INTRODUCTION

Top Abstract Introduction Materials and methods Results Discussion References

ANGIOGENESIS REPRESENTS a complex physiological response of vascular cells to physical or chemical stimuli. In hypoxic or injured tissues, vascular endothelial growth factor (VEGF) is released and initiates a cascade of events including endothelial barrier opening, endothelial cell migration and proliferation, and tube formation (27). Recent experiments have demonstrated that the production of nitric oxide (NO) comprises a common mechanism underlying VEGF-stimulated morphological and functional changes in endothelial cells (1, 9, 11, 21, 22, 27). Our previous study in isolated and perfused coronary venules showed that the NO synthase (NOS) inhibitor N^G-monomethyl-L-arginine (L-NMMA) prevented the increases in albumin permeability induced by VEGF (29). Although NO seems to be a critical signaling molecule in the mediation of different cellular responses to the growth factor, the sequence of events that occurs upstream from NO production has not been elucidated. Within this context, it remains unclear whether the intracellular molecules that have long been known to participate in the inflammatory hyperpermeability response (7, 12, 16, 17, 32), such as phospholipase C (PLC), endothelial cytosolic Ca²⁺ ([Ca²⁺]_i), and protein kinase C (PKC), are located in the postreceptor signaling pathway of VEGF.

VEGF binds to two different receptor tyrosine kinases, designated Flt-1 and Flk-1/KDR (27). Although information about Flt-1 is limited, VEGF stimulation of KDR is known to result in phosphorylation of the Src family of protein kinases and the -isoform of phospholipase C (PLC-) (10, 15). KDR transduction has been implicated in the development of endothelial actin reorganization, membrane ruffling, and chemotactic contraction (28). These morphological changes are often observed in endothelial monolayers undergoing barrier changes in response to inflammatory mediators, implying that KDR-triggered intracellular cascade may be involved in the permeability response to VEGF.

In this study, we hypothesized that VEGF binding to the KDR receptor phosphorylated PLC-, which subsequently elevated the intracellular Ca²⁺ as well as upregulated PKC, leading to NOS activation and NO production. We measured the endothelial permeability to albumin and the [Ca²⁺]_i response to VEGF in intact perfused coronary venules. Moreover, immunoprecipitation followed by Western blot analysis was performed using cultured vascular endothelial cells to verify the phosphorylation status of the signaling molecules. The results support a KDR-PLC-Ca²⁺/PKC-ecNOS cascade in the venular hyperpermeability reaction elicited by VEGF.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and methods Results Discussion References

Materials. An albumin-physiological salt solution (APSS) was used as a bathing solution while the microvessels were being dissected. It contained the following (in mM): 145.0 NaCl, 4.7 KCl, 2.0 CaCl₂, 1.17 MgSO₄, 1.2 NaH₂PO₄, 5.0 glucose, 2.0 pyruvate, 0.02 EDTA, and 3.0 MOPS buffer. After 1% bovine serum albumin was added, the solution was buffered to a pH of 7.40 at 4°C and then filtered through a Millex-PF 0.22-µm filter unit (Millipore, Bedford, MA). The APSS used to perfuse the vessels during permeability measurements had the same composition as mentioned above but was buffered to a pH of 7.40 at 37°C. The chemicals used to make the perfusate, including FITC-albumin, were purchased from Sigma (St. Louis, MO). Bovine serum albumin was obtained from United States Biochemical (Cleveland, OH). Cell culture supplies including the culture media DMEM and fetal bovine serum were from GIBCO (Gaithersburg, MD). VEGF was from R & D System (Minneapolis, MN). The PLC inhibitor U-73122 and the PKC inhibitor bisindolylmaleimide (BIM) were ordered from Calbiochem (San Diego, CA). Antibodies including monoclonal anti-PLC-, polyclonal antiphosphotyrosine, and polyclonal antiphosphoserine were from Transduction Laboratory (Lexington, KY). Monoclonal antibody against the extracellular domain of the VEGF receptor KDR was from Sigma, and anti-IgG conjugated with horseradish peroxidase was from Jackson ImmunoResearch Laboratory (West Grove, PA). Fura 2 was ordered from Molecular Probes (Eugene, OR).

Isolation and perfusion of coronary venules. Pigs weighing 9-13 kg were anesthetized with pentobarbital sodium (30 mg/kg iv) and heparinized (250 units/kg iv). After a tracheotomy and intubation, the animal was ventilated with room air. A left thoracotomy was performed, and the heart was electrically fibrillated, excised, and placed in 4°C physiological saline. The coronary sinus was cannulated, and 5 ml india ink-gelatin-physiological salt solution were infused to clearly define venular microvessels. This solution was prepared by adding 0.2 ml india ink (Koh-I-Noor, Bloomsbury, NJ) and 0.35 g porcine skin gelatin to 10 ml warm physiological salt solution and filtering through P8 filter paper (Fisher Scientific, Pittsburgh, PA). Information regarding the validation and limitation of the ink-perfusion procedure has been provided in our previous publications (12, 31). The method for isolating and cannulating coronary venules has also been described in detail in these studies. Briefly, a suitable venule (length 0.8-1.2 mm, diameter 20-60 µm) was dissected from the surrounding myocardium in a dissecting chamber containing APSS at 4°C with the aid of a Zeiss stereo dissecting microscope. The vessel was transferred to a cannulating chamber that was mounted on a Zeiss axiovert microscope. The isolated vessel was cannulated with an inflow and outflow micropipette on each end and secured with 11-0 suture (Alcon, Fort Worth, TX). A third smaller pipette was inserted into the inflow pipette. The vessel was perfused with either APSS through the outer inflow pipette or APSS containing FITC-albumin through the inner inflow pipette. Each cannulating micropipette was connected to a reservoir, and the vessel was perfused at a relatively constant intraluminal pressure and flow rate. The bath solution in the chamber was maintained at 37°C and pH 7.4 throughout the experiments. The image of the vessel was projected onto a Hamamatsu charged coupled device-intensified camera and was displayed on a high-resolution monochromtic video monitor and recorded onto a VHS video recorder. Diameter of the vessel was measured on-line with a video caliper (Microcirculation Research Institute, Texas A & M University, College Station, TX).

Measurement of venular permeability. The permeability of the vessel was measured with a fluorescence ratio technique (13). With the use of an optical window of a video photometer positioned over the venules and adjacent space on the monitor, the fluorescent intensity from the window was measured and digitized on-line by a Power Macintosh computer. In each measurement, the isolated venule was first perfused with APSS through the outer inflow pipette to establish a baseline intensity. The venular lumen was then rapidly filled with APSS containing FITC-albumin by switching the perfusion to the inner inflow pipette. This produced an initial step increase, followed by a gradual increase, in fluorescence intensity. There was a step decrease of intensity when the fluorescently labeled molecules were washed out of the vessel lumen by switching the perfusion back to the outer inflow pipette. The apparent solute permeability coefficient of albumin (P_a) was calculated using the equation P_a = (1/I_f)(dI_f /dt)_o(r/2), where I_f is the initial step increase in fluorescent intensity, (dI_f /dt)_o is the initial rate of gradual increase in intensity as the fluorescently labeled solutes diffuse out of the vessel into the extravascular space, and r is the venular radius.

10

10

5

5

Measurement of [Ca²⁺]_i in isolated venules. Changes in [Ca²⁺]_i were measured using a fluorescence ratio technique with the aid of a microscope photometry system (Photon Technology International), which consisted of a PowerFilter high-speed dual-wavelength illuminator, a high-grade quartz fiber-optic bundle, a D-104B single-channel microscope photometer, and a 710 photon-counting photomultiplier tube (PMT). The cell-permeable form of a fluorescent dye, fura 2-AM, was used as the Ca²⁺ indicator. To load the endothelium with the indicator, the porcine venule was cannulated as described above and perfused for 20 min at a pressure gradient of 20 cmH₂O through the inner inflow pipette with APSS containing 5 µM fura 2-AM and 0.05% DMSO. After being loaded, the dye remaining in the extracellular space was washed off by switching the perfusion to the outer inflow pipette containing APSS for 20 min. The loading and washing procedures were performed in the dark at room temperature. The cannulated vessel was then aligned on the optical axis of a Zeiss Axiovert microscope that was connected to the PMT equipped with the photometry system. Emission fluorescence at 510 nm during excitation at 340 and 380 nm was detected by the PMT photon-counting system through the measuring window positioned over the venule and recorded with FeliX, computer software associated with the photometer system. In each experiment, the excitation wavelength alternated between 340 and 380 nm at a rate of 650/s, and the fluorescence intensities at both wavelengths were recorded for 10 min in 5-s intervals. Background correction was automatically controlled by the program through subtraction of the background values from the measured values in real time. Because the wall of isolated venules mainly consisted of endothelial cells, the contribution of other cells to the fluorescence signals was small and therefore ignored. The ratio of fluorescent intensities at 340 vs. 380 nm was calculated with the computer program and presented as an index of the intracellular Ca²⁺ level.

10

6

6

Cell culture and treatment. Human umbilical vein endothelial cells (HUVEC) were ordered from Clonetics (San Diego, CA). Cells were routinely maintained in gelatin-coated dishes containing EGM-2 culture media with 2% fetal bovine serum (Clonetics). Protein assays indicated that confluent HUVEC grown on a 60-mm dish contained proteins at a level of 150-170 µg.

10

5

5

4

5

5

10

Immunoprecipitation and Western blot analysis. After the above treatment, confluent cell monolayers in 60-mm dishes were lysed by incubation for 20 min in 1 ml of ice-cold lysis buffer (20 mM Tris · HCl, pH 7.4, 150 mM NaCl, 1% Triton X-100, 1 mM phenylmethylsulfonyl fluoride, 0.15 units/ml aprotinin, 10 µg/ml leupeptin, 10 µg/ml pepstatin A, 1 mM sodium orthovanadate, and 10% glycerol). The lysate was clarified by centrifugation at 14,000 g for 10 min at 4°C. For immunoprecipitation, the lysate was incubated overnight at 4°C with monoclonal antibodies directed against specific proteins, such as anti-PLC- and anti-ecNOS. The extract was then incubated with protein G plus Sepharose 4B (Zymed, San Francisco, CA) for 2 h at 4°C. The immunocomplex was collected by centrifugation at 15,000 g for 10 s and washed three times with cold immunoprecipitation buffer containing 0.1% Triton X-100 and once with 10 mM Tris · HCl at pH 7.4.

Data analysis. In the intact vessel studies, P_a was measured two to three times for each venule at each experimental intervention, and the values were averaged. For all experiments, n is given as the number of vessels studied, with each vessel representing a separate animal. For each experimental condition, the values of P_a from different venules were averaged, normalized to the control values obtained before drug treatments, and reported as percent control in means ± SE. ANOVA was used to evaluate the significance of intergroup differences. A value of P < 0.05 was considered significant for the comparisons.

	RESULTS

Top Abstract Introduction Materials and methods Results Discussion References

VEGF-induced permeability changes in venules. In control venules ranging from 20 to 60 µm in diameter (n = 6), VEGF (10¹⁰ M) significantly increased P_a from the basal value of 3.6 ± 0.7 × 10⁶ to 7.5 ± 0.9 × 10⁶ cm/s, which was 211.8 ± 14.3% of the basal permeability (Fig. 1). Blockage of the VEGF receptor KDR or inhibition of the intracellular signaling molecules PLC and PKC abolished the hyperpermeability effect of VEGF. As shown in Fig. 1, in venules treated with the KDR receptor antibody at 1 µg/ml (n = 9), VEGF (10¹⁰ M) did not significantly increase P_a; P_a values were 2.7 ± 0.2 × 10⁶ cm/s before VEGF and 3.1 ± 0.5 × 10⁶ cm/s after VEGF. Moreover, the venules treated with the selective PLC antagonist U-73122 at 10⁵ M displayed no hyperpermeability response to VEGF as indicated by a P_a value of 2.3 ± 0.3 × 10⁶ cm/s before and 2.3 ± 0.4 × 10⁶ cm/s after VEGF (n = 7). Similarly, administration of VEGF did not cause significant increases in venular permeability in the presence of the highly selective PKC inhibitor BIM at 10⁵ M (P_a was 3.1 ± 0.3 × 10⁶ cm/s before VEGF vs. 3.4 ± 0.3 × 10⁶ cm/s after VEGF, n = 7) (Fig. 1). The results indicated that KDR, PLC, and PKC are involved in the mechanism underlying VEGF's effect on venular permeability.

View larger version (30K): [in this window] [in a new window]   Fig. 1.   Topical application of vascular endothelial growth factor (VEGF; 1010 M) significantly increased albumin permeability (Pa) in intact perfused coronary venules under control conditions. Treatment of vessels with monoclonal antibodies directed against VEGF receptor KDR (1 µg/ml), selective phospholipase C inhibitor U-73122 (105 M), or specific protein kinase C inhibitor bisindolylmaleimide (BIM; 105 M) blocked VEGF-induced hyperpermeability. § Significant difference vs. basal. * Significant difference vs. control in presence of VEGF.

VEGF-elicited Ca²⁺ response in venules. The change in the endothelial [Ca²⁺]_i in response to topical application of VEGF (10¹² to 10⁹ M) was measured in intact isolated coronary venules. In 12 vessels studied, all of them responded to bradykinin (10⁶ M, Fig. 2A) and ionomycin (10⁶ M, Fig. 2B) with typical kinetics. However, an increase in [Ca²⁺]_i was observed in only four vessels with VEGF at the dose of 10¹⁰ M (Fig. 2, C-F).

View larger version (27K): [in this window] [in a new window]   Fig. 2.   Changes in content of endothelial cytosolic Ca2+ ([Ca2+]i) in intact perfused coronary venules. A: typical [Ca2+]i response to bradykinin (106 M) from 12 vessels. B: typical [Ca2+]i response to ionomycin (106 M) from 12 vessels. C-F: individual [Ca2+]i transients from 4 different venules upon administration of VEGF (1010 M).

VEGF-stimulated phosphorylation of PLC and ecNOS. Western blot analyses showed an increase in the phosphotyrosine content of PLC- at 3 and 5 min after administration of 10¹⁰ M VEGF (Fig. 3, top). This effect was not observed in cells treated with the selective PLC inhibitor U-73122 (Fig. 3, middle and bottom), indicating the specificity of the inhibitor. Treatment of the cells with the anti-KDR antibodies before VEGF stimulation blocked the hyperphosphorylation of PLC-, but inhibition of PKC with BIM or blockage of ecNOS with L-NMMA did not significantly reduce the phosphorylation of PLC to VEGF (Fig. 3, middle). The treatments did not significantly alter the protein level of PLC- (Fig. 3, bottom). The data suggest that KDR served as a tyrosine kinase receptor to trigger PLC- tyrosine phosphorylation, whereas PKC and ecNOS are downstream from PLC activation. Furthermore, the same dose of VEGF was found to cause no changes in tyrosine phosphorylation but an increase in serine phosphorylation of ecNOS (Fig. 4). The hyperphosphorylation of serine in ecNOS was greatly attenuated by treating the cells with either U-73122 or BIM, suggesting that the activation of PLC and PKC occurred before ecNOS phosphorylation upon VEGF stimulation.

View larger version (39K): [in this window] [in a new window]   Fig. 3.   Effect of VEGF (1010 M) on tyrosine phosphorylation of phospholipase C (PLC)-. Top: confluent cultures of human umbilical vein endothelial cells were stimulated with vehicle (Cont) and VEGF for 3 or 5 min and lysed. Lysates were immunoprecipitated (IP) with a monoclonal antibody directed against PLC- and then blotted (IB) with an antiphosphotyrosine antibody. First lane on left (P Cont) was a positive control using manufacturer-provided samples. Middle: cells were treated with vehicle (A), anti-KDR (B), U-73122 (C), BIM (D), or NG-monomethyl-L-arginine (L-NMMA; E) and then stimulated with VEGF. Immunoprecipitation followed by Western blot was performed in cell lysates as described above. To test protein level, PLC- immunoprecipitates were blotted with anti-PLC- antibodies. Bottom: densitometric quantification of bands in middle panel. OD, optical density. Data are representative of 3 independent experiments.

View larger version (37K): [in this window] [in a new window]   Fig. 4.   Top: serine phosphorylation of endothelial constitutive nitric oxide synthase (ecNOS) in human umbilical vein endothelial cells under control conditions (Cont) and after treatment with VEGF at 1011 M (A) and 1010 M (B). In other groups, cells were first incubated with BIM (C) or U-73122 (D) and then subject to VEGF (1010 M) stimulation. Cell lysates were immunoprecipitated with an anti-ecNOS antibody and blotted with an antiphosphoserine antibody to detect serine phosphorylation of ecNOS or blotted with an anti-ecNOS antibody to detect ecNOS protein level. Bottom: densitometric quantification of bands in top panel. Data are representative of 3 independent experiments.

	DISCUSSION

Top Abstract Introduction Materials and methods Results Discussion References

We previously reported that VEGF induced an increase in venular permeability through the production of NO. This study further examined the role of the VEGF receptor KDR and several postreceptor signaling molecules in the transduction of microvascular permeability response to VEGF. There are three new findings: 1) the effect of VEGF on endothelial barrier function is triggered by VEGF binding of KDR receptor and subsequent activation of PLC; 2) PKC is a major intermediate between activated PLC and ecNOS, although [Ca²⁺]_i may also be involved; and 3) VEGF activates PLC and ecNOS by upregulating PLC- tyrosine phosphorylation and ecNOS serine phosphorylation. From our current and previous findings (28), we suggest that VEGF increases microvascular permeability via a signaling cascade initiated by KDR activation of PLC, mediated by PKC upregulation and/or [Ca²⁺]_i elevation, and culminated with ecNOS-catalyzed NO production.

The central role of NO in mediating the angiogenic action of VEGF has recently been documented. NO is considered to be a chemoattractant and mitogen to bovine coronary venular endothelial cells (34). In HUVEC, VEGF promotes cell growth and formation of networklike structure through a NO-dependent mechanism in which long-term exposure to VEGF increases ecNOS protein level and short-term exposure stimulates the release of NO (11, 21). Direct measurements of NO production confirm that VEGF produces a dose- and time-dependent rise in NO concentration from various vascular segments or cultured endothelial cells (1, 11, 33). Consistent with the mitogenic action, the VEGF effect on microvascular permeability is linked to NO production, for VEGF-induced protein leakage can be prevented by the NOS inhibitors as shown in isolated venules studies (29) and in vivo Mile's assays (9). Although the NO cascade seems to serve as a common pathway leading to VEGF-stimulated endothelial proliferation and hyperpermeability, information is limited regarding the signal transduction from cell membrane to cytosol before NO production.

Chemical regulation of endothelial barrier function is mediated by soluble agonist occupancy of PLC-associated receptors (16, 24, 32). Two typical intermediates that are downstream of activated PLC are endothelial cytosolic Ca²⁺ and PKC (12, 16). For histamine-type inflammatory agonists, PLC hydrolysis of phosphotidylinositol bisphosphate and subsequent Ca²⁺ mobilization are the primary triggers of transendothelial flux of macromolecules (16, 24, 27). The relationship between [Ca²⁺]_i and NO production is well established based on the fact that elevated Ca²⁺ is a potent stimulator of the constitutive NO synthase (25). On the other hand, a group of agonists, such as phorbol esters, bradykinin, and platelet-activating factor, alters endothelial barrier function via a PKC-dependent mechanism (12, 14, 17, 20). The underlying enzymatic reaction involves PLC-catalyzed diacylglycerol formation and subsequent PKC activation (16). Recent evidence indicates a potential linkage between PKC and NOS in that PKC can activate NOS to produce NO in different types of cells (18, 26).

In this study, we tested whether the KDR-PLC-Ca²⁺/PKC-ecNOS cascade contributed to the venular permeability response to VEGF, a unique hyperpermeability factor that is well known for its tyrosine kinase receptor activity (27). Our finding that the anti-KDR antibody blocked VEGF's effects on venular permeability and PLC- tyrosine phosphorylation emphasizes the importance of KDR receptor. In support of the KDR hypothesis, experiments show that KDR-expressing cells display striking changes in cell morphology, actin reorganization, membrane ruffling, chemotaxis, and mitogenicity upon VEGF stimulation, whereas Flt-1-expressing cells lack such responses (28). Adding to this is that the K_d value of Flt-1 is in the picomolar range, whereas KDR has a K_d value at the subnanomolar level (5); the latter is consistent with the permeability dose response observed in our studies. Evidence is accumulating that VEGF binding to its tyrosine kinase receptor KDR results in tyrosine phosphorylation of various signaling proteins (10, 15, 22). Among the effector proteins, PLC- has been identified as important for its ability to propagate the signal to the downstream events characterized by Ca²⁺ mobilization and PKC upregulation. In this regard, PLC- is activated through tyrosine phosphorylation by VEGF (27), whereas KDR may serve as the primary kinase receptor that phosphorylates PLC- directly or indirectly through other SH2-containing molecules (10, 27).

The intracellular Ca²⁺ study, in which only one-third of venules responded to VEGF with Ca²⁺ spikes, argues against the primary or sole role of Ca²⁺ in VEGF action. Previous work on cultured endothelial cells reported that VEGF induced a transient elevation of [Ca²⁺]_i, which was interpreted as a PLC-stimulated internal Ca²⁺ release followed by a predominant external Ca²⁺ influx (4, 6). However, a recent study on intact perfused frog mesenteric microvessels (3) demonstrated that in eight vessels with typical Ca²⁺ responses to ionomycin, VEGF produced a significant increase in four, a small increase in two, and no response in [Ca²⁺]_i in two vessels. Apparently, the Ca²⁺ kinetics vary in different preparations under different experimental conditions. Compared with the above studies, the Ca²⁺ response to VEGF in coronary venules occurred in a smaller fraction of vessels. Taking the data from the current observation together with our previous results (29), the majority (>90%) of venules responded to the same concentration of VEGF with an increased permeability, but only one-third exhibited a Ca²⁺ transient. Thus the elevation of [Ca²⁺]_i does not seem to be a necessary step in the permeability response to VEGF. However, the presence of Ca²⁺ is required for the maintenance of basal barrier function and NOS activity (7, 26).

In contrast to Ca²⁺, PKC may play a role in mediation of VEGF's effect on vascular permeability, as indicated by the efficient inhibition of VEGF-induced protein leakage by a selective PKC inhibitor. Although our current technique did not allow direct quantification of PKC activity in intact perfused microvessels, the pharmacological approaches combined with the in vitro Western blot analysis provided evidence for the role of PKC as an integral component in transducing the VEGF signal from PLC to ecNOS. Indeed, other studies on bovine aortic endothelial cells (30) showed that VEGF induced a concentration- and time-dependent increase in PKC activity, and the effect was preceded by the activation of PLC-. Within this context, a parallel increase in PLC- tyrosine phosphorylation, inositol phosphate production, and diacylglycerol formation was detected in VEGF-stimulated cells (30). In retina, administration of a PKC-selective inhibitor did not affect histamine-induced but inhibited at 95% level VEGF-induced microvascular leakage (2). With regard to the signals downstream of PKC, a potential effector protein is ecNOS. As a serine kinase, PKC may induce serine phosphorylation of ecNOS and subsequent NO production, which in turn leads to venular hyperpermeability. This hypothesis is supported by a correlation of the current finding that inhibition of PKC attenuated the VEGF-stimulated serine phosphorylation of ecNOS with our previous data that L-NMMA greatly attenuated the permeability response to PKC activators (12). Adding to this is a report of in vivo measurements of microvascular permeability in which the NOS inhibitor antagonized PKC-regulated macromolecular transport (23). More recently, serine phosphorylation of ecNOS was detected in the hamster cheek pouch treated with platelet-activating factor, a PKC-dependent hyperpermeability mediator (8, 14).

In summary, this study reports that VEGF stimulation of endothelial cells upregulates tyrosine phosphorylation of PLC- and serine phosphorylation of ecNOS through a KDR receptor-triggered pathway. Consistently, the effect of VEGF on venular permeability is blocked during inhibition of KDR, PLC, PKC, or ecNOS, suggesting that the KDR-PLC-Ca²⁺/PKC-ecNOS cascade is involved in signaling mechanisms underlying the microvascular hyperpermeability response to VEGF.