Identification of amyloidogenic peptide sequences using a coarse-grained physicochemical model

Area of Science:

• Biophysics
• Computational Biology
• Molecular Biology

Background:

• Cross-beta amyloid fibrils are associated with over 20 human diseases.
• Specific sequence elements within amyloidogenic proteins are critical for initiating amyloid formation (seeding).
• Experimental structure determination of amyloid species is challenging, necessitating theoretical approaches.

Purpose of the Study:

• To identify amyloidogenic sequences in proteins using computational methods.
• To rationalize the molecular mechanisms underlying amyloid formation.
• To aid in the development of targeted therapeutic strategies against amyloid diseases.

Main Methods:

• Utilized a coarse-grained physicochemical protein model.
• Employed an efficient Monte Carlo sampling technique.
• Defined amyloidogenic sequences based on predicted stable, three-stranded beta-sheet structures of interacting peptides.

Main Results:

• Successfully identified amyloidogenic sequences in four proteins with existing experimental peptide fragmentation data.
• Calculated free energies to quantify amyloid propensity.
• Achieved good agreement with experimental data, identifying key self-recognition motifs in two proteins.

Conclusions:

• The combined coarse-grained model and Monte Carlo sampling efficiently predicts amyloidogenic sequences.
• This computational approach is significantly faster than atomistic molecular dynamics simulations.
• The findings provide valuable insights into cross-beta amyloid fibril formation and potential therapeutic targets.