PMC

(A) Sequence alignment of HMHA1 with the typical RhoGAP, p50RhoGAP, and the structurally-related BAR-GAPs, GRAF1 and OPHN1. Green indicates two matching amino acids. Pink indicates three matching amino acids. Purple indicates four matching amino acids. The arginine finger region is indicated with a black bar. (B) 3D model of the protein-protein complex between RhoA and the HMHA1 RhoGAP domain highlighting the catalytic residues (in sticks, colour coding as indicated; P-loop-Switch I-Switch II of RhoA in light green). The homology model for the GAP domain of human HMHA1 is based on the structure of the human p50RhoGAP domain (PDB ID: 1tx4), using Phyre. The position of the HMHA1 GAP domain in the complex with human RhoA (from RhoA⋅GDP⋅AlFx⋅p50RhoGAP; PDB ID: 1tx4) was obtained through its overlay on the p50RhoGAP domain. The RhoGAP domain of GRAF1 from Gallus gallus (PDB ID: 1f7c) was superimposed onto the model of the HMHA1 GAP domain. (C) HMHA1 C1-GAPtail has in vitro GAP activity towards Rac1, Cdc42, and RhoA but not towards Ras (purple bars). p50RhoGAP was used as a positive control (red bars). GTPases or HMHA1 only were used as a control and as a measure for intrinsic nucleotide hydrolysis (yellow bars). Data are mean values of two independent experiments. Error bars indicate SD. (D) HMHA1 GAP activity is inhibited by the N-terminal BAR domain as full-length HMHA1 has no GAP activity while C1-GAPtail, lacking the N-terminal region, shows GAP activity (purple bars). GTPases or HMHA1 only were used as a control and as a measure for intrinsic hydrolysis (yellow bars). Data are mean values of two independent experiments. Error bars indicate SD.

Intracellular localization of myc-tagged HMHA1 (and mutant constructs) was studied by Confocal Laser Scanning Microscopy following expression in HeLa cells. Myc-tagged HMHA1 was detected by immunostaining for the myc epitope in combination with detection of F-Actin with phalloidin. Full-length HMHA1 (FL) as well as HMHA1 N-term are partially localized at membrane ruffles as well as in the cytoplasm. For HMHA1 N-term localization at vesiculo-tubular structures is occasionally observed (arrows). Cells expressing FL or the N-term constructs are morphologically similar to control cells and no effects are seen on F-Actin (upper two rows). HMHA1 constructs lacking the C-terminal tail (GAP and C1-GAP) are partly mislocalized into protein aggregates. In cells expressing HMHA1 C1-GAP, C1-GAPtail, and GAPtail (marked with asteriks), F-Actin distribution is altered and cell morphology is dramatically changed. Higher magnification images of the boxed area are included. Scale bars, 10 µm.

(A-D) Colocalization of myc-tagged HMHA1 with endogenous Rac1 (A), Rac1 Q61L (B), Cdc42 G12V (C) and RhoA V14 (D) was studied by Confocal Laser Scanning Microscopy. Myc-tagged HMHA1 and HMHA-tagged Cdc42 and RhoA were detected by immunostaining in combination with detection of F-Actin. HMHA1 colocalized with endogenous Rac1 (A) and Rac1 Q61L (B) in membrane ruffles (arrows). A partial colocalization of HMHA1 with Cdc42 G12V (C) and RhoA V14 (D) was observed (arrows) although less clear than for Rac1. Higher magnification images of the boxed areas are included. Scale bars, 10 µm.

(A) HeLa cells were transfected as indicated. After 24 hours, a CRIB pull-down assay was performed to measure levels of Rac1GTP. HMHA1 full-length (blue bar) and N-term (purple bar) do not significantly decrease Rac1GTP loading. The N-terminal BAR domain auto-inhibits GAP function as HMHA1 C1-GAP (green bar) and C1-GAPtail (black bar), which lack the N-terminal BAR domain, show a significant decrease in Rac1GTP loading compared to control cells (red bar). Data are mean values of five independent experiments. Error bars indicate SEM. ns, not significant. * p<0.05, ** p<0.01. (B) HeLa cells were transfected as indicated. After 24 hours, a RhoA G-LISA (Cytoskeleton) was used to measure levels of RhoA-GTP. HMHA1 full-length (red bar) and N-term (purple bar) did not significantly decrease RhoA-GTP loading. In contrast, HMHA1 C1-GAPtail (pink bar), which lacks the N-terminal BAR domain, showed a significant decrease in RhoA-GTP loading compared to control cells (red bar). Data are mean values of three independent experiments. Error bars indicate SEM. ns, not significant, * <0.05, *** p<0.001. (C) Jurkat cells were transduced with control (shC) or HMHA1 (shHMHA1 #2 or #3) short hairpin RNAs. After 72 hours, a CRIB pull-down assay was performed to measure levels of Rac1GTP. Knock-down of HMHA1 significantly increases Rac1GTP loading compared to control cells. Data are mean values of two (for shHMHA1 #3) or three (for shHMHA1 #2) independent experiments. Error bars indicate SD. * p<0.05. (D) Jurkat cells transduced as indicated were lysed and RhoA-GTP levels were measured using a RhoA G-LISA. No significant differences in levels of RhoA-GTP loading were observed between control cells and cells treated with shRNAs against HMHA1. Data are mean values of four measurements of two independent experiments. Error bars indicate SD.

(A) Schematic overview of the organization of HMHA1 and of the different constructs used in this study. (B) Morphology of HeLa cells, transfected as indicated, was analyzed by phase contrast microscopy. Cells expressing HMHA1 full-length (FL), GAP, or N-term did not show any changes in morphology compared to control cells. HMHA1 C1-GAP, C1-GAPtail, and GAP-tail induce dramatic changes in cell morphology. In addition, these cells are less adhesive than control cells. Scale bars, 50 µm.

(A) Jurkat T-cells were fixed and immunostained for endogenous HMHA1 and Rac1 and stained for F-actin and DNA. Left panel shows three examples of the distribution of HMHA1, Rac1 and F-actin revealing colocalization of HMHA1 and Rac1 in F-actin rich areas. The nucleus (DNA) and cell morphology (phase image) are included to show the integrity of the cell. Right panel shows intensity distribution of Rac1 (Y-axis) and HMHA1 (X-axis) signals, underscoring the fact that most cells are double positive. (B) Jurkat cells were stimulated for the indicated time-points with 100 ng/ml CXCL12 and analyzed as in A. Two examples of each condition are shown in the left panels. Changes in F-actin distribution in response to CXCL12 can be observed, in particular after 1 and 5 minutes. Right panels show the extent of colocalization (AU, arbitrary units) quantified by the image stream software. Ave, average colocalization, n, number of cells.

The effects of myc-tagged HMHA1 (and deletion constructs) on focal adhesion distribution was studied by Confocal Laser Scanning Microscopy following expression in HeLa cells. Similar to the effects on F-Actin distribution, cells expressing full-length HMHA1 (FL), N-term, or GAP (first, second and fifth rows) constructs show normal focal adhesion distribution as detected using Paxillin immunostainings. Expression of HMHA1 C1-GAP, C1-GAPtail, or GAPtail (marked with asteriks) induces loss of focal adhesions. In the merged images, HMHA1 constructs appear in red, paxillin in green and nuclei in blue. Higher magnification images of the boxed area are included. Scale bars, 10 µm. (B) Mean +/− SD of the average-per-cell paxillin staining intensity (10–20 cells per condition), quantified following background subtraction, is indicated. Statistical differences compared to the Full-Length control are indicated by the respective p-values.

a, b Stable transfectants of HeLa cells with HMHA1 overexpression expression vector or its corresponding empty vector (EV) were subjected to Western blotting for the indicated proteins and to invasion assay under normoxia (20% O₂). Results are shown as mean ± s.d. (n = 3); ***P < 0.001 (Student’s t test). Representative images are shown. c, d HeLa cells (c) and HT1080 cells transfected with siRNA against HMHA1 (#1, #2) or its scramble control (Scr) (d) were used for the invasion assay under normoxia (20% O₂) or hypoxia (1% O₂). Results are shown as mean ± s.d. (n = 3 for c; n = 4 for d); **P < 0.01, ***P < 0.001 (Student’s t test). Representative images are shown. e Viability of HT1080 cells after HMHA1 knockdown treatment as in (d) was assessed. Results are shown as mean ± s.d. (n = 5); ns: not significant (Student’s t test). f, g HeLa cells overexpressing HMHA1 and the corresponding empty vector (EV)-transfected cells were stained with phalloidin (green). Scale bar: 20 µm. Reproducibility was confirmed in three independent set-ups and representative images are shown (f). Percentage of cells with prominent stress fibre structures (stress fibre⁺) in (f) was quantified (g). For each set-up, 5 random fields were studied; number of cells per field ranged between 7 and 21 cells, and at least 50 cells were examined. Results are shown as mean ± s.d. (n = 5); *** P < 0.001 (Student’s t test). h Conditioned media from HT1080 cells transiently transfected with HMHA1 overexpression vector or the corresponding empty vector (EV) were subjected to gelatin zymographic analyses. i qPCR (for MMP2 and MMP9 mRNA) using HeLa cells overexpressing HMHA1 and the corresponding empty vector (EV)-transfected cells. Results are shown as mean ± s.d. (n = 6); ns: not significant (Student’s t test). j HeLa cells overexpressing HMHA1 and the corresponding empty vector (EV)-transfected cells were subjected to invasion assay under normoxia (20% O₂) in the presence of MMP-2/-9 inhibitor (MMPi, 2 µM) or DMSO control. Results are shown as mean ± s.d. (n = 4); * < 0.05, ****P < 0.0001, ns: not significant (Student’s t test). Representative images are shown. k Cell viability after MMP-2/-9 inhibitor (MMPi) treatment as in (j) was assessed. Results are shown as mean ± s.d. (n = 4); ns: not significant (Student’s t test).

a qPCR (for HMHA1, CA9, and VEGFA mRNAs) using HeLa cells cultured with or without actinomycin-D treatment (10 μg/mL) under the indicated oxygen conditions for 24 h. Results are shown as mean ± s.d. (n = 3); ****P < 0.0001 (Student’s t-test). b, c qPCR (for HMHA1 mRNA) and Western blotting (for the indicated proteins) using HeLa cells with or without deferoxamine (DFO, 100 μM) (b) or dimethyloxalylglycine (DMOG, 2 mM) (c) treatment for 24 h at 20% O₂. Results are shown as mean ± s.d. (n = 3); ****P < 0.0001 (Student’s t test). d qPCR (for HMHA1 mRNA) and Western blotting (for the indicated proteins) using parent HeLa cells (WT) or HIF-1β knockout clones (KO#1 and #2) exposed to the indicated oxygen conditions for 24 h. Results are shown as mean ± s.d. (n = 3); ***P < 0.001, ****P < 0.0001 (Student’s t test). e HIF-1β knockout HeLa cells transiently transfected with HIF-1β expression vector or its corresponding empty vector (EV) were cultured under the indicated oxygen conditions for 24 h and subjected to Western blotting (for the indicated proteins). f Same experiments as in (d) were performed using HIF-1α knockout clones. Results are shown as mean ± s.d. (n = 3); ****P < 0.0001 (Student’s t test). g Parent HeLa cells (WT) or HIF-1α knockout HeLa cells stably transfected with HIF-1α expression vector or its corresponding empty vector (EV) were cultured under the indicated oxygen conditions for 24 h and subjected to Western blotting for the indicated proteins. h qPCR (for HMHA1 mRNA) and Western blotting (for the indicated proteins) using parent HeLa cells (WT) and HIF-1α knockout HeLa cells treated with siRNA against HIF-2α or that with negative control (Scr) under the indicated oxygen conditions for 24 h. Results are shown as mean ± s.d. (n = 4); ***P < 0.001, ****P < 0.0001 (Student’s t test). I Same experiments as in (h) were performed using HIF-2α knockout HeLa cells and siRNA against HIF-1α. Results are shown as mean ± s.d. (n = 4); ****P < 0.0001 (Student’s t test).

Cell spreading was measured by Electrical Cell-substrate Impedance Sensing (ECIS) following seeding of 100.000 cells on fibronectin-coated electrodes. Left panel: A significant decrease in electrical resistance, as a measure of cell spreading, was observed in HeLa cells expressing HMHA1 C1-GAPtail (black), C1-GAP (light green), and GAPtail (grey) compared to control cells (red). Ectopic expression of HMHA1 full-length (blue), N-term (dark green), and GAP (magenta) did not affect cell spreading. Right panel: Relative cell spreading at 60 minutes post-seeding. Data are mean values of three independent experiments. Error bars indicate SEM. ns, not significant, ** p<0.01, *** p<0.001.

a HeLa cells were cultured at normoxic (N, 20% O₂) or severely hypoxic (H, < 0.1% O₂) condition for 24 h, followed by genome-wide DNA microarray and RNA sequencing (RNA-Seq) analyses. Differentially expressed genes with induction ratio > 5.0 were identified and the Venn diagram shows the overlapping candidate genes. b, c qPCR (for HMHA1 and CA9 mRNA in (b) and HMHA1 mRNA in (c) and Western blotting (for the indicated proteins) using HeLa cells (b) and the indicated cells (c) cultured under the indicated oxygen conditions for 24 h. Results are shown as mean ± s.d. (n = 3); ns: not significant, **P < 0.01, ****P < 0.0001 (Student’s t-test). d qPCR (for HMHA1 and CA9 mRNA) using HeLa tumour xenografts from untreated or phenylhydrazine-treated mice. Results are shown as mean ± s.d. (n = 6); *P < 0.05, **P < 0.01 (Student’s t-test). e FFPE sections of HeLa tumour xenografts were double-stained with antibodies against the hypoxia marker, pimonidazole (green), and HMHA1 (red). Nucleus was counterstained with Hoechst 33342 (blue). Scale bar: 50 µm. *: blood vessel. Reproducibility was confirmed in xenografted tumours from three independent mice and representative images are shown.

a, Mutations of key residues in the canonical CDAN1 B-domain (BD_A2 mutated to B×) and H3 mimic helix (HMH_A1 mutated to H×) (top) and scheme of ASF1A and ASF1B N-terminal domain (NTD) swaps analyzed. b, ASF1B interaction with CDAN1 is more reliant on HMH_A1 than ASF1A. Flp-In 293 T-REx cells transiently transfected with Strep-tagged CDAN1 (ST-CDAN1) variants without or with mutations in BD_A2 (B×) and/or HMH_A1 (H×) were lysed (input) and subjected to Strep-Tactin pulldowns (PD) before analysis by SDS-PAGE and immunoblotting; representative of 3 independent replicates. c, The indicated ST-CDAN1 variants were co-transfected with FLAG-tagged ASF1A (F-ASF1A, left) or ASF1B (F-ASF1B, right) with the indicated N-terminal domain (NTD). Cell lysates before (input) or after Strep PD were analyzed by SDS-PAGE and immunoblotting; representative of 2 independent replicates.

(A,B) Rescue experiments with constitutively active Rac1 Q61L and G12V (A) or Cdc42 G12V (B), co-expressed with the HMHA1 C1-GAPtail protein were performed in HeLa cells and analyzed by Confocal Laser Scanning Microscopy. Ectopically expressed proteins were visualized in combination with F-Actin. (A) Constitutively active Rac1 Q61L (middle panels) and G12V (bottom panels) were able to (partially) rescue the phenotype induced by C1-GAPtail. As a control, mCherry empty vector (EV; upper panels) was unable to rescue the phenotype. (B) Ectopic expression of constitutively active Cdc42 G12V did not rescue the phenotype induced by C1-GAPtail. Scale bars, 10 µm.

The prognostic value of VAVs expression level in AML (GEPIA and PrognoScan). (A) The prognostic value of VAVs expression level in AML, which was analyzed by GEPIA. (B) The prognostic value of VAVs expression level in AML, which was analyzed by PrognoScan. (C) Prognostic analysis of SIPA1, SH2D3C， and HMHA1 in AML (LinkedOmics).

a HeLa cells incubated under the indicated oxygen conditions for 24 h were irradiated with the indicated doses of γ-ray, reoxygenated, further cultured for 6 h, and subjected to qPCR analysis for HMHA1 mRNA. Results are shown as mean ± s.d. (n = 4); **P < 0.01 (Student’s t test). b, c Same experiments as (a) were performed using HIF-1β knockout HeLa cells (b), or parent HeLa cells (WT) in the presence or absence of NAC (5 μM) (c). Results are shown as mean ± s.d. (n = 4); ns: not significant, ****P < 0.0001 (Student’s t test). d Parent HeLa cells (WT) and two clones of HMHA1 knockout HeLa cells (KO#1 and #2) were cultured under the indicated oxygen conditions for 24 h and subjected to Western blotting for the indicated proteins. e Parent HeLa cells (WT) and HMHA1 knockout clones (KO#1 and #2) were irradiated under hypoxic conditions (< 0.1% O₂) with the indicated doses of γ-ray, reoxygenated, and subjected to the Boyden chamber invasion assay. Results are shown as mean ± s.d. (n = 3); *P < 0.05 (Student’s t test). f Cell viability after irradiation-reoxygenation treatment as in (e) was assessed. Results are shown as mean ± s.d. (n = 6); ns: not significant (Student’s t test).

Gene correlation expression analysis of VAVs (LinkedOmics). (A) The scatter plot shows the Pearson correlation between VAV1 expression and SIPA1, SH2D3C and HMHA1 expression. (B) The scatter plot shows the Pearson correlation between the expression of VAV2 and SRGAP2, MAPK7 and RELL2. (C) Scatter plot showing the Pearson correlation between VAV3 expression and CNST, GUCY1A3 and SLC9A7 expression.

a HMHA1 expression levels in normoxic (with low hypoxia-signature) and hypoxic (with high hypoxia-signature) tumours. Results are shown as mean ± s.d. (The numbers of sample are indicated below each graph.); *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 (Student’s t test). b TCGA-based Kaplan-Meier analysis of overall survival of patients with ovarian cancer, stomach cancer, bladder cancer, glioblastoma (GBM), and low-grade glioma stratified by the intratumoral oxygen environments (normoxia: low hypoxia signature; hypoxia: high hypoxia signature). The censored cases are shown as ticks on the line. P-values, hazard ratios, and confidence intervals (95% CI) are indicated in each graph (Log-rank test).

(a,b) Adjustment improves recovered signals. Integrated signal intensity for each gene in each cell (individual dots) from direct measurements (x axis) and from estimates recovered by the autoencoder decompressed images (y axis) either before (a) and after (b) co-measurement correction. (c) Example correction. Segmented cell intensities before (left) and after (right) correction for two co-measured genes (Hmha1 and Slc17a7) that were not correlated in snRNA-Seq.

a,b, Structural models of CDAN1 B-domain interactions with a, ASF1A-1 (BD_A1) and b, ASF1A-2 (BD_A2), colored as in . c, Sequence alignments of CDAN1 B-domains (BD, top) and H3 mimic helices (HMH, bottom) with other ASF1 interactors or histone H3. CDAN1 sequences are colored according to Consurf conservation values as indicated. d, Structural model of the CDAN1 HMH interaction with ASF1A-1 (HMH_A1) superposed with the corresponding X. laevis histone H3 helix (transparent gray; residues 121-136, PDB 2IO5). e, Alphafold3 model of the putative CDAN1 HMH interaction with ASF1A-2 (HMH_A2) superposed with the H3 helix (transparent gray).

The perivascular infiltrates seen in iPAH lung sections were highly positive for CD8 ((a) 10× objective, (b) 40×). Scattered cells in the perivascular infiltrates stained positive for CD45RA ((c) and (d)). CD45RA is a protein tyrosine phosphatase and is believed to be a marker for cytolytic T lymphocytes. HMHA1 in perivascular infiltrates in APAH SSC lung sample ((e) 10×, (f) 40× objective) vs. control sample ((g) 10×, (h) 40× objective). Cells surrounding a pulmonary arteriole were highly positive for HMHA-1 in a PAH lung sample.