NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
Coffin JM, Hughes SH, Varmus HE, editors. Retroviruses. Cold Spring Harbor (NY): Cold Spring Harbor Laboratory Press; 1997.
Humoral Immune Responses
Neutralizing Antibodies
Neutralizing antibodies are defined by their ability to block, or neutralize, the infectivity of virus. Neutralizing antibodies can act by preventing the virus from binding to its receptor or by blocking entry events subsequent to receptor binding. Several lines of evidence argue for an important role for neutralizing antibodies in limiting lentiviral replication during early and asymptomatic stages of infection (for review, see Haynes 1992). The observation that variants which are resistant to neutralization are selected in infected individuals (see below) argues that neutralizing antibodies serve to limit the replication of HIV, SIV, and other lentiviruses in vivo (Albert et al. 1990), although the extent of this limitation cannot be determined. Passive treatment with human immunoglobulin containing HIV-neutralizing activity or with a specific neutralizing monoclonal antibody has protected chimpanzees against intravenous challenge by live HIV-1 (Prince et al. 1991; Emini et al. 1992). Definition of the action and targets of neutralizing antibodies is of considerable importance for understanding the mechanisms of clinical latency and disease progression and for the development of prophylactic vaccines and immune-based therapies (see Chapter 12.
The targets of neutralizing antibody to HIV-1 have been more extensively studied than those of any other lentivirus. The first target, or epitope, to be identified (and the best defined) is variable region 3 (V3) of the surface glycoprotein gp120 (Javaherian et al. 1989; for review, see Haynes 1992). This region is contained within disulfide-linked cysteines between amino acids 300 and 330 in the linear sequence. Synthetic peptides based on this sequence can elicit antibodies in animals that neutralize laboratory strains of HIV-1 and inhibit the fusion of infected and uninfected CD4+ cells (Javaherian et al. 1990). In some infected individuals, at least a portion of the HIV-1-neutralizing activity appears to be directed to the V3 loop. The V3 regions of some but not all SIVsm/HIV-2/SIVmac isolates have also been found to serve as targets of neutralizing antibodies (Bjorling et al. 1991; Javaherian et al. 1992). Neutralizing antibodies to V3 frequently recognize linear epitopes, but some V3 epitopes may be conformational or discontinuous in nature (Ho et al. 1991; Steimer et al. 1991). Because of the high level of V3 sequence variation, neutralizing antibodies to V3 are usually type- or strain-specific, i.e., they are capable of neutralizing only a limited number of isolates. However, sera with high-titer antibody to HIV-1 V3 peptides do not always neutralize HIV-1 with a homologous V3 (Warren et al. 1992). The ability of V3 to serve as a target for neutralization may be dependent on the other sequences in the envelope; additional studies are needed to clarify this issue.
Another class of neutralizing antibody blocks the binding of gp120 to CD4. These antibodies do not appear to belong to a single class with a single target but rather are a complex group capable of blocking interaction with CD4 by reacting with any of a number of distinct or overlapping epitopes on gp120. Antibodies that recognize determinants in constant region 4 (C4) are members of this class. The vast majority of CD4-blocking antibodies, however, do not recognize linear epitopes but instead react with complex, discontinuous, or conformational targets. Sequences important for recognition by antibody can be derived from regions of the gp120 molecule that are widely separated in the linear sequence, including C1, C2, C3, C4, and V1–V2 (see Chapter 3. gp120 residues that contact DD4 are important for recognition by such antibodies (Thali et al. 1991). CD4-blocking antibodies are usually more broadly reactive than the anti-V3 antibodies. Neutralizing antibodies have also been identified that bind epitopes that are better exposed after gp120 binds CD4 (Sattentau and Moore 1991; Moore et al. 1994).
Although anti-V3 and CD4-blocking antibodies are the most extensively studied, other types of neutralizing antibodies have been described. These include neutralizing antibodies to V1, V2, V5, C2, and gp41 epitopes of HIV-1 and V1–V2, V4, and gp41 epitopes of SIVmac/HIV-2/SIVsm (for review, see Nixon et al. 1992). Although the V3 loop has often been referred to as the “principal neutralizing determinant” of HIV-1, there is no clear evidence to suggest that V3 neutralization epitopes are more important than other epitopes during the natural course of HIV-1 infection in humans. It is not yet known to what extent the neutralizing antibodies that can be found in patient sera are directed at linear versus conformational epitopes, variable versus conserved, or V3 versus other domains. Some studies have suggested that conformational and discontinuous epitopes may predominate in patients over linear epitopes (Steimer et al. 1991; Moore and Ho 1993; Moore et al. 1994). Since neutralizing antibodies represent only a small fraction of the total antibody response to gp120 and since neutralization titers in human sera are generally low, it has been difficult to quantify these different types of antibody responses. The situation is clouded by the fact that neutralization assays have generally been run with one or a few strains of virus which may or may not be representative of the complex mix of viral genotypes present in the infected individual. Primary HIV-1 isolates are much more resistant to neutralization than the laboratory strains most frequently used for these assays and may differ in their principal targets of neutralization (Moore et al. 1995; Moore and Ho 1995; T. Matthews et al., unpubl.).
Escape from Neutralization
During the past two decades, the concepts of antigenic variation and humoral immune selection during persistent infection have been developed experimentally in three lentivirus systems: visna, CAEV, and EIAV (see Chapter 10. EIAV disease is episodic in both naturally infected and experimentally infected horses. EIAV isolates obtained at different time points following primary infection, corresponding to different clinical episodes, can be neutralized by serum samples taken sequentially from the same horse. Successive viral isolates are not susceptible to previously neutralizing antibody obtained earlier in the course of infection, but they are susceptible to antibody obtained several weeks later (Kono et al. 1973). Thus, the appearance of new variants that are resistant to neutralization always precedes the appearance of antibody capable of neutralizing the variants. These data suggest that the neutralizing antibody response of the host can select for neutralization-resistant mutants during the course of infection of a single animal and imply that antigenic variation has an important role in persistent viral replication in chronically infected animals. The relative importance of neutralization-resistant HIV to the persistence of the infection remains to be established; the nonepisodic nature of HIV persistence suggests that there are important infectious differences between HIV and EIAV.
Analyses of sequence data from monkeys experimentally infected with cloned SIV and from humans infected with HIV-1 indicate that variation in the five discrete variable domains of gp120 results from selective pressure (for review, see Burns and Desrosiers 1994). In SIV, greater than 95% of the nucleotide substitutions in the five variable regions are nonsynonymous, i.e., they cause a change in the encoded amino acid. Statistically significant differences are seen between observed frequencies and frequencies expected from random genetic drift. The differences in the frequencies of nonsynonymous and synonymous mutations in env are markedly different from the rates for eukaryotic genes. In eukaryotic genes, nonsynonymous mutations accumulate at about one-fifth the rate of synonymous mutations. The HIV-1 gag gene behaves more like a cellular gene than does env and shows a bias toward synonymous mutations. Since random mutations cause synonymous changes about 30% of the time, these data imply that the env genes have been subjected to strong selective pressure to change, whereas this selective pressure for cellular genes remains unchanged. The synonymous mutations in gag presumably result from a more stringent limitation on the amino acid changes that still allow optimal protein function. In contrast, variant Env proteins appear to be able to tolerate almost any amino acid change in the discrete variable domains. This appears to provide a significant selective advantage in vivo. The host immune response is probably the major source of the selective pressure for change, but some of the changes may reflect selection for altered tropism.
Attempts to demonstrate the appearance of neutralization-resistant HIV-1 variants in infected humans have been hampered by lack of information on the genetic sequence of the initial infecting strain and by the complex mix of genotypes present in infected humans when samples were taken for study. Nevertheless, several studies have indicated that neutralization-resistant HIV-1 variants emerge during the course of infection (Albert et al. 1990; Tremblay and Wainberg 1990). Much clearer evidence for the emergence of escape mutants has been obtained in rhesus monkeys experimentally infected with molecularly cloned SIV (Burns et al. 1993). Sequential sera from infected animals showed much higher neutralizing antibody titers to the cloned virus used for infection than to the variants obtained 1–2 years after infection. Monkeys were subsequently infected with cloned variants obtained from the original animal and each developed antibodies that preferentially neutralized the particular variant with which they were infected. Only 20 amino acid changes in gp120, mostly in the discrete variable domains, were responsible for the resistance to cross-neutralization.
In summary, there is clear evidence for selective pressure in discrete variable regions of the envelope during the course of SIV and HIV infection. The neutralizing antibody response of the host is likely to be a major selective force driving sequence variation in env. These data also suggest that there may be a role for neutralizing antibodies in limiting the replication of HIV and SIV during the lengthy asymptomatic stage of infection. The coincidence of neutralization domains of HIV-1 and/or SIVmac/HIV-2 with variable regions in gp120 and gp41 also suggests that there is a direct relationship between neutralization domains and the sequence variants that emerge.
Other Types of Antibody Responses
Antibody-dependent cytotoxicity (ADC), interfering antibodies, and enhancing antibodies have also been detected in lentiviral infections. Antibodies can lyse infected cells and virions via complement-mediated ADC. Interfering antibodies may bind to virions or infected cells and in so doing block the action of antibodies that can neutralize infectivity.
Some antibodies may bind to virions and in so doing enhance viral infection. Enhanced infection can be mediated by either complement or Fc receptors. The importance of enhancing antibodies has been clearly documented in dengue virus infection; previous exposure to one serotype of dengue can enhance the infection and disease during a secondary infection with a different serotype (Haase 1986; Takeda et al. 1988; Halstead 1989; Kurane and Ennis 1992). Complement or Fc receptor may only bring virus-antibody complexes into close proximity to cellular CD4 for binding and entry into the target cell. Serum from SIV-infected rhesus monkeys stimulates SIV infection in vitro, suggesting that enhancing antibodies might be important for the spread of HIV in vivo (Montefiori et al. 1990). The potential for induction of enhancing antibodies needs to be carefully considered in the development of an HIV vaccine since it is possible that experimental vaccines may induce antibodies that enhance infection rather than prevent it (see Chapter 12.
Cellular Immune Responses
Antigen-specific cellular immune responses include the activities of T lymphocytes that lyse target cells expressing a foreign antigen and T lymphocytes that provide help for B cells and effector T cells. Although some CTLs may have a CD4+ phenotype, the great majority of CTLs are CD4–, CD8+. CD4+ T lymphocytes provide help to some CTLs and B cells in the form of secreted lymphokines, growth factors, and other facilitatory molecules. The specificity of these T-cell responses is determined not only by the specificity of the TCR, but also by the MHC molecules found on most cell types that present peptide antigens processed from viral proteins for recognition by the TCR (see Chapter 12. Each individual expresses a set of genetically determined polymorphic MHC molecules that with rare exception (monozygotic twins and extraordinary coincidence) distinguish each individual. The range of epitopes to which an individual is able to mount T-cell responses is thus said to be “MHC-restricted.” T-cell receptors on the surface of CD8+ CTLs recognize foreign antigens when they are presented as 8–11-amino-acid peptides in the exposed groove of class I MHC molecules present on most cell types. T-cell receptors on the surface of CD4+ T cells recognize foreign peptides presented in the exposed groove of class II MHC molecules, which are found only on specialized antigen-presenting cells (B cells and macrophage-related cells). The peptides presented by class I MHC molecules are generally produced from processing of intracellular proteins (the endogenous pathway). Those presented by class II MHC molecules are produced by internalization of extracellular proteins (the exogenous antigen-processing pathway).
The importance of CD8+ CTLs in limiting or suppressing viral replication has been documented in a number of systems. Adoptive transfer of CD8+ CTLs into syngeneic, inbred strains of mice provides protection against certain viral infections, such as lymphocytic choriomeningitis virus (Baenziger et al. 1986; Oldstone et al. 1986; Koszinowsky et al. 1987; Oukka et al. 1996). A role for CTLs in limiting lentiviral infections has largely been inferred from such studies of other viral systems. The difficulty in approaching the role of CTLs in lentiviral systems is that there are no mouse lentiviruses, and the lack of inbred strains in most other species has precluded adoptive transfer experiments. The adoptive transfer of human anti-HIV CTLs into SCID mice reconstituted with human lymphoid cells provides some degree of protection against subsequent HIV infection (Van Kuyk et al. 1993).
CTL activity against HIV proteins can be readily detected in vitro by incubation of CD8+ T cells (or clones of such cells) with autologous (or MHC class I-matched) cells expressing HIV epitopes. Epitope-expressing cells can include HIV-infected cells, cells expressing one or more HIV gene products expressed for an appropriate vector (vaccinia vectors are commonly used for this purpose), or cells incubated with HIV-derived peptides (for review, see Walker and Plata 1990; Levy 1993). HIV antigen-specific CTL killing of cells can be detected by standard methods or by inhibition of viral replication in appropriate cell cultures (Yang et al. 1996).
CD8+ CTL activity can be directed against virtually all of the HIV-encoded protein products. MHC-restricted CD8+ CTLs can be found in large numbers during the asymptomatic stage of HIV-1 infection but decline with disease progression. Since CTL activity is dependent at least in part on help (cytokines, growth factors, etc.) from CD4+ T cells, the decline in CTL activity may be related to the decline in the number (and functionality) of CD4+ T lymphocytes (see above, Immunopathogenic Mechanisms of HIV Infection). Why CTLs fail to control HIV infection is not clear. Maintenance of the steady state requires that an infected cell must be able to release at least enough virus to infect one more cell before being killed. Recent data suggest that sequence variation allows the virus to escape the CTL response (Phillips et al. 1991; Nowak et al. 1995; Klenerman et al. 1996). However, additional studies are needed to confirm and extend these observations. In some cases, CTL epitopes can persist in the viral genome in the face of demonstrable CTLs directed against them (Bevan and Braciale 1995), implying that the CTL response may be simply inadequate even in the absence of escape. Highly conserved residues should be effective targets for MHC-restricted CTL responses. There may be a limited capacity for viral sequence variation to escape CTL responses, and there is good evidence for the presence of CTLs that recognize conserved elements. Recent studies, both with HIV and SIV, imply a correlation between the quality and magnitude of the CTL response in infected individuals and the rate of progression; higher CTL levels correlate with reduced viral load and a longer period of clinical latency (Cao et al. 1995; Pantaleo et al. 1995).
Some activated CD8+ cells suppress HIV and SIV replication by a mechanism other than cytolysis. This suppressive activity is mediated by a soluble factor or cytokine released by CD8+ cells and is not MHC-restricted. At least one set of relevant factors are the chemokines MIP-1α, MIP-1β, and RANTES (Cocchi et al. 1995), which act by interfering with the interaction of gp120 and the chemokine receptor CCR5. Other, as yet unidentified, factors may also be involved in the CD8 suppressive effect (Mackewicz et al. 1996).
Major expansions of CD8+ T cells, predominantly the Vβ class, occur transiently during the primary immune response to HIV (Pantaleo et al. 1994). HIV-specific cytolytic T cells are found within this population, and molecular cloning of the TCRs from this expanded population demonstrates that the expanded cells are oligoclonal, suggesting a response to a specific antigen. Individuals with a more pronounced and restricted expansion, i.e., major expansions of only one or two Vβ families, seem to have a more fulminant clinical course than those with a less pronounced and more diffuse expansion. These results provide experimental support for the idea that a more broadly reactive CTL response provides increased protection.
Antibody-dependent Cellular Cytotoxicity
Infected cells coated with antibody reacting to antigen can be recognized and killed by effector cells, usually natural killer cells or monocyte/macrophages bearing Fc receptors. Such ADCC has been documented in HIV-infected individuals. Although it is easy to imagine a role in controlling infection for this type of immune response, its relative importance has not been clearly defined.
- Immune Responses to HIV and SIV - RetrovirusesImmune Responses to HIV and SIV - Retroviruses
Your browsing activity is empty.
Activity recording is turned off.
See more...