PROJECT SUMMARY
This proposal aims to elucidate the molecular circuitry that limits the efficacy of chimeric antigen receptor
(CAR)-engineered T cells. CARs are synthetic proteins that, when expressed by T cells, can mediate potent
activity against cancer. The most successful clinical application of this technology targets CD19 on the surface
of B cell leukemia and lymphoma. While curative for some, durable remissions are seen in less than half of
patients treated. Emerging data point to CAR T cell failure, as opposed to cancer cell resistance, as the driver of
poor outcomes and disease progression. These findings highlight that understanding how CARs direct T cell
function and dysfunction is of critical importance.
The form of T cell dysfunction that is best understood is exhaustion, a cell state that develops as a result of
chronic stimulation of the endogenous T cell receptor (TCR) in the setting of chronic viral infection. All CARs
contain the primary signaling domain of the TCR as well as an additional costimulatory domain to enhance T cell
activation. Of the six CAR products approved for clinical use in the US, four contain the costimulatory domain
from 41BB while two contain the domain from CD28. To identify the CAR-activated molecular pathways
responsible for T cell failure we developed an in vitro system that induces T cell dysfunction through chronic
CAR stimulation. We observed that while 41BB and CD28-based CAR T cells both lose function as a result of
chronic stimulation, these receptors activate divergent molecular circuitry as they lose function. CD28-based
CAR T cells are phenotypically, transcriptionally and epigenetically similar to exhausted T cells. In contrast,
41BB-based CAR T cells do not bear the hallmarks of classic exhaustion but are distinct at all levels of
programming. We observed activation of this unique transcriptional signature of 41BB-driven dysfunction in CAR
T cells that failed to control disease in patients. We further identified that the transcription factor FOXO3 is highly
active in dysfunctional 41BB-based CAR T cells and that deletion of FOXO3 improves CAR T cell function.
Based on these observations, we hypothesize that chronic stimulation of 41BB triggers signaling pathways
that aberrantly activate FOXO3 and that FOXO3 activates several distinct programs that suppress CAR T cell
function. In the proposed studies we will use biochemical techniques and high-resolution microscopy to identify
the intermediate signaling proteins that link 41BB and FOXO3 (Aim 1), use conditional expression systems to
determine how FOXO3 suppresses T cell function (Aim 2) and identify other engineered T cell therapies that are
vulnerable to 41BB-driven dysfunctional programs (Aim 3). Collectively, these studies will elucidate how chronic
activation of 41BB leads to a novel state of T cell dysfunction and evaluate translationally-relevant strategies for
advanced engineering that overcome this dysfunction-inducing circuitry.