Thrombospondin-1 regulates blood flow via CD47 receptor-mediated activation of NADPH oxidase 1

Abstract

Objective: Although the matricellular protein thrombospondin-1 (TSP1) is highly expressed in the vessel wall in response to injury, its pathophysiological role in the development of vascular disease is poorly understood. This study was designed to test the hypothesis that TSP1 stimulates reactive oxygen species production in vascular smooth muscle cells and induces vascular dysfunction by promoting oxidative stress.

Methods and results: Nanomolar concentrations of TSP1 found in human vascular disease robustly stimulated superoxide (O(2)(•-)) levels in vascular smooth muscle cells at both cellular and tissue level as measured by cytochrome c and electron paramagnetic resonance. A peptide mimicking the C terminus of TSP1 known to specifically bind CD47 recapitulated this response. Transcriptional knockdown of CD47 and a monoclonal inhibitory CD47 antibody abrogated TSP1-triggered O(2)(•-) in vitro and ex vivo. TSP1 treatment of vascular smooth muscle cells activated phospholipase C and protein kinase C, resulting in phosphorylation of the NADPH oxidase organizer subunit p47(phox) and subsequent Nox1 activation, leading to impairment of arterial vasodilatation ex vivo. Further, we observed that blockade of CD47 and NADPH oxidase 1 gene silencing in vivo in rats improves TSP1-induced impairment of tissue blood flow after ischemia reperfusion.

Conclusions: Our data suggest a highly regulated process of reactive oxygen species stimulation and blood flow regulation promoted through a direct TSP1/CD47-mediated activation of Nox1. This is the first report, to our knowledge, of a matricellular protein acting as a ligand for NADPH oxidase activation and through specific engagement of integrin-associated protein CD47.

Figure 1

TSP1 is a rapid and potent stimulator of VSMC O₂^•-. A) Human aortic VSMCs were treated with vehicle or TSP1 (2.2 nM, 60 min). Superoxide production was measured in the 28,000 g membrane fraction using cytochrome c reduction (n=8). B) Rat aortic VSMCs were incubated with vehicle or TSP1 (0.22, 1.1, 2.2, 11 or 22 nM) for 60 min. Superoxide production was measured using EPR (n=4). C) Rat aortic VSMCs were incubated with vehicle or 2.2 nM TSP1 for 10, 30, 60, and 180 min. Superoxide production was measured using L‐012 chemiluminescence (n=4). D) Rat aortic VSMCs were treated with vehicle, TSP1 (2.2 nM, 60 min), PMA (5 μM, 60 min) or AngII (100 nM, 120 min). Superoxide production was measured using cytochrome c (n=6-7). Data represent the mean ± SEM. *p < 0.05 indicates significant difference between vehicle- vs. TSP1-/PMA-/AngII-treatment.

Figure 2

TSP1-stimulated O₂^•- production requires CD47. A) Rat aortic VSMCs were treated with a CD47 or scrambled (Scrmb) morpholino (10 μM) as per the manufacturer's instructions (GeneTools). Western blot images are representative of 6 independent experiments. B) CD47 and Scrmb morpholino-treated VSMCs were incubated with TSP1 (2.2 nM, 60 min) and O₂^•- production was measured using cytochrome c (n=3). C) VSMCs were treated with a CD47 monoclonal antibody [2 μg/ml, CD47 (OX101), Santa Cruz Biotechnology] or isotype control immunoglobulin IgG₁ (2 μg/ml, Santa Cruz Biotechnology) for 30 min, and treated with vehicle or TSP1 (2.2 nM, 60 min). Superoxide production was measured using cytochrome c (n=4). D) VSMCs were stimulated with a TSP1-based peptide (7N3) that binds CD47 (10 μM, 60 min) and O₂^•- production was measured was measured using cytochrome c (n=5). E) Endothelium-denuded wild type mouse aortic rings were pre-treated with a CD47 monoclonal antibody [2 μg/ml, CD47 (MIAP301), Santa Cruz Biotechnology] or isotype control immunoglobulin IgG_2α (2 μg/ml, Santa Cruz) for 30 min, and treated with vehicle or TSP1 (2.2 nM, 60 min). CM^• radical formation was measured for 60 min at 37 °C using EPR. Representative CM^• spectra are presented. F) Cumulative and averaged CM^• radical formation in vehicle- and TSP1-treated endothelium-denuded aortic rings (n=7). Data represent the means ± SEM. *p < 0.05 indicates significant difference between vehicle- and TSP1/7N3-treatment. ^†p < 0.05 indicates significant difference between control IgG- and CD47 antibody-treatment.

Figure 3

Activated CD47 increases O₂^•- production via Nox1 stimulation. A) VSMCs were transfected with Nox1 or Scrmb siRNA and treated with vehicle or TSP1 (2.2 nM, 60 min). Superoxide production was measured using cytochrome c (n=3). B) Endothelium-denuded wild type and Nox1 null aortic rings were treated with vehicle or TSP1 (2.2 nM, 60 min) and CM^• radical formation was measured using EPR. Images are representative of 3 independent experiments. C) Cumulative and averaged CM^• radical formation in vehicle- and TSP1-treated endothelium-denuded aortic rings. The specificity of CMH for O₂^•− was confirmed by the addition of SOD (150 U/ml) (n=3). D) Endothelium-denuded aortic rings from control (Morpholino-vehicle) and Nox1-morpholino treated rats were treated with vehicle or 7N3 (10 μM, 60 min). CM^• radical formation was measured for 60 min at 37 °C using EPR (n=3‐4). Data represent the means ± SEM. *p < 0.05 indicates significant difference between vehicle- and TSP1/7N3-treatment. ^†p < 0.05 indicates significant difference between control and Nox1 siRNA/morpholino treatment.

Figure 4

TSP1 targets Nox subunit assembly via PKC activation and p47^phox phosphorylation. A) VSMCs were incubated with vehicle (control), the PLC inhibitor U-73122 (1 μM, 1 min), the PKC inhibitor calphostin C (1 μM, 30 min) or the PLD inhibitor VU0155069C (10 μM, 30 min), and treated with vehicle or TSP1 (2.2 nM, 60 min) (n=5-6). Superoxide production was measured using cytochrome c. B) Cells were incubated with vehicle or TSP1 (2.2 nM, 60 min). Precipitated p47^phox was probed with an anti-phosphoserine antibody (n=6). Data represent the means ± SEM. *p < 0.05 indicates significant difference between vehicle- and TSP1-treatment.

Figure 5

TSP1 activates arterial Nox1 to inhibit vasodilatation and IRI-blood flow. A) Endothelium-denuded thoracic aortas from control (Morpholino-vehicle) and Nox1 vivo-morpholino-treated (Nox1 morpholino) rats were preincubated with vehicle (PBS) or TSP1 (2.2 nM, 60 min) and preconstricted with phenylephrine (PE, 3×10⁻⁷ M). Endothelium-independent vasorelaxation was stimulated by the NO donor SNP (10⁻¹⁰ to 10⁻⁵ M) (n=4). B) Laser Doppler analysis of hindlimb blood flow during ischemia (45 min) and subsequent reperfusion (80 min) was performed in control (Morpholino-vehicle) and Nox1 vivo-morpholino-treated rats. Fifty min prior to ischemia, rats were injected via the tail vein with TSP1 (60 μg/kg body weight). Representative color Doppler images of the ligated hindlimbs at baseline, following 20 minutes of ischemia and 40 minutes of reperfusion are presented. Red coloration of laser Doppler images indicates maximum and blue coloration minimum blood flow. Regions of interest (ROI) selected for blood flow analysis are defined by white-outlined boxes. White arrow indicates femoral artery. C) Changes in hindlimb perfusion at indicated time points are presented as the flux ratio between the ischemic and non-ischemic limbs (n=3-4). D) Ninety minutes before ischemia, the animals were treated with a CD47 monoclonal antibody (clone OX101; 0.4 μg/g body weight) via a single intraperitoneal injection. Laser Doppler analysis of hindlimb blood flow during ischemia (45 min) and subsequent reperfusion (80 min) was performed in vehicle- and TSP1-treated rats. Data represent the means ± SEM. *p < 0.05 indicates significant difference in relaxation between Morpholino-vehicle+PBS and Morpholino-vehicle+TSP1. ^‡p < 0.05 indicates significant differences at individual concentrations between Morpholino-vehicle+PBS and Morpholino-vehicle+TSP1. ^†p < 0.05 indicates significant difference between Morpholino-vehicle+TSP1 and Nox1 morpholino+TSP1.