Each time our body’s immune system responds effectively to an invading pathogen, such as a cold or the flu, it creates a population of T cells that “remember” the infection and can respond swiftly if the invader dares return.

But in chronic infections like HIV and malaria, these memory T cells do not develop. Instead, the body goes into a drawn-out standoff with the pathogen, and T cells become sluggish. A similar situation can arise in cancer, with the T cells never quite able to extinguish the malignant cells.

Some of the most exciting new cancer drugs, checkpoint inhibitors, act to rev up these tired T cells, but the impact fades over time. And now we understand why.

A study out last week in Science led by researchers from Penn’s Perelman School of Medicine reveals that so-called exhausted T cells are in fact quite distinct from their more-effective memory T cell counterparts—so different that even powerful checkpoint inhibitor drugs cannot fully revive them. The differences lie at the epigenetic level; that is, the two cell lineages have different sets of markers that determine what genes the cells can turn on and off.

“These findings do explain some of the failures we’ve seen in using checkpoint inhibitor drugs,” says senior author John Wherry, director of the Institute for Immunology and the Barbara and Richard Schiffrin President’s Distinguished Professor of Microbiology at Penn Medicine. “But they also present a huge opportunity. If we can reprogram the epigenetics of exhaustion in T cells, we open up the possibility to go after these cancers in a powerful and lasting way.”

For a few years leading up to this research, Wherry, who is also co-director of the newly launched Parker Institute for Cancer Immunotherapy at Penn, and his colleagues had an inkling that the exhausted T cells were fundamentally different from memory T cells. But it wasn’t until technological advances made it possible to study the epigenetic profiles of small numbers of cells that they were able to test their hypothesis.

Working with epigenetics expert Shelley Berger, a Penn Integrates Knowledge professor, Wherry’s team compared the epigenetic profiles of exhausted T cells to those of memory T cells and effector T cells (another T cell type that mounts an attack on pathogens), and found that the exhausted T cells were distinct from the other two. Applying a checkpoint inhibitor drug to the exhausted T cells resulted in only minor, transient changes in gene expression, not a permanent, durable transformation.

Armed with a catalog of these epigenetic characteristics of exhausted T cells, Wherry says it would be useful to develop biomarkers to determine when checkpoint blockades cease being effective. The findings are also generating ideas for how to pursue therapies that could reprogram the tired T cells and make them fight back with renewed strength.