Computational design of ligand-binding proteins with high affinity and selectivity

Area of Science:

• Protein engineering
• Computational biology
• Molecular recognition

Background:

• Designing proteins with specific small molecule binding capabilities is challenging.
• Current methods like antibody production and directed evolution offer limited control over binding interactions.
• Computational design of protein-small-molecule interfaces remains an unsolved problem.

Purpose of the Study:

• To develop a general computational method for designing pre-organized, shape-complementary small-molecule-binding sites.
• To generate novel protein binders for the steroid digoxigenin (DIG).

Main Methods:

• A novel computational method was employed to design protein binding sites.
• Seventeen designs were experimentally characterized.
• Library selections and deep sequencing were used to generate a binding-fitness landscape.
• X-ray co-crystallography was used to validate computational models.

Main Results:

• Two out of seventeen designed proteins successfully bound DIG.
• The highest affinity binder showed energetically favorable and pre-organized interfaces.
• Optimization achieved picomolar binding affinity.
• Designed binders demonstrated selectivity for DIG over related steroids and allowed for reprogramming of binding preference.

Conclusions:

• The presented computational design method enables the creation of proteins with tailored small molecule binding properties.
• This approach can be used to develop next-generation biosensors, therapeutics, and diagnostics.
• The method offers precise control over molecular recognition through designed interactions.