Thrombospondin-1 limits ischemic tissue survival by inhibiting nitric oxide-mediated vascular smooth muscle relaxation

Abstract

The nitric oxide (NO)/cGMP pathway, by relaxing vascular smooth muscle cells, is a major physiologic regulator of tissue perfusion. We now identify thrombospondin-1 as a potent antagonist of NO for regulating F-actin assembly and myosin light chain phosphorylation in vascular smooth muscle cells. Thrombospondin-1 prevents NO-mediated relaxation of precontracted vascular smooth muscle cells in a collagen matrix. Functional magnetic resonance imaging demonstrated that an NO-mediated increase in skeletal muscle perfusion was enhanced in thrombospondin-1-null relative to wild-type mice, implicating endogenous thrombospondin-1 as a physiologic antagonist of NO-mediated vasodilation. Using a random myocutaneous flap model for ischemic injury, tissue survival was significantly enhanced in thrombospondin-1-null mice. Improved flap survival correlated with increased recovery of oxygen levels in the ischemic tissue of thrombospondin-1-null mice as measured by electron paramagnetic resonance oximetry. These findings demonstrate an important antagonistic relation between NO/cGMP signaling and thrombospondin-1 in vascular smooth muscle cells to regulate vascular tone and tissue perfusion.

Figure 1

TSP1 antagonizes NO-dependent alterations in F-actin and dephosphorylation of MLC in VSMCs. HAVSMCs plated on glass chamber slides were incubated in basal medium with 0.1% BSA (A-B) or 2.2 nM TSP1 (C-D) ± DEA/NO (10 μM). Cells were then fixed, permeabilized, and stained with Oregon Green-phalloidin to visualize F-actin. Photomicrographic images were acquired on a Nikon Eclipse E1000 microscope (Nikon, Melville, NY) using a Plan Apo objective lens. Low-magnification images were taken at 20× and a numeric aperture of 0.75; high-power images, at 40× and a numeric aperture of 0.95. No imaging medium or solution was used. A Cool Snap FX camera (Roberts Scientific, Tucson, AZ), IP Lab 3.5 software (Scanalytics, Fairfax, VA), and Photoshop C5 (Adobe Systems, San Jose, CA) were used for image acquisition and processing. Photomicrographs representative of 3 separate experiments are presented. Scale bar = 50 μm. HAVSMCs in 96-well plates were similarly treated and stained, and the fluorescence was signal quantified (E). *P < .05 versus BSA − NO, Student t test. #P < .05 versus BSA + NO, $P < .05 versus BSA − NO, 2-way ANOVA. &P < .05 versus S1P − NO, one-way ANOVA. Lysates of HAVSMCs in growth medium with 2% serum and treated with the indicated combinations of 100 nM S1P, 10 μM DEA/NO, and 2.2 nM TSP1 for 5 minutes were separated by SDS–polyacrylamide gel electrophoresis (PAGE) and analyzed by Western blot to determine the levels of MLC phosphorylation and total MLC (F). The blot shown is representative of 4 independent experiments. Results are presented as the mean ± SD.

Figure 2

NO-stimulated VSMC contraction is blocked in the presence of exogenous and endogenous TSP1 and S1P. Type I collagen gels (3 mg/mL) were prepared and seeded with either HAVSMCs (A-B) (50 000 cells in 75 μL gel/well) or VSMCs harvested from aortic segments from WT or TSP1-null mice (C-D) (75 000 cells in 75 μL gel/well) and divided into aliquots in 96-well plates (Nunc, Denmark) and incubated overnight. Wells treated with TSP1 were preincubated overnight with 2.2 nM TSP1. Following release of the gels, contraction was initiated with either 10% FCS or 100 nM S1P ± 10 μM DETA/NO, and contraction was determined. *P < .05 versus FCS + NO, #P < .05 versus S1P + NO, &P < .05 versus TSP ± S1P, Student t test. Results are presented as the mean ± SD.

Figure 3

Endogenous TSP1 limits tissue perfusion responses to NO in vivo. BOLD MRI images for (A) WT and (B) TSP1-null mice were obtained from T₂* weighted sequences. DEA/NO (100 nmol/g body weight) was injected with saline via an intrarectal cannula 5 minutes after starting the scan. Green and red colors show positive and negative BOLD MRI signals, respectively, at the indicated times after NO administration. The BOLD images were superimposed with the corresponding anatomic images to determine exact locations in the lateral thigh sections. (C) BOLD MRI signal changes as a function of time after NO challenge. The green and red plots show increased and decreased BOLD MRI signals, respectively. Values are presented as mean ± SD from 5 and 4 experiments in WT and TSP1-null mice, respectively.

Figure 4

Endogenous TSP1 and NO modulate tissue survival under ischemic conditions. (A) Representative random flaps were photographed 7 days following surgery for untreated WT and TSP1-null mice, WT and TSP1-null mice receiving L-NAME (500 mg/L; B), or mice receiving ISDN (1 mg/mL; C) in the drinking water during the postoperative period. (D) Flap survival is expressed as the percentage of the total involved area. Results are the mean ± SD of 24 animals (12 age- and sex-matched pairs) of untreated WT and TSP1-null mice, 16 animals (8 matched pairs) treated with L-NAME, and 16 animals (8 matched pairs) treated with ISDN. *P < .05 versus control, one-way ANOVA. #P < .05 versus wild type, 2-way ANOVA.

Figure 5

Increased angiogenic and spindle cell responses in random ischemic flaps in the absence of endogenous TSP1. Sections from necrotic areas of the excised skin flap in WT (A) and TSP1-null (B) mice are shown. In WT mice the epidermis (E) is necrotic and heavily infiltrated by polymorphonuclear leukocytes (P). A layer of loose granulation tissue (G) is present under the muscular layer (M). The layer of granulation tissue is significantly thicker and more heavily vascularized in the skin flap of the TSP1-null mouse. H + E; original magnification, × 4. Higher magnification of the granulation tissue in the WT (C) and TSP1-null (D) flaps shows more prominent spindle-cell proliferation and capillary formation in the TSP1-null flap. H + E; original magnification, × 20. Immunohistochemical staining with a TSP1 monoclonal antibody of wild-type flaps at 4 hours (E) and 72 hours (F) after surgery was performed. Tissue obtained 4 hours after surgery demonstrated diffuse TSP1 staining of the epidermis, subcutaneous arterioles (arrow), extracellular matrix, striated muscle, and inflammatory cells. At 72 hours after surgery staining was localized to muscle cell borders and extracellular matrix with less staining in other areas. Original magnification, × 20.

Figure 6

Tissue pO₂ in WT and TSP1-null mice after flap treatment using EPR oxymetry. (A) Schematic showing LiPc crystal placement in relation to a dorsal random myocutaneous flap. LiPc crystals were implanted in the dorsal subdermal area of mice 7 days prior to flap elevation. Initial measurements were performed by 700-MHz EPR spectroscopy with a small surface coil to confirm crystal location and calculate basal pO₂ levels. Body temperature of the animals was maintained at 37.5°C ± 0.5°C. Following flap elevation and suturing, measurements were recorded at the indicated times (B). Data represent the mean ± SE of measurements from 4 animals in each group. $P < .05 between proximal and distal against control in WT versus TSP^−/−. *P < .05 between proximal and distal.