CRISPR Activation Screen to Optimize Chimeric Antigen Receptor (CAR) T Cell Immunophenotype

Background

Adoptive cellular therapies (ACT) offer potential cures for a wide range of human diseases. Immune cells genetically engineered with chimeric antigen receptors (CAR) have a clear role in B cell malignancies, and broader applications to other malignancies are under investigation. Significant obstacles to the long-term durability and potency of these therapies remain. As CAR T-cells remain persistently activated and exposed to the tumor microenvironment, they may become functionally impaired. This exhaustion can result in reduced production of cytokines, cytotoxic molecules, and other factors necessary for effective tumor cell killing.

While CAR T-cell exhaustion remains a challenge, ongoing research and clinical trials aim to further understand the underlying mechanisms and develop strategies to overcome this limitation. Clinical studies have demonstrated that sustained in vivo persistence and function of CAR T-cells is correlated with infusion product phenotype. Persistence is associated with a less differentiated, memory-like phenotype. Naive (T _N), Central Memory (T _CM) and Stem Cell Memory (T _SCM) are related to good proliferative potential and durability. CRISPR technologies offer powerful, functional genomic screens that permit the inference and modification of gene regulatory networks to optimize CAR T-cell persistence and function. Genome-wide CRISPR knock out screens have identified the CD2-CD58 axis as critical to T-cell potency and the knock-out of the mediator kinase module as beneficial to potency. Here we report on in vitro CRISPR activation screening of human CAR T-cells for gene expression that enhances T cell phenotypes associated with durable function.

Method

We designed 9,983 single guide RNA (sgRNA) to target the transcription start sites (TSS) of all known human transcription factors as well as a select set of genes with critical functions in T-cells. The vast majority of genes was targeted with 5 different sgRNA. To this we added a pool of 1,000 non-targeting sgRNA. We lentivirally transduced the sgRNA in tandem to a second-generation anti-CD20 CAR into primary T cells from healthy human donors. In a second lentiviral transduction we integrated catalytically dead Cas9 (dCas9) fused to the strong transcription activator, VP64 and in tandem to a puromycin resistance gene. After puromycin selection, we cultured the CRISPR activated CAR T-cells either alone or together with Raji or K562 cells for nine days before immunophenotypically sorting the CAR T-cells into CD8 ⁺ Naive, CD8 ⁺ Transitional Memory, CD8 ⁺ Central Memory, CD8 ⁺ Stem Cell Memory, CD8 ⁺ Transitional Effector, CD8 ⁺ Effector Memory, CD4 ⁺ Naive and CD4 ⁺ Memory T-cell populations. We extracted DNA from the sorted cells and prepared sequencing libraries targeting the sgRNA. We carried out two-way ANOVA analysis on the experimental conditions (trivial, Raji or K562 co-culture versus addition of CD2/CD28 activation beads) and T-cell subset to determine which genes targeted by CRISPRa resulted in favorable phenotypes.

Result

Across the different culture conditions, sgRNA activating IL23A and KLF4 were enriched in Transitional Memory versus Stem Cell Memory T-cells. Across the different CAR T-cell immunophenotypes, sgRNA targeting the TGFß (including KLF10), TNFα (including NFATC4) and FOXO signaling pathways were enriched in CAR T-cells cultured alone compared to CAR T-cells activated with CD2/CD28.

Conclusion

In this study we have investigated genes whose activation results in CAR T-cell immunophenotypes that are associated with more potent and durable CAR T-cell therapies. Our experiment suggests that the activation of genes, such as IL23 and KLF4, enriches for a Transitional Memory over Stem Cell Memory phenotype in a manner that is largely independent of environment (culture conditions). The activation of other genes, such as KLF10 and NFATC4, results in enrichment of CAR T-cells across all T-cell subsets, but in a manner that is dependent on environment. Gene expression requires coordinated control of gene-proximal and -distal cis-regulatory elements (CREs), whose state varies as cells differentiate or respond to environmental signals. Abrogating or adding CRE may result in more durable and potent adoptive cellular therapies independent of environment.