TP53 Deficiency in AML Confers Resistance to CAR T-Cells That Can be Overcome By Targeting the Cholesterol or Wnt Pathways

Introduction

TP53-mutant acute myeloid leukemia (AML) and myelodysplastic neoplasms (MDS) are distinct clinicogenomic entities characterized by chemotherapy resistance and high relapse rates after allogeneic stem cell transplantation, and thus, represent an unmet clinical need. Chimeric antigen receptor (CAR) T-cell therapy might be a promising therapeutic option for TP53-mutant AML/MDS. However, the AML-intrinsic determinants of response to T-cell-based approaches in AML, in particular the impact of TP53-deficiency, are largely unknown. In the present study, we addressed the biologically and clinically highly relevant question as to whether, and, if so, how TP53 deficiency in AML cells might confer resistance to CAR T-cell therapy.

Results

We took advantage of recently developed, CRISPR/Cas9-engineered TP53 isogenic AML cell lines harboring TP53 null, missense, or wildtype alleles to exclude other potential genetic confounders of therapy resistance. Flow cytometry-based co-culture assays using these isogenic cell lines together with CAR T-cells targeting the common AML surface antigens CD33, CD371, CD123, and CD117, showed that AML cells harboring TP53 null or missense mutations were relatively resistant to CAR T-cells in vitro. CAR T-cells engaging TP53-deficient leukemia cells proliferated less, upregulated exhaustion markers, displayed increased trogocytosis, and were ultimately unable to control AML cell outgrowth. Live-cell imaging revealed a longer engagement time of CAR T-cells with TP53-deficient AML cells before killing occurred as judged by propidium iodide influx. Immunodeficient mice xenografted with TP53-deficient AML cells and treated with CAR T-cells exhibited significantly shortened survival compared to TP53-wildtype AML controls. Transcriptional profiling revealed that the cholesterol pathway was upregulated in TP53-deficient AML cells under CAR T-cell attack. Simultaneously, CAR T-cells engaging TP53-deficient AML demonstrated a downregulated Wnt pathway, in particular the master regulators of T-cell fate TCF7 and EOMES. Rational pharmacological targeting of either of these pathways, blocking by Simvastatin and activation by BIO-acetoxime, respectively, rescued TP53-deficient AML cell sensitivity to CAR T-cell-mediated killing in vitro.

Discussion and Outlook

Using a combination of in vitro co-incubation assays, live-cell imaging, gene expression profiling and a therapeutic in vivo model we could demonstrate that TP53 deficiency in AML cells confers relative resistance to CAR T-cell therapy and elucidate some of the underlying molecular mechanisms in leukemia target cells and CAR T-cells. We thus propose a model, in which the intrinsic apoptotic defect in TP53-deficient AML/MDS cells results in a longer duration of the cellular interaction between CAR T-cells and TP53-deficient AML/MDS cells. This longer temporal interaction eventually leads to reduced CAR T-cell proliferation and enhanced CAR T-cell exhaustion with an overall decrease in AML cell killing (figure 1). Furthermore, we identified the cholesterol pathway as a potential therapeutic vulnerability of TP53-deficient AML cells engaged by CAR T-cells, and the Wnt pathway as a promising avenue to enhance the efficacy of CAR T-cell therapy in TP53-deficient AML/MDS. Our data suggest that the combination of CAR T-cell therapies with pharmacological co-interventions – as exemplified in this study – may be a preferable strategy towards more efficacious and tolerable cellular therapies for patients with TP53-mutant myeloid neoplasms.