Search

Circulating cells of several lineages are thought to participate in angiogenesis and tumor growth. Experimental studies in tumor-bearing mice have pointed to the potential importance of VEGF-responding circulating (endothelial) progenitor cells in tumor growth.We have studied circulating CID31- and/or CD34-positive cell populations with a low to moderate VEGFR2 expression in human volunteers and cancer patients. We recognized four cell populations, which were further characterized by their content of major hematopoetic progenitor, monocytic, endothelial and platelet markers. After establishing the test-retest stability of the measurements in nine patients, we determined the frequencies of the various cell populations in a group of 20 volunteers and 14 cancer patients. Two populations were markedly increased in cancer patients. Small CD45neg/CD34bright/VEGFR2+ cells amounted to 12 and 64 cells/ml (P+/CD45bright (/CD14+) monocytic cells were present in controls and cancer patients, respectively (P=0.0 17). A third population of CD45dim/CD34bright/VEGFR2low cells amounted to 25 and 30 cells/ml (P=0.38). Unexpectedly, a population of mainly anucleated CD45low/CD31bright/CD41bright cells was present in numbers of 9,076 and 16,697/ml (P=0.04) in volunteers and cancer patients, which contained a VEGFR2low (compared to IgG isotype control) expressing population amounting to 1,142 and 1,642 cells/ml (P=0.12). This fourth population probably reflects large platelets. The role of the herein identified VEGFR2+ circulating cell populations deserve further investigation in cancer patients treated with VEGF(R)-targeted therapies. Quantification of such cell populations in the blood of tumor patients may be valuable to monitor the efficacy of anti-angiogenic treatment.

SummaryIn diabetes mellitus an increased risk exists for vascular complications. A role for advanced glycation endproducts (AGEs) in the acceleration of vascular disease has been suggested. Nε(carboxymethyl)lysine (CML)- and methylglyoxal (MGO)-modified proteins have been identified as majorAGEs. The interaction of these AGEs with the human endothelial cells and macrophages was studied. Changes in adhesion molecule expression, i. e. vascular cell adhesion molecule-1 (VCAM-1), intercellular adhesion molecule-1 (ICAM-1) and E-selectin were determined by cell-bound Elisa on human endothelial cells after incubation with CML-modified albumin and MGO-modified albumin. The presence of the full-length receptor of AGEs (RAGE) and splice variants of RAGE was determined by specific RT-PCR. In addition, binding studies were performed with CML- and MGOmodified albumin to endothelial cells and P388D1 macrophages. We demonstrated that CML-albumin or MGO-albumin did not induce activation of endothelial cells as measured by the expression of adhesion molecules, while, under the same conditions, TNF-α did. No specific binding of CML-albumin and MGO-albumin on these cells was found. In contrast to endothelial cells, a specific binding of MGO-albumin to P388D1 macrophages was demonstrated, which could be competed by ligands of scavenger receptors. In human umbilical vein and microvascular endothelial cells we found the N-truncated and C-truncated splice variants of RAGE.In conclusion, under our experimental conditions no CML- or MGO-albumin-induced increase in adhesion molecule expression was found on endothelial cells. In agreement with this, no binding of these AGEs was found to endothelial cells. The existence of splice variants of RAGE in endothelial cells might explain the lack of interaction of extracellular AGEs with these cells.

SummaryThe transmigration of leukocytes across the endothelium is a prerequisite for the inflammatory process. Leukocyte-endothelium interaction is regulated by several endothelial adhesion molecules, such as intercellular adhesion molecule-1 (ICAM-1), vascular cell adhesion molecule-1 (VCAM-1) and E-selectin. Their expression is enhanced by inflammatory mediators, such as TNFα. In vivo a small part of these adhesion molecules is shed by proteases, and can be detected in the circulation. In this study, the time course of the TNFα-induced accumulation in the blood of three circulating soluble adhesion molecules, sE-selectin, sICAM-1 and sVCAM-1, was studied in plasma samples of tumor patients enrolled in a phase I trial receiving TNFα. Two different cohorts were studied. Eleven patients received a continuous 24-hour TNFα infusion while 10 other patients received a 5-day continuous TNFα infusion. After 24 hours of TNFα infusion sE-selectin levels increased by 1985 ± 312 %, sVCAM-1 by 301± 19 % and sICAM-1 by 445 ± 82 %. Differences in accumulation patterns were observed after 5 days of continuous TNFα infusion. sVCAM-1 and sICAM-1 levels showed an increase during the infusion with a maximum at 3 to 5 days and stayed elevated after discontinuation of the TNFα infusion. In contrast, sE-selectin reached its peak at day 1 and declined thereafter. In conclusion, sVCAM-1 and sICAM-1 show a different accumulation pattern upon TNFα infusion as compared to sE-selectin in man.

SummaryThe aminopeptidase inhibitor bestatin has been shown to have anti-angiogenic effects in a number of model systems. These effects are thought to result from inhibition of CD13 activity. Because tumor angiogenesis can evolve in a fibrin-rich stroma matrix we have studied for the first time the effects of bestatin on microvascular endothelial capillary-like tube formation in a fibrin matrix. Bestatin enhanced the formation of capillary-like tubes dose-dependently. Its effects were apparent at 8 µM; the increase was 3.7-fold at 125 µM; while high concentrations (>250 µM), that were shown to have anti-angiogenic effects in other systems, caused extensive matrix degradation. Specific CD13-blocking antibodies WM15 and MY-7, and the aminopeptidase inhibitors amastatin and actinonin also enhanced capillary-like tube formation (maximally 1.5-fold), but these effects did not reach statistical significance. The effect of bestatin was not due to a change in uPAR availability because the relative involvement of the u-PA/u-PAR activity was not altered by bestatin. In view of the present findings we hypothesize that aminopeptidases other than CD13 predominantly contribute to the observed pro-angiogenic effect of bestatin in a fibrin matrix. The identification of this novel effect of bestatin is important in the light of the proposed use of bestatin as anti-angiogenic and/or anti-tumor agent.

SummaryAmong other proteolytic enzymes, the urokinase-type plasminogen activator (u-PA)/plasmin cascade contributes to cell migration and the formation of capillary-like structures in a fibrinous exudate. The u-PA receptor (u-PAR) focuses proteolytical activity on the cell surface of the endothelial cell and hereby accelerates the pericellular matrix degradation. Vascular endothelial growth factor (VEGF) and fibroblast growth factor (FGF)-2 enhance u-PA receptor expression in human endothelial cells. In this paper we show that the protein kinase C (PKC) inhibitors Ro31-8220 and GF109203X inhibit VEGF165-induced u-PAR antigen expression in human endothelial cells, whereas PKC inhibition had no effect on FGF-2-induced u-PAR antigen enhancement. In addition, inhibition of PKC activity had no effect on VEGF165-or FGF-2-induced proliferation in human endothelial cells. We conclude that VEGF165 induces u-PAR via a PKC-dependent pathway, whereas proliferation is induced via a different pathway probably involving tyrosine phosphorylation of proteins downstream of the VEGF receptors.

SummarySmooth muscle cell migration plays a role in the development of intimal hyperplasia. Given the established role of the plasminogen activation system in cell migration, an approach to therapy is to overexpress an inhibitor of plasmin. Therefore, an adenoviral vector was constructed encoding the hybrid protein ATF.BPTI, which contains the active domain of bovine pancreas trypsin inhibitor (BPTI), fused to ATF, the amino terminal fragment or receptor-binding domain of u-PA. Adenoviral vectors expressing ATF and BPTI individually were also constructed, and a fourth vector was constructed encoding ATF.BPTI linked by an internal ribosomal entry site to Green Fluorescent Protein (ABIG). Both the expression and functionality of the recombinant proteins were established in human vascular smooth muscle cells. Adenoviral gene transfer of ATF.BPTI inhibited SMC migration more efficiently than the expression of ATF or BPTI individually. Expression of ABIG resulted in the co-expression of ATF.BPTI and Green Fluorescent Protein, thereby providing a tool to monitor transfection efficiency and the behavior of the transfected cells.

SummaryDysfunction of the vascular endothelium is considered an early step in the development of diabetic angiopathy. Hyperglycaemia results in endothelial dysfunction, both through direct effects of glucose and through formation of advanced glycosylation end-products (AGEs). We hypothesized that the effects of glucose and AGEs on endothelial function in insulin-dependent diabetes mellitus (IDDM) are distinct and are reflected by distinct plasma markers of endothelial function. We therefore measured plasma levels of von Willebrand factor (vWF), soluble (s) E-selectin and vascular cell adhesion molecule-1 (sVCAM-1), and evaluated the relationship with HbA1c and urinary excretion of pentosidine, an AGE product, in 56 patients with IDDM. Urinary pentosidine excretion was higher in the diabetic than in a control group (n = 60) of similar age (P

SummaryOver a decade ago the idea of endothelial cell seeding was introduced in an attempt to improve the function of small caliber vascular prostheses. Although endothelial cell seeding is currently being applied clinically, several questions regarding the functional properties of the seeded endothelial cells remain. Evaluation of functional properties of endothelial cells on various types of vascular prostheses can be performed partly in vitro, but it is hampered by the fact that commonly used methods to quantify endothelial cells do not adequately apply to these cells on prosthetic materials.An accurate quantification method is described that is rapidly and easily applicable to endothelial cells attached to vascular prostheses. The method can also be used to quantify endothelial cells attached to culture dishes or microcarriers. Colorless, non-fluorescing, fluorescein-di-acetate was used, which was taken up by the attached endothelial cells, and which was then intracellularly converted to yellow fluorescein, emitting green fluorescence. Subsequently, triton-X-100 was appli-cated to release fluorescein and levels of fluorescence were measured with the automated aperture-defined microvolume (ADM) method, using an inverted fluorescence microscope to which a photometer was connected. The measured level of fluorescence is linearly related to endothelial cell numbers attached to prostheses. The accuracy and the reproducibility of cell countings are high.

SummaryThe report describes the effect of an in vitro infection of human umbilical vein endothelial cells with human Cytomegalovirus (CMV). The parameters studied are cellular procoagulant activity, secretion of plasminogen activator inhibitor (PAI-1) and urokinase-type plasminogen activator (u-PA), activation and internalization of factor X and Merocyanine 540 staining.The infection does not result in an increase in PAI-1 and u-PA secretion, but it brings about a procoagulant response, which is relatively rapid compared to the tissue factor mediated response induced by inflammatory mediators. The time course and the coagulation factor dependency suggest a facilitated interaction of coagulation factors on the surface of infected cells. Chromogenic activity measurements after the addition of purified factor X and electron microscopic examination of the cells after addition of colloidal gold-factor X conjugates both point to an internalization of factor X and/or Xa after interaction with the endothelial cell surface. Merocyanine 540 staining suggests that CMV infection leads to membrane pertubations.

SummaryIn previous kinetic studies, the catalytic efficiency of the activation of human coagulation factors IX and X by factor VIIa in the presence of purified tissue factor apoprotein was found to be essentially equal. These activation reactions were now studied on the surface of human umbilical vein endothelial cells. The cells were stimulated with endotoxin to express tissue factor. This tissue factor activity was saturable with factor VIIa and could be inhibited by rabbit antibodies against human tissue factor apoprotein. Only stimulated cells supported factor VIIa activity. No difference in the reactivity of factor VII and VIIa was observed in the presence of factor X, due to rapid feedback activation of factor VII by factor Xa. However, the activation of factor IX by factor VII shows a 10 min lag-phase, which reflects that the activation of factor VII by factor IXa is a less efficient process. The kinetic parameters for the factor VIIa dependent activation of factor IX and factor X on the endothelial surface were: Km 0.09 εM, Vmax 0.13 pmol/min, and Km 0.071 εM, Vmax 0.41 pmol/min, respectively. The same ratio between the Vmax for factor X and factor IX activation was observed as in a cell free system. However, the Km of factor IX was 4-fold higher on the endothelial surface than in the cell free system. Together, these kinetic parameters will favour factor X activation 5-fold over factor IX activation at physiological concentrations of these proteins.The activation of factor X by factor VIIa on the endothelial surface was characterized by a short lag-phase, which was absent in factor IX activation. Further, both the activation of factor X and factor IX were down regulated by factor Xa.

Human foreskin microvascular endothelial cells synthesize and release tissue-type plasminogen activator (t-PA) in similar amounts as do endothelial cells from umbilical cord artery and vein. Human thrombin increases the production of t-PA by these cells, which could be visualized from 8 h after addition of 0.1-5 units/ml thrombin by fibrin autography after SDS polyacrylamide gel electrophoresis of the endothelial cell conditioned media. Thrombin also increased the secretion of t-PA antigen. Together with t-PA, human microvascular cells release urokinasetype plasminogen activator (u-PA) antigen and endothelial cell type PA inhibitor, PA inhibitor-1, which were both demonstrated by specific immunoprecipitation from radiolabeled endothelial cell conditioned medium. Thrombin increases the release of u-PA antigen, but no u-PA activity could be demonstrated. Thrombin induced a two-fold stimulation of the synthesis and secretion of PA inhibitor-1 antigen. At 0.1 unit/ml thrombin also an increase in PA inhibitor activity was found. At high concentrations of thrombin a decrease of PA inhibitor activity was found, due to the conversion of the active 46 kD PA inhibitor-1 into a 42 kD product without PA inhibitor activity. Our data indicate that interaction of thrombin with microvascular endothelial cells will shift the balance between t-PA, u-PA and PA inhibitor-1, and thus affects the regulation of fibrinolysis. Chemicals/CAS: plasminogen activator inhibitor, 105844-41-5; thrombin, 9002-04-4; Glycoproteins; Plasminogen Activators, EC 3.4.21.-; Plasminogen Inactivators; Thrombin, EC 3.4.21.5

SummaryThe effects of human activated protein C (APC) and thrombin on plasminogen activator inhibitor (PAI-1) released from cultured human umbilical endothelial cells, grown in serum-free 35S-methionine containing medium, were studied in two ways: 1) measurement of PAI-1 activity with an amidolytic assay, and 2) immunoprécipitation of the medium with anti-PAI-1 IgG, antiprotein C IgG or anti-thrombin IgG followed by SDS-PAGE and autoradiography.Addition of APC or thrombin to the endothelial cell conditioned medium results in a time and concentration dependent loss of PAI-1 activity and in the degradation of PAI-1 from 46 kD into a 42 kD product.After incubation of the medium with APC in the presence of cells, an additional band of 95 kD was found, which could be immunoprecipitated with both anti-PAI-1 IgG and anti-protein C IgG, indicating the formation of an APC-PAI-1 complex before degradation occurs. No complex could be detected after incubation of the medium with thrombin in the presence of endothelial cells.Blocking the active sites of APC and thrombin prevented both the formation of APC-PAI-1 complexes and the inactivation and degradation of PAI-1. After removal of the active PAI-1 from the medium, no degradation of the inactive PAI-1 by APC or thrombin could be found.It is concluded that both APC and thrombin react with the active PAI-1, resulting in inactivation and degradation of PAI-1.

SummaryA model has been developed to study the transport of fluid and macromolecules through human arterial umbilical cord endothelial cell monolayers in vitro. Cells were cultured on fibronectin- coated polycarbonate filters and formed within a few days a tight monolayer, with an electrical resistance of 17 ± 4 Ohm · cm2. The cells were connected by close cell contacts with tight junctions. The passáge-rate of horse radish peroxidase (HRP) through these filters was 20-40 fold lower than through filters without an endothelial monolayer. The continuous presence of 10% human serum was needed to maintain the electrical resistance of the monolayer and its barrier function towards macromolecules. Chelation of extracellular calcium resulted in an increased permeability and a decreased electrical resistance of the monolayer. This process was reversible by re-addition of calcium ions to the cells. The permeation rate of dextrans of various molecular weights (9-480 kD) was inversely related to the molecular mass of the molecule. No difference was measured between the passage rate of dextran of 480 kD and dextran of 2,000 kD. Incubation of the endothelial cell monolayer with 2-deoxy-D-glucose resulted in a decreased permeability but it had no effect on electrical resistance. This suggests that the passage-process is energy- dependent.Fluid permeation through the endothelial cell monolayer on filters was measured in a perfusion chamber under 20 mmHg hydrostatic pressure. It was decreased by the presence of serum proteins and responded reversibly on the chelation and readdition of extracellular calcium ions.