The synthesis of five chiral DBFOX (dibenzofuran-oxazoline) ligands with either aryl or benzyl substituents will be presented. The requisite amino alcohols were obtained with high enantioselectivity either commercially (DBFOX/Bn), via Sharpless asymmetric aminohydroxylation (DBFOX/Nap, DBFOX/t-BuPh, DBFOX/Pip), or via phase-transfer catalyzed asymmetric alkylation (DBFOX/MeNap). These ligands, complexed with Mg(NTf2)2, were used as Lewis acid promoters of enantioselective radical conjugate additions to α/β-unsaturated nitro-amides/esters. A summary of these results is presented and discussed. These findings led us to believe that our initial binding model between metal, ligand, and substrate was flawed. Thus, we figured that if we started with a functionality known to bind to both nitro groups and carbonyls, and then introduced a chiral element for control, we may be able to improve the beta-carbon enantioselectivity. We have tried to accomplish this via hydrogen-bonding ligands (ureas and thioureas). Initial studies on achiral versions of this concept are discussed.

Celogentin C and theonellamide F are a class of natural products that have potential antimitotic behavior. They both contain interesting bicyclic structures with unusual linkages within a central moiety. Celogentin C's highly functionalized tryptophan moiety has two unusual linkages, a β-substituted Leu connection to the C6 of the indole structure that makes up the left-hand ring, and a τ-N connection of the imidazole to the C2 of the indole constructing the right-hand ring. This right-hand ring connection was solved via a novel oxidative coupling procedure developed in our group and the left-hand ring was initially constructed via a radical conjugated addition of an isopropyl group. Due to stereoselective concerns, our group explored hydrogen bond donors as potential catalyst candidates. Unfortunately, there were challenges in limiting the background reaction and obtaining reproducible results. We then designed an alternative route to solve this left-hand ring connection which would have utilized MacMillan asymmetric hydrogenation and α-chlorination procedures. Further work towards a second generation synthesis of the β-Leu-(C6)Trp connection was halted with the publication of two formal syntheses of celogentin C. Theonellamide F contains a τ-L-histidino-D-alanine (τ-HAL) bridging unit that separates the left-and right-hand rings. Previous efforts in the synthesis of this natural product were hindered due to an inefficient regioselective synthesis of τ-HAL. Our proposed synthesis of τ-HAL began with commercially available L- and D-Ser methyl esters which were then chemically transformed and coupled to one another to create a bis-amino subunit. Further preparations afforded us with an important cyclic intermediate which should readily lead to the first regioselective synthesis of a τ-HAL.

Project I Celogentin C, isolated by Kobayashi from the seeds of Celosia argentea in 2001 exhibits strong inhibitory behavior towards polymerization of tubulin – globular protein, which plays crucial role during the cell division cycle. We have attempted synthesis of the left-hand ring of celogentin C via the intramolecular alkylation and the Knoevenagel condensation strategies. Utilizing synthetic methodologies in the field of nitroacetyl chemistry, developed by Kornblum and Rajappa we have successfully prepared the intermolecular Knoevenagel condensation product – the late-stage precursor to the left-hand ring of celogentin C. Synthesis of this key intermediate subsequently led to the preparation of the left-hand ring and the total synthesis of celogentin C by other members of Castle's group. Project II Thioviridamide is the potent apoptosis inducer isolated by Hayakawa from the bacterial broth of Streptomyces olivoviridis. Unusual structural features of the thioviridamide macrocycle contain two novel amino acids, never before encountered in a natural product - S-(2-aminovinyl)cysteine (avCys) and β-hydroxy-N1,N3-dimethylhistidinium (hdmHis). No stereochemical information except for the Z-configuration of the β-thioenamide linkage was reported in the literature. We have performed a computational study to predict the thioviridamide stereochemistry. Initial populations of conformers for the likely candidate structures were produced using OPLS-AA force field. Prediction of the NMR properties was accomplished at the mPW1PW91/6-311+G(2d,p) level of theory with the polarizable continuum model of salvation. Utilizing Boltzmann averaging and statistical analysis we have determined that the only possible cases of stereochemical inversion occur at the sites of the two novel amino acids. Project III Model studies towards the synthesis of the β-thioenamide subunit of thioviridamide were performed. The radical addition reaction of thiyl radicals to ynamides produces Z- (kinetic products) or E- β-thioenamides (thermodynamic products) depending on the reaction conditions. Two distinct sets of reaction conditions allowing kinetic or thermodynamic control of β-thioenamide formation were developed. Synthesis of the model β-thioenamide subunit of thioviridamide was attempted.

Project I. Cinchona alkaloid-derived quaternary ammonium salts have been successfully used as phase-transfer catalysts, particularly in asymmetric alkylations. Our group applied this type of catalyst in the synthesis of β-hydroxy α-amino acids via aldol reactions and discovered that the Park-Jew catalyst afforded good yields and good enantiomeric excess of the syn diasteromers, but negligible diastereoselectivity. This project was therefore focused on the synthesis of novel cinchonidine-derived catalysts with the Park-Jew catalyst as the lead structure. The C3 position of cinchonidine nucleus was modified to achieve dimers and catalysts possessing electron-deficient alkyne and alkene moieties. Synthesized catalysts were tested in the asymmetric aldol reactions, with some of them yielding improvements relative to the Park-Jew catalyst. Project II. Celogentin C is a natural product that was isolated from the seeds of Celosia argentea by Kobayashi in 2001. It is the most potent inhibitor of the polymerization of tubulin from among the celogentin family. The novel bicyclic octapeptide structure contains unusual linkages between leucine β-carbon and indole C-6 of tryptophan and between tryptophan indole C-2 and imidazole N-1 of histidine. The project culminated in the first total synthesis of celogentin C. Reaction conditions were developed by synthesizing the left-hand ring and the right-hand ring separately, and the total synthesis was accomplished via a left to right strategy. Key transformations in the construction included intermolecular Knoevenagel condensation, radical conjugate addition, macrolactamization, and oxidative coupling.