AI-assisted proofreading of RNA splicing

Area of Science:

• Molecular Biology
• RNA Splicing
• Protein Interactions

Background:

• Accurate intron removal from pre-mRNAs is crucial and involves RNA helicases.
• The recruitment mechanisms of these helicases to spliceosome/pre-mRNA complexes are not well understood.

Purpose of the Study:

• To elucidate the molecular mechanism by which SF3B1 interacts with SUGP1 to recruit the DHX15 RNA helicase.
• To develop a model explaining how this interaction influences splicing accuracy and aberrant splice site activation.

Main Methods:

• Biochemical experiments
• AI-based structure prediction
• Protein-protein interaction modeling

Main Results:

• A model was generated for the interaction between SF3B1, SUGP1, and DHX15.
• SF3B1 interaction exposes SUGP1's G-patch domain, enabling DHX15 binding and activation.
• The model explains cryptic 3' splice site activation linked to SF3B1/SUGP1 mutations in cancer.

Conclusions:

• The study provides a mechanistic model for RNA helicase recruitment in splicing.
• This finding sheds light on the role of SF3B1, SUGP1, and DHX15 in maintaining splicing fidelity.
• The model offers insights into cancer-associated splicing defects.