Solution nuclear magnetic resonance (NMR) spectroscopy is a valuable analytical technique that is nondestructive, highly reproducible, and relatively quick to identify and quantify many chemical compounds. Quantitative NMR is a technique commonly used in many medical applications such as drug analysis, metabolomics, and protein-nanoparticle (P-NP) interactions. The most common technique used is the proton (1H) NMR experiment. The 1H NMR analysis provides a quick snapshot of the interested compounds in solution. However, as the compounds become more complex the spectrum becomes overpopulated. This dissertation focuses on various quantitative NMR techniques applied to metabolic and protein competition studies. Specifically, we investigated the effect of biochar on Escherichia coli (E. coli) growth to provide insight on how the metabolic pathways were influenced with the addition of biochar in the RPMI media. A 1H NMR spectrum was recorded at various time points to monitor the metabolic changes over time as E. coli grew in the presence and absence of biochar. The spectra were compared to an in-house metabolite library to identify and quantify the metabolic changes in E. coli. To enhance our metabolic library analysis, we utilized a pure shift analysis attached to the TOCSY pulse program to deconvolute spin systems by using a second dimension for analysis. DIPSI-PSYCHE TOCSY was applied to investigate a metabolite mixture sample and Streptococcus pneumoniae (S. pneumoniae) extracellular metabolites to better resolve the spin systems that significantly overlap each other in the 1H NMR spectra. Our novel approach suggests that adding a pure shift to the TOCSY pulse program is extremely beneficial to investigate various metabolic profiles. Finally, we investigated the protein competition to the AuNP surfaces using a 2D 1H-15N HSQC pulse program. Specifically, we used 1H-15N HSQC technique to quantify the binding capacity for each protein to the AuNP surface before we investigated the competition of two proteins, GB3-Ubq (model protein mixture) or AM-R2ab (biofilm forming protein mixture) to the surface. We also employed a model to study the kinetics of the protein competition to the surface. Our model suggests that GB3-Ubq does not specifically behave kinetically but AM-R2ab is strictly kinetically controlled.

Doctor of Philosophy / Department of Chemical Engineering / Peter H. Pfromm / One area of considerable importance in modern biotechnology is the preparation of highly active and selective enzyme based biocatalysts for applications in organic solvents. A major challenge is posed by the tendency of enzymes to cluster when suspended in organic solvents. Because the clusters obstruct the transport of substrates to the active site of the enzyme, the observed activity is often severely reduced. Over the past two decades, many strategies have been proposed to mitigate this problem. We have tackled this major hurdle by devising an immobilization strategy that utilizes fumed silica as carrier for the enzyme molecules. Fumed silica is a non-porous nanoparticulated fractal aggregate with unique absorptive properties. The enzyme/fumed silica preparation is formed in two steps. The buffered enzyme molecules are physically adsorbed on the fumed silica and then lyophilized. This protocol was shown to be successful with two enzymes of industrial relevance, Candida antarctica Lipase B (CALB) and subtilisin Carlsberg. The maximum observed catalytic activity in hexane reached or even exceeded commercial immobilizates and nonbuffer salt based preparations. The results demonstrated that catalytic activity has an intricate relationship with the nominal surface coverage (%SC) of the support by the enzyme molecules. s. Carlsberg exhibited an ever increasing activity as more surface area was provided per enzyme molecule. The activity leveled off when a sparse surface population was reached. CALB showed a maximum in catalytic activity at an intermediate surface coverage with steep decreases at both lower and higher surface coverage. It was shown that this maximum results from the presence of three distinct surface loading regimes after lyophilization: 1. a low surface coverage where opportunities for multi-attachment to the surface likely lead to detrimental conformational changes, 2. an intermediate surface coverage where interactions with neighboring proteins and the surface help to maintain a higher population of catalytically competent enzyme molecules, and 3. a multi-layer coverage where mass transfer limitations lead to a decrease in the apparent catalytic activity. Conformational stability analyses with both fluorescence and CD spectroscopy showed evidence that these regimes are most likely formed during the adsorption step of our protocol. A low conformational stability region was detected at low surface coverage while adsorbates with highly stable enzyme ensembles were observed at high surface coverage. Secondary structural analysis of the lyophilized nanobiocatalysts with FTIR confirmed a substantial decrease in the alpha-helical components at low surface coverage. In summary, the work presented here traces the phenomenological observation of the catalytic behavior of a nanobiocatalyst to molecular-level: enzyme-enzyme and enzyme-support interactions, which are specific to the intricate properties of the enzyme molecules.

This thesis describes different aspects of protein interactions. Initially the function of peptides and their conjugates with small molecule inhibitors on the surface of Human Carbonic Anhydrase isoenzyme II (HCAII) is evaluated. The affinities for HCAII of the flexible, synthetic helix-loop-helix motif conjugated with a series of spacered inhibitors were measured by fluorescence spectroscopy and found in the best cases to be in the low nM range. Dissociation constants show considerable dependence on linker length and vary from 3000 nM for the shortest spacer to 40 nM for the longest with a minimum of 5 nM for a spacer with an intermediate length. A rationale for binding differences based on cooperativity is presented and supported by affinities as determined by fluorescence spectroscopy. Heteronuclear Single Quantum Correlation Nuclear Magnetic Resonance (HSQC) spectroscopic experiments with 15N-labeled HCAII were used for the determination of the site of interaction. The influence of peptide charge and hydrophobicity was evaluated by surface plasmon resonance experiments. Hydrophobic sidechain branching and, more pronounced, peptide charge was demonstrated to modulate peptide – HCAII binding interactions in a cooperative manner, with affinities spanning almost two orders of magnitude. Detailed synthesis of small molecule inhibitors in a general lead discovery library as well as a targeted library for inhibition of α-thrombin is described. For the lead discovery library 160 members emanate from two N4-aryl-piperazine-2-carboxylic acid scaffolds derivatized in two dimensions employing a combinatorial approach on solid support. The targeted library was based on peptidomimetics of the D-Phe-Pro-Arg showing the scaffolds cyclopropane-1R,2R-dicarboxylic acid and (4-amino-3-oxo-morpholin-2-yl)- acetic acid as proline isosters. Employing 4-aminomethyl-benzamidine as arginine mimic and different hydrophobic amines and electrophiles as D-phenylalanine mimics resulted in 34 compounds showing IC50 values for α-thrombin ranging more than three orders of magnitude with the best inhibitor showing an IC50 of 130 nM. Interestingly, the best inhibitors showed reversed stereochemistry in comparison with a previously reported series employing a 3-oxo-morpholin-2-yl-acetic acid scaffold.