HMGA1 chromatin regulators drive transcriptional networks governing cell cycle progression, immune escape, and menin-inhibitor resistance in KMT2A-r Acute Myeloid Leukemia Free

Introduction: KMT2A-rearranged acute myeloid leukemias (KMT2A-r AML) are highly lethal leukemias caused by rearrangements of the KMT2Agene, which encode abnormal fusion proteins that activate pro-leukemogenic genes such as HOXAgenes andMEIS1. Prior studies focused on identifying drugs to target proteins that form complexes with KMT2A fusions led to the recent FDA approval of the first menin inhibitor for refractory KMT2A-r AML. However, resistance remains a significant clinical challenge. The gene encoding the High Mobility Group A1 (HMGA1) chromatin regulator is up-regulated in refractory hematologic malignancies and solid tumors where high levels portend adverse outcomes (Resar et al., Cancer Res 2018; Li et al., Blood 2022; Chia et al., JCI2023; Luo et al., JCI2025). In preclinical models, HMGA1 drives tumor progression and a highly plastic, stem-like state by modulating chromatin structure to activate stem cell gene networks in diverse tumors (Li et al., Blood 2022; Chia et al., JCI2023; Luo et al., JCI2025). HMGA1is also enriched in the leukemic stem cell signature of KMT2A-r mouse models (Somervaille et al., Cell Stem Cell 2009), and a recent study identified HMGA1 among the gene signature of menin inhibitor resistance in KMT2A-r AML patient-derived xenografts (Soto-Feliciano et al., Cancer Discov2023). We therefore hypothesized that: 1) HMGA1 drives refractory disease and menin inhibitor resistance in KMT2A-r AML by modulating transcriptional networks, 2) Elucidating HMGA1 function in KMT2A-r AML will reveal novel therapeutic vulnerabilities.

Methods: To define HMGA1 function in KMT2A-r AML, we silenced HMGA1 using CRISPR or short hairpin RNA (shRNA) in human KMT2A-r AML cell lines (MOLM-14, THP-1, MV4;11). Leukemogenic properties were tested in vitroand in mouse implantation models. To elucidate mechanisms underlying HMGA1, we integrated multiomic studies (RNA/ATACseq). Artificial intelligence (AI) was used to predict drugs to target HMGA1 gene networks.

Results: HMGA1 is overexpressed in primary KMT2A-r AML caused by diverse fusions in two large patient cohorts compared to age-matched, healthy bone marrow by RNAseq (P<0.0001). Moreover, HMGA1 transcripts correlate positively with downstream targets of KMT2A-r fusions, including HOXA9,HOXA10, andMEIS1 (r>0.4; P<0.0001). HMGA1 gene silencing in KMT2A-r AML cell lines disrupts proliferation (P<0.0001) and clonogenicity (P<0.0001) while inducing differentiation markers (CD11b, CD14; P<0.01). Strikingly, HMGA1deficiency also decreases leukemic engraftment (P<0.0001) and prolongs survival (P<0.0001) following implantation in immunodeficient mice. Intriguingly, AML cells that engraft after HMGA1 silencing express higher HMGA1, suggesting that the escape from gene silencing and a specific HMGA1 level are required for leukemogenesis in this model. RNAseq in KMT2A-r AML cell lines (MOLM-14, THP-1) followed by GSEA revealed that HMGA1 activates transcriptional networks that govern: 1) HOX gene expression, 2) DNA repair, and 3) cell cycle progression (E2F, G2M, Mitotic Spindle). Moreover, HMGA1 activates the signature of menin inhibitor-resistance genes. Conversely, HMGA1 represses MHC-II genes involved in antigen presentation and an anti-tumor immune attack. ATACseq revealed that HMGA1 reduces chromatin accessibility within the MHC-II locus. Leveraging drug prediction algorithms, we identified candidate compounds to target HMGA1 networks, which include the histone deacetylase inhibitor (HDACi) entinostat (ENT), that promotes chromatin opening. Treatment with ENT represses HMGA1 and cell-cycle genes while inducing MHC-II gene expression. Preliminary results suggest that ENT enhances sensitivity to menin inhibitors, supporting further investigation into potential combination strategies.

Conclusions: HMGA1 is required for salient leukemogenic properties in KMT2A-r AML preclinical models. Mechanistically, HMGA1 activates genes governing cell cycle progression, particularly those implicated in menin inhibitor resistance, while repressing anti-tumor, immune attack genes. Further, these networks can be targeted with clinically available drugs. Together, these findings implicate HMGA1 as a novel epigenetic regulator and promising therapeutic target in KMT2A-r AML.