Engineered T cells have been used to treat cancer, autoimmune diseases and infections. These cells are altered to express chimeric antigen receptors (CARs) that combine an antigen recognition domain with intracellular T cell signaling components which directs T cell specificity and functionality. By using CAR T cells, researchers can target antigen specific cells and then initiate signaling pathways that trigger effector T cell function. The antigen targeting portion of the CAR is an extracellular single-chain variable immunoglobulin fragment (scFv) that is then fused to intracellular signaling domains, this allows for T cell activation upon ligand binding. Two common costimulatory domains used are CD28 and 4-1BB, which are needed for normal T cell activation and survival. Several current treatments with CAR T cells have been approved and more are in development, however there are some risks involved. These treatments include toxicities, such as cytokine release syndrome and neurotoxicity caused by intracellular signaling pathways triggered by CAR activation. Work to better understand the signaling cascades and effects of co-stimulatory domains on CAR function is being done at Fred Hutch in the Vaccine and Infectious Disease and Clinical Research Divisions and was recently published in the journal Science Signaling. As Dr. Salter, first author of the study explains, “Scientists who create engineered CAR T cell therapies often incorporate one of two costimulatory domains — known as CD28 and 4-1BB — into CARs. Prior work has shown that CD28 CAR T cells and 4-1BB CAR T cells behave differently in the lab and in the clinic, but researchers have not known how the costimulatory domain impacts the identity or quality of the T cell activating signals delivered by CARs”.
In order to identify why the different costimulatory domains have differing phenotypes and functions, the group used liquid chromatography-tandem mass spectrometry (LC-MS/MS) to produce comprehensive quantitative data. This allowed for interrogation of the signaling pathways that were activated by CD28 or 4-1BB CARs. CD28 or 4-1BB CAR T cells were stimulated with matched antigen then LC-MS/MS was performed to map phosphorylation events to signaling pathways. LC-MS/MS identified 26,804 phosphorylation sites with activation of both CARs producing nearly identical changes in protein phosphorylation. This suggested that differences seen in CAR effector function and phenotype was not dictated by unique pathways. However, the CD28 CAR had significantly higher intensity and kinetics of signaling compared to 4-1BB. This increased signaling strength lead to a robust fast acting effector cell-like profile with more cytokine production which was short lived leading to T cell exhaustion and reduced antitumor activity. On the other hand, 4-1BB CARs had lower signaling intensity which lead to more longevity of the population. “Overall, our findings challenge an assumption in the field that altering the costimulatory domain changes the identity of the signaling cascades. Our data indicate that signaling cannot be fully predicted by the identity of the CAR signaling domains; one also needs to account for what molecules may be associating with CARs on the T cell surface.” stated Dr. Salter.
This data suggests that the current idea of selecting CAR T cells based on maximal effector function and proliferation may not be the way to go. Like most biological systems, there are more interactions that should be considered, including binding partners, signal strength and T cell function. Future work will focus on modifications of the CD28 CAR T cell to alter signaling strength as well as the development of a high throughput multiplexed mass spectrometry assay to be used to characterize future CAR T cell therapies.
Salter AI, Ivey RG, Kennedy JJ, Voillet V, Rajan A, Alderman EJ,Voytovich UJ, Lin C, Sommermeyer D, Liu L, Whiteaker JR, Gottardo R, Paulovich AG, Riddell SR. 2018. Phosphoproteomic analysis of chimeric antigen receptor signaling reveals kinetic and quantitative differences that affect cell function. Science Signaling, 11(544).
Funding was provided by the National Institutes of Health and Fred Hutch.
Fred Hutch/UW Cancer Consortium faculty members Raphael Gottardo, Amanda Paulovich, and Stanley Riddell contributed to this work.