Public Release: 

Researchers Discover How The Immune System Shuts Down Faulty T Cells

Harvard Medical School

When This Quality Control Mechanism Fails, Tumors Escape The Body's Defenses Or Autoimmune Diseases Develop

Defective T cells run afoul of a quality control system operating in the peripheral immune system, Harvey Cantor's lab has discovered. Incompetent CD 8 cells, including some that respond to antigen (red) presented by class II rather than class I MHC proteins (top panel)--and others that fail to see MHC at all--sometimes elude capture in the thymus and reach the periphery. Yet if their T cell receptors recognize the wrong antigen and their CD 8 coreceptors are therefore unable to bind, internal changes in gene expression soon disarm and kill these duds. By contrast, competent T cells bind continuously at a low level, suppressing genetic death controls (middle panel). The T cells whose CD 8 and T cell receptors bind tightly to foreign antigens (red diamond) that are displayed on infected cells (bottom panel) trigger an immune response.

Full size image available through contact

BOSTON -- May 11, 1999 -- Researchers at Dana-Farber Cancer Institute and Harvard Medical School have identified a mechanism the immune system uses to detect and eliminate dysfunctional T cells. Reported in the May 14 Science, the study clarifies a long-standing puzzle about the development of systemic autoimmune diseases and has immediate practical implications for how to design cancer vaccines.

Harvey Cantor, chair of the Department of Cancer Immunology and AIDS at DFCI and an HMS professor of pathology, and his colleagues have found that the peripheral immune system plays an important role in the selection of properly functioning T cells. Until recently, this process was thought to occur only in the thymus.

As nascent T lymphocytes emerge from the bone marrow, they enter the thymus gland. The thymus identifies promising infection fighters by selecting T cells able to recognize and lightly bind the body's own peptides, presented by MHC proteins. The thymus then eliminates those cells that bind self antigen too tightly, to prevent autoimmune reactions. At this point, typically only two types of mature T cells are left: Helper T cells and cytotoxic cells. Helper T cells carrying the CD 4 coreceptor recognize antigen bound by class 2 MHC molecules and trigger antibody production upon activation by foreign peptides. Cytotoxic cells carry CD 8, recognize antigen bound to class 1 proteins, and kill cells bearing foreign peptides.

But nothing in biology is foolproof. The thymus does allow into the periphery some CD 8 misfits that recognize class 2-bound antigen, as well as incompetent cells that cannot recognize MHC molecules at all. Since these cells could be dangerous, immunologists have wondered how the body deals with them.

Cantor and his colleagues isolated mouse CD 8 cells unable to recognize class 1 MHC and tracked their fate. Suspecting that repeated binding of the CD 8 coreceptor to peripheral class 1 proteins was somehow important, the researchers in a parallel experiment deprived normal CD 8 cells of that binding by transferring them into mice engineered to lack class 1 proteins.

To their surprise, they found that in both experiments, CD 8 cells unable to bind class 1 essentially disarmed themselves within 48 hours. That is, the cells initially left the thymus carrying CD 8 proteins on their surface. Sensing, however, that their coreceptors were not being properly engaged, the cells promptly shut down further transcription of the CD 8 gene, allowing normal protein turnover to deplete its CD 8 supply. The cells also stopped making an enzyme needed to kill a target cell.

The CD 8 cells, starved of class 1 contact, also began transcribing their genes for two surface proteins known to touch off a genetic cell death program. Another two days later, the cells were dead. Cantor's team also discovered the genetic switch that may have unleashed the cell death activators: While immature T cells undergo testing in the thymus, they teeter on the brink of death, but as they graduate, they receive protection in the form of the apoptosis-suppressing transcription factor. Yet without seeing class 1 in the periphery, this protection is stripped away and the factor disappears.

This means that the default mode for CD 8 cells is death, which is kept at bay only by continuous contact of its CD 8 protein and T cell receptor with class 1-bound antigen, says Cantor. He adds that this quality control pathway extends a process started in the thymus.

The study's clinical implications are twofold, says Cantor.

Tumors sometimes frustrate oncologists with a phenomenon dubbed "tumor escape," in which tumor cells shed class 1 molecules along with tumor antigens. Cantor's findings suggest that this trick is even more insidious than was thought, because not only does the tumor evade its supposed killers, but by withholding class 1 proteins from the CD 8 cells, the tumor cell induces the death of the very cells that are supposed to destroy them. Indeed, when the researchers allowed CD 8 cells to interact with tumor cells that had cast off some class 1 proteins, the level of cell death among the CD 8 cells doubled whereas the tumor cells remained increasingly unscathed.

"This problem becomes more serious as we get better at developing cancer vaccines," says Cantor, adding that researchers need to define the whole range of tumor antigens with peptide libraries and then deploy a cocktail of tumor antigens in their vaccines rather than counting on one or two. That should work, since researchers know tumor cells cannot downregulate all antigens.

The study also helps explain observations that have long perplexed immunologists trying to understand systemic autoimmune diseases such as lupus, says Cantor. One in four patients with lupus contains T cells without either CD 4 or CD 8, sometimes in huge numbers suggestive of a lymphoma.

Though scientists have suspected this pool of cells to be the autoreactive culprits, Cantor's work now shows they are not. Instead, they are the detritus of the immune system -- impotent CD 8 cells that have started down the quality control pathway by taking down CD 8 but cannot conclude it due to mutations in programmed cell death genes. They end up accumulating, possibly exacerbating inflammation by producing cytokines. However, researchers studying the cause of lupus will need to look elsewhere for a culprit, says Cantor.

The current work reflects a change in immunologists' thinking, he says. Increasingly, they view autoimmune disease not simply as a cartoon version of renegade T cell clones, but as a broader dysfunction of the body's intricate controls over the life and death of its T cells.

By that reasoning, he adds, autoimmune diseases and cancer can no longer be considered unrelated. They appear to be two sides of one coin in that these life -- death controls are out of kilter in opposite ways in the two conditions. In lupus, useless CD 8 cells that ought to die stay on, whereas in cancer, urgently needed CD 8 cells are tricked into suicide.

"We have stood on its head the current paradigm for understanding the life and death of a peripheral T cell," Cantor says.

The research was funded in part by the National Institutes of Health.


Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.