AlphaFold2 SLiM screen for LC3-LIR interactions in autophagy

Area of Science:

• Molecular Biology and Biochemistry
• Structural Biology
• Computational Biology

Background:

• Selective autophagy relies on proteins with LC3-interacting regions (LIRs) or Atg8-family interacting motifs (AIMs) to recruit cargo to the phagophore.
• Post-translational modifications, particularly phosphorylation, significantly influence the binding of these motifs to LC3/Atg8 proteins.
• Current computational methods for predicting LIRs face challenges with non-canonical motifs and large protein structures, limiting their accuracy and applicability.

Purpose of the Study:

• To develop and validate a fragment-based computational approach for improved prediction of LIRs and related protein-protein interactions.
• To overcome the limitations of existing tools like AlphaFold2 in predicting functional LIRs, especially for complex or non-canonical motifs.
• To provide a framework for optimizing LIR screening, including guidance on fragment selection, sequence tuning, and scoring.

Main Methods:

• A novel fragment-based computational strategy was developed to predict protein-protein interactions involving short linear motifs (SLiMs).
• Systematic fragment screening was performed on various target proteins to generate structural models.
• The approach was tested for its ability to predict interactions missed by full-length protein modeling with AlphaFold2 and AlphaFold3, and its transferability to SUMO-SIM interactions was assessed.

Main Results:

• The fragment-based approach successfully predicted structural models for interactions that full-length AlphaFold2 and AlphaFold3 failed to capture.
• Fragment length and phosphomimetic mutations were found to modulate the predicted interactions, providing insights into regulatory mechanisms.
• The study yielded guidance on optimizing fragment choice, sequence modifications, and scoring for effective LIR screening.

Conclusions:

• The fragment-based method offers a powerful alternative for predicting LIRs and other SLiM-mediated interactions, addressing limitations of current computational tools.
• This approach enhances the structural modeling of autophagy-related protein interactions and can be extended to other SLiM-based interactions.
• The findings provide a valuable framework for researchers studying autophagy and related cellular processes, improving the identification of functional motifs.