Tripwire Uncovered for “Exhausted” T Cells in HIV Infection

Blocking Protein Pairing May Pick Up Immune Response

Almost 20 years ago, Bruce Walker, working as a postdoc in the lab of Robert Schooley at Massachusetts General Hospital, published a paper in Nature describing CD8, or killer, T cells that respond specifically to HIV in the blood of patients infected with the virus. Since then, studies have shown that killer T cells are important in controlling the virus in the acute stages of infection, but they lose their ability to combat it over the long term. Why they persist in the body without doing their job has been a mystery. “How can we reconcile the fact that these cells are there in large numbers but are not controlling virus?” asked Walker, now a Howard Hughes investigator and an HMS professor of medicine. “We’ve basically spent two decades trying to understand this fundamental question.”

Photo by Graham Ramsay

Gordon Freeman (left) and Daniel Kaufmann were part of an international team that found a molecular explanation for T cell “exhaustion” in HIV infection.

Because CD8 T cells seem to lose their killer instinct in the presence of chronic HIV infection, their state has been described as “T cell exhaustion.” This exhaustion is also seen in CD4, or T helper, cells in HIV, and it is a phenomenon that is common to chronic infections. But “exhaustion” is the kind of vague diagnosis used to explain why a Hollywood starlet disappears from the set, a euphemism more than an explanation. So it has been particularly unsatisfying for describing a behavior in cells that must have some underlying mechanism.

Recently, however, a molecular player has come to light: programmed death 1, or PD-1, a receptor protein on the surface of T cells. In the Sept. 21 Nature, a team led by Walker and scientists at Emory University, Oxford University, and the University of KwaZulu-Natal in South Africa, shows that PD-1 gets switched on in exhausted cells in patients infected with HIV. Two studies by other groups, published in Nature Medicine and the Journal of Experimental Medicine, confirm the findings. These studies point to a specific, reversible mechanism responsible for the T cells’ ailment.

The Shut-off Signal
PD-1 is one of the general pathways in T cells responsible for shutting down immune responses. Normally, an infection causes killer T cells that respond specifically to a pathogen to expand rapidly in number. Once the infection is controlled, most of these cells die, while a small number of them persist as “memory” cells. Yet in chronic infections, the killer T cells slowly dwindle, and the ones that remain lose their ability to kill infected cells. They also seem to receive less help from CD4 cells. The lab of Rafi Ahmed at Emory University first observed that in most infections in mice, PD-1 expression in killer T cells was high in the acute stages, but was reduced in the memory cells. But when mice were given a chronic viral infection, PD-1 levels remained high, as did the levels of PD-1’s binding partner on antigen-presenting cells, PD-L1.

This research, published last year in Nature, pointed to a mechanism that might explain T cell exhaustion in human chronic infections. “It was really compelling reading that paper to think that something similar might be operating in HIV infection,” Walker said. “It was a potential opportunity to explain why the cells weren’t functioning.”

Waking up tired T cells. In a normal infection, when killer T cells are activated by interacting with antigen­presenting cells (left), they work to resolve the infection by destroying infected cells and releasing cytokines; once the infection is resolved, relatively few of the T cells persist as memory cells (top). In an uncontrolled chronic infection, the T cells persist but gradually lose function (bottom). Blocking the interaction between the T cell receptor PD-1 and its ligand on antigen-presenting cells, PD-L1, may help reinvigorate T cells that have become exhausted due to chronic infection.

Gordon Freeman, HMS associate professor of medicine at Dana–Farber Cancer Institute, who has been studying the role of PD-1 and PD-L1 in the immune system, explained that T cells normally activate when a receptor on their surface binds to an antigen on an antigen-presenting cell. When the T cell also expresses PD-1 and binds to PD-L1 on the antigen-presenting cell, “it activates molecules that reduce the strength of the T cell receptor signal,” effectively putting a brake on the T cell’s response. Freeman said that such a mechanism is useful because “the immune system can destroy tissues,” so a prolonged immune response is potentially dangerous.

A Reversible Collapse
Walker worked with Ahmed and Freeman, who developed antibodies to block the interaction between PD-1 and PD-L1, to look at the role of the PD-1 pathway in the most devastating chronic infection in humans, HIV. Cheryl Day, co–first author of the current paper and instructor in medicine at MGH, led a team to examine the status of PD-1 in the T cells of patients at the Doris Duke Medical Research Institute in Durban, South Africa, part of a partnership between the Partners AIDS Research Center and the University of KwaZulu-Natal. In the killer T cells of 71 patients who had not yet begun antiretroviral therapy, PD-1 expression was higher and was particularly high in cells specific to HIV compared to those specific to controlled infections. Cells with higher PD-1 expression responded less readily to HIV proteins.

PD-1 expression was highest in patients with more advanced disease; it correlated with viral load and loss of helper T cells. To see whether treatment affected PD-1 levels, the team also looked at four patients who began antiretroviral therapy during the study. PD-1 expression dropped in HIV-specific killer T cells as the viral loads of the patients fell.

Daniel Kaufmann, co–first author and instructor in medicine at MGH, found similar patterns of PD-1 expression in CD4 cells from samples in Boston. Kaufmann said that it has long been known that the helper T cells not only diminish in number in HIV infection, but also lose their ability to function. “It’s thought that one of the reasons why the HIV-specific CD8 cells don’t function properly is that they don’t receive proper help from CD4 cells,” he said.

Blocking the interaction between PD-1 and PD-L1 made both CD4 and CD8 cells proliferate, indicating that the exhaustion of these cells could be overcome. “We know these CD8 and CD4 T cells don’t work properly,” said Walker. “Until now, it’s been unclear whether cells were irreversibly damaged or whether it was a reversible mechanism.”

In HIV infection, switching on the PD-1 pathway may be the body’s way of abandoning a battle that can’t be won. “You can think of this as the body prematurely giving up the fight,” Walker said.

Antibodies that block PD-1 are already being investigated as cancer treatments. PD-L1 is expressed at high levels on many solid tumors, and seems to help cancer cells escape destruction by the immune system. Kaufmann said that such studies will help speed the pace of studies that look at blocking the pathway in HIV and other chronic infections. But the study authors caution that it is not yet clear how promising this approach will be. Blocking this pathway completely in mice can accelerate autoimmune disease, and it may be tricky to manipulate the PD-1 pathway without causing disease. “It’s definitely something that can be manipulated,” Freeman said. “The main question is to learn how to do it effectively but safely.”

—Courtney Humphries

top