The Duke team discovered that certain broadly protective antibodies, which recognize and latch onto the HIV protein gp41, resemble antibodies made in autoimmune diseases. In most people, the immune system destroys these types of antibodies to prevent attacks against self.
The Duke study suggests HIV vaccines may have failed in part because certain proteins on HIV's protective outer coat trigger only short-lived, self-reactive antibodies instead of long-lasting, HIV-specific antibodies. The results also imply that during the initial infection stage in humans, HIV may escape destruction by the immune system because these seemingly vulnerable outer coat proteins activate self-reactive antibodies.
"The fundamental problem in all of HIV vaccine research has been that when you inject the envelope of the HIV virus into people or animals, no broadly neutralizing antibodies ?those antibodies that kill most HIV strains ?are made. This provides a plausible explanation for why broadly protective antibodies have not been made in response to currently tested HIV vaccines," said Barton Haynes, M.D., lead author of the study and director of the Human Vaccine Institute at Duke University Medical Center.
The researchers will report their findings in a forthcoming issue of Science. The results were published online April 28, 2005, in Science Express.
The antibody-producing portion of the human immune system is broadly divided into two categories. The first, innate B cell immunity, comprises fast-acting but weak antibodies that fight a broad range of pathogens. These antibodies can also attack the body itself, as in autoimmune disorders such as systemic lupus erythematosus. When viruses activate innate B cells, the body destroys the B cells to protect against autoantibodies that could cause autoimmune disease or other harm.
The second immune system category is adaptive B cell immunity, a slower response that creates powerful, pathogen-specific antibodies and provides lasting immunity. The body's normal response to infection is to produce adaptive antibodies that target only the invading virus or other pathogens. Many widely used non-HIV vaccines "train" adaptive antibodies to seek out a unique protein on the protective outer coating of viruses. HIV researchers have attempted to induce broadly neutralizing antibodies ?long-lived, HIV-specific antibodies that can kill all or most HIV strains ?with a similar vaccine design.
Some broadly neutralizing antibodies have been isolated from HIV-infected humans, although the antibodies are rare, with less than five identified. "We know these antibodies can exist, but we have not been able to give a vaccine to people or animals that stimulates the production of these types of antibodies," said Haynes, who has studied HIV vaccines for 15 years.
In their experiments, Haynes and his colleagues demonstrated that some of these rare broadly neutralizing antibodies are actually polyspecific autoantibodies that react with many proteins, including one's own tissues, like the antibodies made by innate B cells. In laboratory tests, the antibodies reacted with multiple types of human molecules, most prominently with a fat molecule called cardiolipin.
"It appears the most vulnerable spots on the outer coat protein of HIV, to which the most protective antibodies bind, are the target of autoantibodies that also react with normal human tissues and are normally destroyed by the immune system," Haynes said.
Haynes, an AIDS researcher who has also studied autoimmune diseases, began to focus on possible similarities between HIV infection an d the biology of autoimmunity after work on an experimental outer coat vaccine failed to produce broadly neutralizing antibodies in animals.
"Recently, we spent two years making an experimental outer coat vaccine candidate that had the correct areas on the outer coat for the good broadly neutralizing antibodies to bind to, and we vaccinated several kinds of animals. In none did we get any of the good antibodies. That frustrating result led me to ask if something was preventing these good antibodies from being made," Haynes said.
"A light went on when I saw that the rare human monoclonal antibodies had physical characteristics very similar to autoantibodies found in autoimmune disease ?in other words, to the antibodies the normal immune system does not allow to be made," Haynes said.
The results provide a new goal for future HIV research, Haynes said. "We can focus on trying to redirect the response to HIV outer coat proteins from innate B cells to adaptive B cells. Alternatively, we can develop ways to induce that first line of polyspecific antibody defense during vaccination, if these antibodies are not harmful to those being vaccinated," Haynes said.
"We now have a window into how to study HIV vaccines from the host side of the problem," he said.
Collaborators on the study include Judith Fleming, William St. Clair, Richard Scearce, Kelly Plonk, Herman Staats, Thomas Ortel, Hua-Xin Liao and Munir Alam of Duke; Herman Katinger, Gabriela Stiegler and Renate Kunert of the Institute of Applied Microbiology, University of Agriculture, Vienna, Austria; and James Robinson of the Tulane University School of Medicine. The National Institute of Allergy and Infectious Diseases of the National Institutes of Health supported the work.