A study of viral entry proteins and receptors using deep mutagenesis

Mutagenesis allows for the analysis and manipulation of protein interactions and can be especially insightful when protein complexes are challenging for structure determination. Both targeted and random mutagenesis have their benefits but also have limitations, especially as tools for testing the phenotypes of large numbers of mutations. Deep mutational scanning can overcome the limitation of finding desired mutations by testing many thousands of mutations for a specific phenotype in a single experiment. In particular, a single site-saturation mutagenesis library is composed of sequence variants where each variant has one mutation, and the library cumulatively contains each possible amino acid substitution for all regions of interest. This Big Data method can be used to discover binding sites or to find desirable mutations. I have used this method to investigate the dimerization interfaces of chemokine receptors CXCR4 and CCR5. CXCR4 has five homodimeric conformations crystallized, but CCR5 has only had the monomeric form crystallized. Using bimolecular fluorescence complementation (BiFC), I found a mutationally less tolerant region on CXCR4 that matched the known dimerization interface and found a similarly mutationally intolerant region for CCR5, which could be a potential dimerization interface. In the process, I found that many mutationally tolerant regions for self-association were disruptive, leading to non-specific aggregation. Mutations that decreased self-associations had increased binding to ligand but decreased calcium signaling activity in response to an agonist. Overall, the data suggest multiple mechanisms are involved in receptor dimerization. I then assisted in the search for mutations in HIV-1 Env trimer that promoted a closed conformational state, as the naturally conformationally flexible viral spike uses this property to evade the immune system by 'directing' antibody production towards strain-specific epitopes rather than generating broadly neutralizing antibodies. In another scan sorting for the ability of HIV-1 Env to bind to co-receptor CD4, I found that the glycosylation site N262 is highly conserved for binding to CD4 by stabilizing a high affinity-conformation, and that mutations to that site resulted in an intrinsic binding deficit to CD4. Lastly, I assisted in investigating the binding interactions between HCMV and the receptor it uses for viral entry, PDGFRα. We found a mutation in PDFGRα that resulted in preferential binding to HCMV over its endogenous ligands PDGFs. Creating a soluble Fc-fused variant containing the mutation, we are developing an antiviral for HCMV. I found that murine PDGFRα is sufficiently homologous to human PDGFRα that the receptors and their ligands cross react, which will allow us to test in a mouse model in the future the tolerability and pharmacokinetics of the decoy receptor, with that data being relevant to humans.

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Biochemistry

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2023

Author dc:creator

• Gill, Kevin Sean

Contributors dc:contributor

• Procko, Erik
• Stadtmueller, Beth M
• Kranz, David M
• Brooke, Christopher B

• Deep Mutagenesis
• Chemokine
• Decoy Receptor
• Hiv
• Hcmv

Statement dc:rights

• Copyright 2023 Kevin Gill

Language dc:language

en, eng

Handle dc:identifier

https://hdl.handle.net/2142/120232

OAI record oai:identifier

https://www.ideals.illinois.edu/oai-pmh?verb=GetRecord&metadataPrefix=etdms&identifier=oai%3Awww.ideals.illinois.edu%3A2142%2F120232