Reversing T cell dysfunction in cancer by multimodal genetic screening

T cell dysfunction is a key problem in cancer, enabling not only tumorigenesis but also causing resistance to immunotherapy.

Induction of immune checkpoints is a hallmark of T cell dysfunction, but clinical blockade by PD-1 and CTLA-4 antibodies solves this problem for few patients only. Dysfunction is driven also by additional mechanisms, including chronic stimulation and metabolic insufficiency. A better mechanistic understanding will be imperative for improving immunotherapy.

My laboratory recently launched ReverT, a genome-wide CRISPR-Cas9 screening program to identify genes, ablation of which reverses dysfunction in primary T cells.

To de-risk this application, we have already successfully completed three such screens in vitro, for chronic antigen stimulation, metabolic stress and PD-1 induction, validating and characterizing several dysfunction genes not previously reported.

Furthermore, an independent follow-up re-screen with a Dysfunction Reversion Candidate (DRC) mini-library containing the top 100 hits of each of these screens validated >100 genes that reversed T cell dysfunction in vitro.

Here, we will carry out the most challenging and important step: to systematically validate and mechanistically characterize this collection of T cell dysfunction genes in vivo.

This will be done in a pooled and multimodal fashion, analyzing multiple dysfunction phenotypes in parallel, specifically immune checkpoints, exhaustion, metabolism, recruitment and proliferation.

Our proof-of-concept results indicate that the DRC library contains nodal factors, operating in several seemingly different dysfunction settings, which may thus in fact be linked.

We will use a collection of adoptive cell transfer mouse and human tumor models for validation and mechanistic characterization, as well as primary human T cells in patient-derived tumor fragments. Lastly, we will translate our findings to a preclinical setting, aiming to achieve more durable clinical responses.