ADAR1 Is a Vulnerability in TP53-Mutant Acute Myeloid Leukemia

Among the various subtypes of AML, each defined by distinct oncogenic mutations, those harboring mutations in the tumor suppressor gene TP53 are the most aggressive, presenting a significant clinical challenge. Notably, while TP53 mutations are present in only 10% of patients at diagnosis, their prevalence increases to 30% in secondary AML and treatment-resistant AML, further underscoring the aggressiveness of TP53-mutated (TP53-Mut) AML. Despite efforts to target TP53-Mut AML cells, no targeted therapies have resulted in significant clinical benefit for this patient population. Thus, identifying therapeutic vulnerabilities in TP53-mutated AML is a pressing need.

To identify these vulnerabilities, we leveraged the Cancer Dependency Map (DepMap), a comprehensive repository of genome-wide loss-of-function CRISPR-Cas9 screens across over 1000 cancer cell lines. We stratified the 26 AML cell lines in DepMap into TP53-wildtype (TP53-WT) and TP53-Mut categories and calculated the differential dependency scores between these two subtypes. As expected, we found that the number of selective dependencies in the TP53-WT population exceeded the number of selective dependencies in the TP53-Mut cell lines, underscoring the more aggressive nature of the TP53-Mut cells. The strongest differential dependencies in TP53-WT cell lines were MDM2 and MDM4, both negative regulators of TP53, thereby providing support for the quality of the screening data and our approach.

Importantly, our analysis uncovered differential dependencies in the TP53-Mut cells, including ADAR1 and PRKRA, which were two of the highest scoring dependencies. Interestingly, ADAR1 and PRKRA dependency scores are significantly correlated (r=0.507, p =0.0081), and are components of the same biological pathway, suggesting that this pathway plays a crucial role in the survival of TP53-Mut cells. We focused on ADAR1 due to its potential druggability and multiple ongoing efforts to develop ADAR1-targeting molecules.

ADAR1 catalyzes the conversion of adenosines to inosines (A-to-I) in double-stranded RNA (dsRNA). By editing endogenous dsRNA, ADAR1 prevents the activation of cytoplasmatic dsRNA sensors that trigger the interferon (IFN) pathway, thereby preventing autoimmune related cell death. To investigate the impact of ADAR1 loss on AML cells, we generated ADAR1 knockout (KO) cells in both TP53-WT (MV4-11, MOLM13, OCIAML3) and TP53-Mut (THP1, SHI-1, OCIM2) cell lines. Consistent with our DepMap analysis, our results showed that TP53-Mut AML cells are more sensitive to ADAR1 loss. Notably, KO of MAVS, a key activator of the IFN pathway, reduced the increase of inflammatory markers observed with ADAR1 KO and rescued ADAR1 KO mediated cell death in THP1 and SHI-1 cells. Of note, TP53-KO isogenic MV4-11 cells also exhibited increased sensitivity to ADAR1 loss compared to non-targeting TP53-WT control cells.

To elucidate the mechanism underlying the differential sensitivity of TP53-Mut AML cells to ADAR1 loss, we performed RNA sequencing (RNAseq). Our results show that ADAR1 loss increased the expression of inflammatory pathways in both TP53-WT and TP53-Mut cell lines. However, loss of TP53 enhanced the expression of inflammatory pathways in basal conditions and after ADAR1 KO. Of note, high expression of inflammatory genes is a biomarker for sensitivity to ADAR1 loss. Importantly, we found that patient AML cells with TP53 mutations are enriched for inflammatory-related pathways compared to TP53-WT patient AML cells mirroring our findings in isogenic cells.

TP53 can repress the expression of transposable element (TE). Thus, we hypothesized that TP53 loss leads to increased expression of immunogenic double-stranded RNA (dsRNA), rendering TP53-Mut cells more reliant on ADAR1 editing. Supporting this hypothesis, we observed higher ADAR1 expression and activity (A-to-I index) in TP53-Mut cells compared to TP53-WT cell lines. These results were also recapitulated in our MV4-11 isogenic cell lines.

In summary, this work reveals that TP53-mut cells develop a selective dependency on ADAR1 for survival, strongly suggesting that ADAR1 is a promising therapeutic target in treating the extremely aggressive TP53-Mut AML.