3 Conclusion

A knowledge gap in the binding environment for heme exists in the present literature. A high-throughput framework employing UCSF Chimera was constructed to process diverse sets of hemoproteins and output information about their binding pockets: amino acid frequencies and distances from heme, volume, surface area, angles. Data was gathered and predicted from representative and varied datasets for heme-b, heme-c, verdoheme, and siroheme, and their respective hemoproteins. R was used to analyze all data.

The results of this study suggest that binding pockets for hemoproteins have some requirements for binding that may have been overlooked to date. The data and their trends observed in this study demonstrate several phenomena.

First, the heme binding environments for heme-b, heme-c, and verdoheme contain high populations of nonpolar amino acids, suggesting nonpolar interactions may be of greater importance than previously thought to providing the necessary interactions to bind heme. The binding environment for siroheme, by contrast, is shown to be extremely enriched with polar amino acids, which is not very surprising; but this binding environment also still contains many nonpolar amino acids, reinforcing the idea that the polar interactions for all heme molecules, while necessary, may be insufficient for heme binding.

Second, most of the volume data for the binding pockets of all heme molecules centers around a value of 1200 A³. Surface areas of heme-b and verdoheme binding pockets are similar, approximately 10000 A², the surface area for heme-c is less, approximately 7500 A², and for siroheme is approximately 21000 A². These values may be useful in the design of artifical metalloenzymes.

Additionally, the seeming conservation of the volume size but the variety in pocket surface areas demonstrates that while the heme molecules may be of similar size and, besides attached groups, similar structure, the attached groups will significantly affect what interactions occur in the binding pockets, and therefore the shape and exposure to solvent in the binding pockets. Siroheme is strongly polar and its binding pocket has a large surface area and is therefore highly solvent exposed, as compared to heme-b which has more nonpolar groups that must be buried and therefore requiring a smaller surface area.

Finally, angular data were generated; but the phenomena observed, such as some residues having tight ranges of angles in relation to heme or the heme iron, cannot be interpreted as useful results, except perhaps for some protein engineering efforts that may have interest in the range or distribution of possible angles for a specific residue.

These results may be useful for the rational design of hemoproteins, with the importance of nonpolar interactions in particular likely of great interest. The framework constructed for this study can be applied to any list of PDBs and their respective ligands, thereby facilitating similar research for other proteins.