Modulating mRNA Translation Fidelity As a Novel Mechanism for AML Chemoresistance

Acute Myeloid Leukemia (AML) is a highly aggressive blood cancer that initially responds to high-dose chemotherapy, but the majority of patients relapse. To understand the mechanisms underlying chemoresistance of AML, we investigated the metabolic adaptations in AML cells that occur following chemotherapy but before overt disease relapse. In AML patient-derived xenografts (PDX), we applied untargeted metabolomics and showed that chemoresistant AML cells have significantly elevated levels of branched-chain amino acids (BCAAs) compared to untreated cells. Elevated BCAA levels and their catabolism have been implicated in AML but their upregulation to drive chemoresistance is not well studied. Transcriptome analysis and stable isotope tracing supported the upregulation of BCAA pathways during chemotherapy in vivo. Our results showed that AML cells, compared to normal granulocyte-monocyte progenitor cells, had significantly higher valine uptake, suggesting a dependency on BCAA metabolism for survival during chemotherapy. When BCAA restriction was combined with chemotherapy, survival was significantly improved with an accompanying clearance of residual AML cells. However, relapse still occurred. Unexpectedly, the surviving cells had significantly higher levels of protein synthesis during maximum response to chemotherapy and BCAA restriction. We anticipated that BCAA restriction after chemotherapy would lower mTOR activity and thus protein synthesis rates. Unexpectedly, we found that ribosomal gene transcripts were enriched, and OPP incorporation in newly synthesized peptides increased in AML cells restricted for BCAAs combined with chemotherapy in vivo. Using low-input Ribo-seq to analyze paired transcriptome and translatome data, we found activated ribosomal protein translation and elevated production of the dioxygenase, 2-oxoglutarate-and-iron-dependent oxygenase 1 (OGFOD1). OGFOD1 modulates the fidelity of ribosomal codon-anticodon pairing through ribosome hydroxylation in proximity to the decoding center where this pairing takes place. Accordingly, AML cells with increased OGFOD1 in vivo had reduced ribosome stalling by ribosome profiling. Conversely, inhibiting OGFOD1 led to translation arrest, reduced protein production and improved animal survival following chemotherapy. Our data therefore suggest that OGFOD1 may enable cells to bypass BCAA shortage by permitting less stringent translation accuracy and targeting OGFOD1 and restricting amino acids can counteract this chemoresistance mechanism. Modifying translation fidelity is a previously unrecognized strategy in AML and has important potential implications across other oncologic contexts.