Investigation of endogenous and adoptively transferred T cell function in a genetic mouse model of lung adenocarcinoma

Abstract

The clinical success of immune checkpoint blockade, including in patients with lung cancer, has clearly demonstrated that the body's immune system is capable of tumor cytotoxicity if modulated appropriately. Unfortunately, only a minority of patients with solid tumors respond to checkpoint blockade, and even fewer respond to other immunotherapeutic modalities like adoptive cell therapy (ACT). The reasons for these low response rates are not well understood, suggesting an improved understanding of the mechanisms which shape a tumor immune response and promote T cell dysfunction is needed. Toward these ends, we first sought to characterize the transcriptional changes leading to toward dysfunction in endogenous T cell responses. Using an autochthonous mouse model of lung adenocarcinoma, we longitudinally profiled CD8⁺ T cells from the lungs of tumor bearing mice using single cell RNAseq (scRNAseq).

We identified significant longitudinal changes within the tumor specific T cell population over the course of more than 20 weeks of tumor development. Among these transcriptional changes was a transition from functional effector cells, to states of T cell exhaustion. We used this information to develop a signature which identifies T cells that are more readily reinvigorated by checkpoint blockade. We also identified multiple factors which promote heterogeneity within the tumor specific T cell response including TCR affinity for antigen and antigen identity within a dominance hierarchy. We close this 2nd chapter by demonstrating that this transcriptional information allows for the identification of mediators limiting anti-tumor T cell responses. Modern adoptive cell therapy has demonstrated remarkable, durable efficacy in treating patients with hematological malignancies but has failed to deliver comparable efficacy in patients with solid tumors like NSCLC.

In Chapter 3, we transcriptionally profiled adoptively transferred cells in our autochthonous mouse model of lung cancer and identified multiple mechanisms which limit ACT efficacy. We used this transcriptional data to perform focused in vivo CRISPR mediated screens to identify mediators of T cell dysfunction. We identify a number of genes limiting T cell persistence and functionality within the tumor microenvironment and demonstrate that by ablating expression of these genes, we can dramatically improve the efficacy of ACT.