Clinical & Experimental Metastasis 18: 239–244, 2000. © 2000 Kluwer Academic Publishers. Printed in the Netherlands.
239
In vitro modulation of human lung cancer cell line invasiveness by antisense cDNA of tissue factor pathway inhibitor-2 Sajani S. Lakka1 , Santhi D. Konduri1, Sanjeeva Mohanam1 , Garth L. Nicolson2 & Jasti S. Rao1 1 Department of
Neurosurgery, The University of Texas M. D. Anderson Cancer Center, Houston, Texas, USA; 2 Institute for Molecular Medicine, Huntington Beach, California, USA Received 20 September 2000; accepted in revised form 12 October 2000
Key words: TFPI-2, lung cancer, invasion
Abstract Human tissue factor pathway inhibitor-2 (TFPI-2) is a Kunitz-type serine protease inhibitor that inhibits plasmin, trypsin, chymotrypsin, cathepsin G and plasma kallikrein but not urokinase (uPA) or tissue-type plasminogen activator and thrombin. Earlier studies from our and other laboratories have shown that the production of TFPI-2 is downregulated during the progression of various cancers. To investigate the role of TFPI-2 in the invasion and metastasis of lung tumors, the human lung cancer cell line A549, which produces high levels of TFPI-2, was stably transfected with a vector capable of expressing an antisense transcript complementary to the full-length TFPI-2 mRNA. Northern blot analysis was used to quantify the TFPI-2 mRNA transcript, and western blot analysis was used to measure TFPI-2 protein levels in parental cells and stably transfected (vector and antisense) clones. The levels of TFPI-2 mRNA and protein were significantly less in antisense clones than in the parental and vector controls. The invasive potential of the parental cells and stably transfected vector clones in vitro, as measured by the Matrigel invasion assay, was also markedly less than that of antisense clones. Further characterization of these clones showed that more cells migrated from antisense clones than from parental and vector clones. These data suggest that TFPI-2 is critical for the invasion and metastasis of lung cancer and that the downregulation of TFPI-2 production may be a feasible approach to increase invasiveness and metastasis. Abbreviations: ECM – extracellular matrix; uPA – urokinase-type plasminogen activator; uPAR – uPA receptor; tPA – tissue-type PA; TFPI-2 – tissue factor pathway inhibitor-2; MMP – matrix metalloproteinase; SDS-PAGE – sodium dodecyl sulfate polyacrylamide gel electrophoresis; GAPDH – glyceraldehyde phosphate dehydrogenase Introduction Plasminogen activators (PAs) are an important means by which tumor cells trigger the degradation of the extracellular matrix (ECM), thereby facilitating the invasiveness and proliferation of the tumor cells [1, 2]. Cell migration and the resultant tissue invasiveness have been attributed to the ability of urokinase plasminogen activator (uPA) to convert plasminogen to plasmin [3]. Plasmin, a serine protease of broad specificity, cleaves several specific proteins at the level of the cell-cell and cell-ECM contact. The presence of the urokinase plasminogen activator receptor (uPAR) and plasminogen in the same cell [4] allows the generation of plasmin on the cell surface [5], which can act as a space-oriented cleaver. Plasmin degrades the ECM and activates many proteases, leading to tumor cell invasion. ECM-degrading proteinases and their inhibitors are thought to play an essential role in the invasiveness and metastasis of certain tumor cell lines. For example the components of Correspondence to: Jasti S. Rao, Division of Cancer Biology, Department of Biomedical and Therapeutic Sciences, University of Illinois College of Medicine, P.O. Box 1649, Peoria, IL 61656, USA. Tel: +1-309-671-3445; Fax: +1-309-671-8403; E-mail:
[email protected]
the plasminogen-activation system present in non-small-cell lung cancer have been studied by different techniques, and all the components are present at different levels in lung cancer tissue [5]. In addition, the presence of uPA and uPAR mRNA in fibrotic lung fibroblasts suggests the involvement of the uPA system in the progression of lung fibrosis [6]. High levels of uPA and uPAR have also been detected in human non-small-cell lung cancer tissue [7]. In fact, the uPAR level is of prognostic importance because it gauges the degree of local proteolysis [8]. Similarly factor VIIatissue factor complex was shown to be necessary for lung metastasis (9) Tissue factor pathway inhibitor-2 (TFPI-2) is a 32-kDa novel serine protease inhibitor [10–12] homologous to TFPI that can inhibit plasmin, trypsin, plasma kallikrien, chymotrypsin, cathepsin G, factor X1a and factor VIIa-tissue factor complex. The gene that codes for TFPI-2 has been mapped to chromosome 7q22 by fluorescence in situ hybridization [13]. An earlier study showed that the recombinant matrix-associated serine protease inhibitor (MSPI) inhibits plasmin, regardless of whether the enzyme is associated with the tumor cells or the ECM, and also that the
240 recombinant MSPI inhibits matrix degradation and Matrigel invasion by HT-1080 fibrosarcoma cells [14]. It has been reported that TFPI mRNA and protein are expressed in the lung [15]. In the study we report here, we stably transfected a lung cancer cell line (A549) with a vector (pcDNA3.1) capable of expressing a transcript complementary to the fulllength TFPI-2 in an antisense (1.0 kb) orientation. We then selected antisense clones that produced low levels of protein and mRNA compared with levels in parental and vectortransfected clones. We observed that the downregulation of TFPI-2 led to increased migration and invasion of these clones compared with the behavior of parental and vector clones.
Materials and methods Cell culture The human lung cancer cell line A549 was obtained from the American Type Culture Collection (Manassas, Virginia) and was maintained in RPMI 1640 medium and supplemented with 10% fetal calf serum in a humidified atmosphere containing 5% CO2 at 37 ◦ C. Antibodies against TFPI-2 were generated in rabbits [10]. Preparation of constructs The human TFPI-2 cDNA fragment (1.0 kb) was amplified by the reverse transcriptase-polymerase chain reaction (RTPCR) using specific primers. The fragment was subcloned in an antisense orientation into pcDNA 3.1 vector (Invitrogen, San Diego, California). The orientation of the fragment was confirmed by restriction digestion followed by sequencing and showed 100% homology with the published sequence of TFPI-2 cDNA [9]. Transfection of cells with TFPI-2 cDNA Lung cancer (A549) cells were transfected with a 1.0-kb transcript in an antisense orientation and with pcDNA3.1 vector alone using lipofectamine (Life Technologies, Gaithersburg, Maryland). The cells were grown overnight (4 × 105 cells per 60-mm dish) and washed twice with phosphate-buffered saline. Plasmid DNA (1–2 µg) was mixed with 100 µl of serum-free medium, and 7 µl of lipofectamine was diluted in 100 µl of serum-free medium and added dropwise to the DNA. Twelve hours later, the medium was replaced with minimal essential medium (MEM) supplemented with NEAA Earl’s salts containing 10% fetal bovine serum. Forty-eight hours after transfection selection was initiated by growing the cells in complete medium containing 800 µg/ml G418 (Life Technologies). After selection, stable transfectants were expanded and used for the study.
S.S. Lakka et al. Western blotting The ECM proteins were extracted with 200 µl of the sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDSPAGE) sample buffers, and TFPI-2 protein levels were assayed by western blot analysis in 25 µl of this extract in parental, vector control and antisense clones. ECM samples were boiled for 5 min, and proteins were separated by SDS-PAGE using 15% polyacrylamide gels [16]. After electrophoresis, proteins were electroblotted onto nitrocellulose membranes, as described elsewhere [17] and blocked with 4% nonfat dry milk in 10 mM Tris-HCl and 150 mM NaCl (pH 7.4) containing 0.1% Tween-20 (TTBS) for 2 h at 23 ◦ C. The membranes were then incubated for 2 h at 23 ◦ C or overnight at 4 ◦ C with anti-TFPI-2 antibody, after which they were diluted 1:3000 in TTBS containing 1% bovine serum albumin. After several washes, the membranes were incubated for 1 h with a peroxidase-conjugated secondary antibody and then diluted 1:3000 in TTBS and 1% bovine serum albumin. The immunoreactive proteins were identified using the enhanced chemiluminescence (ECL) reagent system, following the manufacturer’s instructions. Northern blotting TFPI-2 mRNA levels were determined by northern blot analysis. The total cellular RNA was extracted from parental, vector and antisense clones using a previously described method [18]. After this, 10 µg of the total RNA was electrophoresed in agarose-formaldehyde gels, transferred overnight to a nylon membrane by capillary action, and cross-linked using ultraviolet irradiation. The membranes were hybridized overnight at 65 ◦ C with a TFPI-2 cDNA probe labeled with 32 P-deoxycytidine triphosphate by random primer labeling. The membranes were washed in 0.5% standard saline citrate and 0.1% SDS for 20 min at room temperature, and then for 15 min at 65 ◦ C, after which they were exposed to X-ray film at −70 ◦ C. After stripping, the membranes were rehybridized with glyceraldehyde phosphate dehydrogenase (GAPDH) cDNA, which was also used as an internal control. Growth curve The growth rate was assessed in parental, vector control and antisense clones plated to a density of 1 × 104 cells in 100mm culture dishes. The cells were counted in triplicate plates every 48 h until 14 days. Matrigel invasion assay Invasiveness was assessed in parental, vector control and antisense clones by an in vitro Matrigel invasion assay [19]. Briefly, transwell inserts with an 8-µm pore size were coated with Matrigel to a final concentration of 1 mg/ml. Cells were then trypsinized, and 1 × 106 cells were added in triplicate wells. The lower compartment chamber contained serum-free conditioned medium (CM) from A549 cells as a chemo-attractant. After incubation for 24 h at 37 ◦ C, cells
Regulation of TFPI-2 in lung cancer cell invasiveness
241
Figure 1. Western blot analysis of TFPI-2 production in parental, vector control and antisense clones (A549) using human anti-TFPI-2 polyclonal antibody, as described in ‘Materials and methods’.
that had passed to the other side of the membrane were fixed, stained with Hema-3 stain (CMS. 122-911) and also quantitated by MTT assay. The percentage invasion was calculated using the formula
Figure 2. Northern blot analysis of TFPI-2 mRNA in parental, vector control and antisense clones (A549). The total cellular RNA was extracted from parental, vector control and antisense clones. RNA (10 µg) was electrophoresed in agarose-formaldehyde gels, transferred to a nylon membrane and labeled with 32 P TFPI-2 cDNA. The membranes were then stripped and rehybridized with glyceraldehyde phosphate dehydrogenase (GAPDH) cDNA, which was used as an internal control.
Low aspect OD × 100 Low aspect OD + Upper aspect OD Migration assay Parental, vector control and antisense clones were cultured as spheroids in 100-mm2 tissue culture plates that had been precoated with 0.75% agar as described elsewhere [20]. Briefly, 3 × 106 cells were seeded onto agar-coated plates and cultured until spheroids formed. Spheroids were explanted onto a plastic surface and a total of 6–8 spheroids were plated for each group. At the end of the migration assay, spheroids were fixed and mounted onto glass microscope slides and photographed. The migration of cells from spheroid to monolayers was also measured using a microscope calibrated with a stage and ocular micrometer.
Results Transfection of A549 cells with expression construct Lung cancer cells (A549) were transfected with the eukaryotic expression vector pcDNA3.1(+) containing TFPI-2 cDNA (1.0 kb) in an antisense, orientation. Stable clones were selected in the presence of G418, and these clones were analyzed by in vitro assays to determine the amount of TFPI-2 and the invasive capabilities of the cells. Determination of TFPI-2 in ECM samples To determine the effect of antisense TFPI-2 cDNA transfection on TFPI-2 we seeded an equal number of antisense, vector control clones that were grown to 70–80% confluency in six-well plates, after which ECM proteins were subjected to SDS-PAGE and analyzed by western blotting. Figure 1 shows that Mr 33,000, 31,000 and 27,000 TFPI-2 protein bands were present in parental and vector clones but that these bands were reduced by more than 80% in antisense clones (Figure 1).
Figure 3. Parental, vector control and antisense clones were plated, and the growth rate was determined every 48 h as described in ‘Materials and methods’. The data shown are the average of values from four separate experiments.
Northern blot analysis of TFPI-2 levels in transfected clones To determine whether the antisense construct reduces TFPI2 mRNA levels, G418 resistance was characterized by northern blot analysis in stable antisense clones, vector clones and the parental cell line. This showed that the levels of TFPI-2 mRNA were significantly lower in antisense clones (Figure 2) than they were in parental and vector control cells. GAPDH, which was used as an internal control, confirmed that A549 antisense transfectants reduced the level of TFPI-2 mRNA. Levels of TFPI-2 mRNA were much higher in vector control and parental cells. Growth curves We examined the growth rate of parental, vector control and antisense clones for 14 days. Figure 3 shows that there was no difference in the growth rate among the parental, vector control and antisense clones.
242
S.S. Lakka et al.
Figure 4. Matrigel invasion assay of parental, vector control and transfected antisense clones (A549). In this experiment, 1 × 106 cells were layered onto Matrigel-coated transwell inserts, and the percentage of invasion was calculated as described in ‘Materials and methods’. After incubation for 24 h at 37 ◦ C, filters were removed and stained with Hema-3 (A) and invasion was quantitated (B) as described in ‘Materials and methods’. Data shown are the average of values from four separate experiments (± SD; P < 0.001).
Figure 5. Migration assay of parental, vector control and antisense clones A549. Plates were precoated with 0.75% agar, and 3 × 106 cells were seeded and cultured until spheroids were explanted on a plastic surface. At the end of the migration assay (24 h), spheroids were fixed, stained with Hema-3 and mounted glass microscope slides (A). The migration of cells from the spheroids was measured using a microscope calibrated with a stage and ocular micrometer (B). The data shown are the mean values ± SD of the results of four independent experiments from each group.
Invasive potentials of the stable clones The effect of antisense TFPI-2 transfection on in vitro invasion was compared among the antisense transfectant clones (AS1) and the parental and vector transfectants by the Matrigel invasion assay. We found no marked difference between parental and vector controls in the intensity of cells that invaded through the Matrigel. However, antisense clone cells that invaded through Matrigel was much higher compared with the invasion of parental and vector clones (Figure 4A). In addition, the 3-(4,5-dimethylthiazol-2yl)2,5-dimethylthiazolium bromide assay showed that 15% of parental cells and 16% of vector control cells invaded to the lower side of the membrane (Figure 4B). In contrast, 58% of antisense clones invaded to the lower side of the membrane, which was significantly higher (P < 0.001) than the invasiveness of parental and vector controls. Migration assay We also examined the migrating capacity of the transfected clones using spheroids made from parental, vector control and antisense clones. The migrating capacity of the antisense clones was significantly higher (P < 0.001) than that of parental and vector control, as shown by the number of cells that migrated out from the spheroids (Figure 5).
Discussion In the present study, we were able to reduce the levels of the TFPI-2 protein (see Figure 1) and mRNA (see Figure 2) in antisense transfected A549 cells, and resultant invasion
capacity of these antisense clones was 3–4 times (Figure 4) greater than that of parental and vector clones. There was, however, no significant difference in the growth rate among parental, vector control and antisense clones (Figure 3). The migrating capacity was also greater in antisense clones (A549) than in parental and vector controls (Figure 5). These collective results strongly suggest that decreased levels of TFPI-2 increase the invasive behavior of lung cancer cells. The invasive potential of tumor cells is related to tumor cell-associated proteolytic activity, which causes degradation of the ECM and basement membranes. In fact, proteolytic degradation of the ECM is considered an essential step in the invasion and metastasis of malignant cells [1, 20]. In general, cell surface-associated uPA and plasmin activity can be detected in tumor cells. Plasmin is responsible for the release of growth factors from the ECM, the activation of transforming growth factor β, the degradation and invasion of the ECM and the activation of certain prometalloproteinases by tumor cells [22, 23]. The inhibition of plasmin and uPA decreases the invasive potential of tumor cells [6, 24]. Proteases and their inhibitors have been shown to be involved in the invasiveness of tumor cells. Protease activities are always regulated by their natural inhibitors in vivo. Some human cancer cell lines secrete serine proteases, including trypsin [25, 26] and multiple forms of trypsin inhibitors. The inhibitors of these matrix proteases seem to be essential in preventing excess activity of the enzymes in vivo. Protease inhibitors capable of protecting the ECM from degradation by malignant cells could therefore be of use as therapeutic agents by preventing metastasis. One such inhibitor, human TFPI-2, is a 32-kDa Kunitz-type inhibitor found in the ECM of endothelial cells and fibroblasts [27, 28]. TFPI-
Regulation of TFPI-2 in lung cancer cell invasiveness 2 transcripts are highly abundant in the full-term placenta and widely produced in various adult human tissues, such as liver, skeletal muscle, heart, kidney and pancreas [12]. TFPI-2 is also an ECM-associated serine protease inhibitor [11] that plays a major role in wound healing and angiogenesis as well as in degrading the ECM during tumor cell invasion and metastasis. It has been shown that ECM-bound TFPI-2 is functional and that recombinant TFPI-2 is active at the cell:ECM interface [14]. Several ovarian carcinoma cell lines as well as T98G contain significant amounts of the transcript [12]. It has been reported that uPA and uPAR are overproduced in lung cancer [6, 7]. However, the study by Eitzman et al. [29] raised the possibility that interfering with the plasminogen-activation system at the level of plasmin rather than at the level of uPA and tPA could be a more effective strategy for blocking tumor cell invasion and metastasis. The reason for this is that TFPI-2 (Kunitz-type inhibitor) directly inhibits plasmin (derived from plasminogen by uPA or tPA, regardless of whether the enzyme is cell or matrixassociated) which is responsible for ECM degradation and invasion and for the activation of pro MMP-1 and pro MMP-3 by tumor cells. Interestingly, in addition to inhibiting plasmin and five other related proteases, TFPI-2 also inhibits tissue factor VIIa complexes, which are the cellular initiator of the extrinsic coagulation pathway [10, 30]. For example, TFPI-2 inhibited the activation of human factor X by the tissue factor VIIa complex on J82 bladder carcinoma cells [30]. Tissue factor also increased the production of vascular endothelial cell growth factor [31] and uPAR [32] in tumor cells, and its levels perfectly correlated with the degree of tumor angiogenesis and progression in gliomas, fibrosarcomas, melanomas, and breast and pancreatic tumors [15, 33–36]. Our results here suggest that decreased levels of TFPI-2 increased the invasive behavior of a lung cancer cell line in vitro. This protease inhibitor may therefore represent a therapeutic strategy for some types of tumors, and its therapeutic potential should therefore be investigated in other studies.
243 3.
4. 5. 6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16. 17.
18.
19.
Acknowledgements This work was supported by National Cancer Institute Grants CA-76350, CA-75692, and CA-75557 (JSR). We thank Lydia Soto for preparing the manuscript and Beth Notzon for manuscript review.
20.
21. 22.
References 23. 1. Dano K, Andreasen PA, Grondahl-Johnson J et al. Plasminogen activators, tissue degradation and cancer. Adv Cancer Res 1985; 44: 139–239. 2. Dano K, Grondhl-Hansen J, Eriksen J. The receptor for uPA: Stromal cell involvement in extracellular proteolysis during cancer invasion. In Bond JS and Barrett AJ (eds): Proteolysis and Protein Turnover. London: Portland Press 1993; 239–45.
24.
25.
Ossowski L. Plasminogen activator dependent pathways in the dissemination of human tumor cells in the chick embryo. Cell 1998; 52: 321–8. Stoppellie M, Tacchetti C, Cubellis MV. Autocrine saturation of prourokinase receptors on human A431 cells. Cell 1988; 45: 675–84. Pappot H, Brunner N. The plasminogen activation system and its role in lung cancer. A review. Lung Cancer 1995; 12: 1–12. Shetty S, Kumar A, Johnson AR et al. Differential expression of the urokinase receptor in fibroblasts from normal and fibrotic human lungs. Am J Respir Cell Mol Biol 1996; 15: 78–87. Sumuhito M, Atsuhiko S, Hiroshi H et al. Cancer cells overexpress mRNA of urokinase type plasminogen activator, its receptor and inhibitors in human non-small cell lung cancer tissue: Analysis by northern blotting and in situ hybridization. Int J Cancer 1998; 78: 286–92. Pedersen H, Brunner N, Francis D. Prognostic impact of urokinase, urokinase receptor and type I plasminogen activator inhibitor in squamous and large cell lung cancer tissue. Cancer Res 1994; 54: 4671–75. Mueller BM, Ruf W. Requirement for binding of catalytically active factor VIIa in tissue factor-dependent experimental metastasis. J Clin Invest 1998; 101: 1372–8. Sprecher CA, Kisiel W, Mathewes S et al. Molecular cloning, expression and partial characterization of a second human tissue-factorpathway inhibitor. Proc Natl Acad Sci USA 1994; 91: 3353–7. Rao CN, Reddy P, Liu YY et al. Extracellular matrix-associated serine protease inhibitors (M 33,000, 31,000 and 27,000) are single-gene products with differential glycosylation: cDNA cloning of the 33-kDa inhibitor reveals its identity to tissue factor pathway inhibitor-2. Arch Biochem Biophys 1996; 335: 82–92. Miyagi Y, Koshikawa N, Yasumitsu H et al. cDNA cloning and mRNA expression of a serine proteinase inhibitor secreted by cancer cells: identification as placental protein 5 and tissue factor pathway inhibitor-2. J Biochem 1994; 116: 939–42. Miyagi Y, Yasumitsu H, Eki T et al. Assignment of the human PP5/TFPI-2 gene to 7q22 by FISH and PCR-based human/rodent cell hybrid mapping panel analysis. Genomics 1996; 35: 267–8. Rao CN, Cook B, Liu YY et al. HT-1080 fibrosarcoma cell matrix degradation and invasion are inhibited by the matrix-associated serine protease inhibitor TFPI-2/33 kDa MSPI. Int J Cancer 1998; 76: 749– 59. Yamanobe F, Mochida S, Ohno A et al. Recombinant human tissue factor pathway inhibitor as a possible anticoagulant targeting hepatic sinusoidal walls. Thromb Res 1997; 85: 493–501. Laemmli UK Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970; 227: 680–5. Towbin H, Staehelin T, Gordon J. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications. Proc Natl Acad Sci USA 1979; 76: 4350–4. Chomczynski P, Sacchi N. Single-step method of RNA isolation by acid guanidium thiocyanate-phenol-chloroform extraction. Anal Biochem 1987; 162: 156–9. Mohanam S, Sawaya R, McCutcheon I et al. Modulation of in vitro invasion of human glioblastoma cells by urokinase type plasminogen activator receptor antibody. Cancer Res 1993; 53: 4143–7. Lund-Johansen M, Bjerkvig R, Humphrey PA et al. Effect of epidermal growth factor on glioma cell growth, migration and invasion in vitro. Cancer Res 1990; 50: 6039–44. Mignatti P, Rifkin DB. Biology and biochemistry of proteinases in tumor invasion. Physiol Rev 1993; 73: 161–95. Rao C N, Mohanam S, Puppala A et al. Regulation of ProMMP-1 and ProMMP-3 activation by tissue factor pathway inhibitor-2/matrix associated serine protease inhibitor. Biochem Biophys Res Commun 1999; 225: 94–8. Kwaan HC. The plasminogen-plasmin system in malignancy. Cancer Metastasis Rev 1992; 11: 291–311. Baker MS, Bleakley P, Woodrow GC et al. Inhibition of cancer cell urokinase plasminogen activator by its specific inhibitor PAI-1 and subsequent effects on extracellular matrix-degradation. Cancer Res 1990; 50: 4676–84. Koshikawa N, Yasumitsu H, Umeda M et al. Multiple secretion of matrix serine proteinases by human gastric carcinoma cell lines. Cancer Res 1992; 52: 5046–53.
244 26. Koshikawa K, Yasumitsu H, Nagashima Y et al. Identification of oneand-two-chain forms of trypsinogen 1 produced by a human gastric adenocarcinoma cell line. Biochem J 1994; 303: 187–90. 27. Lino M, Foster DC, Kisiel W. Quantification and characterization of human endothelial cell derived tissue factor pathway inhibitor-2. Arterioscler Thromb Vasc Biol 1998; 18: 40–6. 28. Rao CN, Liu Y, Peavey CL et al. Novel extracellular matrix associated serine proteinase inhibitors from human skin fibroblasts. Arch Biochem Biophys 1995; 317: 311–4. 29. Eitzman DT, Krauss JC, Shen T et al. Lack of plasminogen activator inhibitor-1 effect in a transgenic mouse model of metastatic melanoma. Blood 1996; 87: 4718–22. 30. Petersen LC, Sprecher CA, Foster DC et al. Inhibitory properties of a novel human Kunitz-type protease inhibitor homologous to tissue factor pathway inhibitor. Biochemistry 1996; 35: 266–72. 31. Ollivier V, Bentolila S, Chabbat J et al. Tissue factor-dependent VEGF production by human fibroblasts in response to activated factor VII. Blood 1998; 91: 2698–703.
S.S. Lakka et al. 32.
33.
34.
35.
36.
Taniguchi T, Kakkar AK, Tuddenham EG et al. Enhanced expression of urokinase receptor induced through the tissue factor VIIa pathway in human pancreatic cancer. Cancer Res 1998; 58: 4461–7. Hamada K, Kuratsu JI, Saitoh Y et al. Expression of tissue factor correlates with grade of malignancy in human glioma. Cancer 1996; 77: 1877–83. Contrini J, Hair G, Kretzer DL et al. In situ detection of tissue factor in vascular endothelial cells: Correlation with the malignant phenotype of human breast disease. Nat Med 1996; 2: 209–15. Mueller BM, Reisfeld RA, Edgington TS et al. Expression of tissue factor by human melanoma cells promotes efficient hematogenous metastasis. Proc Natl Acad Sci USA 1992; 89: 11832–6. Nakagawa K, Zhang Y, Tsuji H et al. The angiogenic effects of tissue factor on tumors and wounds. Semin Thromb Hemost 1998; 24: 207– 10.