Structure-guided analysis and prediction of human E2-E3 ligase pairing specificity.

Area of Science:

• Biochemistry and Molecular Biology
• Structural Biology
• Computational Biology

Background:

• Protein ubiquitination, regulated by E2 and E3 enzymes, is crucial for selective protein degradation and cellular processes.
• Dysregulation of ubiquitination is linked to cancer and other diseases, highlighting therapeutic potential.
• Existing knowledge lacks a systematic understanding of E2-E3 enzyme interactions and substrate specificities.

Purpose of the Study:

• To develop a predictive model for functional E2-E3 enzyme pairings in ubiquitination networks.
• To provide structural insights into ubiquitin-E2-E3 ternary complexes.
• To create a publicly accessible resource for E2-E3 pairing predictions and structures.

Main Methods:

• Analysis of experimental structures from the Protein Data Bank (PDB).
• Structure generation of thousands of ubiquitin-E2-E3 ternary complexes using AlphaFold.
• Development of a machine learning model to predict functional E2-E3 pairings.
• Bioinformatic analysis of experimentally determined structures and AlphaFold modeling.

Main Results:

• A machine learning model was developed to predict functional E2-E3 pairings.
• The model predicted E2 partners for 88 E3 ligases lacking known interactors.
• A potential functional link between UBE2C and RNF214 in hepatocellular carcinoma was identified.
• The UbiqCore web resource was created, providing access to predictions and structures.

Conclusions:

• The study advances the mapping of ubiquitination networks by predicting functional E2-E3 pairings.
• Structural insights into E2-E3 complexes are provided, aiding in understanding ubiquitination mechanisms.
• The UbiqCore resource facilitates future biological and therapeutic discoveries in ubiquitination.
• The findings support the potential of targeting ubiquitination pathways for therapeutic interventions.