AFFILIATIONS
1 Memorial Sloan Kettering Cancer Center, New York, New York, United States
2 Cambridge Stem Cell Institute, Cambridge, United Kingdom
3 Cambridge Stem Cell Institute, Cambridge, United Kingdom
4 Memorial Sloan Kettering Cancer Center, New York, New York, United States
5 Weill Cornell Medicine, New York, United States
6 Weill Cornell Medicine of Cornell University, New York, United States
7 Cambridge Stem Cell Institute, Cambridge, United Kingdom
8 Memorial Sloan Kettering Cancer Center, New York, United States
9 Memorial Sloan Kettering Cancer Center, New York, New York, United States
10 Cambridge Stem Cell Institute, Cambridge, United Kingdom
11 Memorial Sloan Kettering Cancer Center, New York, New York, United States
12 Memorial Sloan Kettering Cancer Center, New York, New York, United States
13 Memorial Sloan Kettering Cancer Center, New York, New York, United States
14 Cambridge Stem Cell Institute, Cambridge, United Kingdom
15 Cambridge Stem Cell Institute, Cambridge, United Kingdom
16 Cambridge Stem Cell Institute, Cambridge, United Kingdom
17 Cambridge Stem Cell Institute, Cambridge, United Kingdom
18 Cambridge Stem Cell Institute, Cambridge, United Kingdom
19 University of Cambridge, Cambridge, United Kingdom
20 Cambridge Stem Cell Institute, Cambridge, United Kingdom
21 Memorial Sloan Kettering Cancer Center, New York, New York, United States
22 Memorial Sloan Kettering Cancer Center, New York, New York, United States
23 Memorial Sloan Kettering Cancer Center, New York, New York, United States
24 Cambridge Stem Cell Institute, Cambridge, United Kingdom
25 Cambridge Stem Cell Institute, Cambridge, United Kingdom
26 Cambridge Stem Cell Institute, Cambridge, United Kingdom
27 Cambridge Stem Cell Institute, Cambridge, United Kingdom
28 Technische Universität München, Munich, Germany
29 Cambridge Stem Cell Institute, Cambridge, United Kingdom
30 Cambridge Stem Cell Institute, Cambridge, United Kingdom
31 University of Cambridge, Cambridge, United Kingdom
32 Technische Universität München, Munich, Germany
33 Storm Therapeutics Ltd, Cambridge, United Kingdom
34 Cambridge Stem Cell Institute, Cambridge, United Kingdom
35 Nanjing University, Nanjing, China
36 Transition Bio, Inc, Cambridge, United States
37 Transition Bio, Inc, Cambridge, United States
38 Transition Bio, Inc, Cambridge, United States
39 Weill Cornell Medicine of Cornell University, New York, United States
40 Weill Cornell Medicine, New York, United States
41 Storm Therapeutics Ltd, Cambridge, United Kingdom
42 Transition Bio, Inc, Cambridge, United States
43 Cambridge Stem Cell Institute, Cambridge, United Kingdom
44 Memorial Sloan Kettering Cancer Center, New York, United States
ABSTRACT
N6-methyladenosine (m6A), deposited by the methyltransferase METTL3, sustains oncogenic programs in myeloid leukemia, and METTL3 inhibitors (METTL3i) are advancing through early-phase clinical development. However, mechanisms of adaptive resistance remain unknown. Genome-wide CRISPR/Cas9 sensitization and resistance screens integrated with published polysome-profiling data from catalytically inactive METTL3 models (Wei et al., Nat Cell Biol 2022) identified the nuclear m6A reader YTHDC1 as a key determinant of METTL3i response. Dual-guide CRISPR/Cas9 targeting of METTL3 and YTHDC1 synergistically impaired fitness in MOLM-13 and MV4-11 AML cells, while YTHDC1 depletion combined with low-dose METTL3i significantly prolonged survival in MV4-11 xenografts.
Using endogenous GFP-YTHDC1 knock-in OCI-AML3 cells, METTL3i treatment reversibly increased YTHDC1 protein, and YTHDC1-high cells acquired resistance to METTL3i re-challenge. YTHDC1 overexpression similarly conferred resistance in THP-1, K562, and KCL-22 cells. METTL3 inhibition increased adaptive nuclear YTHDC1 condensate (nYAC) number, size, and intensity while reducing liquid-liquid phase separation (LLPS) dynamics in primary AML samples and five myeloid leukemia cell lines. In contrast, normal cord blood CD34+ cells retained dynamic LLPS, suggesting a leukemia-specific adaptive condensate response. An image-based screen identified Polyadenylation-Binding Nuclear Protein 1 (PABPN1), a YTHDC1 interactor and component of the polyA exosome targeting (PAXT) complex, as a negative regulator of nYACs. PABPN1 overexpression restored LLPS dynamics of nYACs and METTL3i sensitivity in YTHDC1-overexpressing THP1 cells (IC50 1.94 μM vs. 41.59 μM), whereas PABPN1 depletion generated larger, less dynamic nYACs and promoted resistance (p<0.001). Likewise, YTHDC1 mutants defective in m6A recognition or condensate formation (W377A, W428A, ΔYTH, and ΔR+P) failed to confer METTL3i resistance, linking nYAC abundance and biophysical state to drug response.
To define how nYAC remodeling alters RNA regulation, we integrated direct RNA nanopore m6A mapping with YTHDC1 iCLIP in YTHDC1-overexpressing THP-1 cells following METTL3 inhibition. Despite ~80% global m6A loss, residual m6A-modified transcripts persisted. YTHDC1 binding was redistributed (74% of sites lost, 22% retained, and 4% gained), with retained sites exhibiting increased YTHDC1 enrichment versus vehicle (p<2.2e-16) and higher residual m6A stoichiometry than lost or unbound sites (p=2.99e-52). NanoSplit assays further demonstrated increased YTHDC1 binding to a DRACH-containing reporter following METTL3 inhibition, supporting a model in which remodeled nYACs selectively concentrate YTHDC1 on residual m6A-marked transcripts despite global m6A loss. Bulk RNA-seq and direct RNA nanopore profiling showed that YTHDC1 overexpression partially restored METTL3i-altered oncogenic programs, including MYC-associated signatures. Quantitative proteomics identified 183 METTL3i-regulated proteins rescued by YTHDC1, including 103 direct YTHDC1 targets enriched for MYC activation and suppression of METTL3-loss signatures. Increased YTHDC1 binding to MYC and BCL-2 transcripts restored their protein expression suppressed by METTL3 inhibition, while ectopic MYC expression phenocopied YTHDC1-mediated resistance in THP-1 and OCI-AML3 cells.
Guided by these findings, we tested a newly developed YTHDC1 inhibitor and found that it synergized with METTL3i across 10 leukemia cell lines and suppressed primary AML proliferation and colony formation while sparing normal CD34+ progenitors. Moreover, because YTHDC1 restored BCL-2, METTL3i was combined with venetoclax, enhancing anti-leukemic activity in 3 myeloid leukemia cell lines, 2 AML PDX-derived samples in vitro, and significantly improving survival while reducing leukemic burden in a chemoresistant AML PDX model in vivo.
Collectively, these findings define an adaptive nYAC-mediated mechanism that preserves MYC/BCL-2 programs and drives resistance to METTL3 inhibition. More broadly, they demonstrate that epitranscriptomic therapies can rewire the biophysical state of RNA-protein condensates, reshaping RNA regulation to promote adaptive therapeutic resistance. These findings nominate YTHDC1 as a biomarker of METTL3i response and support YTHDC1 or BCL-2 co-targeting as rational strategies to enhance METTL3-directed therapy in myeloid leukemia.