Proline-rich tyrosine kinase 2 via enhancing signal transducer and activator of transcription 3-dependent cJun expression mediates retinal neovascularization

Despite the involvement of proline-rich tyrosine kinase 2 (Pyk2) in endothelial cell angiogenic responses, its role in pathological retinal angiogenesis is not known. In the present study, we show that vascular endothelial growth factor A (VEGFA) induces Pyk2 activation in mediating human retinal microvascular endothelial cell (HRMVEC) migration, sprouting and tube formation. Downstream to Pyk2, VEGFA induced signal transducer and activator of transcription 3 (STAT3) activation and cJun expression in the modulation of HRMVEC migration, sprouting and tube formation. Consistent with these observations, hypoxia induced activation of Pyk2-STAT3-cJun signaling axis and siRNA-mediated downregulation of Pyk2, STAT3 or cJun levels substantially inhibited hypoxia-induced retinal endothelial cell proliferation, tip cell formation and neovascularization. Together, these observations suggest that activation of Pyk2-mediated STAT3-cJun signaling is required for VEGFA-induced HRMVEC migration, sprouting and tube formation in vitro and hypoxia-induced retinal endothelial cell proliferation, tip cell formation and neovascularization in vivo.

effects of growth factors and cytokines in endothelial cells 14,17 . Furthermore, STAT3 has been reported to be involved in the regulation of VEGF expression, tumor angiogenesis and metastasis 18 . It has also been shown that binding of hypoxia-inducible factor (HIF)-1α and STAT3 to VEGFA promoter is necessary for hypoxia-induced VEGFA expression 19 . In contrary to a large number of studies pointing a role for STAT3 in angiogenesis 18,19 , some reports have shown that STAT3 inhibits VEGFA-induced angiogenic events 20 . Given the presence of STAT3 in the retina and its involvement in tumor angiogenesis, we examined its role in retinal neovascularization.
In the present study, we report that Pyk2 via STAT3-cJun signaling mediates VEGFA-induced HRMVEC migration, sprouting and tube formation in vitro and hypoxia-induced retinal endothelial cell proliferation, tip cell formation and neovascularization in vivo.

Results
Pyk2 mediates VEGFA-induced HRMVEC migration, sprouting and tube formation. Both developmental and pathological angiogenesis are mostly governed by VEGFA 21 . Various non receptor tyrosine kinases such as Janus kinase 2 (Jak2) and Pyk2 are involved in VEGFA-induced endothelial cell migration and proliferation 22,23 . In order to understand the role of these non-receptor tyrosine kinases in VEGFA-induced angiogenic events, we have studied the time course effect of VEGFA on Pyk2 tyrosine phosphorylation in HRMVECs. VEGFA induced Pyk2 tyrosine phosphorylation in a time-dependent manner with maximum effect at 10 min (Fig. 1a). Furthermore, dominant negative mutant-mediated blockade of Pyk2 activation without affecting DNA synthesis attenuated VEGFA-induced HRMVEC migration, sprouting and tube formation (Fig. 1b-f). These results indicate that activation of Pyk2 is required for VEGFA-induced HRMVEC migration, sprouting and tube formation but not proliferation.

Hypoxia-induced retinal neovascularization requires Pyk2 activation.
To emphasize the significance of the in vitro findings in vivo, we next studied the role of Pyk2 in retinal neovascularization using a mouse model of oxygen-induced retinopathy (OIR). Mice pups were kept at 75% oxygen from P7 to P12 and at P12 they were returned to room air to develop the relative hypoxia. Retinal extracts were prepared at various time periods of hypoxia (starting from P12) and analyzed by Western blotting for Pyk2 tyrosine phosphorylation. Hypoxia induced tyrosine phosphorylation of Pyk2 in a time-dependent manner with maximum effect at 12 hrs (Fig. 2a). To validate hypoxic condition, we also measured HIF-1α levels. HIF-1α was induced by several-fold in response to hypoxia with a maximum effect at 12 hrs (Fig. 2b). To test the role of Pyk2 in retinal neovascularization, Pyk2 levels were depleted by its siRNA molecules and tested for its effects in hypoxia-induced retinal neovascularization. Downregulation of Pyk2 levels profoundly reduced hypoxia-induced retinal neovascularization as observed by a decrease in endothelial cell proliferation, a decrease in the number of tufts and filopodia and disappearance of blood vessel dilatation with increased avascular area (Fig. 2c-i). These findings indicate that Pyk2 activation is required for hypoxia-induced retinal EC proliferation, tip cell formation and neovascularization. STAT3 activation is required for VEGFA-induced HRMVEC migration, sprouting and tube formation. Since STAT3 plays a role in VEGFA-induced tumor angiogenesis 24 , and Pyk2 mediates epidermal growth factor-induced STAT3 activation 25 , we explored the potential role of Pyk2 in VEGFA-induced STAT3 activation in HRMVECs. VEGFA induced both serine and tyrosine phosphorylation of STAT3 in HRMVECs (Fig. 3a). Since Janus kinases (JAKs) mediate STAT3 activation in tumor angiogenesis, we first tested the effect of VEGFA on Jak1 and Jak2 tyrosine phosphorylation in HRMVECs. VEGFA had little or no effect on Jak1 or Jak2 tyrosine phosphorylation, suggesting lack of a role for these tyrosine kinases in STAT3 activation (Fig. 3b). We next tested the role of Pyk2 in VEGFA-induced STAT3 activation. Interestingly, dominant negative mutant-mediated blockade of Pyk2 activation inhibited VEGFA-induced STAT3 phosphorylation at both serine and tyrosine residues in HRMVECs (Fig. 3c). This observation indicates that Pyk2 mediates VEGFA-induced STAT3 activation. To understand the functional importance of STAT3 activation in VEGFA-induced angiogenic responses in HRMVECs, we studied the effects of its serine (S727A) and tyrosine (Y705F) mutants in VEGF-induced HRMVEC migration, proliferation, sprouting and tube formation. Interference with STAT3 activation by forced expression of its S727A or Y705F mutant without affecting DNA synthesis abolished VEGFA-induced HRMVEC migration, sprouting and tube formation ( Fig. 3d-g). To confirm these findings, we also tested the effect of its double mutant (Y705F/ S727A) in VEGFA-induced HRMVEC angiogenic responses. Blockade of STAT3 phosphorylation both at serine and tyrosine residues simultaneously without affecting DNA synthesis inhibited VEGFA-induced HRMVEC migration, sprouting and tube formation (Fig. 4a-e).
Pyk2-mediated STAT3 activation is required for hypoxia-induced retinal neovascularization. In order to understand the role of STAT3 in retinal neovascularization, we first tested its activation by OIR. Hypoxia induced both serine and tyrosine phosphorylation of STAT3 (Fig. 5a). To understand the mechanisms of STAT3 activation in hypoxic retina, we studied the role of Pyk2. Depletion of Pyk2 levels by its siRNA molecules without affecting HIF-1α levels reduced hypoxia-induced STAT3 phosphorylation both at serine and tyrosine residues suggesting that Pyk2 mediates STAT3 activation in hypoxic retina as well (Fig. 5b). To understand the role of STAT3 in retinal neovascularization, we used siRNA approach. Downregulation of STAT3 levels using its siRNA molecules significantly reduced retinal neovascularization with increased avascular area compared to mice injected with control siRNA (Fig. 5c-g). Depletion of STAT3 levels also attenuated retinal EC proliferation as well as tip cell formation (Fig. 5h,i). To rule out the possibility that the decreased retinal EC proliferation by STAT3 depletion was due to increased apoptosis, we measured cleaved poly (ADP-ribose) polymerase (PARP) levels, a marker of apoptosis 26 . Although hypoxia induced cleaved PARP levels, no significant differences were found in its levels between control and STAT3-depleted retinas (Fig. 5c). These observations confirm that (a) Quiescent HRMVECs were treated with and without VEGF (40 ng/ml) for the indicated periods and cell extracts were prepared. Equal amounts of protein from control and each treatment were analyzed by Western blotting for pPyk2 levels using its specific antibodies. The blot was reprobed for total Pyk2 or β -tubulin for normalization. (b) HRMVECs were transduced with Ad-GFP or Ad-dnPyk2 (40 moi), growth-arrested, treated with and without VEGF (40 ng/ml) for 10 min, cell extracts were prepared and equal amounts of protein from control and each treatment were analyzed for pPyk2 levels and the blot was reprobed for total Pyk2 or β -tubulin to show the over expression of dnPyk2 or normalization. (c-f) All the conditions were the same as in panel b except that after growth-arrest, cells were subjected to VEGFA-induced migration (c), DNA synthesis (d), sprouting (e), or tube formation (f) assays. The bar graphs represent quantitative analysis of three independent experiments. The values are presented as Mean ± SD. *p < 0.01 vs control or Ad-GFP; **p < 0.01 vs Ad-GFP ± VEGFA.  An equal amount of protein from control and various time periods of VEGFA (40 ng/ml)-treated HRMVECs was analyzed by Western blotting for pSTAT3, pJak1 and pJak2 levels and the blots were reprobed for total STAT3, Jak1, Jak2 or β -tubulin for normalization. (c) Cells were transduced with Ad-GFP or Ad-dnPyk2 (40 moi), growth-arrested, treated with and without VEGFA (40 ng/ml) for 10 min or 60 min, cell extracts were prepared and analyzed by Western blotting for pSTAT3 and pPyk2 levels and the blots were reprobed for total STAT3, Pyk2 or β -tubulin to show the over expression of dnPyk2 or normalization. (d) Upper panel: Cells were transfected with empty vector or mutant STAT3 (STAT3S727A or STAT3Y705F), growth-arrested, treated with and without VEGFA (40 ng/ml) for 10 min or 60 min, cell extracts were prepared and analyzed by Western blotting for pSTAT3 levels and the blots were reprobed for total STAT3 or β -tubulin to show the over expression of mutant STAT3 or normalization. Lower panel: All the conditions were the same as in upper panel except that after growth-arrest cells were subjected to VEGFA-induced cell migration assay. (e-g) All the conditions were the same as in panel d except that after growth-arrest cells were subjected to VEGFA-induced DNA synthesis (e), sprouting (f) or tube formation (g) assays. The bar graphs represent quantitative analysis of three independent experiments. The values are presented as Mean ± SD. *p < 0.01 vs control or vector or Ad-GFP; **p < 0.01 vs vector ± VEGFA or Ad-GFP ± VEGFA.
the decreased retinal EC proliferation by STAT3 downregulation was not due to increased EC apoptosis. These results indicate that STAT3 activation downstream to Pyk2 is required for retinal neovascularization.  except that pups were injected with control or STAT3 siRNA intravitreally at P12, P13 and P15 and at P17 the retinal extracts were analyzed for STAT3, PARP and β -tubulin levels to show the efficacy of the siRNA on its target and off target molecules or the enucleated eyes were fixed, cross-sections made and stained for CD31 and STAT3 or the retinas were isolated, fixed, stained with isolectin B4, flat mounts were made and examined for retinal neovascularization. Retinal vascularization is shown in the first column. Neovascularization is highlighted in red in the second column. The third column shows the selected rectangular areas of the images in the first column under 10x magnification. (e-g) Retinal vasculature (e), neovascularization (f) and avascular area (g) were determined as described in "Materials and Methods. " (h) All the conditions were the same as in panel c except that pups were injected with control or STAT3 siRNA intravitreally at P12 and P13 and at P15 the enucleated eyes were fixed, cross-sections made and stained for CD31 and Ki67. The right column shows the higher magnification of the selected areas in the left column images. (i) Conditions were the same as in panel c except that at P17 the eyes were enucleated, fixed, retinas isolated, stained with isolectin B4, flat mounts made and examined for EC filopodia formation. The first column shows the whole flat mounts and the second and third columns show the microscopic and digital magnifications of the selected rectangular areas in the first column. The bar graphs represent quantitative analysis of 6 retinas. The values were presented as Mean ± SD. *p < 0.01 vs normoxia; **p < 0.01 vs siControl + hypoxia.
Scientific RepoRts | 6:26480 | DOI: 10.1038/srep26480 Pyk2-STAT3 signaling enhances cJun expression in the mediation of retinal neovascularization. STAT3 mediates the growth promoting effects of angiogenic factors such as VEGFA. In addition, many reports showed that cJun/cFos interact with STAT3 in the transcriptional induction of several genes. To understand whether VEGFA-induced STAT3 activation has any role in cJun or cFos expression, we tested the time course effect of VEGFA on the expression of these proto-oncogenes. VEGFA induced cJun but not cFos expression in a time-dependent manner in HRMVECs (Fig. 6a). Next, we tested the role of Pyk2-STAT3 signaling in VEGFA-induced cJun expression. Dominant negative mutant-mediated inhibition of either Pyk2 or STAT3 activation blocked VEGFA-induced cJun expression (Fig. 6b). As expected, upon treatment with VEGFA, VEGFR2 levels were found to be downregulated, perhaps, due to its phopshorylation and internalization 27 . However, blockade of Pyk2-STAT3 signaling had no effect on VEGFR2 levels (Fig. 6b). To understand the functional aspects of cJun in VEGFA-induced angiogenic responses, we tested its role in VEGFA-induced HRMVEC migration, sprouting, and tube formation. Suppression of cJun levels by its siRNA molecules attenuated VEGFA-induced HRMVEC migration, sprouting, and tube formation (Fig. 6c-f).
To understand the role of cJun in hypoxia-induced retinal neovascularization, we have first studied the effect of hypoxia on its expression. Without affecting cFos levels, hypoxia induced cJun expression in the retina in a time-dependent manner (Fig. 7a). Furthermore, downregulation of Pyk2 or STAT3 levels by their siRNA molecules inhibited hypoxia-induced cJun expression (Fig. 7b,c). Downregulation of STAT3 levels had no effect in hypoxia-induced HIF-1α expression (Fig. 7c). Since cJun regulates ischemia-induced angiogenesis 28 , we next explored its role in hypoxia-induced retinal neovascularization. Depletion of cJun levels by its siRNA molecules while having no effect on HIF-1α expression, blunted hypoxia-induced retinal EC proliferation and filopodia formation leading to a reduction in retinal neovascularization with increased avascular area (Fig. 7d-j). Since inhibition of Pyk2-STAT3-cJun signaling does not affect VEGFA-induced HRMVEC proliferation but blocks hypoxia-induced retinal EC proliferation in mice, we asked the question whether these differential effects of Pyk2, STAT3 and cJun in the modulation of HRMVEC versus mouse retinal EC proliferation were due to species variations. To address this postulation, we have studied the role of Pyk2-STAT3-cJun signaling in VEGFA-induced mouse retinal microvascular endothelial cell (MRMVEC) proliferation. Dominant negative mutant-mediated Pyk2 or STAT3 activation or siRNA-mediated downregulation of cJun levels substantially inhibited VEGFA-induced DNA synthesis in MRMVECs (Fig. 7k).

Discussion
Retinal neovascularization is a common manifestation of various retinal disorders, including retinopathy of prematurity (ROP), proliferative diabetic retinopathy (DR) and age related macular degeneration 2,3,29 . All these ocular pathologies are associated with increased expression of angiogenic factors that stimulate neovascularization in the retina 30,31 . Most of these diseases are characterized by increased endothelial cell proliferation, vascular permeability and inflammation [2][3][4] . VEGFA is a pro-angiogenic cytokine that mediates all these processes and is the primary factor involved in neovascularization 32 . VEGFA signaling in retinal neovascularization is complex and involves various signal transduction pathways 7,12,23 . In the present study, we show that Pyk2-STAT3-dependent cJun expression plays a role in VEGFA-induced HRMVEC migration, sprouting and tube formation and hypoxia-induced retinal endothelial cell proliferation, tip cell formation and neovascularization. Since interference with activation of Pyk2, STAT3 or cJun had no effect on hypoxia-induced HIF-1α levels, it is likely that Pyk2-STAT3-cJun signaling plays a role downstream to, or independent of HIF-1α in hypoxia-induced retinal neovascularization. These findings provide the first conclusive evidence on the role of Pyk2-STAT3-cJun signaling in pathological retinal angiogenesis.
Pyk2 is a focal adhesion tyrosine kinase, which regulates multiple signaling events required for focal adhesion turnover and migration of cells 33 . Pyk2 knockout mice have exhibited defective endothelial cell migration and tubulogenesis following hind-limb ischemia 10 . The role of Pyk2 in endothelial cell sprouting has also been reported 34,35 . In addition, we have shown that Pyk2 is involved in endothelial tight junction disruption 36 . In the present study, we present evidence that Pyk2 mediates VEGFA-induced HRMVEC migration, sprouting and tubulogenesis and hypoxia-induced retinal endothelial cell proliferation, tip cell formation and neovascularization. However, it was noted that downregulation of Pyk2 levels only inhibited hypoxia-induced EC proliferation in the mouse retina, but not HRMVEC proliferation. These differential effects of Pyk2 in mouse versus human retinal microvascular endothelial cell proliferation may be attributed to species differences as clarified later in the "Discussion. " Pyk2 has been shown to mediate epidermal growth factor-induced STAT3 activation in HeLa cells 37 . In the present study, we demonstrate that Pyk2 mediates both VEGFA and hypoxia-induced STAT3 serine and tyrosine phosphorylation. Many studies have shown that STAT3 plays a role in angiogenesis, tumor development and metastasis in response to various stimulants [38][39][40] . It has also been demonstrated that STAT3 is activated by hypoxia in microglia, astrocytes, and neurons 41,42 . Using rat 50/10 oxygen-induced retinopathy model 43 , previous studies have shown that neovascular retinal vessels exhibit increased STAT3 activity 44 . Using HRMVECs and mouse OIR model, our present observations demonstrate that depletion of STAT3 levels significantly reduces endothelial cell migration, sprouting and tube formation in vitro and retinal endothelial cell proliferation, tip cell formation and neovascularization in vivo. Thus, while a large body of data indicates a role for STAT3 in angiogenesis [38][39][40] , some studies have reported that STAT3 activation blocks retinal vascularization 45 . It should be noted that if STAT3 was involved in vessel regression then one would expect that depletion of STAT3 levels results in enhanced vessel growth in response to hypoxia, which was not the case in the present study. Therefore, it is likely that STAT3 activation is required for enhanced neovascularization rather than vessel regression.
Upon activation, STAT3 translocates form the cytoplasm to the nucleus and increases the expression of its target genes involved in cell proliferation, migration, invasion or angiogenesis depending on the stimulus [46][47][48] . It has also been reported that cJun interacts with STAT3 and enhances IL-6 response element transactivation 49  to Pyk2-STAT3 signaling in the regulation of HRMVEC migration, sprouting and tube formation. A role for cJun in 15(S)-hydroxyeicosatetraenoic acid and epoxyeicosatrienoic acid-induced angiogenesis has also been  reported 26,50 . Corroborating these in vitro findings, cJun expression in hypoxic retina was dependent on both Pyk2 and STAT3 activation and its downregulation suppresses hypoxia-induced endothelial cell proliferation, tip cell formation and retinal neovascularization. Since inhibition of Pyk2-STAT3-cJun signaling only blocks VEGFA-induced MRMVEC but not HRMVEC proliferation, it should be pointed out that this signaling modulates EC growth in a species-specific manner. Many studies have shown that AP-1 plays a role in the regulation of expression of matrix metalloproteinases (MMPs) involved in cell migration and invasion [51][52][53] . Since cJun via homodimerization or heterodimerization with the members of the Fos family of proto-oncogene proteins can form an active AP-1 complex 54 , it is possible that cJun might be involved in the regulation of expression of MMPs in mediating VEGFA and hypoxia-mediated EC angiogenic responses. The reports that Pyk2 and STAT3 modulate the expression of MMPs 55,56 also support this view as Pyk2-STAT3 signaling axis is required for both VEGFA and hypoxia-induced cJun expression.
In summary, as depicted in Fig. 8, the present study demonstrates the importance of Pyk2-STAT3-dependent cJun expression in mediating retinal endothelial cell migration, proliferation, sprouting and neovascularization.

Adenoviral vectors.
The construction of Ad-GFP, Ad-dnPyk2, and Ad-dnSTAT3 was described previously 40,57,58 . Animals. C57BL/6 pregnant mice at E16 were obtained from Charles River Laboratories, Wilmington, MA and bred at University of Tennessee Health Science Center's vivarium. All the experiments involving animals were approved by the Animal Care and Use Committee of the University of Tennessee Health Science Center, Memphis, TN. The methods were carried out according to approved Institutional guidelines.

Cell culture. Human retinal microvascular endothelial cells (HRMVECs) were purchased from Applied Cell
Biology Research Institute (Kirkland, WA) and cultured in medium 131 containing microvascular growth supplements (MVGS), 10 μ g/ml gentamycin, and 0.25 μ g/ml amphotericin B. Cultures were maintained at 37 °C in a humidified 95% air and 5% CO 2 atmosphere. HRMVECs with passage numbers between 5 and 10 were synchronized in medium 131 without MVGS for 24 hrs and used to perform the experiments unless otherwise indicated.
Mouse retinal microvascular endothelial cell isolation. Eyes from 3 wk-old C57BL/6 mice pups were enucleated, retinas dissected out, minced, and digested in 4 ml of collagenase type I (1 mg/ml in serum-free DMEM) for 60 min at 37 °C. Following digestion, equal volume of DMEM containing 10% FBS was added and the cellular digest was filtered through a sterile 40 μ m nylon mesh and centrifuged at 1000 rpm for 5 min to pellet the cells. The cell pellet was washed twice with DMEM containing 10% FBS, resuspended in 1.0 ml of DMEM with 10% FBS and incubated with sheep anti-rat IgG magnetic beads pre-coated with rat anti-mouse CD31 antibody column. The third column shows the selected areas of the images in the first column under 10x magnification. (f-h) Retinal vasculature (f), neovascularization (g) and avascular area (h) were shown. (i) Pups were injected with the indicated siRNA intravitreally at P12 and P13 and at P15 eyes were fixed, cross-sections made and stained for CD31 and Ki67. The right column shows the higher magnification of the selected areas in the left column images. (j) Pups were injected with the indicated siRNA intravitreally at P12, P13 and P15 and at P17 eyes were enucleated, fixed, retinas isolated, stained with isolectin B4, flat mounts made and examined for EC filopodia formation. The first column shows the whole flat mounts and the second and third columns show the microscopic and digital magnifications of the selected areas in the first column images. (k) MRMVECs that were transduced with the indicated adenovirus or transfected with the indicated siRNA and growth-arrested were subjected to VEGFA-induced DNA synthesis. The bar graphs represent quantitative analysis of 6 retinas. The values were presented as Mean ± SD. *p < 0.01 vs normoxia; **p < 0.01 vs siControl + hypoxia; # p < 0.01 vs Ad-GFP or siControl; @ p < 0.01 vs Ad-GFP + VEGFA or siControl + VEGFA.
Scientific RepoRts | 6:26480 | DOI: 10.1038/srep26480 [the beads were washed three times with serum-free DMEM and incubated with rat anti-mouse CD31 antibody for overnight at 4 °C (10 μ l antibody/50 μ l beads in DMEM)]. After affinity binding, magnetic beads were washed six times with DMEM containing 10% FBS and the bound cells were plated onto a single well of a 24-well plate. Mouse retinal endothelial cells were cultured in EGM-2 medium containing 20% FBS at 37 °C in a humidified 95% air and 5% CO 2 and growth-arrested in EBM-2 medium without any supplements for 24 hrs to perform the experiments.
Cell migration. HRMVEC migration was measured by wound-healing assay. Briefly, endothelial cells were plated in ibidi culture insert at 2 × 10 4 cells/chamber, grown to full confluence and growth-attested in medium 131 without any growth supplements for 24 hrs. After the growth-arresting period, the ibidi inserts were removed and 1 ml of medium 131 containing 5 mM hydroxyurea was added. Cells were treated with and without VEGFA (40 ng/ml) for 24 hrs and the migrated cells were observed under Nikon Eclipse TS100 microscope with 4X/0.13 magnification and the images were captured with a Nikon Digital Sight DS-L1 camera. The cell migration was calculated using ImageJ and expressed as percent wound closure (total area at 0 hrs-area not occupied by the cells at 24 hrs/total area at 0 hrs X 100). 59 . HRMVECs were plated onto 6-well plates, allowed to grow to 70-80% confluence, growth-arrested for 24 hrs, treated with and without VEGFA (40 ng/ml) for 30 hrs and pulse-labeled with 1 μ Cl/ml of [ 3 H]-thymidine for the last 24 hrs of the 30-hrs incubation period. Cells were collected by trypsinization followed by centrifugation and the cell pellet was resuspended in cold 10% (w/v) trichloroacetic acid. After vortexing the mixture was kept on ice for 30 min and was then passed through a GF/F glass microfiber filter. The filter was washed once with cold 5% tricholoroacetic acid, once with cold 70% ethanol, dried, placed in a scintillation vial containing the scintillation cocktail and the radioactivity was measured in liquid scintillation counter (Beckman LS 3801). DNA synthesis was expressed as counts/min/dish. Tube formation. Three hundred microliters of HRMVEC suspension at a concentration of 5 × 10 5 cells/ml were added to each well of a twenty four-well culture plate that was pre-coated with 280 μ l of growth factor-reduced Matrigel. Cell were treated with vehicle or VEGFA (40 ng/ml) for 6 hrs at 37 °C. Tube formation was observed under an inverted microscope (Eclipse TS100, Nikon, Tokyo, Japan) and the images were captured using a CCD color camera (KP-D20AU: Hitachi, Ibaraki, Japan) and Apple iMovie 7.1.4 software. The tube length was calculated using NIH ImageJ and expressed in micrometers. Western blotting. Cell or tissue extracts containing an equal amount of protein were resolved by electrophoresis on 0.1% SDS and 10% polyacrylamide gels. The proteins were transferred to a nitrocellulose membrane and after blocking in either 5% nonfat dry milk or BSA, the membrane was probed with the appropriate primary antibodies followed by incubation with horseradish-peroxidase-conjugated secondary antibodies. The antigen-antibody complexes were detected using an enhanced chemiluminescence detection reagent kit (Amersham Biosciences).

DNA synthesis. DNA synthesis was measured by [ 3 H]-thymidine incorporation
Transfection. HRMVECs at 70% confluence in medium 131 lacking MVGS were transfected with control or target siRNA at a final concentration of 100 nM using Lipofectamine 2000 transfection reagent following the manufacturer's instructions. Six hours after transfection, medium was replaced with fresh medium 131 containing MVGS and antibiotics for 24 hrs and growth-arrested in medium 131 without any supplements for 24 hrs before using for experimentation.
Transduction. HRMVECs were transduced with control or target adenovirus at a final concentration of 40 moi in medium 131 containing MVGS and antibiotics for overnight. Medium was replaced with fresh medium 131 containing both MVGS and antibiotics and after 24 hrs of incubation cells were growth-arrested overnight in medium 131 without any supplements and used for experimentation.
Sprouting. Three-dimensional sprouting was performed as described previously 59 . Briefly, the transduced/ transfected HRMVECs were coated onto cytodex beads overnight and embedded in fibrin (2 mg/ml fibrinogen + 0.625 units/ml of thrombin) gels. Fibroblasts were plated on top of the fibrin gels and incubated at 37 °C for the indicated time periods. The beads were examined on day 3 and day 5 for sprouting under Ziess inverted microscope (Observer.Z1: original magnification X10/NA 0.45) and the fluorescence images were captured by Ziess AxioCam MRm camera using microscope operating and image analysis software AxioVision 4.7.2 (Carl Ziess Imaging Solutions GmbH). The number of sprouts on each bead were counted using NIH imageJ. Sprouting was expressed as number of sprouts/bead.

Oxygen-induced retinopathy (OIR).
OIR was performed and quantified as described by Connor et al. 60 .
C57BL/6 mice pups along with dams were exposed to 75% oxygen from P7 to P12 and at P12 they returned to room air to develop the relative hypoxia. Mice pups of the same age kept at room air were used as controls. At P17 mice pups were sacrificed, eyes were enucleated and fixed in 4% (w/v) paraformaldehyde for 1 hr at room temperature. Retinas were isolated, stained with isolectin B4, flat mounts were made, coverslips were placed and examined under a Zeiss inverted fluorescence microscope (Zeiss Observer. Z1). Retinal vasculature was quantified by calculating the ratio of fluorescence intensity to the total retinal area. Retinal neovascularization was measured by first setting a scale with a tolerance point of 50 pixels based on the fluorescence intensity in the screenshot using Nikon NIS-Elements software version AR 3.1. Neovascularity (values above 50 pixels tolerance set point) was highlighted in red and then quantified by dividing the fluorescence intensity in the highlighted area by the total fluorescence intensity in the screenshot (n = 6 eyes).
Intravitreal injections. During the indicated time periods of hyperoxia or hypoxia, pups were administered with control, or the targeted siRNA at 1 μ g/0.5 μ l/eye) by intravitreal injections using a 33G needle.
Immunofluorescence staining. After hyperoxia, mouse pups were returned to room air for 3 days, after which time they were sacrificed, eyes enucleated, fixed in optimal cutting temperature compound, and cryosections were prepared. To identify proliferating ECs, after blocking in normal goat serum, the cryosections were probed with rabbit anti-mouse Ki67 antibodies (1:100) and rat anti-mouse CD31 antibodies (1:100) followed by incubation with Alexa Fluor 568-conjugated goat anti-rabbit and Alexa Fluor 488-conjugated goat anti-rat secondary antibodies. The sections were observed under Zeiss inverted microscope (Observer.Z1: original magnification X40/NA 0.6 or original magnification X10/NA 0.45) and the fluorescence images were captured by AxioCam MRm camera using the microscope operating and image analysis software AxioVision 4.7.2 (Carl Zeiss Imaging Solutions GmbH). The retinal EC proliferation was quantified by counting Ki67-and CD31-positive cells that extended anterior to the inner limiting membrane per section (n = 6 eyes, 3 sections/eye).