SEARCH

SEARCH BY CITATION

Keywords:

  • Angiogenesis;
  • Hypoxia;
  • Murine marrow stromal cells;
  • Cell migration;
  • Matrix metalloproteinase

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Recent evidence indicates that bone-marrow-derived stromal cells (MSCs) have a histology coherent with endothelial cells that may enable them to contribute to tumor angiogenesis through yet undefined mechanisms. In this work, we investigated the angiogenic properties of murine MSCs involved in extracellular matrix degradation and in neovascularization that could take place in a hypoxic environment such as that encountered in tumor masses. MSCs were cultured in normoxia (95% air and 5% CO2) or in hypoxia (1% oxygen, 5% CO2, and 94% nitrogen). We found that hypoxic culture conditions rapidly induced MSC migration and three-dimensional capillary-like structure formation on Matrigel. In vitro, MSC migration was induced by growth-factor- and cytokine-enriched conditioned media isolated from U-87 glioma cells as well as from MSCs cultured in hypoxic conditions, suggesting both paracrine and autocrine regulatory mechanisms. Although greater vascular endothelial growth factor levels were secreted by MSCs in hypoxic conditions, this growth factor alone could not explain their greater migration. Interestingly, matrix metalloproteinase (MMP)-2 mRNA expression and protein secretion were downregulated, while those of membrane-type (MT)1-MMP were strongly induced by hypoxia. Functional inhibition of MT1-MMP by a blocking antibody strongly suppressed MSC ability to migrate and generate capillary-like structures. Collectively, these data suggest that MSCs may have the capacity to participate in tumor angiogenesis through regulation of their angiogenic properties under an atmosphere of low oxygen that closely approximates the tumor microenvironment.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Bone-marrow-derived stromal cells (MSCs) are a population of pluripotent adherent cells residing within the bone marrow microenvironment and are frequently referred to as mesenchymal stem cells due to their ability to differentiate into many mesenchymal phenotypes [1]. In contrast to their hematopoietic counterparts, MSCs demonstrate a strikingly enhanced ability to adhere to tissue-culture surfaces and to differentiate in culture into cells of the osteogenic, chondrogenic, tendonogenic, adipogenic, and myogenic lineages [2]. In their undifferentiated state, MSCs do not express hematopoietic or endothelial cell surface markers such as CD11, CD14, CD31, CD34, or CD45 [3]. Intriguingly, however, recent work has shown that altering culture conditions could render these CD34 cells capable of differentiating into endothelial cells, and this further highlights the potential role of MSCs in neovascularization [3, 4]. All these attributes thus make the MSC an interesting cell phenotype to investigate in light of their recently reported potential to differentiate into mesoderm-derived endothelial cells [5] and their ability to transdifferentiate in vivo into CD31+ endothelial cells [6, 7].

The possible involvement of human MSCs in neovascularization was recently proposed due to their capacity to contribute to tumor angiogenesis in vivo [8]. In addition, MSCs also exhibited selective interaction toward epithelial tumor cells [9], and implication of growth and survival of myeloma plasma cells was also recently reported and attributed to MSCs [10]. Consequently, a potential role of MSCs in response to tumor angiogenic factors may thus impact on tumor cell growth. However, whether MSC angiogenic properties enable these cells to participate in extracellular matrix (ECM) proteolysis in order to migrate, proliferate, and form capillary-like structures either within or at distal sites of the bone marrow environment remains to be evaluated. In the present study, we investigated critical molecular mechanisms regulating the phenotypical and functional properties of MSCs that further define their role in neovascularization and in microvascular network remodeling.

The molecular regulation of ECM proteolysis that occurs during angiogenesis and of MSC migration/invasion properties is largely accomplished through the action of the soluble and membrane-bound matrix metalloproteinases (MMPs) [11, 12]. More recently, the contribution of soluble MMPs supplied by bone-marrow-derived cells was documented in response to several growth factors, cytokines, and chemokines [13, 14], as well as skin carcinogenesis [15]. In light of these published data, one can thus predict a critical role for MMPs in support of MSC cell-cell and cell-matrix interactions that would be required for tumor neovascularization. Although there has been much emphasis on the role of soluble secreted MMPs, the precise role and regulation of the membrane-type MMPs at the cell surface of MSCs in neovascularization have, however, been largely underscored. Of particular interest, targeted disruption of the membrane-type 1 MMP (MT1-MMP) gene in mice resulted in disorders in connective tissue growth [16], illustrating the pivotal roles of MT1-MMPs in stromal remodeling and during cellular growth and differentiation processes involved in development.

Hypoxia alters fundamentally and physiologically important intracellular pathways and has long been recognized as a stimulus for tumor angiogenesis [17]. Recent evidence demonstrated that long-term cultures of MSCs under an atmosphere of low oxygen that closely approximates documented in vivo oxygen tension enabled these cells to optimally proliferate and differentiate [18]. More recently, a three-dimensional (3-D) in vivo model was used where MSCs were resuspended in Matrigel and the plugs were implanted subcutaneously in isogenic mice recipients. At 2 weeks postimplantation, Matrigel plugs were removed, and MSCs demonstrated functional vascular plasticity and transdifferentiated into CD31 endothelial cells, possibly due to the low oxygen tension of Matrigel implants [7]. In this paper, we investigated the effects of a low oxygen tension on the ability of MSCs to exhibit angiogenic-like properties and characterized some molecular players involved in these processes. Our data indicate that: A) MSC migration and capillary-like structure formation are rapidly induced by growth-factor-mediated regulation in hypoxic culture conditions; B) greater MSC migratory potential, however, cannot be solely attributed to vascular endothelial growth factor (VEGF), and C) MT1-MMP plays a key role, independent of pro-MMP-2 activation, in MSC 3-D capillary-like structure formation. These observations provide molecular and cellular rationales to potential MSC involvement at sites of active microvascular network remodeling, possibly in response to tumor angiogenic factors.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Antibodies and Chemicals

The following materials were purchased from the indicated sources. Anti-tissue inhibitor of MMPs (TIMP)-2 polyclonal antibody (AB801), anti-MMP-2 monoclonal antibody (mAb) (MAB3308), anti-MT1-MMP catalytic-domain mAb (AB8102), anti-MT1-MMP raised against the hinge domain (AB815), and the enhanced chemiluminescence Western blot kit were from Chemicon (Temecula, CA; http://www.chemicon.com). Anti-VEGF polyclonal antibody (SC-507) was from Santa Cruz Biotechnology (Santa Cruz, CA; http://www.scbt.com). Horseradish peroxidase-conjugated donkey anti-rabbit or anti-mouse IgG were obtained from Jackson Immunoresearch Laboratories (West Grove, PA; http://www.jacksonimmuno.com). Polyvinylidene difluoride (PVDF) membranes were from Roche Diagnostics (Laval, QC; http://www.roche-applied-science.com). All products for electrophoresis and zymography were from Bio-Rad (Mississauga, ON; http://www.bio-rad.com). TRIzol reagent, trypsin, penicillin, and streptomycin were from GIBCO/BRL (Burlington, ON; http://www.invitrogen.com). Fetal bovine serum (FBS) was from HyClone Laboratories (Logan, UT; http://www.hyclone.com). Agarose, gelatin, sodium dodecyl sulfate (SDS), sphingosine-1-phosphate (S1P), and Triton X-100 were from Sigma (St. Louis, MO; http://www.sigmaaldrich.com). Matrigel was from Becton Dickinson Labware (Bedford, MA; http://www.bd.com). Basic fibroblast growth factor (bFGF) and human recombinant VEGF were from R&D Systems (Minneapolis, MN; http://www.rndsystems.com).

Cell Culture and Experimental Hypoxic Conditions

The human U-87 glioblastoma cell line was purchased from American Type Culture Collection, maintained in modified Eagle's medium containing 10% FBS, 2 mM glutamine, 100 U/ml penicillin, and 100 mg/ml streptomycin, and cultured under a humidified atmosphere of 5% CO2. MSCs were isolated from mouse bone marrow and cultured as follows. Whole bone marrow was harvested from the femurs and tibias of 18–22 g female C57Bl/6 mice (Charles River; Laprairie Co., QC; http://www.criver.com) sacrificed by CO2 inhalation. Cells were plated in high glucose Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% FBS and 50 U/ml penicillin/streptomycin. After 5–7 days of incubation in a humidified incubator at 37°C with 5% CO2, the nonadherent hematopoietic cells were discarded. The adherent MSCs were further maintained in a humidified incubator at 37°C with 5% CO2 in the absence of any exogenous growth factor or anchoring materials such as fibronectin or collagen. MSCs were kept subconfluent and expanded in number over 14 passages by a 1:2 split on a weekly basis. Our culture-expanded murine MSCs were CD31, CD34, and CD45 in vitro as assessed by immunohistochemistry (data not shown), consistent with other reports [1921]. Culture-expanded MSCs were trypsinized and subsequently stained with the following mAbs: phycoerythrin (PE)-labeled anti-CD44 (clone IM7), anti-CD45 (clone 30-F11), anti-Flk1 (clone Avas12a1), biotinylated anti-CD31 (clone 390), anti-CD34 (clone RAM34) (all from BD Biosciences; San Diego, CA; http://www.bdbiosciences.com), biotinylated anti-Flt4 (clone AFL4), and anti-Tie2 (clone Tek 4) (eBioscience; San Diego, CA; http://www.ebioscience.com). Biotinylated mAbs were revealed by CyChrome streptavidin (BD Biosciences). PE-labeled or biotinylated rat isotypic control immunoglobulins were from BD Biosciences. Events were acquired on a Coulter EPICS flow cytometer (Beckman Coulter; Fullerton, CA; http://www.beckman.com) and analyzed by means of Win MDI 2.7 software. In vitro culture-expanded adherent MSCs were uniformly fibroblast-like in appearance. Analysis by flow cytometry performed at passage 14 revealed that MSCs expressed CD44 but were negative for CD45, CD31, KDR/flk1 (VEGF-R2), flt-4 (VEGF-R3), and Tie2 (angiopoietin receptor) (data not shown). Progeny derived from this population were utilized for all subsequent analyses. Hypoxic conditions were attained by incubation of confluent cells in an anaerobic box. The oxygen was maintained at 1% by a compact gas oxygen controller Proox model 110 (Reming Bioinstruments Co., Redfield, NY) with a residual gas mixture composed of 94% N2 and 5% CO2.

Cell Migration/Invasion Assay

Cells were dislodged after brief trypsinization and dispersed into homogeneous single-cell suspensions that were washed extensively and resuspended in DMEM to a concentration of 106 cells/ml. To assess migration from established monolayers, cells (105) were dispersed onto 0.15% gelatin/phosphate-buffered saline-coated chemotaxis filters (Corning/Costar; Acton, MA; http://www.corning.com; 8-μm pore size) within Boyden chamber inserts and allowed to adhere for 1 hour at 37°C, after which they were challenged by the addition of 600 μl of a chemoattractant solution composed of conditioned media isolated from the indicated serum-deprived cell lines to the lower compartments. Migration was allowed to proceed for 4 hours at 37°C in either 5% CO2/95% air (normoxia) or 1% O2/5% CO2/94% N2 (hypoxia). Cells remaining attached to the upper surfaces of the filters were carefully removed with cotton swabs. Cells that had migrated to the lower surfaces of the filters were fixed with 3.7% formaldehyde, stained with 0.1% crystal violet/20% MeOH, and counted by microscopic examination. The average number of migrating cells per field was assessed by counting at least four random fields per filter. Data points indicate the mean obtained from three separate chambers within one representative experiment.

Capillary-Like Structure Formation Assay

Induction of tubulogenesis was performed using Matrigel [22]. Matrigel was thawed on ice to prevent premature polymerization; aliquots of 50 μl were plated into individual wells of 96-well tissue culture plates (Costar) and allowed to polymerize at 37°C for at least 60 minutes. Cells were removed from confluent cultures by treatment with trypsin 0.05%-EDTA 0.53 mM. The cells were washed in serum-containing medium then resuspended to 106 cells/ml. Into each culture well was added 100-μl cell suspension with or without additional test substances. Each dose of control or test compound was assayed in duplicate, and all experiments were performed at least three times. For quantitation of tube formation, the total length of the tubes formed in a unit area was digitized and measured using ECLIPSE software. For each test, five randomly chosen areas were measured and averaged.

Total RNA Isolation and Reverse Transcriptase-Polymerase Chain Reaction (RT-PCR) Analysis

Total RNA was extracted from cultured MSCs using the TRIzol reagent. One microgram of total RNA was used for first-strand cDNA synthesis followed by specific gene product amplification with the MasterAmp One Tube RT-PCR Kit (Epicentre; Madison, WI; http://www.epicentre.com). Primers for hypoxia-induced factor-1 (HIF-1α), MMP-2, and MT1-MMP were all derived from mouse sequences, and PCR conditions were optimized so that the gene products were found to be at the exponential phase of amplification. β-actin was used as an internal control and was found to be constant between all tested conditions. PCR products were resolved on 2% agarose gels containing 1 μg/ml ethidium bromide.

Gelatin Zymography

Gelatinolytic activity in culture media from monolayer cultures was detected by gelatin zymography, as described previously [23]. Briefly, an aliquot (30 μl) of the culture medium was subjected to SDS-PAGE using a 9% (w/v) acrylamide gel containing 0.1 mg/ml gelatin. The gels were then incubated for 30 minutes at room temperature twice in 2.5% Triton X-100 to remove SDS and rinsed five times in double-distilled H2O. The gels were further incubated at 37°C for 20 hours in 20 mM NaCl, 5 mM CaCl2, 0.02% Brij-35, and 50 mM Tris-HCl buffer (pH 7.6), then stained with 0.1% Coomassie Brilliant blue R-250 and destained in 10% acetic acid/30% methanol in H2O. Gelatinolytic activity was detected as unstained bands on a blue background.

Western Blot Analysis

The preparation of whole-cell lysates was performed by resuspending cells in a buffer containing 1% Triton X-100, 0.5% Nonidet P-40, 150 mM NaCl, and 20 mM Tris-HCl pH 7.4. Cell lysates (30 μg/well) or aliquots of culture media were subjected to SDS-PAGE under reducing conditions and transferred to PVDF membranes. Immunoblotting procedures were performed as previously described in detail [23]. PVDF membranes were incubated with primary antibody, washed, and incubated with a secondary antibody conjugated to horseradish peroxidase. Bound IgG were detected using a chemiluminescent substrate. Immunoreactive bands were quantified by densitometric measurement using a Personal Densitometer (Molecular Dynamics, Sunnyvale, CA).

Statistical Data Analysis

Data are representative of three or more independent experiments. Statistical significance was assessed using Student's unpaired t-tests to compare migration and extent of capillary-like structure formation with untreated control basal migration or spontaneous MSC tube formation. Probability values of less than 0.05 were considered significant.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Low Oxygen Tension Rapidly Induces MSC Migration and Capillary-Like Structure Formation

We first tested the effect of hypoxia on MSC capacity to manifest angiogenic features. MSCs were cultured under either normoxic or hypoxic conditions for 4 hours. In vitro migration of MSCs was assessed using gelatin-coated filters and found to be more than fourfold greater (normoxia: 132 ± 9 cells/field; hypoxia: 571 ± 20 cells/field) when cell migration proceeded under low oxygen tension conditions (Fig. 1, upper panels). When cells were plated onto a 3-D Matrigel basement membrane model that closely mimics the structure, composition, physical properties, and functional characteristics of the basement membrane in vivo [24] and incubated under hypoxic conditions, 3-D capillary-like structure formation was also found to be rapidly induced (Fig.1, lower panels).

thumbnail image

Figure Figure 1.. Hypoxic culture conditions rapidly stimulate both MSC migration and their ability to generate capillary-like structures.MSCs were assayed for basal migration in normoxic or hypoxic conditions as described inMaterials and Methods. In parallel, MSCs were also seeded on Matrigel, and capillary-like structure formation was monitored in normal and hypoxic conditions within a 4-hour period.

Download figure to PowerPoint

Hypoxia Downregulates pro-MMP-2 Expression and Secretion from MSCs

Extracellular matrix proteolysis is an essential step in the angiogenic process that requires MMP activation. Confluent MSCs were thus serum deprived and cultured in normal and hypoxic conditions to monitor the effect on soluble MMP secretion. Gelatin zymography was performed and the hydrolytic activity of MMP-2 was assessed (Fig. 2). Intriguingly, hypoxic culture conditions resulted in a marked time-dependent lower secreted latent pro-MMP-2 gelatinolytic activity (Fig. 2A). Using Western blotting and immunodetection with a specific anti-MMP-2 antibody, this lower activity was further correlated with ∼40% lower extracellular pro-MMP-2 protein levels compared with normoxia (Fig. 2B). Since the MSC proliferation rate was not affected by these short-term incubation conditions (not shown), one can thus hypothesize that hypoxia either decreased MMP-2 gene expression and protein synthesis rate or affected vesicular trafficking components responsible for pro-MMP-2 secretion.

thumbnail image

Figure Figure 2.. Hypoxic culture conditions downregulate extracellular MMP-2 gelatinolytic activity through diminished protein secretion.MSCs were serum starved and cultured in normal or hypoxic conditions for up to 48 hours. A) Aliquots (30 μl) of the conditioned media were assayed for MMP-2 activity by gelatin zymography, or B) assessed for MMP-2 protein content by Western blotting. A representative densitometric analysis of pro-MMP-2 protein levels derived from the Western blots data is shown for normal (open circle) and hypoxic (closed circle) culture conditions.

Download figure to PowerPoint

Hypoxia Promotes the Expression of MT1-MMPs and Angiogenesis-Related Markers in MSCs

In order to assess hypoxia-induced modulations of gene expression levels, total RNA was isolated from MSCs cultured under normal and hypoxic conditions, and RT-PCR was used to amplify cDNA from both MMP-2 and its known cell surface activator, MT1-MMP. Results in Figure 3 show that, under normoxia, MMP-2 gene expression was stable after 24 hours of serum deprivation (MSCs were cultured in high-glucose DMEM without addition of 10% FBS) but started to decline at 48 hours. In accordance with the zymography and Western blot data in Figure 2, hypoxia also downregulated MMP-2 gene expression, while that of β-actin remained unaffected. Interestingly, these MMP-2 modulations were also recently reported in hypoxic-differentiated endothelial cells [25], further strengthening the potential common endothelial-like phenotype that may characterize MSCs.

thumbnail image

Figure Figure 3.. Hypoxia upregulates MT1-MMP gene expression but not that of MMP-2.MSCs were cultured in serum-deprived media (high-glucose DMEM without addition of 10% FBS) under normal (N) or hypoxic (H) conditions for up to 48 hours. Total RNA was isolated and gene expression levels of MMP-2, MT1-MMP, and β-actin assessed by RT-PCR using murine primer sequences as described inMaterials and Methods.

Download figure to PowerPoint

In contrast to MMP-2, MT1-MMP gene expression was strikingly greater when MSCs were exposed to hypoxic conditions, and maximal induction was already observed after as few as 24 hours of incubation (Fig. 3). Longer (48-hour) serum deprivation of MSCs resulted in greater MT1-MMP gene expression that was, at that point, comparable whether cells were cultured in normal or hypoxic culture conditions. Hypoxia-induced MT1-MMP gene expression was also further correlated with greater levels of MT1-MMP protein, as assessed by Western blotting in cell lysates (Fig. 4A). This was demonstrated by greater levels of the respective 63-kDa proactive forms of MT1-MMP (corresponding to the recombinant MT1-MMP protein isolated from transfected COS-7 cells [26]) and 60-kDa active forms of MT1-MMP. Interestingly, the intracellular protein expression of early growth response-1 (Egr-1), a transcription factor known to tightly regulate MT1-MMP gene expression, was also found to be greater in response to hypoxia (Fig. 4A).

thumbnail image

Figure Figure 4.. Hypoxia-induced VEGF secretion correlates with MT1-MMP protein expression.MSCs were serum starved as described in the legend of Figure 3 for 24 or 48 hours in normoxic or hypoxic culture conditions. A) Cell lysates were monitored for MT1-MMP (using AB815 antibody; rMT1-MMP protein was isolated from transfected COS-7 cells [40]) and Egr-1 protein expression by Western blotting as described inMaterials and Methods. B) Conditioned media were assessed for secreted MMP-2 activity by gelatin zymography and for MMP-2, VEGF, and TIMP-2 protein levels by immunoblotting.

Download figure to PowerPoint

Many other genes involved in cellular differentiation can also be directly or indirectly regulated by hypoxia in MSCs. While hypoxic conditions resulted in a lower secretion of pro-MMP-2 into the extracellular media (Fig. 2 and Fig. 4B), that of its inhibitor, TIMP-2, was found to be unaffected (Fig. 4B). This, indeed, contrasted with that of VEGF, which was highly expressed and secreted by MSCs cultured under hypoxic conditions up to 48 hours (Fig. 4B). VEGF protein upregulation was correlated with a sustained gene expression of HIF-1α for up to 48 hours (not shown).

Growth Factor-Dependent Autocrine Regulation of MSC Migration

Besides VEGF, other growth factors could be secreted by MSCs under hypoxia. To investigate the effect of hypoxia-secreted growth factors on cell migration, conditioned media were isolated from serum-starved MSCs cultured under either normal or hypoxic conditions. These respective media would very likely contain a mixture of growth factors that may up- or downregulate the basal migratory potential of MSCs. When MSCs were allowed to migrate in modified Boyden chambers under normoxia for 4 hours, their migration was strongly induced by serum (Fig. 5). When cell migration was assessed in the presence of conditioned media isolated from serum-deprived MSCs added to the lower compartment of the Boyden chambers, migratory activity gradually enhanced from 24 hours to 48 hours. Interestingly, conditioned media isolated from serum-deprived MSCs cultured in hypoxic conditions induced cell migration to levels comparable with those obtained with serum. An approximately fourfold greater migration was seen at 24 hours, but similar levels were obtained when conditioned media were isolated at 48 hours in hypoxia and in normoxia (Fig. 5). Equivalent induction of MSC migration was also observed when conditioned media were isolated from serum-deprived U-87 glioma cells, a highly vascularized tumor-derived cell line, that had been cultured in normal and hypoxic conditions (not shown). Together, these results suggest that both autocrine and paracrine regulation of MSC migration may take place.

thumbnail image

Figure Figure 5.. Soluble growth factors are secreted in hypoxic conditions and regulate MSC migration.Conditioned media were isolated from serum-starved MSCs cultured in normal or hypoxic conditions. MSCs were trypsinized and seeded (105 cells) on gelatin-coated filters of modified Boyden chambers. Migration was allowed to proceed for 4 hours under normal conditions in the presence of either serum-free media (-), or serum-containing media (+). Conditioned media isolated at 24 or 48 hours from serum-starved MSCs that were cultured in normal or hypoxic conditions. Four different fields were counted. * indicates statistically significant differences (p < 0.05) compared with the control (serum-free conditioned media) using the Student's t-test.

Download figure to PowerPoint

Serum contains several polypeptide growth factors that regulate cell motility [27]. We thus decided to assess the effects of the two major angiogenic growth factors, VEGF and bFGF, or a combination of both growth factors, on MSC migration. While these two growth factors are known to induce differentiated endothelial cell migration [28], we observed that maximal VEGF concentrations (50 ng/ml) only induced a marginal increase in MSC migration, in agreement with previous reports where expression of the VEGF receptor Flk1 was very low [8]. In contrast, bFGF (10 ng/ml) alone (Fig. 6A) or in combination with VEGF (not shown), markedly induced MSC migration. Interestingly, however, VEGF- and bFGF-induced MSC migration was much lower than that induced by serum itself or by S1P (Fig. 6B), a bioactive lysophospholipid that is released from platelets, macrophages, and epithelial cells that triggers invasion of primitive hematopoietic cells [29], suggesting that other alternative growth factors or cytokines are needed. Furthermore, MSC migration was also assessed in normoxic and hypoxic conditions in the presence of a blocking antibody directed against the MT1-MMP catalytic domain. As shown in Figure 6C, MT1-MMP-dependent MSC migration was shown to be effectively antagonized by the blocking MT1-MMP antibody whether migration proceeded in a normal or low oxygen tension environment. This suggests that MT1-MMP can contribute to MSC migration.

thumbnail image

Figure Figure 6.. Modulation of MSC migration by growth factors.MSCs were trypsinized, seeded (105cells) on gelatin-coated filters, and allowed to migrate in normoxia for 4 hours at 37°C. The lower compartment of the Boyden chamber either contained serum-free media, serum-containing media, or the indicated growth factors (50 ng/ml VEGF, 10 ng/ml bFGF, 1 μM S1P) resuspended in serum-free media. A representative staining of the filters is shown in A, and a mean of three independent experiments is plotted in B. MSC migration on gelatin-coated filters was also assessed in the presence of 1 μg anti-mouse IgG (-) or an equivalent amount of an antibody directed against the MT1-MMP catalytic domain (+), left to proceed in normoxic or hypoxic conditions (C) as described inMaterials and Methods. * indicates statistically significant difference (p < 0.05) compared with control (serum-free conditioned media) in normoxic conditions using the Student's t-test. **indicates statistically significant differences (p < 0.05) compared with control (serum-free conditioned media) in hypoxic conditions using the Student's t-test.

Download figure to PowerPoint

MSC Capillary-Like Structure Formation is Mediated by Functional MT1-MMP

We herein report that MSCs spontaneously generated 3-D capillary-like structures, and that this capability was stimulated by hypoxia (Fig. 1). While MMP-2 and MT1-MMP expressions are regulated in MSCs by hypoxia, we decided to examine the extent of implication of these MMPs in the formation of such structures under basal and hypoxic conditions. MSCs were plated onto Matrigel and overlaying culture media were replaced by fresh serum-free media containing, respectively, anti-mouse IgG (control); anti-MT1-MMP mAb directed against the catalytic domain; anti-MT1-MMP polyclonal antibody raised against the noncatalytic hinge domain; 10 μM marimastat, a nonspecific MMP inhibitor reported to tightly bind the recombinant catalytic domain of MT1-MMP [30, 31]; or 25 μM epigallocatechin gallate (EGCg), a green tea catechin known to specifically affect MT1-MMP-driven cell migration and gene expression [32]. The results shown in Figure 7A clearly demonstrate that the ability of MSCs to form capillary-like structures was antagonized or strongly diminished when the MT1-MMP catalytic domain was inhibited either through the action of an antibody directed toward its functional extracellular catalytic domain or through the action of agents known to inhibit MMP hydrolytic activity, such as marimastat or EGCg. Quantification of the MSC capillary-like structure networking revealed a 75% decrease when MT1-MMP activity was inhibited or complexed by the antibody, and a 40% inhibition by marimastat, while it was almost completely inhibited by EGCg (Fig. 7B). The rapid formation of MSC capillary-like structures under hypoxic conditions (Fig. 1) was similarly antagonized with the anti-MT1-MMP catalytic antibody (not shown). In addition, we provide evidence that the catalytic function of MT1-MMP involved in the processing of latent pro-MMP-2 into its active form is also affected by the anti-MT1-MMP catalytic antibody in a cell-based system. Recombinant MT1-MMP was overexpressed in COS-7 cells, as previously reported [32], and was shown to convert an exogenous source of latent pro-MMP-2 into active MMP-2 (Fig. 7C). This reaction was specifically antagonized by marimastat, EGCg, and anti-MT1-MMP catalytic antibody, but not by EGC or anti-MT1-MMP antibody directed against the hinge region. These observations further validate the involvement of MT1-MMP catalytic activity in both pro-MMP-2 activation and capillary-like structure formation by MSCs. Inhibition of MSC-secreted soluble MMP-2 was not assessed for two main reasons: A) no specific anti-MMP-2 antibody raised against its catalytic domain was commercially available and B) the high levels of Matrigel-derived endogenous MMP-2 would have saturated the antibody and thus interfered with secreted MMP-2 attributable to MSCs [24]. Collectively, our observations thus indicate that the upregulation of MT1-MMP observed under hypoxia may be essential to elicit the formation of capillary-like structures and, consequently, to the angiogenic properties of MSCs.

thumbnail image

Figure Figure 7.. MSC capillary-like structure formation is mediated by functional MT1-MMP.MSCs were trypsinized, seeded (105cells) on Matrigel-coated filters, and allowed to adhere for 1 hour at 37°C. Overlaying culture media were then removed and replaced by fresh serum-free media containing 4.5 μg/ml of, respectively, anti-mouse IgG (control), anti-MT1-MMP mAb raised against the catalytic domain, anti-MT1-MMP polyclonal antibody raised against the hinge domain, 10 μM marimastat, or 25 μM EGCg. Cells were further incubated for 18 hours in normal conditions. A) Capillary-like structure formation was monitored, as described inMaterials and Methods, and B) quantified by measuring the total relative length of the capillary network. COS-7 cells were transfected with a cDNA encoding for MT1-MMP [32], and activation of an exogenous source of pro-MMP-2 was monitored in the presence or absence of the indicated agents (C). Mock (-) and MT1-MMP (+) transfected COS-7 cells. *indicates statistically significant differences (p < 0.05) compared with control (no test substance added) using the Student's t-test.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The recently reported phenotypic plasticity of MSCs into endothelial-like cells provides a rationale for their potential role in neovascularization and microvascular network remodeling [4, 5]. Moreover, we have recently shown that our in vitro cultured MSCs retained their progenitor cell potential in vivo. Specifically, when CD31 MSCs were either mixed with Matrigel then implanted s.c. [6, 7] or coinjected s.c. with U-87 glioblastoma cells in nude mice [unpublished data], they demonstrated, in both experimental approaches, the plasticity to differentiate into CD31+ endothelial cells [6, 7, 33]. Interestingly, these processes were recently shown to be modulated in response to tumor angiogenic factors [8]. Molecular and cellular understanding of MSC behavior thus needs to be promptly resolved before these cells can be used for safe and effective clinical application. In the present study, we focused on molecular events that regulate MSC migration and tube formation under hypoxia, which are essential events involved in angiogenesis. In this matter, we suggest that MSCs are able to participate in the production of vessels through their ability to migrate and generate capillary-like structures and that these angiogenic properties are modulated through growth-factor-mediated paracrine regulation.

Under hypoxic conditions, the major transcription factor affecting gene regulation is HIF-1 [33]. This transcription factor in turn upregulates expression of several genes involved in angiogenesis, such as bFGF, VEGF, and the VEGF receptors Flk-1 and Flt-1 as well as components of the plasminogen system including the urokinase receptor, and the inhibition of plasminogen activator 1 [33, 34]. Furthermore, other transcription factors activated by hypoxia interestingly include Egr-1, which regulates MT1-MMP transcription [35, 36]. In the present study, we blocked MSC ability to generate tube formation using an antibody directed against the extracellular MT1-MMP catalytic domain. This suggests that MT1-MMP is a crucial protease for MSCs in the process of tube formation. Indeed, this is supported by previous reports where the blocking of MMP activity in both in vitro and in vivo models resulted in inhibition of angiogenesis [37]. In particular, the MT-MMPs (MT-1, -2, -3, -4, -5, -6) are tethered to the plasma membrane [26], and such localization is known to enable pericellular proteolysis associated with cell migration, proliferation, and capillary tube formation required for neovascularization [38]. Interestingly, we and others have observed that MT1-MMP overexpression triggered cell migration [39] and was activated in endothelial cells where increased cell motility, matrix remodeling, and formation of capillary tubes were also reported [40]. Furthermore, MT1-MMP was recently shown to cleave CD44, one of the cell adhesion molecules expressed at the cell surface of MSCs, and to promote cell migration [41]. Of interest is the fact that the multifactorial agents used in the present study, such as marimastat and EGCg, have also been shown to modulate other molecular targets that potentially participate in tubulogenesis. Indeed, hydroxamate-based zinc MMP inhibitors, such as batimastat or marimastat, blocked the activity of members of the “a disintegrin and MMP-like” (ADAM) family [42], while EGCg not only directly inhibited MT1-MMP activity [43] and MT1-MMP-dependent pro-MMP-2 activation [32] but also specifically inhibited tubulogenesis in differentiated endothelial cells [44]. We thus propose that MT1-MMP regulation may be a key modulator in the proteolytic and cell migratory angiogenic responses of MSCs that are modulated by hypoxic conditions. More specifically, we also provide evidence for MSC ability to generate MT1-MMP-dependent capillary-like structures that may ultimately enable them to contribute to neovascularization. Future studies aimed at demonstrating the roles of other partners in ECM degradation (such as the TIMPs) will certainly give further insight into the mechanisms of MSC capillary-like structure formation on Matrigel. The use of differentiated endothelial cells isolated from MT1-MMP- and/or MMP-2-null mice eventually may be correlated and compared with the molecular events involved in tubulogenesis that were identified in MSCs.

While recently emerging views provide evidence that MT1-MMP and VEGF can be functionally linked in angiogenesis [45, 46], compelling data also provide evidence that low oxygen levels have the potential to optimize progenitor cell expansion and survival in culture [47], possibly through a hypoxia-dependent increase in VEGF expression [48, 49]. Interestingly, our results indicate that hypoxia-induced upregulation of MT1-MMP in MSCs correlates with a stimulation of VEGF production. This may translate in vivo to an increase in surrounding endogenous endothelial cell response and eventually to their subsequent recruitment and involvement in tumor vascularization and growth. In support of a role for VEGF in MSC differentiation and involvement in neovascularization is also the fact that human marrow-derived multipotent adult progenitor cells (MAPCs) depleted of hematopoietic cells expressed von Willebrand factor and markers of mature endothelium (CD31, CD34, CD36, and CD62-P) upon the presence of VEGF in long-term cultures [5, 8]. Remarkably, injection of undifferentiated MAPCs resulted in human endothelial cells vascularizing a spontaneous murine tumor, indicating that undifferentiated MAPCs can respond to local differentiation-inducing cues in vivo [8]. Experimentally, it was also shown, by the use of neutralizing antibodies, that bFGF released from intracellular stores of hypoxic endothelial cells was a crucial mediator of this effect [50].

Taken together, our results indicate that hypoxia strongly affects the regulatory pathways of MSCs, leading to the activation of several molecular and cellular events that result in the formation of conditions favorable for proliferative and ECM-remodeling activities. Moreover, these findings support the hypothesis that hypoxia-driven angiogenesis may be a critical condition for ECM formation and remodeling by MSCs that would be potentially necessary for successful applications using engineered MSCs to deliver therapeutic agents into sites of active or pathological angiogenesis. MSCs may thus play an important regulatory role in hypoxic microenvironments such as those found within tumor masses. Future studies involving immunohistochemical analysis of the regulation of MSC surface markers expressed during their differentiation process may offer greater insight into the specific mechanisms mediating and regulating MSC angiogenic properties.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

B.A. holds a CIHR-Canada Research Chair in Molecular Oncology. This work was supported by grants from Valorisation Recherche Québec to the groupe de Thérapie Expérimentale du Cancer and from the Fondation Charles-Bruneau.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References