Nitric oxide synthase lies downstream from vascular endothelial growth factor-induced but not basic fibroblast growth factor-induced angiogenesis.

Systemic administration of the nitric oxide (NO) synthase inhibitor Nomega-nitro--arginine methyl ester (L-NAME) to rabbits bearing a corneal implant blocked vascular endothelial growth factor (VEGF), but not basic fibroblast growth factor (bFGF)-induced angiogenesis. L-NAME completely blocked angiogenesis induced by VEGF-transfected MCF-7 breast carcinoma cells and the cells remained dormant in the cornea. Postcapillary endothelial cell migration and growth induced by VEGF were blocked by both the NO synthase inhibitor Nomega-mono-methyl--arginine and by the guanylate cyclase inhibitor LY 83583. We conclude that NO is a downstream imperative of VEGF-, but not bFGF-induced angiogenesis, and propose that the NO synthase/guanylate cyclase pathway is a potential target for controlling tumor angiogenesis in response to VEGF. Our studies support recent evidence that VEGF and bFGF induce angiogenesis by different mechanistic pathways using the alphavbeta5 and alphavbeta3 integrins, respectively.

Nitric oxide mediates angiogenesis in vivo and endothelial cell growth and migration in vitro promoted by substance P.

We evaluated the effects of nitric oxide (NO) generators and endogenous production of NO elicited by substance P (SP) in the angiogenesis process. Angiogenesis was monitored in the rabbit cornea in vivo and in vitro by measuring the growth and migration of endothelial cells isolated from coronary postcapillary venules. The angiogenesis promoted in the rabbit cornea by [Sar9]-SP-sulfone, a stable and selective agonist for the tachykinin NK1 receptor, and by prostaglandin E1 (PGE1), was potentiated by sodium nitroprusside (SNP). Conversely, the NO synthase inhibitor N omega-nitro-L-arginine methyl ester (L-NAME), given systemically, inhibited angiogenesis elicited by [Sar9]-SP-sulfone and by PGE1. Endothelial cells exposed to SNP exhibited an increase in thymidine incorporation and in total cell number. Exposure of the cells to NO generating drugs, such as SNP, isosorbide dinitrate, and glyceryl trinitrate, produced a dose-dependent increase in endothelial cell migration. Capillary endothelial cell proliferation and migration produced by SP were abolished by pretreatment with the NO synthase inhibitors N omega-mono-methyl-L-arginine (L-NMMA), N omega-nitro-L-arginine (L-NNA), and L-NAME. Exposure of the cells to SP activated the calcium-dependent NO synthase. Angiogenesis and endothelial cell growth and migration induced by basic fibroblast growth factor were not affected by NO synthase inhibitors. These data indicate that NO production induced by vasoactive agents, such as SP, functions as an autocrine regulator of the microvascular events necessary for neovascularization and mediates angiogenesis.

Visualization of fluorescent erythrocytes in the microcirculation.

Erythrocytes loaded internally with FITC-BSA can be readily visualized in the microcirculation by using a television camera and a fluorescence microscope. Flow properties of the erythrocytes and their adherence to the vascular endothelium or erythrophagocytic cells can be observed. This procedure should also be useful to delineate the microcirculation under circumstances where infused free FITC-BSA can escape into the interstitial tissue.

Permeability of intestinal microvessels in chronic arterial hypertension.

We studied changes in intestinal microvascular permeability resulting from chronic arterial hypertension. Normotensive dogs and dogs made chronically hypertensive utilizing the one-kidney, one clip Goldblatt technique were used to obtain values for: arterial pressure, portal pressure, intestinal lymph flow, and the lymph-to-plasma protein concentration ratio (CL/Cp). Values for the normotensive dogs were 111 mm Hg, 7.1 mm Hg, 6.2 ml/hr, and 0.64, respectively, while values for the chronically hypertensive dogs were 165 mm Hg, 7.3 mm Hg, 12.5 ml/hr, and 0.66, respectively. Control lymph flow in the hypertensives was 100% greater than in the normotensives, while there was no significant difference in control CL/Cp between the two groups. When portal venous pressure was acutely increased to 30 mm Hg, lymph flow increased to approximately the same maximum value in both groups. This represents an eightfold increase in normotensive and a fourfold increase in hypertensive lymph flows. The reflection coefficient determined as 1 - (CL/Cp) for total proteins at maximum lymph flow was 0.78 for the normotensives and 0.56 for the hypertensives. An electrophoretic analysis indicated sieving of large molecular weight protein fractions was considerably reduced in the hypertensives when compared to the normotensives. Our results indicate a significant increase in intestinal microvascular permeability to macromolecules resulting from the one-kidney, one clip Goldblatt model of chronic arterial hypertension.

Power dissipation as a measure of peripheral resistance in vascular networks.

Peripheral resistance was examined in the microcirculation of the rat cremaster muscle using a network-conserved parameter, power dissipation. Previous studies of peripheral resistance used network-sensitive parameters, and their interpretation is limited by tacit assumptions about the structure of the peripheral vasculature. Power dissipation is directly linked to the resistive process, providing a measure of resistance based on the actual hemodynamics of the network. The dissipation parameter was quantified with the usual vascular parameters of velocity and vessel segment length; 991 segment lengths were measured in 12 normotensive Wistar-Kyoto rats and 16 spontaneously hypertensive rats. Arterial power dissipation was significantly elevated over a wide range of vessel segments; blood flow ranged from 0.08 to 80 nl/sec. Since the largest vessels showed the greatest power dissipation, the organ resistance elevation seen in hypertension in the cremaster apparently is mediated by the larger vessels in the high flow range. Vessel segment length and number of dissipative vessels were unchanged. The increase in power dissipation was due to a network-averaged reduction in mean vessel diameter. Power dissipation also increased significantly in the fastest flowing venous microvessels (greater than 25 nl/sec), also due to a reduction in vessel segment diameter.

Acute local effects of angiotensin II on the intestinal vasculature.

The vasotoxic properties of angiotensin II (AII) on vascular endothelial cells have been implicated in the increased extravasation of proteins and water observed in certain forms of chronic arterial hypertension. Since microvascular permeability in the small intestine is increased in one-kidney, one clip hypertension, we tested the hypothesis that AII decreases the permselectivity of intestinal microvessels to plasma proteins. The acute intestinal vascular effects of AII were studied by locally infusing AII into isolated canine jejunal segments. A measure of vascular permeability in control and infused segments was obtained by estimating the osmotic reflection coefficient as 1-lymph/plasma protein concentration ratio when the ratio reached a plateau at high lymph flows. Lymph flows were increased by raising the venous pressure from a control of 5 to 30-35 mm Hg in 5 mm Hg steps. Resting control blood and lymph flows were reduced by AII, but these flows were not different at the higher venous pressures due to the apparent blunting of the venous-arteriolar response by AII. The estimated osmotic reflection coefficient of 0.85 for the control animals was less than that obtained in the infused segments (0.93). This effect was substantiated by electrophoretic separation of protein fractions. Thus, AII per se does not cause an acute increase in intestinal vascular permeability, and may, in fact reduce it.

Autoregulation and vasoconstriction in the intestine during acute renal hypertension.

The purpose of this study was to investigate whether local mechanisms of blood flow autoregulation mediate vasoconstriction during the early development of renal hypertension. Anesthetized rats were instrumented with Doppler flow probes on the celiac (CA), superior mesenteric (SMA), and renal arteries to measure flow velocity in these vessels. Acute two-kidney, one clip renal hypertension was produced by inflating a pneumatic occluder on the left renal artery to reduce flow velocity by 50%. Two hours after renal artery stenosis (RAS), femoral artery pressure (AP) was increased by 35%, CA resistance by 45%, and SMA resistance by 57%. No increases were observed in AP or in CA and SMA resistances for sham-operated, control rats. To determine if autoregulation contributed to the increase in SMA resistance, we protected the SMA vasculature from the increased arterial pressure by servocontrolled inflation of a pneumatic cuff implanted around the SMA. Although normalizing SMA pressure with the protective cuff significantly reduced (p less than 0.05) the increase in SMA resistance that occurred after RAS, SMA resistance remained elevated above control levels. These results suggest that (1) reduced intensity of SMA constriction produced by protection of the SMA is due to inhibition of a local autoregulatory mechanism that is contributing to the increase in SMA resistance during the acute development of renal hypertension, and (2) maintenance of elevated SMA resistance during protection from increased AP is the result of pressure-independent mechanisms that are activated subsequent to renal artery stenosis.

Hemodynamic characteristics of the intestinal microcirculation in renal hypertension.

This study investigated the microvascular changes that affect vascular resistance in the rat small intestine during two-kidney, one clip renal hypertension 4 weeks after renal artery stenosis. To study the intestinal microcirculation, a loop of the small intestine was exteriorized with intact circulation and innervation and a section of the bowel wall was prepared for observation with an intravital video microscopy system. Microvascular diameter, pressure, and flow velocity were measured for first, second, and third branch order arterioles and venules, using an image shearing monitor, servo-null micropipette system, and an optical Doppler velocimeter, respectively. The diameters of the first order arterioles and venules were significantly (p less than 0.05) reduced in hypertensive rats; however, diameters were unaltered in smaller second and third order arterioles and venules as compared with normotensive vessels. In hypertensive rats, mean arterial pressure was significantly (p less than 0.05) elevated (47%) and pressures also were elevated significantly (p less than 0.05) throughout the microcirculation, although by a proportionally smaller amount. Total network flow (i.e., first order arteriole flow) was significantly (p less than 0.05) reduced (40%) in hypertensive rats, but volume flows in individual second and third order arterioles were similar to flows measured in normotensive rats. Calculated total network resistance was increased (124%) in hypertensive rats. Thus, the intestinal microcirculation in rats with two-kidney, one clip renal hypertension is disturbed by elevated pressure and decreased total flow. The presence of normal flows in individual second and third order arterioles without any demonstrable difference in their diameters suggests that the predominant cause of elevated resistance across this segment of the intestinal microcirculation is a reduction in the number of perfused small arterioles.

Distinct capillary density and progression promoted by vascular endothelial growth factor-A homodimers and heterodimers.

The aim of this study was to characterize the capillary density, progression and persistence of new capillaries induced by different isoforms of vascular endothelial growth factor (VEGF)-A. They were produced and purified using the same protocol and assessed in the same experimental model, the rabbit cornea assay. Monogenic homodimers for VEGF121 and VEGF165 together with the heterodimer VEGF121/165 were tested as slow release polymer pellets implanted into the avascular rabbit cornea and examined up to 18 days post-implantation. The implants consistently stimulated angiogenesis in the absence of inflammation. The VEGF121 isoform produced the strongest increase of new capillary vessels which rapidly and persistently progressed into the corneal stroma. VEGF165 promoted the growth of a smaller number of capillaries which ten-ded to regress over time. Heterodimers of VEGF121/165 produced intermediate in vivo activities between the two homodimers. In vitro endothelial cell proliferation, mobilization and adhesion were promoted by all VEGF isoforms under serum-free or serum-reduced conditions with the same order of potency. Anti-soluble KDR (sKDR) antibody completely inhibited the effects of all the isoforms. These results indicate that monogenic homodimer preparations of VEGF121 and VEGF165 can display distinct biological effects which are functionally retained after the heterodimeric assembly of VEGF121 and VEGF165. The observed different biological behavior of the VEGF isoforms reveals the possibility that in vivo the assembly of dimers derived from splicing of a single gene may yield molecules with either different matrix or receptor interaction, stability or diffusion rate according to specific needs.

The effect of linomide on the migration and the proliferation of capillary endothelial cells elicited by vascular endothelial growth factor.

In order to assess the mechanism of action of the quinoline-3-carboxyamide linomide as an antiangiogenic drug, the effect of linomide was studied in vitro on postcapillary endothelial cells exposed to vascular endothelial growth factor (VEGF). Linomide did not block the spontaneous replication of endothelial cells, but significantly suppressed endothelial cell growth and migration elicited by VEGF. It is concluded that linomide appears to be an effective tool to inhibit VEGF-dependent angiogenesis.

B1 receptor involvement in the effect of bradykinin on venular endothelial cell proliferation and potentiation of FGF-2 effects.

1. Bradykinin (BK) contributes to the inflammatory response inducing vasodilation of postcapillary venules and has been demonstrated to induce neovascular growth in subcutaneous rat sponges. 2. In this study the ability of BK to stimulate cell growth and migration in cultured endothelium from coronary postcapillary venules (CVEC) has been investigated. 3. [3H]-thymidine incorporation in subconfluent and synchronised CVEC was used to monitor DNA synthesis over 24 h. BK promoted a concentration-dependent increase of DNA synthesis with maximal activity at 100 nM. At this concentration BK also induced 18 fold accumulation of c-Fos protein immunoreactivity in the nucleus within 1 h from peptide exposure. 4. The total number of cells recovered after 48 h exposure to BK was increased in a concentration-dependent manner. Maximal effect was produced by 100 nM concentration of the peptide which produced 50% increase in cell number. The selective B1 receptor agonist Des-Arg9-BK mimicked the proliferative effect of BK, while the B2 receptor agonist kallidin was devoid of any activity. The proliferation induced by BK was abolished in a concentration-dependent manner by the addition of the B1 selective antagonist Des-Arg9-Leu8-BK, while the selective B2 receptor antagonist HOE140 did not modify BK-induced growth. 5. DNA synthesis and growth promoted by a threshold concentration of fibroblast growth factor-2 (FGF-2) (0.25 nM) were potentiated by increasing concentrations of BK and Des-Arg9-BK. 6. Endothelial cell migration assessed by the Boyden Chamber procedure was not promoted by BK or the selective B1 and B2 receptor agonists. 7. These data are the first demonstration that BK promotes growth of endothelial cells from postcapillary venules. The mitogenic activity of BK involves c-Fos expression and potentiates the growth promoting effect of FGF-2. Only the B1 receptor appears to be responsible for the proliferation induced by BK and suggests that this type of receptor might be implicated in favouring angiogenesis of coronary venules.

Endothelial cell migration is induced by soluble P-selectin.

We have assessed the role of soluble P-selectin in promoting endothelial cell migration. Human endothelial cells (HUVEC) and bovine microvascular endothelial cells (CVEC) were assessed for migration in the Neuroprobe 48-well microchemotaxis chamber. Soluble P-selectin promoted a dose-dependent (0.1-10 nM) migration of both cell types, with maximal response at 10 nM, producing approximately 60% increment over basal migration. Anti-P-selectin monoclonal antibody (5 microg/ml) selectively blocked P-selectin induced migration. Fibronectin and collagen were essential to disclose the migration induced by P-selectin. It is suggested that at vascular level in the presence of modifications of the extracellular matrix milieu, the production of soluble P-selectin could contribute to angiogenesis by promoting endothelial cell migration.

Substance P increases cyclic GMP levels on coronary postcapillary venular endothelial cells.

The vasodilating effect of substance P (SP) at the microvascular level is endothelium-dependent. In the present study we evaluated whether SP activates nitric oxide (NO) production by venular endothelial cell. We evaluated NO activation by measuring cyclic GMP levels in cultured endothelial cells isolated from coronary postcapillary venules of bovine origin (CVEC). Our results indicate that 5 min exposure of CVEC to 10 nM SP doubled basal cyclic GMP levels. Cell treatment with the NO synthase inhibitor L-NMMA reduced the basal levels of cyclic GMP and abolished the effect of SP but did not modify the increase in cyclic GMP in response to exogenous NO. These data indicate that a) microvascular endothelium responds in an autocrine fashion to NO with increased cyclic GMP levels, b) SP activates cyclic GMP pathway through NO production.

Nuclear accumulation of exogenous basic fibroblast growth factor in endothelial, fibroblast, and myoblast cell lines results in diverse biological responses.

During studies comparing 125I-bFGF internalization between endothelial cells and other cell types, we found, unexpectedly, internalization and nuclear translocation of exogenously added 125I-bFGF in two cell lines: Chinese hamster ovary cells (CHO) and rat L6 myoblasts. These cell lines were previously reported to be devoid of FGF receptors. Furthermore, CHO cells showed a weak mitogenic response to added bFGF, while L6 cells were mitogenically unresponsive. By comparison, coronary venular endothelial cells (CVEC), BALB/c 3T3 fibroblasts, and BHK-21 cells, demonstrated internalization and nuclear translocation of added 125I-bFGF, and mitogenic responsiveness to the growth factor. Insulin alone stimulated DNA synthesis in all cell types, yet augmented bFGF-dependent DNA synthesis only in CVEC, 3T3, and BHK. All five cell types expressed FGF receptors as assessed by covalent crosslinking with 125I-bFGF and immunoblotting with anti-FGF receptor antibodies. Differing rates of cytoplasmic and nuclear accumulation of 125I-bFGF and partial inhibition of internalization by pretreatment of CVEC with chlorate support a recent model that bFGF can internalize by two mechanisms. Insulin did not significantly affect 125I-bFGF internalization or metabolism in any cell type. bFGF treatment resulted in weak inhibition of RNA synthesis in L6 cells. bFGF appears firmly bound to the nuclear matrix as little nuclear-bound 125I-bFGF in CVEC is released by DNAse I or RNAse A digestion, while washes with 0.5 M NaCl result in partial release. Nuclear bFGF may thus be involved in regulation of nuclear events (e.g., gene transcription and/or DNA replication).

Immunochemical comparison of peptide angiogenic factors.

A panel of 40 monoclonal antibodies was constructed in response to cationic endothelial cell growth factor (c-ECGF), the cationic peptide mitogen isolated from endothelial mitogen. The monoclonal antibodies were assayed by dot blot for immunoreactivity to various other peptide angiogenic factors. The panel of monoclonal antibodies to c-ECGF exhibited complete cross-reactivity with pituitary fibroblast growth factor and sarcoma-derived growth factor. A group of 28 monoclonal antibodies was found to exhibit reactivity to anionic endothelial mitogen (a-ECGF), brain fibroblast growth factor, endothelial cell growth factor, and retina-derived growth factor. None of the monoclonal antibodies was found to react with epidermal growth factor or platelet-derived growth factor. These data provide an immunological basis for grouping heparin-binding endothelial cell growth factors into anionic and cationic groups.

Lymph flow transients following elevation of venous pressure in the dog hindpaw.

Lymph flow transients were studied in a dog paw preparation when venous pressure was elevated by 15 and 25 mmHg. The lymph flow transients showed a very rapid initial increase which then declined to a steady-state value that was one-half the peak lymph flow response for both pressure changes. Lymph flow increased in the initial 5.3 minutes following venous pressure elevation to 10.4 +/- 2.0 and 18.1 +/- 4.5 times the normal lymph flow (mean +/- standard deviation) for the 15 and 25 mmHg increases in venous pressure, respectively. However, approximately 9 minutes after attaining the maximal flow rate, the lymph flow declined to only 5.5 +/- 0.7 and 9.8 +/- 1.8 times the control values. These data demonstrate another condition in which lymph flow is not maintained at the maximal capability. Possible mechanisms causing the observed biphasic lymph flow response to capillary pressure elevation are: 1) changes in Starling forces oppose an increase in capillary pressure; 2) the rate of change in tissue fluid pressure affects lymph flow to a greater extent than does the absolute change in tissue fluid pressure; or, 3) the lymphatics empty upon elevation and refill as the capillaries filter.