Dissertation
Sequence-based prediction of virus phenotype, evolutionary potential, and selective pressures.
University of Iowa
Doctor of Philosophy (PhD), University of Iowa
Spring 2024
DOI: 10.25820/etd.007731
Abstract
The HIV-1 and SARS-CoV-2 pandemics have driven significant advancements in therapeutic interventions, including antiretroviral medications and vaccination campaigns. Despite these efforts, HIV-1 remains incurable, and the emergence of new variants of SARS-CoV-2 continues to challenge vaccine efficacy. The high mutation rates of these viruses contribute to their ongoing diversity, presenting clinical challenges that necessitate a deeper understanding of viral evolution. To address these challenges, we leveraged sequence information to investigate the selective pressures driving virus evolution, the emergence of clinically relevant mutations, and the phenotypic diversity of these viruses.
We examined the historical changes in amino acid frequencies at different positions of the HIV-1 envelope glycoproteins (Envs) among subtype B isolates. The Fusion Peptide Proximal Region (FPPR) of Env showed particularly interesting evolutionary patterns at some sites, characterized by a rapid increase in frequency of new variants during the early years of the AIDS pandemic, followed by a steady state frequency that was maintained thereafter. In vitro tests of the emergent FPPR variants were conducted to determine their effects on viral fitness and on Env conformation. We found that mutations at some sites induced exposure of otherwise-cryptic epitopes at the base of the Env trimer. Such epitopes are targeted by antibodies commonly elicited during infection, which we suggest apply weak selection pressure on Env. This study demonstrated that the balance between fitness of each variant and its sensitivity to commonly elicited antibodies determines its set point frequency in the population and the historical pattern of change.
My second project examined the evolutionary dynamics of SARS-CoV-2 lineages in the population, focusing on mutations in the spike glycoprotein associated with emerging variants of concern (VOCs). We found that different SARS-CoV-2 lineages have distinct evolutionary spaces (a probabilistic definition of the sequence states that can be occupied by expanding virus subpopulations). These spaces can be accurately inferred from patterns of amino acid variability at the whole-protein level. Robust networks of co-variable sites identify the highest-likelihood mutations in new VOC sublineages and predict remarkably well the emergence of subvariants with resistance mutations to COVID-19 therapeutics. Our studies reveal that low frequency variant patterns at heterologous sites across the protein accurately forecast the mutation potential at each position of interest. Accurate estimations of the potential changes in SARS-CoV-2 lineages can contribute to design of therapeutics that maintain their efficacy against future forms of this virus.
My third project focused on developing sequence-based machine learning models to predict resistance of HIV-1 to the Env-targeting agent Temsavir (TMR), an FDA-approved small molecule attachment inhibitor. Development of the models was based on use of sequence data collected from the patients, each associated with an in vitro-measured resistance value to TMR. Our approach incorporated features such as amino acid occupancy at Env positions, HIV-1 clade association, Env variable loop features, and a novel in silico estimation of the Env-TMR interaction. We show that our model performs well at predicting resistance to TMR at clinically relevant concentrations. Furthermore, it demonstrates robust generalizability when validated against an external dataset. Our models represent a significant advancement towards a practical clinical solution. By sequencing a patient's viral pool, we may precisely identify the most suitable set of antiretrovirals for treatment with the lowest likelihood of therapy failure. Taken together, our studies offer a simple framework to understand protein evolution, predict mutation emergence, and facilitate therapeutic development and administration.
Details
- Title: Subtitle
- Sequence-based prediction of virus phenotype, evolutionary potential, and selective pressures.
- Creators
- Roberth Anthony Rojas Chavez
- Contributors
- Hillel Haim (Advisor)Wendy Maury (Committee Member)Li Wu (Committee Member)Aloysius Klingelhutz (Committee Member)Benjamin Darbro (Committee Member)
- Resource Type
- Dissertation
- Degree Awarded
- Doctor of Philosophy (PhD), University of Iowa
- Degree in
- Microbiology
- Date degree season
- Spring 2024
- DOI
- 10.25820/etd.007731
- Publisher
- University of Iowa
- Number of pages
- xvii, 224 pages
- Copyright
- Copyright 2024 Roberth Anthony Rojas Chavez
- Language
- English
- Date submitted
- 04/22/2024
- Description illustrations
- illustrations, tables, graphs
- Description bibliographic
- Includes bibliographical references (pages 199-224).
- Public Abstract (ETD)
- The pandemics caused by HIV-1 and SARS-CoV-2 have led to advancements in treatments and vaccines. However, HIV-1 remains incurable and new SARS-CoV-2 variants challenge vaccine efficacy. The high mutation rates of these viruses contribute to their ongoing diversity, presenting clinical challenges that necessitate a deeper understanding of viral evolution. To address these challenges, we leveraged sequences information to investigate the evolutionary pressures applied to the virus, the potential for the emergence of mutations, and the effect of mutations onresistance to therapeutics. We identified interesting evolutionary patterns of mutations within HIV-1’s FPPR. We aimed to understand the selection pressures driving the evolution of this region. We found that some mutations increase infectivity while also exposing rare antibody neutralization targets. During infection, weak antibodies targeting these regions exert selective pressure against FPPR mutations, impacting mutation frequencies in the population. In addition, we explored how sequence information from early in the SARS-CoV-2 pandemic predicts the emergence of mutations in spike that were sampled within Variants-of-Concern lineages. While predicting future mutations may aid vaccine design, assessing their clinical impact is paramount. To address this, we developed a proof-of-concept sequenced based model aimed at predicting whether a given HIV-1 envelope sequence will be resistant to an FDA approved drug. Our tool achieves high predictive performance at clinically relevant concentration of the drug, offering the potential to expedite optimal clinical decisions regarding therapy regimens. Our studies offer a simple framework to understand protein evolution, predict mutation emergence, and facilitate therapeutic development and administration.
- Academic Unit
- Microbiology and Immunology
- Record Identifier
- 9984647149702771
Metrics
33 File views/ downloads
25 Record Views