Background:  ART drug resistance is common in treated patients; however, it is not a significant predictor of HIV morbidity and mortality. Drug-resistance mutations may have an important effect on viral fitness that could allow sustained high levels of CD4⁺ T cell counts in the presence of significant levels of viral replication. Using a Bayesian hierarchical model, we found that the D30N mutation of HIV protease had the largest decrease in in vitro replication capacity relative to other resistance-associated mutations. We wanted to test the hypothesis that the D30N mutation has an effect on viral fitness in vivo in a primary lymphoid organ.

Methods:  To test the effect of the D30N mutation in vivo, we constructed mutants of HIVNL4-3 containing either the D30N mutation in protease, the M184V mutation in reverse transcriptase, or both the D30N + M184V mutations. We then used the SCID-hu mouse model to determine the effects of virus containing these mutations compared to wild type virus in vivo. Wild type and the mutant virus were injected into the thymus implants in the SCID-hu mice and the ability of the virus to replicate and induce CD4-depletion was assessed 3, 5, and 7 weeks following infection (n = 5 mice per group, viral load was measured by polymerase chain reaction of cells from biopsies and CD4 depletion was determined by flow cytometry). Groups were compared using Kruskal-Wallis.

Results:  We found that virus containing the D30N mutation (D30N or D30N+M184V) had a significant defect in the ability to replicate and induce CD4 depletion in vivo compared with wild type virus or virus containing the M184V mutation (p = 0.02). By week 7, thymus tissue infected with either the D30N or the D30N + M184V did not differ significantly in CD4⁺ thymocyte levels from uninfected implants (CD4 percentages:  92.5, 93.7, and 94.1, respectively, p = 0.56). While viral DNA was detected, mice infected with the D30N mutant or the D30N + M184V at week 7 had significantly lower viral DNA levels than wild type (p = 0.017) or M184V (p = 0.004).

Conclusions:  We identified the D30N mutation as having a significant defect in RC in vitro. We then demonstrated impaired viral fitness in virus containing the D30N mutation in vivo as determined by its decreased ability to replicate and deplete thymocytes. These results suggest changes in viral fitness due to resistance mutations could partially explain the sustained immunologic and virologic benefits seen in patients with drug-resistant HIV.