AUTHORS
Xiaoxu Zhang, Lei Dong2, Sadick Karma3, Alexandra N Huang4, Lili Ren, PhD5, Honghai Zhang, MD6, Hongjie Bi7, Xueer Wang, PhD8, Chaoling Wan9, Yanni Ma10, Nora Heisterkamp11, Steven T Rosen, MD12, Jianjun Chen, PhD, DPhil13, Seán O'Leary14, Eric Wang, PhD15, Jia Yu16, Rui Su, PhD17
AFFILIATIONS
1 City of Hope, Monrovia, United States
2 UT Sounthwestern Medical Research Center, Dallas, United States
3 The Jackson Laborotory for Genomic Medicine, Farmington, United States
4 University of California, Riverside, Riverside, United States
5 City of Hope, Monrovia, United States
6 City of Hope, Monrovia, United States
7 City of Hope, Monrovia, United States
8 City of Hope, Monrovia, United States
9 City of Hope, Monrovia, United States
10 Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China
11 City of Hope, Monrovia, United States
12 City of Hope, Monrovia, United States
13 City of Hope, Monrovia, United States
14 University of California, Riverside, Riverside, United States
15 The Jackson Laborotory for Genomic Medicine, Farmington, United States
16 Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China
17 City of Hope, Monrovia, United States
ABSTRACT
Liquid-liquid phase separation (LLPS) drives the assembly of diverse membranelle organelles, including stress granules (SGs) and processing bodies (PBs), which orchestrate the spatiotemporal regulation of gene expression. Cancer cells exploit these RNA granules to adapt to and survive chemotherapy-induced stress. However, whether and how the material properties of RNA granules are dynamically regulated in cancer, especially in acute myeloid leukemia (AML), remain largely unknown. Here, we uncover a previously unrecognized role of eukaryotic initiation factor 4A1 (eIF4A1), particularly its eIF4F-free pool, in regulating de novo serine biosynthesis and therapeutic drug resistance in AML through restraining the liquid-to-solid phase transition of RNA granules.
To assess molecular regulators governing RNA granule dynamics, we integrated pan-cancer proteomics with unbiased functional screen, identifying eIF4A1 as the predominant RNA helicase across cancer types and a potent suppressor of RNA granule assembly. eIF4A1 knockout (KO) significantly increased SG and PB formation and condensation while restricting granule motility. Live-cell spatiotemporal half-bleach assay and fluorescence recovery after photobleaching assay orthogonally showed that either genetic depletion or pharmacological suppression of eIF4A1 notably abolished intragranular redistribution and facilitated a liquid-to-solid transition of RNPs. Moreover, in vitro droplet formation assays independently confirmed that eIF4A1 partitioned into condensates and suppressed RNA-dependent phase separation. Together, these findings establish eIF4A1 as a key determinant of RNA granule fluidity.
We next investigated the biological function of eIF4A1 in AML. eIF4A1 KO profoundly inhibited leukemogenesis and prolonged overall survival in vivo. Consistently, eIF4A1 depletion significantly impaired AML cell growth, induced apoptosis, suppressed mitochondrial respiration, and reduced global translation efficacy in vitro. Strikingly, eIF4A1 KO sensitized AML cells to chemotherapy drugs. These results pinpoint eIF4A1 as a promising vulnerability of AML.
Mechanically, we uncovered a novel translation-independent function of eIF4A1. Although eIF4A1 is best known as a core component of eIF4F complex, we showed that >90% of eIF4A1 existed outside eIF4F. BioID-MS, followed by PLA and co-IP assays revealed a robust RNA-independent interaction between eIF4F-free eIF4A1 and mRNA stabilizer Y-box-binding protein 1 (YB-1). Metabolic profiling established a crucial role of eIF4A1 in amino acid biosynthesis, with de novo serine metabolism emerging as a notable downstream pathway. Integrative CLIP-seq, transcriptomics and proteomics identified thousands of direct eIF4A1/YB-1 target transcripts characterized by relatively shorter CDS/3’UTR regions and higher GC content.
Among these targets, PHGDH, encoding the rate-limiting enzyme of de novo serine biosynthesis, represents a prominent example. Depletion of either eIF4A1 or YB-1 markedly decreased PHGDH expression and accelerated PHGDH mRNA decay. CLIP-qPCR, with cell-free single-molecule RNA pulldown assays confirmed the direct binding of eIF4A1 and YB-1 to PHGDH. Furthermore, in vitro droplet-formation assays and single-molecule fluorescence in situ hybridization (smFISH) revealed that eIF4A1 prevented the sequestration of PHGDH into RNA granules. Collectively, these findings interrogate eIF4F-free eIF4A1 and YB-1 as core components of RNA chaperones that protect PHGDH from aberrant condensation and maintain its translation-ready state, thereby promoting chemoresistance.
Pharmacologically, Zotatifin, a clinical-stage eIF4A1 inhibitor, dramatically reduced the leukemogenesis and significantly prolonged survival of AML mouse models (immunodeficient xenograft model: 63 vs. 150 days, P = 0.0006; immunocompetent bone marrow transplantation model: 40 vs. 75 days, P = 0.001). Moreover, genome-wide CRISPR screens identified a strong synergistic activity of Zotatifin with YB-1 inhibitor SU056 in eradicating AML both in vitro and in vivo (P < 0.0001), primarily by disrupting amino acid biosynthesis.
Overall, our findings identify eIF4A1, especially its eIF4F-free pool, as a key regulator of AML pathogenesis through preserving RNA granule fluidity and governing metabolic plasticity. Targeting eIF4A1, particularly with Zotatifin, represents a promising therapeutic strategy for AML.
DISCLOSURES
Xiaoxu Zhang: No relevant conflicts of interest to declare.