Development and application of a phage-based deep mutational scanning system for mapping the fine epitopes of HIV-1 and SARS-CoV-2 specific antibodies

Abstract

The rise of pandemic viruses such as HIV-1 and SARS-CoV-2 in modern times has highlighted the need for effective vaccines to combat these public health threats. In the case of HIV-1, development of an effective vaccine has been hampered by the virus’ ability to rapidly evolve and evade detection by the immune system. While SARS-CoV-2 is not nearly as diverse a virus as HIV-1 and effective vaccines against SARS-CoV-2 virus do exist, early reports indicate that viral variants that escape vaccine-elicited immunity are emerging and may reduce the protective effect of vaccines. For both viruses there is a clear need to develop vaccines that elicit immunity that remains effective in the face of mutations.Antibodies are a crucial component of immune response against HIV-1 and SARS-CoV-2. The protective effect of antibodies that develop after infection or immunization relies on their ability to either neutralize virus or direct the immune system to kill infected cells. Antibodies typically bind to the viral entry protein present on the surface of a virion or an infected cell, and mapping these binding sites is a key step needed to predict whether mutations on viral entry proteins could lead to loss of antibody function. A myriad of methods can be used to map antibody epitopes, and each have various pros and cons. Structural methods such as X-ray crystallography or cryogenic electron microscopy (cryo-EM) are the gold standard for mapping antibody binding sites but are low-throughput and often slow. Other methods such as alanine scanning or peptide arrays do not give a complete picture of the effect of all amino acid mutations within antibody binding sites. To accelerate mapping of antibody epitopes and potential escape mutations we developed Phage-DMS, a method of rapidly mapping the fine epitopes of monoclonal or polyclonal antibodies that combines phage display technology with deep mutational scanning (DMS). Phage-DMS involves the generation of a library of phage displaying peptides that correspond to either the wild-type sequence of a protein of interest or a sequence containing a single amino acid mutation. These libraries are incubated with antibody and then phage enriched by the antibody are sequenced, allowing us to determine epitope regions and sites sensitive to mutation. In this thesis, I first describe the development and validation of Phage-DMS using four HIV-1 monoclonal antibodies that have been previously well characterized in the literature (Chapter 2). We compared the sites of escape as determined by Phage-DMS for these HIV-1 antibodies and found that our method recapitulated, and in some cases refined, the known epitopes of these antibodies that had been mapped using other methods. Additionally, we found that the effect of individual mutations as determined by Phage-DMS was a relatively quantitative measurement of the loss of binding as determined by a low throughput assay (peptide ELISA). We then applied Phage-DMS to the study of antibodies in the plasma of people who had recovered from SARS-CoV-2 infection (Chapter 3). We found that convalescent antibodies commonly bound linear epitopes in the fusion peptide and heptad-repeat 2 regions of the Spike protein. Interestingly, examination of mutations that lead to escape from antibody binding revealed that there was person-to-person variation between the sites that were important for antibody binding. Additionally, we found that mutations within these conserved epitopes were not under selection in nature, suggesting either a lack of selective immune pressure or inability to evolve due to functional constraints at these sites. Finally, we compared antibody epitopes and escape mutations in the plasma and sera of SARS-CoV-2 infected and mRNA vaccinated individuals (Chapter 4). We found that antibodies from vaccinated or severely infected individuals bound to linear epitopes in both the S1 and S2 subunits of Spike, whereas antibodies from mildly infected individuals bound to linear epitopes in the S2 subunit alone. Antibody binding changed over time after vaccination within individuals, but was not affected by participant age, mRNA vaccine type, or vaccine dose. We also examined the effect of mutations within identified epitope regions and found that in many cases, vaccination induced a highly uniform escape profile across individuals. This finding has implications for the selection of escape variants on a global level. In summary, the following chapters detail the creation and validation of Phage-DMS as a method of mapping antibody epitopes and escape mutations and describe its application with samples derived from SARS-CoV-2 infected and vaccinated individuals. The speed at which the COVID-19 pandemic has unfolded has underscored the need for methods that can provide answers equally as rapidly. With the capacity to screen hundreds of antibody samples in parallel and map epitopes down to the amino acid level in a short span of time, Phage-DMS is a significant step forward towards this goal. The studies in this thesis leverage this approach to define epitopes and pathways of escape for two of the greatest pandemic viruses in modern times.