De novo high-resolution protein structure determination from sparse spin-labeling EPR data

Area of Science:

• Structural Biology
• Biophysics
• Computational Biology

Background:

• Many critical proteins are challenging to study using traditional methods like X-ray crystallography or Nuclear Magnetic Resonance (NMR) spectroscopy due to difficulties in crystallization or large size.
• Electron paramagnetic resonance (EPR) spectroscopy, particularly with site-directed spin labeling, offers a complementary technique for probing protein structure and dynamics.
• Translating EPR-derived distance information into usable geometric restraints for computational modeling is essential for determining protein structures.

Purpose of the Study:

• To develop and validate a computational approach for converting Electron Paramagnetic Resonance (EPR) data into atomic-detail protein models.
• To assess the accuracy and feasibility of this method for structure determination of proteins intractable by other techniques.
• To highlight the utility of distance restraints derived from EPR for de novo protein structure prediction.

Main Methods:

• Site-directed spin labeling was employed to introduce paramagnetic probes at specific protein sites.
• Distances between spin labels were calculated using a 'motion-on-a-cone' model, translated into distance ranges between beta carbons.
• A linear-correlation model was used to link spin-label accessibility to neighboring residues, integrated with the de novo structure prediction algorithm Rosetta.

Main Results:

• The approach successfully generated highly accurate, full-atom protein models with atomic resolution (1.0 Å and 2.6 Å) for T4-lysozyme and alphaA-crystallin.
• Distance restraints between residues distant in sequence but proximal in 3D space proved most critical for accurate structure determination.
• The methodology demonstrated the feasibility of obtaining precise structural information from EPR data.

Conclusions:

• The integration of EPR spectroscopy with computational modeling provides a powerful strategy for atomic-detail protein structure determination.
• This method is particularly valuable for proteins that are difficult to analyze using conventional structural biology techniques.
• The approach is extensible to other biophysical methods like fluorescence spectroscopy and mutational studies, broadening its applicability in structural biology.