Date of Award

Summer 2021

Publication Type


Degree Name



Chemistry and Biochemistry

First Advisor

J. Gauld

Second Advisor

J. Trant

Third Advisor

S. Goodwin


Sulfur-containing biomolecules, Enzyme cysteine desulfurase, Maleamate amidohydrolase



Creative Commons License

Creative Commons Attribution 4.0 International License
This work is licensed under a Creative Commons Attribution 4.0 International License.


Sulfur-containing biomolecules display incredible functional diversity. Indeed, in addition to thiols and thioethers, S-nitrosothiols, 3,4-coordinate, sulfoxides, persulfides and now even polysulfides are commonly observed intermediates. Unfortunately, however, their biological synthesis and roles remain poorly understood. In addition, sulfur-containing species can access a broad range of oxidation states and thus can act as either an electrophile or nucleophile giving rise to an even more diverse set of sulfur-derived functional groups. However, these unique properties can lead to difficulties in characterizing such compounds experimentally and reinforces the need for computational studies to reliably predict their structural and energetic properties. In this dissertation, we have applied a broad range of computational methodologies to important sulfur-containing biomolecules to assess their applicability and to elucidate the chemistry of such species.

The first two chapters highlight some of the roles of sulfur and challenges associated with studying sulfur systems. We detail current methods in computational enzymology, discussing examples of their application and outline the next generation of approaches.

In Chapter 3, we demonstrate the importance of selecting a proper computational method for application to biorelevant sulfur/selenium-containing molecules. We conduct a detailed benchmark study and find that sulfur-sulfur bonds in particular are sensitive to changes in basis set. The ωB97XD/6-311G(2d,p) level of theory is found to be the most accurate and reliable for obtaining geometries and energetics.

We apply the knowledge gained from these benchmark studies in Chapter 4 to study biological persulfide/polysulfide formation pathways. Radical-containing charged intermediates are commonly proposed which can prove challenging for DFT methods to accurately describe. These species were found to be transient due to small contributions towards their stability from high-energy Rydberg states. Extending to an active site model showed that sulfur-sulfur bond formation is favoured with radical intermediates, particularly when electron transfer cofactors such as FAD are involved.

Chapter 5 examines the catalytic mechanism of maleamate amidohydrolase (NicF) which converts an amide substrate to a carboxylic acid with assistance of a cysteine nucleophile. The reaction took place in two stages: a stepwise pathway for release of ammonia by concerted attack of a water molecule to form the acid product. Stabilization of the oxyanion hole by a critical threonyl is thought to facilitate the reaction instead of a metal ion. Then, in Chapter 6, several approaches for selecting a suitable structure from molecular dynamics simulations for use in QM/MM calculations were examined. Building upon our mechanistic studies on NicF in Chapter 5, we show that poor structure selection can result in problematic surfaces due to the erroneous energetic contributions from the MM layer.

To help alleviate some of the user burden, Chapter 7 builds upon the conclusions of Chapter 6 to investigate the ability of artificial intelligence to guide the structure selection process. The program “Pose Selector” was created which uses agglomerative machine learning to select a set of structures for analysis. Evaluation of this approach was done on the NicF system and revealed that this approach can be used to select a structure, and that different catalytic pathways may be identified.

Finally, in Chapter 8, we apply these combined learnings to the enzyme cysteine desulfurase (SufS) and examine how the active site prepares for sulfur transfer to an active site cysteinyl to produce an enzyme-bound persulfide. Protonation states of the key active site residues were clarified with dynamic pKa calculations on these residues.

Throughout this dissertation we developed new protocols and demonstrated their application to sulfur-containing biomolecular systems. These insights can be extended beyond the realm of computational enzymology and may have impact in other fields such as the design of sulfa-drugs and calculating their properties using these approaches.