Improving engineered TCR-T cell therapy for solid tumors by targeting Tcf1+ stem-like reservoirs in tumor-draining lymph nodes

PROJECT SUMMARY Adoptive transfer of engineered T cells targeting tumor antigens is ineffective against most solid tumors, due in-part to poor persistence and the rapid onset of dysfunction. Recent studies of endogenous tumor- specific T cells (Endog-T) have described a population of stem-like PD-1+Tcf1+ progenitor exhausted T cells

(Tpex) that can self-renew and differentiate into effector-like T cells in the tumor. Tpex preferentially reside in tumor-draining lymph nodes (TdLNs), where they are maintained through activation by dendritic cells (DCs). Importantly, Tpex in TdLNs are necessary for promoting persistent and functional T cell responses, as well as

conferring response to PD-L1 blockade. However, it is not clear whether engineered T cells form PD-1+Tcf1+ populations in TdLNs, and how this impacts their long-term persistence and function in the tumor. To test this, I adapted the clinically relevant KrasG12D-LSL/+;p53fl/fl murine model (KP) of lung adenocarcinoma

to overexpress the model neoantigen Ovalbumin (Ova) to compare Ova-specific engineered T cells (TCR-T) and Endog-T cells. While the majority of Endog-T cells in TdLNs were PD-1+Tcf1+, TCR-T cells were predominantly PD-1-Tcf1+, markers associated with lack of activation and maintenance of a pre-exhausted

central-memory phenotype. Consistent with these phenotypes, PD-L1 blockade failed to increase numbers of TCR-T cells and enhance tumor control in comparison to Endog-T. Thus, I hypothesize that factors unique to TCR-T therapy, such as their in vitro pre-activation and/or infusion at a high cell number, promote their

formation of PD-1-Tcf1+ reservoirs in TdLNs instead of PD-1+Tcf1+. I also hypothesize that formation of PD-1- Tcf1+ reservoirs in the TdLN is actually an advantage for TCR-T therapy, since the PD-1-Tcf1+ population may possess superior functional potential that can be mobilized through vaccination targeting DCs in the TdLN.

In this project, I seek to determine 1) why TCR-T cells form PD-1-Tcf1+ reservoirs in TdLNs instead of PD- 1+Tcf1+ like what is described for Endog-T, as well as 2) whether the PD-1-Tcf1+ TCR-T cells in the TdLN have stem-like characteristics and functional potential that can be unleashed with vaccination. In Aim 1, I will use

fluorescently-labeled tumors in the KP model, flow cytometry, and high-parameter fluorescence microscopy to determine how infusion of a large number of cells and/or in vitro pre-activation affect TCR-T cell formation of PD-1+Tcf1+ populations in TdLNs, and how this correlates with their localization near DCs. In Aim 2, I will use

RNA-seq, ATAC-seq, and ex vivo assays to describe the stem-ness and functional potential of the PD-1-Tcf1+ TCR-T cells in the TdLN compared to PD-1+Tcf1+ Endog-T. I will also evaluate whether a novel T cell vaccination approach targeting DCs in TdLNs can activate PD-1-Tcf1+ TCR-T cells to promote tumor control.

Through this work, I will define the mechanisms by which TCR-T cells form Tcf1+ stem-like reservoirs in TdLNs and potentially identify a therapeutic strategy to enhance TCR-T efficacy by targeting these reservoirs that could be translated to the clinic to improve patient outcomes.