Cysteine proteases (CPs) are enzymes with a nucleophilic thiol in their active sites. Inhibitors of cysteine proteases (ICPs) occur naturally in bacterial pathogens and some protozoa. In parasites, ICPs are often virulence factors, contributing to the formation and survival of amastigotes within host cells. These amastigotes have higher CP activity, therefore making both ICPs and CPs potential drug targets. Despite great genetic variability, ICPs contain highly conserved structural features, including a series of defined loops that play a significant role in binding CPs. Papain, a CP from Carica papaya, complexes with ICP from Leishmania mexicana. Although the individual 3-D structures of ICP and papain have been determined, as of this work, the structure of the papain-ICP complex has only been predicted, not solved. This research details the development of a technique for determining quaternary structure of the papain-ICP complex using paramagnetic relaxation enhancement NMR (PRE-NMR). A paramagnetic tag (MTSL) was added to various cysteine-mutants of ICP to measure distances to reductively 13C-methylated papain. The modification of ICP with MTSL was quantified using EPR, and the effects of labeling on the binding kinetics of papain and ICP were determined using SPR. 13C-methyl peak perturbations due to PRE were observed when papain was bound to spin-labeled E102C-ICP and K27C-ICP. Intermolecular distances were predicted using modeling software and a working model of the complex was created. Data from additional mutants will help to further determine complex structure and perfect the model.The penultimate chapter of this dissertation includes work towards the development of a method for studying protein-protein interactions using atomic force microscopy. Papain-ICP was used as a model system, with the intention to apply this method to the study of another system: filamentous actin (f-actin) and the actin-binding domain of abelson tyrosine-protein kinase (ABL2-FABD). The creation of nanopores on an AFM sensor chip surface was successful. ICP monomers bound selectively into the pores. Attempts to form the papain-ICP complex on the chip surface were unsuccessful, and future work is needed to perfect this method. The final chapter of this dissertation is a literature review outlining previous work in this area.

The observation that multi-electron activation of small molecule substrates occurs at polynuclear reaction sites, common to both metalloenzymes and heterogeneous catalysts, has led to the articulation of the polynuclear hypothesis - the idea that the expanded redox reservoir afforded by M-M interactions in polynuclear systems stabilizes multiple oxidation states and facilitates multi-electron transformations. Currently, examples of synthetic clusters that test the viability of polynuclear reaction sites towards effecting multi-electron activation of small molecule substrates are lacking. To test the polynuclear hypothesis, we targeted a system that embodies design elements common to metaloenzyme cofactors: polynuclear reaction sites that feature high-spin, coordinatively unsaturated metal centers. Metallation of tbsLH6 [tbsLH6 = 1,3,5-C6H9(NHC6H4-o-NHSiMe2tBu)3] yields high-spin trinuclear Fe(II) complex (tbsL)Fe3(THF). The filled anti-bonding orbitals in high-spin cluster (tbsL)Fe3(THF) renders ligand reorganization facile, which allows for a range of metal-substrate binding modes. The polynuclear site within the (tbsL)Fe3(THF) cluster cooperatively binds anionic donors and allows 2e- reduction of substrates including inorganic azide and hydrazines, yielding μ3-nitrido and μ3-imido products, respectively. The 4e- reductive N=N bond cleavage of azobenzene is also achieved in the presence of (tbsL)Fe3(THF) to yield Fe3 bis-imido complex ((tbsL)Fe3(μ3-NPh)(μ2-NPh), which has been structurally characterized. Cyclic voltammograms of a series of selected triiron imido and nitrido clusters suggest that oxidation states up to [Fe(IV)][Fe(III)]2 are electrochemically accessible. Addition of neutral pi-acidic molecules including tert-butylisonitrile (tBuNC) and carbon monoxode (CO) to trinuclear cluster (tbsL)Fe3(THF) led to the formation of a new series of coordination compounds, where binding to a single metal center is favored over cooperative substrate binding. Coordinated substrates are activated toward further reactivity, highlighted by the reductive coupling of isonitriles by (tbsL)Fe3(μ1-CNtBu)3 in the presence of phenylsilane. Finally, efforts to synthesize of a family of mixed Fe-Mn clusters that differ by single metal-site substitutions are presented. Substitutionally homogeneous (tbsL)Fe2Mn(THF) cluster is accessed from binuclear complex (tbsLH2 )Fe2. Attempts to synthesize similar Mn2Fe clusters results in isolation of a mixture of heterotrinuclear species. In conjunction with NMR, EPR, Mössbauer, and X-ray fluorescence spectroscopies, anomalous scattering measurements were critical for the unambiguous assignment of the metal substitution products that were synthesized. / Chemistry and Chemical Biology

Structural ion mobility-mass spectrometry (IM-MS) is often paired with computational conformational space studies to provide more insight into the experimental measurements. From the IM-MS or gas phase electrophoresis experiment, drift times are measured for the gas phase ions, which are converted into collision cross section (CCS) values. This CCS value can be interpreted as a rotationally averaged surface area of the gas phase ion. By computationally modeling these molecular ions, actual three-dimensional structures can be assigned to these CCS values. Within this work a combination of IM-MS techniques with computational modeling have been used to study polymer precursors, natural products, and metabolites. With the addition of tandem MS, fragmentation pathways were suggested for polyurethane hard block precursors. A conformational sampling protocol based on distance geometry methods, which selects random, pairwise interatomic distances to generate three-dimensional conformations, was developed to generate conformations of gas phase ions in a more time efficient manner as compared to current molecular dynamics approaches. This method was benchmarked on set of natural product molecules and then used to generate theoretical CCS ranges for metabolites. Distance geometry generates all possible three-dimensional conformations for the molecular ion and therefore the theoretical ranges based on these conformations can serve as a guide for developing new databases of experimental metabolite CCS values. Selecting appropriate computational strategies provides important structural details that help interpret experimental IM-MS results.

The mesocyclic trithioether 1,4,7-trithiacyclononane shows neighboring group participation upon oxidation as evidenced by the peak potential for oxidation in cyclic voltammetry, 0.9 V versus Ag/0.1 M AgNO₃, and by the formation of its monosulfoxide by controlled potential electrolysis in an overall two electron process. Pulse radiolysis of 1,4,7-trithiacyclononane leads to the formation of a two-center three-electron bond of moderate stability. Naptho [1,8-b,c]-1,5-dithiocin shows neighboring group participation upon electrochemical oxidation. A low peak potential is observed, 0.47 V versus Ag/0.1 M AgNO₃, in the cyclic voltammetry. Controlled potential electrolysis leads to formation of the monosulfoxide in a two electron process. The rigid methionine analog (±)-2-exo-amino-6-endo methylthiobicyclo [2.2.1] heptane-2-endo carboxylic acid exhibits neighboring group participation upon electrochemical oxidation. The degree of participation depends on the acid/base chemistry of the α-amino acid. The zwitterion shows the strongest neighboring group participation. Controlled potential electrolysis led to the formation of two diastereomeric sulfoxides in a two electron process. The diastereomer ratio was suggestive of participation by the carboxylate in the oxidation. The structures of the sulfoxides were confirmed by comparison with chemically prepared sulfoxides and sulfoxide amino acid derivatives. The stereochemistry was assigned by comparison with related compounds upon which x-ray crystallographic analysis had been performed. Single-electron pulse radiolytic oxidation of the amino acid led to the formation of an absorbing transient which was assigned to a two-center three-electron sulfur-carboxylate oxygen species. The yield of the species formed was pH dependent. At a pH above 3.5 decarboxylation becomes a major decomposition pathway in contrast to the electrochemical experiments where no decarboxylation was seen. The alcohol (±)-endo-2-hydroxy-6-endo-methylthiobicyclo [2.2.1] heptane shows significant neighboring group participation upon electrochemical oxidation in comparison with its exo-hydroxy isomer as indicated by a 350 mV difference in the peak potentials. Controlled potential electrolysis in a two electron process leads to formation of a mixture of diastereomeric alkoxysulfonium salts.

Hall's Bond Forming Initiation Theory, derived for [2 + 2] systems, was applied to [4 + 2] systems. Polymerizable electron-rich and electron-poor olefins were reacted to obtain evidence for reactive intermediates using polymerization as a trap. A diradical intermediate was successfully trapped in reactions of 1-methoxy-1,3-butadiene with methyl 2,2-dicyanoacrylate, dimethyl cyanofumarate, trimethyl ethylenetricarboxylate, maleic anhydride and acrylonitrile. Diradical intermediates were also trapped in reactions of isoprene, 1,3-butadiene, 2,3-dimethyl-1,3-butadiene, cis and trans-piperylene and 2,5-dimethyl-2,4-hexadiene with acrylonitrile. Copolymers were obtained in all cases. Copolymerization was accompanied by [4 + 2] cycloaddition. A zwitterionic intermediate was successfully trapped in the reactions of p-anisyl-1,3-butadiene and 1-phenyl-1,3-butadiene with various leaving groups in the β-position. Homopolymer of the arylbutadiene was obtained in all cases. Polymerization was initiated from a cationic species which arose from the elimination of the leaving group from a zwitterionic hexamethylene species. Polymerization was accompanied by [4 + 2] cycloaddition. Reactions of p-anisyl-1,3-butadiene and 1-phenyl-1,3-butadiene with methyl 2,2-dicyanoacrylate, dimethyl cyanofumarate, trimethyl ethylenetricarboxylate gave only the [4 + 2] cycloadduct. No polymer was formed in this reaction and no evidence for a reactive intermediate was obtained. The BFI theory was also applied to Group Transfer Polymerization reactions to determine whether the reaction mechanism involved a tetramethylene diradical intermediate. No styrene was incorporated into the methyl acrylate polymer, and so no evidence for a radical intermediate was obtained. LUMO energies calculated by AM1 were compared to reduction potential and UV data for tetrasubstituted and trisubstituted electrophilic olefins. The trends exhibited by the calculated LUMO energies agreed well with the experimental data, implying that calculations may be a reasonable method of predicting reactivity.

Scanning tunneling microscopy has been used to investigate the structural characteristics of thin film dimolybdenum tetracetate on single crystal graphite and molybdenum disulphide substrates. The results indicate that the observed surface structures of the films correspond to two-dimensional crystal faces of the three-dimensional X-ray crystal structure of dimolybdenum tetraacetate. The electronic structure of the films was investigated using photoelectron spectroscopy and correlated to the observed surface structures. Furthermore, thin films of dichromium tetraacetate as well as mixed films of the two dimers on graphite and molybdenum disulphide were also investigated by valence photoelectron spectroscopy.

The initial photochemical and photophysical events of the Bacteriorhodopsin (BR) photocycle are investigated using resonance Raman spectroscopy. The salt water bacterium, Halobacterium halobium, converts light into chemical energy via this cycle. Light induced isomerization of the all-trans retinal chromophore causes proton translocation across the lipid membrane containing the protein. Absorption experiments reveal red shifts in BR absorption on a picosecond time scale. Picosecond time-resolved resonance Raman spectroscopy (PTR³) provides a vibrational probe of these changes. PTR³ utilizes two tunable dye lasers in a pump-probe configuration. One initiates photochemistry while a second probes the chromophore. The vibrational spectrum of the K-590 intermediate present 50 ps after the initiation of the photocycle is obtained by PTR³ spectroscopy. The ability to separate photolytic excitation from the Raman probe facilitates the application of a quantitative model of the optical excitation process to time resolved vibrational measurements of K-590. These spectra are analyzed to find the isomerization state of retinal in K-590 by comparison with the resonance Raman spectra of model compounds. These resonance Raman results are compared to earlier measurements of the K intermediate. PTR³ spectra of K-590 present later in the photocycle are also obtained. These spectra remain unchanged over the period investigated (40 ps-26 ns). These results confirm that isomerization of the chromophore is one of the primary events following initiation of the photocycle. Changes in relative Raman intensities observed earlier than 40 ps are discussed with reference to the photophysics of the optical excitation process. PTR³ techniques are applied to antistokes Raman measurements of BR. The existence of a significant vibrationally excited population is revealed. Differences in the Raman band positions in the stokes and antistokes spectra demonstrate that several quanta of the higher frequency modes in the BR Raman spectrum are excited. These modes decay with a time constant of ≈7 ps. These observations suggest the retinal chromophore does not experience rapid uniform internal vibrational redistribution following the internal conversion producing the vibrationally excited species.

The enzyme ribulose-1,5-bisphosphate carboxylase/oxygenase (rubisco, E.C. 4.1.1.39) is required by all photosynthetic organisms. Recombinant rubisco, generated by heterologous expression of cloned genes in E. coli, was examined to determine enzyme structure/function relationships. Recombinant higher plant rubisco LSU was insoluble and nonfunctional. This was probably due to the absence of appropriate protein assembly apparatus in the heterologous host. Recombinant higher plant rubisco SSU was present in two forms, as soluble protein and as insoluble material. Recombinant Anacystis rubisco was present as soluble protein and as insoluble material. The soluble enzyme had carboxylase activity of 2.2 μmol CO₂ fixed/min/mg rubisco. The kinetically determined parameters with respect to RuBP were K(m)(RuBP) of 38 μM and V(max) of 3.5 μmol CO₂ fixed/min/mg rubisco. Insoluble SSU from either pea or Anacystis was manipulated into soluble, functional protein by a denaturation-renaturation procedure. Insoluble LSU from either spinach or Anacystis could not be manipulated into soluble protein by the denaturation-renaturation procedure. Recombinant Anacystis LSU alone, without SSU, assembled into the octameric core structure, was able to bind CABP, and could catalyze the carboxylation reaction at a rate of approximately 1% of holoenzyme. When SSU was added it assembled with LSU to form holoenzyme, stabilized the conformation of the active site region, and increased the rate of the carboxylation reaction. Mutant LSU with a 4 aa internal addition, near an active site Lys residue, was functional but carboxylase activity was reduced to 0.4 μmol CO₂ fixed/min/mg rubisco. Mutant LSU with large additions or extensions were expressed as insoluble material. Heterologous holoenzyme, composed of Anacystis LSU and the cyanobacterial-type SSU from Cyanophora or higher plant SSU from either pea or Mesembryanthemum, was enzymetically active. Heterologous holoenzymes had reduced carboxylase activity and higher K(m)(RuBP) compared to Anacystis rubisco. Enzyme activation was inhibited when non-activated heterologous holoenzyme was incubated with RuBP. This suggests that higher plant SSU plays a role in the tight binding of RuBP to unactivated enzyme. The function of a higher plant SSU specific region (HPIN) was investigated by examining chimeric SSU in chloroplast protein import assays. Precursor SSU proteins that contained HPIN were imported, processed, and assembled into holoenzyme in isolated chloroplasts. Precursor forms of chimeric SSU without HPIN were imported and processed but did not assemble with endogenous higher plant LSU. This resulted in the identification of HPIN as a putative assembly domain. Carboxylase assays of rubisco composed of Anacystis LSU and the chimeric SSU showed that the chimeric SSU were enzymatically functional.

Intermediates in the cyclotrimerization of alkynes have been isolated using the tantalum phenoxide reagents, Ta(DIPP)₂Cl₃(OEt₂) and Ta(DIPP)₃Cl₂(OEt₂) (DIPP = 2,6-diisopropylphenoxide). The degree of cyclization has been controlled by effecting either the sterics of the metal center or the alkyne itself. Reduction of the less congested bis-phenoxide complex, Ta(DIPP)₂Cl₃(OEt₂), by two electrons in the presence of progressively smaller alkynes allowed the selective synthesis of successively higher coordinated cyclooligomers (alkyne adducts, metalacyclopentadienes, and 7-metalanorbornadienes, respectively). This complex also catalytically cyclotrimerizes phenylacetylene upon reduction. Besides resembling proposed intermediates in the catalytic cyclotrimerization of alkynes by transition metals, the direct conversion of each of these cyclooligomers to the next higher or lower step in the proposed mechanism was demonstrated. The alkyne adduct reacted with additional alkyne to provide metallacyclopentadienes in a very regioselective fashion (i.e. (DIPP)₃Ta(PhC=CPh) reacted with Me₃CC=CH to provide (DIPP)₃Ta(CPh=CPhCH=CCME₃)). The metallacyclic complex, (DIPP)₂ClTa(CCMe₃=CHCH=CCMe₃) was shown to undergo an unprecendented dissociation into a bis-alkyne complex upon thermolysis before rearranging to the less-congested metallacycle, (DIPP)₂ClTa(CCMe₃=CHCCMe₃=CH). This complex then reacted with an additional equivalent of t-butylacetylene to provide the arene complex, (η⁶- C₆H₃ᵗBu₃)Ta(DIPP)₂Cl, or with Me₃CC=N to yield η²-(N,C)-NC₅H₂ᵗBu₃Ta(DIPP)₂Cl. The alkyne adduct, (DIPP)₃Ta(PhC=CPh), also undergoes regioselective cross-coupling reactions with benzaldehyde to provide (DIPP)₃Ta(CPh=CPhCH(Ph)O). This metallacyclic alkoxide reacted with an additional equivalent of benzaldehyde and then undergoes a hydride transfer to provide the Meerwein-Ponndorf-Verley/Oppenauer type redox intermediate, (DIPP)₃(PhCH₂O)Ta(η²-CPh=CPhCPh=O). Nitriles coordinate to (DIPP)₃Ta(PhC=CPh) to provide the complexes, (DIPP)₃Ta(PhC=CPh) (RC=N). Nitriles which contain α-hydrogen react further to provide the metallacycloenamine complexes (DIPP)₃Ta(CPh=CPhC(=CHR)NH).$ These complexes have been shown to arise through a metallacycloimine followed by an unusual intermolecular tautomerization, as inferred from deuterium labeling and crossover experiments.

Several series of complexes containing metal-metal and metal-heteroatom multiple bonds have been examined via gas-phase photoelectron spectroscopy to study the electronic structure effects resulting from simple ligand addition or substitution. Two types of complexes containing metal-metal multiple bonds form the basis of this study: complexes containing the M₂X₄P₄ core (where M = Mo, Re; X = halogen, Me; and P₄ = (PMe₃)₄ or [R₂P(CH₂)ₓPR₂]₂) and the dinuclear cyclopentadienyl metal carbonyl complexes [CpM(CO)ₓ]₂ (where M = Cr, Mo, W and x = 2,3). In the M₂X₄P₄ dimers two basic substitutions were examined. One involved changing the degree of rotation between the metal centers by varying the chelating phosphine ligand bridging the two centers. The other substitution involved varying the X ligand among Cl, Br, I, and Me on Mo₂ and Re₂ complexes with a monodentate phosphine, PMe₃. The synthesis and characterization of Re₂I₄ (PMe₃)₄, a previously unreported member of this series, is also described. For both of these substitutions, only very small changes were seen in the energy of the ionizations from orbitals involved in the metal-metal multiple bond. In the chelating phosphine complexes, changes in bandshape of the metal-based ionizations can be explained through the reduction in symmetry caused by placing the metals in a "staggered" configuration. For the monodentate phosphine complexes, the direction of the energy shifts which do occur indicate that changes in π-donor capability of the X ligand affect the metal-metal bonding to a greater extent than the changes in σ-donor capability. The effects of ligand addition to a metal-metal multiple bond (as in (CpM(CO)₂)₂) were examined, first by addition of one CO/metal, and then by insertion of a heteroatom into the multiple bond. In both cases, substantial energy shifts occur in related metal-based ionizations upon ligand addition. With CO addition, the M-M σ ionization is destabilized by ca. 1-2 eV, correlating with the longer M-M distances and increased reactivity in the (CpM(CO)₃)₂ systems. With the insertion of a chalcogen (S,Se) atom into the M-M triple bond of (CpCr(CO)₂)₂, the π-donor capability of the chalcogen results in M-heteroatom multiple bond character, though the PE spectra suggest that a full triple bond is not present.