HIV vaccine technologists edge nearer to effective designs

A vaccine symposium held at the sixth International AIDS Society conference (IAS 2011) in Rome heard how designers are slowly developing HIV vaccines designed to overcome problems that had prevented the creation of an effective vaccine in the past.

Barbara Ensoli of the Italian Istituto Superiore di Sanitá outlined the barriers to success in a field which has seen only three large efficacy trials in humans in the last 15 years.

Vaccine trials past and present

In 2003, a trial of a first-generation vaccine, AIDSVAX, a simple vaccine made from HIV proteins and designed to generate antibodies to the virus, failed to demonstrate efficacy. The problem with HIV is that the antibodies that are generated in infections, and in response to a conventional vaccine, do not contain the virus’s capacity to mutate its way round the antibody response. 

In 2007 STEP, a trial of a more complex vaccine consisting of HIV components encased in the shell of another virus (a ‘vector’, designed to take it into cells), also failed. This second-generation vaccine was designed to simulate the other branch of the immune system and generate cells that destroy HIV-infected cells, but here is some evidence that instead it may have made some people who already had immunity to the viral shell more vulnerable to HIV.    

In 2009 RV144, a trial using a composite vaccine of a ‘prime’ dose of a vector vaccine plus a ‘boost’ dose of the original AIDSVAX finally produced a weakly positive result. It reduced HIV infections by 31% compared with placebo, though it was not effective against people who had had multiple HIV exposures, and protection faded with time. This reinvigorated the search for a vaccine.

However, Barbara Ensoli commented, it was still unclear how the vaccine generated an immune response and because of this, no large efficacy trials are currently being conducted until developers are more confident that they have something that will generate an effective response.

She said that there were currently about 50 trials of 42 vaccine candidates taking place worldwide, though 40 of these trials were small phase 1 safety trials. There were 22 trials of vaccines using non-replicating viral vectors (plus or minus other boost doses), ten using artificially-generated DNA, three using HIV proteins, two using replicating viral vectors, and a handful of miscellaneous candidates. Twenty-eight used a multiple-dose prime-boost design.

Promising results – a review

The most promising results in recent years have been seen in animal studies, though the STEP vaccine is an example of one that produced positive results in some animal studies but failed in humans.

A vaccine consisting of the HIV cell-entry protein gp41 inside an artificial vector, dosed twice as an intramuscular infection then twice as a nasal spray, protected five monkeys from vaginal challenge with 13 doses of the monkey/human spliced virus SHIV (Barnett 2008). The interesting thing about this vaccine was that it did not produce antibodies to SHIV in the blood, but only in the mucous membranes lining the vagina – a potential ‘vaccine microbicide’.

A vaccine reported on aidsmap consisted of SIV (monkey HIV) components packaged inside an actively replicating vector consisting of the shell of the common CMV virus (Hansen). Whereas non-replcating vectors only carry vaccine components into cells, replicating vectors spread the vaccine like an infection into other cells - an approach with some safety concerns. This vaccine, given to 26 monkeys, did not protect them from SIV infection but in 13 of the monkeys resulted in an infection characterised by a low viral load – which gradually turned into a situation in which there was no detectable virus in their bodies. This, if achieved in humans, would essentially be a ‘vaccine cure’. But we need to understand why only half of the monkeys responded and why.

A vaccine consisting of the two HIV components tat and env inside another actively-replicating viral vector produced a similar effect, preserving the CD4 count and bringing the viral load down to undetectable in five monkeys, and down by a factor of 10,000 in the other three out of eight infected with a highly pathogenic variety of SHIV (Demberg). By comparison, monkeys receiving placebo maintained viral loads of a million and their CD4 count crashed to 10 cells/mm³ within two months. A phase 1 trial of a human version of this vaccine in 50 people is planned.      

How to generate antibodies

The most elusive but potentially effective goal in HIV vaccine development is to design a vaccine that would produce what are called broadly neutralising antibodies (BnAbs).  These are antibodies that neutralise a wide variety of different viral strains – in contrast with most anti-HIV antibodies, whether vaccine-generated or natural, which are too specific to individual strains. Broadly neutralising antibodies have been isolated from the blood of some people who control HIV naturally and even reproduced in the laboratory, but how to design a vaccine that induces the body to generate then has been a huge challenge.

Susan Zoller-Pasner of the New York University School of Medicine described how her team was developing vaccines by analysing which parts of the HIV surface protein, env, were ‘highly conserved’. These are parts of the protein that cannot mutate without affecting viral viability. They were developing vaccines based on highly-conserved sequences from two of the prominent ‘bulges’ on the surface of the env protein, the V3 and V2 loops. The ‘V’ in these terms stand for ‘variable’ as they are amongst the areas of HIV most quick to mutate away from immune surveillance, but in fact there are sequences that are common to nearly all strains and should in theory generate BnAbs. An antibody response to the V2 loop appears to be part of what gave the RV144 vaccine its efficacy.

So far although they have managed to generate test-tube responses that protect against 50% of a panel of viruses susceptible to neutralisation, they have yet to succeed against more evolved viruses that shrug off antibodies. The important thing, however, is that Zoller-Pasner’s team has found a way to rationally design vaccines that may generate BnAbs, rather than rely on serendipity.    

In connection with this work, Susan Barnett, senior scientist at the drug company Novartis, described further details of the intramuscular and intranasal vaccine described above, which induced a protective mucosal response in monkeys. This work has led on to developing HIV protein-based vaccines which are now going to be used as the ‘boost’ for in the vaccine intended to be used in the follow-on to the RV144 trial.  

The company are going to produce 50,000 of an artificially-generated composite viral envelope protein isolated from 22 different subtypes of HIV to use in the follow-on vaccine.

DNA vaccines

David Weiner of the University of Pennsylvania described the latest steps in the development of a different kind of vaccine – DNA vaccines. These vaccines do not employ any naturally occurring component, however adapted, but consist of ring-shaped packages of artificial sequences of DNA.  DNA vaccines are relatively simple to manufacture but so far have had poor immunogenicity in humans; it is difficult to get cells to mount a strong reaction to the DNA and difficult to get the DNA into the cells in the first place. Weiner described how his team is developing technologies to increase the amount of DNA produced, using immune-stimulating chemicals to provoke cells into reacting to the DNA, and even using electrical stimulation of the same type used in in vitro fertilisation to ensure the vaccine becomes incorporated into cells that can then produce anti-HIV antibodies.

A number of phase 1 studies of DNA vaccines had produced long-lasting immune responses in humans – partly because there is no viral vector that the immune system can ‘get used’ to – and more studies of what will hopefully prove to be increasingly potent DNA vaccines are planned.