YOU ARE HERE:
Recombinant sub-unit vaccines
A range of HIV proteins and peptides have been produced, usually in cell-culture systems, as potential vaccines for HIV. Initially, the main targets for vaccine developers were the viral envelope protein gp120, and its precursor gp160. It was hoped that antibodies targeting these proteins would prevent HIV entering human cells. While these viral proteins have been produced in a range of different cell types, including insect cells, it seems that to produce proteins closely resembling the virus in its natural state, they should be grown in mammalian cell lines.
More recently, vaccine developers have experimented with other HIV proteins, including regulatory proteins such as Tat, or modified versions of them such as 'Tat toxoids'. These are presented in ways that promote cellular immune responses rather than antibodies, and there are some studies which suggest that they can modify the course of disease in monkeys.
Envelope proteins
The first HIV vaccine to enter full-scale efficacy testing was the AIDSVAX recombinant gp120-based vaccine developed by the company VaxGen, headed by Dr Don Francis and created as a spin-off from the American biotechnology company Genentech.
This vaccine was designed to induce neutralising antibodies in the hope of preventing or aborting infection with HIV. It is based on gp120 proteins produced in Chinese hamster ovary cell cultures, using alum as an adjuvant. This strategy was used for the first successful hepatitis vaccine. Company researchers report that AIDSVAX is capable of generating strong memory responses although, in the initial trials, booster injections were given every six months so as not to rely on this.
One AIDSVAX version, based on two different isolates of subtype B viruses, was tested among people at risk of sexual transmission of HIV in the United States, Canada, Puerto Rico and Amsterdam. This trial recruited 5400 volunteers, of whom two thirds received the vaccine and one third had placebo injections. The trial began in 1998 and results were reported early in 2003. There was no evidence of protection among the trial volunteers as a whole. The company has expressed some excitement about supposedly better results among non-white, specifically African-American trial volunteers, but the biological rationale for this is not clear and the number of patients is relatively small, casting doubt on these conclusions. Furthermore, the statistical and biological assumptions behind the company's claim have been vigorously disputed.
A second trial, of an AIDSVAX formulation based on Thai subtype E and subtype B viruses, began in March 1999 and reported at the end of 2003. This study recruited 2500 volunteers in Thailand from among injecting drug users attending methadone clinics, but found absolutely no evidence of protection.
There is considerable scepticism among researchers outside the company as to whether useful levels of long-lasting antibodies can be induced. Early reports claimed ‘strong’ antibody responses and a safe and well tolerated vaccine (Belshe 2000). The company claimed that chimpanzees have been protected by such vaccines against some HIV strains and that with a series of vaccinations, increasing levels of increasingly broadly-active neutralising antibodies are observed in human volunteers. What is more, there are plenty of non-HIV vaccines which give useful levels of protection (through 'memory' B cells) even when detectable levels of antibodies are very low.
An alternative use for these vaccines, as a booster to be combined with another vaccine aimed at generating a cellular immune response to HIV, is discussed in the section below on canarypox vaccines (see Recombinant vectored vaccines ). One clinical trial using AIDSVAX B/E in this way is ongoing, although the likelihood of success is now considered to be extremely low.
Modified envelopes
The idea of a vaccine designed to induce neutralising antibodies has not ended with AIDSVAX. Two fundamentally different ideas have been developed, as pathways towards such a vaccine.
The first arises from understanding of the structure of the HIV envelope proteins and how they interact with proteins on the surface of the cells they infect (Liao 2004). It is clear that some regions of the HIV envelope must be conserved – in other words, stay very much the same among all versions of HIV – so that they can bind efficiently to their targets. However, these regions are effectively hidden from the immune system in several ways.
Firstly, there are sugar molecules attached to the on the surface of the viral protein. These stimulate much weaker immune responses than the amino acids from which proteins are made, so they act as a shield from the immune system. Consequently, one strategy for modifying the envelope protein is to remove some of the sugar molecules from the protein's surface and see if that leads to neutralising antibodies that can act against the virus (Hu 2004).
Secondly, there are ‘variable loop’ regions within the virus's proteins. In these regions, mutations and changes in the protein's structure have no effect on the virus's ability to replicate and cause disease, but they enable it to escape from immune responses directed against those regions. In effect, they act as decoys, distracting the immune system from more vital parts of the virus.
Because of this, a further strategy for modifying the virus envelope is to delete large parts of the variable loops, which are known as V1, V2 and V3. Studies have shown that removing parts of V2 and V3 seems to make for stronger antibody responses.
Thirdly, some of the most important parts of the protein are buried deep inside the molecule, and only become ‘visible’ to the immune system when the virus docks with the CD4 receptor. This gives rise to another strategy, where either a part of the CD4 protein or some other protein chosen to have a similar effect is combined with the HIV envelope so that the sensitive parts of the virus are fully exposed for an immune response.
There remains some doubt over whether this strategy will work, because the interaction only happens when the virus is very close to the cell surface, which may not leave enough room or enough time for an antibody to get in the way and block infection.
A different approach, which may or may not arrive at the same result, is to start from the discovery of broadly-active anti-HIV antibodies, which are produced in small amounts by many people living with HIV. Advances in methods for producing antibodies have made it relatively easy to take B-cells from individuals and keep them alive outside the body in cell cultures, leading to a small explosion of research in this area. Although the natural antibodies produced by B-cells tend to be weak by the standards of antibodies against non-HIV viruses, it has been possible to show that combinations of three antibodies can protect monkeys – especially newborn monkeys – against SHIV transmission (Ferrantelli 2004).
Recently, a novel system to develop and screen for potential vaccines has been developed. The process, patented as 'MolecularBreeding’ or 'gene shuffling' by the American biotechnology company Maxygen, is being used to develop new vaccines based on HIV envelope proteins in partnership with the Scripps Institute (Xu 2004).
The process starts with a set of viral gene sequences, which are artificially chopped up and recombined in the test tube. Each of these gene fragments is then grown up and allowed to express its corresponding proteins. The researchers then select the proteins which have the desired properties, such as the ability to bind strongly to a particular HIV antibody.
The selected sequences are then used to start the whole process over again, and the cycle is repeated a number of times. In effect, this allows selective breeding for viral gene sequences, resulting in the production of sequences that are not found anywhere in nature, but which give rise to much stronger immune responses than any of the proteins which are found in nature.
One possible reason for this stronger response is that HIV has not been able to evolve into strains that stimulate stronger anti-HIV immune responses, as such strains would quickly be eliminated. By artificially applying selective pressure in the opposite direction, it may be possible to arrive at an effective vaccine sequence, without pre-judging what that sequence should be and without first having to understand exactly how it works.
Initial results from rabbits suggest that proteins produced with this system produce higher levels of broadly active neutralising antibodies than are seen in nature. If these results can be repeated in monkeys, there will be a great deal of interest in taking such candidates into clinical trials.
Passive protection of breast-fed babies from HIV using a mixture of antibodies is due to be tested in a clinical trial in Durban, South Africa. Should this prove to be effective, it will add to the interest in pursuing this area of research.
Peptide vaccines
Instead of vaccinating with a whole protein, another approach is to use a fragment of a protein, called a peptide, which consists of a few amino acids.
To date, the V3 sequence of the envelope protein has been the most commonly used peptide in vaccine research. However, this looks increasingly more problematic in view of the new understanding of how gp120 interacts with human cells. However, a vaccine containing the V3 sequences from several strains of HIV has been used in animals and produced antibodies able to neutralise several laboratory-adapted virus strains. A vaccine of this kind has been developed in Cuba with a view to clinical trials.
A therapeutic peptide vaccine has been tested in eight HIV-infected individuals. After 36 weeks, participants had increased antibody response to the HIV peptides used in the vaccine (Pinto 1999).
Another peptide-based vaccine candidate, this time designed for a cellular immune response, has been tested in an open-label clinical trial. Forty HIV-positive people taking effective antiretroviral therapy were randomised to receive one of two doses of peptides matching parts of the p24 (Gag) protein of HIV-1. The peptides were chosen for their ability to bind to human leukocyte antigen (HLA) A-2, one of the proteins of the immune system present in some, but not all of the volunteers. Responses were compared between dose groups and between volunteers who did and did not have HLA A-2. Although specific immune responses were reported, their clinical significance, if any, is unclear (Kran 2004).
A modification of peptide vaccines, linking a peptide with a lipid, has been explored by the French Agence Nationale de Recherches sur le SIDA (ANRS) with initial industrial support from Aventis Pasteur as a way of inducing cellular immunity. The lipid carries the peptide directly into cell membranes where it can be presented to the immune system with maximum efficiency. A number of preliminary clinical trials of such vaccines have been carried out in France and a phase II trial using lipopeptides as boosters for canarypox vaccines is planned as a joint project between the French ANRS and the United States-funded HIV Vaccine Trials Network.
Another more novel way of using a particular peptide, identified as the target for a broadly active neutralising antibody against HIV, is to express it on the surface of the common cold virus. This allows the generation of a range of different varieties and the selection of those that give the strongest immune responses (Arnold 2004).
Regulatory proteins
Several groups of researchers have been investigating the use of HIV proteins other than the envelope proteins in vaccines. Most attention has been given to the structural (e.g. Gag) and regulatory proteins (e.g. Tat, Nef) produced by HIV.
The regulatory protein Tat is produced early after a cell is infected with HIV, which means that cellular immune responses may destroy infected cells before they have a chance to release more virus particles. A further advantage of this approach is that Tat is vital to the functioning of HIV and seems to vary little between different HIV subtypes.
The Nef protein is also of interest, despite the fact that some HIV strains can infect and cause disease without it. If cellular immune responses target cells expressing Nef, they could select for less virulent viruses.
A group led by Dr Barbara Ensoli in Italy has reported animal studies of the Tat protein, and of a DNA vaccine coding for the tat gene, which appeared to show some protection in monkeys. This has led to proposals for trials of a Tat-based vaccine in Uganda. However, some researchers have pointed out that Tat is active in the immune system, so it might be preferable to use a modified version of the protein, a so-called 'Tat toxoid'. The use of Tat toxoids in therapeutic vaccines has been assessed in trials.
More recently, researchers from GlaxoSmithKline Biologicals have reported animal studies using a vaccine consisting of an envelope protein plus a 'Nef-Tat fusion protein' in an adjuvant that promotes cellular immune responses. The combination of these two elements was able to protect monkeys against disease, though not against infection, after challenge with a highly pathogenic simian/human immunodeficiency virus (SHIV). This vaccine has now begun to be evaluated in clinical trials in the United States and Belgium.
Virus-like particles
'Virus-like particles' (VLPs) are combinations of a range of viral and non-viral genes that are produced using genetic engineering in cell cultures. VLPs are non-infectious particles that can be used as vaccines. A range of VLPs has been evaluated in animals and has been found to produce immune responses. However, the failure of trials of one VLP produced by British Biotech, based on the p24 protein of HIV, as a therapeutic vaccine in people with HIV, led to the abandonment of that project. Issues with this approach include both the potential cost and fragility of the products (Smith 2001).
In practice, recombinant vectored vaccines and replicons probably amount to better ways of delivering VLPs within the body, rather than manufacturing them directly in cell cultures.