murin e jak2 v617f  (Addgene inc)

Journal: bioRxiv

Article Title: Targeting HMGA1-driven leukemic transformation in myeloproliferative neoplasms with pacritinib

doi: 10.1101/2025.06.01.657170

Figure Lengend Snippet: (a) Box plots showing HMGA1 expression in bone marrow mononuclear cells (BMMCs) from healthy individuals (n = 10) and sAML patients (n = 3; malignant cells and microenvironment components depicted separately), based on CITE-seq data (GSE185381). Statistical analysis by Kruskal-Wallis test with Benjamini-Hochberg (BH) correction. (b) Left: Correlation between mean HMGA1 transcript expression and percentage of HMGA1-positive cells from healthy controls and sAML patients (GSE185381). Right: Bar plot comparing mean HMGA1 expression in total BMMCs between healthy controls (n = 10) and sAML patients (n = 3). Pearson correlation coefficient ( R ) and P -value shown. Two-sample t -test with BH correction. (c) Log2-transformed HMGA1 expression in CD34 + peripheral blood mononuclear cells (PBMCs) from patients with chronic-phase MPN (ET, n = 2; PV, n = 1; MF, n = 5), derived from dataset GSE189979. Wald test with BH correction. (d) Relative HMGA1 mRNA expression, determined by qRT-PCR, in (i) PBMCs and (ii) CD34 + cells from MPN patients (stratified by JAK2 mutation status: JAK2 wile-type PBMCs, n = 21; JAK2 V617F PBMCs, n = 23; JAK2 wile-type CD34 + , n = 20; JAK2 V617F CD34 + , n = 16) and sAML patients (stratified by JAK2 mutation status: JAK2 wile-type sAML PBMCs, n = 4; JAK2 V617F PBMCs, n = 7; JAK2 wile-type CD34 + , n = 4; JAK2 V617F CD34 + , n = 5). Data are presented as mean ± SD. One-way ANOVA. (e) Relative HMGA1 mRNA expression, determined by qRT-PCR, in (i) PBMCs from sAML patients (stratified by TP53 mutation status: TP53 wild-type, n = 6; TP53 mutated, n = 5) and (ii) CD34 + cells from sAML patients ( TP53 wild-type, n = 5; TP53 mutated, n = 4). Data are presented as mean ± SD. Two-sample t -test. (f) Expression trajectory plot depicting protein expression of HMGA1 and key stem/progenitor markers (CD34, PROM1 /CD133, KIT /CD117, FUT4 /CD15, CD82) across diverse hematopoietic cell subsets, ordered by developmental hierarchy, within (i) total BMMCs and (ii) the malignant sAML cell compartment (GSE185381). (g) Violin plots comparing (i) transcript levels (CITE-seq) and (ii) surface protein expression (ADT) of CD34, PROM1 (CD133), KIT (CD117), FUT4 (CD15), and CD82 between malignant and non-malignant BMMCs from healthy controls (n = 10) and sAML patients (n = 3) (GSE185381). Wald test with BH correction.

Article Snippet: Lentiviral vectors for overexpression of human HMGA1 (NM_145899.3), murine Hmga1 (NM_016660.3), murin e Jak2 WT (NM_008413.4), or murin e Jak2 V617F were generated by cloning the respective open reading frames (ORFs) into the pLenti-CMV-Puro vector (Addgene plasmid #39481, a gift from Ie-Ming Shih).

Techniques: Expressing, Transformation Assay, Derivative Assay, Quantitative RT-PCR, Mutagenesis

Journal: bioRxiv

Article Title: Targeting HMGA1-driven leukemic transformation in myeloproliferative neoplasms with pacritinib

doi: 10.1101/2025.06.01.657170

Figure Lengend Snippet: (a) UMAP visualization of 9,191 Lin - CD34 + cells from 14 sAML patient samples (GSE226340), showing pseudotime trajectories. Cells are colored by (left to right): differentiation trajectory score (LSC-like vs. erythroid), HMGA1 expression level, and TP53 mutational status (wild type, single-hit, multi-hit). Arrows indicate inferred differentiation direction. (b) Violin plots showing HMGA1 transcript expression in CD34 + HSPCs from sAML patients (GSE226340), stratified by TP53 mutational status (wild type, single-hit, multi-hit). (c) Violin plots showing HMGA1 transcript expression in CD34 + HSPCs from sAML patients (GSE226340), stratified by specific combinations of JAK2 (wild-type or V617F) and TP53 (wild-type or mutated) genotype. (d) HMGA1 expression in CD34 + HSPCs from sAML patients (GSE226340) stratified by combinations of TP53 genotype (wild-type or mutated) with other co-occurring gene mutations ( CALR , TET2 , U2AF1 ) versus TP53 mutation alone. (e) HMGA1 expression in BMMCs from sAML patients (GSE185381) stratified by TP53 mutational status (wild type, n = 2 vs. mutated, n = 1). (f) HMGA1 expression in BMMC from patients with sAML (GSE185381) stratified by karyotype complexity (non-complex karyotype, n = 1 vs. complex karyotype, n = 2) (g) Hmga1 transcript expression in murine hematopoietic progenitors from various Jak2 / Trp53 mutant mouse models (GSE180851). (i) Megakaryocyte-erythorid progenitors (MEP) from JAK2 V617F ( J VF ) mice versus J VF mice with different Trp53 genotypes ( Trp53 -/- , Trp53 +/- , Trp53 R172H/- or Trp53 R172H/+ ) (n = 3 per group). (ii) Lin-Sca-1 + c-kit + (LSK) stem cells from the same genotypes. (h) HMGA1 expression in MEPs from mice at sAML stage ( Jak2 V617F Trp53 -/- , n = 3; Jak2 V617F Trp53 R172H/- , n = 3) (GSE180851). (i) Cellular composition and marker expression in representative sAML cases from GSE185381. (i) Stacking bar charts showing the percentage of annotated cell types in two sAML cases (AML3762, AML3948). (ii) Violin plots depicting HMGA1 , CD34 , and KIT transcript expression in the top 5 most abundant malignant cell types and microenvironment (non-malignant) cells for each patient. For panels b-h, statistical significance was assessed by Kruskal-Wallis test with Benjamini-Hochberg (BH) correction (b-d), Wilcoxon rank-sum test with BH correction (e,f), likelihood ratio test (g), or Wald test (h) as appropriate.

Article Snippet: Lentiviral vectors for overexpression of human HMGA1 (NM_145899.3), murine Hmga1 (NM_016660.3), murin e Jak2 WT (NM_008413.4), or murin e Jak2 V617F were generated by cloning the respective open reading frames (ORFs) into the pLenti-CMV-Puro vector (Addgene plasmid #39481, a gift from Ie-Ming Shih).

Techniques: Expressing, Mutagenesis, Marker

Journal: bioRxiv

Article Title: Targeting HMGA1-driven leukemic transformation in myeloproliferative neoplasms with pacritinib

doi: 10.1101/2025.06.01.657170

Figure Lengend Snippet: (a) HMGA1 protein expression (percentage of IHC-positive cells) in bone marrow biopsies from a patient cohort, stratified by sex (female, n = 140; male, n = 67) and age (<60 years, n = 56; ≥60 years, n = 75). Statistical analysis by two-sample t-test. (b-c) Association of HMGA1 protein levels (IHC) with established clinical risk scores in (b) essential thrombocythemia (ET, n = 53) using ELN, MIPSS-ET, IPSET, and Revised-IPSET scoring systems, and (c) polycythemia vera (PV, n = 33) using ELN, MIPSS-PV, and IWG-PV scoring systems. Two-sample t -test or one-way ANOVA, as appropriate. (d-e) HMGA1 protein levels (IHC) in (d) primary myelofibrosis (PMF, n = 20) stratified by IPSS, DIPSS, and MIPSS-70 clinical risk scores, and (e) secondary myelofibrosis (sMF, n = 12) stratified by MYSEC-PM score. One-way ANOVA. (f) Differential HMGA1 protein expression (IHC) between MPN ( JAK2 wild-type, n = 53; JAK2 V617F , n = 88) and sAML ( JAK2 wild-type, n = 9; JAK2 V617F , n = 12). Representative IHC images for JAK2 wild-type MPN and sAML are inset. One-way ANOVA. (g) HMGA1 protein expression (IHC) in sAML patient samples stratified by TP53 mutation status ( TP53 wild-type, n = 14; TP53 mutated, n = 7). A representative IHC image for TP53 wild-type sAML is inset. Two-sample t -test. (h-j) Elevated Hmga1 expression in a Jak2 V617F knock-in MPN mouse model. (h) Representative images of Hematoxylin and Eosin (H&E) staining and IHC for Hmga1 in spleen and bone marrow tissues from Jak2 wild-type (control) and Jak2 V617F mice. Scale bars, 50 µm. (i) Quantification of Hmga1 IHC-positive cells (%) in spleen (n = 6 per group) and bone marrow (n = 4 per group). (j) Relative Hmga1 mRNA levels in PBMCs from Jak2 wild-type (n = 3) and Jak2 V617F (n = 3) mice, determined by qRT-PCR. Data are presented as mean ± SD. Two-sample t -test. (k-l) Expression profiling of HMGA1, CD34, and CD117 in HEL and UKE-1 MPN cell lines. (k) Representative images of Wright-Giemsa staining (left panels) and immunocytochemical (ICC) staining for HMGA1, CD34, and CD117 (right panels). Scale bars, 80 µm (overview), 20 µm (inset). (l) Flow cytometric analysis of intracellular HMGA1 and surface CD34/CD117 expression. (m) Multiplex immunofluorescence (mIF) demonstrating HMGA1 co-expression with stem/progenitor markers in bone marrow biopsies from MPN patients progressing to sMF (representative PV-sMF and ET-sMF cases). Images show H&E staining alongside fluorescence channels for DAPI (nuclei, blue), Vimentin (cyan), HMGA1 (yellow), CD34 (green), and CD117 (red). Scale bars, 80 µm (overview), 20 µm (inset). (n) Flow cytometric analysis revealing HMGA1, CD34, and CD117 expression within CD45 + blasts from an sAML patient’s peripheral blood (PB). (o) Confocal microscopy demonstrating co-localization of HMGA1 (green) with CD34 (red) and CD117 (yellow) in bone marrow (BM) cells from an sAML patient. Nuclei are counterstained with DAPI (blue). Scale bar, 10 µm.

Article Snippet: Lentiviral vectors for overexpression of human HMGA1 (NM_145899.3), murine Hmga1 (NM_016660.3), murin e Jak2 WT (NM_008413.4), or murin e Jak2 V617F were generated by cloning the respective open reading frames (ORFs) into the pLenti-CMV-Puro vector (Addgene plasmid #39481, a gift from Ie-Ming Shih).

Techniques: Expressing, Mutagenesis, Knock-In, Staining, Control, Quantitative RT-PCR, Multiplex Assay, Immunofluorescence, Fluorescence, Confocal Microscopy

Journal: bioRxiv

Article Title: Targeting HMGA1-driven leukemic transformation in myeloproliferative neoplasms with pacritinib

doi: 10.1101/2025.06.01.657170

Figure Lengend Snippet: (a) Efficient shRNA-mediated knockdown of HMGA1 (sh1, sh2 vs. sh-NC control) in HEL and UKE-1 cells. Left: Relative HMGA1 mRNA levels by qRT-PCR (mean ± SD, n = 3). Right: Western blot analysis of HMGA1 protein; ACTB served as loading control. (b) Lentiviral-mediated overexpression of HMGA1 (OE vs. CMV-NC control) in HEL and UKE-1 cells. Left: Relative HMGA1 mRNA levels by qRT-PCR (mean ± SD, n = 3). Right: Western blot analysis of HMGA1 protein; Tubulin served as loading control. (c) Lentiviral-mediated overexpression of Hmga1 (J/OE vs. J/NC control) in murine Ba/F3 ( Jak2 wild type, or Jak2 V617F ) and 32D-cl3 ( Jak2 wild type, or Jak2 V617F ) cells. Left: Relative Hmga1 mRNA levels by qRT-PCR (mean ± SD, n = 3). Right: Western blot analysis of Hmga1 protein; Tubulin served as loading control. Statistical analyses for (a-c) by two-sample t-test or one-way ANOVA, as appropriate. (d) HMGA1 overexpression exacerbates disease phenotype in a HEL xenograft model. Hematological parameters (WBC, white blood cell count; HGB, hemoglobin; HCT, hematocrit; PLT, platelet count) in NSG mice engrafted with HEL cells stably expressing control vector (CMV-NC, n = 6) or HMGA1 (OE, n = 6) at 35 days post-transplantation. Data are presented as mean ± SD. Two-sample t -test. (e) HMGA1 knockdown alters chromatin accessibility and HMGA1 binding at key cell cycle regulatory gene loci. Integrative Genomics Viewer (IGV) snapshots displaying ATAC-seq and HMGA1 CUT&Tag signals at representative E2F target genes ( E2F1 , CCNE1 , CCNE2 , CDK2 , RB1 ), G2M checkpoint genes ( CCNB1 , CCNB2 , CDC2 , WEE1 , CDC25C , PLK1 , AURKA , AURKB ), and common cell cycle regulators ( CCNA2 , CDKN1A / p21 , CDKN1B / p27 ) in HEL cells following control (NC) versus HMGA1 knockdown (KD). (f) Enhanced E2F target and G2M checkpoint gene signatures in sAML patient cells. UMAP projections of single-cell CITE-seq data (GSE185381) from control and sAML patients, with cells colored by enrichment scores for E2F target and G2M checkpoint gene sets. Corresponding density plots illustrate score distributions.

Article Snippet: Lentiviral vectors for overexpression of human HMGA1 (NM_145899.3), murine Hmga1 (NM_016660.3), murin e Jak2 WT (NM_008413.4), or murin e Jak2 V617F were generated by cloning the respective open reading frames (ORFs) into the pLenti-CMV-Puro vector (Addgene plasmid #39481, a gift from Ie-Ming Shih).

Techniques: shRNA, Knockdown, Control, Quantitative RT-PCR, Western Blot, Over Expression, Cell Counting, Stable Transfection, Expressing, Plasmid Preparation, Transplantation Assay, Binding Assay

Journal: bioRxiv

Article Title: Targeting HMGA1-driven leukemic transformation in myeloproliferative neoplasms with pacritinib

doi: 10.1101/2025.06.01.657170

Figure Lengend Snippet: (a) Colony-forming units (CFU) assays in human (HEL, UKE-1) and murine (Ba/F3- JAK2 V617F , 32D- JAK2 V617F ) cell lines following HMGA1 knockdown (shHMGA1) or overexpression (OE-HMGA1) relative to respective controls (NC). Data are mean ± SD. (n=5 per group). Representative colony images are shown (scale bar 300 μm).Two-sample t -test and one-way ANOVA with Tukey’s post-hoc test. (b) Schematic of the HEL cell xenograft experiment in NSG mice, comparing HMGA1-OE cells to vector control cells. (c) Representative bioluminescence images (left panel) and quantification of total photon flux (right panel) from NSG mice at day 35 post-engraftment with luciferase-tagged HEL cells transduced with vector control (CMV-NC, n=6) or HMGA1 overexpression constructs (OE, n=6). Data are mean ± SD. Two-sample t -test. (d) Correlation between HMGA1 mRNA expression (snRNA-seq) and the number of open chromatin peaks (scATAC-seq) in CD34 + cells from a patient during primary MF and after sAML transformation (GSE221946). Pearson correlation coefficient ( R ) and P -value from Wilcoxon ranksum test with BH correction. (e) Genomic distribution of HMGA1 binding sites lost upon HMGA1 knockdown in HEL cells, as determined by CUT&Tag (i), and chromatin regions with significantly reduced accessibility upon HMGA1 knockdown, as determined by ATAC-seq (ii). Stacked bar chats show the percentage of peaks located in promoter (±3kb of TSS), intronic, intergenic, and other genomic regions. (f) GSEA enrichment plots for E2F targets and G2M checkpoint Hallmark gene sets. Comparisons shown in figure. Normalized Enrichment Score (NES) and False Discovery Rate (FDR) q-value are indicated. (g) Cell cycle distribution (Propidium Iodide staining followed by flow cytometry) of HEL and UKE-1 cells with HMGA1 knockdown (sh1, sh2) or non-targeting control (NC). Data are mean ± SD. (n = 5 per group). One-way ANOVA with Tukey’s post-hoc test. (h) Western blots analysis of key E2F pathway, G2M checkpoint, and common cell cycle regulatory protein levels in HEL and UKE-1 cells following overexpression (OE vs. CMV-NC vector control; left panels) or HMGA1 knockdown (shHMGA1 vs. shNC control; right panels). GAPDH served as loading controls. Blots are representative of at least three independent experiments.

Article Snippet: Lentiviral vectors for overexpression of human HMGA1 (NM_145899.3), murine Hmga1 (NM_016660.3), murin e Jak2 WT (NM_008413.4), or murin e Jak2 V617F were generated by cloning the respective open reading frames (ORFs) into the pLenti-CMV-Puro vector (Addgene plasmid #39481, a gift from Ie-Ming Shih).

Techniques: Knockdown, Over Expression, Plasmid Preparation, Control, Luciferase, Transduction, Construct, Expressing, Transformation Assay, Binding Assay, Staining, Flow Cytometry, Western Blot

Journal: bioRxiv

Article Title: Targeting HMGA1-driven leukemic transformation in myeloproliferative neoplasms with pacritinib

doi: 10.1101/2025.06.01.657170

Figure Lengend Snippet: (a) Relative proliferation curves of human (HEL, UKE-1) and murine (Ba/F3, 32D-cl3 transduced with Jak2 wild-type or Jak2 V617F ) cell lines following HMGA1/Hmga1 overexpression (OE) or shRNA-meidated knockdown (sh1, sh2) compared to respective controls (CMV-NC or sh-NC)NC.) 32D-cl3 cells were cultured with IL-3. Data are mean ± SD. (n = 5 per group). Two-way ANOVA. (b) Flow cytometric analysis of CD11b expression on 32D-cl3 cells transduced with Jak2 wild-type (J WT ) or Jak2 V617F (J VF ), and co-transduced with control vector (NC) or HMGA1 overexpression (OE), following G-CSF (100 ng/mL) induced differentiation. (i) Representative histograms of CD11b-FITC fluorescence. (ii) Quantification of HMGA1-PE mean fluorescence intensity (MFI). (iii) Quantification of CD11b-FITC MFI (n = 5 per group). Data are mean ± SD. Two-sample t -test. (c) Quantification of human CD45 + CD117 + HEL cells in peripheral blood of NSG mice at day 35 post-transplant, comparing HMGA1-OE versus vector control (CMV-NC) groups (n=6 per group). Data are mean ± SD. Two-sample t -test. (d) Wright-Giemsa stained peripheral blood smears from NSG mice engrafted with HMGA1-OE or CMV-NC HEL cells at day 35. Quantification of HEL cells (% of total nucleated cells) is shown (n = 6 per group). Data are mean ± SD. Two-sample t -test. (e-f) Representative H&E and HMGA1 IHC staining (left panels of e and f, respectively) and quantification of HMGA1-positive cells (%) (right panels fo e and f, respectively) in (e) femur bone marrow and (f) spleen sections from NSG mice engrafted with HMGA1-OE or CMV-NC HEL cells. Scale bars: 50 µm. Data are mean ± SD. Two-sample t -test. (g) Representative images of spleens (left) and relative spleen weights (spleen weight/body weight %, right) from NSG mice at day 35 post-engraftment with HMGA1-OE or CMV-NC HEL cells (n = 6 per group). Data are mean ± SD. Two-sample t -test. (h) Kaplan-Meier survival curves for NSG mice injected with HMGA1-OE ro CMV-NC HEL cells (n = 6 per group). Median survival times are indicated. Log-rank (Mantel-Cox) test. (i) Heatmaps showing HMGA1 binding intensity (CUT&Tag, left) and chromatin accessibility (ATAC-seq, right) centered on transcription start site (TSS ± 3kb) for genes in HEL cells transduced with shNC or shHMGA1. Color scale indicates normalized read counts (Max/Min normalized). (j) Top de novo motifs identified by HOMER analysis within ATAC-seq peak regions that either lose accessibility (left) or gain accessibility (right) upon HMGA1 knockdown in HEL cells. P -value for motif enrichment are indicated. (k) Quantification of apoptosis by Annexin V-APC/7-AAD staining and flow cytometry in HEL and UKE-1 cells after transduction with shNC or HMGA1 shRNAs (sh1, sh2). Representative flow cytometry plots are shown. Data are mean ± SD. (n = 5 per group). One-way ANOVA with Tukey’s post-hoc test.

Article Snippet: Lentiviral vectors for overexpression of human HMGA1 (NM_145899.3), murine Hmga1 (NM_016660.3), murin e Jak2 WT (NM_008413.4), or murin e Jak2 V617F were generated by cloning the respective open reading frames (ORFs) into the pLenti-CMV-Puro vector (Addgene plasmid #39481, a gift from Ie-Ming Shih).

Techniques: Transduction, Over Expression, shRNA, Knockdown, Cell Culture, Expressing, Control, Plasmid Preparation, Fluorescence, Staining, Immunohistochemistry, Injection, Binding Assay, Flow Cytometry

Journal: bioRxiv

Article Title: Targeting HMGA1-driven leukemic transformation in myeloproliferative neoplasms with pacritinib

doi: 10.1101/2025.06.01.657170

Figure Lengend Snippet: (a-f) HMGA1 expression levels modulate sensitivity to diverse therapeutic agents. Dose-response curves showing viability of HEL, UKE-1, Ba/F3 ( Jak2 wild type, or Jak2 V617F ), and 32D-cl3 ( Jak2 wild type, or Jak2 V617F ) cells with engineered HMGA1/Hmga1 expression (OE vs. NC; sh1/sh2 vs. sh-NC) following 72-hour treatment with (a) IFNα, (b) 5-Azacytidine, (c) Decitabine, (d) Cytarabine, (e) Venetoclax, and (f) Hydroxyurea. Calculated IC50 values are shown. Data represent mean ± SD from n = 3 independent experiments. Two-way ANOVA. (g) Ruxolitinib treatment, particularly long-term exposure, alters key signaling and cell cycle protein expression. Western blot analysis of indicated JAK-STAT, E2F pathway, G2M checkpoint, and cell cycle regulatory proteins in HEL and UKE-1 cells treated with vehicle, short-term ruxolitinib (4 hours), or in ruxolitinib-persistent (Rux-P) lines. GAPDH served as loading control. (h-j) HMGA1/Hmga1 expression status influences sensitivity to JAK inhibitors. Dose-response curves assessing viability of (h) UKE-1 cells, (i) Ba/F3 cells (J VF /NC: Jak2 V617F /control vector; J VF /OE: Jak2 V617F /Hmga1 OE; J WT /NC: Jak2 wild-type/control vector; J WT /OE: Jak2 wild-type /Hmga1 OE), and (j) 32D-cl3 cells (similarly engineered) with engineered HMGA1/Hmga1 expression, following treatment with ruxolitinib, fedratinib, pacritinib, or momelotinib. Calculated IC50 values are shown. Data represent mean ± SD from n = 3 independent experiments. Two-way ANOVA. (k) Pacritinib mitigates weight loss in mice bearing HMGA1-overexpressing HEL xenografts. Percent body weight change in NSG mice engrafted with HEL-Luc cells (CMV-NC or HMGA1-OE) and treated with vehicle or pacritinib (100 mg/kg, BID, 14 days). Data are mean ± SD (n = 6 per group). One-way ANOVA. (l) Pacritinib treatment improves hematological parameters in the HMGA1-overexpressing HEL xenograft model. Peripheral blood counts (WBC, HGB, HCT, PLT) in xenografted mice at day 35 endpoint. Data are mean ± SD (n = 6 per group). One-way ANOVA.

Article Snippet: Lentiviral vectors for overexpression of human HMGA1 (NM_145899.3), murine Hmga1 (NM_016660.3), murin e Jak2 WT (NM_008413.4), or murin e Jak2 V617F were generated by cloning the respective open reading frames (ORFs) into the pLenti-CMV-Puro vector (Addgene plasmid #39481, a gift from Ie-Ming Shih).

Techniques: Expressing, Western Blot, Control, Plasmid Preparation

Journal: bioRxiv

Article Title: Targeting HMGA1-driven leukemic transformation in myeloproliferative neoplasms with pacritinib

doi: 10.1101/2025.06.01.657170

Figure Lengend Snippet: (a) UMAP projections of bone marrow mononuclear cells (BMMCs) from healthy donors (n = 10) and MPN patients progressing to sAML (n = 3) (GSE185381), colored by (i) sample origin, (ii) annotated cell type, and (iii) HMGA1 transcript expression level. (iv) Correlation plot showing the fraction of HMGA1 -positive cells versus the fraction of anueploid-define malignant cells in paried MPN-sAML samples. Two-sample t -test. (b) Quantitative RT-PCR analysis of HMGA1 mRNA expression in (i) peripheral blood mononuclear cells (PBMCs; n = 7, 13, 12, 12, 11 for healthy, ET, PV, MF, sAML respectively) and (ii) CD34 + cells (n = 7, 11, 10, 8, 9 for healthy, ET, PV, MF, sAML respectively). Data are presented as mean ± standard deviation (SD). One-way ANOVA with Tukey’s post-hoc test. (c) Oncoprint depicting HMGA1 expression and mutational landscape in paired MF and subsequent sAML samples (n = 7 pairs) (GSE210253). (i) Heatmap illustrates relative HMGA1 expression (color intensity) and mutation burden. (ii) Statistical comparison of HMGA1 expression, driver mutation load, and high-risk mutation load between MF and sAML phase. Paired two-sample t -test. (d) Identifying of five surface markers (CD34, PROM1, KIT, FUT4, CD82) via CITE-seq, whose corresponding mRNA transcripts are significantly elevated in malignant cell populations from sAML patients (GSE185381). (e) Stacked bar chart illustrating the proportions of distinct hematopoietic cell subsets identified by unsupervised clustering of BMMCs from healthy controls and sAML patients (GSE185381). (f) Expression trajectory plot depicting the distribution of transcript expression for HMGA1 , CD34 , PROM1 (CD133), KIT (CD117), FUT4 (CD15), and CD82 across annotated cell subsets, ordered by developmental hierarchy, for (i) all cells and (ii) malignant cells only.

Article Snippet: To investigate the impact of endogenous Jak2 V617F expression on Hmga1, Jak2-Flox-V617F/Vav1-Cre-Tg mice (C57BL/6J background; Shanghai Model Organisms Center, Inc., Cat# NM-XA-242440) were utilized.

Techniques: Expressing, Quantitative RT-PCR, Standard Deviation, Mutagenesis, Comparison

Journal: bioRxiv

Article Title: Targeting HMGA1-driven leukemic transformation in myeloproliferative neoplasms with pacritinib

doi: 10.1101/2025.06.01.657170

Figure Lengend Snippet: (a) Box plots showing HMGA1 expression in bone marrow mononuclear cells (BMMCs) from healthy individuals (n = 10) and sAML patients (n = 3; malignant cells and microenvironment components depicted separately), based on CITE-seq data (GSE185381). Statistical analysis by Kruskal-Wallis test with Benjamini-Hochberg (BH) correction. (b) Left: Correlation between mean HMGA1 transcript expression and percentage of HMGA1-positive cells from healthy controls and sAML patients (GSE185381). Right: Bar plot comparing mean HMGA1 expression in total BMMCs between healthy controls (n = 10) and sAML patients (n = 3). Pearson correlation coefficient ( R ) and P -value shown. Two-sample t -test with BH correction. (c) Log2-transformed HMGA1 expression in CD34 + peripheral blood mononuclear cells (PBMCs) from patients with chronic-phase MPN (ET, n = 2; PV, n = 1; MF, n = 5), derived from dataset GSE189979. Wald test with BH correction. (d) Relative HMGA1 mRNA expression, determined by qRT-PCR, in (i) PBMCs and (ii) CD34 + cells from MPN patients (stratified by JAK2 mutation status: JAK2 wile-type PBMCs, n = 21; JAK2 V617F PBMCs, n = 23; JAK2 wile-type CD34 + , n = 20; JAK2 V617F CD34 + , n = 16) and sAML patients (stratified by JAK2 mutation status: JAK2 wile-type sAML PBMCs, n = 4; JAK2 V617F PBMCs, n = 7; JAK2 wile-type CD34 + , n = 4; JAK2 V617F CD34 + , n = 5). Data are presented as mean ± SD. One-way ANOVA. (e) Relative HMGA1 mRNA expression, determined by qRT-PCR, in (i) PBMCs from sAML patients (stratified by TP53 mutation status: TP53 wild-type, n = 6; TP53 mutated, n = 5) and (ii) CD34 + cells from sAML patients ( TP53 wild-type, n = 5; TP53 mutated, n = 4). Data are presented as mean ± SD. Two-sample t -test. (f) Expression trajectory plot depicting protein expression of HMGA1 and key stem/progenitor markers (CD34, PROM1 /CD133, KIT /CD117, FUT4 /CD15, CD82) across diverse hematopoietic cell subsets, ordered by developmental hierarchy, within (i) total BMMCs and (ii) the malignant sAML cell compartment (GSE185381). (g) Violin plots comparing (i) transcript levels (CITE-seq) and (ii) surface protein expression (ADT) of CD34, PROM1 (CD133), KIT (CD117), FUT4 (CD15), and CD82 between malignant and non-malignant BMMCs from healthy controls (n = 10) and sAML patients (n = 3) (GSE185381). Wald test with BH correction.

Article Snippet: To investigate the impact of endogenous Jak2 V617F expression on Hmga1, Jak2-Flox-V617F/Vav1-Cre-Tg mice (C57BL/6J background; Shanghai Model Organisms Center, Inc., Cat# NM-XA-242440) were utilized.

Techniques: Expressing, Transformation Assay, Derivative Assay, Quantitative RT-PCR, Mutagenesis

Journal: bioRxiv

Article Title: Targeting HMGA1-driven leukemic transformation in myeloproliferative neoplasms with pacritinib

doi: 10.1101/2025.06.01.657170

Figure Lengend Snippet: (a) UMAP visualization of 9,191 Lin - CD34 + cells from 14 sAML patient samples (GSE226340), showing pseudotime trajectories. Cells are colored by (left to right): differentiation trajectory score (LSC-like vs. erythroid), HMGA1 expression level, and TP53 mutational status (wild type, single-hit, multi-hit). Arrows indicate inferred differentiation direction. (b) Violin plots showing HMGA1 transcript expression in CD34 + HSPCs from sAML patients (GSE226340), stratified by TP53 mutational status (wild type, single-hit, multi-hit). (c) Violin plots showing HMGA1 transcript expression in CD34 + HSPCs from sAML patients (GSE226340), stratified by specific combinations of JAK2 (wild-type or V617F) and TP53 (wild-type or mutated) genotype. (d) HMGA1 expression in CD34 + HSPCs from sAML patients (GSE226340) stratified by combinations of TP53 genotype (wild-type or mutated) with other co-occurring gene mutations ( CALR , TET2 , U2AF1 ) versus TP53 mutation alone. (e) HMGA1 expression in BMMCs from sAML patients (GSE185381) stratified by TP53 mutational status (wild type, n = 2 vs. mutated, n = 1). (f) HMGA1 expression in BMMC from patients with sAML (GSE185381) stratified by karyotype complexity (non-complex karyotype, n = 1 vs. complex karyotype, n = 2) (g) Hmga1 transcript expression in murine hematopoietic progenitors from various Jak2 / Trp53 mutant mouse models (GSE180851). (i) Megakaryocyte-erythorid progenitors (MEP) from JAK2 V617F ( J VF ) mice versus J VF mice with different Trp53 genotypes ( Trp53 -/- , Trp53 +/- , Trp53 R172H/- or Trp53 R172H/+ ) (n = 3 per group). (ii) Lin-Sca-1 + c-kit + (LSK) stem cells from the same genotypes. (h) HMGA1 expression in MEPs from mice at sAML stage ( Jak2 V617F Trp53 -/- , n = 3; Jak2 V617F Trp53 R172H/- , n = 3) (GSE180851). (i) Cellular composition and marker expression in representative sAML cases from GSE185381. (i) Stacking bar charts showing the percentage of annotated cell types in two sAML cases (AML3762, AML3948). (ii) Violin plots depicting HMGA1 , CD34 , and KIT transcript expression in the top 5 most abundant malignant cell types and microenvironment (non-malignant) cells for each patient. For panels b-h, statistical significance was assessed by Kruskal-Wallis test with Benjamini-Hochberg (BH) correction (b-d), Wilcoxon rank-sum test with BH correction (e,f), likelihood ratio test (g), or Wald test (h) as appropriate.

Article Snippet: To investigate the impact of endogenous Jak2 V617F expression on Hmga1, Jak2-Flox-V617F/Vav1-Cre-Tg mice (C57BL/6J background; Shanghai Model Organisms Center, Inc., Cat# NM-XA-242440) were utilized.

Techniques: Expressing, Mutagenesis, Marker

Journal: bioRxiv

Article Title: Targeting HMGA1-driven leukemic transformation in myeloproliferative neoplasms with pacritinib

doi: 10.1101/2025.06.01.657170

Figure Lengend Snippet: (a) Flow cytometric quantification of nuclear HMGA1 protein in CD34 + blast cells versus matrue hematopoietic lineages from representative MPN patient bone marrow aspirates (n = 7). Data presented as percentage of HMGA1-positive cells. One-way ANOVA with Tukey’s post-hoc test. (b) HMGA1 positive percentage in peripheral blood leukocytes from healthy donors (n = 12), chronic-phase MPN patients (n = 11), and sAML patients (n = 3). One-way ANOVA with Tukey’s post-hoc test. (c) HMGA1 immunohistochemistry (IHC) scores in bone marrow biopsies from patients with ET (n = 57), PV (n = 33), MF (n = 51), sAML (n = 21), and non-MPN controls (n = 78). One-way ANOVA with Tukey’s post-hoc test. Scale bar: top overview, 80 µm; bottom inset, 20 µm. Insets show representative HMGA1 IHC staining. (d, e) Representative HMGA1 IHC staining in paried bone marrow biopsies from patients at chronic phase MPN versus post-myelofibrosis (post-MF, n = 3) and (e) chronic phase MPN vsrsus sAML transformation (n = 5). Scale bars as in (c). Paired two-sample t -test. (f) (i) Receiver operating characteristic (ROC) curve analysis of of HMGA1 IHC score for discriminating sAML (n = 21) from non-sAML MPN (n = 141). Area under the curve (AUC) is indicated. Right panel: contingency table at the optimal HMGA1 cut-off. (ii) Sankey diagram illustrating the 1-year clinical course of 15 n on-blast-phase MPN patients with high baseline HMGA1 expression (> 40.78%) but < 20% bone marrow blasts. (g) Longitudinal analysis of two representative MPN patients who progressed to sAML. Serial bone marrow biopsy images (HMGA1 IHC, scale bar: 10 µm); bone marrow smear cytology (Wright-Giemsa staining, scale bar: 20 µm); and corresponding flow cytometry dot plots (CD45 vs. side scatter) at indicated time points, demonstrating HMGA1 upregulation and moprpholoigcal changes preceding overt blast crisis.

Article Snippet: To investigate the impact of endogenous Jak2 V617F expression on Hmga1, Jak2-Flox-V617F/Vav1-Cre-Tg mice (C57BL/6J background; Shanghai Model Organisms Center, Inc., Cat# NM-XA-242440) were utilized.

Techniques: Immunohistochemistry, Transformation Assay, Expressing, Staining, Flow Cytometry

Journal: bioRxiv

Article Title: Targeting HMGA1-driven leukemic transformation in myeloproliferative neoplasms with pacritinib

doi: 10.1101/2025.06.01.657170

Figure Lengend Snippet: (a) HMGA1 IHC score in MPN patients with (n = 95) versus without (n = 67) splenomegaly. Representative IHC images of normal spleen and splenomegaly are shown. Data are mean ± SD. Two-sample t -test. Scale bars: top, 80 µm; bottom, 20 µm. (b) HMGA1 IHC scores in MPN patients stratified by the latency period (in months) between initial suspicion of MPN and bone marrow biopsy (< 10 months, n = 66; ≥10 months, n = 90). Data are mean ± SD. Two-sample t -test. (c) HMGA1 IHC score in MPN patients stratified by bone marrow reticulin fibrosis grades (MF-0, n = 53; MF-1, n = 33; MF-2, n = 32; MF-3, n = 13). Data are mean ± SD. One-way ANOVA with Tukey’s post-hoc test. (d) Correlation plots of HMGA1 IHC scores versus peripheral blood parameters: white blood cell coutn (WBC), hemoglobin (HGB), hematocrit (HCT), and platelet count (PLT) in the total MPN cohort (n = 162). Pearson correlation coefficient ( r ) and P -value are indicated. (e) HMGA1 IHC score in sAML patients (n = 21) stratified by European Leukemia Net (ELN) 2022 risk categories (Intermediate vs. Adverse). Data are mean ± SD. Two-sample t -test. (f) Representative multiplex immunofluorescence (mIF) images of bone marrow sections from primary MF and sAML patients, co-stained for HMGA1 (yellow) with CD117 (red), CD34 (green). Nuclei are counterstained with DAPI (blue). HMGA1 is highly expressed in megakaryocytes in PMF and in blasts in sAML. Scale bar: 20 µm. (g-j) Quantification of positive %, intensity and a composite coefficient (positive % × intensity) for (g) HMGA1, (h) CD34, (i) CD117, across different MPN subtypes (ET, n = 12; PV, n = 15; MF, n = 12) and, sAML (n = 11) and (j) three marker in sAML (n = 11). Data are mean ± SD. One-way ANOVA with Tukey’s post-hoc test. (k) Serial H&E and IHC (HMGA1, CD34, CD117) staining of bone marrow biopsies from a representative patient progressing from ET (August 2020) thourgh secondary MF (sMF; September 2022, November 2022, Februrary 2023, May2023) to sAML (August 2023). Scale bars: top, 80 µm; bottom, 20 µm.

Article Snippet: To investigate the impact of endogenous Jak2 V617F expression on Hmga1, Jak2-Flox-V617F/Vav1-Cre-Tg mice (C57BL/6J background; Shanghai Model Organisms Center, Inc., Cat# NM-XA-242440) were utilized.

Techniques: Multiplex Assay, Immunofluorescence, Staining, Marker

Journal: bioRxiv

Article Title: Targeting HMGA1-driven leukemic transformation in myeloproliferative neoplasms with pacritinib

doi: 10.1101/2025.06.01.657170

Figure Lengend Snippet: (a) HMGA1 protein expression (percentage of IHC-positive cells) in bone marrow biopsies from a patient cohort, stratified by sex (female, n = 140; male, n = 67) and age (<60 years, n = 56; ≥60 years, n = 75). Statistical analysis by two-sample t-test. (b-c) Association of HMGA1 protein levels (IHC) with established clinical risk scores in (b) essential thrombocythemia (ET, n = 53) using ELN, MIPSS-ET, IPSET, and Revised-IPSET scoring systems, and (c) polycythemia vera (PV, n = 33) using ELN, MIPSS-PV, and IWG-PV scoring systems. Two-sample t -test or one-way ANOVA, as appropriate. (d-e) HMGA1 protein levels (IHC) in (d) primary myelofibrosis (PMF, n = 20) stratified by IPSS, DIPSS, and MIPSS-70 clinical risk scores, and (e) secondary myelofibrosis (sMF, n = 12) stratified by MYSEC-PM score. One-way ANOVA. (f) Differential HMGA1 protein expression (IHC) between MPN ( JAK2 wild-type, n = 53; JAK2 V617F , n = 88) and sAML ( JAK2 wild-type, n = 9; JAK2 V617F , n = 12). Representative IHC images for JAK2 wild-type MPN and sAML are inset. One-way ANOVA. (g) HMGA1 protein expression (IHC) in sAML patient samples stratified by TP53 mutation status ( TP53 wild-type, n = 14; TP53 mutated, n = 7). A representative IHC image for TP53 wild-type sAML is inset. Two-sample t -test. (h-j) Elevated Hmga1 expression in a Jak2 V617F knock-in MPN mouse model. (h) Representative images of Hematoxylin and Eosin (H&E) staining and IHC for Hmga1 in spleen and bone marrow tissues from Jak2 wild-type (control) and Jak2 V617F mice. Scale bars, 50 µm. (i) Quantification of Hmga1 IHC-positive cells (%) in spleen (n = 6 per group) and bone marrow (n = 4 per group). (j) Relative Hmga1 mRNA levels in PBMCs from Jak2 wild-type (n = 3) and Jak2 V617F (n = 3) mice, determined by qRT-PCR. Data are presented as mean ± SD. Two-sample t -test. (k-l) Expression profiling of HMGA1, CD34, and CD117 in HEL and UKE-1 MPN cell lines. (k) Representative images of Wright-Giemsa staining (left panels) and immunocytochemical (ICC) staining for HMGA1, CD34, and CD117 (right panels). Scale bars, 80 µm (overview), 20 µm (inset). (l) Flow cytometric analysis of intracellular HMGA1 and surface CD34/CD117 expression. (m) Multiplex immunofluorescence (mIF) demonstrating HMGA1 co-expression with stem/progenitor markers in bone marrow biopsies from MPN patients progressing to sMF (representative PV-sMF and ET-sMF cases). Images show H&E staining alongside fluorescence channels for DAPI (nuclei, blue), Vimentin (cyan), HMGA1 (yellow), CD34 (green), and CD117 (red). Scale bars, 80 µm (overview), 20 µm (inset). (n) Flow cytometric analysis revealing HMGA1, CD34, and CD117 expression within CD45 + blasts from an sAML patient’s peripheral blood (PB). (o) Confocal microscopy demonstrating co-localization of HMGA1 (green) with CD34 (red) and CD117 (yellow) in bone marrow (BM) cells from an sAML patient. Nuclei are counterstained with DAPI (blue). Scale bar, 10 µm.

Article Snippet: To investigate the impact of endogenous Jak2 V617F expression on Hmga1, Jak2-Flox-V617F/Vav1-Cre-Tg mice (C57BL/6J background; Shanghai Model Organisms Center, Inc., Cat# NM-XA-242440) were utilized.

Techniques: Expressing, Mutagenesis, Knock-In, Staining, Control, Quantitative RT-PCR, Multiplex Assay, Immunofluorescence, Fluorescence, Confocal Microscopy

Journal: bioRxiv

Article Title: Targeting HMGA1-driven leukemic transformation in myeloproliferative neoplasms with pacritinib

doi: 10.1101/2025.06.01.657170

Figure Lengend Snippet: (a) Efficient shRNA-mediated knockdown of HMGA1 (sh1, sh2 vs. sh-NC control) in HEL and UKE-1 cells. Left: Relative HMGA1 mRNA levels by qRT-PCR (mean ± SD, n = 3). Right: Western blot analysis of HMGA1 protein; ACTB served as loading control. (b) Lentiviral-mediated overexpression of HMGA1 (OE vs. CMV-NC control) in HEL and UKE-1 cells. Left: Relative HMGA1 mRNA levels by qRT-PCR (mean ± SD, n = 3). Right: Western blot analysis of HMGA1 protein; Tubulin served as loading control. (c) Lentiviral-mediated overexpression of Hmga1 (J/OE vs. J/NC control) in murine Ba/F3 ( Jak2 wild type, or Jak2 V617F ) and 32D-cl3 ( Jak2 wild type, or Jak2 V617F ) cells. Left: Relative Hmga1 mRNA levels by qRT-PCR (mean ± SD, n = 3). Right: Western blot analysis of Hmga1 protein; Tubulin served as loading control. Statistical analyses for (a-c) by two-sample t-test or one-way ANOVA, as appropriate. (d) HMGA1 overexpression exacerbates disease phenotype in a HEL xenograft model. Hematological parameters (WBC, white blood cell count; HGB, hemoglobin; HCT, hematocrit; PLT, platelet count) in NSG mice engrafted with HEL cells stably expressing control vector (CMV-NC, n = 6) or HMGA1 (OE, n = 6) at 35 days post-transplantation. Data are presented as mean ± SD. Two-sample t -test. (e) HMGA1 knockdown alters chromatin accessibility and HMGA1 binding at key cell cycle regulatory gene loci. Integrative Genomics Viewer (IGV) snapshots displaying ATAC-seq and HMGA1 CUT&Tag signals at representative E2F target genes ( E2F1 , CCNE1 , CCNE2 , CDK2 , RB1 ), G2M checkpoint genes ( CCNB1 , CCNB2 , CDC2 , WEE1 , CDC25C , PLK1 , AURKA , AURKB ), and common cell cycle regulators ( CCNA2 , CDKN1A / p21 , CDKN1B / p27 ) in HEL cells following control (NC) versus HMGA1 knockdown (KD). (f) Enhanced E2F target and G2M checkpoint gene signatures in sAML patient cells. UMAP projections of single-cell CITE-seq data (GSE185381) from control and sAML patients, with cells colored by enrichment scores for E2F target and G2M checkpoint gene sets. Corresponding density plots illustrate score distributions.

Article Snippet: To investigate the impact of endogenous Jak2 V617F expression on Hmga1, Jak2-Flox-V617F/Vav1-Cre-Tg mice (C57BL/6J background; Shanghai Model Organisms Center, Inc., Cat# NM-XA-242440) were utilized.

Techniques: shRNA, Knockdown, Control, Quantitative RT-PCR, Western Blot, Over Expression, Cell Counting, Stable Transfection, Expressing, Plasmid Preparation, Transplantation Assay, Binding Assay

Journal: bioRxiv

Article Title: Targeting HMGA1-driven leukemic transformation in myeloproliferative neoplasms with pacritinib

doi: 10.1101/2025.06.01.657170

Figure Lengend Snippet: (a) Colony-forming units (CFU) assays in human (HEL, UKE-1) and murine (Ba/F3- JAK2 V617F , 32D- JAK2 V617F ) cell lines following HMGA1 knockdown (shHMGA1) or overexpression (OE-HMGA1) relative to respective controls (NC). Data are mean ± SD. (n=5 per group). Representative colony images are shown (scale bar 300 μm).Two-sample t -test and one-way ANOVA with Tukey’s post-hoc test. (b) Schematic of the HEL cell xenograft experiment in NSG mice, comparing HMGA1-OE cells to vector control cells. (c) Representative bioluminescence images (left panel) and quantification of total photon flux (right panel) from NSG mice at day 35 post-engraftment with luciferase-tagged HEL cells transduced with vector control (CMV-NC, n=6) or HMGA1 overexpression constructs (OE, n=6). Data are mean ± SD. Two-sample t -test. (d) Correlation between HMGA1 mRNA expression (snRNA-seq) and the number of open chromatin peaks (scATAC-seq) in CD34 + cells from a patient during primary MF and after sAML transformation (GSE221946). Pearson correlation coefficient ( R ) and P -value from Wilcoxon ranksum test with BH correction. (e) Genomic distribution of HMGA1 binding sites lost upon HMGA1 knockdown in HEL cells, as determined by CUT&Tag (i), and chromatin regions with significantly reduced accessibility upon HMGA1 knockdown, as determined by ATAC-seq (ii). Stacked bar chats show the percentage of peaks located in promoter (±3kb of TSS), intronic, intergenic, and other genomic regions. (f) GSEA enrichment plots for E2F targets and G2M checkpoint Hallmark gene sets. Comparisons shown in figure. Normalized Enrichment Score (NES) and False Discovery Rate (FDR) q-value are indicated. (g) Cell cycle distribution (Propidium Iodide staining followed by flow cytometry) of HEL and UKE-1 cells with HMGA1 knockdown (sh1, sh2) or non-targeting control (NC). Data are mean ± SD. (n = 5 per group). One-way ANOVA with Tukey’s post-hoc test. (h) Western blots analysis of key E2F pathway, G2M checkpoint, and common cell cycle regulatory protein levels in HEL and UKE-1 cells following overexpression (OE vs. CMV-NC vector control; left panels) or HMGA1 knockdown (shHMGA1 vs. shNC control; right panels). GAPDH served as loading controls. Blots are representative of at least three independent experiments.

Article Snippet: To investigate the impact of endogenous Jak2 V617F expression on Hmga1, Jak2-Flox-V617F/Vav1-Cre-Tg mice (C57BL/6J background; Shanghai Model Organisms Center, Inc., Cat# NM-XA-242440) were utilized.

Techniques: Knockdown, Over Expression, Plasmid Preparation, Control, Luciferase, Transduction, Construct, Expressing, Transformation Assay, Binding Assay, Staining, Flow Cytometry, Western Blot

Journal: bioRxiv

Article Title: Targeting HMGA1-driven leukemic transformation in myeloproliferative neoplasms with pacritinib

doi: 10.1101/2025.06.01.657170

Figure Lengend Snippet: (a) Relative proliferation curves of human (HEL, UKE-1) and murine (Ba/F3, 32D-cl3 transduced with Jak2 wild-type or Jak2 V617F ) cell lines following HMGA1/Hmga1 overexpression (OE) or shRNA-meidated knockdown (sh1, sh2) compared to respective controls (CMV-NC or sh-NC)NC.) 32D-cl3 cells were cultured with IL-3. Data are mean ± SD. (n = 5 per group). Two-way ANOVA. (b) Flow cytometric analysis of CD11b expression on 32D-cl3 cells transduced with Jak2 wild-type (J WT ) or Jak2 V617F (J VF ), and co-transduced with control vector (NC) or HMGA1 overexpression (OE), following G-CSF (100 ng/mL) induced differentiation. (i) Representative histograms of CD11b-FITC fluorescence. (ii) Quantification of HMGA1-PE mean fluorescence intensity (MFI). (iii) Quantification of CD11b-FITC MFI (n = 5 per group). Data are mean ± SD. Two-sample t -test. (c) Quantification of human CD45 + CD117 + HEL cells in peripheral blood of NSG mice at day 35 post-transplant, comparing HMGA1-OE versus vector control (CMV-NC) groups (n=6 per group). Data are mean ± SD. Two-sample t -test. (d) Wright-Giemsa stained peripheral blood smears from NSG mice engrafted with HMGA1-OE or CMV-NC HEL cells at day 35. Quantification of HEL cells (% of total nucleated cells) is shown (n = 6 per group). Data are mean ± SD. Two-sample t -test. (e-f) Representative H&E and HMGA1 IHC staining (left panels of e and f, respectively) and quantification of HMGA1-positive cells (%) (right panels fo e and f, respectively) in (e) femur bone marrow and (f) spleen sections from NSG mice engrafted with HMGA1-OE or CMV-NC HEL cells. Scale bars: 50 µm. Data are mean ± SD. Two-sample t -test. (g) Representative images of spleens (left) and relative spleen weights (spleen weight/body weight %, right) from NSG mice at day 35 post-engraftment with HMGA1-OE or CMV-NC HEL cells (n = 6 per group). Data are mean ± SD. Two-sample t -test. (h) Kaplan-Meier survival curves for NSG mice injected with HMGA1-OE ro CMV-NC HEL cells (n = 6 per group). Median survival times are indicated. Log-rank (Mantel-Cox) test. (i) Heatmaps showing HMGA1 binding intensity (CUT&Tag, left) and chromatin accessibility (ATAC-seq, right) centered on transcription start site (TSS ± 3kb) for genes in HEL cells transduced with shNC or shHMGA1. Color scale indicates normalized read counts (Max/Min normalized). (j) Top de novo motifs identified by HOMER analysis within ATAC-seq peak regions that either lose accessibility (left) or gain accessibility (right) upon HMGA1 knockdown in HEL cells. P -value for motif enrichment are indicated. (k) Quantification of apoptosis by Annexin V-APC/7-AAD staining and flow cytometry in HEL and UKE-1 cells after transduction with shNC or HMGA1 shRNAs (sh1, sh2). Representative flow cytometry plots are shown. Data are mean ± SD. (n = 5 per group). One-way ANOVA with Tukey’s post-hoc test.

Article Snippet: To investigate the impact of endogenous Jak2 V617F expression on Hmga1, Jak2-Flox-V617F/Vav1-Cre-Tg mice (C57BL/6J background; Shanghai Model Organisms Center, Inc., Cat# NM-XA-242440) were utilized.

Techniques: Transduction, Over Expression, shRNA, Knockdown, Cell Culture, Expressing, Control, Plasmid Preparation, Fluorescence, Staining, Immunohistochemistry, Injection, Binding Assay, Flow Cytometry

Journal: bioRxiv

Article Title: Targeting HMGA1-driven leukemic transformation in myeloproliferative neoplasms with pacritinib

doi: 10.1101/2025.06.01.657170

Figure Lengend Snippet: (a) Prognostic significance of HMGA1 expression in the OHSU BeatAML MPN-sAML cohort (n = 31). Genes are ranked by their hazard ratio (HR) for overall survival (OS). Points are colored based on FDR significance: grey ( FDR > 0.05), blue ( FDR < 0.05 & HR < 1, good prognosis), red ( FDR < 0.05 & HR ≥ 1, poor prognosis) (b) Gene Set Enrichment Analysis (GSEA) showing Hallmark pathways enriched among genes associated with poor prognosis (ranked by HR) in the OHSU BeatAML MPN-sAML cohort. Heatmap displays NES for selected pathways. Asterisks (*) indicate FDR < 0.05. (c) Kaplan-Meier OS curves for MPN-sAML patients from the OHSU BeatAML cohort (n = 31, top panel) and an independent in-house cohort (n = 21, bottom pnel), stratified by high versus low HMGA1 expression (HMGA1 expression levels for BeatAML in-house cohort using median cut-off). Log-rank (Mantel-Cox) test. (d) Representative immunohistochemical (IHC) staining for HMGA1 in bone marrow biopsies from in-house MPN-sAML cohort patients, illustrating HMGA1 expression changes with therapy and clinical outcome. (i) HMGA1-low patient achieving complete remission (CR) post-ruxolitinib. (ii) HMGA1-low patient achieving CR post-decitabine + venetoclax. (iii) HMGA1-high patient with progressive disease (PD) despite 5-azacytidine + venetoclax, showing increased HMGA1 at relapse. (iv) HMGA1-high patient achieving durable remission with decreased HMGA1 staining post-allogeneic hematopoietic stem cell transplantation (allo-HCT). Scale bars: 80µm (overview), 20µm (insets). (e) Comparison of HMGA1 expression levels between MPN-sAML patients achieving complete remission (CR, includes CRh, CRi) and those not achieving CR (Non-CR). Top panel: HMGA1 transcript levels (Log2 normalized counts) in the OHSU BeatAML cohort (n=31). Bottom panel: Percentage of HMGA1-positive cells (IHC score) in the in-house cohort (n=21). Data are mean ± SD. Two-sample t -test. (f) Heatmap illustrating Hallmark GSEA of differentially expressed genes in HEL cells treated with DMSO (vehicle), ruxolitinib (Rux), fedratinib (Fed), pacritinib (Pac), or momelotinib (Mmb) for 4 hours or 48 hours (GSE229712) and in HEL cells with acquired resistance to ruxolitinib (Rux-Persistent, GSE190517) compared to DMSO control. Color intensity represents NES. * indicate FDR < 0.05. (g) Dose-response curves depicting cell viability of parental (NC) control versus ruxolitinib-persistent (Rux-P) HEL (left) and UKE-1 (right) cells treated with indicated concentrations of ruxolitinib for 72 hours. IC 50 values (mean± SD) are shown. Two-way ANOVA comparing IC 50 values. (h) HMGA1 mRNA expression (RNA-seq, normalized counts) in HEL cells: non-targeting control (NC), ruxolitinib-persistent (Rux-P), and fedratinib-persistent (Fed-P). (I) Immunoblot analysis of HMGA1 protein levels in parental (NC) and and ruxolitinib-persistent (Rux-P) HEL and UKE-1 cells. Tublin serves as a loading control. (J) Dose-response curve showing cell viability of HEL cells transduced with control vector (NC), HMGA1 overexpression (OE), or HMGA1 shRNA (Sh1) constructs, treated with indicated concentrations of ruxolitinib (Rux), fedratinib (Fed), pacritinib (Pac), and momelotinib (Mmb) for 72 hours. IC50 values (mean ± SD) are shown. Two-way ANOVA comparing IC50 values between OE/Sh1 and respective NC. (k) Schematic representation of the in vivo pacritinib treatment study in NSG mice engrafted with luciferase-expressing HEL cells (transduced with CMV-NC vector or HMGA1-OE construct). Following leukemia engraftment (Day 0-21), mice received pacritinib (100 mg/kg, BID) or vehicle orally for 14 days (Day 21-35). Endpoint analyses included bioluminescence imaging, spleen weight, flow cytometry, Wright-Giemsa staining, H&E, and IHC, alongside survival monitoring. (l) Representative bioluminescence image (left) and quatification of total tumor bioluminescence (total flux, right) at day 35 in NSG mice engrafted with CMV-NC or HMGA1-OE HEL cells and treated with vehicle or pacritinib (n=6 mice/group). Data are shown in mean ± SD. One-way ANOVA with Tukey’s post-hoc test.

Article Snippet: To investigate the impact of endogenous Jak2 V617F expression on Hmga1, Jak2-Flox-V617F/Vav1-Cre-Tg mice (C57BL/6J background; Shanghai Model Organisms Center, Inc., Cat# NM-XA-242440) were utilized.

Techniques: Expressing, Immunohistochemical staining, Immunohistochemistry, Staining, Transplantation Assay, Comparison, Control, RNA Sequencing, Western Blot, Transduction, Plasmid Preparation, Over Expression, shRNA, Construct, In Vivo, Luciferase, Imaging, Flow Cytometry

Journal: bioRxiv

Article Title: Targeting HMGA1-driven leukemic transformation in myeloproliferative neoplasms with pacritinib

doi: 10.1101/2025.06.01.657170

Figure Lengend Snippet: (a) Forest plot illustrating multivariable Cox proportional hazards analysis of overall survival (OS) in the OHSU BeatAML cohort (n = 31) and the in-house MPN-sAML cohort (n = 21). Harzard ratio (HR) and 95% confidence intervals (CI) are shown for HMGA1 epxression (high vs. low), age, gender and ELN 2022 risk category. (b) Kaplan-Meier OS curves for MPN-sAML patients from the OHSU BeatAML cohort stratified by specific therapies received (ruxolitinib, 5-azacytidine, decitabine, cytarabine,and hydroxyurea). Log-rank (Mantel-Cox) test. (c) Western blot analysis key proteins in JAK-STAT, E2F, and G2/M pathways in HEL and UKE-1 cells with vector control (CMV-NC) or HMGA1 overexpression (OE), following 4-hour treatment with vehicle (DMSO) or pacritinib (500nM). (d) Quantification of human HEL cells in peripheral blood of NSG mice at day 35 post-transplantation by flow cytometric analysis (CD45 + CD117 + ) or Writght_Giemsa smear, from mice engrafted with vector control (CMV-NC) or HMGA1-OE HEL cells and treated with vehicle or pacritinib (n=6 mice/group). Data are mean ± SD. One-way ANOVA. (e-f) Pacritinib reduces leukemic infiltration and HMGA1 expression in bone marrow and spleen of HMGA1-OE xenografted mice. Representative IHC staining for HMGA1 and quantification of human HEL cell engraftment (% of HEL cells) and HMGA1-positive cells (%) in (e) bone marrow and (f) spleen sections. NSG mice (n=6 per group) were engrafted with luciferase-expressing HEL cells transduced with CMV-NC or HMGA1-OE constructs, and subsequently treated with vehicle or pacritinib (100 mg/kg, BID) for 14 days. Scale bars for IHC images, 50 µm. Bar graphs depict mean ± SD. One-way ANOVA with Tukey’s post-hoc test. (g) Representative images of spleens and relative spleen weights (spleen weight/body weight %) from NSG mice engrafted with CMV-NC or HMGA1-OE HEL cells, treated with vehicle or pacritinib. One-way ANOVA with Tukey’s post-hoc test. (h) Kaplan–Meier survival curves for NSG mice engrafted with vector control HEL cells (CMV-NC) or HMGA1-OE HEL cells, treated with vehicle or pacritinib (n=6 per group). Log-rank (Mantel-Cox) test.

Article Snippet: To investigate the impact of endogenous Jak2 V617F expression on Hmga1, Jak2-Flox-V617F/Vav1-Cre-Tg mice (C57BL/6J background; Shanghai Model Organisms Center, Inc., Cat# NM-XA-242440) were utilized.

Techniques: Western Blot, Plasmid Preparation, Control, Over Expression, Transplantation Assay, Expressing, Immunohistochemistry, Luciferase, Transduction, Construct

Journal: bioRxiv

Article Title: Targeting HMGA1-driven leukemic transformation in myeloproliferative neoplasms with pacritinib

doi: 10.1101/2025.06.01.657170

Figure Lengend Snippet: (a-f) HMGA1 expression levels modulate sensitivity to diverse therapeutic agents. Dose-response curves showing viability of HEL, UKE-1, Ba/F3 ( Jak2 wild type, or Jak2 V617F ), and 32D-cl3 ( Jak2 wild type, or Jak2 V617F ) cells with engineered HMGA1/Hmga1 expression (OE vs. NC; sh1/sh2 vs. sh-NC) following 72-hour treatment with (a) IFNα, (b) 5-Azacytidine, (c) Decitabine, (d) Cytarabine, (e) Venetoclax, and (f) Hydroxyurea. Calculated IC50 values are shown. Data represent mean ± SD from n = 3 independent experiments. Two-way ANOVA. (g) Ruxolitinib treatment, particularly long-term exposure, alters key signaling and cell cycle protein expression. Western blot analysis of indicated JAK-STAT, E2F pathway, G2M checkpoint, and cell cycle regulatory proteins in HEL and UKE-1 cells treated with vehicle, short-term ruxolitinib (4 hours), or in ruxolitinib-persistent (Rux-P) lines. GAPDH served as loading control. (h-j) HMGA1/Hmga1 expression status influences sensitivity to JAK inhibitors. Dose-response curves assessing viability of (h) UKE-1 cells, (i) Ba/F3 cells (J VF /NC: Jak2 V617F /control vector; J VF /OE: Jak2 V617F /Hmga1 OE; J WT /NC: Jak2 wild-type/control vector; J WT /OE: Jak2 wild-type /Hmga1 OE), and (j) 32D-cl3 cells (similarly engineered) with engineered HMGA1/Hmga1 expression, following treatment with ruxolitinib, fedratinib, pacritinib, or momelotinib. Calculated IC50 values are shown. Data represent mean ± SD from n = 3 independent experiments. Two-way ANOVA. (k) Pacritinib mitigates weight loss in mice bearing HMGA1-overexpressing HEL xenografts. Percent body weight change in NSG mice engrafted with HEL-Luc cells (CMV-NC or HMGA1-OE) and treated with vehicle or pacritinib (100 mg/kg, BID, 14 days). Data are mean ± SD (n = 6 per group). One-way ANOVA. (l) Pacritinib treatment improves hematological parameters in the HMGA1-overexpressing HEL xenograft model. Peripheral blood counts (WBC, HGB, HCT, PLT) in xenografted mice at day 35 endpoint. Data are mean ± SD (n = 6 per group). One-way ANOVA.

Article Snippet: To investigate the impact of endogenous Jak2 V617F expression on Hmga1, Jak2-Flox-V617F/Vav1-Cre-Tg mice (C57BL/6J background; Shanghai Model Organisms Center, Inc., Cat# NM-XA-242440) were utilized.

Techniques: Expressing, Western Blot, Control, Plasmid Preparation

Journal: bioRxiv

Article Title: Targeting HMGA1-driven leukemic transformation in myeloproliferative neoplasms with pacritinib

doi: 10.1101/2025.06.01.657170

Figure Lengend Snippet: (a) Box plots showing HMGA1 expression in bone marrow mononuclear cells (BMMCs) from healthy individuals (n = 10) and sAML patients (n = 3; malignant cells and microenvironment components depicted separately), based on CITE-seq data (GSE185381). Statistical analysis by Kruskal-Wallis test with Benjamini-Hochberg (BH) correction. (b) Left: Correlation between mean HMGA1 transcript expression and percentage of HMGA1-positive cells from healthy controls and sAML patients (GSE185381). Right: Bar plot comparing mean HMGA1 expression in total BMMCs between healthy controls (n = 10) and sAML patients (n = 3). Pearson correlation coefficient ( R ) and P -value shown. Two-sample t -test with BH correction. (c) Log2-transformed HMGA1 expression in CD34 + peripheral blood mononuclear cells (PBMCs) from patients with chronic-phase MPN (ET, n = 2; PV, n = 1; MF, n = 5), derived from dataset GSE189979. Wald test with BH correction. (d) Relative HMGA1 mRNA expression, determined by qRT-PCR, in (i) PBMCs and (ii) CD34 + cells from MPN patients (stratified by JAK2 mutation status: JAK2 wile-type PBMCs, n = 21; JAK2 V617F PBMCs, n = 23; JAK2 wile-type CD34 + , n = 20; JAK2 V617F CD34 + , n = 16) and sAML patients (stratified by JAK2 mutation status: JAK2 wile-type sAML PBMCs, n = 4; JAK2 V617F PBMCs, n = 7; JAK2 wile-type CD34 + , n = 4; JAK2 V617F CD34 + , n = 5). Data are presented as mean ± SD. One-way ANOVA. (e) Relative HMGA1 mRNA expression, determined by qRT-PCR, in (i) PBMCs from sAML patients (stratified by TP53 mutation status: TP53 wild-type, n = 6; TP53 mutated, n = 5) and (ii) CD34 + cells from sAML patients ( TP53 wild-type, n = 5; TP53 mutated, n = 4). Data are presented as mean ± SD. Two-sample t -test. (f) Expression trajectory plot depicting protein expression of HMGA1 and key stem/progenitor markers (CD34, PROM1 /CD133, KIT /CD117, FUT4 /CD15, CD82) across diverse hematopoietic cell subsets, ordered by developmental hierarchy, within (i) total BMMCs and (ii) the malignant sAML cell compartment (GSE185381). (g) Violin plots comparing (i) transcript levels (CITE-seq) and (ii) surface protein expression (ADT) of CD34, PROM1 (CD133), KIT (CD117), FUT4 (CD15), and CD82 between malignant and non-malignant BMMCs from healthy controls (n = 10) and sAML patients (n = 3) (GSE185381). Wald test with BH correction.

Article Snippet: Lentiviral vectors for overexpression of human HMGA1 (NM_145899.3), murine Hmga1 (NM_016660.3), murin e Jak2 WT (NM_008413.4), or murin e Jak2 V617F were generated by cloning the respective open reading frames (ORFs) into the pLenti-CMV-Puro vector (Addgene plasmid #39481, a gift from Ie-Ming Shih).

Techniques: Expressing, Transformation Assay, Derivative Assay, Quantitative RT-PCR, Mutagenesis

Journal: bioRxiv

Article Title: Targeting HMGA1-driven leukemic transformation in myeloproliferative neoplasms with pacritinib

doi: 10.1101/2025.06.01.657170

Figure Lengend Snippet: (a) UMAP visualization of 9,191 Lin - CD34 + cells from 14 sAML patient samples (GSE226340), showing pseudotime trajectories. Cells are colored by (left to right): differentiation trajectory score (LSC-like vs. erythroid), HMGA1 expression level, and TP53 mutational status (wild type, single-hit, multi-hit). Arrows indicate inferred differentiation direction. (b) Violin plots showing HMGA1 transcript expression in CD34 + HSPCs from sAML patients (GSE226340), stratified by TP53 mutational status (wild type, single-hit, multi-hit). (c) Violin plots showing HMGA1 transcript expression in CD34 + HSPCs from sAML patients (GSE226340), stratified by specific combinations of JAK2 (wild-type or V617F) and TP53 (wild-type or mutated) genotype. (d) HMGA1 expression in CD34 + HSPCs from sAML patients (GSE226340) stratified by combinations of TP53 genotype (wild-type or mutated) with other co-occurring gene mutations ( CALR , TET2 , U2AF1 ) versus TP53 mutation alone. (e) HMGA1 expression in BMMCs from sAML patients (GSE185381) stratified by TP53 mutational status (wild type, n = 2 vs. mutated, n = 1). (f) HMGA1 expression in BMMC from patients with sAML (GSE185381) stratified by karyotype complexity (non-complex karyotype, n = 1 vs. complex karyotype, n = 2) (g) Hmga1 transcript expression in murine hematopoietic progenitors from various Jak2 / Trp53 mutant mouse models (GSE180851). (i) Megakaryocyte-erythorid progenitors (MEP) from JAK2 V617F ( J VF ) mice versus J VF mice with different Trp53 genotypes ( Trp53 -/- , Trp53 +/- , Trp53 R172H/- or Trp53 R172H/+ ) (n = 3 per group). (ii) Lin-Sca-1 + c-kit + (LSK) stem cells from the same genotypes. (h) HMGA1 expression in MEPs from mice at sAML stage ( Jak2 V617F Trp53 -/- , n = 3; Jak2 V617F Trp53 R172H/- , n = 3) (GSE180851). (i) Cellular composition and marker expression in representative sAML cases from GSE185381. (i) Stacking bar charts showing the percentage of annotated cell types in two sAML cases (AML3762, AML3948). (ii) Violin plots depicting HMGA1 , CD34 , and KIT transcript expression in the top 5 most abundant malignant cell types and microenvironment (non-malignant) cells for each patient. For panels b-h, statistical significance was assessed by Kruskal-Wallis test with Benjamini-Hochberg (BH) correction (b-d), Wilcoxon rank-sum test with BH correction (e,f), likelihood ratio test (g), or Wald test (h) as appropriate.

Article Snippet: Lentiviral vectors for overexpression of human HMGA1 (NM_145899.3), murine Hmga1 (NM_016660.3), murin e Jak2 WT (NM_008413.4), or murin e Jak2 V617F were generated by cloning the respective open reading frames (ORFs) into the pLenti-CMV-Puro vector (Addgene plasmid #39481, a gift from Ie-Ming Shih).

Techniques: Expressing, Mutagenesis, Marker

Journal: bioRxiv

Article Title: Targeting HMGA1-driven leukemic transformation in myeloproliferative neoplasms with pacritinib

doi: 10.1101/2025.06.01.657170

Figure Lengend Snippet: (a) HMGA1 protein expression (percentage of IHC-positive cells) in bone marrow biopsies from a patient cohort, stratified by sex (female, n = 140; male, n = 67) and age (<60 years, n = 56; ≥60 years, n = 75). Statistical analysis by two-sample t-test. (b-c) Association of HMGA1 protein levels (IHC) with established clinical risk scores in (b) essential thrombocythemia (ET, n = 53) using ELN, MIPSS-ET, IPSET, and Revised-IPSET scoring systems, and (c) polycythemia vera (PV, n = 33) using ELN, MIPSS-PV, and IWG-PV scoring systems. Two-sample t -test or one-way ANOVA, as appropriate. (d-e) HMGA1 protein levels (IHC) in (d) primary myelofibrosis (PMF, n = 20) stratified by IPSS, DIPSS, and MIPSS-70 clinical risk scores, and (e) secondary myelofibrosis (sMF, n = 12) stratified by MYSEC-PM score. One-way ANOVA. (f) Differential HMGA1 protein expression (IHC) between MPN ( JAK2 wild-type, n = 53; JAK2 V617F , n = 88) and sAML ( JAK2 wild-type, n = 9; JAK2 V617F , n = 12). Representative IHC images for JAK2 wild-type MPN and sAML are inset. One-way ANOVA. (g) HMGA1 protein expression (IHC) in sAML patient samples stratified by TP53 mutation status ( TP53 wild-type, n = 14; TP53 mutated, n = 7). A representative IHC image for TP53 wild-type sAML is inset. Two-sample t -test. (h-j) Elevated Hmga1 expression in a Jak2 V617F knock-in MPN mouse model. (h) Representative images of Hematoxylin and Eosin (H&E) staining and IHC for Hmga1 in spleen and bone marrow tissues from Jak2 wild-type (control) and Jak2 V617F mice. Scale bars, 50 µm. (i) Quantification of Hmga1 IHC-positive cells (%) in spleen (n = 6 per group) and bone marrow (n = 4 per group). (j) Relative Hmga1 mRNA levels in PBMCs from Jak2 wild-type (n = 3) and Jak2 V617F (n = 3) mice, determined by qRT-PCR. Data are presented as mean ± SD. Two-sample t -test. (k-l) Expression profiling of HMGA1, CD34, and CD117 in HEL and UKE-1 MPN cell lines. (k) Representative images of Wright-Giemsa staining (left panels) and immunocytochemical (ICC) staining for HMGA1, CD34, and CD117 (right panels). Scale bars, 80 µm (overview), 20 µm (inset). (l) Flow cytometric analysis of intracellular HMGA1 and surface CD34/CD117 expression. (m) Multiplex immunofluorescence (mIF) demonstrating HMGA1 co-expression with stem/progenitor markers in bone marrow biopsies from MPN patients progressing to sMF (representative PV-sMF and ET-sMF cases). Images show H&E staining alongside fluorescence channels for DAPI (nuclei, blue), Vimentin (cyan), HMGA1 (yellow), CD34 (green), and CD117 (red). Scale bars, 80 µm (overview), 20 µm (inset). (n) Flow cytometric analysis revealing HMGA1, CD34, and CD117 expression within CD45 + blasts from an sAML patient’s peripheral blood (PB). (o) Confocal microscopy demonstrating co-localization of HMGA1 (green) with CD34 (red) and CD117 (yellow) in bone marrow (BM) cells from an sAML patient. Nuclei are counterstained with DAPI (blue). Scale bar, 10 µm.

Article Snippet: Lentiviral vectors for overexpression of human HMGA1 (NM_145899.3), murine Hmga1 (NM_016660.3), murin e Jak2 WT (NM_008413.4), or murin e Jak2 V617F were generated by cloning the respective open reading frames (ORFs) into the pLenti-CMV-Puro vector (Addgene plasmid #39481, a gift from Ie-Ming Shih).

Techniques: Expressing, Mutagenesis, Knock-In, Staining, Control, Quantitative RT-PCR, Multiplex Assay, Immunofluorescence, Fluorescence, Confocal Microscopy

Journal: bioRxiv

Article Title: Targeting HMGA1-driven leukemic transformation in myeloproliferative neoplasms with pacritinib

doi: 10.1101/2025.06.01.657170

Figure Lengend Snippet: (a) Efficient shRNA-mediated knockdown of HMGA1 (sh1, sh2 vs. sh-NC control) in HEL and UKE-1 cells. Left: Relative HMGA1 mRNA levels by qRT-PCR (mean ± SD, n = 3). Right: Western blot analysis of HMGA1 protein; ACTB served as loading control. (b) Lentiviral-mediated overexpression of HMGA1 (OE vs. CMV-NC control) in HEL and UKE-1 cells. Left: Relative HMGA1 mRNA levels by qRT-PCR (mean ± SD, n = 3). Right: Western blot analysis of HMGA1 protein; Tubulin served as loading control. (c) Lentiviral-mediated overexpression of Hmga1 (J/OE vs. J/NC control) in murine Ba/F3 ( Jak2 wild type, or Jak2 V617F ) and 32D-cl3 ( Jak2 wild type, or Jak2 V617F ) cells. Left: Relative Hmga1 mRNA levels by qRT-PCR (mean ± SD, n = 3). Right: Western blot analysis of Hmga1 protein; Tubulin served as loading control. Statistical analyses for (a-c) by two-sample t-test or one-way ANOVA, as appropriate. (d) HMGA1 overexpression exacerbates disease phenotype in a HEL xenograft model. Hematological parameters (WBC, white blood cell count; HGB, hemoglobin; HCT, hematocrit; PLT, platelet count) in NSG mice engrafted with HEL cells stably expressing control vector (CMV-NC, n = 6) or HMGA1 (OE, n = 6) at 35 days post-transplantation. Data are presented as mean ± SD. Two-sample t -test. (e) HMGA1 knockdown alters chromatin accessibility and HMGA1 binding at key cell cycle regulatory gene loci. Integrative Genomics Viewer (IGV) snapshots displaying ATAC-seq and HMGA1 CUT&Tag signals at representative E2F target genes ( E2F1 , CCNE1 , CCNE2 , CDK2 , RB1 ), G2M checkpoint genes ( CCNB1 , CCNB2 , CDC2 , WEE1 , CDC25C , PLK1 , AURKA , AURKB ), and common cell cycle regulators ( CCNA2 , CDKN1A / p21 , CDKN1B / p27 ) in HEL cells following control (NC) versus HMGA1 knockdown (KD). (f) Enhanced E2F target and G2M checkpoint gene signatures in sAML patient cells. UMAP projections of single-cell CITE-seq data (GSE185381) from control and sAML patients, with cells colored by enrichment scores for E2F target and G2M checkpoint gene sets. Corresponding density plots illustrate score distributions.

Article Snippet: Lentiviral vectors for overexpression of human HMGA1 (NM_145899.3), murine Hmga1 (NM_016660.3), murin e Jak2 WT (NM_008413.4), or murin e Jak2 V617F were generated by cloning the respective open reading frames (ORFs) into the pLenti-CMV-Puro vector (Addgene plasmid #39481, a gift from Ie-Ming Shih).

Techniques: shRNA, Knockdown, Control, Quantitative RT-PCR, Western Blot, Over Expression, Cell Counting, Stable Transfection, Expressing, Plasmid Preparation, Transplantation Assay, Binding Assay

Journal: bioRxiv

Article Title: Targeting HMGA1-driven leukemic transformation in myeloproliferative neoplasms with pacritinib

doi: 10.1101/2025.06.01.657170

Figure Lengend Snippet: (a) Colony-forming units (CFU) assays in human (HEL, UKE-1) and murine (Ba/F3- JAK2 V617F , 32D- JAK2 V617F ) cell lines following HMGA1 knockdown (shHMGA1) or overexpression (OE-HMGA1) relative to respective controls (NC). Data are mean ± SD. (n=5 per group). Representative colony images are shown (scale bar 300 μm).Two-sample t -test and one-way ANOVA with Tukey’s post-hoc test. (b) Schematic of the HEL cell xenograft experiment in NSG mice, comparing HMGA1-OE cells to vector control cells. (c) Representative bioluminescence images (left panel) and quantification of total photon flux (right panel) from NSG mice at day 35 post-engraftment with luciferase-tagged HEL cells transduced with vector control (CMV-NC, n=6) or HMGA1 overexpression constructs (OE, n=6). Data are mean ± SD. Two-sample t -test. (d) Correlation between HMGA1 mRNA expression (snRNA-seq) and the number of open chromatin peaks (scATAC-seq) in CD34 + cells from a patient during primary MF and after sAML transformation (GSE221946). Pearson correlation coefficient ( R ) and P -value from Wilcoxon ranksum test with BH correction. (e) Genomic distribution of HMGA1 binding sites lost upon HMGA1 knockdown in HEL cells, as determined by CUT&Tag (i), and chromatin regions with significantly reduced accessibility upon HMGA1 knockdown, as determined by ATAC-seq (ii). Stacked bar chats show the percentage of peaks located in promoter (±3kb of TSS), intronic, intergenic, and other genomic regions. (f) GSEA enrichment plots for E2F targets and G2M checkpoint Hallmark gene sets. Comparisons shown in figure. Normalized Enrichment Score (NES) and False Discovery Rate (FDR) q-value are indicated. (g) Cell cycle distribution (Propidium Iodide staining followed by flow cytometry) of HEL and UKE-1 cells with HMGA1 knockdown (sh1, sh2) or non-targeting control (NC). Data are mean ± SD. (n = 5 per group). One-way ANOVA with Tukey’s post-hoc test. (h) Western blots analysis of key E2F pathway, G2M checkpoint, and common cell cycle regulatory protein levels in HEL and UKE-1 cells following overexpression (OE vs. CMV-NC vector control; left panels) or HMGA1 knockdown (shHMGA1 vs. shNC control; right panels). GAPDH served as loading controls. Blots are representative of at least three independent experiments.

Article Snippet: Lentiviral vectors for overexpression of human HMGA1 (NM_145899.3), murine Hmga1 (NM_016660.3), murin e Jak2 WT (NM_008413.4), or murin e Jak2 V617F were generated by cloning the respective open reading frames (ORFs) into the pLenti-CMV-Puro vector (Addgene plasmid #39481, a gift from Ie-Ming Shih).

Techniques: Knockdown, Over Expression, Plasmid Preparation, Control, Luciferase, Transduction, Construct, Expressing, Transformation Assay, Binding Assay, Staining, Flow Cytometry, Western Blot

Journal: bioRxiv

Article Title: Targeting HMGA1-driven leukemic transformation in myeloproliferative neoplasms with pacritinib

doi: 10.1101/2025.06.01.657170

Figure Lengend Snippet: (a) Relative proliferation curves of human (HEL, UKE-1) and murine (Ba/F3, 32D-cl3 transduced with Jak2 wild-type or Jak2 V617F ) cell lines following HMGA1/Hmga1 overexpression (OE) or shRNA-meidated knockdown (sh1, sh2) compared to respective controls (CMV-NC or sh-NC)NC.) 32D-cl3 cells were cultured with IL-3. Data are mean ± SD. (n = 5 per group). Two-way ANOVA. (b) Flow cytometric analysis of CD11b expression on 32D-cl3 cells transduced with Jak2 wild-type (J WT ) or Jak2 V617F (J VF ), and co-transduced with control vector (NC) or HMGA1 overexpression (OE), following G-CSF (100 ng/mL) induced differentiation. (i) Representative histograms of CD11b-FITC fluorescence. (ii) Quantification of HMGA1-PE mean fluorescence intensity (MFI). (iii) Quantification of CD11b-FITC MFI (n = 5 per group). Data are mean ± SD. Two-sample t -test. (c) Quantification of human CD45 + CD117 + HEL cells in peripheral blood of NSG mice at day 35 post-transplant, comparing HMGA1-OE versus vector control (CMV-NC) groups (n=6 per group). Data are mean ± SD. Two-sample t -test. (d) Wright-Giemsa stained peripheral blood smears from NSG mice engrafted with HMGA1-OE or CMV-NC HEL cells at day 35. Quantification of HEL cells (% of total nucleated cells) is shown (n = 6 per group). Data are mean ± SD. Two-sample t -test. (e-f) Representative H&E and HMGA1 IHC staining (left panels of e and f, respectively) and quantification of HMGA1-positive cells (%) (right panels fo e and f, respectively) in (e) femur bone marrow and (f) spleen sections from NSG mice engrafted with HMGA1-OE or CMV-NC HEL cells. Scale bars: 50 µm. Data are mean ± SD. Two-sample t -test. (g) Representative images of spleens (left) and relative spleen weights (spleen weight/body weight %, right) from NSG mice at day 35 post-engraftment with HMGA1-OE or CMV-NC HEL cells (n = 6 per group). Data are mean ± SD. Two-sample t -test. (h) Kaplan-Meier survival curves for NSG mice injected with HMGA1-OE ro CMV-NC HEL cells (n = 6 per group). Median survival times are indicated. Log-rank (Mantel-Cox) test. (i) Heatmaps showing HMGA1 binding intensity (CUT&Tag, left) and chromatin accessibility (ATAC-seq, right) centered on transcription start site (TSS ± 3kb) for genes in HEL cells transduced with shNC or shHMGA1. Color scale indicates normalized read counts (Max/Min normalized). (j) Top de novo motifs identified by HOMER analysis within ATAC-seq peak regions that either lose accessibility (left) or gain accessibility (right) upon HMGA1 knockdown in HEL cells. P -value for motif enrichment are indicated. (k) Quantification of apoptosis by Annexin V-APC/7-AAD staining and flow cytometry in HEL and UKE-1 cells after transduction with shNC or HMGA1 shRNAs (sh1, sh2). Representative flow cytometry plots are shown. Data are mean ± SD. (n = 5 per group). One-way ANOVA with Tukey’s post-hoc test.

Article Snippet: Lentiviral vectors for overexpression of human HMGA1 (NM_145899.3), murine Hmga1 (NM_016660.3), murin e Jak2 WT (NM_008413.4), or murin e Jak2 V617F were generated by cloning the respective open reading frames (ORFs) into the pLenti-CMV-Puro vector (Addgene plasmid #39481, a gift from Ie-Ming Shih).

Techniques: Transduction, Over Expression, shRNA, Knockdown, Cell Culture, Expressing, Control, Plasmid Preparation, Fluorescence, Staining, Immunohistochemistry, Injection, Binding Assay, Flow Cytometry

Journal: bioRxiv

Article Title: Targeting HMGA1-driven leukemic transformation in myeloproliferative neoplasms with pacritinib

doi: 10.1101/2025.06.01.657170

Figure Lengend Snippet: (a-f) HMGA1 expression levels modulate sensitivity to diverse therapeutic agents. Dose-response curves showing viability of HEL, UKE-1, Ba/F3 ( Jak2 wild type, or Jak2 V617F ), and 32D-cl3 ( Jak2 wild type, or Jak2 V617F ) cells with engineered HMGA1/Hmga1 expression (OE vs. NC; sh1/sh2 vs. sh-NC) following 72-hour treatment with (a) IFNα, (b) 5-Azacytidine, (c) Decitabine, (d) Cytarabine, (e) Venetoclax, and (f) Hydroxyurea. Calculated IC50 values are shown. Data represent mean ± SD from n = 3 independent experiments. Two-way ANOVA. (g) Ruxolitinib treatment, particularly long-term exposure, alters key signaling and cell cycle protein expression. Western blot analysis of indicated JAK-STAT, E2F pathway, G2M checkpoint, and cell cycle regulatory proteins in HEL and UKE-1 cells treated with vehicle, short-term ruxolitinib (4 hours), or in ruxolitinib-persistent (Rux-P) lines. GAPDH served as loading control. (h-j) HMGA1/Hmga1 expression status influences sensitivity to JAK inhibitors. Dose-response curves assessing viability of (h) UKE-1 cells, (i) Ba/F3 cells (J VF /NC: Jak2 V617F /control vector; J VF /OE: Jak2 V617F /Hmga1 OE; J WT /NC: Jak2 wild-type/control vector; J WT /OE: Jak2 wild-type /Hmga1 OE), and (j) 32D-cl3 cells (similarly engineered) with engineered HMGA1/Hmga1 expression, following treatment with ruxolitinib, fedratinib, pacritinib, or momelotinib. Calculated IC50 values are shown. Data represent mean ± SD from n = 3 independent experiments. Two-way ANOVA. (k) Pacritinib mitigates weight loss in mice bearing HMGA1-overexpressing HEL xenografts. Percent body weight change in NSG mice engrafted with HEL-Luc cells (CMV-NC or HMGA1-OE) and treated with vehicle or pacritinib (100 mg/kg, BID, 14 days). Data are mean ± SD (n = 6 per group). One-way ANOVA. (l) Pacritinib treatment improves hematological parameters in the HMGA1-overexpressing HEL xenograft model. Peripheral blood counts (WBC, HGB, HCT, PLT) in xenografted mice at day 35 endpoint. Data are mean ± SD (n = 6 per group). One-way ANOVA.

Article Snippet: Lentiviral vectors for overexpression of human HMGA1 (NM_145899.3), murine Hmga1 (NM_016660.3), murin e Jak2 WT (NM_008413.4), or murin e Jak2 V617F were generated by cloning the respective open reading frames (ORFs) into the pLenti-CMV-Puro vector (Addgene plasmid #39481, a gift from Ie-Ming Shih).

Techniques: Expressing, Western Blot, Control, Plasmid Preparation