The antiangiogenic function of the tissue inhibitors of metalloproteinases (TIMPs) has been attributed to their matrix metalloproteinase inhibitory activity. Here we demonstrate that TIMP-1 but not Ala+TIMP-1 inhibits both basal and vascular endothelial growth factor (VEGF)-stimulated migration of human microvascular endothelial cells (hMVECs), suggesting that this effect is dependent on direct inhibition of matrix metalloproteinase (MMP) activity. In contrast, TIMP-2 and mutant Ala+TIMP-2, which is devoid of MMP inhibitory activity, block hMVEC migration in response to VEGF-A stimulation. TIMP-2 and Ala+TIMP-2 also suppress basal hMVEC migration via a time-dependent mechanism mediated by enhanced expression of RECK, a membrane-anchored MMP inhibitor, which, in turn, inhibits cell migration. TIMP-2 treatment of hMVECs increases the association of Crk with C3G, resulting in enhanced Rap1 activation. hMVECs stably expressing Rap1 have increased RECK expression and display reduced cell migration compared with those expressing inactive Rap1(38N). RECK-null murine embryo fibroblasts fail to demonstrate TIMP-2–mediated decrease in cell migration despite activation of Rap1. TIMP-2–induced RECK decreases cell-associated MMP activity. Anti-RECK antibody increases MMP activity and reverses the TIMP-2–mediated reduction in cell migration. The effects of TIMP-2 on RECK expression and cell migration were confirmed in A2058 melanoma cells. These results suggest that TIMP-2 can inhibit cell migration via several distinct mechanisms. First, TIMP-2 can inhibit cell migration after VEGF stimulation by direct inhibition of MMP activity induced in response to VEGF stimulation. Secondly, TIMP-2 can disrupt VEGF signaling required for initiation of hMVEC migration. Third, TIMP-2 can enhance expression of RECK via Rap1 signaling resulting in an indirect, time-dependent inhibition of endothelial cell migration.

Angiogenesis is an invasive process that requires proteolysis and remodeling of the extracellular matrix, proliferation and migration of endothelial cells, as well as the synthesis of new matrix components (1, 2, 3). The regulated proteolysis of extracellular matrix required for endothelial cell proliferation, migration and invasion is principally mediated by the activity of members of the matrix metalloproteinase (MMPs) family. MMPs are a family of zinc-dependent endopeptidases that can collectively degrade all components of extracellular matrix, and their requirement during angiogenesis has been demonstrated in MMP-deficient mice (4, 5, 6). Several distinct classes of proteins have been found to inhibit MMP activity and thereby modulate the angiogenic process. These include the tissue inhibitors of metalloproteinases (TIMPs) as well as a new class of membrane protease inhibitors represented by the reversion-inducing–cysteine-rich protein with Kazal motifs or RECK (7, 8).

The TIMPs are well-studied inhibitors of MMPs and consist of a family of four structurally related proteins (TIMP-1–4), with core proteins of ∼21 kDa (9, 10). TIMPs inhibit MMP activity by a common mechanism involving interaction of the amino-terminal cysteine residue with the zinc atom at the MMP active site (11). The TIMPs inhibit MMP activity associated with tumor invasion and angiogenesis, and these observations led to the development of synthetic MMP inhibitors for potential therapeutic application in cancer (12, 13). In addition to their MMP-inhibitory activity, it is now widely appreciated that TIMPs have direct effects on cellular behaviors such as cell growth, apoptosis, migration, and differentiation (10, 12). TIMP-1, -2, and -3 have all been shown to inhibit angiogenesis. The ability of TIMP-4 to inhibit angiogenesis in vivo has not been demonstrated. TIMPs 1–3 inhibit angiogenesis via mechanisms that may involve both inhibition of MMP activity and MMP-independent actions.

TIMP-1 inhibits endothelial cell migration but not proliferation via a mechanism that involves direct inhibition of MMP activity and can be mimicked by synthetic MMP inhibitors (14, 15). In support of these findings, a recent report demonstrates that TIMP-1–blocking antibody enhances endothelial cell migration in vitro and angiogenesis in vivo(16). TIMP-3 inhibits capillary morphogenesis in vivo and endothelial cell migration in vitro, both effects reportedly mediated by inhibition of MMP activity. However, TIMP-3 has been shown recently to function as an antagonist for the vascular endothelial growth factor receptor (VEGFR)-2, inhibiting vascular endothelial growth factor (VEGF)-A binding to this receptor (17, 18). It has also been reported that TIMP-2 inhibits cell migration after TIMP-2 transfection of human microvascular endothelial cells (hMVECs; ref. 19). TIMP-2 has been reported to inhibit cell migration through inhibition of MMP activity, specifically MT1-MMP inhibition (19, 20). However, many of these reports involve forced expression of TIMP-2 and, as a result, have not examined the kinetics of this effect. Interestingly, recent reports also suggest that the effects of TIMP-2 on cell migration may be more complex. TIMP-2 has also been reported to stimulate cell migration and accelerate wound closure (21). Several reports have demonstrated that TIMP-2 can directly inhibit the proliferation of endothelial cells in response to angiogenic stimuli such as fibroblast growth factor 2 or VEGF-A (22, 23); this effect is independent of MMP inhibitory activity and is not observed with other members of the TIMP family or synthetic MMP inhibitors. We have demonstrated recently that the growth-inhibitory activity of TIMP-2 for hMVECs is mediated through binding to α3β1 integrin and induction of protein tyrosine phosphatase activity (23). This orthovanadate-sensitive mechanism appears to be the principal mechanism of TIMP-2 inhibition of angiogenesis in vivo and suggests that TIMP-2 is a unique member of the TIMP family, not only through its ability to mediate the cellular activation of pro-MMP-2 by MT1-MMP but also in having a cell surface receptor that mediates direct growth inhibition (23).

Takahashi et al. described recently a novel membrane-anchored MMP inhibitor, RECK (7). When HT1080 cells with forced-expression RECK were injected s.c. into nude mice, tumors arose but with reduced angiogenic responses, suggesting that RECK controls angiogenesis (8). RECK serves as a crucial regulator of MMP activity at the cell surface, as demonstrated by examination of RECK homozygous-null mouse embryos (8). In contrast to the embryonic lethal phenotype of null RECK animals, the ablation of the TIMP-1–3 genes reportedly has little effect on embryonic development, but the effects of these gene knockout experiments on chronic disease processes in adult animals have not been investigated (10). That the effects of TIMP gene ablation may be subtle but nevertheless important is suggested by recent studies of TIMP-3 knockout animals, which show decreased alveolar branching morphogenesis, increased pulmonary air space development, and enhanced apoptosis of mammary epithelial cells in response to lactational stimuli (24, 25). The effects of TIMP-1, -2, and -4 gene deletion in adult animals or in animal models of chronic diseases, such as cancer, arthritis, or atherosclerosis, have not been examined.

Previous reports from our laboratory and others have demonstrated a differential effect of TIMP-1 and TIMP-2 on endothelial cell proliferation that led to identification of an MMP-independent, α3β1-mediated mechanism for these effects (22, 23). In the present study, we examine the effects of TIMP-1 and TIMP-2 on basal and stimulated hMVEC migration. In these studies we have focused on the effects of TIMP-1 and -2 on quiescent hMVECs, with or without subsequent VEGF-stimulation. We again find differential effects of these two metalloproteinase inhibitors, and our results provide evidence of a novel mechanism involving TIMP-2 stimulation of RECK expression, which, in turn, functions to reduce endothelial cell migration.

Cell Culture and Antibodies

hMVECs derived from lung (Clonetics, San Diego, CA) were grown in EGM-2 MV BulletKit medium containing 5% fetal bovine serum and growth factors (Clonetics). A2058 human melanoma cells (American Type Culture Collection, Rockville, MD) were cultured as described previously (26). Retroviral transduction of TIMP-2 expression in A2058 human melanoma cells and characterization of effects on cell adhesion are as described previously (26). TIMP-2, TIMP-1, and their mutant forms, Ala+TIMP-2 and Ala+TIMP-1, were expressed in Escheria coli, purified, and characterized as described previously (27). Signal transduction inhibitors were purchased from commercial sources. Antibodies against Crk and RECK were purchased from BD Transduction Laboratories (San Diego, CA); antibodies against C3G, Rap1, and actin were obtained from Santa Cruz Biotechnology (Santa Cruz, CA).

Cell Growth and Migration Assays

Growth Assays.

hMVECs were grown in basal medium before treatment with TIMP-2, Ala+TIMP-2, fibroblast growth factor 2, or VEGF-A. After treatment, cell numbers were quantified using Cell Titer 96 Aqueous One Solution reagent (Promega, Madison, WI) as described previously (23). The results from triplicate determinations (mean ± SD) are presented relative to unstimulated hMVECs.

Boyden Chamber Assays.

Migration assays were performed as described (28) using hMVEC cells (1 × 105 cells/well), Nucleopore polycarbonate PVP-free filters (10 μm pore size, 25 × 80 mm, Neuroprobe, Cabin John, MD) coated in a solution of fibronectin (10 μg/mL in PBS) overnight. After assembly and introduction of cell to the upper compartment, the chambers were then incubated for 4 hours at 37°C. At the end of the incubation period, cells on the underside of the filters were quantified as described previously (28).

Monolayer Wounding Assay.

Cells were plated, grown to confluence, and a wound was introduced by scratching the confluent monolayer with a pipette tip (29). Tissue culture medium was replaced to remove detached cells, and migration was observed with still images taken at the indicated times.

Detection of GTP-Bound Rap1

Detection of GTP-bound Rap1 was performed as described previously (30), with slight modification. Cells were washed twice with ice-cold PBS. Lysates were clarified by centrifugation and supernatants were incubated with GST-RalGDS-RBD for 1 hour at 4°C. Beads were washed with lysis buffer and resuspended in SDS sample buffer. Proteins bound to the beads were separated by SDS-PAGE and analyzed by Western blotting with anti-Rap1 antibody.

Assay of Cell-Associated MMP Activity

hMVECs, 2.5 × 105, were plated onto collagen I-coated 24-well plates and incubated with or without TIMP-2 (100 nmol/L) for 24 hours. When indicated, anti-RECK antibody (12 μg/mL) was added to the medium and incubated for 30 minutes. The cells were then washed with PBS twice and assayed for cell-associated MMP activity in the mixture containing 50 mmol/L 4-morpholinepropanesulfonic acid (pH 7.0), 1 mmol/L 5,5′Dithiobis(2-nitrobenzoic acid) (DTNB), 1 mmol/L CaCl2, 10 μmol/L ZnCl2, and 500 μmol/L Ac-Pro-Leu-Gly-(2-mercapto-4-methyl-pentanoyl)-Leu-Gly-Oet. After 1 hour of incubation, absorbance changes at 405 nm were measured.

Retroviral Infection of hMVECs

Retroviral expression vectors pLXSN, containing cDNA for human Rap1 or its inactive mutant, Rap1(38N), were as described previously (31). Retroviral-containing culture media were harvested, passed through sterile 0.22-μm membrane filter, and added directly onto hMVECs with 8 μg/mL Polybrene (Sigma, St. Louis, MO). After 20 to 30 minutes of incubation in a CO2 incubator, hMVEC recipient cells were refed with growth medium followed by selection in the presence of geneticin.

Immunoprecipitation and Western Blotting

After serum starvation for 24 hours and treatment with or without TIMP-2 or Ala+TIMP-2 as indicated, hMVECs were rinsed twice with ice-cold PBS and harvested by scraping in 50 mmol/L Tris-HCl (pH 7.4), 150 mmol/L NaCl, 10% glycerol, 1% NP40, 1 mmol/L EDTA, 100 μg/mL 4-(2-aminoethyl)benzene-sulfonyl fluoride, 10 μg/mL aprotinin, 1 μg/mL pepstatin A, 0.5 μg/mL leupeptin, 80 mmol/L β-glycerophosphate, 25 mmol/L NaF, and 1 mmol/L sodium orthovanadate. Equivalent amounts of protein (700 μg) were incubated with anti-Crk antibody at 4°C for >6 hours, and immune complexes were then precipitated with Protein A/G PLUS-agarose. Immunoprecipitates were washed with the lysis solution, separated using SDS-PAGE, and analyzed by Western blotting as described previously (8).

Differential Effects of TIMP-1 and TIMP-2 on hMVEC Migration.

We examined the differential effects of TIMP-1 and TIMP-2 on migration of hMVECs using the monolayer wounding assay (29). VEGF-A stimulation of hMVECs results in an ∼60% increase in endothelial cell migration over that observed in untreated cells (Fig. 1 A). The addition of recombinant human TIMP-1 or TIMP-2 before stimulation with VEGF-A results in suppression of hMVEC migration to levels observed in nonstimulated hMVECs. However, the effects of NH2 terminally substituted TIMPs, which lack MMP inhibitory activity (27), reveal that Ala+TIMP-2 but not Ala+TIMP-1 retains suppressive activity for VEGF-stimulated endothelial cell migration. These differential effects of Ala+TIMP-1 and Ala+TIMP-2 on hMVEC migration suggest that TIMP-1 inhibition of hMVEC migration in response to VEGF-A is mediated directly by MMP inhibition. In contrast, the observation that both TIMP-2 and Ala+TIMP-2 are equipotent in inhibiting VEGF-A–stimulated hMVEC migration suggests that the mechanism of this effect is, at least in part, independent of MMP inhibition. The observation that both TIMP-2 and Ala+TIMP-2 inhibit VEGF-A–simulated hMVEC migration is consistent with our previous report demonstrating that both TIMP-2 and Ala+TIMP-2 but not TIMP-1 can block VEGF-A signaling by dephosphorylation/inactivation of VEGFR-2 (23).

We also examined the effects of TIMP-1 and TIMP-2 on the spontaneous or basal migration (non-VEGF-A–stimulated migration) of hMVECs. As shown in Fig. 1,A, hMVECs migrate in the absence of VEGF-A, and this spontaneous migration constitutes ∼60% of the maximal VEGF-A–stimulated migration observed in these assays. Treatment with TIMP-1 reduced the basal migration of hMVECs in this assay by ∼60% (Fig. 1,B). This TIMP-1 inhibition is also dependent on MMP inhibitory activity, because Ala+TIMP-1 failed to inhibit migration (either VEGF-A–stimulated or spontaneous; Fig. 1,B). In contrast, TIMP-2 only slightly inhibited hMVEC migration in the absence of VEGF-A stimulation. These differential effects of TIMP-1 and TIMP-2 on non-stimulated hMVEC migration were confirmed in Boyden chamber assays (Fig. 1 C).

To additionally examine the effect of TIMP-2 on endothelial cell basal migration we preincubated hMVECs with TIMP-2 for 12 to 24 hours before initiating the migration assays. In contrast to the marginal effect on hMVEC migration after the immediate addition of TIMP-2, preincubation with TIMP-2 for 12 to 24 hours results in a significant (∼50%) inhibition of basal hMVEC migration (Fig. 2,A). A similar effect was observed using Ala+TIMP-2, suggesting that this effect is also independent of TIMP-2 inhibition of MMP activity. In contrast, similar experiments using TIMP-1 demonstrate that the ability of this inhibitor to block spontaneous migration of hMVECs is the same after 12 or 24 hours preincubation as that observed immediately after TIMP-1 addition at zero time (Fig. 2 B). These findings clearly suggest that TIMP-1 and TIMP-2 inhibit spontaneous hMVEC migration via distinct mechanisms.

To investigate the mechanism of this delayed TIMP-2 inhibition of hMVEC migration we examined the effect of TIMP-2 and Ala+TIMP-2 on cell-associated MMP activity critical for cell migration. hMVECs were untreated or incubated with TIMP-2 for 24 hours and then washed with PBS before assaying for MMP activity. The results show that both TIMP-2 and Ala+TIMP-2 treatments result in a decrease in cell-associated MMP activity (Fig. 2,C). These findings demonstrate that the apparent increase in cell-associated MMP inhibitory activity after TIMP-2 treatment is not simply the result of enhanced TIMP-2 cell surface association. The findings that both TIMP-2 and Ala+TIMP-2 can reduce cell-associated MMP activity suggest that these effects are not mediated by TIMP-2 inhibition of MMP activity. Furthermore, TIMP-2 (or Ala+TIMP-2) did not suppress the basal growth of hMVECs (non-VEGF-A–stimulated; Fig. 2 D) demonstrating that the delayed inhibitory effect of TIMP-2 on basal hMVEC migration is not a result of cytotoxicity as this growth assay used in dependent on cell viability. These findings suggest that although TIMP-2 suppresses the mitogenic response of hMVECs to angiogenic stimuli (21), it does not alter the growth or viability of quiescent endothelial cell cultures. This finding shows that the ability of TIMP-2 to suppress mitogenic signaling in response to growth factors such as VEGF-A or fibroblast growth factor-2 (21) does not account for TIMP-2 inhibition of hMVEC migration under nonstimulated conditions (spontaneous migration).

TIMP-2 Induces RECK Expression.

To determine the effect of TIMP-2 on cell-associated MMP activity we examined the expression of RECK, a novel cell membrane-associated protease inhibitor. hMVECs were serum starved for 24 hours; treated with recombinant human TIMP-2, Ala+TIMP-2, or TIMP-1; and analyzed for RECK expression. Northern and Western blot analyses showed that hMVECs have basal levels of RECK expression that are increased >2.8–3-fold by TIMP-2 or Ala+TIMP-2 treatment (Fig. 3,A and B). In contrast, TIMP-1 treatment did not significantly alter RECK expression (Fig. 3,A). Time course analysis reveals that the increase in the levels of RECK mRNA and protein reaches their maxima at 4 and 24 hours, respectively, after TIMP-2 treatment (Fig. 3 B). The results obtained with Ala+TIMP-2 demonstrate that the MMP-inhibitory activity of TIMP-2 is not required for induction of RECK.

To investigate the molecular mechanism underlying TIMP-2 induction of RECK expression we used two experimental approaches. First, we sought to determine whether the mechanism of TIMP-2 or Ala+TIMP-2–mediated increase in RECK expression involved the previously described TIMP-2 receptor α3β1. To this end we examined the effect of TIMP-2 on RECK expression in β1-null murine embryo fibroblasts (MEFs). Although β1-null MEFs (GD25 cells) have higher basal levels of RECK expression, implying that endogenous β1 integrins may suppress basal RECK expression, it is apparent that incubation with TIMP-2 (100 nmol/L) does not up-regulate RECK expression in these cells (Fig. 3,C). Furthermore, reconstituted β1 expression in these cells (GD25-β1A) results in down-regulation of basal RECK levels and restores TIMP-2 induction of enhanced RECK expression (Fig. 3 C). These findings are consistent with our previous report demonstrating that α3β1 functions as a cell surface receptor for TIMP-2 and is required for multiple cellular effects of TIMP-2, including suppression of cell growth and, as shown in the current report, enhanced expression of RECK.

In a second approach we pretreated hMVECs with inhibitors of various signaling pathways before TIMP-2 stimulation and analyzed their effects by northern blotting. Pretreatment of hMVECs with actinomycin D (5 μg/mL), a RNA synthesis inhibitor, abrogated the increase in mRNA levels of RECK, suggesting that TIMP-2 stimulates RECK expression via transcriptional activation (Fig. 3,D). Inhibitors of MEK (PD98059, 50 μmol/L), phosphatidylinositol 3′-kinase(LY294002, 10 μmol/L), adenylate cyclase (SQ22536, 100 μmol/L), or protein kinase C (calphostin C, 0.5 μmol/L) showed no significant effects on TIMP-2 enhancement of RECK expression (Fig. 3 D). However, orthovanadate (1 μmol/L), a protein tyrosine phosphatase inhibitor, considerably impairs TIMP-2–mediated increase in RECK mRNA expression. It is interesting to note that TIMP-2–mediated inhibition of hMVEC growth in vitro, suppression of angiogenesis in vivo(23), and the increase in RECK expression observed in the present report are all independent of MMP-inhibitory activity and sensitive to orthovanadate. This suggest that, like TIMP-2–mediated suppression of hMVEC proliferation and angiogenesis, TIMP-2 induction of RECK is mediated at some level by protein tyrosine phosphatase activity.

Rap1 Mediates TIMP-2 Induction of RECK Expression.

Rap1 is known to induce inside-out signaling across integrin receptors resulting in enhanced cell attachment, which can limit cell migration (32). Interestingly, both Rap1 and RECK genes were isolated by the same expression cloning strategy designed to identify genes that suppress cell transformation by K-ras (7, 33). Like other Ras family small GTPases, activation of Rap1 can be regulated by specific guanine nucleotide exchange factors and GTPase activating proteins (34). C3G, identified as a Crk SH3 domain-binding protein, has been shown to exhibit Rap1 guanine nucleotide exchange factor activity (35, 36). We examined the relationship between the TIMP-2–stimulated increased RECK expression and the Rap1 signaling pathway.

To determine whether TIMP-2 modulates this established activation pathway for Rap1, we treated hMVECs with recombinant TIMP-2, immunoprecipitated cell lysates with anti-Crk antibody, and analyzed by Western blotting. As shown in Fig. 4,A (top), Crk associates with C3G in the control cells, and TIMP-2 enhances this interaction ∼2-fold. Interestingly, orthovanadate pretreatment, before TIMP-2 stimulation, reverses TIMP-2 effects, reducing Crk-C3G interaction to basal levels. The Western blot profile of Crk pull-down experiments shows that Rap1 binding to Crk parallels that of C3G (Fig. 4,A, top), suggesting that Rap1 is activated in response to TIMP-2 treatment. Furthermore, the levels of GTP-bound Rap1 (activated Rap1) after TIMP-2 treatment exactly parallels the observed increase in Rap1 complexed with C3G and Crk. Pull-down assay using GST-RalGDS-RBD demonstrated basal levels of Rap1 in the GTP-bound form present in control (untreated) hMVECs and that TIMP-2 treatment results in a ∼40% increase activated Rap1 (GTP-Rap1; Fig. 4,A, bottom). Again, orthovanadate pretreatment ablates TIMP-2 induced Rap1 activation (Fig. 4,A, bottom), which is consistent with our previous findings (Fig. 3 D).

Our findings suggest that TIMP-2 stimulates increased RECK expression via a mechanism involving Rap1. To confirm these findings, we established hMVEC lines stably expressing human Rap1 or its inactive mutant, Rap1(38N; ref. 31) by retroviral infection. We then determine the RECK protein levels with or without TIMP-2 treatment in these infectant hMVECs. Western blot analysis shows that RECK levels are induced ∼3-fold in the TIMP-2–treated control hMVECs (Fig. 4,B). Forced expression of Rap1 induced an ∼3-fold increase in RECK expression in hMVECs, which is unaffected by TIMP-2 treatment (Fig. 4 B). In contrast, the expression of Rap1(38N) completely blocks the ability of TIMP-2 to enhance RECK expression. Our finding clearly supports the role of Rap1 as an effector protein in TIMP-2 induction of RECK expression. It is interesting to note that RECK gene expression is reportedly down-regulated by several oncogenes, including Ras (7), but is found up-regulated by Rap1 in this study. This observation may be another example of the antagonizing effects of Ras and Rap1 (34).

Rap1-Mediated Increase in RECK Inhibits hMVEC Migration.

It was reported that the deficiency in C3G, an activator of Rap1, inhibits the activation of Rap1 on cell attachment and increases cell motility (37). To evaluate how the TIMP-2–induced Rap1 affects endothelial cell migration, the parental hMVECs and Rap1 infectant cells were tested for the ability to migrate using the monolayer wounding assay. As shown in Fig. 4 C, cell migration of hMVEC-Rap1 cells is reduced by ∼50%, whereas hMVEC-Rap1(38N) cells display ∼20% enhanced ability to migrate in the monolayer wounding assay compared with the parental cells. These findings suggest that TIMP-2 activation of Rap1 leads to enhancement of RECK expression that results in a reduction in endothelial cell migration.

To confirm that the reduction in hMVEC migration is due to TIMP-2–mediated increase in RECK expression and not activation of Rap1 alone, we examined the effects of TIMP-2 in RECK+/− and RECK−/− MEFs. Interestingly, TIMP-2 increased the levels of active (GTP bound) Rap1 in both RECK+/− and RECK−/− MEFs (data not shown). However, despite this TIMP-2–mediated increase of Rap1 activation in RECK−/− MEFs, TIMP-2 fails to inhibit migration of these cells under conditions in which there is a ∼50% reduction in RECK+/− MEF migration (Fig. 4 D). These findings confirm that TIMP-2 reduction of cell migration requires enhancement of RECK expression and that activation of Rap1 itself is not sufficient to inhibit endothelial cell migration.

TIMP-2 Suppression of hMVEC Migration Requires MMP-Inhibitory Activity of RECK.

RECK has been characterized as a membrane-bound inhibitor of MMPs (8). We sought to determine whether the MMP-inhibitory activity of RECK is required for suppression of hMVEC migration. The total cell-associated MMP activities of control and TIMP-2–treated hMVECs were assayed as described in Fig. 2,C, both in the presence and absence of anti-RECK antibodies. The TIMP-2–mediated increase in RECK expression is associated with an ∼40% reduction in total MMP activity as shown in Fig. 5,A. Treatment with anti-RECK antibody, specifically directed against the Kazal motifs (7), significantly increased MMP activity in both control and TIMP-2–treated hMVECs. It should be noted that hMVEC cultures have low endogenous expression of both TIMP-2 and RECK and that blocking endogenous RECK also results in enhancement of cell-associated MMP activity. We then used this antibody to examine RECK-mediated inhibition of MMP activity in hMVEC migration. TIMP-2–transduced hMVECs show a reduction in cell migration similar to that demonstrated previously in cells treated with exogenous recombinant TIMP-2 for 12 to 24 hours (Fig. 5 B). However, inclusion of anti-RECK antibodies in this assay results in complete normalization of migration of TIMP-2–transduced hMVECs. This finding suggests that the effects of TIMP-2–mediated increase in RECK expression on hMVEC migration require the MMP-inhibitory activity of RECK.

To determine whether these effects of TIMP-2 are endothelial cell-specific we also examined the effects of TIMP-2 on the expression of RECK and migration of the human melanoma cell line A2058. Fig. 5,C demonstrates that parental A2058 cells do not produce detectable levels of RECK protein. However, after forced expression of TIMP-2 in these cells (A2058-T21 cells), there is a significant increase in RECK expression that is detectable at both the protein and mRNA levels (Fig. 5,C). In addition, there is ∼3-fold increase in detectable levels of Rap1 activation. Like hMVECs, addition of purified recombinant TIMP-2 did not immediately alter the ability of A2058 cells to migration (Fig. 5,D). However, stable A2058-T21 cells show >50% reduction in cell migration (Fig. 5 D), which is not additionally reduced by addition of exogenous TIMP-2. The reduced cell migration observed in A2058-T21 cells could also be reversed by inclusion of anti-RECK antibodies as demonstrated for hMVECs (data not shown). These observations suggest that TIMP-2–mediated induction of RECK expression and subsequent inhibition of cell migration are not endothelial cell-specific.

The results of the present study additionally highlight the differences between members of the TIMP family, in particular TIMP-1 and TIMP-2. Studies have demonstrated that via their MMP-inhibitory activity TIMPs can negatively influence cell migration. However, in the present study we have focused on the effects of elevated TIMP levels before angiogenic factor stimulation of cell migration. The ability of TIMP-1 to inhibit endothelial cell migration, VEGF-A–stimulated or nonstimulated, appears entirely dependent on the ability to inhibit MMP activity, as evidenced by the lack of effect of Ala+TIMP-1 in these assays. Furthermore, the synthetic, broad spectrum MMP inhibitor BB-94 mimics the effect of TIMP-1. In contrast, the effects of TIMP-2 on endothelial cell migration are identical to that of Ala+TIMP-2 in that both inhibit the VEGF-A–stimulated migration of hMVECs but have no immediate effect on cell migration under non-VEGF–stimulated conditions. We have demonstrated previously that TIMP-2 suppresses angiogenic factor receptor signaling and prevents VEGFR-2 activation (21). In the present study we demonstrate that TIMP-2 and Ala+TIMP-2 inhibit VEGF-stimulated migration of hMVECs. Because VEGF-stimulated migration also requires VEGF-receptor activation we propose that TIMP-2 and Ala+TIMP-2 suppression of hMVEC migration is mediated by a similar signaling mechanism (21).

We have focused on the effect of TIMP-2 and Ala+TIMP-2 on the basal migration of hMVECs. Additional analysis demonstrates that preincubation of endothelial cells with TIMP-2 for 12 to 24 hours results in significant down-regulation of endothelial cell migration. This delayed inhibitory effect on cell migration is accompanied by a decrease in cell-associated MMP activity that could be induced by both TIMP-2 and Ala+TIMP-2. Because Ala+TIMP-2 has no intrinsic MMP-inhibitory activity (27), these findings suggested to us that TIMP-2 treatment of quiescent hMVECs may induce expression of an endogenous cell surface MMP inhibitor. Additional investigation demonstrates that TIMP-2 promotes enhanced expression of the metalloproteinase regulator RECK (Fig. 3), which accounts for the decrease in cell-associated MMP activity. The TIMP-2–mediated increase in RECK expression is independent of MMP-inhibitory activity (observed with both TIMP-2 and Ala+TIMP-2 treatment) and is also sensitive to orthovanadate inhibition. These characteristics are shared with the TIMP-2 inhibition of hMVEC proliferation. Similarly, β1-null MEFs fail to enhance RECK expression in response to TIMP-2 treatment, demonstrating the requirement of β1 integrins in mediating this effect. These findings show that TIMP-2 interaction with its receptor, α3β1, in the absence of VEGF stimulation, leads to enhanced expression of the cell-surface MMP-inhibitor RECK. These findings suggest that in addition to proximal inhibition of VEGFR signaling (23), TIMP-2 interaction with α3β1 also results in down-stream signaling that results in enhanced RECK expression.

TIMP-2 promotes the association of Crk and C3G, as well as subsequent Rap1 activation, i.e., increase in Rap1-GTP. Activation of Rap1 leads to the induction of RECK gene expression. The requirement for Rap1 is demonstrated by loss of enhanced RECK expression after TIMP-2 treatment of cells with mutant Rap1–38N (Fig. 4,B). Furthermore, activation of Rap1 without a concomitant increase in RECK is not sufficient to inhibit hMVEC migration, as demonstrated using RECK-deficient MEFs (Fig. 4 D).

The presumptive pathway for the TIMP-2 signaling, TIMP-2–Crk-C3G-Rap1–RECK, is also supported by the previous reports that C3G, Rap1, and RECK all suppress the transformed phenotype of Ki-ras-NIH3T3 cells (7, 33, 35). Thus, our study supports the functional significance of TIMP-2–initiated signaling pathways that are independent of MMP-inhibitor activity. Furthermore, our findings suggest that TIMP-2 can regulate MMP activity indirectly via RECK, as well as through direct interaction with MMP active sites. Additional study is required to identify the downstream targets of Rap1, the protein tyrosine phosphatase(s) involved in the RECK induction, and also to address how the TIMP-2 signal generated via interaction with α3β1 is transduced to Crk.

It has been demonstrated that Crk, an adaptor protein containing both SH2 and SH3 domains, mediates various signaling pathways through its SH3 binding partners, such as C3G and DOCK180 (36, 38). Recent studies showed that the signaling pathway involving the interaction of Crk with DOCK180 promotes cell migration and invasion via activation of Rac (39, 40). In this study, we showed that TIMP-2 signaling reduces cell migration through Crk-C3G pathway involving activation of Rap1. We could not detect any complex formation of Crk with DOCK180 in human microvascular endothelial cells (data not shown). Thus, it is interesting to note that cells may make completely opposing migratory responses depending on the SH3 binding partner (i.e., C3G versus DOCK180).

Cell migration requires membrane extension, assembly of extracellular matrix-cell contacts at the leading cell edge, destabilization of those in the rear of the cell, and increased locomotive forces (41). MMP activity, specifically MT1-MMP, has been implicated in mediating cell migration (42, 43). However, details of this mechanism remain unclear. In this study, we show that endothelial cell migration is MMP-dependent in that neutralizing anti-RECK antibody enhances cell-associated MMP activity in hMVEC-T2 cells and promotes cell migration. Collectively these results support the conclusion that RECK inhibits cell migration through suppression of MMP activity. However, clarification of this mechanism awaits identification of the MMP-inhibitory domain of RECK and preparation of RECK mutants devoid of MMP-inhibitor activity.

The differential effects of TIMP-1 and TIMP-2 on basal endothelial cell migration present an interesting but not unique phenomenon. As reported by several laboratories TIMP-1, TIMP-2, and TIMP-3 are known to differentially alter cell migration and proliferation in a variety of cell types. TIMP-2 selectively inhibits hMVEC proliferation, an effect that is not observed with TIMP-1 (23). TIMP-3 but not TIMP-1 or TIMP-2 antagonizes VEGF-A binding to its receptor, VEGFR-2, and reportedly mediates induction of cell death (10, 18). TIMP-1 but not TIMP-2 or TIMP-3 reportedly protects against induction of programmed cell death in Burkitt’s lymphoma and human breast cancer cells (10, 44, 45). These findings suggest that members of the TIMP family each have unique biological functions and/or that members of the TIMP family have unique protease inhibitory profiles.

The selective ability of TIMP-2 to induce RECK expression represents a novel biological activity of TIMP-2 compared with TIMP-1. Our data suggest that the TIMP-2 inhibition of hMVEC migration is more complex than reported previously and involves three possible mechanisms. First, it is well-appreciated that TIMP-2 inhibits endothelial cell migration and invasion by direct inhibition of MMP activities involved in this process (8, 10, 12, 13, 14, 15, 16, 17). VEGF stimulation of hMVEC results in significant induction of a variety of protease activities, including several members of the MMP family that are sensitive to inhibition by TIMP-2, as well as other members of the TIMP family. However, if quiescent endothelial cells are treated with TIMP-2 before VEGF stimulation, as in the present report, the effects of TIMP-2 are mediated through binding to α3β1 resulting in orthovanadate-sensitive, proximal inhibition of VEGF-receptor activation. This second mechanism for TIMP-2 inhibition of hMVEC migration is identical to that reported previously for TIMP-2 inhibition of VEGF-stimulated hMVEC proliferation (21). The present report describes a third mechanism for TIMP-2 inhibition of hMVEC migration that is independent of VEGF-stimulation or direct inhibition of MMP activity by TIMP-2 but involves an indirect MMP-inhibitor effect that requires transcription, synthesis, and cell surface localization of the RECK gene product. These antiangiogenic effects, as well as the high level of TIMP-2 expression observed in quiescent endothelial cells of mature capillaries, are also consistent with the proposal that TIMP-2 functions to maintain endothelial cells in a quiescent state and maintain vascular homeostasis. Delineation of TIMP-2 signaling pathways provides important insights into the mechanism underlying its antiangiogenic functions and may facilitate the design of effective approaches for disruption of tumor angiogenesis.

Fig. 1.

Differential effects of TIMP-1 and TIMP-2 on hMVEC migration. A and B, hMVECs were grown to confluence and a wound introduced in the monolayer using a pipette tip. After incubation with the indicated factors, relative migration distance of treated cell into the monolayer defect was measured using untreated hMVECs as control. Treatments are indicated as follows: VEGF-A (V, 10 ng/mL), TIMP-1 (T-1, 25 nmol/L), TIMP-2 (T2, 100 nmol/L), Ala+TIMP-1 (AT1, 25 nmol/L), and Ala+TIMP-2 (AT2, 100 nmol/L). The data are presented as the mean of triplicate determinations, and the results are representative of two independent experiments; bars,±SD. ∗∗ P < 0.01 compared with untreated controls. C. Cell migration was measure in modified Boyden chamber assays using 10-μm pore size membranes incubated at 37°C for 4 hours. hMVECs were treated with indicated concentrations of TIMP-1 (white bars) or TIMP-2 (black bars).

Fig. 1.

Differential effects of TIMP-1 and TIMP-2 on hMVEC migration. A and B, hMVECs were grown to confluence and a wound introduced in the monolayer using a pipette tip. After incubation with the indicated factors, relative migration distance of treated cell into the monolayer defect was measured using untreated hMVECs as control. Treatments are indicated as follows: VEGF-A (V, 10 ng/mL), TIMP-1 (T-1, 25 nmol/L), TIMP-2 (T2, 100 nmol/L), Ala+TIMP-1 (AT1, 25 nmol/L), and Ala+TIMP-2 (AT2, 100 nmol/L). The data are presented as the mean of triplicate determinations, and the results are representative of two independent experiments; bars,±SD. ∗∗ P < 0.01 compared with untreated controls. C. Cell migration was measure in modified Boyden chamber assays using 10-μm pore size membranes incubated at 37°C for 4 hours. hMVECs were treated with indicated concentrations of TIMP-1 (white bars) or TIMP-2 (black bars).

Close modal
Fig. 2.

TIMP-2–mediated reduction in hMVEC migration. A and B. hMVECs were preincubated with 100 nmol/L TIMP-2 (A) or 25 nmol/L TIMP-1 (B) for indicated times, and the monolayer wounding assay was performed. Relative migration was determined as above using untreated hMVECs (□) as control. C. hMVECs were untreated or incubated with 100 nmol/L TIMP-2 or Ala+TIMP-2 for 24 hours, and total cell-associated MMP activity was measured as described previously (28). D. hMVECs were serum starved to allow synchronization and incubated with indicated factors for 24 hours. Proliferation of viable cells was determined as describedpreviously (20). The data are presented as the average of six determinations; bars,±SD. ∗∗P < 0.01 compared with untreated cells.

Fig. 2.

TIMP-2–mediated reduction in hMVEC migration. A and B. hMVECs were preincubated with 100 nmol/L TIMP-2 (A) or 25 nmol/L TIMP-1 (B) for indicated times, and the monolayer wounding assay was performed. Relative migration was determined as above using untreated hMVECs (□) as control. C. hMVECs were untreated or incubated with 100 nmol/L TIMP-2 or Ala+TIMP-2 for 24 hours, and total cell-associated MMP activity was measured as described previously (28). D. hMVECs were serum starved to allow synchronization and incubated with indicated factors for 24 hours. Proliferation of viable cells was determined as describedpreviously (20). The data are presented as the average of six determinations; bars,±SD. ∗∗P < 0.01 compared with untreated cells.

Close modal
Fig. 3.

TIMP-2–mediated induction of RECK expression. A. hMVECs were cultured in serum-free medium for 24 hours and treated with 25 nmol/L TIMP-1, 100 nmol/L TIMP-2, or Ala+TIMP-2 for 4 hours. Cells were then harvested and their mRNAs analyzed by Northern blotting. B. hMVECs were treated with 100 nmol/L TIMP-2 for the time indicated, and their mRNA and protein samples from whole cell lysates were analyzed by Northern (left) and Western blotting (right), respectively. C. Both GD25 β1-deficient murine embryonic fibroblasts and GD25-β1A cells with reconstituted β1 expression were treated with TIMP-2 for 24 hours and analyzed by Western blotting. D. hMVECs were pretreated with indicated inhibitors for 15 minutes before the addition of 100 nmol/L Ala+TIMP-2. After 4 hours of Ala+TIMP-2 treatment, their mRNAs were analyzed by Northern blotting. Similar results were obtained with TIMP-2. These figures are representatives of at least two independent experiments. Relative band intensities were measured by NIH image software and are noted at the bottom of the blot. G3PDH and actin were used as loading controls.

Fig. 3.

TIMP-2–mediated induction of RECK expression. A. hMVECs were cultured in serum-free medium for 24 hours and treated with 25 nmol/L TIMP-1, 100 nmol/L TIMP-2, or Ala+TIMP-2 for 4 hours. Cells were then harvested and their mRNAs analyzed by Northern blotting. B. hMVECs were treated with 100 nmol/L TIMP-2 for the time indicated, and their mRNA and protein samples from whole cell lysates were analyzed by Northern (left) and Western blotting (right), respectively. C. Both GD25 β1-deficient murine embryonic fibroblasts and GD25-β1A cells with reconstituted β1 expression were treated with TIMP-2 for 24 hours and analyzed by Western blotting. D. hMVECs were pretreated with indicated inhibitors for 15 minutes before the addition of 100 nmol/L Ala+TIMP-2. After 4 hours of Ala+TIMP-2 treatment, their mRNAs were analyzed by Northern blotting. Similar results were obtained with TIMP-2. These figures are representatives of at least two independent experiments. Relative band intensities were measured by NIH image software and are noted at the bottom of the blot. G3PDH and actin were used as loading controls.

Close modal
Fig. 4.

TIMP-2 reduces hMVEC migration via Rap1-mediated RECK expression. A, top. hMVECs were treated with 100 nmol/L TIMP-2 for 15 minutes with or without the pretreatment of 1 μmol/L orthovanadate (V) for 5 minutes. Cells were then lysed and immunoprecipitated with anti-Crk antibodies followed by immunoblotting with antibodies indicated. Fold increase in band intensity of C3G and Rap1 associated with Crk are noted. A, bottom. After the TIMP-2 treatment described above, cell lysates were immunoblotted with anti-Rap1 antibody to detect total Rap1. To detect GTP-bound Rap1, cell lysates were precipitated with GST-RalGDS-RBD and then blotted with anti-Rap1 antibody. Fold increase in the ratio of GTP-Rap1 to total Rap1 are indicated. B. The control (mock-transduced) hMVECs and either Rap1- or Rap1(38N)-transduced hMVECs were treated with 100 nmol/L TIMP-2 for 24 hours, and cell lysates were blotted with antibodies against RECK, actin, or Rap1. Fold increase in the ratios of RECK/actin and Rap1/actin are denoted. C. The control hMVECs and infectant cells were grown to confluence, and monolayer wounding assay was performed. D. Cell migration of both TIMP-2–transduced RECK+/− and −/− MEFs was examined by monolayer wounding assay. Relative migration of TIMP-2 transduced cells (▪) was determined relative to control (mock-transduced) RECK+/− and RECK−/− MEFs (□), respectively. It should be noted however, that RECK−/− MEFs have increased migration (>60%) when compared with RECK+/− cells, but this enhanced migration is not inhibited by TIMP-2.

Fig. 4.

TIMP-2 reduces hMVEC migration via Rap1-mediated RECK expression. A, top. hMVECs were treated with 100 nmol/L TIMP-2 for 15 minutes with or without the pretreatment of 1 μmol/L orthovanadate (V) for 5 minutes. Cells were then lysed and immunoprecipitated with anti-Crk antibodies followed by immunoblotting with antibodies indicated. Fold increase in band intensity of C3G and Rap1 associated with Crk are noted. A, bottom. After the TIMP-2 treatment described above, cell lysates were immunoblotted with anti-Rap1 antibody to detect total Rap1. To detect GTP-bound Rap1, cell lysates were precipitated with GST-RalGDS-RBD and then blotted with anti-Rap1 antibody. Fold increase in the ratio of GTP-Rap1 to total Rap1 are indicated. B. The control (mock-transduced) hMVECs and either Rap1- or Rap1(38N)-transduced hMVECs were treated with 100 nmol/L TIMP-2 for 24 hours, and cell lysates were blotted with antibodies against RECK, actin, or Rap1. Fold increase in the ratios of RECK/actin and Rap1/actin are denoted. C. The control hMVECs and infectant cells were grown to confluence, and monolayer wounding assay was performed. D. Cell migration of both TIMP-2–transduced RECK+/− and −/− MEFs was examined by monolayer wounding assay. Relative migration of TIMP-2 transduced cells (▪) was determined relative to control (mock-transduced) RECK+/− and RECK−/− MEFs (□), respectively. It should be noted however, that RECK−/− MEFs have increased migration (>60%) when compared with RECK+/− cells, but this enhanced migration is not inhibited by TIMP-2.

Close modal
Fig. 5.

TIMP-2 reduces cell migration via MMP-inhibitory activity of RECK. A. hMVECs were untreated or 100 nmol/L TIMP-2–treated for 20 hours, and total cell associated MMP activity was measured in the absence (□) or presence (▪) of anti-RECK antibody (12 μg/mL) as described previously (27). B. Monolayer wounding assay was performed using TIMP-2 transduced hMVECs (hMVEC-T2) preincubated with or without anti-RECK antibody for 30 minutes. Relative migration was determined using mock-transduced hMVECs (□) as controls. C. Parental A2058 cells and TIMP-2 transduced A2058 (A2058-T21) were harvested and analyzed for RECK expression by Western (top) and Northern blotting (middle). Levels of total and GTP-bound Rap1 were also examined, and fold increase in the ratio of GTP-Rap1 to total Rap1 is denoted. D. Migration of A2058 and A2058-T21 cells were measured in modified Boyden chamber assays. Cells treated with recombinant TIMP-2 are shown as ▪. Data are presented as relative migration (percentage of cells migrating) using untreated A2058 cells as control. The data are presented as the mean of triplicate determinations, and the results are representative of two independent experiments; bars,±SD. ∗∗P < 0.01 compared with untreated cells.

Fig. 5.

TIMP-2 reduces cell migration via MMP-inhibitory activity of RECK. A. hMVECs were untreated or 100 nmol/L TIMP-2–treated for 20 hours, and total cell associated MMP activity was measured in the absence (□) or presence (▪) of anti-RECK antibody (12 μg/mL) as described previously (27). B. Monolayer wounding assay was performed using TIMP-2 transduced hMVECs (hMVEC-T2) preincubated with or without anti-RECK antibody for 30 minutes. Relative migration was determined using mock-transduced hMVECs (□) as controls. C. Parental A2058 cells and TIMP-2 transduced A2058 (A2058-T21) were harvested and analyzed for RECK expression by Western (top) and Northern blotting (middle). Levels of total and GTP-bound Rap1 were also examined, and fold increase in the ratio of GTP-Rap1 to total Rap1 is denoted. D. Migration of A2058 and A2058-T21 cells were measured in modified Boyden chamber assays. Cells treated with recombinant TIMP-2 are shown as ▪. Data are presented as relative migration (percentage of cells migrating) using untreated A2058 cells as control. The data are presented as the mean of triplicate determinations, and the results are representative of two independent experiments; bars,±SD. ∗∗P < 0.01 compared with untreated cells.

Close modal

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Note: J. M. Ray is currently at Capital Genomix, 2 Cessna Ct., Gaithersburg, MD 20879.

Requests for reprints: William G. Stetler-Stevenson, National Cancer Institute, Center for Cancer Research, Vascular Biology Faculty, Cell and Cancer Biology Branch, Bethesda, MD 20892. Phone: 301-402-1521; Fax: 301-402-7575; E-mail: sstevenw@mail.nih.gov

1
Kalluri R Basement membranes: structure, assembly and role in tumour angiogenesis.
Nat Rev Cancer
2003
;
3
:
422
-33.
2
Folkman J Fundamental concepts of the angiogenic process.
Curr Mol Med
2003
;
3
:
643
-51.
3
Hanahan D, Folkman J Patterns and emerging mechanisms of the angiogenic switch during tumorigenesis.
Cell
1996
;
86
:
353
-64.
4
Vu TH, Werb Z Matrix metalloproteinases: effectors of development and normal physiology.
Genes Dev
2000
;
14
:
2123
-33.
5
Itoh T, Tanioka M, Yoshida H, Yoshioka T, Nishimoto H, Itohara S Reduced angiogenesis and tumor progression in gelatinase A- deficient mice.
Cancer Res
1998
;
58
:
1048
-51.
6
Sternlicht MD, Werb Z How matrix metalloproteinases regulate cell behavior.
Ann Rev Cell Dev Biol
2001
;
17
:
463
-516.
7
Takahashi C, Sheng Z, Horan TP, et al Regulation of matrix metalloproteinase-9 and inhibition of tumor invasion by the membrane-anchored glycoprotein RECK.
Proc Natl Acad Sci USA
1998
;
95
:
13221
-6.
8
Oh J, Takahashi R, Kondo S, et al The membrane-anchored MMP inhibitor RECK is a key regulator of extracellular matrix integrity and angiogenesis.
Cell
2001
;
107
:
789
-800.
9
Visse R, Nagase H Matrix metalloproteinases and tissue inhibitors of metalloproteinases: structure, function, and biochemistry.
Circ Res
2003
;
92
:
827
-39.
10
Baker AH, Edwards DR, Murphy G Metalloproteinase inhibitors: biological actions and therapeutic opportunities.
J Cell Sci
2002
;
115
:
3719
-27.
11
Gomis-Rüth FX, Maskos K, Betz M, et al Mechanism of inhibition of the human matrix metalloproteinase stromelysin-1 by TIMP-1.
Nature (Lond)
1997
;
389
:
77
-81.
12
Jiang Y, Goldberg ID, Shi YE Complex roles of tissue inhibitors of metalloproteinases in cancer.
Oncogene
2002
;
21
:
2245
-52.
13
Coussens LM, Fingleton B, Matrisian LM Matrix metalloproteinase inhibitors and cancer: trials and tribulations.
Science
2002
;
295
:
2387
-92.
14
Ikenaka Y, Yoshiji H, Kuriyama S, et al Tissue inhibitor of metalloproteinases-1 (TIMP-1) inhibits tumor growth and angiogenesis in the TIMP-1 transgenic mouse model.
Int J Cancer
2003
;
105
:
340
-6.
15
Guedez L, McMarlin AJ, Kingma DW, Bennett TA, Stetler-Stevenson M, Stetler-Stevenson WG Tissue inhibitor of metalloproteinase-1 alters the tumorigenicity of Burkitt’s lymphoma via divergent effects on tumor growth and angiogenesis.
American J Pathol
2001
;
158
:
1207
-15.
16
Reed MJ, Koike T, Sadoun E, Sage EH, Puolakkainen P Inhibition of TIMP1 enhances angiogenesis in vivo and cell migration in vitro.
Microvasc Res
2003
;
65
:
9
-17.
17
Spurbeck WW, Ng CY, Vanin EF, Davidoff AM Retroviral vector-producer cell-mediated in vivo gene transfer of TIMP-3 restricts angiogenesis and neuroblastoma growth in mice.
Cancer Gene Ther
2003
;
10
:
161
-7.
18
Qi JH, Ebrahem Q, Moore N, et al A novel function for tissue inhibitor of metalloproteinases-3 (TIMP3): inhibition of angiogenesis by blockage of VEGF binding to VEGF receptor-2.
Nat Med
2003
;
9
:
407
-15.
19
Vincent L, Varet J, Pille JY, et al Efficacy of dendrimer-mediated angiostatin and TIMP-2 gene delivery on inhibition of tumor growth and angiogenesis: in vitro and in vivo studies.
Int J Cancer
2003
;
105
:
419
-29.
20
Ahn SM, Jeong SJ, Kim YS, Sohn Y, Moon A Retroviral delivery of TIMP-2 inhibits H-ras-induced migration and invasion in MCF10A human breast epithelial cells.
Cancer Lett
2004
;
207
:
49
-57.
21
Terasaki K, Kanzaki T, Aoki T, Iwata K, Saiki I Effects of recombinant human tissue inhibitor of metalloproteinases-2 (rh-TIMP-2) on migration of epidermal keratinocytes in vitro and wound healing in vivo.
J Dermatol
2003
;
30
:
165
-72.
22
Fernandez CA, Butterfield C, Jackson G, Moses MA Structural and functional uncoupling of the enzymatic and angiogenic inhibitory activities of tissue inhibitor of metalloproteinase-2 (TIMP-2): loop 6 is a novel angiogenesis inhibitor.
J Biol Chem
2003
;
278
:
40989
-95.
23
Seo DW, Li H, Guedez L, et al TIMP-2 mediated inhibition of angiogenesis: an MMP-independent mechanism.
Cell
2003
;
114
:
171
-80.
24
Fata JE, Leco KJ, Voura EB, et al Accelerated apoptosis in the Timp-3-deficient mammary gland.
J Clin Investig
2001
;
108
:
831
-41.
25
Leco KJ, Waterhouse P, Sanchez OH, et al Spontaneous air space enlargement in the lungs of mice lacking tissue inhibitor of metalloproteinases-3 (TIMP-3).
J Clin Investig
2001
;
108
:
817
-29.
26
Ray JM, Stetler-Stevenson WG Gelatinase A activity directly modulates melanoma cell adhesion and spreading.
EMBO J
1995
;
14
:
908
-17.
27
Wingfield PT, Sax JK, Stahl SJ, et al Biophysical and functional characterization of full-length, recombinant human tissue inhibitor of metalloproteinases-2 (TIMP-2) produced in Escherichia coli. Comparison of wild type and amino-terminal alanine appended variant with implications for the mechanism of TIMP functions.
J Biol Chem
1999
;
274
:
21362
-8.
28
Aznavoorian S, Stracke ML, Krutzsch H, Schiffmann E, Liotta LA Signal transduction for chemotaxis and haptotaxis by matrix molecules in tumor cells.
J Cell Biol
1990
;
110
:
1427
-38.
29
Tamura M, Gu J, Matsumoto K, Aota S, Parsons R, Yamada KM Inhibition of cell migration, spreading, and focal adhesions by tumor suppressor PTEN.
Science (Wash DC)
1998
;
280
:
1614
-7.
30
Franke B, Akkerman JW, Bos JL Rapid Ca2+-mediated activation of Rap1 in human platelets.
EMBO J
1997
;
16
:
252
-9.
31
Noda M Mechanisms of reversion.
FASEB J
1993
;
7
:
834
-40.
32
Li L, Okura M, Imamoto A Focal adhesions require catalytic activity of Src family kinases to mediate integrin-matrix adhesion.
Mol Cell Biol
2002
;
22
:
1203
-17.
33
Kitayama H, Sugimoto Y, Matsuzaki T, Ikawa Y, Noda M A ras-related gene with transformation suppressor activity.
Cell
1989
;
56
:
77
-84.
34
Stork PJ Does Rap1 deserve a bad Rap?.
Trends Biochem Sci
2003
;
28
:
267
-75.
35
Gotoh T, Hattori S, Nakamura S, et al Identification of Rap1 as a target for the Crk SH3 domain-binding guanine nucleotide-releasing factor C3G.
Mol Cell Biol
1995
;
15
:
6746
-53.
36
Feller SM Crk family adaptors-signalling complex formation and biological roles.
Oncogene
2001
;
20
:
6348
-71.
37
Ohba Y, Ikuta K, Ogura A, et al Requirement for C3G-dependent Rap1 activation for cell adhesion and embryogenesis.
EMBO J
2001
;
20
:
3333
-41.
38
Kiyokawa E, Mochizuki N, Kurata T, Matsuda M Role of Crk oncogene product in physiologic signaling.
Crit Rev Oncog
1997
;
8
:
329
-42.
39
Hsia DA, Mitra SK, Hauck CR, et al Differential regulation of cell motility and invasion by FAK.
J Cell Biol
2003
;
160
:
753
-67.
40
Klemke RL, Leng J, Molander R, Brooks PC, Vuori K, Cheresh DA CAS/Crk coupling serves as a “molecular switch” for induction of cell migration.
J Cell Biol
1998
;
140
:
961
-72.
41
Lauffenburger DA, Horowitz AF Cell Migration: A Physically Integrated Molecular Process.
Cell
1996
;
84
:
359
-69.
42
Galvez BG, Matias-Roman S, Albar JP, Sanchez-Madrid F, Arroyo AG Membrane type 1-matrix metalloproteinase is activated during migration of human endothelial cells and modulates endothelial motility and matrix remodeling.
J Biol Chem
2001
;
276
:
37491
-500.
43
Galvez BG, Matias-Roman S, Yanez-Mo M, Sanchez-Madrid F, Arroyo AG ECM regulates MT1-MMP localization with beta1 or alphavbeta3 integrins at distinct cell compartments modulating its internalization and activity on human endothelial cells.
J Cell Biol
2002
;
159
:
509
-21.
44
Guedez L, Stetler-Stevenson WG, Wolff L, et al In vitro suppression of programmed cell death of B cells by tissue inhibitor of metalloproteinases-1.
J Clin Investig
1998
;
102
:
2002
-10.
45
Liu XW, Bernardo MM, Fridman R, Kim HR Tissue inhibitor of metalloproteinase-1 protects human breast epithelial cells against intrinsic apoptotic cell death via the focal adhesion kinase/phosphatidylinositol 3-kinase and MAPK signaling pathway.
J Biol Chem
2003
;
278
:
40364
-72.