Targeting myeloid cell coagulation signaling blocks MAP kinase/TGF-β1–driven fibrotic remodeling in ischemic heart failure

Figures

• 

  Figure 1. The proteomic signature of human IHF. Proteins isolated from cardiac tissue from human IHF (n = 5) and non-IHF donor hearts (controls, n = 5) were analyzed by label-free quantitative proteomics. Samples were measured in quadruplicate (proteins) and triplicate (phosphopeptides) liquid chromatography–tandem mass spectrometry (LC-MS/MS) runs. Heatmaps show the quantity profile for 208 proteins and 685 phosphopeptides significantly differing (P < 0.05, fold change > 2 or < 0.5, and identified in >60% of all LC-MS/MS replicates in either control or IHF group), log10-transformed and row-wise normalized using z score and sorted by descending fold change.

• 

  Figure 2. Enrichment of innate immune and coagulation pathways and of MAPK1 signaling in human IHF. (A) Reactome pathway enrichment analysis of differentially abundant proteins from Figure 1. (B) Differentially regulated (P < 0.05) nonphosphorylated and phosphorylated proteins were analyzed for their previously known protein-protein interactions between each other by Cytoscape STRING-DB search and sorted by the number of interactions in a circular layout. Among the present kinases, MAPK1 shows the most interactions, followed by mTOR, TTN, CDK13, PI4KA, and TAOK2. (C) Volcano plot depicting significantly differing Integrative Inferred Kinase Activity (InKA) scores between IHF and control groups.

• 

  Figure 3. Infiltrating myeloid cells are the major source for increased MAPK1 activation in a preclinical model of non-reperfused MI. Confocal microscopy of whole LV myocardial cryosections obtained from n = 3–5 C57BL/6J mice after 7 days of permanent LAD ligation. (A) Representative images of p-ERK1/2+ cells costained for CD31, CD45, α-SMA, cTNT. Scale bars: 500 μm. (B) Top: Quantification of p-ERK1/2+ intensity in remote, border, and infarction regions at days 1, 3, and 7 after MI. Bottom: Quantification of colocalization analysis of p-ERK1/2 signal intensity with CD31, CD45, α-SMA, and cTNT in border and infarct regions at day 7 after MI using Pearson’s correlation coefficient. RAU, relative arbitrary units. (C) Experimental design: C57BL/6J mice were subjected to permanent LAD ligation versus sham surgery and given trametinib (1 mg/kg/d) or vehicle treatment once daily via oral gavage from day 1 to day 7. (D) High-frequency ultrasound echocardiography obtained in parasternal long axis with measurement of LV ejection fraction (LVEF, %) and LV end-diastolic volume (LVEDV, μL) on day 7 after operation. Ordinary 1-way ANOVA, Šidák’s multiple-comparison test; n = 4–6 animals per group. Data…

• 

  Figure 4. Inhibition of ERK1/2 activation attenuates myocardial remodeling and inflammation in permanent MI. Trametinib- and vehicle-treated mice were studied for 7 days after permanent LAD ligation. (A) Relative mRNA expression analysis of Il6, Tnf, Ccl2, and Ccr2 from the infarcted myocardium. (B) Flow cytometry analysis of the infarcted myocardium obtained from vehicle- or trametinib-treated mice normalized to heart weight. Representative gating strategies for quantification of CD45+ leukocytes: CD45+CD90.2–B220–NK1.1–CD11b+ myelomonocytic cells, CD45+CD90.2–B220–NK1.1–CD11b+Ly6G–F4/80–Ly6Chi monocytes, and CD45+CD90.2–B220–NK1.1–CD11b+Ly6G–F4/80–Ly6Clo macrophages. (C) Protein expression analysis of p-ERK1/2 (normalized to total ERK1/2) and activated TGF-β1 (normalized to GAPDH) in infarcted myocardium obtained from vehicle- or trametinib-treated mice. IOD, integrated optical density. (D) Protein expression analysis of p-ERK1/2 (normalized to ERK1/2) and activated TGF-β1 (normalized to GAPDH) in PBMCs isolated from vehicle- or trametinib-treated mice. Ordinary 1-way ANOVA, Šidák’s multiple-comparison test; n = 4–6 animals per group. Data are shown as mean ± SEM. *P < 0.05, **P < 0.01,…

• 

  Figure 5. A profibrotic MEK1/2–TGF-β1 pathway is linked to PAR2-mediated ROS signaling in monocytes. (A and B) Protein expression analysis of monocytes isolated from WT mice and pretreated in vitro with trametinib (10 μM) for 1 hour (A), or isolated from PAR2–/– versus WT mice (B). Cells were stimulated with an inflammatory cytokine cocktail containing IL-6, TNF-α, and CCL2 at a concentration of 20 ng/mL with and without hypoxia for 4 hours. Western blotting of p-ERK1/2 (normalized to total ERK1/2) and activated TGF-β1 (normalized to GAPDH). Ordinary 1-way ANOVA, Šidák’s multiple-comparison test; n = 5 replicates (2–3 mice were pooled for each sample). (C–E) PAR2fl/fl and PAR2fl/fl LysMCre littermates were subjected to permanent LAD ligation and investigated after 7 days; n = 5–10 animals per group. (C) Western blot analysis of activated TGF-β1 (normalized to GAPDH) and p-SMAD2 (normalized to total SMAD2) in the infarcted myocardium. Representative blots and quantification of biological replicates. (D) High-frequency ultrasound echocardiography obtained from PAR2fl/fl LysMCre and PAR2fl/fl littermate control mice with measurement of LVEF (%) and LVEDV (μL). Mann-Whitney test. (E)…

• 

  Figure 6. Myeloid cell–derived TF-PAR2 complex is required for TGF-β1 activation. (A) Confocal microscopy of myocardial cryosections obtained from n = 5 sham-operated and n = 5 LAD-ligated WT (C57BL/6J) mice at day 7. Representative images and quantification of TF+ cells costained for CD45. Unpaired, 2-sided t test. Scale bar: 50 μm. (B–D) TFfl/fl LysMCre and TFfl/fl littermates were subjected to permanent LAD ligation versus sham surgery and investigated after 7 days; n = 5–7 animals per group. (B) Protein expression analysis of p-ERK1/2 (normalized to total ERK1/2) in the infarcted myocardium. Ordinary 1-way ANOVA, Šidák’s multiple-comparison test. (C) Western blot analysis of activated TGF-β1 (normalized to GAPDH) and p-SMAD2 (normalized to total SMAD2) in the infarcted myocardium obtained from TFfl/fl LysMCre and TFfl/fl littermates. Representative blots and quantification of biological replicates. (D) High-frequency ultrasound echocardiography obtained from TFfl/fl LysMCre and TFfl/fl littermates. Ordinary 1-way ANOVA, Šidák’s multiple-comparison test. (E) Sirius red staining and deconvoluted images of fibrotic area on paraffin-embedded heart sections 4 weeks after permanent LAD…

• 

  Figure 7. Myeloid cell to myofibroblast crosstalk for SMAD2 phosphorylation. (A) Confocal microscopy analysis of CD45+p-SMAD2+ and α-SMA+p-SMAD2+ cells in the infarcted myocardium compared with remote myocardium obtained from TFfl/fl LysMCre and TFfl/fl littermates at day 7 after MI. Scale bars: 500 μm. (B) Protein expression analysis of α-SMA (normalized to GAPDH) in the infarcted myocardium compared with myocardium of sham-operated TFfl/fl LysMCre mice at day 7 after MI. Ordinary 1-way ANOVA, Šidák’s multiple-comparison test; n = 4–7 animals per group. Data are shown as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

• 

  Figure 8. TF cytoplasmic tail deletion attenuates ROS production and ERK1/2–TGF-β1 signaling–dependent cardiac fibrosis and improves cardiac function. WT and TFΔCT mice were subjected to permanent LAD ligation versus sham surgery and investigated after 7 days and 4 weeks. (A) Confocal microscopy of myocardial cryosections. Representative images and quantification of MFI of CD45/gp91phox double-positive cells. Scale bars: 75 μm. (B) Assessment of superoxide formation in infarcted myocardium by dihydroethidium-HPLC analysis. Representative chromatogram of 2-hydroxyethidium (2-HE), the oxidation product of DHE, and quantification normalized to weight of the infarcted tissue. Ordinary 1-way ANOVA, Šidák’s multiple-comparison test; n = 5–6 animals per group. (C) Protein expression analysis of monocytes isolated from NOX2–/– animals and stimulated with an inflammatory cytokine cocktail containing IL-6, TNF-α, and CCL2 at a concentration of 20 ng/mL with and without hypoxia for 4 hours. Western blotting of p-ERK1/2 (normalized to ERK1/2) and activated TGF-β1 (normalized to GAPDH). Ordinary 1-way ANOVA, Šidák’s multiple-comparison test; n = 4 replicates per group (2–3 mice were pooled for each…

• 

  Figure 9. Myeloid cell TF cytoplasmic domain phosphorylation mediates ERK1/2–TGF-β1–dependent cardiac remodeling in permanent LAD ligation. (A) Confocal microscopy of myocardial cryosections obtained from WT (C57BL/6J) and TFΔCT mice. Representative images and quantification of MFI of Ly6C+TGFβ-1+ and CD31+TGF-β1+ cells. Ordinary 1-way ANOVA, Šidák’s multiple-comparison test; n = 5 animals per group. Scale bars: 25 μm. (B) Mice with transplanted BM were subjected to permanent LAD ligation versus sham surgery and investigated 7 days later. Western blot analysis of NOX2 (normalized to GAPDH), p-ERK1/2 (normalized to total ERK1/2), and TGF-β1 (normalized to GAPDH) in infarcted myocardium obtained from chimeric mice. Ordinary 1-way ANOVA, Šidák’s multiple-comparison test; n = 5–7 animals per group. Data are shown as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001.

• 

  Figure 10. TF cytoplasmic domain phosphorylation–dependent increased TGF-β1 activation in clinical setting of MI. (A) Representative confocal images of phosphorylation status of TF in infarcted myocardium obtained from WT or TFΔCT mice after 7 days. Representative images and quantification of biological replicates. Kruskal-Wallis test and Dunn’s multiple-comparison test; n = 3–4 animals per group. Scale bars: 50 μm. (B) Representative immunofluorescence confocal microscopy images of CD45+ and CD45/p-TF double-positive cells in human myocardium specimens obtained from n = 5 nonischemic (NI) donor hearts and n = 7 IHF patients. Quantification of biological replicates. Mann-Whitney test. Scale bars: 50 μm. (C and D) Western blot analysis and quantification of human LV tissue obtained from n = 5 nonischemic donor hearts and n = 9 IHF patients for p-TF (normalized to total TF) and TF (C) or TGF-β1 (normalized to GAPDH) and p-SMAD2 (normalized to total SMAD2) (D). Mann-Whitney test. Data are shown as mean ± SEM. *P < 0.05.

• 

  Figure 11. Phosphorylation of TF cytoplasmic domain on circulating monocytes in ongoing cardiac ischemic damage. Circulating PBMCs were isolated from C57BL/6J mice at day 7 after permanent LAD ligation. (A) Western blot analysis of TF expressed by PBMCs. Unpaired, 2-sided t test; n = 5–7 animals per group. (B) Flow cytometry analysis of the PBMCs. Representative counter-plots and quantification of CD45+TF+, CD45+TF+CD115+Ly6C+, and TF+CD115+Ly6C– monocytes per 1 mL whole blood. Mann-Whitney 2-sided t test; n = 5–6 animals per group. (C–E) Representative confocal microscopy of isolated monocytes stained for p-TF (red), TF (red), and DAPI (blue) (C) followed by Western blot analysis of monocytic TF cytoplasmic domain phosphorylation (4G6) and TF (10H10) (D) and plasma levels of activated TGF-β1 (E) in samples obtained from patients with subacute MI (n = 6) and stable coronary artery disease (CAD) (n = 6) as described in Supplemental Table 5. Mann-Whitney unpaired, 2-sided t test. Scale bars: 25 μm. Data are shown as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001.

• 

  Figure 12. Pharmacological targeting of TF-FVIIa improves cardiac function by preventing TGF-β1 activation. (A) Protein expression analysis of p-ERK1/2 (normalized to total ERK1/2), NOX2, and TGF-β1 (normalized to GAPDH) on isolated human monocytes exposed to hypoxia in the presence of cytokine cocktail mix (20 ng/mL) with and without NAPc2 (200 ng/mL). Representative blots and quantification; n = 6 healthy individuals per group. (B) Experimental design: Mice were injected with NAPc2 (1 mg/kg/d) versus sham once daily by i.p. injection from day 1 through day 7. (C) Western blot analysis of NOX2 and TGF-β1 (normalized to GAPDH) obtained from PBMCs of the experimental animals 7 days after MI. Representative images and quantification of replicates. (D and E) High-frequency echocardiography obtained in parasternal long axis (PLAX) with measurement of LVEF and LVEDV on day 7 after LAD ligation (D) and representative blots for protein expression analysis of p-ERK1/2, NOX2, TGF-β1, and α-SMA (E) in infarcted myocardium obtained from vehicle- or NAPc2-treated mice. Ordinary 1-way ANOVA, Šidák’s multiple-comparison test; n= 5–7 animals per group. (F–H) Mice were injected with NAPc2 (1 mg/kg/d) or…