Structure-based characterization of the binding of peptide to the human endophilin-1 Src homology 3 domain using position-dependent noncovalent potential analysis

Machine Learning in Quantitative Protein–peptide Affinity Prediction: Implications for Therapeutic Peptide Design

Background:

Protein–peptide recognition plays an essential role in the orchestration and regulation of cell signaling networks, which is estimated to be responsible for up to 40% of biological interaction events in the human interactome and has recently been recognized as a new and attractive druggable target for drug development and disease intervention.

Methods:

We present a systematic review on the application of machine learning techniques in the quantitative modeling and prediction of protein–peptide binding affinity, particularly focusing on its implications for therapeutic peptide design. We also briefly introduce the physical quantities used to characterize protein–peptide affinity and attempt to extend the content of generalized machine learning methods.

Results:

Existing issues and future perspective on the statistical modeling and regression prediction of protein– peptide binding affinity are discussed.

Conclusion:

There is still a long way to go before establishment of general, reliable and efficient machine leaningbased protein–peptide affinity predictors.

BIN1 membrane curvature sensing and generation show autoinhibition regulated by downstream ligands and PI(4,5)P2

In striated muscles, invaginations from the plasma membrane, termed transverse tubules (T-tubule), function in the excitation–contraction coupling machinery. BIN1 (isoform8) plays a critical role in the biogenesis of T-tubules. BIN1 contains an N-terminal BAR domain to sense and induce membrane curvature, an isoform8-specific polybasic motif (exon10) as the phosphoinositide binding module and a C-terminal Src homology 3 (SH3) domain for the recruitment of downstream proteins such as dynamin 2. Previous studies of N-BAR domains focused on elucidating mechanisms of membrane curvature sensing and generation (MC-S&G). Less is known about how MC-S&G is regulated. We found that the SH3 domain binds to the exon10 motif more strongly compared to the proline-rich domain (PRD) of dynamin 2. Furthermore, we found that the MC-S&G ability of full-length BIN1 is inhibited on membranes lacking PI(4,5)P₂. Addition of PI(4,5)P₂ in the membrane activates BIN1 to sense and induce membrane curvature. Co-presence of the SH3 domain and exon10 motif leads to the strongest phosphoinositide-mediated control of BIN1 function. Addition of SH3 domain ligand (such as PRD peptides), as well as addition of the water-soluble PI(4,5)P₂ analogue, can both enhance the MC-S&G ability of BIN1 on membranes without PI(4,5)P₂, indicating that the key to activate BIN1 is to disrupt the exon10–SH3 interaction. The nonsense mutation K436X, found in centronuclear myopathy (CNM) patients, abolishes SH3 domain binding with either exon10 or the PRD motif, resulting in increased membrane deformation capacity. Our results suggest an autoinhibition model for BIN1 that involves a synergistic regulation by membrane composition and protein–protein interactions.