HIV vaccines: the future looks promising.

If I had a dollar for every time a person asked me despairingly, "How come no HIV vaccine?" I'd be rich and eating bonbons in Tahiti. But I'd rather forgo the prospect of riches and additional pounds and actually have an effective and broadly used HIV vaccine.

There are several major reasons why we don't yet have a vaccine that meets our needs. The most basic reason is that HIV is a wily pathogen, with surprises at every turn. The genetic diversity of HIV is a big problem. HIV also conceals important regions in its proteins that would normally elicit a robust immune response. This concealment leads to a key problem in the development of neutralizing antibodies, both in infected human beings and in animal models of HIV. (1)

What are neutralizing antibodies? These are antibodies that develop in the body after exposure to an organism and actually prevent infection by "neutralizing" the agent in question via a number of mechanisms. The questions are: How do we create these for HIV? and How can we get humans to make such antibodies?

To even begin to address these questions, a broad review of how vaccines are made and delivered to humans is necessary. When the average person thinks of vaccines, they might think of the polio or smallpox vaccines, which are vaccines against diseases that have been conquered. The advantage of these vaccines is that they are live but attenuated (see Table 1). That is, the organism used as a vaccine is a less pathogenic form, but nevertheless causes infection and hence induces great long-lasting protective immunity. However, the downside is that in some cases, both the smallpox vaccine and the oral polio vaccine can cause disease, which, of course, is undesirable. (2)

A live, attenuated vaccine is unlikely to be made against HIV because of worries about in vivo mutations and the fact that HIV integrates into cellular DNA. This means that once it's there, it's in the DNA for the life of the cell. The principle protective mechanism for live, attenuated vaccines is thought to be production of neutralizing antibodies. Unfortunately, the reality is that the actual mechanism of protection for any vaccine is seldom completely understood. Indeed, this is a crucial issue for the design of any vaccine. How do you know if you have the right antigen? How do you know if the immune responses you can easily measure are protective? Having a "challenge" animal model to test these ideas is extremely important. At this juncture, most agree that targeting both envelope and structural proteins from the virus will be necessary for an effective vaccine against HIV. Sterilizing immunity, however, has only been observed with antibodies given before or at the time of challenge with Simian Immunodeficiency Virus (SIV). (3) T-cell vaccines induce good T-cell responses, but of course the animals still get infected with SIV. Thus far, the best control of virus replication occurred using an MVA (modified vaccinia virus Ankara) DNA prime/boost method. (2)

Most of the other vaccines used to immunize humans are killed agents, or proteins, or subunit vaccines--those made from components of an organism, sometimes against a single protein, as is the case for Hepatitis B vaccine. But, get real. Preventive and/or protective immunity induced by immunization with a single protein is rare. For HIV, it is likely that responses to multiple proteins will be required.

The modern age of vaccine design is centered around immunizing by delivery methods using viral vectors that are either replication competent (meaning alive) or replication defective (meaning dead). None of these is really in use yet, although a Phase I study was recently published. (4) The replication-competent vaccines are most like the live, attenuated vaccines mentioned above. They usually use other viral agents, which have been genetically engineered to contain HIV proteins. However, they do have the potential to revert (similar to live, attenuated vaccines).

A key issue with replication-defective vectors is that there may be pre-existing immunity to the vector, such as with adenovirus. Most of us have had various infections caused by this organism in our lives, so that when our immune system sees this virus again, there will be a memory response that will reduce the effectiveness of the responses against HIV proteins the vector is carrying. Naked DNA-encoding proteins from various organisms have also been found to induce immunity, especially cellular responses by T cells.

Both of these strategies--virus-based vectors and naked DNA-based vectors--are under evaluation for HIV vaccines. The virus-based vectors include those that are based on adenovirus and pox virus, both of which are being studied in advanced clinical trials. Several other formulations of vaccines are in advanced trials based on adeno-associated virus vectors or the use of HIV peptides mixed with lipids. These later vaccines primarily elicit cellular immunity. In addition, an HIV multiclade DNA vaccine and an adenovirus vaccine have induced antibodies. Now these 2 vaccines will be used together to elicit an optimal response. (4,5)

The current strategy of choice, arising from multiple preclinical trials, is called prime/boost, meaning priming with DNA (for example), then following with a boost from a vectored vaccine for the same organism. This primarily results in the generation of CD8 T cells, which are key in antiviral immunity. Variables important in generating good CD8 T-cell responses include the way vaccine is given (subcutaneous is best), the dosage, and the type of delivery vector. (6,7)

However, one thing should be made clear. No one believes that CD8 T cells by themselves can result in sterilizing immunity to HIV (ie, no infection at all). Rather, a good CD8 response typically contains the pathogen through the elimination of infected cells, but does not eliminate the pathogen per se. To eliminate or prevent infection, antibodies that are broadly neutralizing are needed and should be reactive with many of the HIV envelope proteins and other HIV proteins found in human populations.

So why don't we just take a bunch of different envelopes from different HIV groups and create a vaccine from them? The answer is that HIV has "unprecedented mechanisms for evading the host antibody response." (1) Plus, a key problem discovered in the last few years concerns how HIV gets into susceptible cells. The viral envelope is a trimer of glycoprotein (gp): gpl20-gp41. The gp120 binds to CD4, which induces a conformational change, exposing a region that then binds to a chemokine receptor on the T cell. After this binding, another conformational change allows gp41 to facilitate entry of HIV RNA into the cell. However, much of the HIV surface of the envelope has N-linked sugars, which actually come from the host cells that HIV buds from; hence much of the gp120, which is highly immunogenic without the sugars (for both mice and humans), is cloaked in carbohydrates! (8,9)

On the other hand, as we have learned more about the structure of HIV, we also have learned that neutralization by antibodies is possible! That is, through the work of a few diligent investigators in the 1990s, and the screening of many thousands of human antibodies, a few neutralizing antibodies were found that came from a few rare HIV-infected people. The fact that these antibodies exist offers proof that, although rare, humans can make neutralizing antibodies to HIV. The best studied antibodies (4 of them) indicate that they work by recognizing physically conserved areas exposed on the gp120 molecule that have conformational restraints and are thus important in allowing HIV to enter cells. (1,8) Many investigators, especially organic chemists, are now at work trying to "mimic" these regions, so that an ordinary host immune system could be induced to respond reliably. Thus, there is real hope on this front.

Another recent area of promise in vaccine research uses virus-like particles or VLPs. These particles consist of various HIV proteins that can self-assemble without the infectious RNA. A key advantage of these particles is that they induce high levels of both T cells and antibody, most likely because these particles are so structurally similar to the real HIV. (10) A key disadvantage is that they are very expensive to make and their stability is questionable, so future work is necessary.

So far, this article has focused on prophylactic vaccines. What about vaccines for those already infected? The good news is that if the immune system is intact enough to respond, several strategies can be used to boost immunity and better suppress HIV or SIV. (11) Moreover, even if a vaccine did not protect against infection, a recent study showed that vaccination before viral challenge could protect memory cells from dying during infection, allowing the immune system to remain intact. (12) Another strategy using a canary pox virus vaccine and interleukin-2 (IL-2) in HIV-infected patients resulted in increased CD4 T cells (because of the IL-2), but did not control viral rebound. (13) There is some indication that the timing of immune intervention is likely to be critical for a therapeutic vaccine to be effective. (14) In summary, there has been real progress conceptually in HIV vaccine design in the last 5 years, as well as new approaches that remain to be tested clinically--the future of HIV vaccines looks promising.

References

(1.) Douek DC, Kwong PD, Nabel GJ. Cell. 2006;124:677-681.

(2.) Robinson HL, Amara RR. Nat Med. 2005; 11 :$25-$32.

(3.) Eda Y, Murakami T, Ami Y, et al. J Virol. 2006;80:5563-5570.

(4.) Catanzaro AT, Koup RA, Roederer M, et al. J Infect Dis. 2006;194:1638-1649.

(5.) Graham BS, Koup RA, Roederer M, et al. J Infect Dis. 2006; 194:1650-1660.

(6.) Harari A, Pantaleo G. Eur J Immunol. 2005;35:2528-2531.

(7.) Estcourt MJ, Letourneau S, McMichael AJ, Hanke T. Eur J Immunol. 2005;35:2532-2540.

(8.) Calarese DA, Lee HK, Huang CY, et al. Proc Natl Acad Sci USA. 2005;102:13372-13377.

(9.) Burton DR, Desrosiers RC, Doms RW, et al. Nat lmmunol. 2004;5:233-236.

(10.) Young KR, Ross TM. AIDS Res Hum Retroviruses. 2006;22:99-108.

(11.) Lu W, Arraes LC, Ferreira WT, Andrieu JM. Nat Med. 2004;10:1359-3165.

(12.) Mattapallil JJ, Douek DC, Buckler-White A, et al. J Exp Med. 2006;203:1533-1541.

(13.) Kilby JM, Bucy RP, Mildvan D, et al. J Infect Dis. 2006;194:1672-1676.

(14.) Hunt PW, Deeks SG. J Infect Dis. 2006;194:1632-1634.

Dorothy E. Lewis is Professor in the Department of Immunology at Baylor College of Medicine in Houston.

Table 1. Types of vaccines

Type                    Examples              Immunity due to:

Live, attenuated        Smallpox              Antibodies
                        Oral polio
Heat killed: protein    Polio                 Antibodies
or subunit              Hepatitis B
Replication competent   Yellow fever          Antibodies
Replication defective   Adenovirus 5          Cellular immunity
DNA                     Many in development   Cellular immunity
Prime/Boost             Optimized strategy    Both antibodies and
                        for HIV               cellular immunity
Virus-like particles    Many in development   Both antibodies and
(VLPs)                                        cellular immunity

Type                    Risks/Problems

Live, attenuated        Reversion to pathogenic

Heat killed: protein    Not very diverse,
or subunit              not as effective
Replication competent   Reversion
Replication defective   Pre-existing immunity
DNA                     No antibodies
Prime/Boost             Complicated,
                        expensive
Virus-like particles    Expensive to produce,
(VLPs)                  unstable