Chapter Nine - Aggregation of disease-related peptides

https://doi.org/10.1016/bs.pmbts.2019.12.002Get rights and content

Abstract

Protein misfolding and aggregation of amyloid proteins is the fundamental cause of more than 20 diseases. Molecular mechanisms of the self-assembly and the formation of the toxic aggregates are still elusive. Computer simulations have been intensively used to study the aggregation of amyloid peptides of various amino acid lengths related to neurodegenerative diseases. We review atomistic and coarse-grained simulations of short amyloid peptides aimed at determining their transient oligomeric structures and the early and late aggregation steps.

Introduction

Numerous diseases affecting either the central nervous system or a variety of peripheral tissues result from the self-assembly of amyloid proteins. Disorders range from Alzheimer's disease (AD, implication of Aβ proteins of 39–43 amino acids and tau protein of 441 amino acids), Parkinson (α-synuclein protein of 140 amino acids) to amyotrophic lateral sclerosis or ALS (superoxide dismutase, SOD of 154 amino acids) and type II diabetes (IAPP or islet amyloid polypeptide of 37 amino acids). All these proteins differ in amino acid sequence and length, yet they all form amyloid fibrils with cross-β structure.1 This propensity to self-assembly into amyloid fibrils under given conditions is also observed for heptapeptides (e.g., Aβ16–22),2 tetrapeptides (e.g., KFFE),3 dipeptides4 and even by single amino acids.5

Experimentally, aggregation kinetics of amyloid proteins display a sigmoidal curve with a lag phase, during which monomers self-assembly into oligomers and undergo structural rearrangements, until the growth phase where fibril elongation and primary/secondary nucleation occur, followed by the saturation phase where the system is in equilibrium between fibrils and a small concentration of monomers. Note that our understanding of amyloid aggregation kinetics goes much beyond classical nucleation theory where primary nucleation event is sufficient to fit the experimental curves.1

Structural determination of all species along amyloid fibril formation pathways is challenging by experimental means because of their transient character despite the use of a large variety of biophysical techniques.6 Complexity comes from the metastable character of each species (a very large number of conformations for each aggregate), but also from the sensitivity of the kinetics to experimental conditions (solution pH, temperature, salt concentration, agitation, ions) and external conditions (presence of membrane, crowding, etc.).7

To add further complexity, amyloid aggregation kinetics is also modulated by the exact composition of the lipid bilayers. For example, although Aβ1–40, Aβ1–42 and tau oligomers are key players in AD,8 dietary PUFA (polyunsaturated fatty acids) supplementation change molecular phospholipids, and there is evidence that increased intake of omega-3 PUFA slows the progression of AD, while omega-6 PUFA is linked to higher risk of AD.9 Another example of the modulation of the aggregation upon membrane presence can be taken from experiments on α-synuclein. On the one hand, DPLS lipid bilayers significantly augment its aggregation rate, while DOPE lipid bilayers have no impact on its aggregation rate.10

Simulations on amyloids at different time and length scales can complement experiment, but require accurate potential energy models ranging from atomistic in explicit aqueous solution/lipid bilayer, coarse-grained with implicit solvent/membrane to mesoscopic or super-mesoscopic representations.11, 12, 13, 14 Molecular dynamics (MD), replica exchange molecular dynamics (REMD) simulations and other sampling methods are often used to generate the conformational ensemble of intrinsically disordered proteins (IDPs). These methods are described in Section 2 and we will see that their results vary with the protein force field used. Section 3 focuses on the application of computer simulations to a better understanding of small oligomers of short amyloid peptides. While an implicit solvent representation reduces the number of degrees of freedom, hydrodynamics, which deals with the motion of fluids and the forces acting on solid bodies immersed in fluids and in motion relative to them, plays a significant role in the early aggregation steps of amyloids using simplified models (Section 4). Section 5 reports on the simulations aimed at understanding the primary nucleation and the surface-catalyzed secondary nucleation. The final section reports on recent advances in the determination of oligomer structures of Aβ40/42 peptides in aqueous solution and lipid bilayers based on computer simulations. Our review will be mainly centered on Aβ peptides. For computer simulations of tau, hIAAP and synuclein, see Ref. [14].

Section snippets

Computer simulation models for amyloid protein aggregation

Numerous conformational sampling methods are used to study amyloid proteins. We will not describe them in detail but rather summarize their main features. Atomistic molecular dynamics (MD) simulations in explicit environment offer the most detailed dynamic and energetic pictures of protein folding and aggregation. The longest simulation on the fastest computer (Anton) does not exceed, however, 1 ms in explicit solvent, sufficient for sampling the monomeric state of amyloid proteins, but far too

Structures of small aggregates

Among all amyloid sequences, the Aβ16–22 peptide has been the most studied. Due to its simplicity, its amyloid fibril at neutral pH consists of antiparallel β-strands and parallel β-sheets,2 and the amino-acids 17–21 constitute the main driving force for the aggregation of the full-length Aβ peptides.8 Based on MD and REMD simulations with all force fields except AMBER99sb-disp, it is clear that the probability of Aβ16–22 fibril formation following the one-step nucleation is very small.

Using

Exploring the early aggregates of amyloid peptides at quasi-atomic level with hydrodynamics

Coarse-grained simplified molecular models with implicit solvent have been extensively used to explore the aggregation process of amyloid systems and to inspect the impact of the peptide β-propensity and amino acid sequence, peptide-peptide interactions, concentration, temperature,46, 47, 52, 53, 71, 72 and crowding.73 However, these simulations based on simplified models neglect the effect of solvent mediated interactions on the kinetic behavior of the system and the diffusion limited

Primary and secondary nucleation from simulations

The foundation of the nucleation for amyloid fibrils dates from the polymerization studies of actin by Oosawa83 and deoxyhemoglobin by Weaton.84 They theoretically explained the nucleation and subsequent polymerization processes by employing the homogeneous nucleation theory developed for vapor condensation. Their theories have served as a background for the development of kinetic and thermodynamic analyses for amyloid nucleation.85

Basically, in the first step, the fluctuations in the system

Recent advances in structures of Aβ40/42 oligomers from simulations

In this last section, we would like to review some recent and significant contributions in the field of Aβ40/42 peptide simulations. We already discussed the simulation results of Aβ42 dimer in aqueous solution (see Section 2).

In the case of Aβ40 dimer in aqueous solution, Nguyen et al. performed atomistic REMD simulations on the wild-type (WT) sequence, the A2V/A2V mutant and the mixed WT/A2V mutant.100, 101, 102 Experimentally, the A2V mutation protects from Alzheimer's disease in its

Conclusions

Understanding how amyloid aggregates actually become toxic is truly a real challenge in developing a treatment for neurodegenerative diseases, as only monomers are nontoxic. Thus far, all molecules (antibodies and drugs) targeting amyloid-β oligomers have failed to pass clinical trials. Many reasons have been put forward to explain this repetitive failure.112, 113

Atomistic and coarse-grained simulations with increased computer efficiency, improved force field accuracy,15, 114 coupling to

Conflict of interest

The authors declare no competing financial interest.

Acknowledgments

We acknowledge support by the “Initiative d'Excellence” program from the French State (Grant “DYNAMO,” ANR-11-LABX-0011-01, and “CACSICE,” ANR-11-EQPX-0008). PhD thanks Université de Paris, CNRS and PSL.

References (119)

  • M. Cheon et al.

    Spontaneous formation of twisted Aβ16  22 fibrils in large-scale molecular dynamics simulations

    Biophys J

    (2011)
  • F. Chiti et al.

    Protein misfolding, amyloid formation, and human disease: a summary of progress over the last decade

    Annu Rev Biochem

    (2017)
  • A. Melquiond et al.

    Following the aggregation of amyloid-forming peptides by computer simulations

    J Chem Phys

    (2005)
  • E. Gazit

    Reductionist approach in peptide-based nanotechnology

    Annu Rev Biochem

    (2018)
  • N. Gour et al.

    Amyloid-like structures formed by single amino acid self-assemblies of cysteine and methionine

    ACS Chem Neurosci

    (2019)
  • J. Nasica-Labouze et al.

    Amyloid beta protein and Alzheimer's disease: when computer simulations complement experimental studies

    Chem Rev

    (2015)
  • G. Meisl et al.

    Quantitative analysis of intrinsic and extrinsic factors in the aggregation mechanism of Alzheimer-associated Aβ-peptide

    Sci Rep

    (2016)
  • D.G. Selkoe et al.

    The amyloid hypothesis of Alzheimer's disease at 25 years

    EMBO Mol Med

    (2016)
  • Y. Lu et al.

    Amyloid-β(29-42) dimeric conformations in membranes rich in omega-3 and omega-6 polyunsaturated fatty acids

    J Phys Chem B

    (2019)
  • C. Galvagnion et al.

    Chemical properties of lipids strongly affect the kinetics of the membrane-induced aggregation of α-synuclein

    Proc Natl Acad Sci USA

    (2016)
  • P.H. Nguyen et al.

    Familial Alzheimer A2V mutation reduces the intrinsic disorder and completely changes the free energy landscape of the Aβ1-28 monomer

    J Phys Chem B

    (2014)
  • M. Zhang et al.

    Molecular simulations of amyloid structures toxicity and inhibition

    Isr J Chem

    (2017)
  • I.M. Ilie et al.

    Simulation studies of amyloidogenic polypeptides and their aggregates

    Chem Rev

    (2019)
  • P. Robustelli et al.

    Developing a molecular dynamics force field for both folded and disordered protein states

    Proc Natl Acad Sci USA

    (2018)
  • C.T. Leahy et al.

    Peptide dimerization-dissociation rates from replica exchange molecular dynamics

    J Chem Phys

    (2017)
  • B. Narayan et al.

    Conformational analysis of replica exchange MD: temperature-dependent Markov networks for FF amyloid peptides

    J Chem Phys

    (2018)
  • A. Irback et al.

    Aggregate geometry in amyloid fibril nucleation

    Phys Rev Lett

    (2013)
  • A. Šarić et al.

    Kinetics of spontaneous filament nucleation via oligomers: insights from theory and simulation

    J Chem Phys

    (2016)
  • N.S. Bieler et al.

    Connecting macroscopic observables and microscopic assembly events in amyloid formation using coarse grained simulations

    PLoS Comput Biol

    (2012)
  • T.T. Tran et al.

    Lattice model for amyloid peptides: OPEP force field parametrization and applications to the nucleus size of Alzheimer's peptides

    J Chem Phys

    (2016)
  • M.S. Li et al.

    Probing the mechanisms of fibril formation using lattice models

    J Chem Phys

    (2008)
  • P. Tuffery et al.

    Flexibility and binding affinity in protein-ligand, protein-protein and multi-component protein interactions: limitations of current computational approaches

    J R Soc Interface

    (2012)
  • G. Stirnemann et al.

    Recovering protein thermal stability using all-atom Hamiltonian replica-exchange simulations in explicit solvent

    J Chem Theory Comput

    (2015)
  • J. Nasica-Labouze et al.

    A multiscale approach to characterize the early aggregation steps of the amyloid-forming peptide GNNQQNY from the yeast prion sup-35

    PLoS Comput Biol

    (2011)
  • A.K. Smith et al.

    Does replica exchange with solute tempering efficiently sample Aβ peptide conformational ensembles?

    J Chem Theory Comput

    (2016)
  • P.H. Nguyen et al.

    Communication: simulated tempering with fast on-the-fly weight determination

    J Chem Phys

    (2013)
  • T. Zhang et al.

    Folding atomistic proteins in explicit solvent using simulated tempering

    J Phys Chem B

    (2015)
  • M. Hoang Viet et al.

    Communication: multiple atomistic force fields in a single enhanced sampling simulation

    J Chem Phys

    (2015)
  • S.J. Bunce et al.

    Molecular insights into the surface-catalyzed secondary nucleation of amyloid-β40 (Aβ40) by the peptide fragment Aβ16-22

    Sci Adv

    (2019)
  • D. Granata et al.

    The inverted free energy landscape of an intrinsically disordered peptide by simulations and experiments

    Sci Rep

    (2015)
  • F. Baftizadeh et al.

    Multidimensional view of amyloid fibril nucleation in atomistic detail

    J Am Chem Soc

    (2012)
  • J. Huang et al.

    CHARMM36m: an improved force field for folded and intrinsically disordered proteins

    Nat Methods

    (2017)
  • R.B. Best

    Atomistic force fields for proteins

    Methods Mol Biol

    (2019)
  • P.H. Nguyen et al.

    Effects of all-atom force fields on amyloid oligomerization: replica exchange molecular dynamics simulations of the Aβ(16-22) dimer and trimer

    Phys Chem Chem Phys

    (2011)
  • V.H. Man et al.

    Effects of all-atom molecular mechanics force fields on amyloid peptide assembly: the case of Aβ16–22 dimer

    J Chem Theory Comput

    (2019)
  • M. Carballo-Pacheco et al.

    On the applicability of force fields to study the aggregation of amyloidogenic peptides using molecular dynamics simulations

    J Chem Theory Comput

    (2018)
  • C.M. Siwy et al.

    Is the conformational ensemble of Alzheimer's Aβ10-40 peptide force field dependent?

    PLoS Comput Biol

    (2017)
  • D.J. Rosenman et al.

    Characterization of Aβ monomers through the convergence of ensemble properties among simulations with multiple force fields

    J Phys Chem B

    (2016)
  • M. Carballo-Pacheco et al.

    Comparison of force fields for Alzheimer's Aβ42: a case study for intrinsically disordered proteins

    Protein Sci

    (2017)
  • O.C. Weber et al.

    How accurate are your simulations? Effects of confined aqueous volume and AMBER FF99SB and CHARMM22/CMAP force field parameters on structural ensembles of intrinsically disordered proteins: Amyloid-β42 in water

    Intrinsically Disord Proteins

    (2017)
  • Cited by (0)

    View full text