Vaccine science in Bangkok

By Richard Jefferys*

Over the past several years, there has been a sense that both the number and overall quality of the basic science presentations at the International AIDS Conference has declined.

As Robert Steinbrook wrote in a post-conference review in the New England Journal of Medicine, “it is no longer the cutting-edge scientific meeting that it once was.” The throng of attendees that braved the sweltering humidity of Bangkok in July faced an additional problem: the US Department of Health & Human Services limited the number of US government-sponsored researchers to just fifty, resulting in withdrawn abstracts, speaker changes and blank walls where scientific poster presentations were intended to hang. Nevertheless, there were still some interesting, thought-provoking presentations on vaccine-related science.


Enhancing DNA vaccines

After an initial wave of excitement generated by studies in small animal models, DNA vaccines have hit something of a wall due to poor immunogenicity results in humans and researchers are now exploring strategies to overcome this problem. H Li from Yi-Ming Shao’s group at the National Center for AIDS/STD Control and Prevention in China described their recent work in this area. Li began by noting that DNA vaccines have to gain access to the nucleus so that the DNA can be transcribed and translated into protein antigens that can then trigger an immune response. There are two important barriers the DNA vaccine must pass through: the outer plasma membrane and the nuclear envelope. Li reviewed published work demonstrating that a short (72 base-pair) sequence of DNA from simian virus 40 (SV40) encodes a signal that promotes nuclear import (Exp. Cell. Res. 233, 713, 1999). Li and colleagues decided to exploit this finding by incorporating the SV40 sequence into a DNA vaccine construct encoding the Gag protein from HIV. They then conducted studies in mice to directly compare the immunogenicity of the original and modified DNA constructs. Animals were immunized at weeks 0, 2, 4 and 6 and responses to the HIV Gag protein evaluated at week 8.

Li showed that both antibody titers and CD8+ T cell responses, measured by both ELISPOT and intracellular cytokine staining (ICS) for interferon (IFN)-γ production, were approximately two-fold higher in mice that received the DNA vaccine containing the SV40 sequence compared to those that received the original construct. While it certainly cannot be assumed that this apparent enhancement of immunogenicity in mice will be mirrored in humans, Li and colleagues are sufficiently encouraged by the data to move the construct into human testing.

A collaborative effort between the research groups of George Pavlakis and Barbara Felber at the US National Cancer Institute is also evaluating novel methods for improving DNA vaccines. Pavlakis presented new data from studies of DNA vaccines that encode the simian immunodeficiency virus (SIV) proteins Gag and Env fused to genes with potential adjuvant effects. One set of constructs (the researchers have created one DNA vaccine for each protein) fuses the SIV protein to the chemokine monocyte chemoattractant protein-3 (MCP-3), to try to improve the secretion of the proteins from cells that take up the DNA vaccine; MCP-3 is able to exit cells particularly easily, using a route called the secretory pathway. The second set of constructs fuses the SIV protein to a fragment of ß-catenin which chaperones the proteins out of the cell via a different route called the proteasomal degradation pathway.

Pavlakis briefly reviewed a preventive vaccine study in juvenile macaques that was presented at the Keystone conference earlier this year. Four groups of four animals each were assigned to receive DNA constructs containing: (a) Gag and Env alone; (b) Gag and Env alone + Gag and Env fused to ß-catenin; (c) Gag and Env alone + Gag and Env fused to ß-catenin + Gag and Env fused to MCP-3; (d) no immunization (controls).

Immunizations were given at weeks 0, 4, 12, 24 and 48, followed by a challenge with the highly pathogenic SIVmac251 at week 54. Surprisingly, given the virulence of the challenge virus and the fact that only DNA vaccines were used, animals in group (c) had viral loads that were statistically significantly lower than controls during both the acute and chronic phases of the infection. Pavlakis next unveiled results from two new studies involving the combination of both fusion constructs. The first was conducted in collaboration with Marta Marthas (University of California at Davis) using a macaque model of breastfeeding transmission (see IAVI Report 5(8), 2001, p4 for background on this model). Neonatal macaques were immunized at weeks 0, 2 & 3 and challenged orally with SIVmac251 (in milk, twice a day for 5 days) beginning at week 4. Six of eight vaccinated animals and seven of eight non-vaccinated controls became infected (a non-significant difference), but the post-infection viral loads in vaccinated animals were significantly lower compared to controls (p=0.029).

Lastly, Pavlakis offered a glimpse of some intriguing results from a study aimed at exploring the therapeutic potential of the DNA vaccines. Thirty one macaques that had been infected with SIVmac251 for 15-70 weeks were treated with combination antiretrovirals (ARVs; ddI, d4T & PMPA) for approximately 20 weeks. Fifteen animals were immunized with the DNA constructs at weeks 8, 12 & 16 of the on-therapy period. At week 20, ARVs were stopped in all macaques. In the group that did not receive vaccines, viral loads quickly returned to baseline levels. Among the immunized group, however, five animals were able to control SIV viremia to below the levels of detection after an initial rebound. Six additional macaques displayed viral loads that were 1-2 logs below baseline values. Viral loads in the remaining vaccinated animals were indistinguishable from controls. While interesting, there are some important caveats to the results. Due to high costs and a shortage of macaques, animals were enrolled as they became available from other studies. Furthermore, a subgroup of both the vaccinees and controls had received DNA immunizations prior to infection with SIVmac251 (although separate analyses suggested that this did not bias the overall outcome). The data does at least suggest that further studies of the approach may be warranted. In a follow-up interview, Pavlakis stated that “we hope these results will stimulate interest in human trials using DNA vaccination, which appears to be very safe and—unlike some vector approaches—can be administered many times.” Pavlakis acknowledged that recent disappointments have cast a pall over the DNA approach, but argued that “it’s too early to abandon DNA vaccines for AIDS.” The two NCI groups currently lack support for GMP manufacturing, but they hope to advance some of these projects through a collaborative agreement with Wyeth Vaccines. In the immediate future, further macaque studies are planned in collaboration with David Weiner at the University of Pennsylvania.

Phase I vaccine studies

Bangkok saw the presentation of the first human immunogenicity data from two Phase I studies of new AIDS vaccine candidates. Guiseppe Pantaleo of the Laboratory of AIDS Immunopathogenesis, Lausanne and the EuroVac consortium described results from a trial of a NYVAC vector, an attenuated vaccinia virus, encoding a Gag-Pol-Nef polyprotein and the Env protein from a Chinese clade C virus. Twenty-four individuals with a low risk of HIV infection (11 women and 13 men) were recruited at sites in London, UK and Lausanne, Switzerland. Immunizations were given intramuscularly at weeks 0 & 4; four participants were given placebo while the remaining 20 received NYVAC. Pantaleo reported that there were no serious adverse events. CD8+ T-cell responses were evaluated at week 6 using IFN-γ ELISPOT and the cut-off for a positive response was defined using the same criteria employed by Merck in their recent vaccine studies: >55 spot-forming cells (SFC) per million PBMCs (peripheral blood mononuclear cells) and at least 4-fold greater than the background number of SFC; assays with a background of >50 SFC/million PBMC were considered invalid. Approximately 45% of participants displayed a positive ELISPOT response, mainly targeted at the Env protein. Responses to Gag and Nef were less common, and no Pol-specific CD8+ T cells were detected. Follow-up is ongoing, with further immunogenicity evaluations planned for the week 8, 24 and 48 timepoints to assess the degree of long term T-cell memory induced by the vaccine. EuroVac is conducting similar studies with additional vectors (including MVA) containing the same insert in order to compare immunogenicity and thus prioritize which candidates will be advanced into Phase II testing.

Representatives from the German company Bavarian Nordic presented preliminary data from an ongoing study of their MVA vector encoding the Nef protein in HIV uninfected volunteers (a therapeutic study was published last year; Vaccine 22, 21, 2003). Fourteen individuals are enrolled, and immunogenicity data was available for eight of them. Immunizations were given subcutaneously at weeks 0, 4 & 16. T-cell responses were evaluated by IFN-γ ELISPOT using overlapping peptides, short peptides corresponding to optimal epitopes, the Nef protein and the MVA vector. All eight participants evaluated so far have developed MVA-specific T cell responses, with a median frequency of 122 SFC/million PBMC at week 6. Six participants have been followed out to week 18; median MVA-specific responses were 177 SFC/million PBMC at this timepoint. In contrast, however, only 3/6 showed any evidence of Nef-specific T-cell responses and at a very low borderline level of 17 SFC/million PBMC (range 7-49). These data, while preliminary, suggest that responses to MVA-based vaccines may skew toward the vector itself rather than the encoded HIV antigen. Studies using MVA vectors encoding a broader array of HIV antigens should help discern whether this is a generalized problem or specific to the Nef protein.


The vast majority of AIDS vaccine candidates currently in human trials aim to induce HIV-specific CD4+ and CD8+ T cell responses. The degree to which such responses may benefit an uninfected individual upon subsequent exposure to HIV remains a critically important question, which can only be answered definitively by efficacy trials of T cell-based vaccines. In the meantime, however, investigators continue to try and understand the role of T cells in controlling HIV replication in infected individuals. In Bangkok, a session on immunology featured reviews of current knowledge regarding HIV-specific T cell responses. Guiseppe Pantaleo addressed CD4+ T cells and Juliana McElrath (University of Washington) discussed CD8+ T cells.

Pantaleo began his talk by reviewing previously presented work by Alexandre Harari from his group regarding the potential importance of interleukin (IL)-2-producing HIV-specific CD4+ T cells in controlling viremia (see IAVI Report, 8(1), 2004). Pantaleo then described a new study that used ICS to analyze the functional profile of tetanus toxoid (TT)-specific CD4+ T cells in individuals with a history of prior TT immunization, both before and after receipt of a TT booster. The take-home message of Pantaleo’s presentation was that the cytokine profile of TT-specific CD4+ T cells changes in response to stimulation with the TT booster. Prior to boosting, TT-specific CD4+ T cells primarily produced IL-2, indicative of a long-term memory population. In the weeks immediately after boosting, the TT-specific CD4+ T-cell population expanded and comprised a mixed population of cells secreting either IL-2 alone, IL-2 + IFN-γ or IFN-γ alone. Pantaleo noted that the population secreting IFN-γ alone were likely to be short-lived effector cells generated from the long term memory CD4+ T cells that were present prior to boosting. In support of this hypothesis, he showed that by day 60 after the booster, the TT-specific CD4+ T cell population had contracted and was again almost entirely made up of cells that secreted IL-2 alone. Some researchers have argued that the functional profile of responding T cells differs based on the particular pathogen that the cells are targeting; Pantaleo argued that the profile largely depends on how recently the T cells have been stimulated by antigen. For analysis of vaccine-induced CD4+ T cell responses, Pantaleo suggested that it will be critical to quantify not just IFN-γ production but also IL-2 since this cytokine appears to be a signature of long term immunological memory (shortly after the conference, Pantaleo and Rick Koup from the US Vaccine Research Center published a Commentary addressing this issue; Nat. Med. 10, 806, 2004).

McElrath highlighted some of the reasons why CD8+ T cells (cytotoxic T-lymphocytes or CTLs) typically fail to control HIV replication in infected individuals. First she discussed the impact of immune escape, wherein regions of HIV targeted by CTLs mutate to avoid recognition. McElrath’s group studied the evolution of the HIV-specific CTL response in 21 seroconverters and (in collaboration with Jim Mullins) conducted full length sequencing of the HIV genome at multiple timepoints post-infection to look for evidence of CTL escape mutations. The first HIV proteins to be recognized by CTLs were Tat and Vpr, and epitopes within these proteins were also the first to show evidence of CTL escape mutations. McElrath pointed out that studies in macaques have suggested that the efficiency of CTL recognition of a particular epitope (the technical term is functional avidity) may be correlated with the incidence of escape mutations, but so far she has been unable to confirm these observations. In McElrath’s study, escape occurred in epitopes within all viral proteins and appeared unrelated to the magnitude or functional avidity of the CTL response.

She went on to discuss the possible role of CTL dysfunction in HIV infection. Following up on published work by Mark Connors (Nat. Immunol. 3, 1061, 2002), McElrath has begun to investigate whether the ability of CTLs to proliferate in response to HIV antigens is correlated with control of viremia. In the two seroconverters studied to date, HIV-specific CTLs appeared to maintain their proliferative capacity for the first few months post infection, but this functional property was lost over time. McElrath hypothesized that the apparent loss of CTL function may relate to the rapid diminution of HIV-specific CD4+ T-cell activity that typically occurs in acute HIV infection. Her preliminary finding was echoed in a late-breaker presentation by Marcus Altfeld from Bruce Walker’s group in Boston; in a study of 18 acutely infected individuals, he also found that HIV-specific CTLs initially possessed the ability to proliferate but that the function was lost over time. Supporting McElrath’s hypothesis, Altfeld also reported that the addition of autologous CD4+ T cells isolated from an early timepoint in infection restored the proliferative capacity of the dysfunctional HIV-specific CTLs while, conversely, depletion of CD4+ T cells almost totally abrogated the proliferation. As McElrath explained at the close of her presentation, these data suggest that the functional properties of HIV-specific T cells may be an important correlate of the immune system’s ability to control viral replication, but the critical question is whether the pre-exposure induction of T cell responses by vaccination will allow seemingly important functions (such as proliferation and IL-2 production) to be better maintained if an individual becomes HIV-infected.


The next International AIDS Conference will take place in 2006 in Toronto (www.aids2006.org). While it’s likely that the meeting will continue to have to juggle the many different topic areas relevant to the global AIDS pandemic, vaccine science and policy will undoubtedly remain on the agenda. The conference co-chairs are Helene Gayle, recently appointed president of the International AIDS Society, and former IAS president Mark Wainberg, both of whom have a strong interest in accelerating the progress of AIDS vaccine research.

*Richard Jefferys is Basic Science Project Director at the Treatment Action Group, a New York-based organization advocating for HIV research.