AIDS vaccine: new ideas from old approaches

Abstract

If the global epidemic of HIV infection is to be controlled, a vaccine will be needed. Large-scale chemotherapy, with its high expense and difficulties with compliance, is for many countries not a realistic option. Unfortunately, progress towards a vaccine has been slow: attempts to generate protective antibodies with the recombinant HIV glyco-protein gp 120 seem to have failed. Attention is now turning to agents that stimulate cell-mediated immunity, but the value of such vaccines, even if they work, may be limited by virus variability and the need for repeated immunization.1 In this short paper I propose reconsideration of two approaches that have been explored by the group at the National Institute for Biological Standards and Control (NIBSC), led first by Jim Stott and now by Neil Almond. Both strategies have been somewhat neglected by the mainstream effort in the USA. Despite concerns about safety, this work may indicate a way forward. One of these approaches is the use of live attenuated vaccine, with a deletion in the nef gene, first described by Desrosiers and colleagues in the USA.2 This gave first-class protection in macaques against both free and cell bound SIV, and also against viruses antigenically quite different from the vaccine strain. The large drawback is that live vaccines of this kind have caused AIDS in recipient macaques—especially in immature animals, for which these viruses are intrinsically pathogenic. Also, reversion to wild type has been observed. Nevertheless, the vaccine is by far the most effective so far produced against SIV. Work proceeds at NIBSC on the mode of protection; neither arm of the conventional immune response provides a convincing explanation.3 The second approach is alloimmunization. When the facility to carry out experiments on SIV in macaques was set up at NIBSC, one of the first experiments was an attempt to replicate a strategy they had used for bovine respiratory syncytial (RS) virus,4 whereby animals were immunized with inactivated virus-infected bovine mucosal cells and then challenged with live virus. When protection was achieved, it was a simple matter to identify the antigens and the immune mechanism. For the SIV experiments they used a human cell line infected with the virus, stabilized with glutaraldehyde, inactivated with beta propiolactone, and administered with a purified saponin (Quil A) as adjuvant. When all the vaccinated macaques proved immune to intravenous SIV challenge, it seemed that a conventional vaccine was achievable5—a view strengthened by the observation that protection could be conferred passively by serum from protected animals. In the SIV experiments, unlike those with RS virus, none of the animals received an uninfected-cell control preparation; a human T cell line (C8166) was used both as substrate for the vaccine and as a vehicle for the challenge virus. This omission was understandable, because of the experience with RS virus infected cells and the cost of macaques. There were also ethical considerations in these experiments with primates. The essential aim was to confer protection; the mechanism would be worked out later. But the omission was regrettable because it gave an unprofessional feel to the early experiments, which prompted the view that the results were artifacts and not worth pursuing. Some workers felt the results cast doubt on the credibility of the SIV/macaque model, and this added to the customary difficulty of gaining funding for novel work. In fact the results have been confirmed by experiments that do include uninfected cells, and the degree of protection is impressive. It soon became apparent that protection correlates with host cell antibodies rather than antibodies to viral antigens; moreover even uninfected cells confer some protection against SIV.6 The identity of the protective host cell antigen has not been established. MHC class II antigens will do the trick7 but others are involved. The use of human cells in macaques represents xenoimmunization, which is not an option in man. Therefore, in the SIV model, the next step is to explore the effect of alloimmunization, by use of simian cells. Some limited progress has been made. For example, fixed inactivated uninfected peripheral blood mononuclear cells from two macaques conferred some protection as did HSC-4 cells (a macaque T cell line) infected with SIV.8 The probable mode of protection for this host-cell-component vaccine is complement-mediated lysis of the challenge virus. An obvious drawback is the potential for unwanted immunoreactivity. With both these vaccine approaches, the mechanisms of protection deserve further vigorous investigation.