Sampling and energy evaluation challenges in ligand binding protein design.

Area of Science:

• Protein Engineering
• Computational Biology
• Biomarker Detection

Background:

• 17α-hydroxylprogesterone (17-OHP) is a critical biomarker for congenital adrenal hyperplasia.
• Development of sensitive sensors for 17-OHP is of significant clinical interest.
• Computational protein design offers a powerful approach for creating novel binding proteins.

Purpose of the Study:

• To design novel proteins with specific binding sites for 17-OHP using computational methods.
• To experimentally validate the binding affinity and structural characteristics of designed proteins.
• To iteratively improve binding affinity and understand design-model discrepancies.

Main Methods:

• Utilized computational protein design to engineer binding pockets complementary to 17-OHP.
• Incorporated nonpolar residues for the steroid core and hydrogen-bonding residues for polar groups.
• Experimental testing, co-crystallography, mutagenesis, and binding selection were employed for validation and optimization.

Main Results:

• Eight out of 16 designed proteins exhibited 17-OHP binding with micromolar affinity.
• Co-crystal structure revealed a unique, rotated binding mode of 17-OHP within the designed pocket.
• Iterative design and selection enhanced binding affinity to the nanomolar range and refined ligand conformation.

Conclusions:

• Computational protein design can successfully generate proteins with specific binding capabilities for steroid hormones like 17-OHP.
• Discrepancies between design and crystal structures highlight the need for improved modeling of backbone flexibility and solvation energies.
• Engineered proteins show promise for developing advanced sensors for clinical biomarkers.