Back

Cheminformatic identification of small molecules targeting acute myeloid leukemia

Daneman, M. R.; Meika, B.; Tjahjono, E.; Revtovich, A. V.; Stolbov, L. A.; Gilbertson, S. R.; Poroikov, V. V.; Kirienko, N. V.

2025-05-24 cancer biology
10.1101/2025.05.20.655224 bioRxiv
Show abstract

Acute myeloid leukemia (AML) is an aggressive hematological malignancy that has poor prognosis and high relapse rates with cytotoxic chemotherapeutics. Previously, we identified modulators of mitochondrial function, PS127-family compounds, that were cytotoxic to AML and were characterized by two predicted functions: apoptotic agonism and thioredoxin/glutathione reductase inhibition (T/GRi). Here, we uncovered a third critical predicted function: autophagic induction. A cheminformatics screening of [~]4.2 million compounds for molecules with high probability of these three functions yielded 93 hits, 81 of which were closely related to PS127-family molecules. In silico hits selected for validation selectively killed AML cells, activated apoptosis, required functional autophagy, and interfered with glutathione metabolism, confirming predicted functions. This increased pools of cytosolic and mitochondrial ROS and decreased oxygen consumption and ATP synthesis. Differential scanning fluorimetry implicated glutathione reductase as a direct target of these molecules. Structurally-unrelated compounds from different clusters caused the same phenotype, validating our structure-blind screening approach. Furthermore, strong synergy between these compounds and the AML treatment midostaurin underscores their therapeutic potential.

Matching journals

The top 11 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.