Date on Master's Thesis/Doctoral Dissertation


Document Type

Doctoral Dissertation

Degree Name

Ph. D.


Biochemistry and Molecular Biology

Degree Program

Biochemistry and Molecular Biology, PhD

Committee Chair

Sabo, T. Michael

Committee Co-Chair (if applicable)

Clem, Brian

Committee Member

Klinge, Carolyn

Committee Member

Blackburn, Jessica

Committee Member

Watson, Corey

Committee Member

Li, Ying

Author's Keywords

NMR; protein dynamics; structure-function; drug discovery; conformational exchange; high-power


The versatility of nuclear magnetic resonance (NMR) spectroscopy is apparent when presented with diverse applications to which it can contribute. Here, NMR is used i) as a screening/ validation tool for a drug discovery program targeting the Phosphatase of Regenerating Liver 3 (PRL3), ii) to characterize the conformational heterogeneity of p53 regulator, Murine Double Minute X (MDMX), and iii) to characterize the solution dynamics of guanosine monophosphate kinase (GMPK). Mounting evidence suggesting roles for PRL3 in oncogenesis and metastasis has catapulted it into prominence as a cancer drug target. Yet, despite significant efforts, there are no PRL3 small molecule inhibitors currently in clinical trials. This work combines screening of an FDA-approved drug panel and the identification of binders by protein-observed NMR. FDA-approved drugs salirasib and candesartan were identified as potent inhibitors in in vitro inhibition and migration assays while a weak inhibitor, olsalazine, was identified by NMR as the first small molecule inhibitor to directly bind PRL3. NMR was also used to validate the binding of additional compounds identified as experimental PRL3 inhibitors. Thienopyridone, a potent experimental inhibitor, did not show direct binding to PRL3 but instead inhibited phosphatase activity via redox mechanism. NMR also revealed that other experimental inhibitors did not engage PRL3. Thus, there remains a need to identify potent PRL3-directed inhibitors. Meanwhile, molecular modeling revealed a putative druggable site that has not been thoroughly explored before. The current study provides some scaffolds such as candesartan and particularly, olsalazine, the only binder identified, that could be the starting point of further drug discovery efforts, as well as a putative site that can be targeted in silico. MDMX, a negative regulator of p53, is another important therapeutic target in cancer, along with the homologous protein, MDM2. Inhibitors that block the MDM2-p53 interaction have been identified and despite similarities in the binding site of these homologous proteins, these inhibitors are ineffective against MDMX. It is hypothesized that the flexibility of MDMX contributes to this significant difference in response to inhibitors, despite comparable affinity to their endogenous target, p53. Examination of available inhibitor-bound structures of MDMX reveal a conserved pharmacophore but the structures adopt distinct conformations away from the binding site. This implies that global motions of the protein might contribute to molecular recognition. The conformational heterogeneity in MDMX was further confirmed by collecting residual dipolar couplings (RDCs). Further investigations on both MDMX and MDM2 are necessary to uncover whether the flexibility of MDMX contributes to the differential binding to inhibitors. Finally, NMR relaxation methods and state-of-the-art high-power Carr-Purcell-Meiboom Gill (CPMG) relaxation dispersion measurements, the first documented application on an enzyme, were used to characterize the solution dynamics of GMPK and the changes in dynamics upon GMP binding. Substrate binding resulted in restricting the amplitudes of motion for backbone amide bonds within the picosecond-nanosecond timescale. Meanwhile, CPMG showed dispersion in both in the absence and presence of GMP, such that substrate binding did not quench dynamics within the microsecond-millisecond timescale. Interestingly, more residues are observed to have dispersion in the bound form, some near the C-terminal of helix 3, which has previously been proposed to be involved in product release. Current studies show that substrate binding affect different timescales of protein motion. Future work shall follow how motions within different timescales are affected as GMPK processes its substrates – such as, for instance, binding of ATP analogs within the ATP binding site or simultaneous occupancy of both substrate binding pockets. This paves the way for a complete picture of the relationship of function and dynamics in the conformational enzymatic cycle of a bi-substrate enzyme using GMPK as a model. The current work illustrates some of the diverse applications of NMR on three unique systems that are also drug targets. Information collected here can be leveraged on future structure and dynamics studies as well as drug discovery efforts targeting any of these proteins.