In a previous work, we immunized eight rhesus macaques with SIV-NEF and -GAG lipopeptides combined with a promiscuous TT 830–846 lipopeptide . In the present study, all animals and 3 control macaques were intravenously challenged with pathogenic SIVmac251. This pathogenic viral isolate consisted of a mixture of several viral quasispecies of the nef gene that display several differences in particular within the NEF epitopes recognized by lipopeptide-induced CTL.
Five macaques had post-immunization anti-NEF CTL and one of them (macaque 125) had CTL directed only against NEF 169–178, which is perfectly conserved within the challenge SIVmac251 quasispecies. No variation was observed in the sequence of this epitope 35 weeks following SIV challenge. Likewise, this epitope was conserved in both macaques 129 and 109. This result is in accordance with our previous data in another macaque immunized with a similar mixture of NEF- and GAG- lipopeptides . These observations suggest that NEF 169–178 is a stable epitope that is not submitted to the pressure of CTL selection.
Recently, Watkins et al.  demonstrated that a high level of CTL against the single GAG 181–189 epitope was not sufficient to control viremia. In rhesus macaques immunized with DNA-gag-pol-IL2, emergence of viral mutants occurred in GAG 181–189 after SIV-challenge under the pressure of mono-epitope CTL [25, 26]. This viral escape was due to the selection of mutant viral epitopic peptides unable to stably bind to MHC class I molecules, as we have previously shown in lipopeptide-vaccinated macaques within epitope NEF 128–136  and in HIV-infected patients . In addition, the emergence of such viral mutants had no effect on the viral load , which suggests no effect on viral fitness.
In the present study, macaque 105 had lipopeptide induced CTL against NEF 128–136, a non-conserved epitope within the pathogenic SIVmac251 isolate, which contains 18% of 136T and 82% of 136A quasispecies. Forty weeks following SIV challenge of this monkey, the percentage of 136T viruses had increased (45%) whereas 136A viruses decreased (55%). The persistence of the two wild type variants within the single vaccine induced CTL epitope did not affect viral replication. The NEF 116–126 epitope recognized by CTL after lipopeptide vaccination in macaque 127 was perfectly conserved in SIV isolate (NEF 116–126) as NEF 169–178 epitope but 122L mutant occurred in 40% of the SIV quasispecies 40 weeks after SIV infection. Nevertheless, the persistence of 60% of wild type viral sequences likely allowed viral replication to remain very high during clinical evolution in this macaque without effect on the high viral fitness.
Three of the 4 epitopes recognized by lipopeptide-induced CTL from macaque 129 were not conserved (NEF 128–136, NEF 201–211 and NEF 211–219) in SIVmac251 isolates. The emergence of the wild type variants 136T (100%) was observed within the CTL epitope NEF 128–136 after SIV challenge. Epitopes NEF 201–211 and 211–219 shifted by acquiring mutations that had no effect on viral load and the persistence of wild type viral sequences (22%) within these epitopes could also have contributed to intense viral replication.
In macaque 109, three of the 6 epitopes recognized by CTL following lipopeptide vaccination, namely epitopes (NEF 116–126, NEF 169–178 and GAG 266–275) were perfectly conserved in SIVmac251 isolates used for the challenge. Interestingly, after challenge, we did not observe any variation within all sequenced NEF epitopes from SIV-infected macaque in particular in epitope NEF 116–126, in contrast to the data in macaque 127. Within epitopes NEF 101–110, NEF 128–136 and NEF 215–225, only one viral variant issued from SIVmac251 was selected and expanded in the absence of emergence of new variations. We hence hypothesize that macaque 109 exerted a selection of few-replicative and non-pathogenic viral variant following SIV challenge. This selection could be the consequence of the vaccine induced CTL. However, we cannot formally exclude the role of an uncontrolled and random process.
CTL responses were evaluated in the two infected macaques 109 and 129, 12 months post-challenge. They were undetectable against all the identified vaccine peptides except in macaque 129. In the latter animal, CTL response against peptide 128–136 disappeared at week 47, following a 100% selection of 136T viral variant as shown in Table 3 and previously observed .
CTL obtained following vaccination could play a key role in the control of viremia. Decrease and control of viral load have also been reported in macaques vaccinated with MVA-gag-pol-env [27, 28], MVA-nef , MVA-gag-pol , ALVAC-gag-pol-env , NYVAC-gag-pol-env , adeno-gag , DNA [34, 35], a combination of DNA and MVA [36, 37] or a prime/boost with DNA/gag-Sendai virus  and challenged with SIV or SHIV. In these studies, the control of SIV/SHIV replication was clearly related to a high magnitude of CTL recognizing NEF  or GAG 181–189 epitope in MamuA1 macaques [30, 35], or to the selection of a non pathogenic viral mutant in GAG 206–216 (216S) CTL vaccine epitope . Indeed, viral escape by mutation in an epitope under CTL pressure can also prevent virus replication. Matano et al  observed that after vaccination with DNA/gag-Sendai and viral challenge, all macaques that controlled viral replication had a mutation in GAG leading to the substitution of one residue in GAG 206–216 (216S) CTL vaccine epitope by week 5 after challenge. This viral escape variant could have a lower fitness than wild type SIVmac239, indicating that the vaccine-induced CTL could have exerted a strong immune pressure leading to clearance of the wild type pathogenic SIV.
In our study, the emergence of several viral mutants in two macaques (127 and 129) within vaccine CTL epitopes was always associated with the persistence of the wild type virus and therefore was not concomitant with the decrease of viral fitness. The occurrence of an exclusive viral escape variant within several vaccine induced CTL epitopes was observed in only one macaque (109) and could be associated either with a selection of a poor replicative virus or with a control of viral replication by CTL.
These results tentatively bring a clue for a better understanding of SIV control and might provide new insight for the development of an effective HIV vaccine.