The reason a highly effective HIV vaccine has eluded medical researchers to date is the speed with which the virus evolves. HIV, it appears, can find ways around any antigen that has been thrown at it to date.

The past few years have seen marked success in identifying broadly neutralizing antibodies, raising hopes that it will eventually be possible to find antigens that can reliably induce those antibodies in an HIV vaccine. (See BioWorld Today, July 30, 2012.)

Now, scientists have suggested an alternate approach. Rather than directing the response of an HIV vaccine against viral proteins, they have used antigens that are expressed only in cells that the virus infects.

Fittingly, those proteins are themselves parts of former retroviruses – viruses that managed to colonize human cells at some point in our evolutionary history. Retroviruses copy themselves into the genome of the cells they infect, and "if a retrovirus can infect a sperm or egg cell without killing it, offspring will have a copy in each cell," the University of Toronto's Brad Jones told BioWorld Today.

Exactly that appears to have happened multiple times – several percent of the human genome is made up of human endogenous retrovirus, or HERV, sequences.

As for why such a large part of the human genome is not human at all, Jones said that once they integrate, endogenous retroviruses "may not be neutral, but may have some benefit for us."

There is evidence, for example, that some endogenous retroviruses protect against infection from other retroviruses. And in a sign of how old the relationship between humans and their retroviruses is, the placental fusion protein, which plays a key role in pregnancy, is derived from a retroviral protein. "Without these endogenous retroviruses, mammalian reproduction as we know it would not exist."

Plenty of other HERV sequences, though, are not expressed in human cells. And then there are some that are not expressed in healthy cells, but come to be expressed when cells are infected with HIV.

Given that the vaccine targets a host protein – albeit one that only appears to be expressed in HIV-infected cells – autoimmunity "is, of course, one concern going forward," Jones acknowledged. "But the value of targeting HIV irrespective of its sequence diversity merits at least having a look."

In their experiments, which they published in the Nov. 12, 2012, online edition of the Journal of Clinical Investigation, first author Jones and his colleagues used T cells that respond to the endogenous retrovirus HML-2. HML-2 has colonized humans relatively recently, and its reversely transcribed DNA has now made itself at home an average of about 90 times in the average human genome. That DNA is not transcribed under normal circumstances.

But when Jones and his colleagues infected human CD4-positive helper T cells with HIV, those cells started expressing HML-2 as well as HIV proteins.

The team next looked at CD8-expressing killer T cells that targeted such HML-2 proteins, and they found that such cells were able to kill helper T cells that were infected with several strains of HIV whose sequences are quite dissimilar, and even cells infected with SIV, the primate equivalent of HIV, though not all strains were attacked with equal effectiveness. Expression of the HIV protein Vif was required for endogenous retrovirus to be expressed, although administering Vif was not enough for such expression.

Jones said Vif might be working by shutting down the cell's countermeasures against viral infection, with HML-2, as a relative newcomer to the HERV family, still similar enough to an outright retrovirus that it could benefit from Vif's silencing of host defenses. Alternately, he added, HML-2 proteins may be transcribed under normal circumstances, but not be ferried to the cytoplasm for translation. Vif might also enable HML-2-derived mRNA to make it out of the nucleus and into protein form.

Jones and his team next plan to look at whether a vaccine that uses HERV-derived antigens can be protective against SIV infections in macaques. He cautioned, however, that those experiments are not completely straightforward to either do or interpret, since "the endogenous retroviruses that we have are not exactly the same as those of rhesus macaques, and HIV is not exactly the same as SIV."