Different ways to look at, model, and understand peptide-protein interactions
Peptide-mediated interactions, namely interactions mediated by short peptides or short linear stretches within a partner, play fundamental roles in biological regulation. The characterization of the structural details of these interactions can provide new insights into their functional role, and new handles for their targeted manipulation. However, their transient character and the peptide flexibility pose considerable challenges to accurate structural modeling of peptide-protein complexes. We have developed several approaches for the high resolution docking of peptide-protein complexes. Given a binding site and an approximate starting conformation, Rosetta FlexPepDock  can generate high-resolution models by optimization of both the peptide conformation as well as its rigid body orientation in the binding site. But what if the binding site is unknown?
By looking at peptide-protein interactions from different viewpoints, we have developed three different ways to solve the global docking problem. Looking at the problem from point of view of the peptide, we have shown that the peptide bound conformation is already sampled in its free form, and that by generating a representative conformer ensemble we can reduce the complexity of the problem to parallel rigid body docking (PIPER-FlexPepDock ). Looking at the problem from the point of view of the receptor, we have shown that by mapping the receptor surface for structurally similar motifs in solved structures, we can extract complementing fragments as starting points for docking (PatchMAN; , under review). Finally, looking at peptide binding as a final step in monomer folding, we have shown that alphafold2 can easily be adapted to peptide docking, providing excellent results, even if the receptor structure undergoes significant changes upon binding (Alphafold2 peptide docking; , under review). Thus, we are able to leverage the deep learning revolution in accurate structural modeling for the modeling of peptide-protein interactions. Each of these strategies has revealed new principles that govern peptide-mediated interactions, and together they provide an excellent starting point for the accurate large-scale structural characterization of these interactions.
- Raveh B, London N, Schueler-Furman O: Sub-angstrom modeling of complexes between flexible peptides and globular proteins. Proteins Struct Funct Bioinforma 2010, 78:2029–2040.
- Alam N, Goldstein O, Xia B, Porter KA, Kozakov D, Schueler-Furman O: High-resolution global peptide-protein docking using fragments-based PIPER-FlexPepDock. PLoS Comput Biol 2017, 13:e1005905.
- Khramushin A, Tsaban T, Varga J, Avraham O, Schueler-Furman O: PatchMAN docking: Modeling peptide-protein interactions in the context of the receptor surface. bioRxiv 2021, doi:10.1101/2021.09.02.458699.
- Tsaban T, Varga J, Avraham O, Ben-Aharon Z, Khramushin A, Schueler-Furman O: Harnessing protein folding neural networks for peptide-protein docking. bioRxiv 2021, doi:10.1101/2021.08.01.454656.
Hosted by Professor Sagar Khare