UC Berkeley

In the biological and medical fields, there is a heavy reliance on using genetic tools to manipulate proteins and investigate cellular functions. While these methods have proven hugely successful, they are limited in their ability to introduce novel functionality into biomolecules or interrogate the distinct chemical entities that control cellular responses. To address these restrictions, chemists have developed a set of chemoselective reactions with minimal cross-reactivity and toxicity in biological systems; termed the bioorthogonal reaction. The work in this dissertation has centered on the use of one such chemistry, oxime formation, which is obtained by the reaction of an aminooxy functionality with the carbonyl group of a ketone or aldehyde. I have utilized this chemistry to create unique protein assemblies (Chapter 1), site-specifically glycosylate proteins (Chapter 3,4), and introduce homogeneous glycans onto cell surfaces (Chapter 5). An emphasis of this work is the generation of homogeneous glycoconjugates that can be utilized to study the role of certain glycan structures in biology. This includes advancing methodology in cell surface modification for introducing multivalent glycomimics onto cellular membranes as well as utilizing protein modification techniques to create homogeneous glycoproteins.

Chapter 1 outlines my work on utilizing the genetically encoded consensus sequence termed the aldehyde tag to introduce a bioorthogonal carbonyl group onto protein backbones. The technique exploits the known function of the formylglycine generating enzyme to oxidize the thiol group of a cysteine residue within a specific peptide sequence to an aldehyde forming the non-natural amino acid, formylglycine. My work has focused on optimizing the conjugation of small molecules and large proteins to the formylglycine residue to create new protein scaffolds for studying protein function and for the creation of improved biotherapeutics.

The remaining chapters reside largely in the realm of glycobiology and thus Chapter 2 is devoted to providing a brief overview of the field specifically as it pertains to the role of glycans in modulating and defining medical treatment. Many therapeutics consist of glycoproteins or glycans themselves and though they have been used for decades, only recently have scientists been able to understand and engineer these treatments at the molecular level. As more tools become available to study how the location and chemical makeup of saccharides affect their function, we have begun to see a better appreciation of the significant part that glycans play in medicine. This will presumably lead to a surgence in novel types of treatment that rely solely on the function of specific glycan structures.

In the pursuit of creating new methods to define the role of glycan structures on protein function, Chapter 3 presents the use of the aldehyde tag technology to conjugate a wide range of chemically synthesized glycans to proteins. This approach relies on the development of an improved synthetic route of aminooxy glycans that utilizes chemical and enzymatic synthesis to create a small library of aminooxy sugars. I also show that these glycans can be conjugated site-specifically to the human growth hormone to create a novel set of chemically glycosylated therapeutic proteins. This system was taken a step further by placing the aldehyde tag on the capsid of human adenovirus as a conjugation point to attach glycan targeting moieties to the virus for gene therapy. The glycosylated protein should then retarget the virus to infect specific leukocyte lineages based on their unique expression of lectins that have specific glycan binding preferences. More work is needed to demonstrate the ability to direct the virus to infect nonnative cellular types such as leukocytes.

The aminooxy glycans generated thus far have been limited to small concise structures that one would normally find on the antennary regions of a full-length glycoconjugate. Chapter 4 outlines new work to create chemical mimics of full-length N-glycans with established dendrimer chemistry. We produced a novel glycodendrimer structure that consists of a piperidine-melamine core decorated with monosaccharides at the antennary and an aminooxy functionality for site-specific conjugation to aldehyde-tagged proteins. We have successfully conjugated a biantennary N-glycan mimic to an hGH construct and have begun work to create a IFN-β "glycodendriprotein" as a biotherapeutic surrogate.

The final chapter is the culmination of my thesis work, where we have utilized phospholipid containing glycopolymers to engineer cell surfaces with chemically defined glycan structures to study their role in regulating cellular activities. Chapter 5 focuses on the use of this approach to define the role of Siglec-7 on regulating the activity of Natural Killer (NK) cells to target and kill cancer cells. The overexpression of sialic acid has been well documented for many human cancers but few reports have provided the molecular details for how this provides a selective advantage to cancer cells to survive. We provide evidence that increasing sialosides on cancer cell surfaces can engage the inhibitory receptor Siglec-7 on NK cells and dampen the ability of NK cells to recognize and kill tumorigenic cells.