Data-Driven Engineering of Thermostable Collagen-Mimetic Peptoid Triple Helices

Area of Science:

• Biomaterials Science
• Computational Chemistry
• Molecular Biology

Background:

• Collagen-mimetic peptides (CMPs) mimic natural collagen's triple helix using a repeating x-y-Gly sequence.
• Modifying x and y positions allows tuning CMP structure and properties.
• Peptoids offer stability and diverse side chains but can reduce triple helix stability due to backbone flexibility.

Purpose of the Study:

• To computationally identify peptoid substitutions that stabilize collagen-like triple helices.
• To develop a generalizable strategy for engineering stable peptoid-based biomaterials.

Main Methods:

• Utilized a computational active learning cycle involving molecular dynamics simulations.
• Employed Gaussian process regression and Bayesian optimization to predict stabilizing peptoid substitutions.
• Synthesized and characterized a top computational candidate using scanning electron microscopy.

Main Results:

• Identified several peptoid substitutions predicted to stabilize collagen-like triple helices through side chain interactions.
• A synthesized peptoid-based CMP candidate formed fibril-like bundles, consistent with stable triple helices.
• Demonstrated the potential of peptoid substitutions to create stable, collagen-like quaternary structures.

Conclusions:

• This study successfully predicts and validates peptoid substitutions for stable collagen-mimetic peptides.
• Presents a novel, generalizable computational design strategy for engineering peptoid-based biomaterials.
• Opens new avenues for developing advanced biomimetic materials with tailored properties.