Adoptive cell therapy (ACT) is an exciting area of cancer immunotherapy research in which a patient's own immune cells are removed from their body, modified in vitro, and re-injected to attack the cancer. The most successful form of modification is to engineer T cells from the patient with chimeric antigen receptor (CAR) proteins, which trigger an immune response upon recognition of a cancer cell. To activate the T cells, CARs are composed of (1) an extracellular domain, derived from an antibody, that can bind to a tumor associated antigen and (2) several intracellular signaling domains, derived from endogenous T cell receptors that can activate an immune response. Currently three generations of CARs have been developed, each increasing the number of intracellular co-stimulatory domains present on the CAR. CARs containing the CD3? signaling domain in combination with either CD28 or 41BB co-stimulatory domains have shown some promise in clinical trials of B cell lymphoma; however, some patients do not respond to therapy, while others experience over activation of the immune system, which can be deadly. It is not clear how further increasing the number of co-stimulatory domains will affect T cell activation or if the increase will be able to improve the control over CAR therapy. Therefore, a deeper understanding of the cell signaling pathways that lead to CAR T cell activation is needed to describe how signaling domain pathways integrate to affect the various fucntions of T cell activation. Systems biology, specifically computational mechanistic modeling, provides a unique platform to understand and optimize the activation of CAR engineered T cells. This proposal aims to explore the effects of different CAR signaling domains, individually and in combination, by developing a set of mechanistic computational models that can predict T cell activation mediated by different CARs. The models will describe the mechanisms by which activation of individual signaling domains, which have not been previously modeled, affect downstream proteins that correlate to specific properties of T cell activation, such as cytokine production, T cell survival, and cell proliferatin. The models will be comprised of ordinary differential equations derived from known interactions in the literature. They will be trained on flow cytometry data of protein binding and activation such that the models are able to predict the effects of CAR stimulation. We will perform in silico experiments to predict how signaling of individual CAR co-stimulatory domains is integrated to affect particular aspects of the T cell response. For example, we can optimize CAR therapy for patients that do not respond to treatment by using the model to determine the optimal signaling domain combination to increase proteins that correlate to cytokine production without affecting those that correlate to survival. These hypotheses will then be validated by in vitro experiments. Thus, the proposed research will provide valuable information to improve the safety and efficacy of CAR engineered T cells, enabling the benefits of this therapy to be expanded to a wider patient population.
Immunotherapy is currently one of the most promising areas of cancer research, but the progress of these new treatments is limited by a poor understanding of the interactions that activate and control the immune response mediated by T cells. Computational models can be used to understand the signaling pathways that lead to T cell activation, predict the complex dynamics of these pathways, and test different hypotheses regarding how to effectively activate T cells. This project aims to develop a mechanistic model of engineered T cell activation and apply the model to optimize immunotherapy treatments for cancer.
|Rohrs, Jennifer A; Wang, Pin; Finley, Stacey D (2016) Predictive Model of Lymphocyte-Specific Protein Tyrosine Kinase (LCK) Autoregulation. Cell Mol Bioeng 9:351-367|
|Rohrs, Jennifer A; Sulistio, Christopher D; Finley, Stacey D (2016) Predictive model of thrombospondin-1 and vascular endothelial growth factor in breast tumor tissue. NPJ Syst Biol Appl 2:|