Once a person is infected with Human Immunodeficiency Virus (HIV), the virus evolves or undergoes changes to its sequence to produce virus variants. But how do these variants differ from the original ‘parent’ virus and what drives this evolution within a person? David Swan, MS, MME, Dr. Joshua Schiffer, and Dr. Dan Reeves from the Schiffer Lab in Fred Hutchinson’s Vaccine and Infection Disease Division sought to provide insight to these questions on intrahost HIV evolution and published their findings in a recent article in PNAS.
In the United States, about 36,000 people were diagnosed as HIV positive in 2019 (HIV: Basic Statistics, CDC). HIV infection causes a chronic illness called acquired immunodeficiency syndrome (AIDS) in which a person’s adaptive immune cells, CD4+ T cells specifically, are significantly reduced. This means that a person suffering from AIDS has a weakened immune system and their ability to fight diseases is less robust. Effective management of HIV-associated disease is possible by taking a multi-drug cocktail consolidated into a single pill taken daily that works to suppress HIV-associated disease, thus restoring a person’s adaptive immune response. However, HIV integrates into the DNA of cells and remains there even with this treatment. For this reason, understanding what drives changes in the genetic code of HIV may provide insight into how this virus can be eradicated from infected individuals and aid in our understanding of how HIV causes disease.
A long-standing question in the field of virus evolution is the extent to which the immune response dictates virus adaption within an infected individual. To address this question, members of the Schiffer lab combined several datasets from human HIV infections that together included data for 1) early infection virus abundance, 2) virus sequence divergence and diversity over time, and 3) sequence data for virus genetic code integrated into human DNA – from individuals taking virus-suppressing treatments. To determine how the adaptive immune response impacts HIV evolution within a person, 24 different mathematical model variations were applied to the combined datasets to determine which models for “viral fitness” and “adaptive immunity” accurately resembled the data.
Before testing these model variations, Dr. Reeves recalled envisioning what might be happening within a newly infected person. He predicted a scenario “in which HIV infects an individual and immediately the host adaptive immune system begins to act to control the virus, which leads to the virus escaping the immune system.” Following this initial round of adaptation, “the immune system has to readjust and tackle the newest “variant”.” With this rationale, one might expect the strain-specific adaptive immune response to be a significant driver of HIV adaptation and a key condition for modeling the intrahost HIV evolution data. The findings from this research however support a simpler model as "strain-specific immunity was not necessary to explain the data,” stated Dr. Schiffer. Interestingly, the ‘best-fit-model’ for the role of adaptive immunity in HIV evolution was one where the immune system broadly reduced virus abundance in a pan-HIV or HIV sequence-independent manner. In short, the adaptive immune response reduces viral load in the infected individual, but the evolution of new variants occurs in a more random fashion, not strictly dictated by strain-specific immunity. “Another way to think about this is that important T cell clones (and maybe antibodies) are cross reactive against multiple viral variants within [the] host,” said Dr. Schiffer.
The ‘best-fit-model’ also informed on another aspect of intrahost, or within the person, viral evolution. As HIV adapted, the fitness of most viral variants either decreased, generating ‘less fit’ or less infectious virus variants or retained their fitness level over time. Dr. Reeves added, “The model implies that the virus can take on many many reasonably good variants of itself (which agrees with in vitro “Deep Mutational Scanning” data from Jesse Bloom’s lab).” To summarize these data, HIV intrahost evolution is driven by mutations that occur by random chance. These random mutations result in a distribution of strain-specific fitness phenotypes including ‘less fit’, ‘similarly fit’, and ‘more fit’ than the ‘parent’ virus over time. These HIV variants are mostly ‘less fit’ or ‘similarly fit’ as the original, ‘parent’ virus, but a small fraction of ‘more fit’ viral variants are also generated, although not selected for as these mutations are driven by chance.
These findings demonstrate for the first time the utility of combining mathematical modeling and virus genetics to understand the dynamics of HIV intrahost evolution. Mr. Swan commented on this method, “We hope this [research] provides a simulation framework for studying the establishment of the latent HIV reservoir in the body and various means of reducing or eliminating that reservoir that are being explored today.” The application of this simulation framework has additional prospects. Dr. Reeves stated, “the logical next step of the project is to now use the model to help understand creation and maintenance of the HIV reservoir, the fundamental barrier to a cure in people living with HIV. We’re really excited this work is out and are looking forward to the next chapter.”
The spotlighted research was funded by a Washington Research Foundation and the National Institutes of Health.
Swan DA, Rolland M, Herbeck JT, Schiffer JT, Reeves DB. 2022. Evolution during primary HIV infection does not require adaptive immune selection. Proc Natl Acad Sci USA. 119(7):e2109172119.