Slow but real progress on antibodies against HIV

There is wide consensus among AIDS vaccine researchers that we need an HIV vaccine to induce high levels of neutralising antibodies against a wide range of HIV strains, of the kind that are actually transmitted from person to person. Beyond this, even moderate levels of imperfect antibodies could be enough to change the course of disease and allow other immune mechanisms to take effect. The bad news from the AIDS Vaccine 04 meeting which has now concluded in Lausanne is that we still have no such vaccine; the good news is that methods to develop one are available and beginning to show results.

As shown by the first Phase III HIV vaccine trials, native versions of the main HIV-1 envelope protein, gp120, as in VaxGen’s AIDSVAX, do not elicit protective immunity against HIV. However, that and the other envelope protein, gp41, can undoubtedly be modified to make them more active as vaccines. Variants with deletions in the "variable loop" regions or with particular sugar groups removed from the surface are being tested, and results from some animal studies have been reported. Another approach is to combine gp120 with a protein designed to act like the CD4 receptor on human cells in exposing parts of the virus which are needed for binding to other cell receptors. These active regions, normally hidden from the immune system, are much the same across all HIV-1 strains - and even HIV-2. However, there is continuing doubt over how effective antibodies can be, given that the exposure of these regions may typically be brief and only when the virus is very close to a target cell.

A reliable method now exists for cloning B cells and making human monoclonal versions of antibodies produced - for example - by people with HIV. This is feeding into a number of promising approaches. In particular, it has greatly facilitated work on “broadly neutralising” monoclonal antibodies, active against many different HIV strains. These are commonly seen - though typically at low levels - in established HIV infection. HIV, once established in the body, can mutate to escape any and every antibody. Nevertheless, animal models show that such antibodies can prevent infection if given before or shortly after exposure to the virus. One study, reported in Lausanne, found that even 24 hours after exposure, antibodies could modify the course of infection. A clinical trial in Durban, South Africa, is about to test a cocktail of three antibodies given to breast-fed babies as a method of preventing transmission from their HIV positive mothers.

The best of the monoclonals show activity against all HIV-1 strains tested, but their potency is weak compared to that seen in neutralising antibodies against other viruses. Dr Dennis Burton of the Scripps Institute in California reviewed two of these in particular, directed at a sequence of 20 amino acids at the “membrane proximal repeat” end of the gp41 envelope protein, which have been studied intensively.

In the vaccine context, such antibodies can be used in screening systems for potential antigens to induce similar antibodies, when given as a vaccine.

The International AIDS Vaccine Initiative, IAVI, is funding a biotech company, MaxyGen, to create and screen new versions of gp120 using what they are calling “gene shuffling”. This starts with a family of related gene sequences, as found in nature, then makes breaks at various points. The fragments re-assemble themselves, combining many short lengths from different versions of the gene in the same order as they were in the original. The difficult step is to separate and screen the many new versions of the gene and find those which have desired characteristics - in this case, strong binding to particular anti-HIV antibodies. The most promising lines are prepared as prototype vaccines and administered to animals (rabbits are being used) to see if they can, in turn, make antibodies against HIV. The process can be repeated, as many times as desired, selecting the best candidates each time.

Preliminary findings are that shuffled genes can generate sequences that have never been seen in natural HIV isolates and which are better than the original ones in generating neutralising antibodies. In a way this is not surprising, since any natural HIV strain which gave rise to a strong and effective antibody response would quickly be eliminated. The next step in this programme will be to take one or more of the best gene lines as identified in rabbits and test them in monkeys.

A related and equally practical approach to creating neutralising antibodies has taken a short peptide from the envelope protein gp41 which is the target for one of the most broadly-active neutralising antibodies, and inserted it - together with a wide variety of flanking sequences - in a rhinovirus (the main cause of the common cold). The location chosen for the gene insertion means that 60 copies of the gene are shown on the outside of the virus particle. Having created a large collection of different variants of these viruses, they are then screened for their ability to stimulate neutralising antibodies. Those antibodies can then be compared for their strength and breadth of activity against HIV. While the approach is still a long way from entering clinical trials, the pathway should be relatively straightforward. That said, this particular vaccine will probably need to be combined with something else, designed to give cell-mediated help to the antibody response to HIV.