Date of Award

December 2017

Degree Type


Degree Name

Doctor of Philosophy



First Advisor

Nicholas R. Silvaggi

Committee Members

Alexander A. Arnold, Sonia Bardy, David N. Frick, Graham R. Moran, Xiaohua Peng


Antibiotic Biosynthesis, Aspartate-Less Regulators, Enduracididine, Oxidase, Pyridoxal-5’-Phosphate, Redox Sensor


The first part of my thesis is focused on a new family of two-component response regulator proteins: Aspartate-Less Regulators (ALR). They lack the catalytic aspartate residue required for the phosphorylation mechanism of typical two component response regulators. We are using biophysical tools to characterize two proteins with redox-sensitive ALR domains: repressor of iron transport regulator (RitR) from Streptococcus pneumonia R6 and diguanylate cyclase Q15Z91 from Pseudoalteromonas atalantica. The structure of inactive RitRC128S monomer showed that the ALR domain and the DNA-binding domain are linked by an α-helix that runs the length of the entire protein, with C128 near the C-terminal end. Bioinformatic analysis of all streptococcal RitR homologs showed that Cys128 is strictly conserved, suggesting that RitR may be a novel redox sensor. Hydrogen peroxide was used to oxidize the cysteine thiol group to determine the structure of the oxidized, dimeric form of RitR. Oxidation of C128 to the disulfide caused a conformational change that caused the DBD to release from the ALR domain. Surprisingly, the freed DBD was observed bound to the ALR domain of the other, disulfide-linked molecule of RitR, recapitulating almost exactly the structure of the inactive, monomeric protein. An extended dimeric conformation was found in the RitRL86A/V93A variant. It binds to the target DNA according to gel filtration and differential scanning fluorimetry. The crystal structure of the RitRL86A/V93A ALR domain showed an unprecedented conformational change for a response regulator protein, where helix α4 is disordered and the two protomers swap their α5 helices to form the dimer. Combined with the C128D mutant in vivo studies, it seems that oxidation of C128 is part of the activation mechanism, but there must be an additional factor that leads to dimerization of the ALR domains. The second ALR protein Q15Z91 has R61 replacing the phosphorylatable aspartate residue in the ALR domain. According to the structure of Q15Z91 with GTP and c-di-GMP, purified Q15Z91 is an activated but product-inhibited dimer. C142 is conserved in the same position as C128 in RitR, and substitution demonstrated that C142 residue is also a redox sensor that involved in Q15Z91 activity regulation.

The second part is a mechanistic enzymology project aimed at understanding the structure and mechanism of the novel pyridoxal-5’-phosphate (PLP)-dependent L-arginine hydroxylase/deaminase, MppP, from Streptomyces wadayamensis (SwMppP). SwMppP is predicted to be a type I/II aminotransferase based on primary sequence identity. However, NMR and ESI-MS results showed that SwMppP is not an aminotransferase, but rather a hydroxylase. The enzyme catalyzes the oxygen-dependent hydroxylation of L-arginine, forming 4-hydroxy-2-ketoarginine and the abortive side-product 2-ketoargine in a ratio of 1.7:1. This is exciting because SwMppP is the first PLP-dependent enzyme to react with oxygen in any context other than oxidative decarboxylation. The discovery of this new activity is especially surprising given that the tertiary structure of SwMppP is very similar to that of the prototypical aminotransferase, the E. coli aspartate aminotransferase (PDB entry 1ARS; RMSD of aligned Cα atoms is 3.7 Å). The major differences between the two enzymes are the disordered N terminus of SwMppP, and changes of a limited number of amino acids near the PLP cofactor. The N-terminal helix transitions from a disordered, random-coil state to a helical conformation covering the active site only if/when the substrate is bound. Specific roles of the un-conserved residues in the active site are being studied by mutagenesis. So far, most of the SwMppP mutants have lost the hydroxylase activity and only produce abortive side product 2-ketoarginine. Our mechanistic studies have revealed that formation of the fully oxidized (hydroxylated) product requires 2 equivalents of dioxygen, while formation of 2-ketoarginine requires only one equivalent of dioxygen. Interestingly, the hydroxyl group of 4-hydroxy-2-ketoarginine comes from H2O, not dioxygen. Mutagenesis, structural and kinetic studies were used to understand how the residues in the active site stabilize the quinonoid form of the L-arginine-PLP complex to promote the reaction with dioxygen. Our structural and kinetic characterization of the wild-type and variant forms of SwMppP have allowed us to propose a model where the oxygen incorporated in the hydroxy-arginine product is derived from water rather than from dioxygen. In addition, SwMppP exhibits very high substrate specificity. Either change on the substrate length or guanidine group would result in no binding affinity or little activity.

Included in

Biochemistry Commons