A biphasic destruction of CD4+ T cells is observed in HIV infection with a massive loss of CD4+ T cells during early infection and a subsequent progressive loss during the chronic stage of infection [29, 30]. Retroviral suppression by ART results in an increase in peripheral CD4+ T cell counts and functional reconstitution of CD4+ T cell responses to many common antigens [31, 32], but HIV-specific CD4+ T cell responses remain deficient . We studied ART-treated, chronic SIV and RhCMV infected rhesus macaques and observed; 1) that both SIV and RhCMV antigen immunizations could induce immune activation and antigen-specific T cell responses in draining LNs. 2) In peripheral blood, the RhCMV-specific response induced by immunization was potent and sustained, whereas the SIV-specific response extinguished quickly. 3) We observed that SIV antigen immunization transiently induced greater levels of SIV replication in draining LNs of all animals compared to RhCMV immunization. In this study, the experimental design of immunizing the same animal with both an SIV antigen and a non-SIV antigen on collateral sides allowed us to directly compare the early immune and viral responses in draining LNs from the same animal at the same time, making it possible to study the effect of antigen stimulation in the context of ART-treated, chronic infection with limited animal numbers. All 3 animals displayed persistent CMV specific T cell responses for 50 days and demonstrated a weak transient SIV specific T cell response in peripheral blood. LNMCs from all animals demonstrated elevated levels of viral replication in response to SIV antigen immunization during the first 7 days after immunization. The data do not prove a causal link between the weak SIV T cell responses and LN viral replication, but are consistent with a hypothesis that SIV antigens induce viral replication that leads to depletion or dysfunction of antigen specific cells leading to a reduction in the strength and longevity of the response.
Pathogenic SIVmac251 infection of rhesus macaques has been well described as one of the preferred experimental models for studying HIV pathogenesis . In this study, all rhesus macaques inoculated with SIVmac251 were Mamu A*01 positive to control for an MHC effect on viral immune responses and disease progression, as well as to aid in the measurement of immune responses by a Mamu A*01 restricted SIV gag tetramer. All macaques established primary SIV infection with typical viral replication dynamics [[26, 27, 35]], and demonstrated responsiveness to ART with rapid control of viremia (Figure 2A). One animal had occasional blips in viral load that returned to undetectable without changes in therapy. It has been shown that even during the most potent regimens of retroviral suppression, the presence of virus in plasma could be measured by some ultrasensitive assays [[14–16]]. We believe the blips in FH40 are similar to those observed during therapy in humans, which are not associated with acute infection of cells  and believe that this is representative of a range of low level virus that can be measured during ART . One study identified that HIV-infected subjects that developed blips in viral load had higher instead of lower levels of CD4+ T cell responses to gag . We do not believe that the macaque with blips in viral load is responding differently to ART compared to the animals without measureable blips.
Therapeutic immunization for HIV infection during ART has been studied in SIV-infected rhesus macaques and the immunological and virologic consequences have been investigated in peripheral blood using DNA immunization [[39–42]], as well as other systems [[43–45]]. After release from ART, variable immunologic and virologic benefits were reported from no control , temporal control , to long-lasting virologic control . In this study, using RhCMV immunization as a non-SIV control in the same animal, we investigated the immunologic and virologic consequences of immunization with SIV antigen in chronic SIV and RhCMV co-infected, ART treated rhesus macaques focusing on the early response in draining LNs. The plasmids encoding immunogens used for immunization were previously demonstrated to induce potent immune responses in uninfected macaques [[19–25]]. The RhCMV pp65 plasmid had a greater number of immunostimulatory motifs, which would bias towards the null hypothesis. Moreover, all the comparisons between LNs at each time point were from the same animal collected at the same time, thus allowing us to use each animal as its own control. Also, our study focused on the local responses in draining LN, where the antigen-specific T cell responses and viral replication occur, rather than systemic responses in peripheral blood, as done in most previous studies. In addition, we chose a range of days post immunization covering the early activation of memory T cells and generation of effector responses. Our data show that both SIV and RhCMV antigen immunizations induced transient immune activation and antigen-specific T cell responses in draining LNs. Further measurement of these responses in peripheral blood showed that the RhCMV-specific responses were sustained in PBMC with a rapid onset of cytokine producing CD8+ T cells 3 days post immunization, which was maintained at day 50-post immunization, whereas the immunization induced SIV-specific T cell responses were transient, appearing only on day 9 post-immunization, and extinguished quickly in blood. The peripheral SIV- and RhCMV-specific CD8+ T cell responses were significantly induced compared to pre-immunization levels, supporting that both immunizations induced immune responses. Of interest, a study that repeatedly immunized macaques with long-standing ART-treated SIVmac251 infection induced stronger SIV-specific CD4+ and CD8+ T cell responses in blood . This study used the MVA vector and delivered three immunizations, whereas in our study, only a single gag DNA immunization was used.
HIV-specific CD4+ memory T cells are preferentially infected by HIV, carrying approximately 2- to 5-fold more viral DNA than total memory cells . We hypothesize that during an HIV-specific response, activation of HIV-specific CD4+ T cells, which bear higher amounts of latent virus, results in activation of viral replication in the LN leading to suppression of CD4+ T cells and the HIV-specific responses through multiple mechanisms, while a non-HIV antigen-specific response activates less viral replication allowing more efficient expansion of the response.