AUTHORS
Phuong Le, Anna Baghnavi2, Ernst Mayerhofer3, Trine A Kristiansen, PhD4, Tobias Feuchtinger, MD5, David T Scadden, MD6, Christina Mayerhofer, MD7
AFFILIATIONS
1 Children’s Hospital, Medical Center and Faculty of Medicine, University of Freiburg, Freiburg, Germany, German Cancer Consortium (DKTK), Partner Site Freiburg; and German Cancer Research Center (DKFZ), Heidelberg, Germany., Freiburg, Germany
2 Children’s Hospital, Medical Center and Faculty of Medicine, University of Freiburg, Freiburg, Germany, German Cancer Consortium (DKTK), Partner Site Freiburg; and German Cancer Research Center (DKFZ), Heidelberg, Germany., Freiburg, Germany
3 Center for Genomic Medicine, Massachusetts General Hospital, Boston, United States
4 Harvard Stem Cell Institute, Department of Stem Cell and Regenerative Biology, Harvard University, Cambridge, United States
5 Children’s Hospital, Medical Center and Faculty of Medicine, University of Freiburg, Freiburg, Germany
6 Center for Cancer Research and Center for Regenerative Medicine, Massachusetts General Hospital, Boston, United States, Harvard Stem Cell Institute, Department of Stem Cell and Regenerative Biology, Harvard University, Cambridge, United States
7 Children’s Hospital, Medical Center and Faculty of Medicine, University of Freiburg, Freiburg, Germany, German Cancer Consortium (DKTK), Partner Site Freiburg; and German Cancer Research Center (DKFZ), Heidelberg, Germany., Freiburg, Germany, Harvard Stem Cell Institute, Department of Stem Cell and Regenerative Biology, Harvard University, Cambridge, United States
ABSTRACT
Background and Objective: Cytotoxic chemotherapy remains the therapeutic backbone for acute myeloid leukemia (AML), yet persistent leukemic subpopulations survive and drive minimal residual disease and relapse. While baseline tumor metabolic dependencies are increasingly defined, how acute chemotherapy-induced metabolic remodeling directs translational control and ribosomal stop-codon readthrough (SCR) remains largely unresolved. Here, we sought to define therapy-induced metabolic changes and their dynamic influence on ribosomal mRNA translation and translation fidelity in AML.
Methods: We analyzed untargeted, targeted, and plasma metabolomics with paired transcriptomics (RNA-seq) and ribosome profiling (Ribo-seq) in primary murine AML (HoxA9/Meis1 and MLL-AF9) treated in vivo with standard induction chemotherapy (cytarabine/doxorubicin) or venetoclax/azacitidine (VenAza). To evaluate human AML and post-chemotherapy dynamics, we profiled patient-derived xenografts (PDX) capturing minimal residual disease and fast relapse settings.
Results: Immediately following acute chemotherapy exposure in vivo, surviving leukemic cells exhibited a profound amino-acid metabolic signature characterized by accumulation of arginine, glutamine, and glutamate and depletion of the TCA-cycle intermediate alpha-ketoglutarate. This metabolic program reflects an acute shift away from glutamine-fueled TCA-cycle anaplerosis toward increasing intracellular amino-acid pools.
Comparing translatomics with transcriptomics revealed distinct differences in chemoresistant cells. While transcriptomes upregulated broad cellular stress-response programs, translatomes revealed direct repression of canonical growth-associated translation initiation factors (eIFs) and 40S/60S ribosomal subunit assembly machinery. Concurrently, translational efficiency (translation/transcription) was selectively prioritized for genome maintenance networks, specifically boosting the translation of homology-directed repair and telomere extension genes.
At the level of translational fidelity, acute chemotherapy stress globally elevated ribosomal SCR efficiency. This readthrough event induced specific SCR targets under chemotherapy, such as pro-survival, epigenetic, and metabolic regulators as well as drivers of the Integrated Stress Response. By generating functional C-terminally extended variants of critical proteins, SCR might serve as a metabolically responsive mechanism of proteome expansion under stress.
To determine if human AML maintains the metabolic amino acid programs required for translational adaptation, we characterized patient-derived xenograft (PDX) models of chemosensitive minimal residual disease and chemoresistant rapid relapse. Following therapy, broader amino-acid pathway alterations were present as a metabolic imprint in both models. Furthermore, chemoresistant rapid relapse uniquely expanded this metabolic supply into a hyper-metabolic state with elevated glutamine, glutamate, and ATP reserves.
Together, these data demonstrate that AML cells deploy an acute metabolic response to chemotherapy that drives stress-induced translational adaptation during leukemic persistence and relapse.
Conclusions: Acute chemotherapy stress induces an amino-acid-driven metabolic program in AML. This program initially represses global protein synthesis while prioritizing translational output for genome maintenance. As this transient state differentiates translation from transcription, it upregulates ribosomal stop-codon readthrough as a potential mechanism of proteome expansion under stress. Validated in patient-derived xenografts, this program leaves a temporary metabolic imprint in residual disease before developing into hyper-activation at relapse. Thus, metabolic control of translation may represent a novel vulnerability in chemoresistant leukemia.
DISCLOSURES
Christina Mayerhofer, MD: No relevant conflicts of interest to declare.