In silico prediction of heme binding in proteins

Area of Science:

• Biochemistry and Structural Biology
• Bioinformatics and Computational Biology
• Molecular Biology

Background:

• Heme binding to proteins is crucial for diverse biological functions, including oxygen transport, electron transfer, and catalysis.
• Heme also acts as a regulatory molecule, binding weakly to proteins for functions like transcriptional regulation and ion channel control.
• Characterizing these regulatory heme-protein interactions is challenging due to weak binding and lack of discrete binding pockets.

Purpose of the Study:

• To develop and validate a computational approach for identifying heme-binding sites in regulatory proteins.
• To facilitate the prediction and understanding of heme-protein interactions, particularly for regulatory roles.
• To provide a tool applicable to both existing structural data and predicted protein models.

Main Methods:

• A modified version of the ProFunc bioinformatics tool was employed.
• The tool was applied to a dataset of heme-dependent regulatory proteins from the Protein Data Bank and AlphaFold models.
• Identified heme-binding sites were visualized using PyMol and could be refined with RosettaDOCK.

Main Results:

• The methodology successfully identified potential heme-binding sites in the test set of regulatory proteins.
• The accuracy of the method is dependent on the quality of the input structural information.
• The tool is effective even when crystal structures are unavailable, relying on predicted models.

Conclusions:

• The modified ProFunc tool offers a valuable approach for identifying heme-binding sites in regulatory proteins.
• This method can aid in the study of heme-protein interactions, especially in cases lacking experimental structures.
• The publicly available tool can be readily adopted for future research on novel heme-binding targets.