Programmed cell death-1 (PD-1) at the heart of heterologous prime-boost vaccines and regulation of CD8+ T cell immunity
© Bot et al. 2010
Received: 11 August 2010
Accepted: 14 December 2010
Published: 14 December 2010
Skip to main content
© Bot et al. 2010
Received: 11 August 2010
Accepted: 14 December 2010
Published: 14 December 2010
Developing new vaccination strategies and optimizing current vaccines through heterologous prime-boost carries the promise of integrating the benefits of different yet synergistic vectors. It has been widely thought that the increased immunity afforded by heterologous prime-boost vaccination is mainly due to the minimization of immune responses to the carrier vectors, which allows a progressive build up of immunity against defined epitopes and the subsequent induction of broader immune responses against pathogens. Focusing on CD8+ T cells, we put forward a different yet complementary hypothesis based primarily on the systematic analysis of DNA vaccines as priming agents. This hypothesis relies on the finding that during the initiation of immune response, acquisition of co-inhibitory receptors such as programmed cell death-1 (PD-1) is determined by the pattern of antigen exposure in conjunction with Toll-like receptor (TLR)-dependent stimulation, critically affecting the magnitude and profile of secondary immunity. This hypothesis, based upon the acquisition and co-regulation of pivotal inhibitory receptors by CD8+ T cells, offers a rationale for gene-based immunization as an effective priming strategy and, in addition, outlines a new dimension to immune homeostasis during immune reaction to pathogens. Finally, this model implies that new and optimized immunization approaches for cancer and certain viral infections must induce highly efficacious T cells, refractory to a broad range of immune-inhibiting mechanisms, rather than solely or primarily focusing on the generation of large pools of vaccine-specific lymphocytes.
Vaccines are arguably the best medical tools we have at our disposal to fight widespread infectious diseases. Despite decades of vaccine research and development against life-threatening infectious diseases with global impact , culminating with the recent licensing of vaccines against human papillomaviruses (HPV) , a key cause of cervical cancer, successes have been confined primarily to prophylaxis. Vaccination has also been extensively researched for the prevention of HIV infection. Therapeutic immunization for cancer or chronic viral infection, however, brings in a new set of lessons and challenges with a few successes to date, such as treatment of HPV-related lesions . It became rapidly evident that the conventional paradigm of eliciting, amplifying, and maintaining immune responses with conventional vectors and homologous prime-boost approaches fell short of expectations in the clinic due to suboptimal immune response results. Two decades since the first cloning of tumor antigens , multiple vaccines are currently in development. Thus far, however, sipuleucel T (Provenge®) is the only approved therapeutic cancer vaccine in the US to date, consisting of autologous DCs expressing prostate acid phosphatase (PAP) and producing granulocyte macrophage colony-stimulating factor (GM-CSF) to treat hormone-refractory prostate cancer .
The HIV vaccine field has unquestionably been at the forefront of vaccine research, exploring potent immunization strategies comprised of synthetic vectors rather than cell-based vaccines. This is in contrast to efforts in cancer vaccine development where cell-based vaccines currently lead the field, while many synthetic and viral vector approaches are in clinical development [6, 7]. Nevertheless, homologous prime-boost approaches for the prophylaxis of HIV, such as the Vaxgene program, showed no significant protective effects in man . While in parallel, emerging evidence over the last two decades showed that novel prime-boost protocols integrating different vectors such as recombinant viruses and proteins [9, 10] did yield considerably higher immune responses with protective capability in several animal models. With the advent of other vectors such as DNA vaccines, and a range of recombinant microbial vectors including alpha virus replicons, research in the area of heterologous prime-boost vaccination against HIV has expanded and resulted in hundreds of preclinical and clinical studies. Interestingly, the most promising clinical regimens to date include: i) the RV144 landmark HIV 'Thai trial' which utilized recombinant viral priming followed by a protein boost and was the first to show modest yet statistically significant evidence of HIV vaccine efficacy in man ; ii) DNA priming coupled with protein ; or iii) DNA priming followed by a recombinant virus boost .
Significant evidence points to two major reasons why heterologous prime-boost vaccination is a more promising strategy compared to homologous prime-boosting: i) diminished anti-vector antibody responses  known to interfere with immunity against target epitopes through the clearance and degradation of vaccine via vaccine-antibody immune complexes; and ii) there is the potential for different vectors to work synergistically by inducing complementary arms of the immune response to jointly control complex pathogenic processes and overcome immune escape mechanisms. For example, while recombinant proteins are quite effective at inducing B and Th immunity, viral vectors can be more effective at inducing cytotoxic T cells .
The optimal positioning of current and future DNA vectors within innovative heterologous prime-boost immunization regimens requires a deeper understanding of the mechanism of action of DNA vaccination. A key observation from many studies to date is that interchanging the order of vectors utilized in these regimens has a dramatic impact on the resulting immune response. For example, while DNA priming followed by a virus boost resulted in significant epitope-specific responses, viral priming followed by DNA boost failed to reproduce this level of specific immunity . A similar result was observed with other vectors in a distinct model, clearly supporting a precise sequence of administration of vectors as a major factor determining the magnitude of immunity , although this hypothesis still requires further testing in other heterologous prime-boost vaccine protocols. This asymmetry between priming and boosting vectors could very well be at the heart of both the mechanism and advantage of heterologous prime-boost regimens. Therefore, the remainder of this review will focus on this key feature and its underlying mechanism, with emphasis on DNA vaccines as priming agents and CD8+ T cell immunity as the desired outcome, as it pertains to the control of cancer and chronic viral infections. Moreover, although we focus on the functionality of CD8+ T cells in this review, we recognize the importance of CD4+ T cells and the possibility that these cells may influence the outcome of vaccine protocols with respect to PD-1 expression by CD8+ T cells.
The fundamental concept behind heterologous prime-boost vaccination is the synergistic contribution of two categories of vectors to induce enhanced immunity against given epitopes. To investigate the immune mechanisms underlying this process, we initiated a systematic evaluation utilizing a reductionist approach that encompasses simple vectors with well-defined MHC class I-restricted epitopes. Using a Melan A/MART-1 preclinical experimental model, we developed a strategy that greatly enhances the immune properties of non-replicating vectors and biological response modifiers by direct intra-nodal administration of plasmid and peptide [19, 41]. We showed that the sequence and the route of administration of plasmid and peptide were absolutely essential to achieve improved antigen-specific CD8+ T cell immune responses . While intra-lymph node priming with DNA (plasmid) and boosting with peptide afforded a robust expansion of epitope-specific CD8+ T cells (on the order of 1/2 - 1/10 specific T cells/total CD8+ T cells), reversing the order of the vectors resulted in a limited overall T cell expansion (~1/100 - 1/1000 or less, of specific T cells/total CD8+ T cells) within the same range of homologous prime-boost vaccination . A closer look at the immunity primed by plasmid showed that, in stark contrast to peptide priming, the epitope-specific CD8+ T cells, although few in numbers (~1/100 specific/total CD8+ T cells), had some strikingly distinguishing features. Within the population of CD8+ T cells initiated by plasmid, we found a significant frequency of the lymphatic migration marker CD62L+ (central/lymphoid-memory) epitope-specific CD8+ T cells with a limited capability to produce proinflammatory cytokines upon peptide stimulation ex vivo. Nevertheless, these DNA vaccine-primed cells showed long-term persistence in vivo and displayed a high expansion potential following in vivo or in vitro re-exposure to antigen, associated with a rapid loss of CD62L and a broadening of their functional capabilities .
The finding that the low PD-1 expression profile afforded by DNA vaccination could be reproduced by intra-lymph node immunization with limited amounts of peptide and TLR stimulation sheds light on the mechanism of action of DNA vaccines and their potency as priming agents in terms of: i) the importance of extended yet reduced levels of antigen exposure; and ii) a role for TCR-independent stimulation through TLRs. However, it should be noted that within this model (Figure 4 and 5) DNA vaccines alone have a limited capability to elicit robust immune responses in homologous prime-boost regimens, as supported by experimental clinical observations as well as mechanistic studies [15–17]. Instead, we argue that the use of DNA vaccines for the purpose of priming high quality antigen-specific CD8+ T cell responses is a viable and highly promising strategy. For example, one could envisage alternating the administration of a DNA vaccine with other vectors such as peptides, recombinant proteins, or viruses for the purpose of inducing and periodically replenishing low PD-1-expressing central-memory T cells and then, through boosting, maintaining a pool of highly functional effector cells. Thus, such heterologous prime-boost regimens would ensure the presence of desirable T cell populations over a longer interval, prevent overall immune exhaustion, and maximize the clinical effect in a therapeutic setting such as cancer, where endogenous antigen exposure alone may not be sufficient to initiate or maintain a clinically relevant immune response.
The body of evidence discussed in this review supports three major conclusions. First, a heterologous prime-boost vaccine should ideally encompass a priming regimen that results in the induction of specific T cells co-expressing low levels of inhibitory receptors. Thus, following a heterologous boost (even within a short time-frame), these cells would expand and differentiate into effector cells rather than being subjected to negative regulatory mechanisms. Secondly, emerging data suggests that DNA vaccines have the capability to elicit low PD-1 expressing CD8+ T cells of central-memory phenotype, a process reproduced by repeat intra-lymph node exposure to minute levels of antigen in the presence of robust TLR9 stimulation. Third, this evidence points to a new dimension of immune homeostasis determined by a tight and synchronized control of inhibitory molecule expression by CD8+ T cells during antigen exposure. This facet of immune homeostasis would shape - as a function of antigen exposure and co-stimulation - the delicate balance between long-lived, readily expandable CD8+ T cells and short-lived T cells that are subject to exhaustion or other negative regulatory mechanisms, in a manner fitting the immunological threat.
Key prerequisites for an effective immune response-to control disseminated tumors for example-are not only the sheer numbers of tumor-associated antigen (TAA)-specific T cells but their quality or capability to recognize and eradicate cancerous cells. The latter depends on the functional avidity of the T cells  as well as their polyfunctionality  in an environment plagued by immune evasion mechanisms . An interesting fact is that the induction of high magnitude immunity, generally requiring exposure to significant antigen doses, may result in a lower proportion of high avidity T cells [61, 62]. This is quite important since tumor cells as well as chronically infected cells may display significantly reduced amounts of antigen which are 'invisible' to vaccine-specific T cells displaying low functional avidity, yet readily quantifiable with current immune monitoring techniques .
have low expression of co-inhibitory receptors (PD-1);
display a central memory phenotype;
have a high TCR functional avidity.
This new paradigm assumes that the selection of vectors is such that it would not result in a deleterious anti-vector immunity. The priming strategy could then be matched with heterologous vectors that expand and/or differentiate the primed cells to therapeutically useful effector T cells or, alternatively, with homologous boosting leading to much higher antigen exposure than during priming. Notably, the latter, which could be a less expensive strategy since it relies only on one vector, is supported by the observation that exposure to gradually higher levels of antigen (starting from minute amounts) over a fairly short interval of just a few days achieved an unexpectedly robust immune response , usually only attainable by live virus infection or heterologous prime-boost vaccination. A similar principle could be applied to homologous prime-boost regimens encompassing naked DNA as primer followed by electroporated DNA as a boosting agent . Effective priming may also be achievable through intradermal delivery of DNA as shown in a model of human skin tattooing .
In light of the scarcity of antigen-specific immune interventions that achieve clear-cut therapeutic benefits in cancer and chronic infections, there is clearly a need for advanced vaccine approaches that undergo rigorous testing and afford objective, quantifiable clinical responses. The paradigm outlined in this review shifts the focus from the overarching objective of inducing high numbers of vaccine-specific lymphocytes to that of generating highly efficacious T cells that are potent in adverse environments brought about by continuous antigen exposure or non-antigen related immune inhibitory mechanisms. Furthermore, these observations warrant a revision of current immune monitoring approaches in an effort to more accurately measure, predict and optimize the efficacy of active immunotherapies.
Mounting evidence supports a different model defining the mechanisms of heterologous prime-boost immunization at the epitope level. In summary, effective priming necessitates low PD-1-expressing central memory T cells and boosting results in their expansion and conversion to effector T cells equipped with broad migratory and functional capabilities. This mechanism is most likely linked to a new dimension of immune homeostasis with a possible role in ensuring the 'response-readiness' of CD8+ T cells, depending on the nature and magnitude of the immunological threat. Finally, this paradigm suggests a series of valuable criteria to guide the design of new immunization regimens.
We acknowledge the contribution of our collaborators: Mayra Carrillo, Diljeet Joea, Xiping Liu, Uriel Malyankar, Brenna Meisenburg, Robb Pagarigan, Angeline Quach, Darlene Rosario, and Victor Tam for generating some of the key experimental evidence in support of the model put forward in this review.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.