AUTHORS
Wallace A. Bourgeois, MD, Reina Takeda, MD, PhD2, Samuel Ojeda3, George Wan4, Alba Sommerschield, MD5, Hannah E Rice6, Hong Yue, PhD7, Brandon D Regalado8, Kaitlyn H. Ko9, Jevon A Cutler, PhD10, Sarah Naomi Olsen, PhD11, Jonathan M Tsai, MD, PhD12, Jerry McGeehan, PhD13, Ross L. Levine, Prof. Dr.14, Sheng F. Cai, MD, PhD15, John G Doench, PhD16, Eric S Fischer, PhD17, Scott A Armstrong, MD, PhD18
AFFILIATIONS
1 Dana-Farber/Boston Children’s Cancer and Blood Disorders Center; Harvard Medical School, Boston, Massachusetts, United States
2 Dana-Farber/Boston Children’s Cancer and Blood Disorders Center; Harvard Medical School, Boston, Massachusetts, United States
3 Department of Biological Chemistry and Molecular Pharmacology, Harvard Medical School; Department of Cancer Biology, Dana-Farber Cancer Institute, Boston, Massachusetts, United States
4 Dana-Farber/Boston Children’s Cancer and Blood Disorders Center; Harvard Medical School, Boston, Massachusetts, United States
5 Dana-Farber/Boston Children’s Cancer and Blood Disorders Center; Harvard Medical School, Boston, Massachusetts, United States
6 Dana-Farber/Boston Children’s Cancer and Blood Disorders Center; Harvard Medical School, Boston, Massachusetts, United States
7 Department of Biological Chemistry and Molecular Pharmacology, Harvard Medical School; Department of Cancer Biology, Dana-Farber Cancer Institute, Boston, Massachusetts, United States
8 Dana-Farber/Boston Children’s Cancer and Blood Disorders Center; Harvard Medical School, Boston, Massachusetts, United States
9 Human Oncology and Pathogenesis Program, Memorial Sloan Kettering Cancer Center, New York, New York, United States
10 Departments of Pediatrics and Cell, Developmental and Cancer Biology, Knight Cancer Institute, Oregon Health and Science University (OHSU), Portland, Oregon, United States
11 Dana-Farber/Boston Children’s Cancer and Blood Disorders Center; Harvard Medical School, Boston, Massachusetts, United States
12 Department of Pathology, Brigham and Women’s Hospital; Harvard Medical School, Boston, Massachusetts, United States
13 Syndax Pharmaceuticals, New York, New York, United States
14 Human Oncology and Pathogenesis Program, Memorial Sloan Kettering Cancer Center, New York, United States
15 Leukemia Service, Division of Hematologic Malignancies, Memorial Sloan Kettering Cancer Center, New York, New York, United States
16 Genetic Perturbation Platform, Broad Institute of MIT and Harvard, Cambridge, Massachusetts, United States
17 Department of Biological Chemistry and Molecular Pharmacology, Harvard Medical School; Department of Cancer Biology, Dana-Farber Cancer Institute, Boston, Massachusetts, United States
18 Dana-Farber/Boston Children’s Cancer and Blood Disorders Center; Harvard Medical School, Boston, Massachusetts, United States
ABSTRACT
Menin inhibitors (MIs) have demonstrated clinical activity in KMT2A-rearranged (KMT2A-r) and NPM1-mutant acute leukemias, leading to regulatory approvals for revumenib and ziftomenib. However, relapses driven by MEN1 mutations have now arisen with several MIs. MEN1 mutations disrupt MI binding while preserving the menin-KMT2A interaction, thereby sustaining oncogenic transcription. Here, we identify a novel resistance mechanism in which mutations at other protein-protein interfaces within the menin-KMT2A-LEDGF complex confer cross-resistance to multiple MIs.
We previously showed that CRISPR base editor screening and patient-derived xenograft (PDX) models can predict clinically observed MEN1 resistance mutations. In PDX models treated continuously with revumenib or bleximenib, relapsing leukemias frequently acquired MEN1 mutations; however, these mutations were sometimes present at low variant allele frequencies (VAFs), and some relapses occurred without detectable MEN1 mutations. Whole-exome sequencing of these specimens identified eight KMT2A mutations affecting seven residues within a critical 28-amino-acid region, including two distinct substitutions at G123 (G123V and G123A). This region includes KMT2A integrase-binding motifs (IBM) 1 and 2, located at amino acids 123–135 and 147–152, respectively, which mediate interactions with menin and LEDGF outside the MI-binding pocket. We also identified LEDGF N421Y within the LEDGF integrase-binding domain, which interacts with both menin and KMT2A. Together, these mutations implicate alternative protein-protein interaction sites within the menin-KMT2A-LEDGF complex as mediators of MI resistance.
As an orthogonal approach to identify KMT2A regions that can drive resistance, we performed CRISPR base editor screens in MV4;11 (KMT2A-r) cells using an sgRNA library tiling exons encoding the first 312 amino acids of KMT2A. We screened four MIs that have entered clinical trials: bleximenib, ziftomenib, revumenib, and enzomenib. After 24 days of treatment, we calculated log2 fold changes in sgRNA abundance for each MI relative to DMSO. Enriched sgRNAs (log2 fold change ≥3.5) clustered almost exclusively within KMT2A regions that interact with menin and/or LEDGF, including menin-binding motifs (MBM1 and 2) and IBM1 and 2. To validate the screens, we individually transduced enriched sgRNAs. 16 of 17 supported growth at 25 nM of all four MIs, a concentration that suppressed WT cells. Five sgRNAs yielded substitutions at three residues also mutated in PDX models, including G123. Across the validated sgRNAs, different MIs generally selected the same or highly similar substitutions from a given sgRNA. Six drug-selected substitutions were tested in viability assays, and each conferred 5- to 14-fold absolute IC50 shifts of similar magnitude across all four MIs. The convergence of structurally distinct MIs on the same substitutions suggests a mechanism other than impaired drug binding.
To elucidate mechanisms of how these KMT2A and LEDGF mutations drive resistance, we used CRISPR homology-directed repair to introduce mutations found in resistance models (KMT2A G123V, V132L, G137D; LEDGF N421Y). As homozygous mutations, they shifted the IC50 of MIs in viability assays 5-to-20-fold. For each mutation, the degree of shift was similar for bleximenib, ziftomenib, and revumenib. RNA-sequencing in vitro and in vivo revealed attenuation of MI-induced transcriptional changes. ChIP qPCR and sequencing revealed retention of menin, KMT2A, and LEDGF on chromatin at critical loci, such as MEIS1. Co-immunoprecipitation using FLAG-tagged mutant or WT KMT2A as bait demonstrated persistent association of menin and LEDGF with mutant, but not WT, KMT2A in the presence of MI. In the presence of these mutations, the chromatin complex acquires a critical dependence on LEDGF, with genetic inactivation or LEDGF protein degradation re-sensitizing to MI treatment. Together, these data support a model in which mutations outside the MI-binding pocket retain the menin-KMT2A-LEDGF complex on chromatin despite MI treatment, thus sustaining oncogenic transcription.
In conclusion, we identify mutations at alternative interfaces within the menin-KMT2A-LEDGF complex as a novel mechanism of MI resistance. Their recurrent emergence in PDX models and independent selection in base editor screens, an approach that previously predicted clinically observed MEN1 mutations, suggest these KMT2A and LEDGF mutations may arise in patients.
DISCLOSURES
Wallace A. Bourgeois, MD: No relevant conflicts of interest to declare.