Gradient bundle analysis of ketosteroid isomerase



Bond bundle analysis is used to investigate enzymatic catalysis in the KSI active site. We identify the unique bonding regions in each system and calculate the precise redistribution of electron density that accompanies either enhancement or inhibition of KSI catalytic activity. In two examples—using direct inspection of bond bundle regional properties, and using correlations between those properties and reaction barrier height—we arrive at similar conclusions, that cat- alytic enhancement results from promoting electron density redistribution between bonds within the KSI-docked substrate molecule in a way that closely resembles our mechanistic understanding of the forward catalyzed reaction. This catalyzing charge redistribution is prevalent in KSI systems catalyzed via electric fields or via amino acid mutation, and are thus suggestive of a significant catalytic role. Bond bundle analysis also generates a wealth of statistically useful bond properties, making it a strong candidate for machine learning approaches to computer- aided artificial enzyme design, and a promising tool for the discovery of relationships between the structure of the electron charge density and enzymatic catalytic activity.

Version notes

Revised manuscript and included SI


Thumbnail image of ksi_gba_v4 chemrxiv submission 2.pdf

Supplementary material

Thumbnail image of ksi_gba_SI_v2.pdf
Supplementary information
• 3D depictions of some of the regions discussed in the manuscript • Coordinates and energies of ab-initio calculations • Atomic basin, bond bundle, and bond wedge tabulated integration values with many more properties than are included in this manuscript • Single and multi-variable bar charts of simplified regional property correlations • Plots of fits of each region and property with reaction barrier energy • Correlations among regional properties

Supplementary weblinks

Source code for analysis used in manuscript
Source code for the gradient bundle decomposition algorithm and for the post-processing statistical analysis and plotting presented in the manuscript.