|Home | About | Journals | Submit | Contact Us | Français|
The HIV-1-specific immune responses of HLA-B*57/5801 patients who spontaneously control viral replication serve as an important model for T cell-based HIV-1 vaccines. Determining the breadth of this response and the extent of virologic escape that occur in primary infection in these patients is therefore critical. Here we document the development of mutations in 3 HLA-B*5801 restricted epitopes in gag, nef and pol in an HLA-B*5801 subject with a viral load of only 1159 copies/ml at day 167 post-infection. Thus, relative control of viral replication can be maintained in spite of the rapid development of multiple escape mutations within CTL epitopes.
Many HIV-1 infected patients who are positive for the closely related HLA-B*57 and HLA-B*5801 alleles have slowly progressive disease [1-3]. While CD8+ T cells that target HIV-1 epitopes restricted by these alleles are thought to be protective, multiple studies have shown that relative or complete control of viral replication can occur in spite of the presence of escape mutations [4-6]. The prototypical example of this is the T242N mutation in the Gag TW10 epitope: this mutation provides effective escape from the HLA-B*57/58 specific CTL response, but does not correlate with a loss of control of viremia, raising the question of how critical the CTL response is in the maintenance of viral control. Evaluating the temporal relationship between the development of escape mutations in HLA-B*57/5801 restricted epitopes and control of viral replication at a genome wide level would help determine the relevance of the CTL response in slowing progression to AIDS.
In this study, we document the transmission of HIV-1 from a patient with progressive disease to an HLA-B*5801 positive individual. Full genome sequence analysis of isolates obtained from both individuals allowed us to determine for the first time the extent of mutations in HLA-B*5801 restricted epitopes compared to the rest of the HIV-1 proteome. Additionally, longitudinal sequence data of specific genes, combined with ELISPOT analysis of critical epitopes and mutations within them, provides us with a more thorough perspective on the importance of escape in HIV-1 infection.
The transmitter was a 45-year-old HIV-1 infected progressor with a history of injection drug use who had a peak viral load of 53,000 copies/ml and a CD4 nadir of 207 cells/μl. He was not on HAART when the transmission occurred. His HLA alleles are HLA-A300101, -A330301, -B4201, and -B530101. Transmission from the transmitter to the subject, a 35-year-old woman who was HLA-A300101, -A680101, -B5801, and - B5802 positive, occurred after unprotected heterosexual vaginal intercourse in August 2006. She developed an influenza-like illness in early September of that year. HIV-1 test ing was negative on 8/18/06, but a repeat test was positive on 10/25/06. The date of transmission was estimated to be 8/15/06. The first viral load and CD4 cell count, obtained on day 102 post-infection, was 4940 copies/ml and 816 cells/μl respectively. On day 137 post-infection viral load was 1159 copies/ml and CD4 cell count was 655 cells/μl. The viral load reached a set point ranging between 3817 and 8285 copies/ml over the next 4 months while the CD4 cell count was stable at 757 and 785 cells/μl (Figure 1A). HAART was initiated in May 2007 as part of the Phase IV Randomized, Open-Label Controlled Trial of Time-Limited Highly Active Antiretroviral Therapy in Patients with Acute or Early HIV Infection (Clinical trial registry # NCT00106171). The clinical protocol and consent form used were reviewed and approved by the Western Institutional Review Board and the Johns Hopkins University Institutional Review Board. Informed consent was obtained before phlebotomy.
Freshly isolated peripheral blood mononuclear cells (PBMC) from the subject were stimulated with overlapping 15-mers peptides obtained from the NIH AIDS Research and Reference Reagent Program as previously described .
Genomic DNA and plasma RNA isolation was performed as previously described . To prevent resampling, nef, pol and gag genes were amplified from provirus in genomic DNA by limiting dilution “digital” nested PCR . RT-PCR was performed from plasma and serum samples as previously described .
Epitopes were defined based both on peptides to which the patients responded in our ELISPOT analysis and on epitopes documented for these patients’ HLA-A and -B alleles on the Los Alamos website CTL epitope map. The term “reversion” refers to differences between the transmitter and subject in which the virus evolved from an apparent escape mutation in the transmitter back to the B Clade consensus sequence in the subject. “Diversion” mutations refer to substitutions which were different both from the transmitter virus and the B Clade consensus sequences, likely representing escape in the subject’s HLA-restricted epitopes. For Env mutation analysis, we exluded sites at which alignment was ambiguous (including portions of V1/V2 and V4/V5). All sequences have been submitted to GenBank (GQ256628-GQ256646).
To confirm that the two patients were indeed a transmission pair, we amplified plasma virus from both patients at day 221 post-transmission and compared full length HIV-1 isolates. The dominant plasma sequences of Vif, Vpr, Vpu, and Tat and Protease were identical in the 2 patients, providing strong evidence that the virus was transmitted to the subject by her partner. Interestingly, almost half of the substitutions in the subject can be explained by the virus evolving in response to pressure from the CTL of the two patients. Four diversion mutations were located in HLA-B*5801-restricted epitopes while six were in epitopes restricted by other HLA alleles (Table). Reversion of potential escape mutations was seen in 3 epitopes that were potentially targeted by the transmitter at some point. Overall, 13 of the 27 differences in the predominant plasma virus amplified from the transmitter and the subject were present in CTL epitopes potentially targeted by either patient. This analysis clearly indicates that the CTL response has a strong influence on viral evolution.
Because of the importance of the HLA-B*5801 allele in viral control, mutations present in HLA-B*57/5801 restricted epitopes are of particular interest. We amplified and sequenced gag from plasma of the transmitter and near full-length gag from a serum sample from the subject on day 67 of infection. There were only 3 differences distinguishing the dominant Gag sequences of the two patients including T242N, a mutation in the HLA-B5801 restricted epitope TW10 (Figure 1B). Inspection of sequence chromatograms revealed double peaks (A, C) at the nucleotide 725 position (the presence of the codon with “C” results in wild type T242 whereas the mutant codon with “A” results in the expression of T242N). Thus it appears that evolution from wild type sequence to T242N was occurring at this time point. We were unable to amplify nef or pol at day 67 due to limited sample availability. However, all 3 genes were amplified from plasma and resting CD4+ T cells at later time points (Figure 1B).
At day 137, the predominant Nef sequence in plasma contained a F90L substitution in KF9. By day 221, this had been replaced by virus containing the well characterized A83G escape mutation  as well as a V85L substitution in KF9, and virus containing just the V85L mutation was also detected. This might reflect selection for a more fit isolate as the F90L mutation is extremely rare. Greater than 99.5% of all HIV-1 Nef sequences from all different clades in the Los Alamos database have F at position 90, and the F90L mutation is seen in just 1 other clade B isolate. In contrast, isolates with A83G and V85L are frequently observed.
The transmitter had 2 distinct plasma isolates with a threonine (T) or an alanine (A) at position 124 in Integrase. The predominant plasma virus in the subject had A at this position, but at day 137, a T125M mutation was detected in the SW10 epitope in some isolates. This mutant later became the dominant plasma and proviral sequence (Figure 1B).
To determine whether the substitutions in HLA-B*5801-restricted epitopes were escape mutations, ELISPOT analysis comparing wild type and autologous variants of epitopes was performed using PBMC from 2 years post-infection. As shown in Figure 1C, there was still a robust response to wild type TW10 but the response to the escape mutant was minimal. A greater response was seen to the transmitter’s variant of the SW10 epitope than to the variant that developed in the subject, suggesting the latter variant was an escape mutant. In contrast there was no detectable response to any of the 3 Nef KF9 variants tested. Thus early escape may have led to a dramatic decline in the frequency of cells specific for the transmitted variant. The use of PBMC from 2 years post-infection may have also resulted in our missing early responses to the escape mutants especially since the HIV-specific immune response can be affected by HAART.
There was substantial evolution in epitopes restricted by alleles other than HLA-B*5801 in the subject. Of the 3 differences seen in Gag on day 67, 2 fell within three overlapping HLA-B*53-restricted epitopes. Sequence of the 3′ end of the gene on day 137 revealed 2 more changes in HLA-B*53 epitopes (Table). Most of these changes likely represent reversion of escape mutations that occurred in the HLA-B*53-positive transmitter (Table). In addition, on day 221, there were changes that probably represented reversion in other epitopes that were targeted by the transmitter, and diversion within epitopes that may have been initially targeted by the subject (Table). No T cell responses were detected to some of these epitopes in the ELISPOT screen, but this could potentially be explained by early virologic escape leading to a loss of reactive CD8+ T cells in the subject.
In this study we document the transmission of HIV-1 from a patient with progressive disease to an HLA-B*5801-positive patient. While previous studies have shown the early development of escape mutations in HLA-B*57/5801 restricted epitopes in Gag [7-10] and Nef [7, 9], we extend these findings by examining the effect of CTL responses on evolution of the entire genome and by documenting the relative predominance of mutations in CTL epitopes. We show for the first time that while escape occurred rapidly in multiple HLA-B*5801-restricted epitopes in three different proteins (Figure 1B), relative control of viral replication was maintained.
Strikingly, all 3 of the substitutions in Gag at day 67 post-infection were in CTL epitopes. However, in spite of the strong influence of the CTL response on viral evolution, escape at CTL epitopes did not seem to affect control of viremia: the subject’s viral load continued to decline from 4940 copies/ml on day 102 post-infection to 1159 copies/ml on day 137. While the T242N mutation in Gag was present at day 67, no escape was seen in the Gag IW9 epitope although substitutions here have been shown to occur frequently in primary infection [9, 10]. Conversely, although the KF9 and SW10 epitopes rarely exhibited rapid escape at the population level in a recent study , mutations developed in both epitopes in our patient by day 137 post-infection.
Of the 8 independent epitopes in the entire HIV-1 proteome that were targeted on day 221, 3 had developed mutations by that time, including 2 of 4 HLA-B*5801 restricted epitopes. While earlier studies showed that the development of escape in a high proportion of CTL epitopes targeted during primary infection was associated with progressive disease [11, 12] our data suggest that early control of viral replication can occur despite the presence of multiple mutations in HLA-B*5801 restricted epitopes. Similar results were seen in a recent study of two acutely infected HLA-B*5701 positive patients .
Interestingly, our data also suggest that the presence of the protective B*5801 allele can supersede the presence of an allele associated with rapid disease progression, in this case B*5802 [3,14]. The subject possessed both of these alleles, yet failed to mount a response to the HLA-B*5802-restricted epitope in Env that is often targeted by chronic progressors . Instead, she exerted impressive though incomplete control over viremia, similar to that previously described in HLA-B*57 patients . While we do not know whether she would have continued to control viral replication without treatment, it is remarkable that her set-point viral load was so low in the presence of multiple escape mutations.
The data presented here advance our understanding of the important early events that occur during the successful spontaneous control which may be critical for the development of an effective T cell-based vaccine.
We thank Susan Langan, Jun Lai, and Dan Elwood for data management and helpful discussions.
This study was supported by the Center for AIDS Research grant number P30 AI42855 (J.B.M.) and NIH grants R01AI056990-01A1 (J.B.M.), R01 DA024565 (S.C.R.), R56 AI73185-01A1 and R01 AI080328 (J.N.B.) as well as the Howard Hughes Medical Institute (R.F.S.) and the General Clinical Research Center at the Johns Hopkins School of Medicine.
The authors report no conflict of interest.