AUTHORS
Christoph Schaniel, PhD, Allison Cote, PhD2, Min Lu, PhD3, Lijuan Xi4, Joseph Tripodi, PhD5, Emmanuel Olivier, PhD6, Vesna Najfeld, PhD7, Bridget Kelly Marcellino, MD, PhD8, Eirini Papapetrou, MD, PhD9, Jeff Vierstra, PhD10, Ronald Hoffman, MD11
AFFILIATIONS
1 Icahn School of Medicine at Mount Sinai, New York, United States
2 Altius Institute for Biomedical Sciences, Seattle, Washington, United States
3 Icahn School of Medicine at Mount Sinai, New York, New York, United States
4 Icahn School of Medicine at Mount Sinai, New York, New York, United States
5 Icahn School of Medicine at Mount Sinai, New York, New York, United States
6 Icahn School of Medicine at Mount Sinai, New York, New York, United States
7 Icahn School of Medicine at Mount Sinai, New York, New York, United States
8 Icahn School of Medicine at Mount Sinai, New York, New York, United States
9 Icahn School of Medicine at Mount Sinai, New York, New York, United States
10 Altius Institute for Biomedical Sciences, Seattle, Washington, United States
11 Icahn School of Medicine at Mount Sinai, New York, New York, United States
ABSTRACT
Background: About 10-20% of myelofibrosis (MF) patients progress to a secondary form of acute myeloid leukemia (MPN-AP/BP), where survival is frequently <6 months. The mechanisms driving this evolution remain poorly defined. While the chromatin regulators HMGA1 and HMGA2 are strongly expressed during embryonic development and largely repressed in normal adult tissues, including hematopoietic stem and progenitor cells (HSPCs), they are reactivated and overexpressed in various malignancies (Mansoori et al., Genes, 2021). We recently identified structural HMGA2 abnormalities, including novel 3’ UTR microdeletions (HMGA2Δ3’) leading to HMGA2 overexpression (OE) in MF and MPN-AP/BP, with a higher prevalence in CALR-mutant patients, which are absent in PV, ET and MDS/MPN overlap patients (Handa et al., Leukemia, 2025). Mechanistically, these lesions disrupt LET7 microRNA binding sites, driving HMGA2OE beyond the levels seen in HMGA2 wildtype (WT) that temporally coincides with transformation to MF and MPN-AP/BP, implicating HMGA2 dysregulation in disease progression. Supporting these findings, a recent report linked HMGA2 abnormalities to elevated blasts in MF (Díaz-González, Blood Adv., 2026). Furthermore, HMGA2OE has been connected to loss of function mutations in EZH2, known to be associated with high risk in MF (Yang et al., Blood, 2016; Shimizu et al., J. Exp. Med., 2016; Ueda et al., Blood Adv., 2017).
Methods: To dissect the consequences of HMGA2OE due to HMGA2Δ3’, we generated clonal induced pluripotent stem cell (iPSC) models: a MF patient-derived CALRdel52+HMGA2Δ3’ line and an isogenic pair of post-ET-MF CALRdel34+HMGA2WT and CRISPR/Cas9-engineered CALRdel34+HMGA2Δ3’. HSPCs and megakaryocytes (MKs), differentiated from each line, were assessed by hematopoietic-colony formation and replating potential, immunophenotyping and RNA sequencing (RNAseq), along with single cell (sc) RNAseq validation in 3 HMGA2Δ3’ and 3 HMGA2WT primary MF patient HSPCs and MKs.
Results: HMGA2Δ3’ iPSC-HSPCs generated significantly more hematopoietic colonies with markedly higher replating potential than HMGA2WT HSPCs (p<0.03), and increased numbers of CD34+CD41a+ MK-primed HSPCs (p<0.009) and CD41a+CD42b+ MKs (p<0.003). RNAseq revealed that HMGA2Δ3’ iPSC-HSPCs transcriptionally cluster separately from HMGA2WT iPSC-HSPCs, with HMGA2 but not HMGA1 expression being selectively increased (p<0.0001). HMGA2Δ3’ signatures showed upregulation of the transcriptional activator SMYD3 which amplifies several cancer promoting genes including HMGA2 (Yang et al. Clin. Epigenetics, 2023) and cell-cycle drivers (CCNA2, CCNB1/2, CCND1, CCNE1/2, CDK4, E2F1/2, CUL3, KPNB1) as well as repression of the negative regulator CDKN2A. In parallel, P53 target genes (CDKN1A, BBC3, PMAIP, GADD45A/B/G) and the modulator PADI4 were repressed, while negative regulators (CCDC8, HERC6, NEDD4, NEDD4L, TRIM24, UBE4B) were upregulated. These patterns were mirrored in scRNAseq of primary MF patient HSPCs and MKs. Interestingly, SSB/LA antigen, which enhances translation of the P53-ubiquitin ligase MDM2 (Nakatake et al., Oncogene, 2012), and LIN28B, a LET7 antagonist which also blocks P53 translation and upregulates HMGA2 (Shi et al., Oncogenesis, 2022), were also elevated in HMGA2Δ3’ versus HMGA2WT iPSC-HSPCs. Consistent with the role of HMGA2 in facilitating MDM2-mediated P53 ubiquitination and degradation (Wang et al., J. Pathol., 2018), HMGA2Δ3’ iPSC-HSPCs showed reduced P53 protein levels compared to HMGA2WT. Depletion of HGMA2 but not HMGA1 reduced colony formation by JAK2V617F+ HEL and CALR-mutant MARIMO cells (p<0.004). Ongoing xenograft studies with our iPSC lines will further determine the effect of HMGA2OE on MF development and progression.
Conclusion: HMGA2OE enhances HSPC self-renewal, MK-lineage priming and cell output by orchestrating a coordinated activation of cell-cycle programs and suppression of P53 activity, mechanistically linking a defined molecular event the clinical progression of MF. Convergent findings across engineered and primary patient cells strengthen the causal link between HMGA2Δ3’/HMGA2OE and disease progression.
Implications: These findings indicate that HMGA2OE plays an important role in MF disease progression and indicate that components of the HMGA2–MDM2-P53 axis and other HMGA2-regulated networks can act as druggable targets in MF and MPN-AP/BP, offering a rational strategy to develop novel therapies to effectively treat MF patients.
DISCLOSURES
Christoph Schaniel, PhD: No relevant conflicts of interest to declare.