AIDS Vaccines

2.1

HIV-Specific Neutralizing Antibodies

Early animal model studies with broadly neutralizing monoclonal antibodies showed that these could protect against transmission. However, the extreme genetic variability, particularly in the envelope protein, not only facilitates escape from these responses, but has generated extreme diversity globally which will require the broadly cross-reactive neutralizing antibodies to be generated by vaccines. In natural infection some degree of cross-reactive antibody responses are generated in 50% of infected persons, but only about 1% of infected persons become “elite neutralizers,” able to potently neutralize viruses across at least four different clades. Adding to the challenge of generating neutralizing antibodies to HIV is the finding that only a limited number of envelope spikes is present on virions, and these are covered by extensive and rapidly shifting glycans, making antibody recognition of conserved sites very difficult.

Despite these challenges, there is renewed optimism about vaccine-induced generation of broadly neutralizing antibodies from studies of natural infection. Major advances in identification of broadly neutralizing antibodies and their targets initially arose from high throughput neutralization assays that screened memory B cells from chronically infected Africans. Parallel advances in B-cell cloning techniques have led to isolation of ever more potent broadly neutralizing antibodies (reviewed in Refs. ), and have revealed unique properties of these antibodies and facilitated identification of sites of vulnerability on the envelope glycoprotein. Potent broadly neutralizing antibodies typically take years to develop, require extensive somatic hypermutation, and have unusual characteristics such as long HCDR3 regions and often framework mutations as well.

With the isolation of broadly neutralizing antibodies, at least five sites of vulnerability on the envelope trimer targeted by these antibodies have been identified, with additional ones likely to be revealed: the CD4 binding site, the V3 glycan, the V1V2 glycan, the membrane proximal external region, and a region spanning gp120 and gp41 that is trimer specific. It is clear that responses to each of these regions can be generated by natural infection, but the ability to produce these through immunization remains elusive. Moreover, there may be genetic limitations since only a subset of variable heavy and variable light chain combinations are able to bind to these sites of vulnerability due to structural constraints. However, adoptive therapy with BNAbs in infected humans has been successful in lowering viral load, giving clear evidence that in the right amounts at the right places these could be effective.

2.2

HIV-Specific Nonneutralizing Antibodies

Although broadly neutralizing antibodies are not induced by current vaccines, nonneutralizing antibodies are readily induced. These antibodies bind to envelope epitopes that are not present on the native envelope trimer, but rather target epitopes that are revealed as the trimer dissociates, including gp120 monomers, nonfunctional conformationally rearranged envelope proteins, and gp41 stumps that are left as gp120 dissociates. Nonneutralizing antibodies have been identified against both gp120 and gp41, and are of particular interest since they have been shown to be a correlate of risk in the HIV vaccine trial that showed at least modest efficacy (given later). In the case of nonneutralizing antibodies, antiviral efficacy is linked to antibody-dependent cellular cytotoxicity (ADCC) mediated through the Fc portion of the antibody, or to Fc-mediated phagocytosis, and other potential mechanisms. Monoclonal nonneutralizing antibodies have recently been tested in a SHIV challenge model, and showed a decrease in the number of transmitted founder viruses that was statistically significant compared to controls. However, other animal model studies have shown lack of protection with nonneutralizing antibodies. Evidence for antiviral activity of nonneutralizing antibodies is very weak and not close to the complete protection against infection and suppression of viremia induced with neutralizing antibodies. The extent to which nonneutralizing antibodies contribute to relative control of natural infection or would contribute to vaccine-mediated protection is unclear.

2.3

HIV-Specific CD8 ⁺ T Cells

Also referred to as cytotoxic T lymphocytes, these cells recognize infected cells through their unique T-cell receptor engaging with the complex of a viral peptide and the cellular HLA Class I molecule on an infected cell. In vitro studies show that infected cells can be recognized before infectious virus progeny are produced, at least under experimental conditions. Even a single viral peptide–HLA complex is sufficient to induce lysis. Genetic studies indicate that certain HLA Class I alleles are associated with protection, implying that there may be genetic limitations on vaccine efficacy. Importantly, the HIV-specific CD8 ⁺ T-cell response does not recognize free virus, so it cannot be expected to prevent the initial round of infection, but at best would be expected to contain viral replication once cells are infected.

Much has been learned about these cells from natural infection. In acute infection, there is a massive induction of these cells, representing up to 80% of circulating CD8 ⁺ T cells in some cases, which then contract despite ongoing viremia. The greater the peak magnitude and the more rapidly peak levels are achieved, the lower the subsequent viral set point; antiviral efficacy is also demonstrated by the rapid generation of mutations with the 8–10 amino acid epitopes targeted by these cells. Numerous studies indicate that enhanced antiviral efficacy is associated with targeting of Gag, possibly due to the combination of early presentation of Gag peptides on infected cells and constraints on mutations in Gag : CD8 ⁺ T cell–induced mutations lead to fitness constraints, which likely contributes to relative immune containment. However, the antiviral efficacy of these responses is limited by escape mutations as well as upregulation of negative immunoregulatory molecules that impair function.

2.4

NK Cells

NK cells are typically considered to be part of the innate immune response, contributing to antiviral control by killing of virus-infected cells through ADCC. In the case of HIV, these cells kill virus-infected cells following HIV-mediated downregulation of HLA Class I, and HIV-induced upregulation of stimulatory NK ligands cells. NK cells also produce antiviral chemokines CCL3, CCL4, and CCL5. The potential relevance of these cells to vaccine strategies comes from an extension of studies in mice suggesting that NK-cell memory can be induced. In a recent study in SHIV- and SIV-infected macaques, it was shown that splenic and hepatic NK cells were able to specifically lyse dendritic cells pulsed with Env and Gag, whereas this was not observed in cells derived from uninfected macaques. Importantly, splenic and hepatic NK cells obtained from macaques immunized 5 years earlier with a recombinant HIV-adenovirus vector lysed antigen-matched but not antigen-mismatched target cells. The demonstration that these memory NK-cell responses can be induced by immunization suggests that they might be exploited by future vaccine regimens.

2.5

CD4 ⁺ T Helper Cells

Virus-specific CD4 T helper-cell responses are critical for induction and maintenance of both CD8 ⁺ T-cell and B-cell responses to viruses. However, these same cells are preferentially infected with HIV. Indeed vaccine-induced CD4 ⁺ T-cell responses have in at least one instance been shown to enhance infection and progression in an animal model of AIDS virus infection. In this case, CD4 ⁺ T-cell responses were induced in the absence of neutralizing antibodies and HIV-specific T-cell responses. Balanced CD4 ⁺ - and CD8 ⁺ T-cell induction may be critical, as suggested in recent studies in mice in which selective induction of virus-specific CD4 ⁺ T-cell responses in the absence of virus-specific CD8 T-cell responses induced massive inflammation and immunopathology. A particular subset of these cells, T follicular helper cells, are critical for providing help to B cells in the process of affinity maturation, and since they correlate with the development of broadly neutralizing antibodies in persons with chronic untreated infection, are likely to be a key component of an effective vaccine.