Toward an HIV type 1 vaccine that generates potent, broadly cross-reactive neutralizing antibodies.

Abstract

P ROGRESS IN T HE D EV EL OPM E NT of an effective vaccine for HIV-1 has been gauged in large part by the ability to induce measurable virus-specific CD8 1 cytotoxic T lymphocytes (CTLs) and neutralizing antibodies as lead correlates of a protective immune response. 1±3 Although there has been success in generating low-level CTLs in a subset of vaccinated volunteers in phase I clinical trials, little progress has been made in generating neutralizing antibodies. Part of the problem for neutralizing antibodies is the high degree of genetic variability in the HIV-1 surface gp120 and transm embrane gp41 glycoproteins. 7 Although conserved neutralization epitopes exist, the current repertoire of candidate HIV-1 vaccines fails to generate sufficient antibodies to them. Critical features of the structure of the oligomeric envelope glycoproteins as they exist in their native form on the virus present yet another level of complexity for immunogen design. Most candidate HIV-1 envelope subunit vaccines are based on monomeric gp120 from T cell line-adapted (TCLA) strains of virus. Those vaccines have generated neutralizing antibodies with variable and sometimes potent activity against the vaccine strain, low activity against heterologous TCLA strains, and poor activity against primary isolates. It is conceivable that this narrow specificity, particularly the inability to neutralize primary isolates, will greatly limit the effectiveness of the antibody component of a vaccine. Scientists have been divided on the issue of whether the weak neutralizing antibody response generated by candidate HIV-1 vaccines is a property of the vaccine or the assay. A major argument has been that current assays underestimate neutralization potency, an argument that stems mostly from the fact that primary isolates are generally less sensitive to neutralization than TCLA strains. A concentrated effort has been made to resolve the vaccine-versus-assay issue by focusing mainly on the assay. We now recognize that primary isolates are relatively insensitive to neutralization under a variety of conditions, 27±30 including those that take into account the differential use of CXCR4 and CCR5 as coreceptors. 31±33 These developments, combined with features revealed by the crystal structure of gp120 and gp41, are making it increasingly clear that the dichotomy in neutralization sensitivity between TCLA strains and primary isolates is unrelated to the assay and actually reflects fundamental differences in the structure of the oligomeric envelope glycoproteins of these two categories of virus. Thus, repeated passage of HIV-1 in T cell lines has the consequence of selecting virus variants on which the neutralization epitopes are readily accessible for efficient antibody binding. Some of these epitopes reside in the V3 cysteine±cysteine loop of gp120, which is a major target for neutralizing antibodies on TCLA strains but not primary isolates. The Division of AIDS, Preventive and Vaccine Branch of the National Institutes of Health (Bethesda, MD), convened a workshop in October of 1998 to discuss new strategies to improve the neutralizing antibody responses generated by candidate HIV-1 vaccines. The goals of the workshop were four-fold: (1) consider the advantages and disadvantages of using genetically engineered cell lines to assess a neutralizing antibody response, (2) examine evidence from passive immunization experiments showing that potent neutralization is required for protection, (3) review the structure of gp120 and gp41 revealed by X-ray crystallography as it relates to neutralization and (4) discuss the merits of novel immunization strategies in early phases of development. A summary of this workshop is presented here.