HIV-1 sequence evolution after dendritic-cell-based immune therapy in a phase I/II clinical trial

Background: Therapeutic vaccination of HIV infected subjects is explored as an alternative to antiretroviral therapy (HAART). In a phase I/II clinical trial, HIV-1 infected subjects were immunized with autologous dendritic cells (DC). We monitored HIV-1 sequence evolution after immune therapy with DC expressing the HIV-1 early antigens Tat, Rev and Nef.

Methods: 17 HIV-1 infected subjects on HAART underwent four monthly vaccinations with autologous DC preparations electroporated with Tat, Rev or Nef expressing mRNA in a phase I/II clinical trial. HAART was interrupted one week after the last vaccination. Participants were regularly screened for clinical, immunological and virological parameters.
HIV RNA was extracted from plasma samples collected before the start of HAART (timepoint PRE), early after vaccination when HAART was stopped (timepoint early post, EP) and about one year post treatment interruption (timepoint late post, LP). RNA was amplified by RT-PCR and bulk-sequenced using standard protocols. Sequences of the vaccine genes tat, rev and nef and control genes vif, vpr, vpu and parts of env were analyzed for variation between pre- and post vaccination time points. Phylogenetic analysis was conducted on the total dataset. Per gene and per subject, base substitution rate (genetic distance) was calculated for the DNA and protein pairwise alignment of PRE vs EP and PRE vs LP. Sequences spanning known and predicted epitopes were analyzed in detail for each subject and for frequent HLA haplotypes.

Results and Conclusion: Immune therapy was well-tolerated and no severe adverse effects occurred. After HAART interruption, plasma viral load became detectable in all patients after 2-6 weeks. Following the viral rebound, a set point was reached, that was lower than the viral load before start of HAART. For 14/17 subjects pre-HAART sequences were obtained, sequences of different timepoints per individual are in the same phylogenetic cluster. With one exception, genetic distance EP vs PRE and LP vs PRE is very small, for both vaccine and control genes, indicating no significant impact of the immune response on virus evolution for the studied population. More focussed analysis on sequences spanning A*0201 HLA restricted epitopes showed that 95% of epitopes were mutated PRE vs vaccine sequence in both A*0201 and non A*0201 subjects. However, for non A*0201 subjects, less mutations occur EP or LP compared to PRE than for A*0201.
For one selected subject, a mutation in a newly discovered Rev B*4403 epitope was observed. This mutation persists in time and no activation of PBMC upon peptide stimulation with the mutated epitope is observed, as opposed to wild-type sequence.
In conclusion, immune therapy had limited impact on sequence evolution in vaccine genes.