Computational Biologists Assemble Atlas of T Cell Dysfunction

Share
MSK computational biologists Christina Leslie and Yuri Pritykin

Christina Leslie and Yuri Pritykin

Our immune system is great at protecting us from infections and cancer. Except when it doesn’t.

For reasons that are not entirely clear, the T cells that normally mount an effective response against infections and cancer sometimes find themselves unable to vanquish these foes and give up.

This flagging, dysfunctional state is a focus of much attention among immunologists who seek to find ways to prod the cells into fighting another day. But there are challenges to making sense of this phenomenon.

“Immunologists who study conditions as varied as chronic infections, cancer, and autoimmunity are all interested in this phenomenon of T cell dysfunction,” says Christina Leslie, a computational biologist at the Sloan Kettering Institute. “But because the biological contexts in which they’re studying dysfunction are so varied, it is not clear if we’re dealing with a single phenomenon. Even the way people talk about it tends to be different.”

Those who study chronic infections, for example, are more likely to refer to T cell “exhaustion” — implying that the cells are basically burnt out from fighting too hard — whereas cancer immunologists are more likely to refer to T cell “dysfunction” — which implies they didn’t try to fight hard in the first place.

Computational & Systems Biology Program
The goal of our research is to build computer models that simulate biological processes, from the molecular level up to the organism as a whole.

To help clear up the matter, Dr. Leslie and a postdoc in her lab, Yuri Pritykin, decided to bring their lab’s particular expertise — mathematical computational biology — to bear on the abundant yet confusing data in the literature about T cell dysfunction. Their results, published April 22, 2021, in the journal Molecular Cell, provide what they call a “unified atlas” of dysfunction, which they argue is a single phenomenon regardless of whether one is talking about infection or cancer. The researchers hope their findings will put to rest a nagging controversy in the field.  

Squeezing More Out of Public Data

The project started as an attempt to computationally understand rapid-fire developments in the field. “We were waiting for results of a collaborator’s experiments and realized there was all this published data in the literature that was hard to compare across studies and therefore difficult to interpret,” Dr. Leslie says. “And we realized that as computational biologists we were well suited to perform the necessary operations on the data to make them comparable — called ‘batch effect correction’ in our field.”

By uniformly reprocessing and correcting for batch effects across 12 previously published studies of T cell exhaustion/dysfunction comprising hundreds of next-generation sequencing experiments, the scientists were able to see a clear pattern running through all of them. 

“All the changes described in the studies line up along a single axis,” says Dr. Pritykin, who is now a professor at Princeton University. “From these data, it’s clear that we are all describing the same thing.”

Instead of the exhaustion model, the findings instead support a view of T cell dysfunction in which the cells figure out early on in the encounter that mounting a full-on fight will not be advantageous, so they stop trying. This latter model implies that immunotherapy is not “reinvigorating” exhausted cells but rather harnessing T cells that have not gone too far down the dysfunctional path.

Immunologists who study conditions as varied as chronic infections, cancer, and autoimmunity are all interested in this phenomenon of T cell dysfunction.
computational biologist

In addition to uncovering this common underlying path to dysfunction, the researchers were able to computationally identify the various transcription factors (proteins that control gene activation) that are active at each successive time point. They also compared the makeup of transcription factors in dysfunctional T cells versus functional T cells actively fighting infections to get a clearer picture of the changes cells undergo as they become dysfunctional. And, through a collaboration with Alexander Rudensky’s lab at SKI, they added rich single-cell data sets to their study, which maps out how functional and dysfunctional T cell responses diverge at an early time point.

The resulting atlas of molecular changes will be useful to researchers who are struggling to understand this dysfunctional state, including those looking for ways to prevent or reverse it. It may be useful, for example, to doctors treating people with CAR T cells, a type of immunotherapy.

“The idea is that if you knocked down transcription factors that are associated with the progression to dysfunction then the CAR T cells might not become exhausted as quickly when you put them in the patient, and so you’ll get a better outcome,” Dr. Leslie says.

Basic researchers will likely find the atlas useful too. “People can just look up their gene of interest and see in detail where it’s expressed and whether it’s associated with a functional or dysfunctional response,” Dr. Pritykin says. “I think the analysis and the dataset is going to be really valuable.”

This study was supported by NIH grants U54 CA209975 and R01AI034206, the Alan and Sandra Gerry Metastasis and Tumor Ecosystems Center, the Ludwig Center at Memorial Sloan Kettering Cancer Center, the Parker Institute for Cancer Immunotherapy, AACR-Bristol Myers Squibb Immuno-oncology Research Fellowship (19-40-15-PRIT), MSK-PICI-Cycle for Survival Fund, the NCI Cancer Center Support Grant (P30 CA008748), Cycle for Survival, and the Marie-Josée and Henry R. Kravis Center for Molecular Oncology. Dr. Rudensky is an investigator with the Howard Hughes Medical Institute.