Thrombospondin-2 Modulates Extracellular Matrix Remodeling during Physiological Angiogenesis

Thrombospondin 2 (TSP2) can inhibit angiogenesis in vitro by limiting proliferation and inducing apoptosis of endothelial cells (ECs). TSP2 can also modulate the extracellular levels of gelatinases (matrix metalloproteases, MMPs) and potentially influence the remodeling of the extracellular matrix (ECM). Here, we tested the hypothesis that by regulating MMPs, TSP2 could alter EC-ECM interactions. By using a three-dimensional angiogenesis assay, we show that TSP2, but not TSP1, limited angiogenesis by decreasing gelatinolytic activity in situ. Furthermore, TSP2-null fibroblast-derived ECM, which contains irregular collagen fibrils, was more permissive for EC migration. Investigation of the role of TSP2 in physiological angiogenesis in vivo, using excision of the left femoral artery in both TSP2-null and wild-type mice, revealed that TSP2-null mice displayed accelerated recovery of blood flow. This increase was attributable, in part, to an enhanced arterial network in TSP2-null muscles of the upper limb. Angiogenesis in the lower limb was also increased and was associated with increased MMP-9 deposition and gelatinolytic activity. The observed changes correlated with the temporal expression of TSP2 in the ischemic muscle of wild-type mice. Taken together, our observations implicate the matrix-modulating activity of TSP2 as a mechanism by which physiological angiogenesis is inhibited. Thrombospondin 2 (TSP2) can inhibit angiogenesis in vitro by limiting proliferation and inducing apoptosis of endothelial cells (ECs). TSP2 can also modulate the extracellular levels of gelatinases (matrix metalloproteases, MMPs) and potentially influence the remodeling of the extracellular matrix (ECM). Here, we tested the hypothesis that by regulating MMPs, TSP2 could alter EC-ECM interactions. By using a three-dimensional angiogenesis assay, we show that TSP2, but not TSP1, limited angiogenesis by decreasing gelatinolytic activity in situ. Furthermore, TSP2-null fibroblast-derived ECM, which contains irregular collagen fibrils, was more permissive for EC migration. Investigation of the role of TSP2 in physiological angiogenesis in vivo, using excision of the left femoral artery in both TSP2-null and wild-type mice, revealed that TSP2-null mice displayed accelerated recovery of blood flow. This increase was attributable, in part, to an enhanced arterial network in TSP2-null muscles of the upper limb. Angiogenesis in the lower limb was also increased and was associated with increased MMP-9 deposition and gelatinolytic activity. The observed changes correlated with the temporal expression of TSP2 in the ischemic muscle of wild-type mice. Taken together, our observations implicate the matrix-modulating activity of TSP2 as a mechanism by which physiological angiogenesis is inhibited. The thrombospondins (TSPs) are a small family of five, secreted, modular glycoproteins (TSPs 1 to 5), with diverse functions.1Adams JC Thrombospondins: multifunctional regulators of cell interactions.Annu Rev Cell Dev Biol. 2001; 17: 25-51Crossref PubMed Scopus (333) Google Scholar, 2Bornstein P Thrombospondins as matricellular modulators of cell function.J Clin Invest. 2001; 107: 929-934Crossref PubMed Scopus (419) Google Scholar TSP1 and TSP2 share a high degree of similarity and are thought to constitute a subfamily. TSP1 has been extensively studied and has been shown to be synthesized by a variety of cells and to interact with a number of receptors such as CD36, CD47, GPIIb/IIIa, heparan sulfate proteoglycan, low-density lipoprotein receptor-related protein (LRP), and several integrins.3Iruela-Arispe ML Luque A Lee N Thrombospondin modules and angiogenesis.Int J Biochem Cell Biol. 2004; 36: 1070-1078Crossref PubMed Scopus (105) Google Scholar TSP2, has not been extensively studied, but because of its similarity to TSP1 it is believed that it can bind to the same receptors.4Bornstein P Devarayalu S Li P Disteche CM Framson P A second thrombospondin gene in the mouse is similar in organization to thrombospondin 1 but does not respond to serum.Proc Natl Acad Sci USA. 1991; 88: 8636-8640Crossref PubMed Scopus (67) Google Scholar, 5Bornstein P O'Rourke K Wikstrom K Wolf FW Katz R Li P Dixit VM A second, expressed thrombospondin gene (Thbs2) exists in the mouse genome.J Biol Chem. 1991; 266: 12821-12824Abstract Full Text PDF PubMed Google Scholar In fact, CD36, heparan sulfate proteoglycan, LRP, and αVβ3 have been shown to be receptors for TSP2.6Chen H Strickland DK Mosher DF Metabolism of thrombospondin 2. Binding and degradation by 3t3 cells and glycosaminoglycan-variant Chinese hamster ovary cells.J Biol Chem. 1996; 271: 15993-15999Crossref PubMed Scopus (58) Google Scholar, 7Chen H Sottile J O'Rourke KM Dixit VM Mosher DF Properties of recombinant mouse thrombospondin 2 expressed in Spodoptera cells.J Biol Chem. 1994; 269: 32226-32232Abstract Full Text PDF PubMed Google Scholar TSPs have also been shown to interact with several extracellular matrix (ECM) proteins including collagen, fibrinogen, and fibronectin.2Bornstein P Thrombospondins as matricellular modulators of cell function.J Clin Invest. 2001; 107: 929-934Crossref PubMed Scopus (419) Google Scholar, 8Adams JC Lawler J The thrombospondins.Int J Biochem Cell Biol. 2004; 36: 961-968Crossref PubMed Scopus (358) Google Scholar Recently, the very low-density lipoprotein receptor was shown to be a receptor for TSP1 and TSP2, and their interaction was shown to inhibit the division of microvascular endothelial cells (ECs).9Oganesian A Armstrong LC Migliorini MM Strickland DK Bornstein P Thrombospondins use the VLDL receptor and a non-apoptotic pathway to inhibit cell division in microvascular endothelial cells.Mol Biol Cell. 2008; 19: 563-571Crossref PubMed Scopus (61) Google Scholar TSP1 was identified as the first endogenous inhibitor of angiogenesis.10Good DJ Polverini PJ Rastinejad F Le Beau MM Lemons RS Frazier WA Bouck NP A tumor suppressor-dependent inhibitor of angiogenesis is immunologically and functionally indistinguishable from a fragment of thrombospondin.Proc Natl Acad Sci USA. 1990; 87: 6624-6628Crossref PubMed Scopus (908) Google Scholar The anti-angiogenic activities of TSP1 and TSP2 have been the focus of rigorous investigation and numerous studies have implicated both proteins in the regulation of tumor angiogenesis.3Iruela-Arispe ML Luque A Lee N Thrombospondin modules and angiogenesis.Int J Biochem Cell Biol. 2004; 36: 1070-1078Crossref PubMed Scopus (105) Google Scholar, 11Lawler J Thrombospondin-1 as an endogenous inhibitor of angiogenesis and tumor growth.J Cell Mol Med. 2002; 6: 1-12Crossref PubMed Scopus (463) Google Scholar, 12Cursiefen C Masli S Ng TF Dana MR Bornstein P Lawler J Streilein JW Roles of thrombospondin-1 and -2 in regulating corneal and iris angiogenesis.Invest Ophthalmol Vis Sci. 2004; 45: 1117-1124Crossref PubMed Scopus (156) Google Scholar, 13Zhang X Lawler J Thrombospondin-based antiangiogenic therapy.Microvasc Res. 2007; 74: 90-99Crossref PubMed Scopus (119) Google Scholar TSP1 and TSP2 have also been shown to have broad anti-angiogenic activities in in vivo and in vitro assays, and a down-regulation of TSP1 synthesis has been implicated in a number of pathological conditions that involve increased angiogenesis. Like TSP1, TSP2 can directly influence ECs by inhibiting basic fibroblast growth factor-induced migration, lysophosphatidic acid-induced mitogenesis, and the formation of focal adhesions.14Volpert OV Stellmach V Bouck N The modulation of thrombospondin and other naturally occurring inhibitors of angiogenesis during tumor progression.Breast Cancer Res Treat. 1995; 36: 119-126Crossref PubMed Scopus (95) Google Scholar, 15Panetti TS Chen H Misenheimer TM Getzler SB Mosher DF Endothelial cell mitogenesis induced by LPA: inhibition by thrombospondin-1 and thrombospondin-2.J Lab Clin Med. 1997; 129: 208-216Abstract Full Text PDF PubMed Scopus (95) Google Scholar, 16Murphy-Ullrich JE Gurusiddappa S Frazier WA Hook M Heparin-binding peptides from thrombospondins 1 and 2 contain focal adhesion-labilizing activity.J Biol Chem. 1993; 268: 26784-26789Abstract Full Text PDF PubMed Google Scholar, 17Murphy-Ullrich JE Hook M Thrombospondin modulates focal adhesions in endothelial cells.J Cell Biol. 1989; 109: 1309-1319Crossref PubMed Scopus (233) Google Scholar A mechanism for the anti-angiogenic effect of TSP1 was shown to involve the interaction of the type I repeats of TSP1 with the scavenger receptor CD36 on ECs.18Dawson DW Pearce SF Zhong R Silverstein RL Frazier WA Bouck NP CD36 mediates the in vitro inhibitory effects of thrombospondin-1 on endothelial cells.J Cell Biol. 1997; 138: 707-717Crossref PubMed Scopus (561) Google Scholar Subsequent studies determined the downstream events that included activation of p59fyn leading to activation of p38MAPK and caspase 3-like proteases and induction of EC apoptosis.19Jiménez B Volpert OV Crawford SE Febbraio M Silverstein RL Bouck N Signals leading to apoptosis-dependent inhibition of neovascularization by thrombospondin-1.Nat Med. 2000; 6: 41-48Crossref PubMed Scopus (870) Google Scholar However, it is unclear if the same mechanism is involved in the regulation of physiological angiogenesis. It is also unclear whether TSP2 can induce the activation of this cascade, even though binding of TSP2 to CD36 on ECs has been reported.20Simantov R Febbraio M Silverstein RL The antiangiogenic effect of thrombospondin-2 is mediated by CD36 and modulated by histidine-rich glycoprotein.Matrix Biol. 2005; 24: 27-34Crossref PubMed Scopus (89) Google Scholar Surprisingly, studies have shown that TSP2 does not induce apoptosis in ECs, but inhibits cell cycle progression independent of CD36.9Oganesian A Armstrong LC Migliorini MM Strickland DK Bornstein P Thrombospondins use the VLDL receptor and a non-apoptotic pathway to inhibit cell division in microvascular endothelial cells.Mol Biol Cell. 2008; 19: 563-571Crossref PubMed Scopus (61) Google Scholar, 21Armstrong LC Bjorkblom B Hankenson KD Siadak AW Stiles CE Bornstein P Thrombospondin 2 inhibits microvascular endothelial cell proliferation by a caspase-independent mechanism.Mol Biol Cell. 2002; 13: 1893-1905Crossref PubMed Scopus (94) Google Scholar Differences between the phenotypes of TSP1-null and TSP2-null mice suggest that the two proteins might function differently. For example, wound-healing studies in TSP1-null, TSP2-null, and double-TSP1/2-null mice indicated that, based on spatiotemporal expression patterns, only the loss of TSP2 could lead to augmented angiogenesis and accelerated repair.22Agah A Kyriakides TR Lawler J Bornstein P The lack of thrombospondin-1 (TSP1) dictates the course of wound healing in double-TSP1/TSP2-null mice.Am J Pathol. 2002; 161: 831-839Abstract Full Text Full Text PDF PubMed Scopus (252) Google Scholar, 23Kyriakides TR Tam JW Bornstein P Accelerated wound healing in mice with a disruption of the thrombospondin 2 gene.J Invest Dermatol. 1999; 113: 782-787Crossref PubMed Scopus (146) Google Scholar TSP1-null mice on the other hand, displayed delayed healing because of a compromised inflammatory response. Studies using models of ischemia have indicated that TSP1, and perhaps TSP2, can limit angiogenesis and hinder recovery of blood flow and tissue survival. Specifically, it was shown that TSP1-null mice have enhanced ischemic tissue survival in a random myocutaneous flap and hindlimb ischemia models.24Isenberg JS Romeo MJ Abu-Asab M Tsokos M Oldenborg A Pappan L Wink DA Frazier WA Roberts DD Increasing survival of ischemic tissue by targeting CD47.Circ Res. 2007; 100: 602-613Crossref PubMed Scopus (96) Google Scholar, 25Isenberg JS Hyodo F Matsumoto K Romeo MJ Abu-Asab M Tsokos M Kuppusamy P Wink DA Krishna MC Roberts DD Thrombospondin-1 limits ischemic tissue survival by inhibiting nitric oxide-mediated vascular smooth muscle relaxation.Blood. 2007; 109: 1945-1952Crossref PubMed Scopus (93) Google Scholar The ability of TSP1 to reduce NO/cGMP signaling in vascular smooth muscle cells was implicated as the mechanism mediating the endogenous response to ischemia. In both the myocutaneous flap and the hindlimb ischemia models, the association of TSP1 with CD47, but not CD36, was shown to be critical for these responses.24Isenberg JS Romeo MJ Abu-Asab M Tsokos M Oldenborg A Pappan L Wink DA Frazier WA Roberts DD Increasing survival of ischemic tissue by targeting CD47.Circ Res. 2007; 100: 602-613Crossref PubMed Scopus (96) Google Scholar In a separate study, double-TSP1/2-null mice were shown to have enhanced recovery from hindlimb ischemia.26Kopp HG Hooper AT Broekman MJ Avecilla ST Petit I Luo M Milde T Ramos CA Zhang F Kopp T Bornstein P Jin DK Marcus AJ Rafii S Thrombospondins deployed by thrombopoietic cells determine angiogenic switch and extent of revascularization.J Clin Invest. 2006; 116: 3277-3291Crossref PubMed Scopus (88) Google Scholar The improved phenotype could be conferred to wild-type (WT) mice and reversed in double-null mice after reverse bone marrow transplantation, suggesting that circulating TSP-producing cells, namely platelets, can influence recovery. Even though the observed changes were attributed to both TSP1 and TSP2, platelets do not contain TSP2.27Kyriakides TR Rojnuckarin P Reidy MA Hankenson KD Papayannopoulou T Kaushansky K Bornstein P Megakaryocytes require thrombospondin-2 for normal platelet formation and function.Blood. 2003; 101: 3915-3923Crossref PubMed Scopus (41) Google Scholar Thus, the participation of TSP2 in ischemic revascularization still remains unclear. Clearly TSP1 and TSP2 can interact directly with ECs and modulate their function in vitro. However, recent studies have shown that the loss of TSP2 can cause an increase in the levels of matrix metalloproteinase (MMP)-2 and/or MMP-9.28Kyriakides TR Zhu YH Yang Z Huynh G Bornstein P Altered extracellular matrix remodeling and angiogenesis in sponge granulomas of thrombospondin 2-null mice.Am J Pathol. 2001; 159: 1255-1262Abstract Full Text Full Text PDF PubMed Scopus (103) Google Scholar, 29Yang Z Kyriakides TR Bornstein P Matricellular proteins as modulators of cell-matrix interactions: adhesive defect in thrombospondin 2-null fibroblasts is a consequence of increased levels of matrix metalloproteinase-2.Mol Biol Cell. 2000; 11: 3353-3364Crossref PubMed Scopus (179) Google Scholar, 30Schroen B Heymans S Sharma U Blankesteijn WM Pokharel S Cleutjens JP Porter JG Evelo CT Duisters R van Leeuwen RE Janssen BJ Debets JJ Smits JF Daemen MJ Crijns HJ Bornstein P Pinto YM Thrombospondin-2 is essential for myocardial matrix integrity: increased expression identifies failure-prone cardiac hypertrophy.Circ Res. 2004; 95: 515-522Crossref PubMed Scopus (161) Google Scholar To date, an increase in the levels of MMP-2 or MMP-9 and changes in ECM assembly in TSP1-null mice have not been reported. Perhaps this is attributable to the spatiotemporal expression of TSP1, which is predominantly secreted by platelets at sites of injury during the acute phase. A possible mechanism for the regulation of the extracellular levels of MMP-2 by TSP2 has been elucidated in dermal fibroblasts.31Yang Z Strickland DK Bornstein P Extracellular matrix metalloproteinase 2 levels are regulated by the low density lipoprotein-related scavenger receptor and thrombospondin 2.J Biol Chem. 2001; 276: 8403-8408Crossref PubMed Scopus (242) Google Scholar Specifically, it was shown that TSP2 could bind MMP-2 and direct it to the catabolic receptor LRP, thus causing its clearance from the ECM. However, it is not clear whether the same mechanism can effectively control the levels of MMPs during tissue remodeling. We postulated that TSPs act as ECM-associated proteins to influence EC function in an ECM-dependent manner. Thus, we analyzed the ability of TSP1 and TSP2 to inhibit angiogenesis in a three-dimensional matrix-based assay. We found that only TSP2 could limit angiogenesis, suggesting a mode of action independent of binding to receptors that are common between TSP1 and TSP2. Furthermore, we found that addition of TSP2 caused a decrease in gelatinolytic activity and in subsequent experiments, discovered ultrastructural abnormalities in the ECM produced by TSP2-null cells. Moreover, ECM derived from TSP2-null fibroblasts could alter EC function. To investigate further the participation of TSP2 in modulating physiological angiogenesis we induced hindlimb ischemia via femoral artery excision in TSP2-null and WT mice and compared their respective abilities to recover blood flow. In addition, we determined the spatiotemporal expression of TSP2 during the recovery process. Ischemic tissues respond to reduced blood flow and hypoxia by mounting arteriogenic and subsequent angiogenic responses that collectively can lead to partial or full restoration of blood flow.32Waters RE Terjung RL Peters KG Annex BH Preclinical models of human peripheral arterial occlusive disease: implications for investigation of therapeutic agents.J Appl Physiol. 2004; 97: 773-780Crossref PubMed Scopus (98) Google Scholar, 33Heil M Schaper W Cellular mechanisms of arteriogenesis.EXS. 2005; 94: 181-191PubMed Google Scholar Experimental hindlimb ischemia in mice has served as a useful model for the identification of molecules that are critical for recovery from ischemia. Specifically, because mice have pre-existing collaterals in the limbs, after ischemia they display an impressive increase in arteriogenesis and angiogenesis.34Heil M Schaper W Influence of mechanical, cellular, and molecular factors on collateral artery growth (arteriogenesis).Circ Res. 2004; 95: 449-458Crossref PubMed Scopus (314) Google Scholar In the present study, we found that the deposition of TSP2 is increased in ischemic tissues and that TSP2-null mice display more baseline collaterals and enhanced recovery of blood flow after ischemia. The improved recovery was associated with increased angiogenesis, and elevated levels of MMP-9 and gelatinolytic activity. Therefore, we conclude that the paracrine function of TSP2 as a modulator of ECM assembly was responsible, in part, for its anti-angiogenic activity. In preparation for the angiogenesis assay, human umbilical vein endothelial cells (HUVECs) or human dermal microvascular endothelial cells (HDMECs) at 80% confluence were cultured overnight in M199 medium supplemented with 1% fetal bovine serum. The cells were collected by trypsinization and incubated at 4°C for 10 minutes in M199 supplemented with 5% fetal bovine serum. Cells (5 × 105) in 250 μl of medium were mixed with 250 μl of growth factor-reduced Matrigel (BD Bioscience, San Jose, CA) and plated in 24-well plates. After 30 minutes of incubation to allow for gel formation, M199 medium containing 5% serum and various concentrations of TSP1 or TSP2 (1 to 10 μg/ml) was added to the wells. Human platelet TSP1 was from a commercial source (EMD Biosciences, San Diego, CA) and recombinant TSP1 and TSP2 were prepared as described previously.9Oganesian A Armstrong LC Migliorini MM Strickland DK Bornstein P Thrombospondins use the VLDL receptor and a non-apoptotic pathway to inhibit cell division in microvascular endothelial cells.Mol Biol Cell. 2008; 19: 563-571Crossref PubMed Scopus (61) Google Scholar Selected wells were treated with 10 μmol/L MMP inhibitor GM6001 (Calbiochem, La Jolla, CA) or dimethyl sulfoxide as vehicle control. Images were collected 24 hours later. The three-dimensional angiogenesis assay was also performed with 250 μl of Vitrogen (Cohesion, Palo Alto, CA) containing 90 μg of human fibronectin (BD Biosciences). For detection of gelatinolytic activity, HUVECs were prepared as described above and mixed in Matrigel supplemented with 25 μg of DQ gelatin (Molecular Probes, Eugene, OR) and plated in 24-well plates. Gels were treated with 2 μg/ml of TSP1 or TSP2 or 10 μmol/L GM6001 for 24 hours and the medium was removed and the gels were fixed with 4% paraformaldehyde for 1 hour. The cells were permeabilized with 0.1% Triton X-100 in 1× phosphate-buffered saline (PBS) for 10 minutes, stained with 1 mg/ml of 4,6-diamidino-2-phenylindole (DAPI), and visualized with fluorescence microscopy. Gelatinase activity was detected as green fluorescence and nuclei appeared blue. Gelatinase activity was quantified from digital images using Metamorph imaging software (Molecular Devices, Sunnyvale, CA). Primary dermal fibroblasts from WT mice or TSP2-null mice were prepared as we described previously and plated into 24-well plates (7.5 × 104 cells per well).29Yang Z Kyriakides TR Bornstein P Matricellular proteins as modulators of cell-matrix interactions: adhesive defect in thrombospondin 2-null fibroblasts is a consequence of increased levels of matrix metalloproteinase-2.Mol Biol Cell. 2000; 11: 3353-3364Crossref PubMed Scopus (179) Google Scholar Cells were incubated for 7 days in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum and 100 μmol/L ascorbic acid to increase the production of ECM. TSP2-null fibroblasts were also cultured in media that was supplemented with 10 μmol/L GM6001 or dimethyl sulfoxide. Cells were removed by incubation and gentle shaking at 37°C for 2 minutes in decellularization buffer (20 mmol/L NH4OH and 0.5% Triton X-100). For morphological analysis, the ECM was fixed with 4% paraformaldehyde and stained with DAPI (1 mg/ml) and rhodamine phalloidin (1:100, Molecular Probes) to confirm the absence of cells. In addition, ECM was stained with anti-fibronectin antibody at 1:400 dilution (rabbit polyclonal; Abcam, Cambridge, UK). Furthermore, mAb HU177 (100 μg/ml), a kind gift from Dr. Peter Brooks, New York University School of Medicine, New York, NY, was used to detect the presence of degraded collagen as described previously.35Xu J Rodriguez D Kim JJ Brooks PC Generation of monoclonal antibodies to cryptic collagen sites by using subtractive immunization.Hybridoma. 2000; 19: 375-385Crossref PubMed Scopus (49) Google Scholar For ultrastructural analysis, samples were prepared according to standard procedures for transmission electron microscopy analysis. Confluent HUVECs were recovered by trypsinization and suspended in M199 medium deprived of serum. Three hundred μl of cell suspension (3 × 105 cells) was added to wells containing decellularized ECM. Cells were allowed to attach for 1 hour at 37°C and unattached cells were removed by washing (3× in PBS). Cell attachment was determined by the analysis of five random high-power (×200) microscopic fields (phase-contrast) per well. To determine the effects of the ECM on cell morphology and growth, cells were cultured on ECM for 12 hours. To visualize the actin cytoskeleton and nuclei, cells were fixed and permeabilized in 4% paraformaldehyde/Triton X-100 (JT Baker, Phillipsburg, NJ) for 20 minutes at room temperature, and stained with rhodamine-phalloidin and DAPI according to standard protocols. All wells were mounted with Vectashield (Vector Laboratories, Burlingame, CA) and examined with the aid of an Axiovert 200M microscope (Carl Zeiss, Thornwood, NY) equipped with fluorescent optics. The effect of the ECM on cell migration was analyzed using a modified Boyden chamber (Transwell; Costar Corp., Cambridge, MA). Specifically, WT- or TSP2-null fibroblasts (5 × 104 cells per well) were plated in the top chamber of a transwell for 7 days and WT and TSP2-derived ECM were prepared by decellularization as described above. Subconfluent HUVECs were serum-starved overnight, collected by trypsinization, and suspended in M199 medium without serum. Cells (1 × 105) were added to the top of each migration chamber and allowed to migrate to the underside of the chamber for 6 hours in the presence or absence of 400 nmol/L sphingosine-1-phosphate. At the completion of the migration assay, cells were fixed and stained (Hema 3 Stain System; Fisher Diagnostics, Pittsburgh, PA) and the upper surface of the filter was scraped with a cotton swab. Four random fields (×100) per well were counted. All experiments were performed in quadruplicate wells and repeated at least twice. All animal studies were approved by the Institutional Animal Care Committee of Yale University. TSP2-null (C57BL6/129SVJ mice) and littermate WT (C57BL6/129SVJ) mice were used for all experiments. Mouse hindlimb ischemia was induced as described previously.36Ackah E Yu J Zoellner S Iwakiri Y Skurk C Shibata R Ouchi N Easton RM Galasso G Birnbaum MJ Walsh K Sessa WC Akt1/protein kinase Balpha is critical for ischemic and VEGF-mediated angiogenesis.J Clin Invest. 2005; 115: 2119-2127Crossref PubMed Scopus (337) Google Scholar, 37Yu J deMuinck ED Zhuang Z Drinane M Kauser K Rubanyi GM Qian HS Murata T Escalante B Sessa WC Endothelial nitric oxide synthase is critical for ischemic remodeling, mural cell recruitment, and blood flow reserve.Proc Natl Acad Sci USA. 2005; 102: 10999-11004Crossref PubMed Scopus (275) Google Scholar Briefly, after anesthesia (100 mg/kg ketamine and 10 mg/kg xylazine), the left femoral artery was exposed under a dissection microscope. The proximal portion of the femoral artery and the distal portion of the saphenous artery were ligated. All branches between the two sites were ligated or cauterized and arteriectomy was performed. Sham operations involved skin incision without femoral artery ligation. A total of six mice per time point per genotype were analyzed. Blood flow was measured by the Periflux system with a laser Doppler perfusion unit (Perimed, North Royalton, OH). A deep measurement probe was placed directly onto the gastrocnemius muscle to ensure a deep muscle flow measurement. Ischemic and nonischemic limb perfusion were measured before and directly and then at 1, 2, 3, and 4 weeks after surgery. The final blood flow value was expressed as the ratio of ischemic to nonischemic hindlimb perfusion. A total of six mice per time point per genotype were analyzed. Two weeks after surgery, mice were anesthetized and heparinized. Mice were perfused with PBS containing vasodilators (papaverine, 4 mg/L; adenosine, 1g/L) for 3 minutes at physiological pressure through the descending aorta, and blood was drained from the inferior vena cava.37Yu J deMuinck ED Zhuang Z Drinane M Kauser K Rubanyi GM Qian HS Murata T Escalante B Sessa WC Endothelial nitric oxide synthase is critical for ischemic remodeling, mural cell recruitment, and blood flow reserve.Proc Natl Acad Sci USA. 2005; 102: 10999-11004Crossref PubMed Scopus (275) Google Scholar The vasculature was fixed with 2% paraformaldehyde (in PBS) for 5 minutes, flushed with PBS for 2 minutes, and infused with contrast agent (bismuth oxychloride in saline and 10% gelatin in PBS, 1:1). Mice were then immersed in ice to solidify the contrast agent. Microangiography was taken with a Kubtec X-ray machine (XPERT80; KUB Technologies Inc., Fairfield, CT) at 25 kV for 70 seconds. Upper limb vascular density (pixel density), vessel length (average length of vessels with diameter >1 pixel), and fractal dimension were analyzed by modified IMAGEJ and MATLAB software (Research Services Branch, National Institute of Mental Health, Bethesda, MD). A total of six mice per genotype were analyzed. Skeletal muscle contraction-stimulated hyperemia was determined as described previously.37Yu J deMuinck ED Zhuang Z Drinane M Kauser K Rubanyi GM Qian HS Murata T Escalante B Sessa WC Endothelial nitric oxide synthase is critical for ischemic remodeling, mural cell recruitment, and blood flow reserve.Proc Natl Acad Sci USA. 2005; 102: 10999-11004Crossref PubMed Scopus (275) Google Scholar Briefly, anesthetized mice were placed on a heated surface and the gastrocnemius and adductor muscle group were exposed via a midline incision of the limb. Gastrocnemius muscle blood flow was measured at baseline (prestimulation), and followed by 2 minutes of stimulation of the adductor muscles with two electrodes at 2 Hz and 5 mA with the aid of an electrostimulator. Measurements were taken at baseline and at 1-minute intervals after stimulation and recorded by Matlab Chart (ADInstruments, Grand Junction, CO). A total of eight mice per time point per genotype were analyzed. Tissues were snap-frozen in liquid nitrogen, pulverized, and resuspended in lysis buffer containing: 50 mmol/L Tris-HCl, pH 7.4, 10 mmol/L ethylenediaminetetraacetic acid, 0.5% Triton X-100, 0.1% sodium deoxycholate, 150 mmol/L NaCl, 10 mmol/L NaF, 1 mmol/L sodium orthovanadate, 1 mmol/L Pefabloc SC, and 2 mg/ml protease inhibitor cocktail (Roche Diagnostics Corp., Indianapolis, IN). The mixture was incubated 30 minutes at 4°C with inversion and centrifuged for 20 minutes at 14,000 rpm. Protein concentrations were determined by using the Bradford protein assay kit (Bio-Rad Laboratories, Hercules, CA). Mice were sacrificed at 0, 1, 2, 3, and 4 weeks after surgery, and muscles of the lower limb were harvested, fixed with zinc fixative, preserved, and embedded in paraffin to generate tissue sections (5 μm thick). Sections were stained using anti-PECAM-1/CD31 antibody (1:100 dilution; BD Bioscience-Pharmingen), anti-smooth muscle actin antibody (Ab15267, Abcam), anti-MMP-9 antibody (1:1000, Abcam), 7/4 neutrophil (1:400, Abcam), anti-ephrin B2 (1:100; Neuromics, Edina, MN) and anti-TSP2 antibody (1:1000, BD Bioscience). The bound primary antibodies were detected by using the peroxidase-based ABC Elite kit (Vector Laboratories). The sections were counterstained with 1% methylene green for 2 minutes and mounted. High-power (×40 objective) and low-power (×10 objective) images from three random areas of each section and two sections per muscle were taken using a Zeiss microscope equipped with a digital camera. Capillary density and collaterals were quantified by counting the number of blood vessels per field from high- and low-power images, respectively. A total of 40 images from six mice per genotype per time