Speaker
Description
Host cells carrying integrated HIV-1 DNA (provirus) can enter a transcriptionally inactive state. The population of these cells with intact provirus comprise the latent viral reservoir (LVR). Empirically, over 90\% of proviruses in a given host are defective, \textit{i.e.}, unable to produce infectious progeny virions. In addition to large mutations introduced during reverse transcription (RT), clonal expansion may also play a significant role in the evolving frequencies of defective proviruses.
In this study we develop a differential equation model to approximate HIV-1 replication within a host, tracking populations of free viruses, intact and defective proviruses denoted by V, I, I', respectively, and latent reservoirs of intact and defective proviruses denoted by L and L'. The model considers the dynamics of free virus binding, clearance, the probability of becoming a defective provirus during reverse transcription (RT error rate), death rates of infected cells, and virus particle release from an infected cell. The model also considers intact and defective proviruses transitioning to/from latency. To compare the effects of RT error rates and clonal expansion on frequencies of defective proviruses, we expanded the model to include a compartment representing the latter process.
The resulting models were fitted by least-squares minimization to longitudinal LVR data from nine HIV-infected individuals, collected by Cho et al.~(2022, PNAS) over the course of up to 10 years following ART initiation. We used the Akaike information criterion to quantify support for one model against another. Based on model parameter estimates, the rate that proviruses are rendered defective by RT mutation is lower than assumed in the literature and clonal expansion seems to have a limited role in the propagation of defective provirus.
