Comparison of force fields for Alzheimer's A : A case study for intrinsically disordered proteins

Intrinsically disordered proteins are essential for biological processes such as cell signalling, but are also associated to devastating diseases including Alzheimer's disease, Parkinson's disease or type II diabetes. Because of their lack of a stable three‐dimensional structure, molecular dynamics simulations are often used to obtain atomistic details that cannot be observed experimentally. The applicability of molecular dynamics simulations depends on the accuracy of the force field chosen to represent the underlying free energy surface of the system. Here, we use replica exchange molecular dynamics simulations to test five modern force fields, OPLS, AMBER99SB, AMBER99SB*ILDN, AMBER99SBILDN‐NMR and CHARMM22*, in their ability to model Aβ₄₂, an intrinsically disordered peptide associated with Alzheimer's disease, and compare our results to nuclear magnetic resonance (NMR) experimental data. We observe that all force fields except AMBER99SBILDN‐NMR successfully reproduce local NMR observables, with CHARMM22* being slightly better than the other force fields.     

Microsecond simulations of mdm2 and its complex with p53 yield insight into force field accuracy and conformational dynamics

The p53‐MDM2 complex is both a major target for cancer drug development and a valuable model system for computational predictions of protein‐ligand binding. To investigate the accuracy of molecular simulations of MDM2 and its complex with p53, we performed a number of long (200 ns to 1 µs) explicit‐solvent simulations using a range of force fields. We systematically compared nine popular force fields (AMBER ff03, ff12sb, ff14sb, ff99sb, ff99sb‐ildn, ff99sb‐ildn‐nmr, ff99sb‐ildn‐phi, CHARMM22*, and CHARMM36) against experimental chemical shift data, and found similarly accurate results, with microsecond simulations achieving better agreement compared to 200‐ns trajectories. Although the experimentally determined apo structure has a closed binding cleft, simulations in all force fields suggest the apo state of MDM2 is highly flexible, and able to sample holo‐like conformations, consistent with a conformational selection model. Initial structuring of the MDM2 lid region, known to competitively bind the binding cleft, is also observed in long simulations. Taken together, these results show molecular simulations can accurately sample conformations relevant for ligand binding. We expect this study to inform future computational work on folding and binding of MDM2 ligands. Proteins 2015; 83:1665–1676. © 2015 Wiley Periodicals, Inc.     

Structural properties of amyloid β(1‐40) dimer explored by replica exchange molecular dynamics simulations

Replica exchange molecular dynamics simulations (300 ns) were used to study the dimerization of amyloid β(1‐40) (Aβ(1‐40)) polypeptide. Configurational entropy calculations revealed that at physiological temperature (310 K, 37°C) dynamic dimers are formed by randomly docked monomers. Free energy of binding of the two chains to each other was ?93.56 ± 6.341 kJ mol^?1. Prevalence of random coil conformations was found for both chains with the exceptions of increased β‐sheet content from residues 16‐21 and 29‐32 of chain A and residues 15‐21 and 30‐33 of chain B with β‐turn/β‐bend conformations in both chains from residues 1‐16, 21‐29 of chain A, 1‐16, and 21‐29 of chain B. There is a mixed β‐turn/β‐sheet region from residues 33‐38 of both chains. Analysis of intra‐ and interchain residue distances shows that, although the individual chains are highly flexible, the dimer system stays in a loosely packed antiparallel β‐sheet configuration with contacts between residues 17‐21 of chain A with residues 17‐21 and 31‐36 of chain B as well as residues 31‐36 of chain A with residues 17‐21 and 31‐36 of chain B. Based on dihedral principal component analysis, the antiparallel β‐sheet‐loop‐β‐sheet conformational motif is favored for many low energy sampled conformations. Our results show that Aβ(1‐40) can form dynamic dimers in aqueous solution that have significant conformational flexibility and are stabilized by collapse of the central and C‐terminal hydrophobic cores with the expected β‐sheet‐loop‐β‐sheet conformational motif. Proteins 2017; 85:1024–1045. © 2017 Wiley Periodicals, Inc.     

Improved side‐chain torsion potentials for the Amber ff99SB protein force field

Recent advances in hardware and software have enabled increasingly long molecular dynamics (MD) simulations of biomolecules, exposing certain limitations in the accuracy of the force fields used for such simulations and spurring efforts to refine these force fields. Recent modifications to the Amber and CHARMM protein force fields, for example, have improved the backbone torsion potentials, remedying deficiencies in earlier versions. Here, we further advance simulation accuracy by improving the amino acid side‐chain torsion potentials of the Amber ff99SB force field. First, we used simulations of model alpha‐helical systems to identify the four residue types whose rotamer distribution differed the most from expectations based on Protein Data Bank statistics. Second, we optimized the side‐chain torsion potentials of these residues to match new, high‐level quantum‐mechanical calculations. Finally, we used microsecond‐timescale MD simulations in explicit solvent to validate the resulting force field against a large set of experimental NMR measurements that directly probe side‐chain conformations. The new force field, which we have termed Amber ff99SB‐ILDN, exhibits considerably better agreement with the NMR data. Proteins 2010. © 2010 Wiley‐Liss, Inc.     

Validation of molecular force field parameters for peptides including isomerized amino acids

Recently, stereoinversions and isomerizations of amino acid residues in the proteins of living beings have been observed. Because isomerized amino acids cause structural changes and denaturation of proteins, isomerizations of amino acid residues are suspected to cause age‐related diseases. In this study, AMBER molecular force field parameters were tested by using computationally generated nonapeptides and tripeptides including stereoinverted and/or isomerized amino acid residues. Energy calculations by using density functional theory were also performed for comparison. Although the force field parameters were developed by parameter fitting for l ‐α‐amino acids, the accuracy of the computational results for d ‐amino acids and β‐amino acids was comparable to those for l ‐α‐amino acids. The conformational energies for tripeptides calculated by using density functional theory were reproduced more accurately than those for nonapeptides calculated by using the molecular mechanical force field. The evaluations were performed for the ff99SB, ff03, ff12SB, and the latest ff14SB force field parameters.     

Evaluating amber force fields using computed NMR chemical shifts

NMR chemical shifts can be computed from molecular dynamics (MD) simulations using a template matching approach and a library of conformers containing chemical shifts generated from ab initio quantum calculations. This approach has potential utility for evaluating the force fields that underlie these simulations. Imperfections in force fields generate flawed atomic coordinates. Chemical shifts obtained from flawed coordinates have errors that can be traced back to these imperfections. We use this approach to evaluate a series of AMBER force fields that have been refined over the course of two decades (ff94, ff96, ff99SB, ff14SB, ff14ipq, and ff15ipq). For each force field a series of MD simulations are carried out for eight model proteins. The calculated chemical shifts for the ¹H, ¹⁵N, and ¹³C^a atoms are compared with experimental values. Initial evaluations are based on root mean squared (RMS) errors at the protein level. These results are further refined based on secondary structure and the types of atoms involved in nonbonded interactions. The best chemical shift for identifying force field differences is the shift associated with peptide protons. Examination of the model proteins on a residue by residue basis reveals that force field performance is highly dependent on residue position. Examination of the time course of nonbonded interactions at these sites provides explanations for chemical shift differences at the atomic coordinate level. Results show that the newer ff14ipq and ff15ipq force fields developed with the implicitly polarized charge method perform better than the older force fields.     

Comparing the folding free‐energy landscapes of Aβ42 variants with different aggregation properties

The properties of the amyloid‐β peptide that lead to aggregation associated with Alzheimer's disease are not fully understood. This study aims at identifying conformational differences among four variants of full‐length Aβ42 that are known to display very different aggregation properties. By extensive all‐atom Monte Carlo simulations, we find that a variety of β‐sheet structures with distinct turns are readily accessible for full‐length Aβ42. In the simulations, wild type (WT) Aβ42 preferentially populates two major classes of conformations, either extended with high β‐sheet content or more compact with lower β‐sheet content. The three mutations studied alter the balance between these classes. Strong mutational effects are observed in a region centered at residues 23–26, where WT Aβ42 tends to form a turn. The aggregation‐accelerating E22G mutation associated with early onset of Alzheimer's disease makes this turn region conformationally more diverse, whereas the aggregation‐decelerating F20E mutation has the reverse effect, and the E22G/I31E mutation reduces the turn population. Comparing results for the four Aβ42 variants, we identify specific conformational properties of residues 23–26 that might play a key role in aggregation. Proteins 2010. © 2010 Wiley‐Liss, Inc.     

Synthesis and receptor binding of opioid peptide analogues containing β3‐homo‐amino acids

β‐Amino acids containing hybrid peptides and β‐peptides show great potential as peptidomimetics. In this paper we describe the synthesis and affinity toward the µ‐ and δ‐opioid receptors of β‐peptides, analogues of Leu‐enkephalin, deltorphin I, dermorphin and α,β‐hybrides, analogues of deltorphin I. Substitution of α‐amino acid residues with β³‐homo‐amino acid residues, in general resulted in decrease of affinity to opioid receptors. However, the incorporation β³h‐D ‐Ala in position 2 or β³hPhe in position 3 of deltorphin I resulted in potent and selective ligand for δ‐opioid receptor. The NMR studies of β‐deltorphin I analogue suggest that conformational motions in the central part of the peptide backbone are partially restricted and some conformational preferences can be expected. Copyright © 2009 European Peptide Society and John Wiley & Sons, Ltd.     

Interaction of a β‐sheet breaker peptide with lipid membranes

Aggregation of β‐amyloid peptides into senile plaques has been identified as one of the hallmarks of Alzheimer's disease. An attractive therapeutic strategy for Alzheimer's disease is the inhibition of the soluble β‐amyloid aggregation using synthetic β‐sheet breaker peptides that are capable of binding Aβ but are unable to become part of a β‐sheet structure. As the early stages of the Aβ aggregation process are supposed to occur close to the neuronal membrane, it is strategic to define the β‐sheet breaker peptide positioning with respect to lipid bilayers. In this work, we have focused on the interaction between the β‐sheet breaker peptide acetyl‐LPFFD‐amide, iAβ5p, and lipid membranes, studied by ESR spectroscopy, using either peptides alternatively labeled at the C‐ and at the N‐terminus or phospholipids spin‐labeled in different positions of the acyl chain. Our results show that iAβ5p interacts directly with membranes formed by the zwitterionic phospholipid dioleoyl phosphatidylcholine and this interaction is modulated by inclusion of cholesterol in the lipid bilayer formulation, in terms of both peptide partition coefficient and the solubilization site. In particular, cholesterol decreases the peptide partition coefficient between the membrane and the aqueous medium. Moreover, in the absence of cholesterol, iAβ5p is located between the outer part of the hydrophobic core and the external hydrophilic layer of the membrane, while in the presence of cholesterol it penetrates more deeply into the lipid bilayer. Copyright © 2010 European Peptide Society and John Wiley & Sons, Ltd.     

Conformational dynamics of a short antigenic peptide in its free and antibody bound forms gives insight into the role of β‐turns in peptide immunogenicity

Earlier immunological experiments with a synthetic 36‐residue peptide (75‐110) from Influenza hemagglutinin have been shown to elicit anti‐peptide antibodies (Ab) which could cross‐react with the parent protein. In this article, we have studied the conformational features of a short antigenic (Ag) peptide (⁹⁸YPYDVPDYASLRS¹¹⁰) from Influenza hemagglutinin in its free and antibody (Ab) bound forms with molecular dynamics simulations using GROMACS package and OPLS‐AA/L all‐atom force field at two different temperatures (293 K and 310 K). Multiple simulations for the free Ag peptide show sampling of ordered conformations and suggest different conformational preferences of the peptide at the two temperatures. The free Ag samples a conformation crucial for Ab binding (β‐turn formed by “DYAS” sequence) with greater preference at 310 K while, it samples a native‐like conformation with relatively greater propensity at 293 K. The sequence “DYAS” samples β‐turn conformation with greater propensity at 310 K as part of the hemagglutinin protein also. The bound Ag too samples the β‐turn involving “DYAS” sequence and in addition it also samples a β‐turn formed by the sequence “YPYD” at its N‐terminus, which seems to be induced upon binding to the Ab. Further, the bound Ag displays conformational flexibility at both 293 K and 310 K, particularly at terminal residues. The implications of these results for peptide immunogenicity and Ag–Ab recognition are discussed. Proteins 2015; 83:1352–1367. © 2015 Wiley Periodicals, Inc.     

The impact of β‐azido(or 1‐piperidinyl)methylamino acids in position 2 or 3 on biological activity and conformation of dermorphin analogues

The synthesis of new dermorphin analogues is described. The (R)‐alanine or phenylalanine residues of natural dermorphin were substituted by the corresponding α‐methyl‐β‐azidoalanine or α‐benzyl‐β‐azido(1‐piperidinyl)alanine residues. The potency and selectivity of the new analogues were evaluated by a competitive receptor binding assay in rat brain using [³H]DAMGO (a μ ligand) and [³H]DELT (a δ ligand). The most active analogue in this series, Tyr‐(R)‐Ala‐(R)‐α‐benzyl‐β‐azidoAla‐Gly‐Tyr‐Pro‐Ser‐NH₂ and its epimer were analysed by ¹H and ¹³C NMR spectroscopy and restrained molecular dynamics simulations. The dominant conformation of the investigated peptides depended on the absolute configuration around C^α in the α‐benzyl‐β‐azidoAla residue in position 3. The (R) configuration led to the formation of a type I β‐turn, whilst switching to the (S) configuration gave rise to an inverse β‐turn of type I′, followed by the formation of a very short β‐sheet. The selectivity of Tyr‐(R)‐Ala‐(R) and (S)‐α‐benzyl‐β‐azidoAla‐Gly‐Tyr‐Pro‐Ser‐NH₂ was shown to be very similar; nevertheless, the two analogues exhibited different conformational preferences. Copyright © 2016 European Peptide Society and John Wiley & Sons, Ltd.     

Force field influences in beta-hairpin folding simulations

All-atom force fields are now routinely used for more detailed understanding of protein folding mechanisms. However, it has been pointed out that use of all-atom force fields does not guarantee more accurate representations of proteins; in fact, sometimes it even leads to biased structural distributions. Indeed, several issues remain to be solved in force field developments, such as accurate treatment of implicit solvation for efficient conformational sampling and proper treatment of backbone interactions for secondary structure propensities. In this study, we first investigate the quality of several recently improved backbone interaction schemes in AMBER for folding simulations of a beta-hairpin peptide, and further study their influences on the peptide's folding mechanism. Due to the significant number of simulations needed for a thorough analysis of tested force fields, the implicit Poisson-Boltzmann solvent was used in all simulations. The chosen implicit solvent was found to be reasonable for studies of secondary structures based on a set of simulations of both alpha-helical and beta-hairpin peptides with the TIP3P explicit solvent as benchmark. Replica exchange molecular dynamics was also utilized for further efficient conformational sampling. Among the tested AMBER force fields, ff03 and a revised ff99 force field were found to produce structural and thermodynamic data in comparably good agreement with the experiment. However, detailed folding pathways, such as the order of backbone hydrogen bond zipping and the existence of intermediate states, are different between the two force fields, leading to force field-dependent folding mechanisms.     

Preparative Enantioseparation of β‐Substituted‐2‐Phenylpropionic Acids by Countercurrent Chromatography With Substituted β‐Cyclodextrin as Chiral Selectors

Preparative enantioseparation of four β‐substituted‐2‐phenylpropionic acids was performed by countercurrent chromatography with substituted β‐cyclodextrin as chiral selectors. The two‐phase solvent system was composed of n‐hexane‐ethyl acetate‐0.10 mol L‐1 of phosphate buffer solution at pH 2.67 containing 0.10 mol L^‐1 of hydroxypropyl‐β‐cyclodextrin (HP‐β‐CD) or sulfobutylether‐β‐cyclodextrin (SBE‐β‐CD). The influence factors, including the type of substituted β‐cyclodextrin, composition of organic phase, concentration of chiral selector, pH value of the aqueous phase, and equilibrium temperature were optimized by enantioselective liquid–liquid extraction. Under the optimum separation conditions, 100 mg of 2‐phenylbutyric acid, 100 mg of tropic acid, and 50 mg of 2,3‐diphenylpropionic acid were successfully enantioseparated by high‐speed countercurrent chromatography, and the recovery of the (±)‐enantiomers was in the range of 90–91% for (±)‐2‐phenylbutyric acid, 91–92% for (±)‐tropic acid, 85–87% for (±)‐2,3‐diphenylpropionic acid with purity of over 97%, 96%, and 98%, respectively. The formation of 1:1 stoichiometric inclusion complex of β‐substituted‐2‐phenylpropionic acids with HP‐β‐CD was determined by UV spectrophotometry and the inclusion constants were calculated by a modified Benesi‐Hildebrand equation. The results showed that different enantioselectivities among different racemates were mainly caused by different enantiorecognition between each enantiomer and HP‐β‐CD, while it might be partially caused by different inclusion capacity between racemic solutes and HP‐β‐CD. Chirality 27:795–801, 2015. © 2015 Wiley Periodicals, Inc.     

Slow formation of aggregation‐resistant β‐sheet folding intermediates

Protein folding has been studied extensively for decades, yet our ability to predict how proteins reach their native state from a mechanistic perspective is still rudimentary at best, limiting our understanding of folding‐related processes in vivo and our ability to manipulate proteins in vitro. Here, we investigate the in vitro refolding mechanism of a large β‐helix protein, pertactin, which has an extended, elongated shape. At 55 kDa, this single domain, all‐β‐sheet protein allows detailed analysis of the formation of β‐sheet structure in larger proteins. Using a combination of fluorescence and far‐UV circular dichroism spectroscopy, we show that the pertactin β‐helix refolds remarkably slowly, with multiexponential kinetics. Surprisingly, despite the slow refolding rates, large size, and β‐sheet‐rich topology, pertactin refolding is reversible and not complicated by off‐pathway aggregation. The slow pertactin refolding rate is not limited by proline isomerization, and 30% of secondary structure formation occurs within the rate‐limiting step. Furthermore, site‐specific labeling experiments indicate that the β‐helix refolds in a multistep but concerted process involving the entire protein, rather than via initial formation of the stable core substructure observed in equilibrium titrations. Hence pertactin provides a valuable system for studying the refolding properties of larger, β‐sheet‐rich proteins, and raises intriguing questions regarding the prevention of aggregation during the prolonged population of partially folded, β‐sheet‐rich refolding intermediates. Proteins 2010. © 2009 Wiley‐Liss, Inc.     

Bayesian Energy Landscape Tilting: Towards Concordant Models of Molecular Ensembles

Predicting biological structure has remained challenging for systems such as disordered proteins that take on myriad conformations. Hybrid simulation/experiment strategies have been undermined by difficulties in evaluating errors from computational model inaccuracies and data uncertainties. Building on recent proposals from maximum entropy theory and nonequilibrium thermodynamics, we address these issues through a Bayesian energy landscape tilting (BELT) scheme for computing Bayesian hyperensembles over conformational ensembles. BELT uses Markov chain Monte Carlo to directly sample maximum-entropy conformational ensembles consistent with a set of input experimental observables. To test this framework, we apply BELT to model trialanine, starting from disagreeing simulations with the force fields ff96, ff99, ff99sbnmr-ildn, CHARMM27, and OPLS-AA. BELT incorporation of limited chemical shift and ³J measurements gives convergent values of the peptide’s α, β, and PP_II conformational populations in all cases. As a test of predictive power, all five BELT hyperensembles recover set-aside measurements not used in the fitting and report accurate errors, even when starting from highly inaccurate simulations. BELT’s principled framework thus enables practical predictions for complex biomolecular systems from discordant simulations and sparse data.     

β‐Hairpin folding and stability: molecular dynamics simulations of designed peptides in aqueous solution

The structural properties of a 10‐residue and a 15‐residue peptide in aqueous solution were investigated by molecular dynamics simulation. The two designed peptides, SYINSDGTWT and SESYINSDGTWTVTE, had been studied previously by NMR at 278 K and the resulting model structures were classified as 3:5 β‐hairpins with a type I + G1 β‐bulge turn. In simulations at 278 K, starting from the NMR model structure, the 3:5 β‐hairpin conformers proved to be stable over the time period evaluated (30 ns). Starting from an extended conformation, simulations of the decapeptide at 278 K, 323 K and 353 K were also performed to study folding. Over the relatively short time scales explored (30 ns at 278 K and 323 K, 56 ns at 353 K), folding to the 3:5 β‐hairpin could only be observed at 353 K. At this temperature, the collapse to β‐hairpin‐like conformations is very fast. The conformational space accessible to the peptide is entirely dominated by loop structures with different degrees of β‐hairpin character. The transitions between different types of ordered loops and β‐hairpins occur through two unstructured loop conformations stabilized by a single side‐chain interaction between Tyr2 and Trp9, which facilitates the changes of the hydrogen‐bond register. In agreement with previous experimental results, β‐hairpin formation is initially driven by the bending propensity of the turn segment. Nevertheless, the fine organization of the turn region appears to be a late event in the folding process. Copyright © 2004 European Peptide Society and John Wiley & Sons, Ltd.     

Folding thermodynamics of β‐hairpins studied by replica‐exchange molecular dynamics simulations

We study the differences in folding stability of β‐hairpin peptides, including GB1 hairpin and a point mutant GB1 K10G, as well as tryptophan zippers (TrpZips): TrpZip1, TrpZip2, TrpZip3‐1, and TrpZip4. By performing replica‐exchange molecular dynamics simulations with Amber03* force field (a modified version of Amber ff03) in explicit solvent, we observe ab initio folding of all the peptides except TrpZip3‐1, which is experimentally known to be the least stable among the peptides studied here. By calculating the free energies of unfolding of the peptides at room temperature and folding midpoint temperatures for thermal unfolding of peptides, we find that TrpZip4 and GB1 K10G peptides are the most stable β‐hairpins followed by TrpZip1, GB1, and TrpZip2 in the given order. Hence, the proposed K10G mutation of GB1 peptide results in enhanced stability compared to wild‐type GB1. An important goal of our study is to test whether simulations with Amber 03* model can reproduce experimentally predicted folding stability differences between these peptides. While the stabilities of GB1 and TrpZip1 yield close agreement with experiment, TrpZip2 is found to be less stable than predicted by experiment. However, as heterogenous folding of TrpZip2 may yield divergent thermodynamic parameters by different spectroscopic methods, mismatching of results with previous experimental values are not conclusive of model shortcomings. For most of the cases, molecular simulations with Amber03* can successfully reproduce experimentally known differences between the mutated peptides, further highlighting the predictive capabilities of current state‐of‐the‐art all‐atom protein force fields. Proteins 2015; 83:1307–1315. © 2015 Wiley Periodicals, Inc.     

Thermodynamics and dynamics of amyloid peptide oligomerization are sequence dependent

Aggregation of the full‐length amyloid‐β (Aβ) and β2‐microglobulin (β2m) proteins is associated with Alzheimer's disease and dialysis‐related amyloidosis, respectively. This assembly process is not restricted to full‐length proteins, however, many short peptides also assemble into amyloid fibrils in vitro. Remarkably, the kinetics of amyloid‐fibril formation of all these molecules is generally described by a nucleation‐polymerization process characterized by a lag phase associated with the formation of a nucleus, after which fibril elongation occurs rapidly. In this study, we report using long molecular dynamics simulations with the OPEP coarse‐grained force field, the thermodynamics and dynamics of the octamerization for two amyloid 7‐residue peptides: the β2m83‐89 NHVTLSQ and Aβ16‐22 KLVFFAE fragments. Based on multiple trajectories run at 310 K, totaling 2.2 μs (β2m83‐89) and 4.8 μs (Aβ16‐22) and starting from random configurations and orientations of the chains, we find that the two peptides not only share common but also very different aggregation properties. Notably, an increase in the hydrophobic character of the peptide, as observed in Aβ16‐22 with respect to β2m83‐89 impacts the thermodynamics by reducing the population of bilayer β‐sheet assemblies. Higher hydrophobicity is also found to slow down the dynamics of β‐sheet formation by enhancing the averaged lifetime of all configuration types (CT) and by reducing the complexity of the CT transition probability matrix. Proteins 2009. © 2008 Wiley‐Liss, Inc.     

Side‐chain hydrophobicity and the stability of Aβ16–22 aggregates

Recent mutagenesis studies using the hydrophobic segment of Aβ suggest that aromatic π‐stacking interactions may not be critical for fibril formation. We have tested this conjecture by probing the effect of Leu, Ile, and Ala mutation of the aromatic Phe residues at positions 19 and 20, on the double‐layer hexametric chains of Aβ fragment Aβ_16–22 using explicit solvent all‐atom molecular dynamics. As these simulations rely on the accuracy of the utilized force fields, we first evaluated the dynamic and stability dependence on various force fields of small amyloid aggregates. These initial investigations led us to choose AMBER99SB‐ILDN as force field in multiple long molecular dynamics simulations of 100 ns that probe the stability of the wild‐type and mutants oligomers. Single‐point and double‐point mutants confirm that size and hydrophobicity are key for the aggregation and stability of the hydrophobic core region (Aβ_16–22). This suggests as a venue for designing Aβ aggregation inhibitors the substitution of residues (especially, Phe 19 and 20) in the hydrophobic region (Aβ_16–22) with natural and non‐natural amino acids of similar size and hydrophobicity.