We analyzed the intrinsic strength of distal and proximal FeN bonds and the stiffness of the axial NFeN bond angle in a series of cytochrome b5 proteins isolated from various species, including bacteria, animals, and humans. Ferric and ferrous oxidation states were considered. As assess- ment tool, we employed local vibrational stretching force constants ka(FeN) and bending force constants ka(NFeN) derived from our local mode theory. All calculations were conducted with the QM/MM method- ology. We found that transition from ferric to ferrous state makes the FeN axial bonds weaker, longer, less covalent, and less polar. Additionally, the axial NFeN bond angle becomes stiffer and less flexible. Local mode force constants turned out as far more sensitive to the protein environment than geometries; unraveling trends across di- verse protein groups and monitoring changes in the axial heme-framework caused a change between ferric and fer- rous oxidation states. These characteristics qualify them as perfect features for machine learning models predicting cytochrome b5 redox potentials, which currently rely more on geometric data and qualitative descriptors of the protein environment. The insights gained through our investigation also offer valuable guidance for strategically fine-tuning ar- tificial cytochrome b5 proteins and designing new, versatile variants.
Read full abstract