Mechanisms Underlying the Impairment of Ischemia-Induced Neovascularization in Matrix Metalloproteinase 2–Deficient Mice
Matrix metalloproteinases (MMPs) have been implicated in the process of neovascularization. However, the exact roles of individual MMPs in vessel formation are poorly understood. To study the putative role of MMP-2 in ischemia-induced neovascularization, a hindlimb ischemia model was applied to MMP-2+/+ and MMP-2−/− mice. Serial laser Doppler blood-flow analysis revealed that the recovery of the ischemic/normal blood-flow ratio in MMP-2−/− young and old mice remained impaired throughout the follow-up period. At day 35, microangiography and anti–l-lectin immunohistochemical staining revealed lesser developed collateral vessels and capillary formation in both old and young MMP-2−/− mice compared with the age-matched MMP-2+/+ mice. An aortic-ring culture assay showed a markedly impaired angiogenic response in MMP-2−/− mice, which was partially recovered by supplementation of the culture medium with recombinant MMP-2. Aorta-derived endothelial cells or bone marrow–derived endothelial progenitor cell (EPC)-like c-Kit+ cells from MMP-2−/− showed marked impairment of invasive or/and proliferative abilities. At day 7, plasma and ischemic tissues of vascular endothelial growth factor protein were reduced in MMP-2−/−. Flow cytometry showed that the numbers of EPC-like CD31+c-Kit+ cells in peripheral blood markedly decreased in MMP-2–deficient mice. Transplantation of bone marrow–derived mononuclear cells from MMP-2+/+ mice restored neovascularization in MMP-2−/− young mice. These data suggest that MMP-2 deficiency impairs ischemia-induced neovascularization through a reduction of endothelial cell and EPC invasive and/or proliferative activities and EPC mobilization.
It is well known that the process of new blood vessel formation is associated with extracellular matrix (ECM) remodeling involving various proteolytic systems. Among such systems, matrix metalloproteinases (MMPs) are a family of zinc-dependent endopeptidases comprising at least 20 members that are collectively capable of degrading all known ECM components.1,2 A number of studies have shown that various kinds of MMPs were upregulated in ischemia-induced angiogenesis.3 Although MMP activity is commonly thought to be involved in the process of angiogenesis, this notion has been challenged by recent studies using genetic or biological target methods. It has been reported that MMP-9 deficiency reduced neovascularization and tumor growth.4 Study of membrane-type1 (MT1)-MMP knockout mice revealed that the deficiency impaired neovascularization in a mouse corneal micropocket model.5 Whereas MMP-1 and MMP-10 appear to control the process of vascular regression rather than morphogenesis.6 On the other hand, certain MMPs, including MMP-12 and MMP-7, are capable of converting plasminogen into angiostatin to inhibit endothelial cell (EC) tubulogenesis in vitro.7 Interestingly, it has been reported that tissue levels of MMP-7 and MMP-9 are increased in α integrin–deficient mice and that these mice have increased circulating levels of angiostatin while also showing decreased angiogenesis.8 These findings collectively suggest that the different influences of various MMPs on wound- or tumor-induced angiogenesis may be directly related to the different functional roles of specific MMPs. Further work is necessary to determine the full spectrum of anti- or proangiogenic activities of the various MMPs expressed during neovascularization.
MMP-2, a major MMP derived from vascular ECs as well as smooth muscle cells (SMCs), degrades various ECM proteins such as the basement membrane.9,10 MMP-2 has been reported to play a role in angiogenesis by the observation that inhibition of MMP-2 by either its antisense oligonucleotide or inhibitors decreased angiogenesis.11,12 These findings, together with the results obtained in the MMP-2–deficient mice,13 have led to the conception that a shift in the net proteolytic balance between MMP-2 and its inhibitors, in favor of MMP-2 inhibition, would result in the suppression of the angiogenic phenotype. However, relatively little is known about the exact mechanism of MMP-2 involved in the angiogenic phenotype.
In this study, we evaluated the influence of the targeted deletion of the MMP-2 gene on the ischemia-induced neovascularization and tried to address the mechanisms underlying the impairment of neovascularization in the hindlimb ischemia models of young and old MMP-2–deficient mice.
Materials and Methods
We performed the following in vivo: ischemic hindlimb perfusion assay, biological analysis (real-time polymerase chain reaction [PCR], Western blotting, and gelatin zymography), histological analysis (capillary density, leukocyte and macrophage infiltration, analysis of endothelial progenitor cells, analysis of collateral vessel formation, bone marrow [BM] cell transplantation). Ex vivo, we performed aortic-ring culture assay; in vitro, cell isolation (BM-derived c-Kit+ cells and aortic ECs), cell adhesion assay, cell migration and invasion assays, cell proliferation assay, gelatinolytic net activity assay, plasma vascular endothelial growth factor ELISA, and immunocytofluorescence as detailed in the online data supplement available at http://circres.ahajournals.org.
Data are expressed as means±SE. Statistical analysis was adequately performed by the unpaired Student’s t test or analysis of variance followed by Scheffe’s multiple-comparison post hoc test. The comparative incidence of limb amputation was evaluated by the χ2 test. A value of P<0.05 was considered statistically significant. Collateral vessel diameter, capillary density, number of infiltrated macrophages and leukocytes, and length and number of endothelial sprouts were evaluated by 2 observers in a blinded manner, and the values they obtained were averaged.
MMP-2 Deficiency Impairs Ischemia-Induced Angiogenesis
The results in Figure 1A are photomicrographs of capillary density staining with endothelial-specific Griffonia Simplicifolia lectin-I (GSL-I) in ischemic or nonischemic tissues. In young mice, capillary density in ischemic muscle, as examined on day 35 after ligation, was significantly lower in MMP-2−/− than in MMP-2+/+ (Figure 1A and 1B). As anticipated, MMP-2 deficiency dramatically reduced capillary density in old MMP-2−/− as compared with the age-matched control (Figure 2A and 2B).
MMP-2 Deficiency Results in Diminished Ischemic Hindlimb Perfusion
In young mice, MMP-2 deficiency inhibited the recovery of hindlimb perfusion throughout the follow-up period, and the ratio of ischemic to normal laser Doppler blood flow (LDBF) was persistently lower in young MMP-2−/− than in MMP-2+/+ mice (Figure 1C and 1D). As anticipated, a poorer recovery of hindlimb perfusion and a lower ratio of ischemic to normal LDBF were also observed in old MMP-2−/− compared with the age-matched MMP-2+/+ during the observation periods, although the ratio in the old MMP-2+/+ was partially recovered in 60% of young MMP-2+/+ at day 14; however, no further improvement was observed up to day 35 (Figure 2C and 2D).
The microangiography on day 35 and the angiographic score revealed well-developed collateral vessels in the ischemic hindlimb of MMP-2+/+ mice (Figure 3A and 3B). The numbers not only of collateral intermediate vessels (<0.3 μm) but also of microvessels (≥0.3 μm) were significantly decreased by MMP-2 deficiency, resulting in angiographic score of 65% and 71% of MMP-2+/+ mice, respectively (Figure 3A and 3B). As shown in Figure 3C and 3D, MMP-2 deficiency reduced ischemia-induced peritoneal collateral vessel formation in MMP-2−/− mice. These data indicated that MMP-2 may be involved in collateral vessel formation that influences ischemic tissue blood perfusion.
MMP-2 Deficiency Promotes Spontaneous Amputation of Feet
As shown in Figure 3E and 3F, spontaneous foot amputation occurred in only 4.7% of 126 young and 15.7% of 38 old MMP-2+/+ mice. Among MMP-2−/− mice, however, spontaneous amputation was significantly increased by 20.3% of 108 young and 41.2% of 34 old mice compared with the relative controls (Figure 3E and 3F). These data provide indirect evidence that MMP-2 deficiency impaired ischemia-induced angiogenic response to influence hindlimb blood flow.
Effects of Acute Hindlimb Ischemia on Expression and Activity of MMPs and Tissue Inhibitors of Metalloproteinase
Gelatin zymographic analysis showed that gelatinolytic activity corresponded to the proform (72 kDa) and active form of MMP-2 (62 kDa) increased and peaked at day 7 after ligation and remained above the basal level for up to 35 days in the ischemic tissue of young MMP-2+/+ (Figure 4A). As expected, no MMP-2 activity was detected in the ischemic tissue of MMP-2−/− mice (Figure 4B). The gelatinolytic activity at 92 kDa (pro–MMP-9) increased as early as day 4 after ligation, and the level of activity was then decreased at days 7 to 35 in the ischemic tissue of both MMP-2+/+ and MMP-2−/− young mice (Figure 4A and 4B). However, quantitative analysis demonstrated higher levels of gelatinolytic activity of MMP-9 in MMP-2−/− mice were observed at days 4 to 7 after ligation as compared with those of MMP-2+/+ mice (Figure 4B). Similar to young mice, the changes for the gelatinolytic activities of MMP-2 and MMP-9 in old MMP-2+/+ or for the gelatinolytic activity of MMP-9 in old MMP-2−/− were observed in ischemic muscle, although the level of MMP-2 was lower in MMP-2−/− than in MMP-2+/+ (Figure I in the online data supplement, available at http://circres.ahajournals.org). There were no changes for MMP-2 and MMP-9 activities in nonischemic muscle from MMP-2+/+ or no change for MMP-9 activity in nonischemic muscle from MMP-2−/− mice during the follow-up periods (data not show).
Expression of MMP-2 and MMP-9 mRNAs increased in MMP-2+/+ from days 4 to 21, with the peak at day 7 after ligation (Figure 4C and 4D). Similar to the compensatory increase of MMP-9 activity, the higher levels of MMP-9 mRNA in ischemic tissues were observed in MMP-2−/− than in MMP-2+/+ during the follow-up periods (Figure 4D). However, there were no significant differences in the expression of other MMP mRNAs (MMP-3, MMP-13, and MT1-MMP mRNA) between both genotypes (data not shown). Tissue inhibitor of metalloproteinase (TIMP)-2 mRNA expression was markedly increased in MMP-2+/+ from days 4 to 14 after ligation, but constant levels of TIMP-1 mRNA were observed in MMP-2−/− during the follow-up period (Figure 4E). In contrast, the mRNA expression of TIMP-1 was lower in MMP-2+/+ than in MMP-2−/− during the follow-up periods (Figure 4F).
Among cysteine proteases, cathepsins L or S were recently shown to play a role in endothelial progenitor cell (EPC) homing and EC migration processes involving angiogenesis.14,15 Real-time PCR analysis of ischemic tissues in the young mice at day 7 after ligation showed that MMP-2 deficiency did not affect expression of cathepsins or their endogenous inhibitor cystatin C mRNAs. Taken together, these findings suggest that a major deferring pattern of expression and activity was attributable to MMP-2 in ischemic muscle following the induction of ischemia between both genotypes.
MMP-2 Deficiency Impairs Ex Vivo Angiogenesis in Aortic-Ring Culture Assay
The photographs in Figure 5A and 5C show microvascular sprouting circumscribing a patent lumen. On day 14 of culture, MMP-2+/+ and MMP-2−/− mouse aortic explants in endothelial basal medium (EBM)-2 without vascular endothelial growth factor (VEGF) mounted a weak tubulogenic response (Figure 5A). VEGF enhanced the tubulogenic response in aortic explants from not only MMP-2+/+ but also MMP-2−/−. However, quantitative analysis revealed that the mean density of the neovessels was significantly higher in the MMP-2+/+ aorta (9.2-fold more than untreated MMP-2+/+ aorta) than in the MMP-2−/− aorta (2.5-fold more than untreated MMP-2−/− aorta; Figure 5B). Interestingly, treatment with recombinant MMP-2 (0.5 μg/mL) in the presence of VEGF markedly enhanced its angiogenic response in the aorta of MMP-2−/− by 5.2-fold more than VEGF alone, whereas this was not the case in MMP-2+/+ (Figure 5B). Furthermore, the VEGF-mediated tubulogenesis in MMP-2+/+ aorta was significantly suppressed by TIMP-1 and TIMP-2 (3 μg/mL for each) and by peptide hydroxamic acid (GM6001, a broad-spectrum MMP inhibitor, 10 μmol/L) (Figure 5C and 5D). Multiple proteinases have been implicated in the angiogenic phenotype14,15; however, consistent with previous reports, neither trans-epoxysuccinyl-l-leucylamido-(4-guanidino)butane (a broad-spectrum cysteine protease inhibitor, 20 μmol/L) nor phenylmethylsulfonyl fluoride (a serine protease inhibitor, 2 mmol/L) showed a significantly reduced VEGF-mediated tubulogenic response compared with control (data not shown).16 These results indicate that MMP-2 plays a pivotal role in the VEGF-mediated angiogenesis in 3D ECM.
VEGF Enhances MMP-2 Expression
Gelatin zymography analysis showed that major bands for latent and active forms of MMP-2 were observed in the aorta-ring cultured media stimulated with VEGF at day 7 (Figure 5E). Consistent with a past report, a small amount of gelatinolytic activity for MMP-9 was observed in the cultured medium.17
Immunohistochemical analysis has shown that MMP-2 was localized predominantly to capillary ECs in the ischemic tissues (supplemental Figure II). As is shown in Figure 6A through 6C, only negligible levels of MMP mRNAs were detected in human umbilical vein endothelial cells (HUVECs) cultured in EBM-2 without specific growth factor. Exposure to VEGF for 24 hours caused the accumulation of MMP-9 (1.9-fold), MMP-2 (6.5-fold), and MT1-MMP (2.5-fold) transcripts. However, VEGF did not significantly affect the mRNA expression of other MMPs or TIMPs in HUVECs compared with controls. Furthermore, Western blotting and gelatin zymographic analyses of conditioned media from cultured ECs and EPC-like c-Kit+ cells demonstrated that VEGF or basic fibroblast growth factor (bFGF) remarkably upregulated MMP-2 protein expression and its gelatinolytic activity (Figure 6D and 6E). These results suggest that VEGF-dependent angiogenesis in aorta-ring culture may be attributable at least in part to the regulation of MMP-2 expression in ECs.
MMP-2 Deficiency Impairs EC and EPC-Like Cell Invasion and Proliferation
Compared with MMP-2+/+ ECs, MMP-2−/− ECs showed no significant difference in adhesion to type IV collagen (P=0.79), fibronectin (P=0.10), or vitronectin (P=0.12). Similarly, EC proliferation showed no difference between the ECs of two genotypes in the presence of bFGF (P=0.56) or VEGF (P=0.45). MMP-2 deficiency also did not significantly affect bFGF- or VEGF-directed EC migration (P=0.11, or P=0.23, respectively). However, MMP-2 deficiency significantly impaired either bFGF- (58±14 versus 175±35 invaded cell numbers/field, P=0.02) or VEGF-directed EC invasion through type I collagen–coated membrane (Figure 5F and 5G). Consistently, lower total gelatinolytic activity was observed in the lysate of cultured ECs from MMP-2−/− mice stimulated with VEGF for 48 hours than in that of MMP-2+/+ ECs (0.18±0.04 versus 0.42±0.11 fluorescence intensity, P<0.01). Taken together, these results suggest that MMP-2 may contribute to angiogenesis specifically by promoting the invasive and ECM degradation activities of EC, as opposed to EC adhesion or migration.
In addition, there were no significant differences in EPC-like c-kit+ cell adhesion and migration between MMP-2−/− and MMP-2+/+ mice (data not shown). Surprisingly, MMP-2 deficiency impaired bFGF- or VEGF-induced EPC-like c-kit+ cell proliferation and VEGF-directed its invasion (Figure 5F to 5H) and ECM degradation (0.25±0.06 versus 0.63±0.14 fluorescence intensity, P<0.05). Interestingly, TIMP-2 as well as GM6001 significantly suppressed invasion and proliferation compared with untreated controls (data not shown).
MMP-2 Deficiency Reduces Injury-Induced Inflammatory Reaction
Immunohistochemical analysis of ischemic muscle sections harvested on day 4 showed that fewer macrophages and leukocytes were detected in the extracapillary space in young MMP-2−/− mice than in control mice (Figure 7A and 7B). These results suggested that MMP-2 may be involved in the recruitment of macrophage and leukocyte into ischemic tissues.
MMP-2 Deficiency Reduces VEGF Expression and Levels in the Ischemic Tissues and Plasma
Quantitative real-time PCR analysis revealed that VEGF mRNA expression in ischemic tissues at day 7 after ischemia was significantly lower in MMP-2−/− mice than MMP-2+/+ mice (Figure 7C). Consistent with this finding, Western blotting analysis demonstrated that MMP-2 deficiency reduced the level of VEGF protein in ischemic tissues to below the level in wild type (Figure 7D). On other hand, immunofluorescence revealed that VEGF was expressed mainly by macrophages in the ischemic tissue of MMP-2−/− mice (supplemental Figure III). Surprisingly, the levels of degraded VEGF protein were higher in the ischemic tissues of the MMP-2−/− than in those of MMP-2+/+ (Figure 7D). ELISA of plasma from the MMP-2−/− and MMP-2+/+ mice for VEGF protein corroborated this observation (Figure 7E). Flt-1 mRNA expression in ischemic tissues was significantly lower in MMP-2−/− mice than in MMP-2+/+. However, these were no significant differences in the expression of Flk-1 and stromal derived factor (SDF)-1α mRNAs between the two genotypes (Figure 7F).
MMP-2 Deficiency Impairs Mobilization of EPC-Like CD31+c-Kit+ Cells From BM
Flow cytometry demonstrated a marked reduction in the number of CD31+c-Kit+ cells in peripheral blood (PB) from young MMP-2−/− at day 10 after ligation (Figure 8A and 8B). Similarly, the ratio of PB to BM for CD31+c-Kit+ cells was significantly lower in MMP-2−/− than in control mice (Figure 8E). Interestingly, there were significantly fewer CD31+c-Kit+ cells not only in the BM but also in the PB in old mice than in young, regardless of genotype. Furthermore, the number of CD31+c-Kit+ cells in PB and the ratio of PB to BM for these cells were lower in the old MMP-2−/− than in old wild type, although there was no statistical significance (Figure 8C to 8E). However, there was no difference between genotypes in the number of EPC-like CD31+c-Kit+ in the BM (Figure 6B and 6D). These findings indicate that MMP-2 may facilitate the mobilization of EPC-like cell from BM into the circulation to support vasculogenesis.
MMP-2+/+ BM Rescues Neovasculogenesis in MMP-2−/−
As is shown in supplemental Figure IVa and IVb, neovascularization following hindlimb ischemia was significantly less impaired in MMP-2−/− mice receiving MMP-2+/+ BM-derived muscle cells (MCs) than in MMP-2−/− mice receiving MMP-2−/− BM-derived MCs. Consistently, the capillary density and number of CD31+c-Kit+ cells in the PB were significantly higher MMP-2−/− mice receiving MMP-2+/+ BM-derived MCs than in MMP-2−/− mice receiving MMP-2−/− BM-derived MCs (supplemental Figure IVc through IVf). Furthermore, the rate of incidence for the autoamputation of ischemic hindlimbs was lower in MMP-2−/− mice receiving MMP-2+/+ BM-derived MCs (11.1%) than in MMP-2−/− mice receiving MMP-2−/− BM-derived MCs (24%). These data suggested that the lack of MMP-2 specially impairs the functional incorporation of BM-derived EPCs. This may account for the impaired neovascularization capacity of MMP-2−/−.
The present study used a gene-targeting strategy to establish a novel role for EC-derived MMP-2 in neovascularization. We provided both in vitro and in vivo evidence that MMP-2 promotes the migratory and angiogenic capacity of cultured aortic EC and BM-derived EPC-like c-Kit+ cells and further neovessel growth. In line with recent studies indicating that certain MMPs are involved in EPC mobilization, our data underscore the critical role of MMP-2 in EPC mobilization into the circulation to support ischemia-induced neovascularization. The mechanisms underlying the impairment of ischemia-induced neovascularization in MMP-2–deficient mice are schematically represented in supplemental Figure X.
The capillary density, collateral artery, and blood flow in the ischemic tissues were much lower in MMP-2−/− mice than in MMP-2+/+ mice, despite the compensatory increase of MMP-9 expression and its activity and TIMP-1 expression. This suggested that endogenous MMP-2 contributes to ischemia-induced angiogenesis at least in this ischemic model. Immunohistochemical analysis has shown that MMP-2 was localized predominantly to capillary ECs. Real-time PCR revealed that the genotypes did not differ in their expression of other MMPs (MMP-3, MMP-13, and MMP-14) or of cysteine protease (cathepsins S, K, and L) mRNAs in the ischemic hindlimbs. Moreover, consistent with past reports,9,17,18 gelatin zymographic analysis revealed major digestive bands for both latent and active forms of MMP-2 in the conditioned medium from the aorta-ring culture in the presence of VEGF, whereas only the weak digestive bands for MMP-9 were detected in the conditioned medium. Furthermore, VEGF stimulated the expression of MMP-2 mRNA (in HUVECs), MMP-2 protein (in ECs), or MMP-2 activity (in c-Kit+ cells). On the other hand, the expression of VEGF mRNA as well as VEGF protein levels were decreased in the local ischemic tissues of MMP-2−/− compared with MMP-2+/+. Based on these observations, we can conclude that VEGF-induced EC-derived MMP-2 is among the most essential MMPs in angiogenesis associated with ECM degradation. This concept is further supported by the results that MMP-2 deficiency markedly abolished the matrix-degrading activity of VEGF-stimulated EC extracts. In addition, the data on the decrease in VEGF receptor Flt-1 mRNA expression in ischemic tissue of MMP-2−/− mice provided indirect evidence that this reduction of Flt-1 expression may further impair the VEGF signal pathway.
Within the MMP family, surface localization of the activated MMPs may be crucial for cellular events, including cell adhesion, migration/invasion, and proliferation, which have been associated with angiogenesis-dependent tumor morphogenesis.12,19 Recently, our studies, as well as those of others, using MMP-2 knockout animals provided evidence that biological effects are specially mediated by MMP-2 in the formation of atherosclerotic and neointimal lesions.10 In line with these findings, ex vivo aortic-ring culture assays demonstrated that MMP-2 deficiency profoundly impairs VEGF-mediated tubulogenic response in aorta explants (15.2% of wild type), and MMP-2−/− aorta explants treated with recombinant MMP-2 had higher (5.2-fold) response than the untreated controls. Consistent with recent findings11,17 that an endogenous inhibitors of MMP-2, TIMP-1 and TIMP-2, as well as a broad-spectrum inhibitor of MMPs, GM6001, inhibited neovessel formation significantly by 72.2%, 81.0%, and 85.1%, respectively. As revealed by EC invasion assay, genetic deletion of MMP-2 markedly abolished EC invasive ability through type I collagen gel. However, MMP-2 deficiency affected neither cell adhesion and migration nor proliferation. These findings suggest that MMP-2–mediated angiogenesis is not attributable to EC adhesion, proliferation, or motility impairment but rather to the impairment of EC-invasive activity. It should be noted that serine proteases and cathepsins were recently shown to contribute to angiogenesis.2,14,15 In our experiments, however, in contrast to MMP inhibitors and in agreement with earlier findings,14,17 VEGF-mediated tubulogenesis was not affected by the pharmacological inhibition of serine proteases or cathepsins. It seems that these proteolytic proteases may not play a major role in tubulogenesis, at least in this type I collagen gel model.
It has been reported that genetic deletion or pharmacological selective inhibition of MMP-2 reduced macrophage and polymorphonuclear leukocyte infiltration in the infarcted myocardium at days 3 to day 7.20 Recently, a few studies have provided important evidence that chemoattractants, such as laminin LGTIPG and fibronectin RGDS peptides generated by the action of MMP-2 in the infarcted myocardium, are responsible for macrophage migration.20 In the present study, immunohistochemical analysis demonstrated that MMP-2 deficiency impaired macrophage and leukocyte infiltration in the ischemic muscle of MMP-2−/− at day 4 after ischemia. It is possible that MMP-2 deficiency reduces macrophage and leukocyte infiltration through impairment of the generation of ECM-degraded fragments such as LGTIPG or RGDS, although we lack direct evidence at this point. Activated macrophages have been shown to predominantly produce VEGF to facilitate ischemia-induced angiogenesis and arteriogenesis.21 Moreover, a previous study demonstrated that the blockade of endogenous MCP-1 activity by a dominant negative mutant of MCP-1 overexpression remarkably inhibited ischemia-induced neovascularization by reducing tumor necrosis factor-α and VEGF inductions through the suppression of macrophage infiltration.21 In the present study, we observed that the levels of VEGF in plasma as well as its mRNA and protein expression in ischemic tissue significantly decreased in MMP-2−/− mice at day 7 after ischemia. Collectively, these findings suggest that MMP-2–mediated impairment of arteriogenesis and angiogenesis may be at least partially reduction of VEGF production resulting from the decreased macrophage infiltration in the ischemic hindlimb of MMP-2−/− mice. It is noteworthy that MMP-9 expression and activity were compensatorily increased as early as days 4 to 7 after ischemia.
Recent studies recognized an additional large family of metalloproteinases, the ADAM (a disintegrin metalloproteinase domain) family, which have a transmembrane domain and can act as sheddases on the surface of the cell.2,22 At least 34 ADAMs are known, of which 19 have proteolytic activity.2,22 Certain ADAMs (ADAM-10, -15, and -17) are known that have properties directly important for the regulation of angiogenesis. Further study will be required to understand the role for the ADAMs in impairment of ischemia-induced neovascularization in MMP-2−/− mice.
The data from flow cytometry demonstrated that there were significantly fewer EPC-like CD31+c-Kit+ MCs in the PB of young MMP-2−/− mice after ischemia than in that of young MMP-2+/+ mice. Immunofluorescence confirmed MMP-2+/+ green fluorescent protein BM-derived c-Kit+ EPCs homing to ischemic tissues (supplemental Figure VI). Moreover, ELISA showed that MMP-2 deficiency significantly reduced the VEGF plasma level at day 7 after ischemia in young MMP-2−/− compared with young MMP-2+/+ mice. VEGF has been shown to modulate EPC mobilization from BM to the circulation and to support vasculogenesis in the adult.23,24 More importantly, in the present study, the impaired recovery of blood flow after the induction of hindlimb ischemia in young MMP-2−/− mice was rescued by the reception of MMP-2+/+ BM-derived MCs, but not by that of MMP-2−/− BM-derived MCs. Furthermore, transplantation with MMP-2+/+ BM-derived MCs significantly protected autoamputation of MMP-2−/− ischemic hindlimbs as compared with transplantation with MMP-2−/− BM-derived MCs. These data suggest that the impairment of vasculogenesis may be attributable to the decrease of EPC mobilization and to its functional incorporation into the vasculature in young MMP-2−/− mice. It is possible that the reduced VEGF concentration in PB of MMP-2−/− mice may contribute to the reduced number of EPC in PB after ischemia. It should be noted that there is no significant difference in peripheral hematopoietic cells between the 2 genotypes.
In the present study, we demonstrated that MMP-2 expression was enhanced with VEGF or bFGF in EPC-like c-Kit+ cells. Inhibition or genetic ablation of MMP-2 specifically reduced the VEGF-mediated invasive and proliferative potential of EPC in vitro. Based on these data, in conjunction with the fact that MMP-2 deletion had no effect on EPC adhesion or migration in vitro, MMP-2 may contribute to vasculogenesis specifically by promoting the invasive and proliferative activities of EPC, as opposed to EPC adhesion and migration. However, our finding of only partial inhibition of ischemia-induced neovascularization by the genetic depletion of MMP-2 indicates that other proteases also participate in cellular events to support ischemia-induced neovascularization. It should be noted that there is a discrepancy in the effects of MMP-2 deficiency on cell proliferation between mature EC and immature EPC. Further study will be required to understand the exact mechanism underlying the effect of MMP-2 on EPC proliferation.
Recently, a few studies using experimental aging models have shown that aging impairs ischemia-induced neovascularization.25 However, the effects of aging on the number of EPCs have not been well examined. In the present study, the number of EPC-like CD31+c-Kit+ MCs was significantly decreased not only in the BM but also in the PB in old mice of both genotypes as compared with younger mice. Importantly, the recovery of capillary density and blood flow in ischemic hindlimb are substantially impaired in old mice of both genotypes, although the ratio of ischemic to normal LDBF was partially recovered in old MMP-2+/+ and in less than 60% of young MMP-2+/+ and was kept at the basal level in old MMP-2−/− during study periods after ischemia. These findings indicated that aging impairs ischemia-induced neovascularization at least partially through the reduction of the EPC mobilization in PB and BM. Furthermore, MMP-2 deficiency reduced the number of EPC-like CD31+c-Kit+ MCs in PB and the ratio of PB to BM for these cells of old mice. This result suggests that the marked reduction of neovascularization in old MMP-2−/− mice compared with old wild-type mice might be partially attributable to an impairment in the functional activity of EPC in old MMP-2−/−, resulting in impaired vasculogenesis.
Taken together, our results support the notion that MMP-2 is necessary not only for angiogenesis but also for vasculogenesis. This novel biological function of MMP-2 may be exploited for the therapeutic control of pathophysiological neovascularization.
We acknowledge the technical assistance of E. Asai, Y. Iwatsuki, M. Aoki, M. Hirata, and A. Inoue. We also thank S. Itohara (RIKEN Brain Science Institute) for providing MMP-2–deficient mouse.
Sources of Funding
This work was supported in part by the Scientific Research Fund of the Ministry of Education, Culture, Sports, Science, and Technology of Japan (no. 17590719 to X.W.C.; 17590723 to M.K.) and by a research grant from the Japan Heart Foundation (no. 26-007508 to X.W.C.).
↵*Both authors contributed equally to this work.
Original received October 22, 2006; revision received January 30, 2007; accepted February 8, 2007.
van Hinsbergh VW, Engelse MA, Quax PH. Pericellular proteases in angiogenesis and vasculogenesis. Arterioscler Thromb Vasc Biol. 2006; 26: 716–728.
Jin DK, Shido K, Kopp HG, Petit I, Shmelkov SV, Young LM, Hooper AT, Amano H, Avecilla ST, Heissig B, Hattori K, Zhang F, Hicklin DJ, Wu Y, Zhu Z, Dunn A, Salari H, Werb Z, Hackett NR, Crystal RG, Lyden D, Rafii S. Cytokine-mediated deployment of SDF-1 induces revascularization through recruitment of CXCR4+ hemangiocytes. Nat Med. 2006; 12: 557–567.
Zhou Z, Apte SS, Soininen R, Cao R, Baaklini GY, Rauser RW, Wang J, Cao Y, Tryggvason K. Impaired endochondral ossification and angiogenesis in mice deficient in membrane-type matrix metalloproteinase I. Proc Natl Acad Sci U S A. 2000; 97: 4052–4057.
Davis GE, Senger DR. Endothelial extracellular matrix: biosynthesis, remodeling, and functions during vascular morphogenesis and neovessel stabilization. Circ Res. 2005; 97: 1093–1107.
Cornelius LA, Nehring LC, Harding E, Bolanowski M, Welgus HG, Kobayashi DK, Pierce RA, Shapiro SD. Matrix metalloproteinases generate angiostatin: effects on neovascularization. J Immunol. 1998; 161: 6845–6852.
Pozzi A, Moberg PE, Miles LA, Wagner S, Soloway P, Gardner HA. Elevated matrix metalloprotease and angiostatin levels in integrin alpha 1 knockout mice cause reduced tumor vascularization. Proc Natl Acad Sci U S A. 2000; 97: 2202–2207.
Haas TL, Davis SJ, Madri JA. Three-dimensional type I collagen lattices induce coordinate expression of matrix metalloproteinases MT1-MMP and MMP-2 in microvascular endothelial cells. J Biol Chem. 1998; 273: 3604–3610.
Kuzuya M, Kanda S, Sasaki T, Tamaya-Mori N, Cheng XW, Itoh T, Itohara S, Iguchi A. Deficiency of gelatinase a suppresses smooth muscle cell invasion and development of experimental intimal hyperplasia. Circulation. 2003; 108: 1375–1381.
Collen A, Hanemaaijer R, Lupu F, Quax PH, van Lent N, Grimbergen J, Peters E, Koolwijk P, van Hinsbergh VW. Membrane-type matrix metalloproteinase-mediated angiogenesis in a fibrin-collagen matrix. Blood. 2003; 101: 1810–1817.
Silletti S, Kessler T, Goldberg J, Boger DL, Cheresh DA. Disruption of matrix metalloproteinase 2 binding to integrin alpha vbeta 3 by an organic molecule inhibits angiogenesis and tumor growth in vivo. Proc Natl Acad Sci U S A. 2001; 98: 119–124.
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: 1048–1051.
Shi GP, Sukhova GK, Kuzuya M, Ye Q, Du J, Zhang Y, Pan JH, Lu ML, Cheng XW, Iguchi A, Perrey S, Lee AM, Chapman HA, Libby P. Deficiency of the cysteine protease cathepsin S impairs microvessel growth. Circ Res. 2003; 92: 493–500.
Urbich C, Heeschen C, Aicher A, Sasaki K, Bruhl T, Farhadi MR, Vajkoczy P, Hofmann WK, Peters C, Pennacchio LA, Abolmaali ND, Chavakis E, Reinheckel T, Zeiher AM, Dimmeler S. Cathepsin L is required for endothelial progenitor cell-induced neovascularization. Nat Med. 2005; 11: 206–213.
Chun TH, Sabeh F, Ota I, Murphy H, McDonagh KT, Holmbeck K, Birkedal-Hansen H, Allen ED, Weiss SJ. MT1-MMP-dependent neovessel formation within the confines of the three-dimensional extracellular matrix. J Cell Biol. 2004; 167: 757–767.
Seals DF, Courtneidge SA. The ADAMs family of metalloproteases: multidomain proteins with multiple functions. Genes Dev. 2003; 17: 7–30.
Asahara T, Takahashi T, Masuda H, Kalka C, Chen D, Iwaguro H, Inai Y, Silver M, Isner JM. VEGF contributes to postnatal neovascularization by mobilizing bone marrow-derived endothelial progenitor cells. EMBO J. 1999; 18: 3964–3972.
Shimada T, Takeshita Y, Murohara T, Sasaki K, Egami K, Shintani S, Katsuda Y, Ikeda H, Nabeshima Y, Imaizumi T. Angiogenesis and vasculogenesis are impaired in the precocious-aging klotho mouse. Circulation. 2004; 110: 1148–1155.