Macromolecules For Protein Binding Using Dendrimer and Graphene Oxide

Abstract

This research explored non-covalently functionalized dendrimers and graphene oxide as macromolecular ligands for protein binding. Their structures were functionalized with amino acids to modulate protein recognition and binding. Developing macro-ligands to target large protein binding surfaces is an effective strategy for inhibiting protein-protein interactions. While covalently functionalized macromolecules have shown potential, their synthesis and control over the positioning of binding groups can be challenging. Therefore, in the first area of research, dendrimers that had been functionalized using non-covalent methods were examined. This involved synthesizing neutral PAMAM dendrimers that could not bind to the surface of α-chymotrypsin (Chy) and a series of linear chains that could bind the protein. These chains were originally synthesised using a simple Boc protection strategy that involved a considerable amount of aqueous workup. However, the use of water resulted in very low yields and poor purity. This thesis describes a new CBz protection method that avoids the use of aqueous workup and was able to generate the required chains in excellent yield and high purity. Linear chains with either tyrosine or valine were prepared and up to 6 of these could be encapsulated within the G3.5 OH ended dendrimer, with a further 4 or 5 remaining dissolved in the bulk water. The resulting complexes were tested for their ability to bind the protein cytochrome-c. To facilitate a quantitative analysis, the quencher zinc tetra(4-hydroxyphenyl) porphyrin (Zn-THPP) was also encapsulated. Using a fluorescence titration technique a dissociation constant (Kd) of 33 nM was measured for the G3.5 dendrimer encapsulated with the tyrosine chains. In contrast, no binding could be detected using the dendrimer alone, or the dendrimer encapsulated with the valine chains. The next part of this project examined a similar “dynamic” approach to protein binding, using graphene oxides (GO) functionalized through both covalent and non-covalent approaches. This involved synthesising a series of anthracenes modified with various amino acids. These could be added to GO covalently, via a Diels-Alder reaction. In this case the functional groups were fixed in specific positions on the GO surface, and any biding would involve the protein moving around, which would limit the binding efficiency. The same anthracenes could also be added to the GO surface using non-covalent π-π interactions (by simply mixing the GO and functionalized anthracenes together in water). In this case the functional groups were free to move around the surface of the GO, allowing for maximum binding efficiency. As expected, the non-covalently functionalized anthracenes demonstrated enhanced binding to chymotrypsin through a series of inhibitory experiments. The research was extended to the study of GO systems functionalized with a mixture of functional groups (using covalent and non-covalent methods). Subsequent binding studies indicated that these mixed systems were more effective than corresponding systems functionalised with a single functional group, highlighting the benefits of combining amino acids for better binding affinity. Overall, the work described presents a proof of principle that addresses the difficulties in controlling functional group placement in precise 3D locations on the surface of protein binding ligands.