The work in this dissertation progressed the research of structural discovery for two targets critical in the fight of infectious disease. Francisella lipoprotein 3 (Flpp3) is a virulent determinant of tularemia and was the first protein of study. The proteins soluble domain was studied using a hybrid modeling theory that used small angle X-ray scattering (SAXS) in combination with computation analysis to generate a SAXS-refined structure. The SAXS-refined structure closely resembled the NMR structure (PDB: 2MU4) which contains a hydrophobic cavity inside the protein that could be used for drug discovery purposes. The full-length domain of Flpp3 purified from the outer membrane of E. coli was also studied with a combination of biophysical characterization methods. Mass spectrometry and western blot analysis confirmed Flpp3 being translocated to the outer membrane, while SDS-PAGE confirmed the purity of Flpp3 in the monomeric form after size exclusion chromatography. Using Circular Dichroism (CD) the monomeric form of Flpp3 was shown to be almost fully refolded into having a primarily β-stranded secondary structure. This information advances the progress of both tularemia research and outer membrane protein research as no natively folded outer membrane protein structures have been solved for F. tularensis.The second protein worked on in this dissertation is the nonstructural protein 15 from SARS-CoV-2, also called NendoU. Nsp15 is an endoribonuclease associated with aiding the virus responsible for the current COVID-19 pandemic in evasion of the immune system. An inactive mutant of Nsp15 was studied with both negative stain electron microscopy and cryogenic electron microscopy (Cryo-EM) in the presence of RNA or without RNA present. The initial findings of negative stain electron microscopy of Nsp15 with and without RNA showed a difference in appearance. Negative stain analysis of Nsp15 is in the presence of a 5nt RNA sequence in low salt conditions shows a conformational change when compared to Nsp15 without RNA present. As well the presence of RNA appeared to shift the electron density in Cryo-EM studies of Nsp15. This work advances the research in how Nsp15 may bind and cleave RNA and aid in the evasion of the host cell immune system.

This work comprises a cumulative effort to provide analysis of proteins relevant to understanding and treating human disease. This dissertation focuses on two main protein complexes: the structure of the Chimp adenovirus Y25 capsid assembly, as used in the SARS-CoV-2 vaccine, Vaxzveria, and the Dbl family RhoGEF (guanosine exchange factor) Syx and its associated small G protein, RhoA. The course of research was influenced heavily by the onset of the Covid-19 pandemic and associated lockdown, which pushed anyone with the means to do meaningful research to shift priorities towards addressing the greatest public health crisis since the 1918 flu pandemic. Analysis of the Syx-RhoA complex for the purposes of structurally guided drug design was initially the focus of heavy optimization efforts to overcome the numerous challenges associated with expression, purification, and handling of this protein. By analyzing E. Coli derived protein new important knowledge was gained about this protein’s biophysical characteristics which contribute to its behavior and may inform drug design efforts. Expression in SF9 insect cells resulted in promising conditions for production of homogeneous and monodispersed protein. Homology modeling and molecular dynamics simulation of this protein support hypotheses about its interactions with both RhoA as well as regions of the cytoplasmic leaflet of the cell membrane. Structural characterization of ChAdOx1, the adenoviral vector used in the AstraZeneca Covid-19 vaccine, Vaxzveria resulted in the highest resolution adenovirus structure ever solved (3.07Å). Subsequent biochemical analysis and computational simulations of PF4 with the ChAdOx1 capsid reveal interactions with important implications for vaccine induced thrombocytic throbocytopenia syndrome, a disorder observed in approximately 0.000024% of patients who receive Vaxzveria.

Cryogenic Electron Microscopy (Cryo-EM) is a method that can be used for studying the structure of biological systems. Biological samples are frozen to cryogenic temperatures and embedded in a vitreous ice when they are imaged by electrons. Due to its ability to preserve biological specimens in near-native conditions, cryo-EM has a significant contribution to the field of structural biology.Single-particle cryo-EM technique was utilized to investigate the dynamical characteristics of various protein complexes such as the Nogo receptor complex, polymerase ζ (Polζ) in yeast and human integrin ⍺vβ8-pro-TGFβ1-GARP complex. Furthermore, I proposed a new method that can potentially improve the sample preparation for cryo-EM. The Nogo receptor complex was expressed using baculovirus expression system in sf9 insect cells and isolated for structural studies. Nogo receptor complex was found to have various stoichiometries and interactions between individual proteins. A structural investigation of the yeast apo polymerase ζ holoenzyme was also carried out. The apo Polζ displays a concerted motions associated with expansion of the Polζ DNA-binding channel upon DNA binding. Furthermore, a lysine residue that obstructs the DNA-binding channel in apo Polζ was found and suggested a gating mechanism. In addition, cryo-EM studies of the human integrin ⍺vβ8-pro-TGFβ1-GARP complex was conducted to assess its dynamic interactions. The 2D classifications showed the ⍺vβ8-pro-TGFβ1-GARP complex is highly flexible and required several sample preparation techniques such as crosslinking and graphene oxide coating to improve protein homogeneity on the EM grid. To overcome challenges within the cryo-EM technique such as particle adsorption on air-water interface, I have documented a collaborative work on the development and application of lipid monolayer sandwich on cryo-EM grid. Cryogenic electron tomography (cryo-ET) along with cryo-EM were used to study the characteristics of lipid monolayer sandwich as a potential protective layer for EM grid. The cryo-ET results demonstrated that the thickness of lipid monolayer is adequate for single-particle cryo-EM processing. Furthermore, there was no appearance of preferred orientations in cryo-EM and cryo-ET images. To establish that this method is actually beneficial, more data must be collected, and high-resolution structures of protein samples must be obtained using this methodology.

Structural-based drug discovery is becoming the essential tool for drug development withlower cost and higher efficiency compared to the conventional method. Knowledge of the three-dimensional structure of protein targets has the potential to accelerate the process for screening drug candidates. X-ray crystallography has proven to be the most used and indispensable technology in structural-based drug discovery. The provided comprehensive structural information about the interaction between the disease-related protein target and ligand can guide the chemical modification on the ligand to improve potency and selectivity. X-ray crystallography has been upgraded from traditional synchrotron to the third generation, which enabled the surge of the structural determination of macromolecular. The introduction of X-ray free electron laser further alleviated the uncertain and time-consuming crystal size optimization process and extenuated the radiation damage by “diffraction before destruction”. EV-D68 2A protease was proved to be an important pharmaceutical target for acute flaccid myelitis. This thesis reports the first atomic structure of the EV-D68 2A protease and the structuresof its two mutants, revealing it adopting N-terminal four-stranded sheets and C-terminal six-stranded ß-barrels structure, with a tightly bound zinc atom. These structures will guide the chemical modification on its inhibitor, Telaprevir. Integrin ⍺Mβ2 is an integrin with the α I-domain, related to many immunological functions including cell extravasation, phagocytosis, and immune synapse formation, so studying the molecular ligand-binding mechanism and activation mechanism of ⍺Mβ2 is of importance. This thesis uncovers the preliminary crystallization condition of ⍺Mβ2-I domain in complex with its ligand Pleiotrophin and the initial structural model. The structural model shows consistency with the previous hypothesis that the primary binding sites are metal iondependent adhesion sites on ⍺Mβ2-I domain and the thrombospondin type-1 repeat (TSR) domains of Pleiotrophin. Drug molecules with high potency and selectivity can be designed based on the reported structures of the EV-D68 2A protease and ⍺Mβ2-I domain in the future.

Ribulose-1,5-bisphosphate carboxylase/oxygenase enzyme (Rubisco) is responsible for the majority of carbon fixation and is also the least efficient enzyme on Earth. Rubisco assists 1,5-ribulose bisphosphate (RuBP) in binding CO2, however CO2 and oxygen have similar binding affinities to Rubisco, resulting in a low enzymatic efficiency. Rubisco activase (Rca) is an enzyme that removes inhibiting molecules from Rubisco’s active sites, promoting the Rubisco activity. The binding of Rubisco and Rca stimulates a high-rate of carbon fixation and lowers the overall CO2 concentration in the atmosphere. To study the interaction between the two complexes, Rubisco was extracted from baby spinach (Spinacia oleracea) and purified using anion-exchange chromatography and size-exclusion chromatography. Rca was designed to use a recombinant gene and overexpressed in Escherichia coli (E. coli). The purified proteins were verified using SDS-PAGE. The two proteins were assembled in vitro and the interaction of the protein complex was stabilized using glutaraldehyde cross-linking. The samples were then deposited on a carbon-coated electron microscopy (EM) grid, stained with uranyl formate, and observed under a transmission electron microscope (TEM). The ultimate goal is to image the specimen and reconstruct the structure of the protein complex at high resolution.

Osteocalcin (Oc) is the most abundant non-collagen protein found in the bone, but its precise function is still not completely understood. Three glutamic acid (Glu) residues within its sequence are sites for vitamin K-dependent post-translational modification, replacing a hydrogen with a carboxylate located at the γ-carbon position, converting these to γ-carboxyglutamic acid (Gla) residues. This modification confers increased binding of Oc to Ca2+ and hydroxyapatite matrix. Presented here, novel metal binding partners Mn2+, Fe3+, and Cr3+ of human Oc were determined, while the previously identified binders to (generally) non-human Oc, Ca2+, Mg2+, Pb2+ and Al3+ were validated as binders to human Oc by direct infusion mass spectrometry with all metals binding with higher affinity to the post-translationally modified form (Gla-Oc) compared to the unmodified form (Glu-Oc). Oc was also found to form pentamer (Gla-Oc) and pentamer and tetramer (Glu-Oc) homomeric self-assemblies in the absence of NaCl, which disassembled to monomers in the presence of near physiological Na+ concentrations. Additionally, Oc was found to form filamentous structures in vitro by negative stain TEM in the presence of increased Ca2+ titrations in a Gla- and pH-dependent manner. Finally, by combining circular dichroism spectroscopy to determine the fraction of Gla-Oc bound, and inductively-coupled plasma mass spectrometry to quantify total Al concentrations, the data were fit to a single-site binding model and the equilibrium dissociation constant for Al3+ binding to human Gla-Oc was determined (Kd = 1.0 ± 0.12 nM). Including citrate, a known competitive binder of Al3+, maintained Al in solution and enabled calculation of free Al3+ concentrations using a Matlab script to solve the complex set of linear equations. To further improve Al solubility limits, the pH of the system was lowered to 4.5, the pH during bone resorption. Complementary binding experiments with Glu-Oc were not possible due to the observed precipitation of Glu-Oc at pH 4.5, although qualitatively if Glu-Oc binds Al3+, it is with much lower affinity compared to Gla-Oc. Taken together, the results presented here further support the importance of post-translational modification, and thus adequate nutritional intake of vitamin K, on the binding and self-assembly properties of human Oc.

Cyanobacteria contribute to more than a quarter of the primary carbon fixation worldwide. They have evolved a CO2 concentrating mechanism (CCM) to enhance photosynthesis because inorganic carbon species are limited in the aqueous environment. Bicarbonate transporters SbtA and BicA are active components of CCM, and the determination of their structures is important to investigate the bicarbonate transport mechanisms. E. coli was selected as the expression host for these bicarbonate transporters, and optimization of expression and protein purification conditions was performed. Single particle electron cryomicroscopy (cryo-EM) or protein crystallography was carried out for each transporter. In this work, SbtA, BicA and SbtB, a regulator protein of SbtA, were heterologously expressed in E. coli and purified for structural studies. SbtB was highly expressed and two different crystal structures of SbtB were resolved at 2.01 Å and 1.8 Å, showing a trimer and dimer in the asymmetric unit, respectively. The yields of SbtA and BicA after purification reached 0.1 ± 0.04 and 6.5 ± 1.0 mg per liter culture, respectively. Single particle analysis showed a trimeric conformation of purified SbtA and promising interaction between SbtA and SbtB, where the bound SbtB was also possibly trimeric. For some crystallization experiments of these transporters, lipidic cubic phase (LCP) was used. In the case of LCP, often times the crystals grown are generally too tiny to withstand radiation damage from the X-ray beam during an X-ray diffraction experiment. As an alternative approach for this research, the microcrystal electron diffraction (MicroED) method was applied to the LCP-laden crystals because it is a powerful cryo-EM method for high-resolution structure determination from protein microcrystals. The new technique is termed as LCP-MicroED, however, prior to applying LCP-MicroED to the bicarbonate transporters, methods needed to be developed for LCP-MicroED. Therefore the model protein Proteinase K was used and its structure was determined to 2.0 Å by MicroED. Additionally, electron diffraction data from cholesterol and human A2A adenosine receptor crystals were collected at 1.0 Å and 4.5 Å using LCP-MicroED, respectively. Other applications of MicroED to different samples are also discussed.

G protein-coupled receptors (GPCRs) are a large family of proteins involved in the cell signaling and regulation of many biological and pathological processes in the human body. To fully understand their functions, various approaches are needed. This work combines several techniques to advance the study of GPCRs with the overarching goal of pursuing X-ray crystallization using lipidic cubic phase (LCP). In meso, or LCP crystallization method involves imbedding the GPCR into a lipid membrane-mimetic material which spontaneously forms when monoacylglycerols (MAGs) are mixed at the correct hydration level and temperature. It provides a stable environment for GPCRs and has been established as the most common method to resolve structural details of GPCRs (Chapter 2). Yet, before crystallization, GPCRs need to be put through several rounds of optimization of the construct design, including truncation of N- and C- termini, fusing different soluble proteins, and mutating the receptor (Chapter 3). Other methods were also used to gain structural insights into GPCR interactions, such as coarse-grained molecular dynamic simulations, which showed the specific regions of interactions with cholesterol molecules imbedded in the membranes (Chapter 4). This study demonstrated β2-adrenergic receptor (β2AR), a GPCR, as a model of a cholesterol-sensitive receptor. Mutations were made to test the effect of removing specific residues of interest on cholesterol stabilization through the LCP-Tm assay, producing results that align with the simulation data. Finally, the goal of the last study is to provide a guide to identify which host lipids form stable LCP phases for different applications (Chapter 5). Small angle X-ray scattering is used to identify phases in hundreds of different precipitant conditions in the search of suitable host lipid for LCP studies. The results present a systematic overview of the compatibility of common MAGs by screening them against different precipitant solutions including varying salts and polyethylene glycol (PEG) concentrations, different PEG sizes, the presence of detergent or protein in the sample, and the addition of cholesterol. Together, these studies present a variety of methods to advance the structural studies of GPCRs using LCP

Since its conception over a century ago, X-ray crystallography (XRC) has become the most successful method used to elucidate the structures and functions of biological molecules at atomic resolution. The extensive use of XRC has led to meaningful discoveries across many scientific fields, notably its contributions to rational drug design. Traditional drug discovery relies on the use of trial-and-error based approaches in cellular and animal models of disease to identify chemical probes that elicit desirable therapeutic effects based off changes in phenotype. However, this approach lacks critical information in regards to the biological target in which the compound interacts with. In contrast, the use of rational drug design presents the opportunity to identify chemical probes that target specific protein targets of known medical importance and study their interactions using three dimensional structures that can be used to suggest new drug candidates. The main focus of my research presented in this dissertation aims to utilize XRC to discover novel therapeutics. In this work, I begin by describing the use of structure-based drug discovery for the rational design of hydrocarbon-stapled peptides that block Focal Adhesion Kinase (FAK) scaffolding in cancer (Chapter 2). FAK is an intracellular tyrosine kinase that has been linked to many cancers through its interaction with Paxillin LD motifs as it relates to tumor growth, invasion, metastasis, and suppression of apoptosis. The results of this study demonstrate the effectiveness hydrocarbon-stapling has on the native Paxillin LD2 sequence with ~50 fold greater binding affinity by surface plasmon resonance (SPR) that can be explained by the unique structural interactions observed by XRC. Next, I present a series of methods which lays the foundations for the discovery of novel anti-bacterial drugs that target 3-Deoxy-D-manno-octulosonate-8-phosphate (KDO8P) Synthase, a critical enzyme in the biosynthesis of gram-negative lipopolysaccharides (Chapter 3).

The field of biomedical research relies on the knowledge of binding interactions between various proteins of interest to create novel molecular targets for therapeutic purposes. While many of these interactions remain a mystery, knowledge of these properties and interactions could have significant medical applications in terms of understanding cell signaling and immunological defenses. Furthermore, there is evidence that machine learning and peptide microarrays can be used to make reliable predictions of where proteins could interact with each other without the definitive knowledge of the interactions. In this case, a neural network was used to predict the unknown binding interactions of TNFR2 onto LT-ɑ and TRAF2, and PD-L1 onto CD80, based off of the binding data from a sampling of protein-peptide interactions on a microarray. The accuracy and reliability of these predictions would rely on future research to confirm the interactions of these proteins, but the knowledge from these methods and predictions could have a future impact with regards to rational and structure-based drug design.