Synthesis of Non-Natural Cyclic Di-Nucleotides for the Investigation of Bacterial Signaling Pathways

Abstract

Humans navigate the world and interact with others through a complex series of communicative tools. We experience both internal and external stimuli, such as pangs of hunger or pain from an injury, and both verbal and nonverbal language. Bacteria also possess the ability to communicate, albeit in more discreet, yet no less complex ways. Bacteria rely on an incredibly diverse signaling system of triggers and responses in order to survive and to thrive. While we perceive language with our eyes and ears, bacteria employ a system of small molecules to relay both intra- and extracellular messages. They utilize this ability, known as quorum sensing to "talk" to their neighbors, express otherwise latent genetic characteristics, and to defend themselves against enemies. It has been suggested that this internal and external activity is linked, however, little is known about their interplay. This family of molecules, the cyclic di-nucleotides, which includes c-di-GMP and c-di-AMP, are critical to regulating bacterial processes such as motility, glucose remediation, and cell wall homeostasis. Their importance has spurred numerous investigations into their mechanism of action. Although found in very low concentrations within cells, they are capable of regulating a multitude of processes due to their ability to adopt variable conformations. To date, analog design by other groups has focused on the modification of the innate phosphate moiety as well as various substitutions or deletions at the 2'-position on the ribofuranose ring. However, these analogs have not been water soluble, limiting them to in vitro investigations only. We propose that by replacing the phosphate linkage entirely we can increase water solubility and have pursued a divergent total synthesis of various cyclic di-nucleotides featuring biomimetic linkages. Herein we address the methods we explored to optimize the synthesis of our three monomers, coupling strategies employed, the novel application of a Staudinger ligation to afford our abasic macrocycles and finally our progress towards implementing a bis-glycosylation strategy to install the desired nucleobase. We are able to efficiently provide large amounts of a di-amino, azide methyl ester, and N,O-substituted furanose monomers in no more than six steps from a common intermediate. These monomers are coupled and cyclized to form our four scaffolds, amide, carbamate, squaramide, and urea. Finally, we have begun to successfully implement our Brønsted acid mediated glycosylation strategy and understand its limitations. It is our goal to develop a general method to afford a diverse array of conformationally unique and water soluble cyclic di-nucleotide analogs with which to probe these essential bacterial signaling pathways.