DomainFit: Identification of protein domains in cryo-EM maps at intermediate resolution using AlphaFold2-predicted...

Area of Science:

• Structural biology
• Biochemistry
• Cell biology

Background:

• Cryo-electron microscopy (cryo-EM) enables structural determination of macromolecular complexes.
• In situ and ex vivo structural biology reveals native complexes, often with unidentified proteins.
• Identifying proteins in cryo-EM maps below 4 Å resolution is challenging due to unreliable side-chain visualization.

Purpose of the Study:

• To present DomainFit, a novel program for semi-automated, domain-level protein identification in cryo-EM maps.
• To address the challenge of identifying proteins in low-resolution cryo-EM data (below 4 Å).

Main Methods:

• DomainFit utilizes AlphaFold2-predicted models, fitting domains into cryo-EM maps.
• The program performs statistical analyses to identify protein domains and candidates contributing to the observed density.
• Application of DomainFit to analyze cryo-EM data of Tetrahymena thermophila.

Main Results:

• DomainFit facilitates semi-automated protein identification from cryo-EM density maps.
• The tool is particularly effective for maps with resolutions lower than 4 Å.
• Two microtubule inner proteins were identified in Tetrahymena thermophila, including one with a CCDC81 domain localized to the proximal region of doublet microtubules.

Conclusions:

• DomainFit is a valuable tool for advancing in situ structural biology by enabling protein identification in low-resolution cryo-EM maps.
• The identification of novel microtubule-associated proteins contributes to understanding microtubule structure and function.
• This method aids in characterizing complex cellular machinery and processes.