SEARCH

SEARCH BY CITATION

Keywords:

  • HGF;
  • Mesenchymal stem cells;
  • Cell migration;
  • Mobilization

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Reverse Transcriptase Polymerase Chain Reaction (RT-PCR)
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References

Human mesenchymal stem cells (hMSC) are adult stem cells with multipotent capacities. The ability of mesenchymal stem cells to differentiate into many cell types, as well as their high ex vivo expansion potential, makes these cells an attractive therapeutic tool for cell transplantation and tissue engineering. hMSC are thought to contribute to tissue regeneration, but the signals governing their mobilization, diapedesis into the bloodstream, and migration into the target tissue are largely unknown. Here we report that hepatocyte growth factor (HGF) and the cognate receptor HGFR/c-met are expressed in hMSC, on both the RNA and the protein levels. The expression of HGF was downregulated by transforming growth factor beta. HGF stimulated chemotactic migration but not proliferation of hMSC. Therefore the HGF/c-met signaling system may have an important role in hMSC recruitment sites of tissue regeneration. The controlled regulation of HGF/c-met expression may be beneficial in tissue engineering and cell therapy employing hMSC.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Reverse Transcriptase Polymerase Chain Reaction (RT-PCR)
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References

A classic concept of tissue repair holds that inflammatory cells enter the damaged tissue and signal resident tissue-specific progenitor cells (e.g., parenchymal cells, fibroblasts) for mitosis. Several studies suggest that multipotent (bone marrow) mesenchymal stem cells can also contribute to tissue repair after mobilization, migration, and engraftment of the damaged tissue [1]. In addition, circulating immature cells seem to participate in regeneration of many different tissues [2, 3]. The role of mesenchymal stem cells in tissue remodeling was shown in different in vivo models—for example, for hepatic regeneration [4, 5], muscle regeneration [6, 7], and infarcted myocardium [8, 9]. Not surprisingly mesenchymal and circulating stem cells have attracted increasing attention because they hold great therapeutic potential for endogenous tissue repair and tissue engineering.

Since the original report from Friedenstein et al. [10], a number of different protocols have been defined to isolate multipotent adult mesenchymal stem cells from bone marrow specimen [11, 12]. Unlike hematopoietic stem cells, human mesenchymal stem cells (hMSC) adhere to cell culture plastic, which is exploited for their isolation [13, 14]. HMSC express CD105 and CD73 but not the lineage-specific surface antigens CD14, CD34, and CD45 [14]. Markers specific for hMSC are not known. Therefore, putative hMSC isolates have to be verified by their capacity to differentiate at least into adipocytes, chondrocytes, and osteoblasts. In addition, bone marrow-derived mesenchymal stem cells can be differentiated in vitro into bone marrow stromal cells and into endothelial, myogenic, hepatic, and neurogenic cells. Cell transplantation studies in human patients and in animals have demonstrated that bone marrow-derived cells can colonize most organs. The colonization was much enhanced by inflammation accompanying, for example, graft rejection or infarction. Hence, successful engraftment of several organs was enhanced by irradiation [15], chemical injury, and genetic diseases [5] or following infarction [8]. Despite major advances in MSC biology, our knowledge of the signals required for MSC mobilization and migration to the injured tissue site lags behind the extensive experience with hematopoetic stem cells, which are in routine clinical use. Therefore we sought to determine which factors may be responsible for mobilization of hMSC. Hepatocyte growth factor/scatter factor (HGF/SF) is a multipotent growth factor that exerts a mitogenic, motogenic, and morphogenic response on cells expressing c-met, the cellular HGF receptor. HGF/SF is essential in paracrine signaling of mesenchymal and epithelial cells, particularly during embryogenesis, repair, and carcinogenesis [16]. In malignant and transfected cells autocrine stimulation has been described [17]. In pathology, HGF/SF has been shown to induce tumor cell invasion in tissues [18].

Studies on HGF and c-met expression in bone marrow cells were reported by Takai et al. [19]. These authors demonstrated that bone marrow stromal cells constitutively express HGF and promote hematopoiesis. In addition, expression of c-met by stromal cells suggested an autocrine stimulation of stromal cells by HGF. However, it was not determined if HGF and c-met were also expressed by hMSC. This study was undertaken to address this question, including functional aspects of HGF and c-met-like cell migration and proliferation. We demonstrate that HGF and c-met are constitutively expressed by hMSC and that the expression of HGF is downregulated by transforming growth factor-β (TGF-β). Furthermore, HGF exerted a strong chemotactic stimulus on hMSC, which may be further enhanced by autocrine signaling through the HGF c-met pathway.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Reverse Transcriptase Polymerase Chain Reaction (RT-PCR)
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References

Cell Culture

HMSC were isolated mechanically from femoral heads of human orthopedic patients (with informed consent) and by adherence to cell culture plastic according to protocols of Haynesworth [13] and Pittenger [14]. Briefly, bone marrow was rinsed thoroughly with stem cell medium (MSCBM Stem Cell BulletKit; CellSystems; St. Katharinen, Germany; http://www.cellsystems.com) several times, then bone marrow was removed and cell suspension was mixed and centrifuged for 10 min (500 g). The cell pellet was resuspended in fresh medium and seeded in a T75 culture flask. Nonadherent cells were removed after 24 hours by medium change. Cells were cultured in stem cell medium at 37°C in a 5% CO2 and 20% O2 humidified atmosphere. Medium was changed every 3–4 days. At confluency, cells were trypsinized (0.05% trypsin, 0.53 mM EDTA, 5 min) with stem cell trypsin (CellSystems) and seeded at a density of 5,000 cells/cm2 to expand cell culture.

Characterization of hMSC

To differentiate hMSC into adipocytes, chondrocytes, and osteoblasts, protocols according to Pittenger et al. [14] were used. For adipogenic differentiation, cells were seeded at a density of 8 × 104 cells/cm2 into stem cell medium. At confluency, the medium was changed every 3–4 days from adipogenic induction to adipogenic maintenance medium and back. This regimen was repeated twice. The adipogenic induction medium consisted of Dulbecco's modified Eagle's medium (DMEM) high glucose (PAA Laboratories; Cölbe, Germany; http://www.paa.at) with 1 μM dexamethasone, 0.2 mM indomethacin, 0.01 mg/ml insulin, 0.5 mM 3-isobutyl-1-methyl-xanthine, and 10% fetal calf serum (FCS) (all Sigma; Steinheim, Germany; http://www.sigmaaldrich.com), while the adipogenic maintenance medium was composed of DMEM high glucose (PAA) with 0.01 mg/ml insulin and 10% FCS. After three complete cycles of induction and maintenance, cells were fixed with 10% formalin and stained for 10 min with Oil red O solution (Sigma) to visualize lipids.

For chondrogenic differentiation, 0.5-ml cell suspension (5 × 105 cells/ml) was placed into a 15-ml polypropylene tube (Becton Dickinson; Heidelberg, Germany; http://www.bd.com) and centrifuged to obtain cell pellets. The pellets were cultured in serum-free chondrocyte induction medium (DMEM high glucose; PAA), 100 nM dexamethasone, 0.17 mM l-ascorbic acid 2-phosphate, 100 μg/ml sodium pyruvate, 40 μg/ml proline (all Sigma), and 1% ITS-Plus (Becton Dickinson). TGF-β3 (CellSystems) was added in a concentration of 10 ng/ml medium at each medium change. After 21 days, pellets were formalin fixed and paraffin embedded. Thin sections were stained with Toluidin blue (Sigma).

For osteogenic differentiation, cells were seeded in a density of 3.1 × 104 cells/cm2 and stimulated with osteogenic induction medium after 24 hours. The medium consisted of DMEM low glucose (PAA), 10% FCS, 100 nM dexamethasone, 10 mM sodium β-glycerophosphate, and 0.05 mM L-ascorbic acid 2-phosphate (all Sigma) and was replaced every 3–4 days. After 16 days of differentiation, cells were fixed with 70% ethanol for 1 hour, washed with aqua bidest, and stained with Alizarin red solution (40 mM, pH 4.1, Sigma) for 10 min. Stained cells were washed three times with phosphate-buffered saline (PBS), and calcium deposits were photographed.

FACS Analysis

An aliquot of 2.5 × 105 cells was preincubated (4°C, 30 min) in PBS containing 1% bovine serum albumin (BSA). A 1-μg primary antibody was added to a 100-μ1 cell suspension (4°C, 60 min). Cells were washed three times in 100 μ1 PBS containing 0.1% BSA and resuspended in PBS containing 1% BSA. The washing step was followed by incubation with secondary antibody (1:20 dilution) in PBS/1% BSA (4°C, 45 min, in the dark). Cells were washed three times as described previously and resuspended in 100-μ1 PBS containing 1 μM propidium iodide. Monoclonal antibodies (CD4, CD14, CD34, CD49a, CD49c, CD49d, CD51, CD54, CD73, CD105, CD117) and corresponding secondary antibodies (conjugated with fluorescein isothiocyanate [FITC]) were purchased from DAKO (Glostrup, Denmark; http://www.dakocytomation.com). Species-matched immunoglobulin Gs (IgGs) served as negative control. Data of 10,000 stained cells were collected and analyzed using a FACSCalibur instrument and FACSCalibur software (Becton Dickinson).

Reverse Transcriptase Polymerase Chain Reaction (RT-PCR)

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Reverse Transcriptase Polymerase Chain Reaction (RT-PCR)
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References

Total RNA was extracted using guanidinium thiocyanate (RNeasy Kit; Qiagen; Hilden, Germany; http://www.qiagen.com). Reverse transcription was accomplished with 5 μg of total RNA using the “first-strand cDNA synthesis kit” (Amersham Pharmacia Biotech; Buckinghamshire, UK; http://www.amershambiosciences.com) with FPLCpure murine reverse transcriptase. PCR was carried out under the following conditions: denaturation at 95°C for 1 min, annealing at 54°C (HGF) or 60°C (c-met) for 1 min, extension at 72°C for 1 min (30 cycles), and a final extension at 72°C for 10 min. PCR amplicon size (266 bp for HGF and 440 bp for c-met) was analyzed by electrophoresis on a 2% agarose gel and visualized with ethidium bromide. The oligonucletoide primer and the GenBank/EMBL identifiers of template sequences were as follows:

  • c-met (NM000245)

    • forward (nt 1398–1423 [20]): 5′-AGAAATTCATCA GGCTGTGAAGCGCG-3′

    • reverse (nt 1814–1838 [20]): 5′-TTCCTCCGATCG CACACATTTGTCG-3′

  • HGF (XM168542)

    • forward (nt 548–568): 5′-GGTAAAGGACGCAGC TACAAG-3′

    • reverse (nt 794–814): 5′-ATAACTCTCCCCATTGC AGGT-3′

Immunohistochemistry

Subconfluent hMSCs were grown in chamber slides and fixed for 30 min using 0.5% paraformaldehyde (in PBS). After washing with wash buffer (Dulbecco's phosphate-buffered saline solution A [PBSSA], 0.5% BSA in PBS), unspecific protein-binding capacity was blocked for 15 min using blocking buffer (PBS with 5% BSA). Cells were incubated for 45 min at RT with the first antibody, diluted 1:50 in PBSSA-NP-40 (1:200 dilution of nonidet P-40 [NP-40] in PBSSA), washed three times for 10 min in PBSSA, followed by an incubation with the secondary antibody (1:250 in PBSSA) for 30 min at RT and again three washing steps. FITC-conjugated streptavidin was added (1:250 in PBSSA) for 30 min, protected from light. Finally, cells were washed three times with PBSSA for 10 min and mounted in 4′,6′-diamidino-2-phenylindole hydrochloride (DAPI)-containing mounting medium.

Polyclonal rabbit c-met antibody (primary antibody) was purchased from Santa Cruz Biotechnology (Santa Cruz, CA; http://www.scbt.com), and secondary antibody (biotin-conjugated goat anti-rabbit) and FITC-conjugated streptavidin from DAKO. As a negative control, cells were incubated with the first antibody, which had been preincubated overnight at 4°C with a specific blocking peptide (Santa Cruz Biotechnology).

Western Blotting

Mesenchymal stem cells were lysed with insect cell lysis buffer (PharMingen International; Hamburg, Germany; http://www.pharmingen.com), and protein concentration was determined using the BCA-Kit (Pierce Biotechnology; Rockford, IL; http://www.piercenet.com). Ten μg of protein lysate and protein marker (Perfect Protein AP WB Marker; Novagen; Darmstadt, Germany; http://www.emdbiosciences.com) were separated on 4%-12% SDS gels (NuPAGE; Karlsruhe, Germany) at 70 V for 2.5 hours. Membranes were fixed with methanol, and after electroblotting (60 min, 150 mA) on polyvinylidene fluoride membrane (Bio-Rad; Munich, Germany; http://www.bio-rad.com), the membranes were blocked overnight with low-fat milk at 4°C. Immunostaining was accomplished by incubation with rabbit polyclonal antibody against c-met (1:200, Santa Cruz Biotechnology) for 90 min at room temperature. After 1 hour of incubation with alkaline phosphatase (AP)-conjugated secondary antibody (1:5,000; Roche; Mannheim, Germany; http://www.roche.com) at room temperature, staining was developed using Sigma Fast 5-bromo,4-chloro,3-indoyl phosphate/nitroblue tetrazolium (BCIP/NBT) tablets (Sigma). As a negative control, the first antibody was preincubated with a fivefold excess of blocking peptide in a small volume of PBS at 4°C overnight.

HGF-ELISA

HGF concentration in hMSC-conditioned media was measured by Quantikine human HGF-ELISA (R&D Systems; Minneapolis, MN; http://www.rndsystems.com), which is based on a sandwich enzyme immunoassay technique with a precoated HGF-specific antibody. To this end, cells were seeded in 96-well plates (20,000 cells/well) and stimulated for 24 hours with TGF-β3 (10, 1, and 0.1 ng/ml), interleukin-1β (IL-1β) (10, 1, and 0.1 ng/ml), and bFGF (10, 1, and 0.1 ng/ml). Unstimulated cells served as a control. Supernatants were harvested and stored at −70°C until they were analyzed.

Scratch Assay

hMSC were grown to confluency in a six-well plate (Becton Dickinson). A scratch in the cell layer was made with a pipette tip over the total diameter of 34.5 mm. HGF was added at 0, 25, 50, and 75 ng HGF/ml medium. Closure of this “wound” was documented photographically (Axiovert 25; Zeiss; Cologne, Germany; http://www.zeiss.com) after 24 hours, and cells in four segments of the scratched area, each of 320 μm × 320 μm, were counted.

Boyden-Chamber Assay

For analysis of cell motility, 1 × 105 hMSCs/ml were seeded in the top compartment of a Boyden chamber (NeuroProbe; Gaithersburg, UK; http://www.neuroprobe.com). The bottom compartment contained different HGF concentrations and was separated from the top compartment by a polycarbonate membrane with 8 μm pores (Corning; Düsseldorf, Germany; http://www.corning.com). Cells were allowed to migrate for 16 hours at 37°C in a humidified atmosphere. After removing cells from the upper side of the membrane with cotton swabs, membranes were fixed, stained with hematoxylin (Merck; Darmstadt, Germany; http://www.merck.com), and transferred onto glass slides. Cells on the bottom side of the membrane were counted in five different highpower fields. Each analysis was performed in triplicate.

Cell Proliferation Assay

Cells were seeded in 24-well plates (3,000 cells per well) and stimulated for 24 hours with HGF (0, 25, and 50 ng HGF/ml medium) in low serum (2% FCS). The medium containing 20% FCS was used as positive control. Proliferation was measured by detection of ATP content of the cells with a luciferase detection system (ViaLight HS; BioWhittaker; Verviers, Belgium; http://www.biowhittaker.be). This bioluminescent method uses luciferase, which catalyzes the formation of light from ATP and luciferin. The emitted light intensity is linearly related to the ATP concentration [21]. ATP content was measured 1, 2, 3, 4, and 7 days after stimulation. The medium with HGF was renewed after day 4.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Reverse Transcriptase Polymerase Chain Reaction (RT-PCR)
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References

Isolation and Characterization of hMSC

Isolated MSC appeared as two distinct morphological phenotypes—long spindle-shaped cells (Fig. 1A) and large, flat cells (Fig. 1B). This mixed appearance is typical of hMSC [22]. As shown by flourescence-activated cell sorter (FACS) analysis (Fig. 1C), isolated cells did not express the hematopoietic markers CD4 (a), CD14 (b), and CD34 (c); the integrins CD49a (d) and CD49d (e); or the stem cell factor receptor CD117 (c-kit, f). In contrast, integrin CD49c (g), CD51 (vitronectin receptor, h), and CD54 (intercellular adhesion molecule 1, i), as well as CD105 (Endoglin, k) and CD73 (1), were readily detected on the cell surface, confirming the results of Pittenger et al. [14].

thumbnail image

Figure Figure 1.. Characterization of hMSC and demonstration of pluripotency.A) hMSC, spindle-shaped morphology. B) hMSC, flat morphology. C) FACS analysis of hMSC: a, CD4; b, CD14; c, CD34; d, CD49a; e, CD49d; f, CD117; g, CD49c; h, CD51; i, CD54; k, CD105;1, CD73. D) Adipogenic differentiation of hMSC: stem cells were cultured for 21 days alternately in adipogenic induction and adipogenic maintenance medium. Lipid vacuoles were visualized by Oil red O staining. E) Chondrogenic differentiation of hMSC: pellet culture of hMSC after 21 days in chondrogenic induction medium. Proteoglycans were stained with Toluidine blue. F) Osteogenic differentiation of hMSC: stem cells were cultured for 16 days with osteogenic induction medium, and calcium deposits were visualized by alizarin red staining.

Download figure to PowerPoint

Next we induced cell differentiation of our isolated MSC, following published protocols [14]. Adipogenic differentiation resulted in lipid vacuole formation (Fig. 1D), chondrogenic differentiation led to the production of proteoglycans in the extracellular matrix (Fig. 1E), and osteogenic differentiation caused the production of Alizarin Red-positive mineral deposits (Fig. 1F). Thus we have shown that the isolated cell population can indeed be differentiated in vitro into adipocytes, chondrocytes, and osteoblasts. Phenotypical and functional characterization was repeated with each isolate and showed virtually identical results.

hMSC Express HGF and c-met on RNA and Protein Level

We analyzed HGF- and c-met-specific RNA transcripts in hMSC of three different donors at several times of culture between passages one and six. Figure 2 shows that both HGF and c-met were constitutively expressed in cultured hMSC. To detect c-met on the protein level, we employed immunoblotting and detected a 145-kDa band representing the c-met protein (Fig. 3). Upon preincubation of the antibody with a blocking peptide, the band at 145 kDa was much reduced in intensity, demonstrating the specificity of the antibody detection. c-met expression was confirmed by immunohistochemistry. Figure 4 illustrates that hMSC stained positive for c-met. Fluorescence labeling of cells appeared uniform without apparent receptor clustering in fluorescent “hot spots” (Fig. 4A, 4B). Consistently, more than 95% of hMSC stained positive, and this positive staining could be abrogated in all cases by preincubation of the c-met antibody with a blocking peptide, indicating that the antibody indeed stained the cellular HGF receptor c-met.

thumbnail image

Figure Figure 2.. Human MSC expresses HGF and its receptor c-met on the RNA-level.Amplicon of HGF gene (266 bp) and c-met gene (400 bp); M, molecular weight marker protein, 100-bp ladder: 1) negative control (without cDNA); 2) hMSC, HGF; 3) liver cells (positive control for HGF); 4) hMSC, c-met; 5) mesothelial cells (positive control for c-met). This figure shows one representative result out of four.

Download figure to PowerPoint

thumbnail image

Figure Figure 3.. Detection of c-met in hMSC on the protein level.Immunoblot analysis of hMSC extracts probed with an antibody directed against the HGF receptor/c-met detected a protein band at 145 kDa, which is characteristic of this receptor (lane 1, arrow). Preincubation of the antibody with a blocking peptide abolished binding to this protein (lane 2). M, molecular weight marker protein. This figure shows one representative result out of four.

Download figure to PowerPoint

thumbnail image

Figure Figure 4.. HGF receptor c-met is detectable on the cell surface.hMSC were incubated with antibody directed against the HGF receptor/c-met (A, B) or with the same antibody preincubated with a blocking peptide (C, D) followed by a secondary FITC-labeled antibody. Consistently, over 95% of all cells stained positive for c-met, albeit at variable levels. This figure shows typical views from one out of four experiments with one brightly staining cell (B) and one moderately bright yet distinctly positive staining cell (A). Cell nuclei were counterstained with DAPI.

Download figure to PowerPoint

Next we analyzed HGF protein amounts in hMSC culture supernatants using enzyme-linked immunosorbent assay. We determined that the amount of HGF present in the medium after a 24-hour culture period of 2 × 104 cells ranged from 770 to 2,000 pg/ml. The amount varied slightly between cell donors. Figure 5 illustrates the combined results obtained with MSC from three different donors. Compared to unstimulated cells, stimulation with TGF-β3 at 0.1, 1.0, and 10 ng/ml caused a downregulation of HGF expression by 33% ± 7%, 54% ± 17%, and 50% ± 19%, respectively. The downregulation of HGF expression was statistically significant in all cases, when compared to the untreated control (p < 0.05), indicating that hMSCs are sensitive to TGF-β3 and that moderate amounts are sufficient to mediate downregulation of HGF in these cells. IL-1β at 0.1, 1, and 10 ng/ml and bFGF at 0.1 and 1 ng/ml did not change HGF expression. bFGF at 10 ng/ml enhanced HGF expression to 180% ± 51% (p = 0.11).

thumbnail image

Figure Figure 5.. Cytokine dependency of HGF production in hMSC.hMSC were stimulated with TGF-β3, IL-1β, and bFGF for 24 hours. HGF expression by hMSC was measured. This figure shows mean values of three different donors, each in duplicate. *p < 0.05

Download figure to PowerPoint

HGF Enhances Cell Migration of Human Mesenchymal Stem Cells

To study potential biological consequences of c-met signaling in hMSC, we analyzed cell migration using a well-established wounding model of a confluent cell monolayer with and without HGF stimulation. In this model a confluent cell layer is disrupted by scraping with a cell scraper of defined width, and the subsequent closure of this “wound” by migrating cells is studied histomorphometrically over time. Figure 6 illustrates that the “wound” afflicted to a confluent layer of about 150 hMSC/mm2 closed faster when HGF was added to the culture medium; this effect was concentration-dependent. At 0 ng/ml added HGF (Fig. 6A), hardly any (1.07 cells/mm2) MSC had migrated into the wounded, cell-free area 24 hours after wounding. Stimulation with 25 and 50 ng/ml HGF enhanced hMSC migrating into the wound to 1.6 and 3.2 cells/mm2, respectively (Fig. 6B, 6C). The addition of 75 ng/ml HGF to the medium further enhanced cell migration to 8.8 cells/mm2, corresponding to an 8-fold increase in cell migration into the wound (Fig. 6D).

thumbnail image

Figure Figure 6.. HGF stimulates migration of hMSC.Confluent cell monolayers of hMSC were wounded by scratching with a pipette tip. After 24 hours of stimulation with HGF at 0 ng/ml cell culture medium (A), 25 ng/ml (B), 50 ng/ml (C), and 75 ng/ml (D), wound closure was documented photographically. This figure shows one representative result out of five.

Download figure to PowerPoint

HGF Is Chemotactic for hMSC

To investigate whether the observed promigratory effect was chemotactic or chemokinetic, we performed Boyden chamber assays. The results of the Boyden chamber assays are shown in Figure 7. HGF at 37.5 ng/ml and 50 ng/ml added to the bottom compartment of the Boyden chamber caused a more than threefold increase of migrating cells (p < 0.005 and p < 0.001, respectively). When identical amounts of HGF were simultaneously added to both the top and the bottom compartments, cell migration of hMSC remained unchanged. Collectively, these results demonstrate that HGF is strongly chemotactic for hMSC.

thumbnail image

Figure Figure 7.. HGF is chemotactic for hMSC.Stem cells were seeded in the top compartment of a Boyden chamber, and different amounts of HGF were added to the top and bottom compartments (ng/ml medium). The cells were allowed to migrate for 16 hours at 37°C. The top and bottom compartments were separated by a polycarbonate membrane with 8-μm pores. Data from one experiment representative of three independent experiments are shown. *p < 0.005, **p ≤ 0.001.

Download figure to PowerPoint

HGF Inhibits Mesenchymal Stem Cell Proliferation

To examine the possibility that the observed promigratory effects of HGF detected in the Boyden chamber and wounding assay were also associated with enhanced proliferation, we analyzed the proliferation of hMSC with and without added HGF. Figure 8 illustrates the amount of ATP measured in hMSC cultures as a surrogate marker for the number of vital cells present in the cells over a period of 7 days. The viability of HGF-stimulated and -unstimulated cells was assessed by trypan blue exclusion and was similar in all cells over the period of 7 days (data not shown). Therefore, changes in ATP content are a measure of proliferation and not of cell death. Compared to the low serum culture, serum stimulated ATP production and hence proliferation by approximately 160% ± 32% during the 7-day incubation period. In contrast, the addition of HGF at 25 ng/ml and 50 ng/ml did not stimulate cell proliferation but inhibited cell proliferation in a dose-dependent fashion. Compared to the low serum culture, 25 ng/ml added HGF decreased proliferation to 72% ± 21%, and 50 ng/ml HGF decreased proliferation to 65% ± 18% of the low serum control incubation. In summary, these results suggest that HGF/c-met signaling may be involved in control of hMSC mobilization by shutting off proliferation and augmenting cell migration. HGF may represent an important cue to attract migratory hMSC to their site of final differentiation, which could be driven by local factors of the respective target organ.

thumbnail image

Figure Figure 8.. HGF inhibits cell proliferation in hMSC.ATP content of hMSC was measured as a surrogate marker of cell proliferation at 1, 2, 3, 4, and 7 days after continuous stimulation with HGF at 0, 25, and 50 ng/ml. The medium containing 20% FCS served as a positive control, while the stimulation medium was supplemented with only 2% FCS. This figure shows mean values of three different donors, each in duplicate.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Reverse Transcriptase Polymerase Chain Reaction (RT-PCR)
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References

Adult pluripotent stem cells are thought to reside in many tissues. They can be isolated from skin [23] and adipose tissue [24], and a baboon study has shown that infused stem cells can colonize a wide range of tissues—for example, gastrointestinal, kidney, lung, liver, thymus, and skin [25]. However, the amount of stem cells in most tissues is exceedingly low, which leads to the hypothesis that injured tissues produce appropriate cues for engraftment and, furthermore, that the local microenvironment would stimulate the differentiation of engrafted (stem) cells into functional, specialized cells [15]. This begs the question of which signals are necessary and sufficient for stem cell mobilization and recruitment. HGF is well known to augment cell migration, scattering (i.e., “scatter factor”), and proliferation of many different cell types, predominantly of epithelial cells. Furthermore, HGF is considered to be a humoral mediator of organogenesis and morphogenesis of various tissues and organs, as well as regeneration of organs, tumor invasion, and metastasis [26].

In the bone marrow, HGF is known as an important hematopoietic regulator [27, 28]. Synthesis and release of HGF precursors could be part of an autocrine mechanism. Pre-HGF has to be activated by proteolytic cleavage—for example, through serine proteases like uPA and tPA, which are also released by hMSC (Neuss et al., unpublished observations). Tissue repair and wound healing are regulated by soluble mediators provided by inflammatory cells. Therefore, we investigated the effect of bFGF, IL-1, and TGF-β on the production of HGF. TGF-β, which is known to be growth inhibitory for epithelial cells, downregulates the production of HGF by hMSC. This effect is not attributed to cell death after TGF-β stimulation (data not shown). In contrast, basic fibroblast growth factor, which stimulates blood vessel formation and angiogenesis [29], and the proinflammatory cytokine IL-1β have no significant influence on the expression of HGF (determined by ELISA). Nevertheless, bFGF upregulates the production of the pre-HGF-activating serine protease tPA in hMSC (data not shown). Thus, production of mature HGF is regulated by the microenvironment, depending on processes like tissue damage or inflammation. Activated monocytes or macrophages, which are replete at damaged tissue sites, are known to produce HGF/c-met [30]. On a more general note, HGF concentration is increased at sites of tissue damage [3133]. After partial hepatectomy in the rat, HGF levels are even measurably elevated in the blood [34]. Taken together, these increases in local and systemic HGF could provide a key chemotactic signal to mobilize and attract hMSC for tissue repair.

Here we identified HGF/SF as a potent regulator of hMSC function, regulating migration and proliferation in vitro. Our study demonstrates the production of HGF and the expression of the HGF receptor c-met by a defined stromal cell population, human mesenchymal stem cells. These cells were shown by FACS analysis to lack any lineage-specific surface marker expression, including CD4, CD14, CD34, and CD117. In addition, the cells were capable of differentiation into adipocytes, chondrocytes, and osteoblasts. Accordingly, these cells represent the bone marrow stromal cell population described by the groups of Pittenger [14] and Haynesworth [13] as MSC. Migration of hMSC was increased by HGF in Boyden chamber assays (chemotactic effect), whereas proliferation of hMSC was negatively influenced. Furthermore, HGF promoted the repopulation of a cell-free “wound” in a cell monolayer wounding model. We attribute this repopulation mainly to the exogenously added HGF, since hMSC did not upregulate HGF production after the cell monolayer was wounded (data not shown).

Expression of HGF and c-met was also reported by Takai [19] in a bone marrow stromal cell population, but this population was not further characterized. Bone marrow stromal cells are widely used as feeder layers for long-term cultures of hematopoietic stem cells. Due to different isolation and culture procedures, these cells are less homogenous than the MSC used here. It is possible, however, that hMSCs are also contained in bone marrow stromal cells and that, in fact, hMSCs are the source of HGF required for hematopoiesis in the coculture system. This conclusion is supported by the fact that our hMSC produced similar amounts of HGF as did the feeder cells described by Takai et al. [19] and, furthermore, that neither the bone marrow stromal cells nor our hMSC showed increased proliferation in response to HGF. The amount of HGF expressed by hMSC (0.7–2 ng/ml) is higher than the normal HGF serum concentration (0.24–0.33 ng/ml [35]). As mentioned, HGF concentration can be elevated in the blood after wounding occurs [34]. We suggest that increased systemic HGF may mobilize bone marrow or tissue resident hMSC to colonize damaged target organs. According to our in vitro results, the elevated HGF should also inhibit hMSC proliferation. We speculate that the local environment will ultimately determine hMSC differentiation into mature tissue cells, as is the case in embryonic development.

HGF could possibly help to mobilize autologous hMSC and to direct possible allogeneic hMSC in future stem cell therapy. The role of HGF and its possible role in dysregulation of wound healing (like that described for TGF-β and other growth factors in diabetic mice, for instance) [36, 37] have to be further clarified. The observation that ischemic or traumatic rat brain extracts induced production of HGF in hMSC [37, 38] also supports the idea of HGF as an important factor for tissue repair by (transplanted or autologous) hMSC. The notion that HGF/c-met signaling may be involved in hMSC mobilization and recruitment to damaged tissues is entirely compatible with previous reports that this important regulator of cell motion and differentiation is critically involved in normal development of epithelial tissues [39], as well as the metastatic spread of tumor cells [40].

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Reverse Transcriptase Polymerase Chain Reaction (RT-PCR)
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References

We thank the Department of Orthopedic Surgery, University Hospital Aachen, and their patients for providing bone spongiosa.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Reverse Transcriptase Polymerase Chain Reaction (RT-PCR)
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References