Investigators from the National Institute of Allergy and Infectious Diseases (NIAID), part of the National Institutes of Health, have conducted studies in mice to gain a new picture of how the immune system's "killer" T cells are prompted to destroy infected cells. Their insights provide a blueprint for rational design of vaccines that induce desired T-cell responses.

The findings are published in this week's "Science." "If we are correct, what we've found will put rational vaccine design on a firmer footing," says Jonathan Yewdell, who led the NIAID team.

T cells belong to the cellular arm of the immune system's two-pronged defense mechanism against foreign invaders -- the other arm features blood-borne antibodies. Historically, vaccines aimed to stimulate antibody production in a bid to prevent specific diseases. More recently, scientists have begun to manipulate T cells to create vaccines effective against pathogens that antibodies alone cannot control. Such T-cell-inducing vaccines are being tested against infectious diseases such as HIV/AIDS and hepatitis and are being studied as treatments for certain cancers.

Once alerted to the presence of infected cells, resting T cells are "awakened" and begin to multiply rapidly. Then they zero in on and destroy infected cells while sparing uninfected ones. Rousing slumbering T cells is the job of dendritic cells, the sentinels of the immune system.

Dendritic cells activate the T cells by displaying peptides--small pieces of virus or other foreign protein--on their surfaces. In a process called direct priming, dendritic cells generate these peptides by themselves after being infected by a virus.

Alternatively, dendritic cells may first interact with other body cells that have been infected by a virus and then activate the T cells. This indirect route is called cross-priming.

Vaccines may exploit either route to T-cell priming, but scientists have not known enough about the mechanisms behind cross-priming to exploit this route in vaccine design.

Test tube experiments suggested that molecular "chaperones" accompany peptides from infected cells to dendritic cells, and a number of experimental vaccines have been designed on this premise. But few studies have been done to determine if chaperoned peptides play any role in animal systems, notes Dr. Yewdell.

If the chaperoned peptide theory is correct, infected cells that make the most peptides should most strongly stimulate cross-priming. Conversely, fewer peptides should mean less cross-priming. To test this prediction, Dr. Yewdell and his colleagues created virus-infected cells that were genetically or chemically prevented from producing peptides and injected those cells into mice. They found the opposite of what they expected: cross-priming correlated directly with levels of whole proteins, rather than levels of peptides, expressed by the virus-infected cells.

This new information could aid vaccine design, says Dr. Yewdell. "Our experiments indicate that two distinct pathways exist to prime T cells," he says. If the rules for T-cell priming suggested by these experiments are correct, vaccines meant to interact with dendritic cells should be designed to generate large amounts of peptides, while vaccines that target other kinds of cells should be designed to generate whole proteins that will go on to be processed in the dendritic cells during T-cell cross-priming.

Prompting a strong and specific T-cell reaction may be the key to vaccines that are effective against certain infectious diseases, including HIV/AIDS and malaria, notes Dr. Yewdell. It is also possible that a therapeutic vaccine might be developed to boost the T cell activity of people who have chronic liver infections caused by hepatitis B or C viruses.