In vitro HIV-1 evolution in response to triple reverse transcriptase inhibitors & in silico phenotypic analysis

• Published in: PloS one 2013-04, Vol.8 (4), p.e61102
• Main Authors: Rath, Barbara A, Yousef, Kaveh Pouran, Katzenstein, David K, Shafer, Robert W, Schütte, Christof, von Kleist, Max, Merigan, Thomas C
• Format: Article
• Language: English
• Subjects: Acquired immune deficiency syndrome; AIDS; Amino Acid Substitution - genetics; Antiretroviral drugs; Biological evolution; Biology; Clinical isolates; Computer Simulation; DNA polymerases; Dose-Response Relationship, Drug; Drug resistance; Drug Resistance, Viral - drug effects; Drugs; Epistasis; Evolution; Experimentation; Fitness; Genetic Fitness; HIV; HIV-1 - drug effects; HIV-1 - growth & development; HIV-1 - isolation & purification; Human immunodeficiency virus; Humans; Kinetics; Mathematics; Medicine; Mutation; Mutation - genetics; Parameter estimation; Parameters; Phenotype; Protease inhibitors; Reproductive fitness; Reverse transcriptase inhibitors; Reverse Transcriptase Inhibitors - pharmacology; RNA-directed DNA polymerase; Serial Passage; Stochastic Processes; Stochasticity; Time Factors; Viruses
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0061102

Effectiveness of ART regimens strongly depends upon complex interactions between the selective pressure of drugs and the evolution of mutations that allow or restrict drug resistance. Four clinical isolates from NRTI-exposed, NNRTI-naive subjects were passaged in increasing concentrations of NVP in combination with 1 µM 3 TC and 2 µM ADV to assess selective pressures of multi-drug treatment. A novel parameter inference procedure, based on a stochastic viral growth model, was used to estimate phenotypic resistance and fitness from in vitro combination passage experiments. Newly developed mathematical methods estimated key phenotypic parameters of mutations arising through selective pressure exerted by 3 TC and NVP. Concentrations of 1 µM 3 TC maintained the M184V mutation, which was associated with intrinsic fitness deficits. Increasing NVP concentrations selected major NNRTI resistance mutations. The evolutionary pathway of NVP resistance was highly dependent on the viral genetic background, epistasis as well as stochasticity. Parameter estimation indicated that the previously unrecognized mutation L228Q was associated with NVP resistance in some isolates. Serial passage of viruses in the presence of multiple drugs may resemble the selection of mutations observed among treated individuals and populations in vivo and indicate evolutionary preferences and restrictions. Phenotypic resistance estimated here "in silico" from in vitro passage experiments agreed well with previous knowledge, suggesting that the unique combination of "wet-" and "dry-lab" experimentation may improve our understanding of HIV-1 resistance evolution in the future.