|Home | About | Journals | Submit | Contact Us | Français|
Vaccine-induced cytotoxic T lymphocytes (CTL) have been implicated in the control of virus replication in simian immunodeficiency virus (SIV)-challenged and simian-human immunodeficiency virus-challenged macaques. Therefore, we wanted to test the impact that vaccine-induced CTL responses against an immunodominant Gag epitope might have in the absence of other immune responses. By themselves, these strong CTL responses failed to control SIVmac239 replication.
While many vaccination regimens have reduced viremia in macaques challenged with simian immunodeficiency virus (SIV) or simian-human immunodeficiency virus (5, 7, 10, 13, 15-17, 23, 29-32, 36, 39, 40), the immunological correlates (22) of this protection are poorly understood. Since most of these vaccine regimens induce both cellular and humoral responses, it is difficult to assess the relative contribution of each of these two arms of the immune system to viral clearance or reduction of virus loads.
We sought, therefore, to determine the role of antigen-specific cytotoxic T lymphocyte (CTL) responses in the control of SIVmac239 replication by employing an epitope-based vaccine regimen to induce an SIV-specific CD8+-T-lymphocyte response in the absence of any significant SIV-specific antibody or T-helper-cell responses. We previously generated strong CTL responses in macaques against a single nonameric SIV CTL epitope, Gag181-189CM9 (4), which is bound by the common rhesus macaque major histocompatibility complex class I molecule Mamu-A*01 (26). These CM9-specific CTL responses were induced by using a DNA priming-modified vaccinia virus Ankara (MVA) boosting regimen, which has been shown to be highly effective at inducing potent CTL responses in both mice and nonhuman primates (18-20). The particular CM9-specific response generated in the macaques represents one of two identified immunodominant Mamu-A*01-restricted responses arising very rapidly during the acute phase of SIV infection (2, 3, 28).
Six Mamu-A*01+ macaques (designated 95058, 96031, 95045, 96123, 96118, and 94004) were immunized intradermally with DNA and MVA as previously described (4). Briefly, animals were given a priming immunization of two DNA vectors which express the CM9 epitope, one in the context of several human immunodeficiency virus (HIV) CTL epitopes (pTH.HW) (4, 20) and the other in the context of a hepatitis B virus core antigen (HBcAg) (4; D. H. Fuller, unpublished data). The HBcAg vector was designed to provide nonspecific helper-T-lymphocyte responses during the induction of SIV-specific CTL without inducing SIV-specific helper-T-lymphocyte responses. Indeed, the magnitude of the CM9-specific responses induced in the macaques in this study correlated with that of antibody and proliferative responses against the HBcAg carrier (Fuller, unpublished), suggesting a beneficial effect from its inclusion in the vaccine. Animals were then administered twice intradermally booster doses of MVA, which encodes the same HIV/SIV polyepitope HW (MVA.HW) used in the pTH.HW vector (19-21), as previously described (4). Given the predilection of SIV for CD4+ lamina propria lymphocytes in the gut (41), two additional Mamu-A*01+ rhesus macaques (95061 and 96114) were given two priming doses of DNA to mucosal tissues (cheek, tongue, and rectal tissues) followed by three priming doses of DNA to both mucosal and skin surfaces. These two mucosally DNA-vaccinated macaques were then given booster doses of MVA (diluted in 500 μl of 1× phosphate-buffered saline) twice intrarectally (i.r.) in an attempt to induce CTL in mucosal tissues.
This DNA priming-MVA boosting vaccination regimen induced significant levels of CM9-specific CTL in several of the intradermally vaccinated Mamu-A*01+ rhesus macaques, as measured with tetrameric major histocompatibility complex class I-peptide complexes 1 week post-MVA vaccination (Fig. (Fig.1)1) (4). These levels were equivalent to, and in some cases higher than, levels of CM9-specific CTL responses detected during the acute phase of infection in SIVmac239-infected macaques (2). Interestingly, in the two macaques which received MVA i.r. (95061 and 96114), the levels of peripheral blood CM9-specific CTL were not boosted (both animals showed an increase to only 0.21%). This may suggest either that the MVA was not effective when delivered i.r. or that there was compartmentalization of the CM9-specific CTL to mucosal tissues. Unfortunately, mucosal tissues in these two animals were not directly assessed for levels of tetramer-positive cells.
Initially, three of the vaccinated Mamu-A*01+ macaques (95045, 95058, and 96031) were challenged i.r. with 1,000 50% tissue culture infective doses (TCID50) of SIVmac239/nef-stop (24, 25, 34) 1 week after administration of the second MVA booster (Fig. (Fig.1A).1A). Animals were challenged when the levels of antigen-specific CTL were expected to be at their highest. Three naïve Mamu-A*01+ macaques (95114, 95115, and 95084) and two naïve Mamu-A*01− macaques (95003 and 95112) were also infected as controls (Fig. (Fig.1A).1A). One of the challenged Mamu-A*01+ control animals (95084) showed no evidence of infection, did not develop CTL or antibodies to SIV, and was therefore considered a technical failure and excluded from further analysis.
The second set of five vaccinated Mamu-A*01+ rhesus macaques (96118, 96123, 94004, 95061, and 96114) were then challenged along with a set of six Mamu-A*01− controls (96072, 96104, 96113, 96020, 96081, and 96093) (Fig. (Fig.1B).1B). This set of animals was challenged i.r. with an in vivo-titered stock of 10 rhesus macaque ID50 (3.16 × 103 TCID50) of cloned SIVmac239/nef-open. This alternative stock of SIVmac239 was chosen over the initial challenge stock (SIVmac239/nef-stop) because it did not demonstrate the delayed kinetics of peak viral replication associated with the SIVmac239/nef-stop virus due to the presence of the stop codon in nef. In addition, the titers of SIVmac239/nef-open stock had been determined in vivo. SIVmac239 challenges generally result in viral set points between 105 and 107 viral copies/ml or higher (in 14 out of 14 control animals infected i.r. in this study and as reported elsewhere) (1). Therefore, the use of two different challenge virus stocks (SIVmac239/nef-open and SIVmac239/nef-stop) should not have had any significant impact on the outcome of this study. It is noteworthy that the levels of SIV RNA in the plasma of the original three vaccinated macaques (95045, 95058, 96031) peaked 1 week earlier than did those in the controls. This phenomenon may be due to the administration of the final MVA immunization dose only 1 week before challenge. Since HIV and SIV preferentially infect actively dividing cells, this may have provided a pool of MVA-activated CD4+ T lymphocytes to serve as additional sites for SIV replication. Therefore, SIV challenge of the second set of animals was delayed for 6 months following the last MVA booster. In addition, to control for the potential effects of MVA itself, four of the six control animals in this set of macaques were also mock immunized either intradermally (96081 and 96093) or i.r. (96104 and 96113) with an empty control MVA vector.
Of the vaccinated macaques, only animals 95058 (Fig. (Fig.1A)1A) and 95061 (Fig. (Fig.1B)1B) demonstrated moderate reductions in peak viremia and/or a reduced viral set point of over 2 logs compared to those of controls by 8 weeks postinfection. Interestingly, animal 95058 also demonstrated the highest levels of prechallenge CM9-specific CTL (4). This apparent viral containment was maintained until 24 weeks postinfection, after which plasma viral RNA began to rebound in this animal to levels above 105 copies/ml. It is noteworthy that CM9-specific tetramer levels also rebounded at the same time (Fig. (Fig.1A),1A), likely in response to the increased viral replication. Sequencing of SIV RNA in plasma from this animal as late as 52 weeks postinfection, however, indicated that the viral rebound was not a result of viral escape from the CM9-specific response (data not shown). In the second set of vaccinees, animal 95061 also demonstrated undetectable plasma viral RNA by 12 through 96 weeks postinfection (the last time point assessed). Interestingly, however, the animal that exhibited the highest levels of post-MVA antigen-specific CTL in this second group (animal 94004) had no significant reduction in plasma SIV RNA (Fig. (Fig.1B).1B). Similarly, none of the other vaccinees demonstrated any reduction in the amount of set point plasma SIV RNA compared to that in controls.
Since long-term nonprogressors mount strong proliferative responses and elaborate gamma interferon in response to HIV proteins (33, 37), we reasoned that proliferative responses should be present in those vaccinees that had reduced viral loads. CD8-depleted peripheral blood mononuclear cells were incubated with whole killed SIV (6, 27, 38), and samples were tested at 10 and 12 weeks postinfection. Considerable proliferative responses to whole inactivated SIV were detected in 6 of 8 vaccinees but in only 4 of 12 controls (data not shown). This difference, however, was not statistically significant (P = 0.065; t test), although animals that made poor vaccine-induced responses to the CM9 CTL epitope also made poor proliferative responses to SIV proteins after challenge. Proliferative responses were also observed to consistently increase over time in the majority of animals (data not shown). Therefore, while the vaccine may have had only a minor impact on reducing the viral replication, some beneficial effect through maintenance of CD4 proliferative responses was observed.
Analysis of antibody responses to both SIV Env and SIV Gag prior to and following challenge revealed no substantial differences in Gag or Env antibody titers between vaccinees and controls, and no neutralizing antibodies were detected in any of the animals (data not shown). Similarly, measurements of the levels of CD4+ T cells following viral challenge revealed that the majority of vaccinees were unable to maintain preinfection levels of CD4+ T cells (data not shown). However, animal 95058, who exhibited partial viral control, did not exhibit any substantial decline in CD4+ T cells in the first 20 weeks of infection.
Challenge of vaccinated and control macaques with a cloned, highly pathogenic viral isolate that reproducibly yields high levels of SIV RNA in peripheral blood may be a more rigorous challenge than humans might encounter. However, the value of this SIVmac239 challenge compared to other macaque challenge viruses such as simian-human immunodeficiency virus SHIV89.6P, against which vaccine-induced protection has recently been reported (5, 7), is that, much like primary HIV type 1 isolates, this strain of SIV is highly resistant to antibody-mediated neutralization (12, 35). The failure of our vaccination with a single CTL epitope in Gag, and the fact that this epitope does not escape in control macaques until well into the chronic phase of infection, suggests that CTL against the Gag CM9 epitope may not exert significant selective pressure on the virus. Since it is unlikely that all CTL responses are created equal with respect to their ability to control viremia in HIV-infected individuals (11), it is possible that vaccination with different CTL epitopes or even multiple CTL epitopes may be more efficient at controlling viral replication. Similarly, induction of CTL responses in the mucosa might also be more efficient at reducing viral replication postchallenge (8, 9, 14). Therefore, data from this study suggest that induction of CTL against this single immunodominant epitope will not reliably control viral replication after challenge with the highly pathogenic SIVmac239 molecular clone.
We thank Jenny Booth for performing the branched DNA assays in a timely fashion and Larry Arthur and Julian Bess for preparation of the inactivated SIV used in lymphoproliferation assays. We also thank the Immunology/Virology Core Laboratory at Wisconsin Regional Primate Research Center for the infection and monitoring of macaques. We thank Deb Fuller (Powderject Vaccines, Inc.) for assistance with immunizations.
This work was supported by grants from the National Institutes of Health (AI49120 and AI46366 to D.I.W., AI38081 to Epimmune, and RR00167 and CFAR CA 79458 to the Wisconsin Regional Primate Research Center) and by the National Cancer Institute, National Institutes of Health, under contract NO1-CO-56000. D.I.W. is an Elizabeth Glaser Scientist.