Quantifying the Kinetics of Immune Responses to Cancer: Applications in Serum Profiling and Single-Cell Biology



Journal Title

Journal ISSN

Volume Title



Immunotherapy that harnesses the body’s immune system to fight cancer has revolutionized the treatment of the disease. Advances in utilizing checkpoint inhibitors that release the potential of the adaptive immune system, and in the design and manufacture of chimeric antigen receptor (CAR) T cells that utilize T cells as living drugs, has served to alter the landscape of cancer treatment. Profiling responses to immunotherapy requires the development of newer methods that can deal with the breadth and complexity of these responses. In this dissertation, I demonstrate the advancement of two approaches that answer central immunological questions: (1) measuring the breadth of humoral responses elicited upon the development of cancer and whether these have prognostic/therapeutic potential, and (2) modeling the kinetics of the single-cell killing mediated by T cells to identify mechanisms for the manufacture of more potent cells for immunotherapy.

In the first part, a combinatorial library of phage-displayed linear dodecapeptides was bio-panned against plasma samples from a cohort of AML patients undergoing checkpoint therapy. Subsequent to recovery of phage particles and high-throughput sequencing, we utilized a novel biodiversity-based analysis to identify candidate peptides. We validated this workflow by profiling humoral responses elicited upon seasonal vaccination against influenza. By utilizing this same methodology for the interrogation of the plasma of AML patients, we demonstrated the discovery and characterization of peptides derived from AML-specific oncoprotein fusions.

In the second part, we sought to understand the mechanistic basis of the failure of CAR T cells to kill antigen-positive target cells. We utilized single-cell timelapse imaging of CD19-targeting CAR T cells and their interactions with leukemia cells to distinguish killer and non-killer CAR T cells. We hypothesized that the nature in which effector-target associations transition from conjugation to killing/detachment occurs in a gamma-distributed fashion, implying a sequential transition across intermediate states during contact. Our modeling results showed that kill events are kinetically homogeneous, characteristic of a single rate-limiting step, whereas no-kill events were heterogeneous mixtures of abortive and persistent contacts. The results of the model were validated by microscopy experiments that illustrated defects in lysosome polarization and degranulation within non-killer CAR T cells.



Phage display, Acute myeloid leukemia, Humoral response, High-throughput sequencing, CAR T cells, Effector-target association, Kinetic modeling