- Email: [email protected]
Von Willebrand’s factor mediates the adherence of human tumoral cells to human endothelial cells and ticlopidine interferes with this effect
Biomed & Pharmacother 2000 ; 54 : 431-6 © 2000 Éditions scientifiques et médicales Elsevier SAS. All rights reserved S0753332200000068/FLA
Original article
Von Willebrand’s factor mediates the adherence of human tumoral cells to human endothelial cells and ticlopidine interferes with this effect M. Morganti1, A. Carpi2, B. Amo-Takyi1, A. Sagripanti2, A. Nicolini2, R. Giardino3, C. Mittermayer1* 1 Institute of Pathology RWTH, Klinikum, Pauwelstrasse 30, Aachen, Germany 52074; 2Departments of Aging and Reproduction, Internal Medicine, University of Pisa, Italy; 3Department of Experimental Surgery, Rizzoli Orthopaedic Institute, Bologna, Italy
Summary – The aim of this study was to explore whether von Willebrand’s factor (vWF) plays a role in the adhesion of human colon tumor cells to human endothelial cells in our coculture system. Cell colony density was evaluated basally (endothelial plus colon tumor cells) and following the addition of: purified vWF, vWF plus vWF-blocking antibodies, antibodies against various integrins and adhesion molecules (α2 b integrin, β1 integrin, β3 integrin, intercellular adhesion molecule-I, intercellular adhesion molecule-II, vitronectin receptor CD61 CD51, laminin α6/β4 receptor), and various drugs inhibiting the hemostatic system (ticlopidine, heparin, acetyl salicylic acid [ASA], defibrotide, indobuphen, dipyridamole, sulfinpyrazone). Furthermore, vWF concentration was measured in the supernatant fluid of the coculture system basally and following the addition of the above-listed drugs. Cell colony density (as measured by light absorption) increased by 33% following the addition of vWF and returned to a value similar to the basal level with antibodies against vWF, while it did not change significantly following the addition of antibodies against the other integrins or adhesion molecules tested. The same parameter was reduced by 35% following the addition of ticlopidine, while it showed a smaller or no change with the other drugs tested. Similarly, vWF concentration in the cell coculture supernatant showed the greatest reduction (from 0.22 to 0.11 mg/mL) following the addition of ticlopidine. These data suggest that vWF mediates the adherence of human tumor cells to human endothelial cells and that ticlopidine interferes with this effect. © 2000 Éditions scientifiques et médicales Elsevier SAS colon tumor cells / endothelial cells / ticlopidine / von Willebrand’s factor
In a previous paper we evaluated the effect of human colon cancer cells (HRT-18) on endothelial cells from human umbilical vein (HUVEC) in a cell coculture system [1]. We showed that the human tumoral cells modify the human endothelial cell activity in such a way as to reduce their production of tissue type plasminogen activator (tPA) and to increase the produc-
*Correspondence and reprints
tion of plasminogen activator inhibitor (PAI-1) as well as von Willebrand’s factor (vWf) [1]. Our observation that the interaction between tumor and endothelial cells led to depression of the fibrinolytic system and enhancement of vWf was in agreement with previous studies, supporting the hypothesis that the hemostatic alterations produced by tumor cells can play a role in cancer progression [2-6]. In this respect, fibrin formation has been studied and is considered to play an important role [7].
432
M. Morganti et al.
MATERIALS AND METHODS The adhesion of two human cell types – endothelial cells from human umbilical vein (HUVEC) and human colon cancer cells (HRT-18) – was evaluated by the cell adhesion assay. This assay was performed without addition of vWF (standard) and with vWF addition. Furthermore, the effect of antibodies aganist vWF or other adhesion molecules on this assay was evaluated. Finally, the effect of various drugs inhibiting the hemostatic system on the adhesion of the two human cell types has been evaluated. Details on the culture cells were presented in our previous paper [1].
sequent steps were those of the standard procedure, as reported above. Cell-adhesion assay in presence of vWf and vWf-blocking antibodies Three 96 multiwell plates were incubated for 60 min at 37° C in a dilution 1:100 vol/vol vWF purified antibodies/HAMS’s medium. The cultures were washed twice with PBS, then 50 µL/well of a suspension of 10,000 tumoral cells HRT-18/mL HAMS’s plus 1:100 vol/vol vWF antibodies was added and the cocultures incubated at 37° C for 6 hours. The subsequent steps were those of the standard procedure.
Cell-adhesion assay, standard procedure Confluent endothelial HUVEC cell cultures grown in 96 multiwell plates were incubated for 60 min at 37° C in serum-free medium (HAMS’s plus glutamine-gentamycin, Gibco, Grand Island, USA) [8-10]. The cultures were washed twice with sterile phosphate buffered saline (PBS), then 50 mL/well of a suspension of 10,000 tumoral cells HRT-18/mL (phenotypically heterogeneous human rectal cancer cell line, Thompkins et al. 1974) HAMS’s was added and the cocultures incubated at 37° C for six hours. The medium containing unattached cells was discarded and cocultures were rinsed in cold PBS to remove unattached cells, then fixed with 70% ethanol and stained with 100 µL/well of a solution of 0.1% crystal violet in distilled water for 25 minutes. The tumoral HRT-18 colonies were counted and measured. Multiwell plates were dried with paper pads and dye was re-suspended with 50 µL of 0.2% Triton X-100 in water plus 50 mL propanol/well. The stain intensity was evaluated by chromatometry with an ELISA reader at 590 nm. Cell-adhesion assay in presence of vWf purified protein Three 96 multiwell plates were incubated for 60 min at 37° C in a dilution 1:100 vol/vol vWf purified protein/HAMS’s medium. The cultures were washed twice with PBS, then 50 mL/well of a suspension of 10,000 tumoral cells HRT-18/mL HAMS’s plus 1:100 vol/vol vWf purified protein was added and the cocultures incubated at 37° C for 6 hours. The sub-
Cell-adhesion assay in presence of surface and adhesion molecules blocking antibodies The HAMS’s medium used for coculture incubation was added with the following antibodies against different integrins and adhesion molecules: β1 integrin (Human β1 integrin monoclonal clone CD29 Endogen, MD, USA) at 1:100 vol/vol, βIII integrin (monoclonal clone, Aachen Laboratory, Germany) at 1:100 vol/vol, α2 β1 integrin (Integrin α2β1 monoclonal clone CD49 Dako SPA, Milan, Italy) at 1:10 vol/vol, intercellular adhesion molecule ICAM-1 (Human ICAM-1 clone CD54 R and D System INC, MN USA) at 1:100 vol/vol, ICAM-2 (ICAM-2 SC Chimera R and D System INC, MN USA) at 1:100 vol/vol, vitronectin receptor (human vitronectin monoclonal clone CD51 Endogen, MD, USA) at 1:100 vol/vol laminin α6β4 receptor (laminin policlonal antibody, Dako, Milan, Italy) at 1:100 vol/vol. The subsequent steps were those of the standard procedure. Cell-adhesion assay following addition of drugs inhibiting the hemostatic system The HAMS’s medium used for coculture incubation (6 hr at 37° C) was respectively added with one of the following drugs: ticlopidine 50 ng/mL, acetyl salicylic acid (ASA) 50 ng/mL, heparin 50 ng/mL, defibrotide 50 ng/mL, indobuphen 50 ng/mL, dipyridamole 50 ng/mL, sulfinpyrazone 50 ng/mL. The subsequent steps were those of the standard procedure. Biomed & Pharmacother 2000 ; 54 : 431-6
vWF in cocultures of human endothelial and colon tumor cells
433
Effect of anti-hemostatic drugs on vWf production A concentration of 160,000 endothelial cell/mL was seeded in 96-well tissue culture plates (BectonDickinson, NJ, USA) at 100 mL/well in a humidified atmosphere at 37° C for 24 h. The cell line HRT-18 was washed with PBS and disaggregated with 2 mL PBS/Trypsin 0.125% at 37° C for three minutes, then the cells were resuspended in 10 mL serum-free RPMI and centrifuged at 1,100 × g for 10 min. The supernatant was discarded and the cells were resuspended in the same medium to obtain a concentration of 10,000 cells/mL. The endothelial cell cultures in 96 wells were washed twice with PBS, then 50 mL of tumor cell suspension (10,000 tumor cells/mL) were added to each well. To seven of the 96 well plates containing cocultures 50 ng of ticlopidine, ASA, heparin, defibrotide, indobuphen, dipyridamole and sulfinpyrazone were added, respectively. Then cocultures were incubated at 37° C in a humidified atmosphere for 24 h. Thereafter the coculture system was terminated and the supernatant separated from the cells; the supernatant was centrifuged twice at 1,100 x g for 10 min to remove cell residues. The cell supernatant was managed according to the standard procedure for measurement of vWF by enzyme-linked immunosorbent assay (ELISA) (Asserachrom Diagnostica Stago, Boehringer, Mannheim, Germany). The spectroscope used was a Titertek Multiskan Plus MKII (Labsystem, Finland). Each probe was repeated three times with three different primary HUVEC cultures.
Figure 1. Cell colony density following the addition of HRT-18 to HUVEC cells and successively of purified vWF to the system.
significant change of the signal following the addition of vWF with antibodies. These data indicate that the effect following vWF addition on the cell coculture system observed in figure 1 is really due to vWF as it was blocked by the specific antibody. The experiment shown in figure 3 compares the reported effect of vWF antibodies with that of antibodies against the above-mentioned integrin and adhesion molecules (α2b integrin, β1 integrin, β3 integrin, intercellular adhesion molecule-I, intercellular adhesion molecule-II, vitronectin receptor CD61 CD51, laminin α6/ b4 receptor), which may be involved in the interaction between HUVEC and HRT-18 tumor cells. The figure shows that the only antibody able to reduce the adhesion and/or the number of cell colonies was that against vWF. The experiment shown in figure 4 explores whether the drugs mentioned above (ticlopidine, ASA, hep-
RESULTS Figure 1 shows the cell colony density following the addition of HRT-18 to HUVEC cells and subsequently of purified vWF to the entire system. It can be observed that the cell colony density increases as expected after addition of HRT-18 cells; however, it shows a further increase following the addition of vWF addition. This suggests that vWF may stimulate attachment and or growth of HRT-18 colonies. Figure 2 shows the results of an experimental setup different from the previous one only for the final addition to the system of antibodies to vWF together with purified vWF. The increase in the chromatometry signal after the addition of HRT-18 to HUVEC cells was the same as in figure 1; however, there was no Biomed & Pharmacother 2000 ; 54 : 431-6
Figure 2. Cell colony density following the addition of HRT-18 to HUVEC cells and successively of vWF plus vWF blocking antibodies to the system.
434
M. Morganti et al.
Figure 3. Cell colony density following the addition of HRT-18 to HUVEC cells and successively of antibodies against vWF or different integrins and adhesion molecules to the system.
arin, defibrotide, indobophen, dipyridamole, sulphinpyrazone) known to influence the hemostatic system exert some effect on the adhesion of the cell colonies in our system. Ticlopidine showed the greatest effect in reducing number and or adhesion of cell colonies so that the chromatometry signal returned to a value very similar to that obtained by the presence of HUVEC cells alone. A similar lighter effect was also shown following the addition of ASA and heparin. Figure 5 shows the concentration value of vWF in the supernatant fluid of the HUVEC-HRT-18 coculture system before and after the addition of the seven drugs. The vWF concentration was markedly reduced following the addition of ticlopidine (from 0.22 to 0.11 ng/mL) and to a lesser extent (about 50% less) after the addition of ASA and heparin.
Figure 4. Cell colony density following the addition of HRT-18 to HUVEC cells and successively of seven different drugs inhibiting haemostasis to the system.
Figure 5. vWF concentration in the supernatant of the HUVECHRT-18 coculture system before and after the addition of seven different drugs inhibiting the haemostatic system.
DISCUSSION Our previous finding that the vWF concentration value in the supernatant of HUVEC cocultures changed from undetectable to an easily measurable value [1] following the addition of human colorectal tumor cells HRT-18 stimulated this second study, which aimed to clarify the meaning of this increased vWF production by endothelial cells. The present study shows that the addition of vWF to our system of tumor and HUVEC cells induces an enhancement of the adhesion among the cells in the coculture or their number. The use of various antibodies against vWF and various integrins and adhesion molecules known to regulate the interaction – clearly intercellular adhesion – shows that vWF is responsible for the enhanced adhesion of tumor to endothelial cell in our system. Therefore, it is likely that tumor cells change some properties of the HUVEC cells by inducing vWF production, which in turn increases the binding of tumor to endothelial cells. It has been reported that vascular growth factor VGEF can induce endothelial cells to release vWF [11]. It is well extablished that vWF promotes platelet adhesion to the subendothelial matrix; moreover, it increases platelet aggregation by interacting with platelet surface glycoproteins. One old, established observation was that platelet aggregates surrounded cancer cells in human pulmonary arterioles [12]. A successive correlation was shown between in vivo malignancy and platelet aggregating activity of substances derived from animal tumor cells [13]. More Biomed & Pharmacother 2000 ; 54 : 431-6
vWF in cocultures of human endothelial and colon tumor cells
recently, a leading hypothesis involves platelets in tumors’ angiogenesis [14]. This hypothesis is based on chemical and preclinical findings that tumor angiogenesis is dependent not only on endothelial and tumour cells but also on platelet-endothelium interaction [14]. Studies have shown that vWF plays an important role in the adhesion of tumor cells to platelets (in vitro) [15] and in facilitating implantation of lung metastasis after inoculation of tumor cells in animals [15-17]. Human and animal colon carcinoma cells have been used for these studies [15-17]. To our knowledge these effects of vWF have been attributed to its participation in the interaction between platelets and tumor cells [15-17]. More recently, vWF has been shown to interact directly with malignant hematopoietic cell lines in an in vitro system devoid of platelets [18]. As the reactions associated with initial tumor cellendothelial attachment remain obscure [18], the principal hypothesis on the metastatic process proposes that vWF may mediate the adhesion of cellular aggregates (tumor cells + platelets) to vascular endothelium [18-20]. Our studies have been performed in a cell coculture system (endothelial and human tumor cells) devoid of platelets [1]. In this system, the presence of tumor cells stimulated endothelial cells to produce vWF [1], and vWF increased the adherence between the two cell types (endothelial and tumor). To the best of our knowledge, this is the first report that vWF can directly modulate the interaction between endothelial and human tumor cells in the absence of platelets. The observation that human tumor cells increase the release of vWF from endothelial cells suggests that vWF can be tested as a prognostic factor, favoring tumor growth and dissemination. As vWF is likely involved in the tumor-induced platelet aggregation, which is counteracted by various anti-platelet agents and heparin in experimental studies [21, 22], we investigated the effect of these agents on the interaction between endothelial and tumor cells in our system. Some of these agents were effective in reducing the adhesion between tumor and endothelial cells, as well as vWF production, ticlopidine being the most effective of the substances tested so far. This result supports prospective trials on cancer treatment with anti-platelet agents, particularly those Biomed & Pharmacother 2000 ; 54 : 431-6
435
blocking the interaction between vWF and its specific glycoprotein platelet receptor. In conclusion, our experimental data show that vWF mediates the adherence of human tumor cells to human endothelial cells, and that its production can be reduced by an anti-platelet drug (ticlopidine), and to some extent by heparin. ACKNOWLEDGEMENTS We thank Professor B. Shapiro from the University of Michigan, Ann Arbor, MI, USA for careful revision of the manuscript. REFERENCES 1 Morganti M, Mittermayer C, Henze U, Carpi A, Sagripanti A. Expression of tissue-type plasminogen activator inhibitor and von Willebrand factor in the supernatant of endothelial cell cultures in response to the seeding of adenocarcinoma cell line HRT-18. Biomed Pharmacother 1996 ; 50 : 373-5. 2 Trousseau A. Phlegmasia alba dolens. Clin Med Hotel Dieu de Paris (London) 1865 ; 3 : 654. 3 Zacharski L, Costantini V. Procoagulant properties of tumor cells. Biomed Prog 1992 ; 5 : 55-8. 4 Sagripanti A, Carpi A, Grassi B. Plasmatic markers of haemostatic system activation in patients with solid meoplasms. J Nucl Med All Sci 1990 ; 34 : 321-32. 5 Sagripanti A, Carpi A, Mittermayer C. The interactions between cancer and the hemostatic system: basic aspects and clinical implications. In: Carpi A, Sagripanti A, Mittermayer Ch, Eds. Progress in clinical oncology. Munich: Sympomed; 1992. p. 294319. 6 Carpi A, Sagripanti A, Poddighe R, Gherardurci G, Nicolini A. Cancer incidence and mortality in patients with heart disease. Am J Clin Onc 1995 ; 18 : 15-8. 7 Costantini V, Zacharski L. Fibrin and cancer. Thromb Haemostas 1992 ; 69 : 406-14. 8 Jaffe EA, Hoyer LW, Nachmann RL. Synthesis of antihemophilic factor antigen by cultured human endothelial cells. J Clin Invest 1973 ; 52 : 2757-64. 9 Gimbrone MA, Shefton EI, Cruise SA. Isolation and primary culture of endothelial cells from human umbilical vessesl. TCA Manual. City: Publisher; 1978. 10 Knedler A, Ham RG. Optimized medium for clonal growth of human microvascular endothelial cells with minimal serum. In Vitro 1987 ; 23 : 481-91. 11 Brock TA, Dvorak HF, Senger DR. Tumour secreted vascular permeability factor increases cytosolic Ca2+ and von Willebrand factor release in human endothelial cells. Am J Pathol 1991 ; 138 : 213-21. 12 Schmidt MB. Die Verbreitungswege der karzinome und die Beziehungen generalisierter Sarkome zu den leukaemischen. Neubildungen. Jena: Fisher; 1903. 13 Pearlstein E, Salk PL, Yogeeswaran G, Karpatkin S. Correlation between spontaneous metastatic potential, platelet aggregating activity of cell surface extracts, and cell surface syalilation in 10 metastatic-variant derivative of a rat renal sarcoma cell line. Proc Natl Acad Sci U S A 1980 ; 77 : 4336-9. 14 Pineto HM, Verheul HMW, D’Amato RJ, Folkman J. Involvement of platelets in tumor angiogenesis? Lancet 1998 ; 352 : 1775-9.
436
M. Morganti et al.
15 Nierodzik ML, Klepfish A, Karpatkin S. Role of platelets, thrombin, integrin IIb-IIIa, fibronectin and von Willebrand factor on tumor adhesion in vitro and metastasis in vivo. Thromb Haemost 1995 ; 74 : 282-90. 16 Karpatkin S, Pearistein E, Ambrogio C, Coller BS. Role of adhesive proteins in platelet tumor interaction in vitro and metastasis formation in vivo. J Clin Invest 1988 ; 81 : 1012-9. 17 Nierodzik ML, Plotkin A, Kajumo F, Karpatkin S. Thrombin stimulates tumor-platelet adhesion in vitro and metastasis in vivo. J Clin Invest 1991 ; 87 : 229-36. 18 Floyd CM, Irani K, Kind PD, Kessler CM. Von Willebrand factor interacts with malignant hematopoietic cell lines: evidence for the presence of specific binding sites and modification of von Willebrand factor structure and function. J Lab Clin Med 1992 ; 119 : 467-76.
19 Sindelar WF, Tralka TS, Ketcham AS. Electron microscopic observation on formation of pulmonary metastasis. J Surg Res 1975 ; 18 : 137-61. 20 Mehta P. Potential role of platelets in the pathogenesis of tumor metastasis. Blood 1984 ; 63 : 55-63. 21 Lenern WA, Pearlstein E, Ambrogio C, Karpatkin S. A new mechanism for tumor-induced platelet aggregation. Comparison with mechanisms shared by other tumors with possible pharmacologic strategy toward prevention of metastases. Int J Cancer 1983 ; 31 : 463-9. 22 Yamamoto K, Kitagawa H, Tanoue K, Tsuruo T, Yamamoto N, Yamazaki H. Role of heparin in tumor cell-induced platelet aggregation. Thromb Haemost 1986 ; 56 : 90-4.
Biomed & Pharmacother 2000 ; 54 : 431-6
Original article
Von Willebrand’s factor mediates the adherence of human tumoral cells to human endothelial cells and ticlopidine interferes with this effect M. Morganti1, A. Carpi2, B. Amo-Takyi1, A. Sagripanti2, A. Nicolini2, R. Giardino3, C. Mittermayer1* 1 Institute of Pathology RWTH, Klinikum, Pauwelstrasse 30, Aachen, Germany 52074; 2Departments of Aging and Reproduction, Internal Medicine, University of Pisa, Italy; 3Department of Experimental Surgery, Rizzoli Orthopaedic Institute, Bologna, Italy
Summary – The aim of this study was to explore whether von Willebrand’s factor (vWF) plays a role in the adhesion of human colon tumor cells to human endothelial cells in our coculture system. Cell colony density was evaluated basally (endothelial plus colon tumor cells) and following the addition of: purified vWF, vWF plus vWF-blocking antibodies, antibodies against various integrins and adhesion molecules (α2 b integrin, β1 integrin, β3 integrin, intercellular adhesion molecule-I, intercellular adhesion molecule-II, vitronectin receptor CD61 CD51, laminin α6/β4 receptor), and various drugs inhibiting the hemostatic system (ticlopidine, heparin, acetyl salicylic acid [ASA], defibrotide, indobuphen, dipyridamole, sulfinpyrazone). Furthermore, vWF concentration was measured in the supernatant fluid of the coculture system basally and following the addition of the above-listed drugs. Cell colony density (as measured by light absorption) increased by 33% following the addition of vWF and returned to a value similar to the basal level with antibodies against vWF, while it did not change significantly following the addition of antibodies against the other integrins or adhesion molecules tested. The same parameter was reduced by 35% following the addition of ticlopidine, while it showed a smaller or no change with the other drugs tested. Similarly, vWF concentration in the cell coculture supernatant showed the greatest reduction (from 0.22 to 0.11 mg/mL) following the addition of ticlopidine. These data suggest that vWF mediates the adherence of human tumor cells to human endothelial cells and that ticlopidine interferes with this effect. © 2000 Éditions scientifiques et médicales Elsevier SAS colon tumor cells / endothelial cells / ticlopidine / von Willebrand’s factor
In a previous paper we evaluated the effect of human colon cancer cells (HRT-18) on endothelial cells from human umbilical vein (HUVEC) in a cell coculture system [1]. We showed that the human tumoral cells modify the human endothelial cell activity in such a way as to reduce their production of tissue type plasminogen activator (tPA) and to increase the produc-
*Correspondence and reprints
tion of plasminogen activator inhibitor (PAI-1) as well as von Willebrand’s factor (vWf) [1]. Our observation that the interaction between tumor and endothelial cells led to depression of the fibrinolytic system and enhancement of vWf was in agreement with previous studies, supporting the hypothesis that the hemostatic alterations produced by tumor cells can play a role in cancer progression [2-6]. In this respect, fibrin formation has been studied and is considered to play an important role [7].
432
M. Morganti et al.
MATERIALS AND METHODS The adhesion of two human cell types – endothelial cells from human umbilical vein (HUVEC) and human colon cancer cells (HRT-18) – was evaluated by the cell adhesion assay. This assay was performed without addition of vWF (standard) and with vWF addition. Furthermore, the effect of antibodies aganist vWF or other adhesion molecules on this assay was evaluated. Finally, the effect of various drugs inhibiting the hemostatic system on the adhesion of the two human cell types has been evaluated. Details on the culture cells were presented in our previous paper [1].
sequent steps were those of the standard procedure, as reported above. Cell-adhesion assay in presence of vWf and vWf-blocking antibodies Three 96 multiwell plates were incubated for 60 min at 37° C in a dilution 1:100 vol/vol vWF purified antibodies/HAMS’s medium. The cultures were washed twice with PBS, then 50 µL/well of a suspension of 10,000 tumoral cells HRT-18/mL HAMS’s plus 1:100 vol/vol vWF antibodies was added and the cocultures incubated at 37° C for 6 hours. The subsequent steps were those of the standard procedure.
Cell-adhesion assay, standard procedure Confluent endothelial HUVEC cell cultures grown in 96 multiwell plates were incubated for 60 min at 37° C in serum-free medium (HAMS’s plus glutamine-gentamycin, Gibco, Grand Island, USA) [8-10]. The cultures were washed twice with sterile phosphate buffered saline (PBS), then 50 mL/well of a suspension of 10,000 tumoral cells HRT-18/mL (phenotypically heterogeneous human rectal cancer cell line, Thompkins et al. 1974) HAMS’s was added and the cocultures incubated at 37° C for six hours. The medium containing unattached cells was discarded and cocultures were rinsed in cold PBS to remove unattached cells, then fixed with 70% ethanol and stained with 100 µL/well of a solution of 0.1% crystal violet in distilled water for 25 minutes. The tumoral HRT-18 colonies were counted and measured. Multiwell plates were dried with paper pads and dye was re-suspended with 50 µL of 0.2% Triton X-100 in water plus 50 mL propanol/well. The stain intensity was evaluated by chromatometry with an ELISA reader at 590 nm. Cell-adhesion assay in presence of vWf purified protein Three 96 multiwell plates were incubated for 60 min at 37° C in a dilution 1:100 vol/vol vWf purified protein/HAMS’s medium. The cultures were washed twice with PBS, then 50 mL/well of a suspension of 10,000 tumoral cells HRT-18/mL HAMS’s plus 1:100 vol/vol vWf purified protein was added and the cocultures incubated at 37° C for 6 hours. The sub-
Cell-adhesion assay in presence of surface and adhesion molecules blocking antibodies The HAMS’s medium used for coculture incubation was added with the following antibodies against different integrins and adhesion molecules: β1 integrin (Human β1 integrin monoclonal clone CD29 Endogen, MD, USA) at 1:100 vol/vol, βIII integrin (monoclonal clone, Aachen Laboratory, Germany) at 1:100 vol/vol, α2 β1 integrin (Integrin α2β1 monoclonal clone CD49 Dako SPA, Milan, Italy) at 1:10 vol/vol, intercellular adhesion molecule ICAM-1 (Human ICAM-1 clone CD54 R and D System INC, MN USA) at 1:100 vol/vol, ICAM-2 (ICAM-2 SC Chimera R and D System INC, MN USA) at 1:100 vol/vol, vitronectin receptor (human vitronectin monoclonal clone CD51 Endogen, MD, USA) at 1:100 vol/vol laminin α6β4 receptor (laminin policlonal antibody, Dako, Milan, Italy) at 1:100 vol/vol. The subsequent steps were those of the standard procedure. Cell-adhesion assay following addition of drugs inhibiting the hemostatic system The HAMS’s medium used for coculture incubation (6 hr at 37° C) was respectively added with one of the following drugs: ticlopidine 50 ng/mL, acetyl salicylic acid (ASA) 50 ng/mL, heparin 50 ng/mL, defibrotide 50 ng/mL, indobuphen 50 ng/mL, dipyridamole 50 ng/mL, sulfinpyrazone 50 ng/mL. The subsequent steps were those of the standard procedure. Biomed & Pharmacother 2000 ; 54 : 431-6
vWF in cocultures of human endothelial and colon tumor cells
433
Effect of anti-hemostatic drugs on vWf production A concentration of 160,000 endothelial cell/mL was seeded in 96-well tissue culture plates (BectonDickinson, NJ, USA) at 100 mL/well in a humidified atmosphere at 37° C for 24 h. The cell line HRT-18 was washed with PBS and disaggregated with 2 mL PBS/Trypsin 0.125% at 37° C for three minutes, then the cells were resuspended in 10 mL serum-free RPMI and centrifuged at 1,100 × g for 10 min. The supernatant was discarded and the cells were resuspended in the same medium to obtain a concentration of 10,000 cells/mL. The endothelial cell cultures in 96 wells were washed twice with PBS, then 50 mL of tumor cell suspension (10,000 tumor cells/mL) were added to each well. To seven of the 96 well plates containing cocultures 50 ng of ticlopidine, ASA, heparin, defibrotide, indobuphen, dipyridamole and sulfinpyrazone were added, respectively. Then cocultures were incubated at 37° C in a humidified atmosphere for 24 h. Thereafter the coculture system was terminated and the supernatant separated from the cells; the supernatant was centrifuged twice at 1,100 x g for 10 min to remove cell residues. The cell supernatant was managed according to the standard procedure for measurement of vWF by enzyme-linked immunosorbent assay (ELISA) (Asserachrom Diagnostica Stago, Boehringer, Mannheim, Germany). The spectroscope used was a Titertek Multiskan Plus MKII (Labsystem, Finland). Each probe was repeated three times with three different primary HUVEC cultures.
Figure 1. Cell colony density following the addition of HRT-18 to HUVEC cells and successively of purified vWF to the system.
significant change of the signal following the addition of vWF with antibodies. These data indicate that the effect following vWF addition on the cell coculture system observed in figure 1 is really due to vWF as it was blocked by the specific antibody. The experiment shown in figure 3 compares the reported effect of vWF antibodies with that of antibodies against the above-mentioned integrin and adhesion molecules (α2b integrin, β1 integrin, β3 integrin, intercellular adhesion molecule-I, intercellular adhesion molecule-II, vitronectin receptor CD61 CD51, laminin α6/ b4 receptor), which may be involved in the interaction between HUVEC and HRT-18 tumor cells. The figure shows that the only antibody able to reduce the adhesion and/or the number of cell colonies was that against vWF. The experiment shown in figure 4 explores whether the drugs mentioned above (ticlopidine, ASA, hep-
RESULTS Figure 1 shows the cell colony density following the addition of HRT-18 to HUVEC cells and subsequently of purified vWF to the entire system. It can be observed that the cell colony density increases as expected after addition of HRT-18 cells; however, it shows a further increase following the addition of vWF addition. This suggests that vWF may stimulate attachment and or growth of HRT-18 colonies. Figure 2 shows the results of an experimental setup different from the previous one only for the final addition to the system of antibodies to vWF together with purified vWF. The increase in the chromatometry signal after the addition of HRT-18 to HUVEC cells was the same as in figure 1; however, there was no Biomed & Pharmacother 2000 ; 54 : 431-6
Figure 2. Cell colony density following the addition of HRT-18 to HUVEC cells and successively of vWF plus vWF blocking antibodies to the system.
434
M. Morganti et al.
Figure 3. Cell colony density following the addition of HRT-18 to HUVEC cells and successively of antibodies against vWF or different integrins and adhesion molecules to the system.
arin, defibrotide, indobophen, dipyridamole, sulphinpyrazone) known to influence the hemostatic system exert some effect on the adhesion of the cell colonies in our system. Ticlopidine showed the greatest effect in reducing number and or adhesion of cell colonies so that the chromatometry signal returned to a value very similar to that obtained by the presence of HUVEC cells alone. A similar lighter effect was also shown following the addition of ASA and heparin. Figure 5 shows the concentration value of vWF in the supernatant fluid of the HUVEC-HRT-18 coculture system before and after the addition of the seven drugs. The vWF concentration was markedly reduced following the addition of ticlopidine (from 0.22 to 0.11 ng/mL) and to a lesser extent (about 50% less) after the addition of ASA and heparin.
Figure 4. Cell colony density following the addition of HRT-18 to HUVEC cells and successively of seven different drugs inhibiting haemostasis to the system.
Figure 5. vWF concentration in the supernatant of the HUVECHRT-18 coculture system before and after the addition of seven different drugs inhibiting the haemostatic system.
DISCUSSION Our previous finding that the vWF concentration value in the supernatant of HUVEC cocultures changed from undetectable to an easily measurable value [1] following the addition of human colorectal tumor cells HRT-18 stimulated this second study, which aimed to clarify the meaning of this increased vWF production by endothelial cells. The present study shows that the addition of vWF to our system of tumor and HUVEC cells induces an enhancement of the adhesion among the cells in the coculture or their number. The use of various antibodies against vWF and various integrins and adhesion molecules known to regulate the interaction – clearly intercellular adhesion – shows that vWF is responsible for the enhanced adhesion of tumor to endothelial cell in our system. Therefore, it is likely that tumor cells change some properties of the HUVEC cells by inducing vWF production, which in turn increases the binding of tumor to endothelial cells. It has been reported that vascular growth factor VGEF can induce endothelial cells to release vWF [11]. It is well extablished that vWF promotes platelet adhesion to the subendothelial matrix; moreover, it increases platelet aggregation by interacting with platelet surface glycoproteins. One old, established observation was that platelet aggregates surrounded cancer cells in human pulmonary arterioles [12]. A successive correlation was shown between in vivo malignancy and platelet aggregating activity of substances derived from animal tumor cells [13]. More Biomed & Pharmacother 2000 ; 54 : 431-6
vWF in cocultures of human endothelial and colon tumor cells
recently, a leading hypothesis involves platelets in tumors’ angiogenesis [14]. This hypothesis is based on chemical and preclinical findings that tumor angiogenesis is dependent not only on endothelial and tumour cells but also on platelet-endothelium interaction [14]. Studies have shown that vWF plays an important role in the adhesion of tumor cells to platelets (in vitro) [15] and in facilitating implantation of lung metastasis after inoculation of tumor cells in animals [15-17]. Human and animal colon carcinoma cells have been used for these studies [15-17]. To our knowledge these effects of vWF have been attributed to its participation in the interaction between platelets and tumor cells [15-17]. More recently, vWF has been shown to interact directly with malignant hematopoietic cell lines in an in vitro system devoid of platelets [18]. As the reactions associated with initial tumor cellendothelial attachment remain obscure [18], the principal hypothesis on the metastatic process proposes that vWF may mediate the adhesion of cellular aggregates (tumor cells + platelets) to vascular endothelium [18-20]. Our studies have been performed in a cell coculture system (endothelial and human tumor cells) devoid of platelets [1]. In this system, the presence of tumor cells stimulated endothelial cells to produce vWF [1], and vWF increased the adherence between the two cell types (endothelial and tumor). To the best of our knowledge, this is the first report that vWF can directly modulate the interaction between endothelial and human tumor cells in the absence of platelets. The observation that human tumor cells increase the release of vWF from endothelial cells suggests that vWF can be tested as a prognostic factor, favoring tumor growth and dissemination. As vWF is likely involved in the tumor-induced platelet aggregation, which is counteracted by various anti-platelet agents and heparin in experimental studies [21, 22], we investigated the effect of these agents on the interaction between endothelial and tumor cells in our system. Some of these agents were effective in reducing the adhesion between tumor and endothelial cells, as well as vWF production, ticlopidine being the most effective of the substances tested so far. This result supports prospective trials on cancer treatment with anti-platelet agents, particularly those Biomed & Pharmacother 2000 ; 54 : 431-6
435
blocking the interaction between vWF and its specific glycoprotein platelet receptor. In conclusion, our experimental data show that vWF mediates the adherence of human tumor cells to human endothelial cells, and that its production can be reduced by an anti-platelet drug (ticlopidine), and to some extent by heparin. ACKNOWLEDGEMENTS We thank Professor B. Shapiro from the University of Michigan, Ann Arbor, MI, USA for careful revision of the manuscript. REFERENCES 1 Morganti M, Mittermayer C, Henze U, Carpi A, Sagripanti A. Expression of tissue-type plasminogen activator inhibitor and von Willebrand factor in the supernatant of endothelial cell cultures in response to the seeding of adenocarcinoma cell line HRT-18. Biomed Pharmacother 1996 ; 50 : 373-5. 2 Trousseau A. Phlegmasia alba dolens. Clin Med Hotel Dieu de Paris (London) 1865 ; 3 : 654. 3 Zacharski L, Costantini V. Procoagulant properties of tumor cells. Biomed Prog 1992 ; 5 : 55-8. 4 Sagripanti A, Carpi A, Grassi B. Plasmatic markers of haemostatic system activation in patients with solid meoplasms. J Nucl Med All Sci 1990 ; 34 : 321-32. 5 Sagripanti A, Carpi A, Mittermayer C. The interactions between cancer and the hemostatic system: basic aspects and clinical implications. In: Carpi A, Sagripanti A, Mittermayer Ch, Eds. Progress in clinical oncology. Munich: Sympomed; 1992. p. 294319. 6 Carpi A, Sagripanti A, Poddighe R, Gherardurci G, Nicolini A. Cancer incidence and mortality in patients with heart disease. Am J Clin Onc 1995 ; 18 : 15-8. 7 Costantini V, Zacharski L. Fibrin and cancer. Thromb Haemostas 1992 ; 69 : 406-14. 8 Jaffe EA, Hoyer LW, Nachmann RL. Synthesis of antihemophilic factor antigen by cultured human endothelial cells. J Clin Invest 1973 ; 52 : 2757-64. 9 Gimbrone MA, Shefton EI, Cruise SA. Isolation and primary culture of endothelial cells from human umbilical vessesl. TCA Manual. City: Publisher; 1978. 10 Knedler A, Ham RG. Optimized medium for clonal growth of human microvascular endothelial cells with minimal serum. In Vitro 1987 ; 23 : 481-91. 11 Brock TA, Dvorak HF, Senger DR. Tumour secreted vascular permeability factor increases cytosolic Ca2+ and von Willebrand factor release in human endothelial cells. Am J Pathol 1991 ; 138 : 213-21. 12 Schmidt MB. Die Verbreitungswege der karzinome und die Beziehungen generalisierter Sarkome zu den leukaemischen. Neubildungen. Jena: Fisher; 1903. 13 Pearlstein E, Salk PL, Yogeeswaran G, Karpatkin S. Correlation between spontaneous metastatic potential, platelet aggregating activity of cell surface extracts, and cell surface syalilation in 10 metastatic-variant derivative of a rat renal sarcoma cell line. Proc Natl Acad Sci U S A 1980 ; 77 : 4336-9. 14 Pineto HM, Verheul HMW, D’Amato RJ, Folkman J. Involvement of platelets in tumor angiogenesis? Lancet 1998 ; 352 : 1775-9.
436
M. Morganti et al.
15 Nierodzik ML, Klepfish A, Karpatkin S. Role of platelets, thrombin, integrin IIb-IIIa, fibronectin and von Willebrand factor on tumor adhesion in vitro and metastasis in vivo. Thromb Haemost 1995 ; 74 : 282-90. 16 Karpatkin S, Pearistein E, Ambrogio C, Coller BS. Role of adhesive proteins in platelet tumor interaction in vitro and metastasis formation in vivo. J Clin Invest 1988 ; 81 : 1012-9. 17 Nierodzik ML, Plotkin A, Kajumo F, Karpatkin S. Thrombin stimulates tumor-platelet adhesion in vitro and metastasis in vivo. J Clin Invest 1991 ; 87 : 229-36. 18 Floyd CM, Irani K, Kind PD, Kessler CM. Von Willebrand factor interacts with malignant hematopoietic cell lines: evidence for the presence of specific binding sites and modification of von Willebrand factor structure and function. J Lab Clin Med 1992 ; 119 : 467-76.
19 Sindelar WF, Tralka TS, Ketcham AS. Electron microscopic observation on formation of pulmonary metastasis. J Surg Res 1975 ; 18 : 137-61. 20 Mehta P. Potential role of platelets in the pathogenesis of tumor metastasis. Blood 1984 ; 63 : 55-63. 21 Lenern WA, Pearlstein E, Ambrogio C, Karpatkin S. A new mechanism for tumor-induced platelet aggregation. Comparison with mechanisms shared by other tumors with possible pharmacologic strategy toward prevention of metastases. Int J Cancer 1983 ; 31 : 463-9. 22 Yamamoto K, Kitagawa H, Tanoue K, Tsuruo T, Yamamoto N, Yamazaki H. Role of heparin in tumor cell-induced platelet aggregation. Thromb Haemost 1986 ; 56 : 90-4.
Biomed & Pharmacother 2000 ; 54 : 431-6