidemiological data have shown that patients suffering from Type 2 Diabetes Mellitus have an increased risk to develop Alzheimer’s disease and vice versa. A possible explanation is the cross-sequence interaction between Aβ and amylin. Because the resulting amyloid oligomers are difficult to probe in experiments, we investigate stability and conformational changes of Aβ–amylin heteroassemblies through molecular dynamics simulations. We find that Aβ is a good template for the growth of amylin and vice versa. We see water molecules permeate the β-strand–turn−β-strand motif pore of the oligomers, supporting a commonly accepted mechanism for toxicity of β-rich amyloid oligomers. Aiming for a better understanding of the physical mechanisms of cross-seeding and cell toxicity of amylin and Aβ aggregates, our simulations also allow us to identify targets for the rational design of inhibitors against toxic fibril-like oligomers of Aβ and amylin oligomers.Keywords: Amyloid oligomer; crosses seeding; molecular dynamics; water channel

Seeding a protein solution with preformed fibrils can dramatically enhance the growth rate of amyloids. As the seeds do not need to be of the same protein, seeding may account for the observed correlations between amyloid diseases. In an effort to understand better the molecular mechanisms behind cross seeding we have studied in silico the effect of mutations on the seeding of amylin fibrils. Our investigations of the structural stability of decamers of wild type amylin peptides, of Y37L mutants, and of heteroassemblies of wild-type and mutant amylin molecules show that the experimentally observed efficient cross seeding can be explained based on similarity in fibril structure of components. We find that amyloids with similar side chains packing at the β-sheet interface are structurally compatible, acting as a good template for the congruent incorporation of homologues peptides. In the Y37L mutants, lack of tyrosine-specific interactions causes significant higher flexibility of the C terminal than observed in the wild-type fibril. This effects elongation of the mutant fibril leading to the longer lag times during aggregation that are observed in experiments. Our study gives guidelines for the design of ligands that could stabilize amylin fibrils.

Structure-based models are an efficient tool for folding studies of proteins since by construction their energy landscape is only minimally frustrated. However, their intrinsic drawback is a lack of structural flexibility as usually only one target structure is employed to construct the potentials. Hence, a Go-model may not capture differences in mutation-induced protein dynamics, if—as in the case of the disease-related A629P mutant of the Menkes protein ATP7A—the structural differences between mutant and wild type are small. In this work, we introduced three implementations of Go-models that take into account the flexibility of proteins in the NMR ensemble. Comparing the wild type and the mutant A629P of the 75-residue large sixth domain Menkes protein, we find that these new Go-potentials lead to broader distributions than Go-models relying on a single member of the NMR ensemble. This allows us to detect the transient unfolding of a loosely formed β1β4 sheet in the mutant protein. Our results are consistent with previous simulations using a physical force field and an explicit solvent and suggest a mechanism by which these mutations cause Menkes disease. In addition, the improved Go-models suggest differences in the folding pathway between the wild type and mutant, an observation that was not accessible to simulations of this 75-residue protein with a physical all-atom force field and explicit solvent.

We study wild type and mutants of the A and B domain of protein G using all-atom Go-models. Our data substantiate the usefulness of such simulation for probing the folding mechanism of proteins and demonstrate that multifunnel versions of such models also allow probing of more complicated funnel landscapes. In our case, such models reproduce the experimentally observed distributions of the GA98 and GB98 mutants which differ only by one residue but fold into different structures. They also reveal details on the folding mechanism in these two proteins.