High-throughput screening uncloaks actionable compounds for ttmv::rara AML

Abstract Abstract Acute myeloid leukemia (AML) with TTMV::RARA fusion, which exhibits clinical features resembling acute promyelocytic leukemia (APL), represents a newly identified and aggressive subtype lacking established therapies. This fusion originates from genomic integration of TTMV, a virus acquired by over 90% of individuals shortly after birth that persists lifelong without causing disease. Clinical data from our multicenter cohorts (n=25) revealed that 20 patients showed initial response to conventional APL regimens including all-trans retinoic acid (ATRA) and arsenic trioxide (ATO), but paradoxically demonstrated poor disease control, evidenced by the event-free survival (EFS) was 53.6% with relapse-free survival (RFS) exceeding to 53.8% beyond 2 years. This highlights an urgent need for optimized therapeutic strategies. To systematically identify effective agents, we established a TTMV::RARA-overexpressing human CD34+ cell model. Compared to the vector group, overexpressing cells showed markedly increased proliferation and enhanced self-renewal activity. Immunophenotyping revealed accumulation of immature myeloid cells (CD45+ CD33+ CD34+) with diminished mature granulocytes (CD45+CD33+CD66b+) and differentiating myeloid cells (CD45+CD33+CD11b+). This CD34+model faithfully recapitulates clinical phenotypes observed in patients, including leukocyte hyperproliferation and myeloid maturation arrest, thus validating the pathogenic role of TTMV::RARA fusion. Using this disease model, we performed high-throughput screening of bioactive compounds to identify actionable vulnerabilities. Primary screening identified the top 10 anti-leukemic compounds, categorized pharmacologically as phosphodiesterase inhibitors, selective estrogen receptor modulators (SERMs), tyrosine kinase inhibitors (TKIs, comprising 7 agents), and DNA synthesis inhibitors, which were subsequently evaluated in CD34⁺ model. Dose-response curves revealed significant half maximal inhibitory concentration (IC50) reductions (>5-fold) relative to vector controls in 5/10 candidates: anagrelide (phosphodiesterase III inhibitor), lenvatinib (multi-kinase inhibitor), cytarabine (nucleoside analog), cyclocytidine (prodrug of cytarabine), and tandutinib (FLT3 inhibitor). Crucially, the established APL therapies ATRA and ATO, as well as the widely used BCL2 inhibitor venetoclax, showed no significant reduction in IC50, suggesting a possible link to the clinical observation of rapid relapse in this disease. Our findings establish the human CD34⁺ model as a genetically system uniquely suited for studying TTMV::RARA-driven leukemogenesis. Combining the CD34⁺ model and high-throughput screening, we identified multiple classes of targeted agents which may against this fusion oncoprotein. This work not only provides a definitive experimental platform for mechanistic investigations but delivers clinically actionable therapeutic candidates to address the this novel AML subtype.