The Human Parts of HIV

Some researchers wonder if targeting the human proteins HIV carries might be a promising vaccine approach

By Andreas von Bubnoff

Two decades ago, there was a sense of optimism in the field of AIDS vaccine research after a few groups of researchers found that vaccinating macaques with inactivated simian immunodeficiency virus (SIV), the monkey equivalent of HIV, which had been grown in human cells could protect the majority of vaccinated animals from challenge with SIV that had been grown in the same human cells (1; 2; 3). “We were excited,” remembers Michael Murphey-Corb, a professor of microbiology and molecular genetics at the University of Pittsburgh School of Medicine and the first author of one of the studies, which showed that eight out of nine animals were protected (2). “It was Christmas time when the paper came out, and people really wanted to believe,” she says.

But the optimism didn’t last long. In 1991, James Stott and colleagues at the National Institute of Biological Standards and Control (NIBSC) in the UK reported that just vaccinating macaques with human cells protected them from challenge with SIV that had been grown in the same cells. This suggested that the impressive protection that had fueled the optimism in the field might have little to do with a virus-specific immune response, but rather with an anti-cell immune response. This finding led many researchers to abandon this research. “I think it was logical to say let’s focus on how to induce a good antiviral response, which we know based on all the vaccines that are available for all other viruses is the way of generating a good protective immunity,” says Adriano Boasso, a Wellcome Trust research fellow at Imperial College London who recently co-authored two review articles on anti-cell vaccines.

But a few stalwarts continued studying it and still believe today that an anti-cell vaccine might be an interesting alternative type of HIV vaccine. One of them is Gene Shearer, a senior associate scientist at the National Cancer Institute (NCI), who says he has tried to revive interest in the approach twice without success, but isn’t giving up. Earlier this year, he teamed up with Boasso, a former post-doc in his lab, to write two articles that again argue that it is still worth studying the approach (4). “We are now 20 years later and still don’t have an effective AIDS vaccine and continue to do the same things over and over again,” says Shearer, who closed his lab and is now semi-retired. “So I came back and thought maybe we should reopen this idea and see if anybody is interested in it.”

Proponents of the anti-cell vaccine approach say that an anti-cell vaccine would have at least one clear advantage. Because it is based on immune responses to the host cells HIV grew in and not to HIV proteins, it would avoid the problem of HIV constantly mutating to escape the immune response. “It clears away the problem of antigenic variation,” says Stott. “That is a huge, huge advantage.” But there are still many open questions about whether this vaccine approach is effective or feasible. For example, people vaccinated this way might not be able to receive organ transplants. There are also concerns that a vaccine that induces immune responses to host cells might induce autoimmune responses. This makes some researchers skeptical as to whether an anti-cell vaccine could be approved. “Autoimmunity is always a concern,” says David Montefiori, a professor in the department of surgery at Duke University Medical Center. “It’s not necessarily insurmountable, but it would take a long time to prove safety.”

The negative control

Stott’s finding on anti-cell immunity came about because he was vaccinating macaques in a slightly different way from the other research groups at the time. Instead of inactivated and purified SIV that was grown in human cells, Stott and colleagues used crude preparations of whole inactivated SIV-infected human cells (a CD4+ T-cell line called C8166) to vaccinate the macaques. He reported that this approach protected all of the vaccinated macaques from challenge with SIVmac251 that had been grown in the same human cells (5).

In another round of experiments, he vaccinated four macaques twice intramuscularly with inactivated SIV-infected human C8166 cells, and again, the majority (three of four macaques) was protected from intravenous challenge with SIVmac251 that had been grown in the same human CD4+ T-cell line. But when Stott vaccinated four macaques with uninfected human cells from the C8166 cell line—a negative control to prove that the virus components in the cells were responsible for protection—he found to his surprise that two of the four macaques were protected. “[This] was clear evidence that the protective component was actually not the virus part of the infected cells, but the host part,” says Stott.

Stott also found that the serum of the animals contained antibodies to the C8166 cell line. The titer of these anti-cell antibodies was significantly higher in the protected animals, Stott says. Reexamination of 49 animals that had been previously vaccinated with inactivated SIV grown in human cell lines or with inactivated SIV-infected human cells and then challenged with human-cell-grown SIV showed the same correlation between anti-cell antibody titer and protection. This suggested that an antibody response to the human cells appeared to be what protected the animals.

Researchers were shocked. At a meeting in Warwick, UK, there was “quite a strong amazed reaction because it did put the whole field in turmoil because everyone was very positive about getting a vaccine at the time,” remembers Mark Page, a principal scientist at NIBSC who was working in Stott’s group at the time. “Clearly this turned things upside down.” At a conference in the US, Stott says, many people “were actually very skeptical as to whether this was really the case and whether there wasn’t some technical [mistake].”

Stott’s findings, which were eventually published (6), led most researchers to abandon this line of research and instead focus on the induction of virus-specific immune responses. Murphey-Corb eventually abandoned the strategy, partly because she found it difficult to study it further, but also because she got promising results when using a vaccine approach that uses SIV DNA. “I gave up not because I didn’t believe in it. I abandoned the inactivated whole virus vaccine approach because I found a more acceptable one,” she says.

But not everyone stopped studying this vaccination strategy. Further evidence that anti-cell antibodies were protective came when Martin Cranage and colleagues reported that macaques vaccinated with human-cell-grown SIV were protected from challenge with human-cell-grown SIV, but not from challenge with monkey-cell-grown SIV (7). They also found that vaccination with human-cell-grown HIV protected against a challenge with human-cell-grown SIV (8). This further supported the notion that anti-virus immune responses were not responsible for protection, Cranage says, because antibodies to HIV Envelope, which was in the vaccine, don’t recognize SIV Envelope, which was in the challenge virus.

Instead, anti-cell immune responses seemed to be responsible for the protection, likely against host cell proteins like major histocompatibility complex (MHC; known as human leukocyte antigens, or HLA, in humans), because in 1992, Larry Arthur reported that when HIV and SIV bud off an infected cell, they take part of the host cell membrane and its host cell proteins, including HLA or MHC, with them (9).

Further studies by Stott and Arthur revealed that, indeed, immunizing macaques with mouse cells expressing HLA proteins or with purified human HLA proteins was sufficient to protect them from SIV that was grown in human cells that expressed the same HLA proteins (10).

Additional support that anti-cell antibody was likely responsible for protection came when Stott, and Murray Gardner’s group found that the protection could be transferred with serum from animals that had antibody to the host cell component to naive animals. Gardner also found that protection could not be transferred from animals that had antibody against the virus (11).

Page says he also has unpublished evidence for the involvement of complement, a protein cascade that gets activated once it binds to the Fc, or tail region of an antibody that is bound to an antigen. Once activated, complement activates a membrane attack complex that punches holes in the virus envelope. Page found that complement binds to anti-host cell antibodies on the virus and then latches onto the virus and lyses it. This suggests that anti-HLA antibodies might protect the vaccinated animals from infection by binding to the HLAs on incoming virus and preventing virus entry into cells (neutralization), or by activating the complement system.

A different mechanism?

Although many researchers believe that anti-HLA antibodies were responsible for protection in Stott’s experiments, Montefiori has proposed that it’s more likely that antibodies to a different host cell protein were involved (12). He and his colleagues found that the anti-cell antibodies in the serum of animals vaccinated with whole inactivated SIV were not important for neutralization of SIV (13). This suggested that the animals were protected by a mechanism that doesn’t involve neutralization of the virus.

Instead, Montefiori and colleagues found antibodies in the vaccinated animals that bound to complement regulatory proteins (CRPs), human host cell proteins the virus takes with it as it buds from the surface of host cells to protect itself from lysis by the complement system. This suggests that anti-CRP antibodies in the vaccinated animals might keep the CRPs from protecting the virus, rendering the virus susceptible to lysis by the complement system. “That was our hypothesis,” Montefiori says.

Consistent with this hypothesis, Montefiori and colleagues showed that complement killed the virus when the CRP antibodies were present, whereas the antibodies alone didn’t have any neutralizing effect on the virus (14). “In the Jim Stott experiment, I believe the monkeys made antibodies to complement regulatory proteins on the virus that blocked the function of those proteins and thereby rendered the virus susceptible to complement-mediated lysis,” Montefiori says. “I am entirely convinced that this was the mechanism of protection.” Still, he concedes that to prove that the CRP model is correct, it needs to be shown that blocking all human CRPs can protect monkeys from challenge with virus that was grown in human cells, without neutralizing the virus in a conventional neutralization assay. “This would be strictly a complement-dependent mechanism,” Montefiori says. “But the experiment to prove that has never been done.”

The CRP mechanism also seems to operate in humans, Montefiori adds, because when researchers studied a woman who had been injected with her husband’s white blood cells to treat spontaneous recurrent abortions, they found that her serum could neutralize HIV in a complement-dependent manner even though there were no HIV-specific antibodies. Instead, there were antibodies to the cells that HIV had been grown in (15).

Stott says Montefiori’s model can’t explain why just vaccinating macaques with purified HLA protein is sufficient to protect them from SIV that carries that same HLA. But Montefiori says that when he looked in monkeys that had been vaccinated with human-cell-grown SIV, he found no evidence that anti-HLA antibodies had anything to do with neutralization of the virus, perhaps because in this case, the levels of anti-HLA antibodies were too low to neutralize.

Other protective responses

Anti-HLA and CRP antibodies aren’t the only potentially protective responses that are induced by exposure to components of whole cells. Lehner and colleagues, for example, found that immunizing macaques with SIV grown in human cells also induces antibodies to the CCR5 receptor that can block HIV entry into its target cells (16). Additionally, Shearer and colleagues showed that stimulating human white blood cells with supernatants of cells with different HLAs induces a ribonuclease called eosinophil-derived neurotoxin (EDN) that inhibits HIV replication (17).

Lehner and colleagues also found HIV replication inhibited in vitro in T cells taken from couples who regularly have unprotected sex, especially women who are often exposed to foreign HLA in the ejaculates of their partner (18; 19). In addition, they found that the CD4+ T cells of women who were injected with their husbands’ white blood cells to treat them for spontaneous recurrent abortion have increased expression of APOBEC3G (which mutates the viral genome during reverse transcription) and inhibited HIV replication (20). They also found that such women show lower levels of CCR5 receptor expression and elevated levels of anti-CCR5 antibodies (21; 22).

The fact that anti-cell vaccination induces many different innate and adaptive immune responses suggests, Shearer says, that an anti-cell vaccine might be able to block the binding of the virus to its target cells at two different steps—at the HLA level by anti-HLA antibodies, and at the CCR5 coreceptor level by anti-CCR5 antibodies or innate factors such as beta chemokines that bind the CCR5 co-receptor (see figure, this page). Anti-cell vaccination also induces other soluble innate factors that inhibit HIV replication such as APOBEC3G and ribonuclease. “No traditional AIDS vaccine will do all of the things that this will,” Shearer says.

Mechanism of Protection of a Successful Alloantigen-based AIDS Vaccine (ABAV)   

Upon exposure to HIV particles carrying allogeneic human leukocyte antigen (ALLO-HLA), pre-formed anti-HLA antibodies in the immunized host will block HIV challenge. Anti-CCR5 antibodies and ß-chemokines will inhibit HIV interaction with its coreceptor and, in case HIV successfully enters target cells, intracellular restriction factors such as APOBEC3G and a ribonuclease called eosinophil-derived neurotoxin (EDN) will prevent productive infection. This allogeneic HLA-induced arsenal of antibodies and antiviral factors may efficiently prevent infection (“sterilizing” immunity) and result in full protection. Originally published in F1000 Med. Rep. 3, 12, 2011.

 Open questions

Before a human anti-cell based AIDS vaccine can be developed, there are many questions that need to be answered. There is still no solid proof that vaccination with cells or cellular components from the same species can protect from virus grown in these cells. In Stott’s initial 1991 study, researchers protected macaques by xenoimmunization, which means they vaccinated the animals with cells from a different species (humans) to protect them from challenge with SIV grown in human cells.

But humans are infected with HIV that comes from other humans. Therefore, if this type of vaccination approach were ever going to be developed into a human vaccine, it would involve alloimmunization, which means using cells or cell components from the same species (humans).

So one major challenge is to show that alloimmunization can protect macaques as well, Shearer says. But that is not so easy. Murphey-Corb says she tried to grow SIV in primary rhesus mononuclear cells for allogeneic vaccination challenge studies but couldn’t get clean enough material to do the experiments. Other attempts had mixed results. According to Page, Stott’s group vaccinated monkeys with fixed SIV-infected monkey cells and achieved partial protection from challenge with monkey-cell-grown SIV in a study that was never formally published.

In contrast, Cranage’s group did not see protection after alloimmunization of macaques with B cells (23). More recently, Page and Neil Almond at the NIBSC did another alloimmunization experiment in Mauritian cynomolgus macaques, which do not have very diverse MHCs because they are inbred and have a relatively homogenous genetic background. They also failed to see any protection, perhaps, Stott says, because the Mauritian cynomolgus macaques used to generate the vaccine had very similar MHCs to the vaccinees, and therefore the vaccine might not have generated much of an immune response.

Given these mixed results, Stott says it still needs to be shown that alloimmunization can protect. If he can get funding, Page says he plans to do further alloimmunization studies in Mauritian cynomolgus macaques, taking advantage of the fact that their MHCs are very well characterized.

Another barrier to developing a vaccine that can protect from HIV infection by inducing anti-HLA antibody responses is that the vaccine would need to contain multiple HLA proteins to match all potential HLAs on viruses a vaccinated person might encounter, Boasso says. Studies are needed to find out how many different HLAs need to be combined in a vaccine to protect against the majority of circulating HIV strains, adds Shearer.

Lehner says that while the type and number of HLAs that are needed to protect from the majority of HLAs is different for different populations, for a given population only a handful of different HLAs might be needed. For Caucasians, analysis of HLA sequences suggests that just four different HLAs could cover 90% of the population, says Lehner.

So far, however, combining several HLAs in a vaccine has not led to the near complete protection researchers observed 20 years ago. Earlier this year, Lehner vaccinated macaques with the four human class I HLA proteins that he says can cover 90% of the Caucasian population, together with one HLA class II protein, HIV-1 gp140, and SIV p27, all linked to dextran (a complex polysaccharide molecule) to keep them together. This protected two out of eight macaques from an intravenous challenge with an SIV/HIV hybrid (SHIV) that was grown in human cells that had at least one HLA class I and one HLA class II protein in common with the HLA proteins used in the vaccine. The remaining macaques had a reduced viral load compared to unvaccinated controls (24).

One possible reason for the incomplete protection is that Lehner used purified proteins and not whole viruses or cells to vaccinate. Lehner says he chose not to use whole cells in the vaccine because that comes with risks, including that they might carry oncogenic viruses. Another possible reason is that not all HLAs in the study were identical in the vaccine and the challenge virus.

Even if all these challenges could be overcome and a human allovaccine was developed, it would have several limitations. For one, it only induces immune responses to HIV particles from another person, Boasso says. Once HIV particles are made by the vaccinee’s own cells, it doesn’t protect anymore. “As soon as the virus becomes part of you, it’s self and every immune reaction induced so far is completely useless. So you would [have to] prevent infection, period.”

Another limitation, Page says, is that because people vaccinated with an allovaccine develop anti-HLA antibodies, they would be excluded from donating blood. They also couldn’t receive an organ transplant, he says, unless they are plasmaphoresed to remove the anti-HLA antibodies.

A vaccine that induces an immune response to HLAs or other proteins that are similar to proteins on the body’s own cells also raises the concern that it might induce inflammation or autoimmunity, Boasso says, adding that HLA is a molecule that is highly immunogenic.

However, there is no clear evidence that exposing people’s immune systems to cells or HLAs from a different person actually leads to autoimmunity, he says. For example, there hasn’t been any sign of autoimmunity in over 3,000 women who have been vaccinated with their partner’s white blood cells as a treatment for recurrent spontaneous abortion. In addition, says Page, women who gave birth multiple times and people who often receive blood transfusions have HLA antibodies without developing autoimmunity.

Also, HIV-infected people who were immunized with inactivated gp120-depleted HIV particles (a therapeutic vaccine candidate called Remune) developed antibodies to the HLA molecules that had been used in the cell line used to grow the HIV particles. But a study by Page and colleagues found that the vaccinees with the HLAs that matched the cells that the virus was grown in didn’t mount an immune response (25), suggesting that their immune systems didn’t lose their tolerance to self proteins as a result of the vaccination. “That lessened any concerns at the time that you would induce autoimmunity to HLA,” says Page, the first author of that study.

Still, Murphey-Corb doesn’t believe the US Food and Drug Administration would ever approve a vaccine that carried even a theoretical risk of inducing an autoimmune response. “The concept that you are deliberately going to induce a response to self is going to kill [this] forever in the US in my opinion,” Murphey-Corb says. “It’s the perception and not the reality.”

Stott acknowledges the obstacles but still hopes that the most recent attempt to revive some interest in the allovaccination approach will succeed. “We really should be trying to look at radically different approaches. But because it’s radically different, there are a whole lot of hurdles that are going to have to be overcome because it’s new territory,” says Stott.

This may be difficult, given that the RV144 trial for the first time showed modest protection against HIV, says Montefiori, who has turned away from studying alloimmunization because he couldn’t get additional funding. “Since RV144 we have a positive signal that by all accounts doesn’t involve anti-cell antibodies or any type of anti-cell immune response,” Montefiori says. “So nowadays I think to try to improve on the existing vaccines that are based on the viral proteins alone makes more sense and avoids all of the potential downsides of alloimmunization.”

1. Proc. Natl. Acad. Sci. 86, 6353, 1989
2. Science 246, 1293, 1989
3. AIDS Res. Hum. Retroviruses 6, 1239, 1990
4. F1000 Med. Rep. 3, 12, 2011; The Scientist, June 2011
5. Lancet 336, 1538, 1990
6. Nature 353, 393, 1991
7. Nature 355, 685, 1992
8. AIDS Res. Hum. Retroviruses 9, 13, 1993
9. Science 258, 1935, 1992
10. J. Virol. 69, 3117, 1995
11. AIDS Res. Hum. Retroviruses 11, 843, 1995
12. AIDS Res. Hum. Retroviruses 11, 1429, 1995
13. Nature 354, 439, 1991
14. Virology 205, 82, 1994
15. Science 263, 737, 1994
16. Eur. J. Immunol. 29, 2427, 1999
17. AIDS 17, 481, 2003
18. Lancet 363, 518, 2004
19. PLoS ONE 4, e7938, 2009
20. Eur. J. Immunol. 39, 1956, 2009
21. Nat. Med. 5, 1004, 1999
22. Clin. Exp. Immunol. 129, 493, 2002
23. AIDS Res. Hum. Retroviruses 13, 923, 1997
24. J. Virol. 85, 6442, 2011
25. AIDS 21, 375, 2007