Patient-informed CRISPR screening reveals gene targets to improve CAR T-cell efficacy

Abstract Background: Chimeric Antigen Receptor (CAR) T-cell therapy targeting CD19 has transformed outcomes for patients with relapsed and refractory B-cell malignancies. Despite its success, more than half of treated patients ultimately experience disease recurrence, underscoring the need to better understand treatment failure and to enhance CAR T-cell effectiveness and persistence. In a prior study (Fraietta et al., Nature, 2018), our group described an exceptional clinical benefit in a patient with chronic lymphocytic leukemia (CLL), driven by the clonal expansion of a CD8+ CAR T-cell harboring a lentiviral integration in the TET2 gene. This observation led us to hypothesize that certain gene disruptions resulting from vector integration may contribute to improved therapeutic efficacy. We analyzed lentiviral vector integration sites (LVIS) from CAR T-cells across 40 patients with acute lymphoblastic leukemia (ALL) and CLL (Nobles et al., JCI, 2020). Several genes were found to be recurrently targeted and clonally dominant across patients with similar outcomes. Methods: To mechanistically interrogate LVIS-based disruptions associated with long-term responders, we designed a focused CRISPR-Cas9 library targeting 180 genes, using 4-8 single guide RNAs (sgRNA) per gene. Our one-shot lentiviral vector was engineered to drive the expression of both the anti-CD19 CAR and gRNAs under the promoters EF1-alpha and human U6, respectively. To generate gene-modified CAR T cells, healthy donor T cells were transduced at a low multiplicity of infection to ensure single-gene edits per cell. Three days post-transduction, CAR-positive cells were purified, electroporated with the Cas9 protein and subsequently expanded in culture for an additional 10 days. As a negative control, CAR T-cells containing the sgRNA library were mock-electroporated without Cas9 and expanded under identical conditions. The impact of the gene edits on CAR T-cell fitness was assessed in an in vitro chronic stimulation assay using CD19-expressing leukemia cells, and in our well-established in vivo immunodeficient xenograft model engrafted with the aggressive leukemia cell line NALM6. Positive hits from this screen were further validated via single gene edits using individual sgRNA-loaded Cas9 ribonucleoproteins (RNPs). Each gene was targeted using 2 sgRNAs to ensure 80% knockout efficiency. The anti-leukemia efficacy, proliferative response, cytokine polyfunctionality, CAR T cell effector and memory differentiation, and global gene expression changes were measured for the top candidates. Results: We identified an overlapping set of highly ranked hits (Log2FC ≥ 0.25, FDR ≤ 0.1) in our in vitro and in vivo CRISPR-Cas9 screens that significantly enhanced proliferation and effector function upon repeated tumor challenges when compared to the control groups. These genes were mostly associated with MAPK signaling, DNA damage, apoptotic pathways, cell migration, and cell cycle. Importantly, during our in vivo screen, although the edited CAR T-cells (+Cas9) presented a delayed anti-tumor response compared to unedited CAR T-cells (-Cas9), mice from the +Cas9 group survived longer with a lower tumor burden when compared to mice from the -Cas9 group. Results from our validation using Cas9-RNPs for our top-ranked genes showed that several genes exhibited increased proliferation and anti-tumor efficacy compared to unedited cells and our safe-harbor control AAVS1. Mice treated with single-edited CAR T-cells, particularly with CARD8 and SRCAP disruptions, displayed improved survival and tumor clearance compared to the control group. Phenotyping of the cells as well as RNA sequencing are currently ongoing and will be presented at ASH 2025. Conclusions: Our CRISPR-library screens validated the functional relevance of LVIS-based disruptions identified in CAR T-cell responders (Nobles et al., JCI, 2020). Our in vitro and in vivo pipelines successfully recapitulate what has been observed in patients through the enrichment of genes implicated with T-cell proliferation and survival. Preliminary assessment of several genes highlights promising candidates which can be modulated to enhance CAR T-cell efficacy.