Modulation of endothelial cell morphogenesis in vitro ... - Springer Link

Clinical & Experimental Metastasis 18: 337–342, 2000. © 2001 Kluwer Academic Publishers. Printed in the Netherlands.

337

Modulation of endothelial cell morphogenesis in vitro by MMP-9 during glial-endothelial cell interactions∗ Nirmala Chandrasekar1 , Sushma Jasti2 , W. K. Alfred-Yung2 , Francis Ali-Osman3 , Dzung H. Dinh4 , William C. Olivero4, Meena Gujrati5, Athanassios P. Kyritsis2, Garth L. Nicolson6, Jasti S. Rao1,4 & Sanjeeva Mohanam1 Division of 1 Cancer Biology, Department of Biomedical and Therapeutic Sciences and Departments of 4 Neurosurgery and 5 Pathology, UIC College of Medicine at Peoria, Peoria, Illinois, USA; Departments of 2 Neuro-Oncology and 3 Neurosurgery, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, USA; 6 The Institute for Molecular Medicine, Huntington Beach, California, USA Received 1 November 2000; accepted in revised form 15 January 2001

Key words: angiogenesis, extracellular matrix, glioblastoma, MMP-9, TIMPs

Abstract The purpose of this study was to investigate the roles of matrix metalloproteinase-9 (MMP-9) and tissue inhibitor of metalloproteinase-1 (TIMP-1) in the formation of capillary structures by human brain microvascular endothelial cells cocultured with SNB19 glioblastoma cells. Unstimulated cocultures did not form capillaries and produce MMP-9 but stimulation with the protein kinase C (PKC) activator 4-phorbol-12-myristate 13-acetate (PMA) produced MMP-9 and capillary networks. Addition of recombinant MMP-9 increased capillary formation. Anti-MMP-9 antibodies, TIMP-1, the synthetic MMPs inhibitor Batimastat (BB-94), and the PKC inhibitor calphostin-C all reduced MMP-9 activity and capillary network formation in these cocultures. Cytochalasin-D in the presence of PMA suppressed MMP-9 expression and capillary formation, but colchicine-B had no such effect. Finally, PMA-induced MMP-9 expression and capillary formation were inhibited by the MEKK-specific inhibitor PD98059. These results suggest that MMP-9 is important in endothelial cell morphogenesis and the formation of capillaries in glial/endothelial cocultures in vitro.

Introduction Angiogenesis, the formation of new capillary blood vessels, involves orchestrated changes in the interactions of endothelial cells with adjacent cells and with components of the underlying basement membrane matrix [1]. The action of matrix metalloproteinases (MMPs), which can cleave basement membrane and interstitial matrix molecules, has been shown to be necessary for angiogenesis in several in vivo and in vitro models. Of these MMPs, MMP-2 and particularly MMP-9 have been implicated in angiogenesis. Both are specific for type IV collagen degradation [2, 3], and MMP9 expression seems to be important in the invasion of the blood vessel wall [4]. The proteolytic activities of MMPs are controlled by a family of proteins called tissue inhibitors of metalloproteinases (TIMPs) [5, 6]. Direct evidence that MMPs are involved in angiogenesis comes from the finding ∗ The work was completed at UT M.D. Anderson Cancer Center, Houston,

Texas, USA. Correspondence to: Sanjeeva Mohanam, PhD, Division of Cancer Biology, Department of Biomedical and Therapeutic Sciences, UIC College of Medicine at Peoria, One Illini Drive, Box 1649, Peoria, IL 61656, USA. Tel: +1-309-671-3445; Fax: +1-309-671-8403; E-mail: [email protected]

that MMP inhibitors, both synthetic and endogenous, also inhibit angiogenesis in vivo and in in vitro models [7, 8]. Angiogenesis involves both stimulatory and inhibitory molecules [9, 10], which often act by modulating protease activity. For example, the protein kinase C (PKC) activator and tumor promoter 4-phorbol-12-myristate 13-acetate (PMA) induced angiogenesis in bovine microvascular endothelial cells [9]. The addition of exogenous type IV collagenases enhanced the formation of capillary tube-like structures by human endothelial cells plated on Matrigel, an extracellular matrix material [11]. Another stimulatory molecule, basic fibroblast growth factor, stimulated migration, protease production, and morphogenesis in endothelial cells [12]. Both vascular endothelial growth factor (VEGF) [13] and hepatocyte growth factor (HGF) are angiogenic in vitro; HGF increased the production of proteases as well [14]. Experimental approaches that reproduce the tumorinduced angiogenic process in vitro are useful for studying the molecular mechanisms involved in the interaction between tumor cells and endothelial cells during angiogenesis. Previous studies have shown that endothelial cells from the central nervous system differentiate morphologically into capillary-like structures in the presence of glial cells in vitro [15, 16]. The purpose of this study was to investigate

338 the role of direct and indirect inhibitors of MMPs on the expression of MMP-9 and TIMP-1 and the formation of capillary networks in cocultures of human brain microvascular endothelial cells and human glioblastoma cells.

Materials and methods Materials

N. Chandrasekar et al. cytochalasin-D, an actin polymerization inhibitor, at 5 µM; colchicine-B, a tubulin inhibitor, at 5 µM; PD98059, a MAPK/ERK kinase (MEKK) inhibitor, at 10 µM; Batimastat (BB-94), a broad MMP inhibitor, at 50, 100, or 200 ng/ml; TIMP-1, an endogenous MMP inhibitor, at 1 or 2 µg/ml; recombinant MMP-9 at 50, 100, or 200 ng/ml; and MMP-9 antibody at 0.5 or 1 µg/ml. All test compounds were added to the cocultures before the addition of PMA unless otherwise indicated.

Human brain microvascular endothelial cells and CS-C serum-free medium kits were purchased from Applied Cell Biology Research Institute (Kirkland, Washington). Factor VIII antibody was purchased from the DAKO Corporation (Carpinteria, California), and mouse monoclonal anti-92kD gelatinase B/type IV collagenase/MMP-9 antibody, human recombinant pro-MMP-9, TIMP-1 antibody, Calphostin-C, and PD98059 were purchased from Calbiochem (San Diego, California). Cytochalasin-D, colchicine-B, PMA, and fibronectin were obtained from Sigma Chemical Co. (St. Louis, Missouri). BB-94 (Batimastat) was obtained from British Biotech (Oxford, UK).

Immunofluorescent staining

Cell cultures

Gelatinolytic enzymes were identified in serum-free medium by gelatin zymography as described elsewhere [17]. Briefly, 20 µg of each sample was subjected to sodium dodecyl sulfate-8% polyacrylamide gel electrophoresis (SDSPAGE). After electrophoresis, the gels were rinsed twice with 2.5% Triton X-100 and incubated at 37 ◦ C for 20 h in 10 mM CaCl2 and 50 mM Tris-HCl (pH 7.5). The gels were then stained with Coomassie blue and destained. Gelatinolytic enzymes were visualized as transparent bands against the blue background of the gel.

Human brain microvascular endothelial cells were grown on tissue culture flasks coated with attachment factor in a humidified 5% CO2 atmosphere at 37 ◦ C in CS-C serumfree medium. The human glioblastoma cell line SNB19 was maintained in DMEM/F12 medium supplemented with 10% fetal bovine serum in a humidified atmosphere containing 5% CO2 at 37 ◦ C.

Endothelial cells were stained for factor VIII antigen by fixing the cells in 3.7% formaldehyde, blocking with 2% bovine serum albumin, and incubating the cells first with primary antibody against factor VIII and next with secondary antibody conjugated with biotin followed by Texas Redconjugated streptavidin (Gibco BRL, Gaithersburg, Maryland). The specimens were then washed and examined under a confocal scanning laser microscope. Gelatin zymography

Transfection of SNB19 cells Western blotting analysis To identify individual tumor cells and to establish a way to monitor tumor cell migration during capillary-like structure formation, we transfected parental SNB19 cells with the expression vector pEGFP-N1 (Clontech Laboratories, Palo Alto, California). To obtain cells with high-level GFP expression, cells were plated in limiting dilution and bright green fluorescent colonies were visually selected, lifted, expanded and subcloned by serial dilution to ensure purity. GFP expressing SNB19 cells were only used in studies involving confocal scanning microscopy.

Equal concentrations of serum-free conditioned medium were separated by SDS-PAGE in 15% polyacrylamide gels and transferred to nitrocellulose membranes. The membranes were incubated first with primary antibody to TIMP-1 and then with secondary antibody coupled to horseradish peroxidase, after which the bands were identified with ECL reagent according to the manufacturer’s protocol (Amersham Pharmacia Biotech, Piscataway, New Jersey).

Cocultures

Results

SNB19 cells were trypsinized, resuspended in stock culture medium, and 2 × 104 cells were plated on tissue culture chamber slides that had been treated with 20-µg/ml fibronectin. After an overnight incubation, the medium was removed and replaced with 4×104 human brain microvascular endothelial cells suspended in CS-C serum-free medium and the cocultures were maintained in 8-well Lab-tech chamber slides (for microscopic analyses) or 24-well plates (for protease/inhibitor analyses). Cocultures were treated with the following test compounds: PMA, a PKC activator, at 200 nM; calphostin-C, a PKC inhibitor, at 300 nM;

Phorbol ester induction of MMP-9 activity Human brain microvascular endothelial cells cultured with SNB19 cells on fibronectin or collagen formed a monolayer. Within 12–24 h after PMA was added to these cocultures, the flattened endothelial cells became more refractile, took on irregular shapes with long cell processes, and formed a capillary-like network (Figure 1b). To determine whether type IV collagenase could participate in this process, conditioned media from cultures were evaluated by zymograms for the presence of gelatinases. Zymographic analysis of the

MMP-9 in cerebral angiogenesis

339

Figure 1. Confocal laser scanning photomicrographs of human brain microvascular endothelial–glioblastoma cell (GFP expressing) cocultures after treatment with the indicated compounds. The endothelial cells are red and the SNB19 cells are green. Ab = antibody; PD = PD098059.

conditioned medium collected after 24 h of coculture without PMA showed the presence of a active and pro-forms of 72-kd gelatinase (MMP-2); medium from cells that had been treated with PMA, in contrast, contained a 92-kd collagenase (pro-MMP-9) and TIMP-1 (Figure 2) suggesting that MMP-9 expression may be associated with the capillary-like network formation in cocultures. Anti-MMP-9 antibodies and inhibition of capillary network formation To determine the role of MMP-9 in capillary formation, we added MMP-9 antibody of 0.5 µg and 1 µg in our cocultures and studied the effect on capillary-like structure formation. A control antibody had no effect (data not shown), whereas the addition of mouse monoclonal anti-MMP-9 antibody to the cocultures significantly reduced the amounts of MMP9 and TIMP-1 present (Figure 2a) and inhibited capillary formation (Figures 1f and 1g). The network-inhibition effect was apparent at antibody concentrations as low as 0.5 µg/ml and the inhibitory effect of antibody paralleled the morphological changes shown in Figures 1f and 1g.

Recombinant MMP-9 and capillary-like structure formation The results from MMP-9 antibody study showed that inhibition of MMP-9 activity decreases the ability of glial/endothelial cocultures to form capillaries. We next tested the effects of exogenous human recombinant MMP-9 on capillary-like structure formation in our cocultures. The addition of 50 ng/ml of recombinant human MMP-9 enhanced the formation of capillary network in the cocultures (Figure 1c), but higher concentrations progressively suppressed capillary formation (Figures 1d and 1e) suggesting that MMP-9 has an apparent bifunctional effect on capillary formation. TIMP-1 levels in the medium were found to progressively decrease as the collagenase concentration increased (Figure 2b), suggesting that TIMP-1 is modulated by the amount of MMP-9 present in the culture. MMP inhibitors and capillary network formation The results obtained using MMP-9 antibodies suggested that blocking the effect of MMP-9 inhibit capillary-like structure formation in our glial/endothelial cocultures model. We examined the effects of biological metalloproteinase inhibitor, TIMP-1 and TIMP-1 at a concentration of 1 or 2 µg/ml, inhibited the formation of capillary tubules in the cocultures by more than 90% (Figures 1h and 1i). At the lower doses,

340

N. Chandrasekar et al. 9 was inhibited by calphostin-C, a specific inhibitor of PKC (Figure 4). Capillary formation was also disrupted by calphostin-C, showing that MMP-9 induction as well as capillary formation by PMA depends on PKC in this system (Figure 1o). Calphostin-C treatment also significantly reduced TIMP-1 levels as indicated by Western blotting (Figure 4). MEKK inhibition, MMP-9 levels, and capillary formation Finally, to examine whether the loss of MAP kinase/ERK kinase activity could affect the induction of MMP-9, we added PD98059, an inhibitor of ERK, to the coculture. This compound completely blocked the induction of MMP9 (Figure 4), significantly reduced the amount of TIMP-1, and inhibited capillary formation (Figure 1p).

Discussion

Figure 2. Human brain microvascular endothelial–glioblastoma cell cocultures were treated with human recombinant MMP-9 and anti-MMP-9 antibody, and 24 h later (A) serum-free medium was collected and analyzed for collagenases by gelatin zymography and (B) for TIMP-1 by Western blotting as described in ‘Materials and methods’.

endothelial cells became aligned but did not form a network. The addition of exogenous TIMP-1 also significantly inhibited MMP-9 activity, as indicated by gelatin zymography (Figure 3). These results suggest that the biological regulation of collagenase activity can affect capillary formation. The addition of Batimastat, a synthetic MMPs inhibitor, inhibited the MMP-9 activity induced by PMA (Figure 3) and suppressed capillary formation (Figures 1j, 1k and 1l) in a dose-dependent manner. TIMP-1 levels also decreased in cocultures as the concentration of the batimastat increased, as indicated by Western blotting (Figure 3). The cytoskeleton and MMP activity Treatment of the cocultures with the actin-polymerization inhibitor cytochalasin-D suppressed both MMP-9 activity (Figure 4) and capillary formation (Figure 1m). On the other hand, adding colchicine-B, a tubulin inhibitor, had no effect on either PMA-induced MMP-9 expression (Figure 4) or capillary formation (Figure 1n). TIMP-1 levels were significantly reduced by cytochalasin-D treatment but were not affected by colchicine-B treatment (Figure 4). Protein kinase C and capillary formation MMP induction in glioma cells is known to depend on PKC expression. The PMA-induced expression of MMP-

Successful targeting of tumor angiogenesis for anticancer therapy requires a thorough understanding of the relationships among tumor cells, stromal cells, and endothelial cells, in particular the intricate intracellular signaling cascades that function during capillary formation. Angiogenesis is a complex phenomenon involving several sequential steps that depend on shifts in the balance between proteases and inhibitors [18]. We have demonstrated here that exogenous signals can change the behavior of cultured microvascular endothelial cells from noninvasive to highly invasive. Under normal conditions, human brain microvascular endothelial cells, when cultured with SNB19 glioblastoma cells, are confined to the surface of the supporting matrix; treating these cocultures with the PKC activator PMA induces the endothelial cells to invade the matrix, where they organize into capillary-like networks. Therefore, in vitro, PMA triggers phenomena that mimic the angiogenic processes that occur in vivo. The induction of capillary formation by PMA suggests that PMA modifies the behavior of the endothelial cells by inducing the synthesis of specific proteins that are essential for invasion to take place. PMA has also been shown to stimulate human umbilical vein endothelial cells to produce MMP-9 [19]. We also found that natural collagenase inhibitors as well as monoclonal antibodies to collagenase suppressed capillary-like structure formation. TIMP-1 and TIMP-2 are known to inhibit angiogenesis [20]. Treatment with PMA produced significant amounts of MMP-9 collagenase in these cocultures; however, the addition of small amounts of recombinant MMP-9 enhanced capillary-tubule formation and the addition of greater amounts suppressed capillary formation. The inhibitory effect of the higher concentrations presumably reflects greater degradation of the substratum that interferes with support of the tube network. Batimastat (BB-94), a synthetic hydroxamate peptidomimetic MMP inhibitor, is being tested in experimental and clinical trials for various conditions known to involve MMP disorders [21, 22]. In this study, Batimastat suppressed the formation of capillary network and MMP-9

MMP-9 in cerebral angiogenesis

341

Figure 3. Human brain microvascular endothelial–glioblastoma cell cocultures were treated with the indicated concentrations of the synthetic MMP inhibitor Batimastat (BB-94), TIMP-1, PMA was added, and (A) serum-free medium was extracted and analyzed by gelatin zymography and (B) Western blotting for TIMP-1 as described in ‘Materials and methods’.

Figure 4. Gelatin zymograms of serum-free conditioned media from brain microvascular endothelial–glioblastoma cell cocultures after treatment with the indicated compounds. Cocultures were plated on 24-well tissue culture plates coated with fibronectin; cytochalasin-D, colchicine-B, calphostin-C, and PD98059 were added before PMA treatment. (A) Serum-free medium was collected from these cultures 24 h later, and gelatin zymography was performed to assess enzyme activities. (B) The gels were also probed with TIMP-1 antibody by Western blotting as described in ‘Materials and methods’.

activity in a dose-dependent fashion. Moreover, our findings that exogenous TIMP-1 also inhibited vascular endothelial cell capillary formation agree with previous findings that in vitro angiogenesis was inhibited by the addition of rTIMP-1 [23]. Incubation of the cocultures with BB-94, resulted in dose-dependent inhibition of MMP-9 activity. Others have also demonstrated that BB-94 completely inhibited MMP-2 and MMP-9 activity in IL-2 activated mouse NK cells [24] and pancreatic cancer cells [25]. Disruption of the actin cytoskeleton by cytochalasin-D previously was shown to decrease PMA-induced MMP9 activity in SNB19 cells [26]. In the current study, we tested whether changes in cytoskeleton polymerization modulate the expression of MMPs by using cytochalasin-D and colchicine-B, agents that change cell shape in vitro. Gelatin zymography analysis of conditioned medium from the cocultures revealed PMA-induced MMP-9, which normally is not present in individual cultures of SNB19 or

brain microvascular endothelial cells. Moreover, treating the cocultured cells with cytochalasin-D suppressed capillary formation and MMP-9 activity, but colchicine-B had no effect. The molecular target of phorbol esters such as PMA is PKC, an enzyme that plays a crucial role in the transduction of extracellular signals in endothelial cells [27]. PKC is also known to play a role in MMP-9 expression in malignant glioma cells [26] and human HL-60 myeloid leukemia cells [28]. Because PKC is central to PMA-induced MMP-9 activity, we were interested in defining the role of PKC in the stimulation of MMP-9 activity as well as that in capillary network formation. The addition of calphostin-C, a PKC inhibitor, to the cocultures significantly reduced the levels of both MMP-9 and TIMP-1. This result suggests that PKC is involved in the induction of MMP-9 during capillary formation in this system. Notably, inhibition of the MMP9 activity paralleled the inhibition of capillary formation.

342

N. Chandrasekar et al.

Others have shown that inhibition of ERK1/2 pathways with PD98059, a specific inhibitor of MEK1, abrogated bFGFmediated tube formation by murine endothelial cells [29]. We also observed inhibition of capillary formation by human brain endothelial cells when those cells were cocultured with SNB19 and stimulated with PMA in the presence of this inhibitor PD98059. In summary these results clearly demonstrated that the significant role of MMP-9 during capillary formation and the specific inhibitor for MMP-9 may be a potential therapeutic target for anti angiogenesis and anti cancer treatment.

13.

Acknowledgements

18.

This work was supported by National Cancer Institute grant CA-85216 to JSR. We thank Lydia Soto for preparing the manuscript and Christine Wogan for editorial review.

19.

14.

15.

16.

17.

20.

References 1. Haas TL, Madri JA. Extracellular matrix-driven matrix metalloproteinase production in endothelial cells: Implications for angiogenesis. Trends Cardiovasc Med 1999; 9: 70–7. 2. Itoh T, Tanioka M, Yoshida H et al. Reduced angiogenesis and tumor progression in gelatinase A-deficient mice. Cancer Res 1998; 58: 1048–51. 3. Vu TH, Shipley JM, Bergers G et al. MMP-9/gelatinase B is a key regulator of growth plate angiogenesis and apoptosis of hypertrophic chondrocytes. Cell 1998; 93: 411–22. 4. Kim J, Yu W, Kovalski K et al. Requirement for specific proteases in cancer cell intravasation as revealed by a novel semiquantitative PCR-based assay. Cell 1998; 94: 353–62. 5. Nagase H, Woessner JF Jr. Matrix metalloproteinases. J Biol Chem 1999; 274: 21491–4. 6. Sang QX. Complex role of matrix metalloproteinases in angiogenesis. Cell Res 1998; 8: 171–7. 7. Wylie S, MacDonald IC, Varghese HJ et al. The matrix metalloproteinase inhibitor batimastat inhibits angiogenesis in liver metastases of B16F1 melanoma cells. Clin Exp Metastasis 1999; 17: 111–7. 8. Anand-Apte B, Pepper MS, Voest E et al. Inhibition of angiogenesis by tissue inhibitor of metalloproteinase-3. Invest Ophthalmol Vis Sci 1997; 38: 817–23. 9. Montesano R, Orci L. Tumor-promoting phorbol esters induce angiogenesis in vitro. Cell 1985; 42: 469–77. 10. Kubota Y, Kleinman HK, Martin GR et al. Role of laminin and basement membrane in the morphological differentiation of human endothelial cells into capillary-like structures. J Cell Biol 1988; 107: 1589–98. 11. Schnaper HW, Grant DS, Stetler-Stevenson WG et al. Type IV collagenase(s) and TIMPs modulate endothelial cell morphogenesis in vitro. J Cell Physiol 1993; 156: 235–46. 12. Bouck N, Stellmach V and Hsu SC. How tumors become angiogenic. Adv Cancer Res 1996; 69: 135–74.

21. 22.

23.

24.

25.

26.

27.

28.

29.

Pepper MS, Ferrara N, Orci L et al. Potent synergism between vascular endothelial growth factor and basic fibroblast growth factor in the induction of angiogenesis in vitro. Biochem Biophys Res Commun 1992; 189: 824–31. Bussolino F, Di Renzo MF, Ziche M et al. Hepatocyte growth factor is a potent angiogenic factor which stimulates endothelial cell motility and growth. J Cell Biol 1992; 119: 629–41. Laterra J, Indurti RR and Goldstein GW. Regulation of in vitro gliainduced microvessel morphogenesis by urokinase. J Cell Physiol 1994; 158: 317–24. Laterra J and Goldstein GW. Astroglial-induced in vitro angiogenesis: requirements for RNA and protein synthesis. J Neurochem 1991; 57: 1231–9. Rao JS, Sawaya R, Gokalsan ZL et al. Modulation of serine proteinases and metalloproteinases during morphogenic glial-endothelial interactions. J Neurochem 1996; 66: 1657–64. De Lorenzo MS, Alonso DF and Gomez DE. Nafoxidine modulates the expression of matrix-metalloproteinase-2 (MMP-2) and tissue inhibitor of metalloproteinases-1 (TIMP-1) in endothelial cells. Anticancer Res 2000; 20: 395–400. Fisher C, Gilberston-Beadling S, Powers EA et al. Interstitial collagenase is required for angiogenesis in vitro. Dev Biol 1994; 162: 499–510. Kostoulas G, Lang A, Nagase H et al. Stimulation of angiogenesis through cathepsin B inactivaiton of the tissue inhibitors of matrix metalloproteinases. FEBS Lett 1999; 455: 286–90. Bottomley KM, Johnson WH and Walter DS. Matrix metalloproteinase inhibitors in arthritis. J Enzyme Inhib 1998; 13: 79–101. Eccles SA, Box GM, Court WJ et al. Control of lymphatic and hematogenous metastasis of a rat mammary carcinoma by the matrix metalloproteinase inhibitor batimastat (BB-94). Cancer Res 1996; 56: 2815–22. Thorgeirsson UP, Yoshiji H, Sinha CC et al. Breast cancer: Tumor neovascularization and the effect of tissue inhibitor of metalloproteinases-1 (TIMP-1) on angiogenesis. In Vivo 1996; 10: 137–44. Kim HM, Kitson RP, Albertsson P et al. Secreted and membraneassociated matrix metalloproteinases of IL-2 activated NK cells and their inhibitors. J Immunol 2000; 164: 5883–9. Jimenez RE, Hartwig W, Antoniu BA et al. Effect of matrix metalloproteinase inhibition on pancreatic cancer invasion and metastasis: An additive strategy for cancer control. Ann Surg 2000; 231: 644–54. Chintala SK, Sawaya R, Aggarwal BB et al. Induction of matrix metalloproteinases-9 requires a polymerized actin cytoskeleton in human malignant glioma cells. J Biol Chem 1998; 273: 13545–51. Yabkowitz R, Meyers S, Black T et al. Inflammatory cytokines and vascular endothelial growth factor stimulate the release of soluble tie receptor from human endothelial cells via metalloprotease activation. Blood 1999; 93: 1969–79. Xie B, Laouar A and Huberman E. Fibronectin-mediated cell adhesion is required for induction of 92-kDa type IV collagenase/gelatinase (MMP-9) gene expression during macrophage differentiation. The signaling role of protein kinase C-beta. J Biol Chem 1998; 273: 11576–82. Tanaka K, Abe M and Sato Y. Roles of extracellular signal-regulated kinases 1/2 and p38 mitogen-activated protein kinase in the signal transduction of basic fibroblast growth factor in endothelial cells during angiogenesis. Jpn J Cancer Res 1999; 90: 647–54.