The HIV-1 integrase genotype strongly predicts raltegravir susceptibility but not viral fitness of primary virus isolates

Objective: Resistance to raltegravir is associated with three genetic pathways defined by the mutations Y143R/C, Q148H/R/K or N155H in integrase, which also infer a viral fitness cost. Additionally, the three major HIV-1 drug-targeted enzymes protease, reverse transcriptase and integrase mature from the same polyprotein, suggesting the potential for interaction between them. This study aims to elucidate the relative contribution of protease-reverse transcriptase, integrase and the rest of the HIV-1 genome to viral fitness and susceptibility to raltegravir. Methods: Recombinant viruses included integrase, protease-reverse transcriptase or the complete pol-coding region from three patients whose raltegravir-containing regimen had failed. The first had the mutations G140S+Q148H+S230N, the second had Y143R+G163R and the third had no evidence of genotypic resistance in integrase. Primary virus isolates were obtained from peripheral blood mononuclear cells. In-vitro phenotypic resistance and changes in replication capacity were assessed. Results: Virus isolates, and integrase-recombinant and pol-recombinant viruses from the patients harboring integrase resistance mutations showed a decrease in raltegravir susceptibility, with no differences between them. Defects in viral fitness were modulated by resistance mutations within protease, reverse transcriptase and integrase, which were further compensated by regions outside pol. Moreover, protease-reverse transcriptase rescued replication capacity of viruses containing integrase resistance mutations, although integrase was unable to compensate defects in replication capacity caused by protease-reverse transcriptase resistance mutations. Conclusion: Susceptibility to raltegravir is driven by resistance mutations in integrase, whereas other viral genes are involved in restoring defects in viral fitness in patients whose raltegravir-containing regimen fails, suggesting the existence of epistatic effects on replication capacity. (C) 2010 Wolters Kluwer Health vertical bar Lippincott Williams & Wilkins