• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Inhibition of bFGF/EGF-Dependent Endothelial Cell Proliferation By the Hyaluronan-Binding Protease From Human Plasma

Posted on: Friday, 30 April 2004, 06:00 CDT

Adhesion - bFGF - EGF - HABP - HUVEC - proliferation

Recently we identified a plasma serine protease with a high affinity to glycosaminoglycans like heparin or hyaluronic acid, termed hyaluronan-binding protease (HABP). Since glycosaminoglycans are found on cell surfaces and in the extracellular matrix a physiological role of this plasma protease in a pericellular environment was postulated. Here we studied the influence of HABP on the regulation of endothelial cell growth. We found that HABP efficiently prevented the basic fibroblast growth factor/epidermal growth factor (bFGF/EGF)-dependent proliferation of human umbilical vein endothelial cells. Proteolytic cleavage of adhesion molecules was found to be involved, but was not solely responsible for the anti-proliferative activity. Pre-treatment of growth factor- supplemented cell culture medium with HABP indicated that no direct contact between the active protease and cells was required for growth inhibition. In vitro studies revealed a growth factor- directed activity of HABP, resulting in complexation and partial hydrolysis and, thus, inactivation of basic fibroblast growth factor, a potent mitogen for endothelial cells. Heparin and heparan sulfate fully protected bFGF from complexation and cleavage by HABP, although these glycosaminoglycans are known to enhance the proteolytic activity of HABP. This finding suggested that free circulating bFGF rather than bFGF bound to heparan sulfate proteoglycans would be a physiologic substrate. In conclusion, down- regulation of bFGF-dependent endothelial cell growth represents an important mechanism through which HABP could control cell growth in physiologic or pathologic processes like angiogenesis, wound healing or tumor development.

Abbreviations. bFGF Basic fibroblast growth factor. - EGF Epidermal growth factor. - FN Fibronectin. - GAGs Glycosaminoglycans. - HABP Hyaluronan-binding protease. - HS Heparan sulfate. - HSPG Heparan sulfate proteoglycan. - MTT 3-[4,5 Dimethyldiazol-2-yl]-2,5-diphenyl tetrazolium bromide. - SFM Serum- free medium.

Introduction

The hyaluronan-binding protease (HABP) was identified in human plasma in its activated form during fractionation of vitamin K- dependent coagulation factors and was purified to homogeneity (Hunfcld et al., 1999). The 65-kDa two-chain enzyme represents the active form of a plasma hyaluronanbinding protein (Choi-Miura et al., 1996). It was later shown that a protease identical to HABP could in vitro activate coagulation factor VH and plasminogen activators (Romisch et al., 1999, 2000), thus attributing procoagulatory and profibrinolytic potencies to this enzyme. In plasma HABP circulates as a single-chain zymogen and in vitro undergoes rapid intcrmolecular autoactivation accelerated by glycosaminoglycans (GAGs) (Etscheid et al., 2000). More recently we demonstrated that HABP cleaves kininogen similar to plasma kallikrein releasing activated kininogen and the vasoactive pcptide bradykinin (Etscheid et al., 2002). The affinity of HABP to hyaluronic acid, heparin or heparan sulfate, as well as to matrix proteins like laminin, fibronectin, or collagen type I (Kre et al., 2002) indicated that the yet unknown physiological function of HABP is directed to cell surfaces and to the extracellular matrix.

Vascular endothelial cells (ECs) form a cell monolayer that lines blood vessels and functions as a barrier between the vessel lumen and the adjacent tissues. During vascular remodelling after vessel injury or during sprouting of new capillaries from existing vessels (angiogenesis) ECs are activated and induced to proliferate and migrate. These processes are regulated by various growth factors, such as vascular endothclial growth factor (VEGF) and basic fibroblast growth factor (bFGF or FGF-2) (Asahara et al., 1995; Cines et al., 1998). Angiogencsis has found a great deal of attention because it is associated with tumor progression. Attracting ECs to tumorigcnic tissue is pivotal to cancer growth, and, therefore, identifying mechanisms that control the proliferation and migration of ECs are of great interest in developing anti-cancer strategies. In this study we investigated the influence of HASP on the growth of cndothclial cells. It is demonstrated that HABP is an effective inhibitor of EC proliferation, and mechanisms underlying this phenomenon are presented. The possible physiologic relevance of these findings is discussed.

Materials and methods

HABP was prepared as described earlier (Hunfeld et al., 1999). Trypsin, horseradish peroxidase-linked secondary antibodies, the MTT reagent 3-[4,5 dimethyldiazol-2-yl]-2,5-diphenyl tetrazolium bromide, gelatine (2%) and bovine serum albumin (BSA) were purchased from Sigma (Deisenhofcn, Germany). All plastic equipment used in cell culture were from NUNC (Wiesbaden, Germany). Fibronectin (FN) was purchased from Bcclon Dickinson (Heidelberg, Germany). All other reagents were obtained from Gibco/Invitrogen (Karlsruhe, Germany) or Merck Eumlab (Darmstadt, Germany).

Human endotheliol tell isolotion

Human umbilical vein endothelial cells (HUVECs) were isolated from umbilical cords as described by Busse and Lamontagne (1991), a modification of the enzymatic method described by Jaffe et al. (1973). Sterile conditions were maintained throughout umbilical vein manipulation. In brief, the vein of the umbilical cord was cannulatcd and perfused with 30 ml of Hanks' balanced salt solution (HBSS) containing 0.35 g/l NaHCO^sub 3^ to wash out blood. Subsequently, the interior of the umbilical vein was digested with 10-20 ml dispase II (2.4 U/ml, Roche, Mannheim, Germany) for 30 min at 37 C. The dispase was washed out and the vein filled with M199 medium containing 0.1 % BSA, 50 g/ml streptomycin, 50 U/ml penicillin and then the umbilical cord was gently kneaded. After perfusion with 30 - 50 ml medium the released cells were centrifugcd (191g, 5 min) and resuspended in serum-free basal growth medium (SFM) supplemented with 10 g/ml streptomycin, 10 U/ml penicillin, 10ng/ml recombinant human epidermal growth factor (EGF), 20 ng/ml recombinant human basic fibroblast growth factor (bFGF) (SFM^sub +bFGF/EGF^) and 20 g/ml human fibronectin. Homogeneity of the cndothclial cell preparation was controlled by visual microscopic inspection.

Cell culture

Freshly isolated HUVECs were cultured to confluence at 37 C in a humidified 5% CO2 atmosphere in SFM^sub +bFGF/EGF^ trypsinized and washed two times in serum-free basal growth medium (SFM). Cells were seeded in a 80-cm^sup 2^ tissue culture flask coated with human fibronectin (1 -2.6 g/cm^sup 2^) in SFM^sub +bFGF/EGF^. HUVECs from passages 1 to 4 were used.

MTT test

Viable cells were monitored by an assay measuring the catalytic activity of cellular oxidoreductases like mitochondrial dehydrogenases, resulting in the cleavage of the substrate MTT [3- (4,5-Dimcthylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide] into water-insoluble formazan crystals. A final concentration of 0.5 mg/ ml MTTreagent was added directly to the cell supernatant. After 2.5 h the medium was removed, the forma/an crystals in adherent cells were dissolved in ethanol/DMSO (1:1 v/v), and the absorhance of the formazan solution was measured at 570 nm in a spectrophotometer (Amcrsham-Pharmacia, Freiburg, Germany). For statistical analysis a two-way ANOVA was applied. A p-value <0.05 was considered statistically significant. In case of statistically significant different time profiles, pairwisc comparisons between dosage groups were performed at day 4. In order to adjust for the multiple comparisons a Bonferroni correction was applied (Sachs 1978).

Cell adhesion assay

HUVECs (0.7-1.4 10^sup 5^ cells/well) were sub-cultured in serum- free medium (SFM) with or without 20 ng/ml hFGFand 10 ng/ml EGF in 24-well plates coated with 2.6 g/cm^sup 2^ fibroneclin or 2% gelatine or in uncoatcd plates. During the entire phase of adhesion (4 h) HABP was present in the medium at the indicated concentrations. Adherent viable cells were quantified using the MTT assay.

Cell proliferation assay

First, cells were allowed to attach to the uncoated or FN-coated surface for 5 h in SFM^sub +bFGF/EGF^ in the absence of HABP. Subsequently, cells were cultured in SFM^sub +bFGF/EGF^ for 4 days in the presence of different concentrations of HABP Where indicated, medium was exchanged on day 1 in order to remove HABP Before adding new HARP-free SFM^sub +bFGF/EGF^ cells were washed once with HABP- free SFM. Finally, adherent viable cells were quantified by the MTTassay or the number of adherent cells was counted in a Neubauer chamber.

Gel and Western blot analysis

For better resolution of the low molecular weight range, samples were separated on a 16.5% tricine gel with 6% cross-linker (Schagger and von Jagow, 1987). Proteins were electro-transferred to a PVDF membrane (Millipore, Eschborn, Germany) and growth factors were visualised by immunodetection using a goat polyclonal anti-human bFGFantibody or a mouse monoclonal anti-human EGF antibody (R&D Systems, Wiesbaden, Germany). After subsequent incubation with horseradish peroxidasc-linked secondary antibodies the labelled proteins were detected using an enhanced chemiluminesccncc (ECL) kit (AmershamPharmacia. Freiburg. Germany\).

Binding studies

To study the interaction of HABP with bFGF, a commercial bFGF detection kit (R&D Systems Inc., Minneapolis, MN. USA) was modified. Instead using the capturing untibody, 1 g/ml HARP was coated in wells of a Maxisorb plate (Nunc, Wiesbaden, Germany) at 4 C overnight in coating buffer (35 mM NaHCO^sub 3^, 15 mM NaCl, pH 9.6). After blocking with PBS containing 2% BSA and 5% sucrose, growth factors were added in a geometric dilution series as indicated. Binding was performed in HBS, pH 7.4. Residual unbound ligand was removed by repeated washing, plates were treated for 2 h at 37 C with PBS containing 2% BSA and 5% sucrose and bound ligand was detected with a biotinylated monoclonal anti-bFGF antibody. Binding of EGF was performed in a similar way, except that an unlabeled monoclonal antibody was used (R&D Systems Inc.) followed by a horseradish peroxidase-lahelled secondary antibody (Sigma- Aldrich, Deiscnhofen, Germany).

Results

Quantification of cell growth

Prior to extensive proliferation and adhesion studies an appropriate method to quantify the number of adherent cells was selected. A quantitative microscopic analysis ofcell growth by counting the cell number and by measuring the MTT signal was performed in preliminary proliferation experiments (Fig. 1). A high correlation between the cell number and the Mil signal intensity (correlation coefficient of 0.97) justified to deduce the number of viable cells by measuring the MTT signal. Since absolute cell numbers are a better presentation of the magnitude of changes, in the following experiments the MTT signal was measured and the respective cell number was calculated based on the correlation of the data shown in Figure 1.

Fig. 1. Correlation between MTT signal intensity and cell number. During a proliferation experiment as shown in Figure 2b every day cell viability was measured using the MTT reagent and additionally the number of adherent HUVECs was determined in a Neubauer chamber. The data set includes proliferation data of cells grown in the presence or in the absence of HABP. The cell number is plotted against the respective MTT signal intensity. Data were obtained from two independent proliferation experiments (n = 70, correlation coefficient 0.97).

Fig. 2. Effect of HABP on HUVEC proliferation. Adherent HUVECs were incubated for 4 days in FN-coated (a) or uncoated (b) 24-well plates in SFM^sub +bFGF/EGF^ without medium exchange at the indicated HABP concentrations, (c) Incubation as in (b) but medium was exchanged on day 1 by fresh SFM^sub +bFGF/EGF^ without new HABP. Each data point represents the mean S.D. of 3 independent observations (n^sub total^ = 105-140). The time profiles of HABP- treated cells differed statistically significantly from the control cells (p < 0.001 in each of the experiments (a-c).

HABP inhibits proliferation of HUVECs

HUVECs freshly prepared from human umbilical veins were cultured in the absence of fetal bovine serum. Replacing serum in the medium by BSA (scrum-free medium, SFM) prevented the rapid premature inhibition of HABP by serum protease inhibitors. Furthermore, defined growth factor conditions could be established because basic fibroblast growth factor (bFGF) and epidermal growth factor (EGF) were the only growth-inducing additives in the SFM (SFM^sub +bFGF/ EGF^).

We analysed the influence of HABP on the proliferation of HUVECs over a period of 4 days. For attachment, cells (passage 1-3) were seeded on fibronectin (FN)-coated plates in SFM^sub +bFGF/EGF^ without HABP and adhesion proceeded for 5 h. HABP was added once (on day 0) at the indicated concentrations. Cell proliferation proceeded for 4 days without medium exchange. Under these conditions no significant influence of HABP on cell proliferation was detectable until day 2, but a reduced proliferation became statistically significant until day 4 (Fig. 2a). Alternatively, freshly prepared HUVECs were adopted to grow on plates lacking a protein surface. Preliminary experiments demonstrated that HUVECs attached somewhat slower and less stringent to uncoated plates but, once attached, proliferation proceeded comparable to cells grown on FN-coated plates. This can be seen by comparing the respective control cells in Figures 2a and 2b.

When on day 0 HABP was added to adherent cells on uncoaled plates, a reduced proliferation rate was visible already within two days and at the lowest HABP concentration used (Fig. 2b). Since from day 3 the signal dropped below the level of day 0, obviously a net loss of cells was involved. When on day 1 the SFM^sub +bFGF/EGF^ was exchanged and cells were allowed to grow in fresh SFM^sub +bFGF/ EGF^ without newly added HABP, cell growth recovered and from day 3 to 4 cells grew at a similar rate as the untreated control cells (Fig. 2c). Obviously, replacing HABP-containing medium by HABP-free SFM^sub +bFGF/EGF^ reversed the anti-proliferative effect.

Fig. 3. Anti-adhesive activity of HABP. Attachment of cells proceeded for 4 h during which HABP was present at the indicated concentrations. Subsequently, the total numer of adherent viable cells was determined by MTT assay. Each data point represents the mean S.D. of 3 independent observations (n^sub total^ = 63). The concentration profiles as well as adhesion on uncoated compared to FN- or gelatine-coated plates at the highest dosage differ statistically significantly (p < 0.001).

Fig. 4. HUVEC proliferation depends on growth factors, a) Cells were kept for 1 day in SFM without growth factors. On day 1 the medium was replaced by fresh SFM^sub +bFGF/EGF^ (without or with 125 nM HABP) or by HABP-treated SFM^sub +bFGF/EGF^. Data are the mean S.D. of 3 independent observations (n^sub total^ = 80). b) After adhesion cells were incubated in SFM supplemented with HABP-pre- treated bFGF/EGF or with HABP-untreated growth factors (control). On day 2, fresh growth factors were added to the cells cultured in SFM with HABP-pre-treated bFGF/EGF as indicated and culturing proceeded until day 4. Each data point represents the mean S.D. of 3 independent observations with 4 measurements each.

When aprotinin-inaclivated HABP was added to HUVECs no significant effect on proliferation was seen (data not shown), indicating that the proteolytic activity of HABP was pivotal for the effects observed. In summary, only cells incubated in SFM^sub +bFGF/ EGF^ free of HABP exhibited efficient proliferation. Already a single dose of 15.6 nM HABP resulted in a significant reduction in cell growth due to detachment of cells. This process was also visible on FN-coated plates, although delayed and less pronounced.

Anti-adhesive function of HABP

The results presented in Figure 2c implicated that removal of HABP and/or addition of fresh SFM^sub +bFGF/EGF^ could restore proliferation. This raised the hypothesis that the reduced proliferation accompanied by the loss of viable cells was caused by proteolytic cleavage of adhesion molecules on the cell surface required for the attachment of HUVECs to the surface. We investigated the influence of HABP on cell-matrix interactions during the initial adhesion of cells to coated or uncoated surfaces. Early passages of HUVECs (pl-3) were chosen for this study. In the absence of HABP the adhesion of HUVECs in SFM^sub +bFGF/EGF^ on uncoated plates was not significantly different to the adhesion of cells to FN-coated plates (Fig. 3). On gelatine-coated plates cells seemed to attach somewhat more efficient but regarding the complete set of data this was not significant. Omitting the growth factors had no effect on the adhesion of endothelial cells in the presence or absence of HABP (data not shown), presumably because during the phase of attachment growth factors were not required or because on the cell surface sufficient residual growth factors were present from the previous culturing period. In the presence of HABP a significant reduction in adhesion was found for HUVECs seeded on uncoated plates, but not on FN- or gelatine-coated cell culture plates. This indicates that HABP disrupts initial cell-surface interactions which arc pivotal for adhesion to uncoated culture dishes but not for adhesion to a gelatine or an FN surface. The more pronounced reduction in proliferation on uncoated plates (compare Figs. 2a and b) is very likely attributed to this anti-adhesive activity of HABP. Western blot analyses of total cell extracts with antibodies specific for [alpha]2, [alpha]5 or [beta]3 identified these integrins but did not indicate cleavage or induction of expression of these proteins (data not shown). The identity of the adhesion molccule(s) involved remains unclear and is subject to ongoing investigations.

Inhibition of proliferation by HABP is dependent on growth factors

Since the anti-adhesive effect of HABP was only observed on uncoated plates, but inhibition of proliferation was also seen on FN- coated plates on day 3 and 4, a second mechanism was postulated to be involved in the clown-regulation of endothelial cell growth. The ability of HUVECs to recover from HABP-induced inhibition of cell growth after replacing the medium by protease-free SFM^sub +bFGF/ EGF^ (cf. Fig. 2c) raised the possibility that growth factors are involved in this process. To address this hypothesis HUVECs were seeded on FN-coated plates in SFM without growth factors. After adhesion the medium was replaced by fresh medium of the same composition ('day 0') and cells were cultured for 24 h during which no growth was visible (Fig. 4a). The medium was exchanged on day 1 against fresh SFM supplemented as indicated. Cells cultured in SFM^sub +bFGF/EGF^ without HABP started immediately to proliferate, demonstrating the requirement of growth factors to initiate cell growth. Addition of SFM^sub +bFGF/EGF^ together with 125 nM HABP (' + bFGF/EGF + HABP') resulted in an initial proliferation that was abolished within one day, reflecting the observations shown \in Figure 2a. When SFM^sub +bFGF/EGF^ (1.2 nM bFGF and 1.6 nMEGF) waspre-incubated with 125 nM HABP for 24 h at 37C prior to addition to the cells on day 1, growth was also initially induced, but was down-regulated after 1 day. Apparently, for the inhibition of proliferation no direct interaction of active HABP with HUVECs was required. This excluded direct disruption of cell-surface interactions as the only reason for the observed growth inhibition. The reduced cell growth was not due to residual HABP activity in the medium because after the preincubation a tenfold molar excess of the serine protease inhibitor aprotinin was added to the HABP-treated SFM^sub +bFGF/EGF^. This excess of aprotinin did not affect HUVEC proliferation but was sufficient to completely abolish the proteolytic activity of HABP (data not shown). When cells were cultured from day 1 to 4 in SFM without growth factors, a continuous detachment of cells was visible because endothelial cells require both growth factors for survival. Interestingly, this process was neither accelerated nor reversed by HABP, indicating that the influence of HABP on endothelial cell proliferation was strongly dependent on growth factors. Similar anti-proliferalive activities were seen with HUVECs seeded on uncoated plates (data not shown), underlining that this process was independent of the surface on which cells were seeded. In contrast, the anti-adhesive activity of HABP was only seen on uncoated plates. From these observations was concluded that the anti-adhesive activity of HABP could not be the major cause for cessation of cell proliferation.

Fig. 5. Binding of growth factors to immobilized HABP. a) Plates were coated with 10 g/ml HABP, and binding of bFGF and EGF, respectively, were studied by ELISA. Each data point represents the mean S.D. of 4 measurements. This experiment has been repeated twice with comparable results. Insert: bFGF binding to immobilized active and aprotinin-inactivated HABP. Data arc the mean S.D. of 3 measurements. b) Effect of heparin and heparan sulfate on binding of bFGF (2.5 ng/ml) to HABP. Each data point represents the mean S.D. of 3 measurements. This experiment has been repeated once with comparable result.

In order to substantiate the finding that growth factors are affected by HABP, cells were sub-cultured in SFM to which only HABP- pretreatcd bFGF/EGF was added (Fig. 4b). Whereas in the experiment shown in Figure 4a an 80 to 105-fold molar excess of enzyme was used during the 24 h pretreatmenl of SFM^sub +bFGF/EGF^ in this study 3.2 M growth factors were pre-incubatcd in 10 mM Hepes, 25 mM NaCl, 10 mM CaCl^sub 2^ pH 7.5 for 24 h with 0.32 M HABP, and HABP was subsequently inactivated by aprotinin. Proliferation in SFM + prctreated growth factors proceeded for 4 days. On day 2 no proliferation of cells that received pre-treated growth factors was detectable. When fresh growth factors were added separately or in combination, only addition of bFGF or the combination of bFGF and EGF could restore proliferation whereas EGF alone was ineffective. This demonstrated unambiguously that HABP affected bFGF but not EGF.

HABP inactivates bFGF

From the results above it was hypothesised that HABP directly affected bFGF e.g. by complcxation or partial hydrolysis. In binding studies it was found that bFGF but not EGF efficiently formed complexes with immobilized HABP in a concentration-dependent way (Fig. 5a). When comparing active and aprotinin-inactivated HABP, it was found that bFGF bound to both with similar efficiency (insert). Furthermore, complex formation between HABP and bFGF was strongly reduced by heparin and HS, suggesting that complex formation is unlikely in a GAG environment (Fig. 5b).

Since aprotinin-inactivated HABP bound bFGF but never inhibited HUVEC proliferation it became clear that complex formation with bFGF alone was not responsible for the observed inhibition of cell growth. Hence, in an in vitro experiment we investigated the possibility that HABP proteolytically cleaves bFGF. When purified bFGF and EGF were incubated with HABP for 24 h a time-dependent reduction in the bFGF band became visible (Fig. 6a). Already after 2 h the intensity of the mature bFGF band was reduced and processing hands were detectable. Cleavage of bFGF did not occur in the presence of aprotinin (cf. Fig. 6b). The limited hydrolysis proceeded within 24 h almost to completeness, yielding two major hands of ~12 and ~5 kDa. In contrast, no cleavage of EGF could be observed under identical conditions. It became evident from these experiments that HABP reduced the bFGF/ EGF-dependent proliferation of HUVECs by complexation and specific cleavage of the pro- angiogenic bFGF. In light of this finding the initial increase of HUVEC proliferation cultured in the HABP-pre-treated medium (cf. Fig. 4) was most likely due to residual growth factor activity which was, however, not sufficient to maintain proliferation.

Since on the one hand the proteolytic activity of HABP is accelerated by heparin-likc molecules (Etscheid et al., 2000), but on the other hand binding of bFGF to HABP was prevented we studied the influence of heparin and heparan sulfate (HS) on the cleavage of bFGF. This was especially interesting because heparan sulfate proteoglycans on the cell surface or in the extracellular matrix can function as an extracellular reservoir where bFGF is protected from degradation by protcases like trypsin or plasmin (Saksela et al., 1988). In in vitro experiments 10 g/ml heparin or HS completely prevented the cleavage of bFGF by HABP (Fig. 6b), presumably due to the prevention of complex formation. Both GAGs strongly reduced binding and prevented cleavage also at concentrations of only 100ng/ ml (data not shown) demonstrating the efficient protective function of heparin and HS against cleavage of bFGF by the hyaluronan- binding protease. It can be concluded that HABP would inhibit cell growth that depends particularly on free bFGF, whereas cell growth triggered by bFGF stored and protected in heparan sulfate proteoglycans would be much less affected.

Fig. 6. In vitro treatment of bFGF and EGF with HABP. Basic FGF or EGF (3.2 M each) were incubated with 0.32 M HABP in 50 mM TrisHCl, 150 mM NaCl, 5 mM CaCl^sub 2^, pH 7.5, for different time intervals. Reaction mixtures were denatured under reducing conditions and were separated on a tricine gel (Schagger and von Jagow, 1987) for better resolution of small polypeptides. bFGF and EGF were visualized by immunoblotting. Results arc representative of at least 3 independent experiments.

Discussion

The results presented here provide further evidence that the physiological role of the hyaluronan-binding protease is not restricted to hemostasis but also affects cell regulation. Previous independent in vitro observations already indicated that HABP has the potency to regulate cellular functions:

i) The intracellular release of Ca^sup 2+^ from internal stores was observed when HUVECs were incubated with HABP. This Ca^sup 2+^ flux was still detectable when the endolhclial protease receptors were desensitised with thrombin, trypsin or FXa, indicating the involvement of a different receptor (Storck et al., 1999).

ii) The consequences of activation of single-chain prourokinase (sc-uPA) by HABP (Romisch et al., 2000) are not restricted to clot lysis but may trigger cell-associated fibrinolysis leading to the degradation of the ECM and promoting cell migration and angiogenesis (Pepper et al., 1987; Mignatti and Rifkin, 1996; Rabbani, 1998).

iii) Similar to plasma kallikrein HABP cleaves kininogen and releases activated kininogen (HKa) and bradykinin (Etscheid et al., 2002). On the surface of endothelial cells which express kininogen and its receptors (Schmaier et al., 1988) this process would result in the release of HKa and bradykinin directly at their sites of action. Bradykinin induces intracellular signalling via specific receptors, HKa has anti-proliferative, anti-migratory and anti- angiogenic properties (Colman et al., 2000).

iv) HABP interacts with GAGs and can directly cleave matrix proteins like vitroneclin, fibroneclin, and fibrinogen (Krc et al., 2002). It is still unclear whether this limited proteolysis of matrix proteins affects cell-matrix interactions or provides exogenous triggers for intracellular signalling.

v) HABP specifically binds to vascular smooth muscle cells and reduces platelet-derived growth factor (PDGF)-dependent smooth muscle cell proliferation (Kannemeier et al., 2002). Complex formation between HABP and PDGF has been reported as the major cause, underlining the results of our studies. No cleavage of PDGF has been found, but prevention of PDGF dimer formation, which is required for receptor activation, has been postulated.

vi) A first immunohistochemical survey identified HABP in many tissues, in particular, various epithelia stained positive (Knoblauch et al., 2002) presenting further evidence for a cell- related role of this plasma protease.

Here we demonstrate that HABP strongly inhibits bFGF/ EGF- induced proliferation of HUVECs and present two mechanisms involved. First, HABP aggravates cell adhesion, most likely by cleavage of adhesion molecules which are required for attachment and for proliferation. This activity is obviously not endothelial cell- specific, because also human lung fibroblasts MRC-5 and adcnocarcinoma cells (A549) showed in the presence of HABP a reduced adhesion to uncoalcd plates but not to fibronectin- or gelatine- coated plates (own observation). Second, HABP strongly binds to and partially hydrolyses bFGF in vitro resulting in almost complete cleavage of a tenfold molar excess of bFGF within 24 h. Under experimental proliferation conditions with 1.2 nM free bFGF and 15.6 to 500 nM HABP, bFGF complexation and cleavage is most likely even more effective. A common property of bFGF and HABP is their affinity to heparin-l\ike molecules which have also been shown to enhance HABP activity (Etscheid et al., 2000, 2002; Romisch et al., 2000). Heparan sulfate proteoglycans (HSPGs) function as co-receptors for bFGF and play an important role in bFGF signalling (Zhang et al., 2001). Additionally, they represent an extracellular storage reservoir for bFGF. In vitro the complexation and cleavage of bFGF by HABP was completely prevented by heparin or HS suggesting that in vivo heparin-like molecules within the ECM and on the cell surface would protect bFGF from direct degradation by HABP. Exogenous free bFGF apparently is a substrate for HABP, implicating that endothelial cell attraction in a paracrine mode would be affected by HABP whereas bFGF stored in cell surface HSPG, e.g. after autocrine secretion, is protected. Studies on lung fihroblasts MRC 5 demonstrated that HABP did not inhibit the proliferation of these cells (unpublished data), raising the question whether inhibition of bFGF-mediated proliferation is endothelial cell-specific. More detailed studies are required to determine the influence of HABP on the proliferation of other cells of the vasculature, e.g. microvascular endothelial cells.

bFGF is a potent mitogenic factor for many cell types. It induces a proangiogenic phenotype in ECs and is expressed in various tissues. This growth factor plays a critical role in physiologic, pathologic and tumor angiogencsis (Li et al., 2000; Rofstad and Halsor, 2000. Tumor development is often correlated with an increased expression of pro-angiogenic factors like bFGF, as reported in pancreatic cancer (Ghaneh et al., 2002), hepatocellular cancer (Qin and Tang, 2002), ovarian carcinoma (Davidson et al., 2002) and 'non-Hodgkin' lymphoma (Pazgal et al., 2002). bFGF exerts its action in different ways: by direct action (Moscatelli et al., 1988), by inducing vascular endothelial growth factor (VEGF) (Stavri et al., 1995) or potentiating VEGF activity (Goto et al., 1993). The inactivation of free circulating bFGF would be a potent mechanism through which endothelial cell proliferation and migration can be controlled and neovascularixation can be prevented. In this regard the potency of HABP to neutralize the pro-angiogenic bFGF deserves special attention. Detailed studies in migration and angiogenesis models arc required to evaluate the physiologic importance of these findings and to evaluate the usefulness of HABP as an agent to control growth factor-mediated endothelial cell growth in a physiologic or pathologic environment.

Acknowledgements. This work was supported by the Federal Ministery of Health. We are grateful to R. Schuhmann, Asklepios Clinics Langen, for his cooperation. The assistance of P. Volkers in statistical data analysis is highly acknowledged.

References

Asahara,T., Bauters, C., Zheng, L. P., Takeshita, S., Bunting, S., Ferrara, N., Symes, J. F., Isner, J. M. (1995): Synergistic effect of vascular endothelial growth factor and basic fibroblast growth factor on angiogenesis in vivo. Circulation 92, II365-II371.

Busse, R., Lamontagne, D. (1991): Endothelium-dcrivcd bradykinin is responsible for the increase in calcium produced by angiotensin- converting enzyme inhibitors in human endothelial cells. Naunyn Schmiedebergs Arch. Pharmacol. 344, 126-129.

Choi-Miura, N. H., Tobe, T., Sumiya, J., Nakano, Y., Sano. Y.. Mazda. T., Tomila, M. (1996): Purification and characterization of a novel hyaluronan-binding protein (PHBP) from human plasma: it has three EGF, a kringle and a serine protease domain, similar to hepatocyte growth factor activator. J. Biochem. 119, 1157- 1165.

Cines, D. B., Pollak, E. S., Buck, C. A., Loscalzo, J., Zimmerman, G. A., McEver, R. P., Pober, J. S., Wick, T. M., Konkle, B. A., Schwartz., B. S., Barnathan, E. S., McCrae, K. R., Hug, B. A., Schmidt, A. M., Stern, D. M. (1998): Endothelial cells in physiology and in the pathophysiology of vascular disorders. Blood 91, 3527-3561.

Colman, R. W., Bradford, A. J., Lin, Y., Johnson, D., Mousa, S. A. (2000): Domain 5 of high molecular weight kininogen (kininostatin) down-regulates endothelial cell proliferation and migration and inhibits angiogenesis. Blood 95, 543-550.

Davidson, B., Goldberg, I., Gotlieb, W. H., Kopolovic, J., Ben Baruch, G., Nesland, J. M., Reich, R. (2002): The prognostic value of metalloproteinases and angiogenic factors in ovarian carcinoma. Mol. Cell. Endocrinol. 187, 39-45.

Etscheid, M., Hunfeld, A., Konig, H., Seitz, R., Dodt, J. (2000): Activation of proPHBSP, the zymogen of a plasma hyaluronan binding serine protease, by an intermolecular autocatalytic mechanism. Biol. Chem. 381, 1223-1231.

Etscheid, M., Been N., Fink, E., Seitz, R., Dodt, J. (2002): The hyaluronan-binding serine protease from human plasma cleaves HMW and LMW kininogen and releases bradykinin. Biol. Chem. 383, 1633-1643.

Ghaneh. P., Kawesha, A., Evans, J. D., Neoptolemos, J. P. (2002): Molecular prognostic markers in pancreatic cancer. J. Hepatobiliary Pancreat. Surg. 9.1-11.

Goto, F., Goto, K., Weindel. K., Folkman, J. ( 1993): Synergistic effects of vascular endothelial growth factor and basic fibroblast growth factor on the proliferation and cord formation of bovine capillary endothelial cells within collagen gels. Lab. Invest. 69, 508-517.

Hunfeld, A.. Etscheid, M., Konig, H., Seitz, R., Dodt, J. (1999): Detection of a novel plasma serine protease during purification of vitamin K-dependent coagulation factors. FEBS Lett. 456, 290-294.

Jaffe, E. A.. Nachman, R. L., Becker, G.G., Minick. C. R. (1973): Culture of human endothelial cells derived from umbilical veins. Identification by morphologic and immunologic criteria. J. Clin. Invest. 52, 2745-2756.

Kannemeier, C, Kanse, S. M., Romisch. J., Preissner, K. T. (2002): Factor VII activating protease (FSAP) regulates the proliferation of human vascular smooth muscle cells. Ann. Hematol. 81, A 38.

Knoblauch. B. Kellert, J., Battmann, A., Preissner, K. T., Romisch, J. (2002): A histological study of FVII-activating protease (FSAP) distribution in human tissue. Ann. Hematol. 81, A 42.

Kre, J., Seitz. R., Dodt, J., Etscheid, M. (2002): Interaction of the plasma hyaluronan-binding serine protease with extracellular matrix proteins. Ann. Hematol. 81, A 47.

Li, Z., Shimada, Y., Uchida, S., Maeda, M., Kawabe, A., Mori, A., Itami, A., Kano, M., Watanabe, G., Imamura, M. (2000): TGF-alpha as well as VEGF, PD-ECGF and bFGF contribute to angiogencsis of csophagcal squamous cell carcinoma. Int. J. Oncol. 17, 453-460.

Mignatti, P., Rifkin, D.B. (1996): Plasminogen activators and matrix metalloproteases in angiogenesis. Enzyme Protein 49, 117- 137.

Moscatelli, D., Joseph-Silverstein, J., Presta. M., Rifkin. D. B. (1988): Multiple forms of an angiogencsis factor: basic fibroblast growth factor. Biochimie 70, 83-87.

Pazgal, L., Zimra, Y., Tzahar, C. Okon, E., Rabizadeh, E., Shaklai, M., Bairey, O. (2002): Expression of basic fibroblast growth factor is associated with poor outcome in non-Hodgkin's lymphoma. Br. J. Cancer 86, 1770-1775.

Pepper, M. S., Vassalli, J. D., Montesano, R., Orci, L. (1987): Urokinase-type plasminogcn activator is induced in migrating capillary endothelial cells. J. Cell Biol. 105, 2535-2541.

Qin, L. X., Tang, Z. Y. (2002): The prognostic molecular markers in hepatoccllular carcinoma. World J. Gastroenterol. 8, 385-392.

Rabbani, S. A. (1998): Metalloproteases and urokinase in angiogenesis and tumor progression. In Vivo 12, 135- 142.

Rofstad, E. K., Halsor, E. F. (2000): Vascular endothelial growth factor, interleukin 8, platelet-derived endothelial cell growth factor, and basic fibroblast growth factor promote angiogenesis and metastasis in human melanoma xenografts. Cancer Res. 60, 4932-4938.

Romisch, J., Feuner, A., Vermohlen, S., Stohr, H.-A. (1999): A protease isolated from human plasma activating factor VII independent of tissue factor. Blood Coagul. Fibrinolysis 10, 471- 479.

Romisch, J., Vermohlen, S., Feuner, A., Stohr, H.-A. (2000): The FVII activating protease cleaves single-chain plasminogen activators. Haemostasis 29, 292-299.

Sachs, L. (1978): Angewandte Statistik, 5th edition. Springer, Berlin.

Saksela, O., Moscatelli, D., Sommer, A., Rifkin, D. B. (1988): Endothelial cell-derived heparan sulfate binds basic fibroblast growth factor and protects it from proteolytic degradation. J. Cell Biol. 107, 743-751.

Schagger, H., von Jagow. G. (1987): Tricine-sodium dodecyl sulfate-polyacrylamide gel clcctrophorcsis for the separation of proteins in the range from 1 to 100 kDa. Anal. Biochem. 166, 368- 379.

Schmaier, A. H., Kuo, A., Lundberg, D., Murray, S., Cines, D. B. (1988): Expression of high molecular weight kininogen on human umbilical vein endothelial cells. J. Biol. Chem. 263, 16327- 16333.

Stavri, G. T, Zachary, I. C., Baskerville, P. A., Martin, J. F., Erusalimsky, J. D. (1995): Basic fibroblast growth factor upregulates the expression of vascular endothelial growth factor in vascular smooth muscle cells. Synergistic interaction with hypoxia. Circulation 92, 11-14.

Slorek, J., Hunfeld, A., Dodt, J. (1999): A new human plasma serine protease activates endothelial cells. Pfluegers Archiv Fur. J. Physiol. 437 (Suppl.), R 164.

Zhang, Z., Coomans, C., David, G. (2001): Membrane heparan sulfate proteoglycan-supported FGFZ-FGFR1 signaling: evidence in support of the "cooperative end structures" model. J. Biol. Chem. 276, 41921-41929.

Michael Etscheid1), Nicole Beer, Julia Anne Kre, Rainer Seitz, Johannes Dodt

Department of Hematology and Transfusion Medicine, Paul-Ehrlich- Institut, Federal Agency for Sera and Vaccines, Langen/Germany

Received August 15, 2003

Received in revised version November 14, 2003

Accepted November 28, 2003

1) Dr. Michael Etscheid, Department of Hematology and Transfusion Medicine, Paul-Ehrlich-Institut, Federal Agency for Sera and Vaccines, Paul-Ehrlich-Str. 51-59, D-63225 Langen/Germany, e-mail: etsmi@ pei.de. Fax: +49610377 1250.

Copyright Urban & Fischer Verlag Jan 2004

More News in this Category