Characterization of dipyridamole as a novel ferroptosis inhibitor and its therapeutic potential in acute respiratory distress syndrome management

Abstract

Rationale: Ferroptosis in lung epithelium and endothelium contributes to the pathogenesis of acute respiratory distress syndrome (ARDS), a critical and often fatal condition marked by acute inflammation and elevated pulmonary vascular permeability. Despite this, there are currently no FDA-approved therapeutics specifically targeting ferroptosis for ARDS management. Methods: A screening of 259 FDA-approved drugs was conducted to identify an effective ferroptosis inhibitor in pulmonary epithelial and endothelial cells. The anti-ferroptotic and therapeutic efficacy of this screened drug was rigorously evaluated using two distinct ARDS mouse models (LPS-induced acute lung injury and CLP-induced sepsis) and human airway organoids (hAOs). The regulatory mechanism of this drug on ferroptosis inhibition was investigated via RNA-sequencing, qRT-PCR, western blotting, IF, luciferase reporter assay, chromatin immunoprecipitation assay, limited proteolysis-mass spectrometry assay, cellular thermal shift assay, and drug affinity responsive target stability assay. Furthermore, a proof-of-concept clinical trial was conducted, wherein ARDS patients were administered with the drug as adjunctive therapy. Results: Dipyridamole (DIPY) was identified as a potent inhibitor of ferroptosis in pulmonary epithelial and endothelial cells. DIPY effectively mitigated ferroptosis and pulmonary damage in both mouse models and hAOs, primarily by downregulating heme oxygenase 1 (HMOX1). The transcription factor cAMP responsive element binding protein 1 (CREB1) was identified as a key transactivator of HMOX1, which DIPY effectively downregulated. Mechanistically, DIPY binds to and activates superoxide dismutase 1 (SOD1), which in turn inhibits the CREB1/HMOX1 pathway, thereby suppressing ferroptosis. Notably, the clinical trial further corroborated the therapeutic potential of DIPY in ARDS patients, demonstrating improved outcomes with DIPY adjunctive therapy. Conclusions: These findings provide compelling evidence that DIPY inhibits ferroptosis in pulmonary epithelial and endothelial cells by modulating the SOD1/CREB1/HMOX1 signaling axis and suggest DIPY as a promising therapeutic strategy for ARDS treatment.

Keywords: acute respiratory distress syndrome; cAMP responsive element binding protein 1; dipyridamole; ferroptosis; heme oxygenase 1.

Figure 1

The FDA-approved drug screening identifies DIPY as a new ferroptosis inhibitor in lung epithelial and endothelial cells. (A) Schematic diagram of drug screening. A549 cells were treated with RSL3 (5 µM) and 259 FDA-approved drugs for 8 hours, followed by CCK-8 assays. (B) Scatter plot showing the viability of A549 cells treated with various compounds from the FDA-approved drug library in the presence of RSL3 normalized to control (left panel). The top five candidate compounds with the highest viability percentages are identified and listed in the table (right panel). (C) A549, HUVEC, and BEAS-2B cells were treated with Ciclopirox, Valbenazine tosylate, DIPY, Lansoprazole, Asenapine maleate, or DMSO, together with RSL3 (5 µM) for 8 hours. Cell viability was assessed and the results were presented as a heat map. (D) The chemical structure of DIPY is shown. (E and F) A549 and HUVEC cells were treated with RSL3, Erastin, STS, or Rapa, together with DIPY (10 µM for A549; 5 µM for HUVEC) or DMSO for indicated times (RSL3 and Erastin for 8 hours; STS for 6 hours; Rapa for 48 hours). The cell viability of A549 cells was measured (E). Representative phase-contrast images of cells were captured (F). Scale bar, 100 μm. (G-I) A549 cells were treated with RSL3 (5 µM), DIPY (10 µM), and RSL3 (5 µM) plus DIPY (10 µM) for 8 hours. HUVEC cells were treated with RSL3 (0.25 µM), DIPY (5 µM), and RSL3 (0.25 µM) plus DIPY (5 µM) for 8 hours. L-ROS levels were assessed using C11-BODIPY (G). MDA contents and Fe²⁺ levels were measured (H). Representative TEM images of cells were taken, with red arrows indicating mitochondria (I). Scale bar, 500 nm. Results are shown as mean ± SD from 3 independent experiments. Statistical significance is indicated as *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

Figure 2

DIPY inhibits ferroptosis in LPS-induced ARDS mouse model. (A) Scheme of the experimental procedure for the LPS-induced ARDS mouse model. DIPY treatment (10 mg/kg, intraperitoneal) was administered 1 hour before the LPS challenge (low dose: 2.5 mg/kg; lethal dose: 20 mg/kg). After 24 hours, the BALF and the lungs of the mice (treated with low-dose LPS) were collected and analyzed. (B) Percent survival after administering lethal doses of LPS instillation is shown (n = 12 per group). (C-J) Low-dose LPS-challenged ARDS mice were pretreated with or without DIPY (n = 6 per group). Representative CT images of mouse lungs (C). Representative images of H&E-stained lung sections, with black arrows indicating infiltrating inflammatory cells, interstitial edema, and alveolar wall thickening (D). Scale bar, 50 μm. Lung injury scores (E). Measurements of lung wet/dry ratio and total protein in BALF (F). Total cell counts and neutrophil percentage in BALF (G). TNF-α and IL-1β expressions in BALF were measured by ELISA (H). MDA contents (upper panel) and Fe²⁺ levels (lower panel) in mouse lung tissues (I). Representative TEM images illustrating mitochondria in the alveolar epithelium, vascular endothelium, and airway epithelium, with black arrows indicating mitochondria (J). Scale bar, 500 nm. Results are shown as mean ± SD. Statistical significance is indicated as *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

Figure 3

DIPY inhibits ferroptosis in hAOs. (A) Workflow for establishing and identifying hAOs from solid lung tissue. (B) Brightfield images of cultured hAOs derived from patients. Scale bar, 100 μm (upper panel) and 50 μm (lower panel). (C) Representative images of H&E-stained sections of hAOs, showing basal and multi-ciliated cells. Scale bar, 20 μm. (D) TEM images reveal detailed longitudinal and cross-sections of the cilia in the organoids, demonstrating the typical microtubule arrangement. Scale bar, 1 μm (left panel) and 250 nm (right panel). (E) Identification of the mucosecretory and ciliated cells in the hAOs by IF assays. Scale bar, 10 μm. (F-I) The hAOs were treated with RSL3 (30 μM), DIPY (20 μM), and RSL3 (30 μM) with DIPY (20 μM) for 24 hours (n = 5 per group). Morphological changes of hAOs (F). Scale bar, 250 μm (upper panel) and 50 μm (lower panel). Assessment of organoid viability (G). Measurement of MDA contents and Fe²⁺ levels (H). TEM images illustrate representative mitochondria (black arrows) in the hAOs (I). Scale bar, 2 μm. Results are shown as mean ± SD. Statistical significance is indicated as **P < 0.01; ***P < 0.001; ****P < 0.0001.

Figure 4

HMOX1 is a target of DIPY. (A) Volcano plots of RNA-seq analysis (n = 3 per group). (B) GSEA analysis of RNA-seq data. (C) Heat maps of the top 20 DEGs in both comparative groups. (D and E) A549 and HUVEC cells were treated with RSL3, DIPY, and RSL3 with DIPY for 8 hours as shown in Figure 1 (n = 3 per group). The expression of HMOX1 was detected by western blot (left panel of D), qRT-PCR (right panel of D), and IF staining (E). Scale bar, 10 μm. (F) hAOs were treated with RSL3, DIPY, and RSL3 with DIPY for 24 hours as shown in Figure 3, and the protein levels of HMOX1 were analyzed by western blot (n = 5 per group). (G and H) The expression of HMOX1 in LPS-induced ALI mouse models with or without DIPY treatment was detected by western blot (G) and IHC (H) assays (n = 6 per group). Scale bar, 50 μm. Results are shown as mean ± SD of 3-6 independent experiments. Statistical significance is indicated as ***P < 0.001; ****P < 0.0001.

Figure 5

DIPY inhibits ferroptosis through suppression of HMOX1. (A-E) A549 and HUVEC cells were infected with lentivirus expressing shRNAs against HMOX1 (shHMOX1^-1 and shHMOX1^-2) or a scrambled sequence (shSc). These cells were stimulated with RSL3 (5 μM for A549; 0.25 μM for HUVEC) for 8 hours (n = 3 per group). HMOX1 expression was analyzed by western blot (upper panel of A) and qRT-PCR (lower panel of A). Cell viability assessment (B). Representative phase-contrast images of cells (C). Scale bar, 100 μm. L-ROS levels measurement (D). MDA contents and Fe²⁺ levels (E). (F-H) The hAOs were infected with lentivirus harboring a vector encoding HMOX1 (Lv-HMOX1) or the empty vector (Lv). These hAOs were treated by RSL3 (30 μM) with or without DIPY (20 μM; n = 5 per group). Representative phase-contrast images of hAOs (F). Scale bar, 250 μm (upper panel) and 50 μm (lower panel). Viability assessment of organoids (G). MDA contents and Fe²⁺ levels (H). (I-M) C57/BL6 mice were administered HMOX1-overexpressed lentivirus (Lv-HMOX1) or the empty vector (Lv) via intratracheal instillation for 7 days. On day 8, the indicated mice were challenged with LPS (2.5 mg/kg) with or without DIPY pretreatment (n = 6 per group). Representative images of H&E-stained lung sections (I). Scale bar, 50 μm. Lung injury scores (J). IHC staining of HMOX1 in mice lung sections (K). Scale bar, 50 μm. ELISA analysis of TNF-α and IL-1β expressions in BALF (L). MDA and Fe²⁺ levels in lung tissues (M). Results are shown as mean ± SD of 3-6 independent experiments. Statistical significance is indicated as *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

Figure 6

DIPY suppresses HMOX1 through the deactivation of CREB1. (A) Schematic illustration for screening TFs for the human HMOX1 gene. (B) A549 and HUVEC cells were treated with RSL3, DIPY, and RSL3 with DIPY for 8 hours as shown in Figure 1. Cell lysates were collected and analyzed with the indicated antibodies (n = 3 per group). (C) hAOs were treated with LPS, DIPY, and LPS with DIPY as shown in Figure S3 (n = 5 per group). The lysates were subjected to immunoblotting with the indicated antibodies. (D-E) Western blot (D) and IHC assays (E) examined the expression of p-CREB1 in LPS-induced ALI mouse models with or without DIPY treatment as shown in Figure 2 (n = 6 per group). Scale bar, 50 μm. (F-K) CREB1 knockout A549 cell lines (sgCREB1) and the control cells (sgCtrl) were treated with RSL3 (5 μM) for 8 hours (n = 3 per group). Confirmation of CREB1 knockout via sequencing (F). Cell lysates were subjected to immunoblotting with the indicated antibodies (G). Cell viability assessment (H). Representative phase-contrast images of cells (I). Scale bar, 100 μm. L-ROS levels measurement (J). MDA contents and Fe²⁺ levels (K). (L-N) sgCREB1 A549 cells or the control cells were infected with lentiviruses carrying Lv-HMOX1 or Lv. The cells were treated with RSL3 (5 μM) for 8 hours (n = 3 per group). Cell lysates were subjected to immunoblotting with the indicated antibodies (L). L-ROS levels (M). MDA contents and Fe²⁺ levels (N). Results are shown as mean ± SD of 3-6 independent experiments. Statistical significance is indicated as **P < 0.01; ***P < 0.001; ****P < 0.0001.

Figure 7

DIPY inhibits the CREB1/HMOX1 pathway through directly binding to and activating SOD1. (A) Schematic diagram illustrating the LiP-MS assay used to identify the direct target of DIPY. A549 cells were pretreated with RSL3 (5 μM) for 8 hours. The lysates were collected and incubated with DIPY, followed by digestion with proteinase K for 10 minutes, and then analyzed using MS. (B) MS analysis showing the fragmentation pattern of SOD1 peptide. The spectrum includes the identification of b-ions (green) and y-ions (orange) corresponding to the amino acid sequence of the SOD1 protein. (C and D) A549 cells were treated with RSL3 (5 μM) for 8 hours. The harvested cell lysates were incubated with DIPY (10 μM) or DMSO for 30 minutes, followed by CETSA (C) and DARTS (D) analysis. (E) Three-dimensional binding mode diagrams between DIPY and SOD1. (F) SPR analysis of DIPY-binding to SOD1. K_D = 8.83 μM. (G) A549 and HUVEC cells were treated with RSL3 (5 μM) and DIPY (0, 5 μM, 10 μM). The enzymatic activity of SOD1 was examined. (H-L) A549 cells were infected with lentivirus expressing shRNAs target SOD1 (shSOD1^-1 and shSOD1^-2) or a scrambled sequence (shSc). The cells were incubated with RSL3 (5 μM), with or without DIPY (10 μM), for 8 hours. Cell lysates were subjected to immunoblotting with the indicated antibodies (H). The level of HMOX1 mRNA was analyzed using qRT-PCR (I). Cell viability assessment (J). L-ROS levels measurement (K). MDA contents and Fe²⁺ levels (L). (M) Schematic illustration of how DIPY suppresses ferroptosis in pulmonary epithelial and endothelial cells by regulating the SOD1/CREB1/HMOX1 pathway in ARDS. Results are shown as mean ± SD of at least 3 independent experiments. Statistical significance is indicated as *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

Figure 8

Clinical evaluation of DIPY in patients with ARDS indicates beneficial effects. (A) Clinical study flow chart. Peripheral blood sampling was performed at baseline before the first dose of DIPY and 1, 2, 3, and 7 days after the initiation of DIPY adjunctive treatment, respectively. Patients were monitored for 14 days after completion of DIPY treatment. (B-D) Clinical parameters post-DIPY treatment. SOFA score at indicated time points after DIPY treatment, normalized to day 0 (pre-DIPY treatment, B). PiO₂/FiO₂ ratio at indicated time points after DIPY treatment, normalized to day 0 (C). CRP levels at indicated time points after DIPY treatment, normalized to day 0 (D). (E) Chest computed tomography scans of two subjects (#1 and #2) before DIPY treatment (Day 0) and 10 days after DIPY treatment (Day 10). (F-J) The mRNA levels of indicated injury markers (F for AGER, G for SFTPD, H for ICAM1, and I for vWF) and HMOX1 (J) in plasma from ARDS patients at indicated time points before and after DIPY treatment were analyzed using qRT-PCR. (K) MDA contents at indicated time points after DIPY treatment, normalized to day 0. (B-D, F-K) Data is shown with individual patient responses as dotted lines and the mean of all subjects as solid lines.