Stem Cells http://www.peprotech.com/
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

First published online January 18, 2007
Stem Cells Vol. 25 No. 5 May 2007, pp. 1241 -1251
doi:10.1634/stemcells.2006-0320; www.StemCells.com
© 2007 AlphaMed Press

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
2006-0320v1
25/5/1241    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Reprints/Permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Fan, V. H.
Right arrow Articles by Griffith, L. G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Fan, V. H.
Right arrow Articles by Griffith, L. G.

TISSUE-SPECIFIC STEM CELLS

Tethered Epidermal Growth Factor Provides a Survival Advantage to Mesenchymal Stem Cells

Vivian H. Fana,b, Ada Aua, Kenichi Tamamac, Romie Littrella, Llewellyn B. Richardsond, John W. Wrighta, Alan Wellsc, Linda G. Griffitha,e

Departments of aBiological Engineering,
dChemical Engineering, and
eMechanical Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts, USA;
bHarvard School of Dental Medicine, Boston, Massachusetts, USA;
cDepartment of Pathology, University of Pittsburgh, Pittsburgh, Pennsylvania, USA

Key Words. Mesenchymal stem cells • Epidermal growth factor • Extracellular signal-regulated protein kinase • Fas ligand • Cell death Cell spreading • Bone graft

Correspondence: Linda G. Griffith, Ph.D., MIT 16-429, 77 Massachusetts Avenue, Cambridge, Massachusetts 02139, USA. Telephone: 617-253-0013; Fax: 617-253-2400; e-mail: griff{at}mit.edu

Received May 26, 2006; accepted for publication January 9, 2007.
First published online in STEM CELLS EXPRESS   January 18, 2007.


   ABSTRACT
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 Disclosure of Potential...
 Acknowledgments
 References
 
MSC can act as a pluripotent source of reparative cells during injury and therefore have great potential in regenerative medicine and tissue engineering. However, the response of MSC to many growth factors and cytokines is unknown. Many envisioned applications of MSC, such as treating large defects in bone, involve in vivo implantation of MSC attached to a scaffold, a process that creates an acute inflammatory environment that may be hostile to MSC survival. Here, we investigated cellular responses of MSC on a biomaterial surface covalently modified with epidermal growth factor (EGF). We found that surface-tethered EGF promotes both cell spreading and survival more strongly than saturating concentrations of soluble EGF. By sustaining mitogen-activated protein kinase kinase-extracellular-regulated kinase signaling, tethered EGF increases the contact of MSC with an otherwise moderately adhesive synthetic polymer and confers resistance to cell death induced by the proinflammatory cytokine, Fas ligand. We concluded that tethered EGF may offer a protective advantage to MSC in vivo during acute inflammatory reactions to tissue engineering scaffolds. The tethered EGF-modified polymers described here could be used together with structural materials to construct MSC scaffolds for the treatment of hard-tissue lesions, such as large bony defects.

Disclosure of potential conflicts of interest is found at the end of this article.


   INTRODUCTION
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 Disclosure of Potential...
 Acknowledgments
 References
 
MSC were first described 40 years ago as bone-forming progenitors isolated from marrow [1]. These cells replicate as pluripotent cells and can differentiate into multiple lineages of connective tissue. In adults, MSC play a crucial role as a possible reservoir for reparative cells during injury and disease [2, 3]. Much attention has been devoted clinically to MSC because of their potential in diverse applications, including regeneration of bone [4], cardiac tissue [5, 6], and treatment of graft-versus-host disease [7]. MSC harvested from marrow aspirates can be expanded and induced into osteogenic, adipogenic, and chondrogenic lineages in vitro [2, 3, 8], although challenges remain in retaining differentiation and in vivo tissue formation potential following extensive expansion [9]. Current therapeutic trials include local and systemic transplantation of MSC [5–7]. For bony defects, certain types of local delivery involve the implantation of a biomaterial scaffold preloaded with patient-specific MSC [10–12]. Despite recent clinical and scientific advances in stem-cell research, there is limited understanding of how survival and function of MSC are governed by interactions with scaffolds used for MSC delivery. Furthermore, it is unclear how MSC fate may be influenced by cytokines present during the inflammatory response to biomaterials used for implantation, and to injury itself.

The use of progrowth and prosurvival cytokines as adjuvants to MSC application may improve implant outcomes. Epidermal growth factor (EGF) is a well characterized cytokine involved in the growth and repair of various tissues. Upon binding to the EGF receptor (EGFR), EGF activates intracellular signals via the extracellular-regulated kinase (ERK) and Akt pathways, among others [13, 14]. These signals promote migration, adhesion, proliferation, and survival in various cell types [13, 15, 16]. Both EGFR expression and EGF responsiveness have been reported in marrow-derived MSC [17–21]. We recently reported that stimulation of MSC with soluble EGF induces EGFR signaling and promotes MSC proliferation and migration without affecting pluripotency [18]. The lack of clinical wound healing products based on EGF underscores the challenges in delivering EGF at physiologically relevant concentrations and durations locally in a wound site. These challenges might possibly be overcome by tethering EGF to the scaffold surface in a manner that allows EGF to stimulate the EGFR but that inhibits internalization [22], that is, a manner akin to a matrix-embedded EGFR ligand [23]. We hypothesized, however, that MSC could respond differently to tethered EGF compared with soluble EGF as a result of binding the ligand on a surface rather than in solution and thus preferentially activate surface-associated signaling pathways.

In addition to prosurvival growth factors, multiple prodeath ligands are upregulated during inflammation of bone tissue, including Fas ligand (FasL), tumor necrosis factor (TNF), TNF-related apoptosis-inducing ligand (TRAIL), interferon {gamma} (IFN-{gamma}), and interleukin 1 (IL-1) [24–28]. For instance, the proinflammatory cytokine FasL is released by activated lymphocytes [29], which causes neutrophil infiltration [30] and further inflammation. Osteoclasts are critical for bone remodeling and express both Fas and FasL [31], suggesting the relevance of FasL in bone tissue. In many cell types, FasL induces apoptosis by activation of caspases [32]. The apoptotic and inflammatory responses of MSC to ligands such as FasL, particularly under conditions that might prevail at an implant site, are as yet poorly understood. However, since net changes in cell number result from the balance between proliferation and cell death, it will be critical to favor the survival of MSC at an implant site. EGFR signaling has been shown to counter the prodeath signaling of various apoptotic cytokines to promote cell survival in many cell types [33, 34].

Here, we show that surface-tethered EGF improved MSC survival upon stimulation with a prodeath stimulus. We used a polymer substrate developed to present clusters of closely spaced ligands on short poly(ethylene oxide) (PEO) tethers [35] to covalently tether EGF via the N terminus, so that EGF is bioactive but restricted to the material surface. Surprisingly, we found that tethered EGF exceeded the ability of soluble EGF to promote cell spreading and survival in the presence of FasL, which we showed to be a potent death factor for human MSC. This study suggests that MSC-implant interactions can be improved by providing local, spatially controlled growth factors to the biomaterial surface.


   MATERIALS AND METHODS
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 Disclosure of Potential...
 Acknowledgments
 References
 
Growth Factors, Antibodies, and Signaling Reagents
Murine EGF and human EGF were purchased from Peprotech (Rocky Hill, NJ, http://www.peprotech.com), and porcine EGF was purchased from GroPep Limited (Thebarton, Australia, http://www.gropep.com.au). EGFR inhibitors AG1478 and PD153035 and mitogen-activated protein kinase kinase (MEK) inhibitor U0126 were from Calbiochem (San Diego, http://www.emdbiosciences.com). The EGFR ligand binding site-blocking antibody C225 was a gift from H.S. Wiley (Pacific Northwest National Laboratory, Richland, WA). Rabbit polyclonal anti-phospho-p44/42 Map kinase, rabbit polyclonal anti-EGF receptor, and rabbit polyclonal anti-phospho-Akt were purchased from Cell Signaling Technology (Beverly, MA, http://www.cellsignal.com); rabbit polyclonal anti-MAP kinase/ERK1/2-CT was from Upstate (Charlottesville, VA, http://www.upstate.com). Mouse monoclonal anti-{alpha}-tubulin was from Calbiochem. Donkey anti-rabbit and sheep anti-mouse IgG peroxidase-linked secondary antibodies were from Amersham Biosciences (Piscataway, NJ, http://www.amersham.com). Western Lightning Chemiluminescence Reagent Plus (PerkinElmer Life and Analytical Sciences, Boston, http://www.perkinelmer.com) or ECL Advanced Western Blotting Detection Kit (Amersham Biosciences) was used to visualize the protein bands with Kodak ISO1000 Image Station (Rochester, NY, http://www.kodak.com). Densitometry was performed with Kodak software.

Cell Culture
Human telomerase reverse transcriptase (hTERT)-immortalized human MSC (hTMSC) were a gift from Dr. Junya Toguchida (Kyoto University, Kyoto, Japan) [36] and were maintained in a standard medium formulation containing Dulbecco's modified Eagle's medium, 10% fetal bovine serum (FBS), 1 mM pyruvate, 1 mM L-glutamine, 1 µM nonessential amino acids, and 100 units/ml penicillin-streptomycin (Invitrogen, Carlsbad, CA, http://www.invitrogen.com), at 37°C in a humidified atmosphere of 5% CO2/95% air. Primary porcine MSC (ppMSC), a gift from Dr. J.P. Vacanti (Massachusetts General Hospital, Boston, MA), were isolated from fresh marrow aspirates using a Percoll gradient centrifugation approach, seeded at an initial density of 1.6 x 106 cells per cm2 in the culture medium described above, and split after reaching confluence [37]. They were used within five passages. Human primary connective tissue progenitors (hCTPs) were obtained as culture-adherent cells from heparinized aspirated human marrow provided by Dr. G. M. Muschler (Cleveland Clinic Foundation, Cleveland, OH) [38]. Cells were obtained and maintained in an osteogenic medium, minimal essential medium-{alpha} buffered with bicarbonate, 10% FBS, 1% penicillin-streptomycin, 10–8 M dexamethasone (Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com), and 50 µg/ml ascorbate (Sigma-Aldrich) at 37°C in 5% CO2/95% air. Cells were cultured for approximately 10 days (~85% confluent within the colonies) before assays. Poietics primary human MSC (phMSC) were purchased from Cambrex (Walkersville, MD, http://www.cambrex.com) and maintained according to the supplier's instructions. phMSC were cultured in MSC growth medium containing MSC growth supplement, L-glutamine, and penicillin-streptomycin (Cambrex). Growth media were changed every 2–3 days, and the cells were split at 5,000–6,000 cells per cm2 every 5–6 days. Cells up to the fifth passage were used in this study.

hCTP Selection
Bone marrow aspirates from human subjects were collected at the Muschler laboratory [38], heparinized and suspended in the culture medium described above, and shipped in sealed 50-ml conical tubes according to Institutional Animal Care and Use Committee- and Institutional Review Board-approved protocols at their respective institutions. Upon arrival, the cells were spun down, resuspended in the culture medium (described above), plated at 2.5 x 105 nucleated cells per cm2 (i.e., 106 nucleated cells per well of a 12-well tissue culture plate), and incubated at 37°C in 5% CO2/95% air. This corresponds to 43 ± 3 alkaline-phosphatase-positive colonies per 106 nucleated cells [39]. A medium change was performed on every third day until the colonies reached sufficient confluence for cytokine stimulation.

Polymeric Substrate Preparation
Two different poly(methyl methacrylate)-graft-poly(ethylene oxide) (PMMA-g-PEO) comb polymers differing in the total weight percent (wt%) PEO (and consequently, the spacing between PEO side chains emanating from the culture surface) were synthesized using general protocols described previously [35] and were mixed to form the culture substrate. The first polymer was designed for presentation of tethered EGF in a locally dense concentration suitable for allowing receptor homodimerization (tethered EGF [tEGF]-polymer) [22]. This tEGF-polymer has a weight-average molecular weight and polydispersity index (PDI) of 96,000 and 3.2, respectively, obtained by gel permeation chromatography using polystyrene standards. It comprises 33 wt% PEO and has ~20 PEO side chains per chain on average (each side chain of 10 EO repeating units, spaced less than 2 nm apart along the backbone). For this PEO content, PMMA-g-PEO is highly resistant to cell adhesion unless covalently modified with cell adhesion ligands. To create cell-adhesive regions on the surface, the tEGF-polymer was diluted with a second PMMA-g-PEO comb polymer (mol. wt., 45,000; PDI, 1.8) comprising only 20 wt% PEO, a composition known to be cell-adhesive in the presence of serum or cell-secreted extracellular matrix (ECM).

Specifically, glass coverslips (12 mm in diameter) were silanized with Siliclad (Gelest Inc., Morrisville, PA, http://www.gelest.com) prior to polymer thin film preparation by spin coating. The tEGF-polymer was activated on the hydroxyl chain ends with 4-nitrophenyl chloroformate, mixed with diluent polymer (40:60 tEGF-polymer:diluent), and spin-coated to form an ~100-nm thin transparent film on the substrate (as measured by ellipsometry). Murine EGF was surface-coupled to the activated side chains of the spin-coated polymer by incubation in 25 µg/ml EGF in phosphate-buffered saline (PBS) (100 mM, pH 8–8.5) at room temperature for 22–24 hours followed by PBS washes. The remaining active groups were blocked in 100 mM Tris buffer (pH 9) at room temperature for 2 hours, followed by PBS washes to achieve a surface density of approximately 5,000–7,000 tethered EGF cells per µm2. Substrates were stored in PBS at 4°C until use. For in vitro experiments, each substrate was placed in individual wells of a 24-well plate and secured with a section of silicone rubber tubing, providing 0.64 cm2 of available surface area.

Activation of EGFR and EGFR Signaling Pathways
Prior to growth factor stimulation, the cells were serum-starved in a serum-free medium (Advanced DMEM [Invitrogen], supplemented with 1 mM pyruvate, 1 mM L-glutamine, 1 µM nonessential amino acids, and 100 units/ml penicillin-streptomycin) for 14–16 hours. The cells were detached with Versene (Invitrogen) and plated in the serum-free medium with the appropriate stimuli at a density of 25 x 103 cells per cm2 on the polymeric substrates and incubated at 37°C. The cells were pretreated for an hour with inhibitors when needed. At the indicated time points after plating, all adherent and nonadherent cells were pooled from two coverslips and incubated on ice for at least 15 minutes in 20 µl of lysis buffer (50 mM β-glycerophosphate at pH 7.3, 10 mM Na-pyrophosphate, 30 mM NaF, 50 mM Tris base at pH 7.5, 1% Triton X-100, 150 mM NaCl, 1 mM benazmidine, 2 mM EGTA, 100 µM Na3VO4, 1 mM dithiothreitol, 10 µg/ml aprotinin, 10 µg/ml leupeptin, 1 µg/ml pepstatin, 1 µg/ml microcystin-LR, and 1 mM phenylmethylsulfonyl fluoride). The supernatants were collected following 15 minutes of centrifugation at 16,000g and stored at –80°C. For the 0-hour time point, the cell mixtures were spun down at 450g for 10 minutes at 4°C. Medium was aspirated, and the cell pellet was lysed. The protein content of the cell lysates was determined using the Micro BCA protein assay (Pierce, Rockford, IL, http://www.piercenet.com). The supernatants were collected and stored at –80°C.

Cell Death Quantification Assay
Serum-starved (24 hours) hTMSC were plated in serum-free media on polymer substrates at 25 x 103 cells per cm2, allowed to attach for 4–8 hours (as indicated in figure legends), and then stimulated with prodeath cytokines at 100 ng/ml by addition to the medium. phMSC cultured to ~80% confluence in serum-containing culture medium were detached with trypsin and plated on polymer substrates in standard (serum-containing medium) at 25 x 103 cells per cm2. The culture medium was changed to a serum-free medium (Cambrex Mesenchymal Stem Cell Basal Medium) 4 or 5 hours after plating (as indicated in figure legends 5 and 6), and FasL was added 1 hour later (5 or 6 hours after plating). Human recombinant superFasL and TRAIL were from Alexis Biochemical (Lausen, Switzerland, http://www.alexis-corp.com). In some experiments with phMSC, substrates were coated with human plasma fibronectin (Sigma-Aldrich) by incubating comb copolymer (22 wt% PEO) substrates with 0.1 µg/ml fibronectin in PBS for 24 hours at room temperature followed by two gentle PBS washes. All adherent and nonadherent cells, detached with trypsin and collected from two coverslips, were spun down at 450g and stained in 10 µg/ml propidium iodide (Molecular Probes Inc., Eugene, OR, http://probes.invitrogen.com) for 15 minutes. The positive staining population was quantified by direct fluorescence using a FACSCalibur flow cytometer.

Cell Morphology Measurements
Static images of cells were captured under differential interference contrast (DIC) optics with a x10 or x20 lens after at least 4 hours of incubation on at least three independent samples to represent the average cell morphology. The acquired images of hTMSC were imported to ImageJ (Version 1.36b) to quantify the spread area per cell on the polymeric substrates. At least 20 cells per field were traced to calculate the average area per cell.

Statistical Analyses
Western blot densitometry data and propidium iodide (PI)+ staining population data were analyzed with analysis of variance and the t test. Cell spread area data were analyzed with Randsum and Kruskal-Wallis tests. Significance was set at p < .05 unless otherwise noted.


   RESULTS
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 Disclosure of Potential...
 Acknowledgments
 References
 
Kinetics of EGFR Signaling Pathways in MSC Stimulated by Soluble and Tethered EGF
Tethered EGF substrates were prepared with murine EGF, which has a single primary amine at the N terminus. This allows it to be precisely linked to the surface via amine-targeting chemistry, tethering the EGF molecules in a configuration competent to bind and activate EGFR and creating a locally high concentration of EGF in the vicinity of the cell-substrate interface [22]. To examine the response of MSC, we used the hTMSC line that expresses approximately 7,300 EGF receptors [18].

We first investigated the kinetics of hTMSC ERK and Akt/PKB [13, 14] responses to the human and murine EGF (the latter species had not been used to date). When hTMSC were stimulated with soluble EGF of either human or murine origin, we found that the phosphorylation of ERK and Akt peaked at 5 minutes and declined to the background level by 120 minutes; the extent and time course of activation was similar by both ligands (Fig. 1). There was basal Akt activity in these hTMSC, possibly due to factors such as insulin in the culture medium (as described in Materials and Methods) [40]. Therefore, murine EGF was comparable to human EGF for hTMSC, similar to other cells types responding to EGFR ligands from other mammalian species. Murine EGF was used for all subsequent experiments as its single primary amine allows for more specific covalent coupling to the substratum.


Figure 1
View larger version (21K):
[in this window]
[in a new window]

  		Figure 1. Soluble EGF transiently activated EGF receptor signaling pathways in human MSC. Human telomerase reverse transcriptase-immortalized human MSC (hTMSC) were plated on tissue culture-grade polystyrene to subconfluence, serum-starved for 14–16 hours, and then stimulated with 100 ng/ml soluble EGF of human (A) or murine (B) origin. hTMSC lysates at the indicated times were analyzed for ERK and Akt phosphorylation by Western blotting and quantified by densitometry. Data indicate the ratio of phosphorylated ERK to total ERK and the ratio of phosphorylated Akt to tubulin upon the stimulation of soluble human EGF (A) and murine EGF (B), respectively. Shown are the average ± SEM of three independent replicates for each comparison; there are no significant differences between the ligands. Abbreviations: EGF, epidermal growth factor; ERK, extracellular-regulated kinase; min, minutes; pAkt, phospho-Akt; pERK, phospho-extracellular-regulated kinase.

 
To exclude the possibility that the EGF response of hTMSC was an artifact of the hTERT immortalization procedure, we also investigated ppMSC, as they are readily obtainable and have metabolic rates comparable to human MSC [41]. These ppMSC express ~11,000 EGFR per cell and respond to a 10-minute stimulation of soluble EGF with an increase in phosphorylation of EGFR, ERK, and Akt that is inhibited by the EGFR tyrosine kinase inhibitor PD153035 [18]. In response to either porcine or murine soluble EGF stimulation, ERK phosphorylation in pMSC peaked at 5 minutes and gradually diminished (supplemental online Fig. 1). This indicated that both immortalized hTMSC and primary pMSC were responsive to murine soluble EGF.

As the hTMSC were similar to the porcine ones (Fig. 1), we chose the immortalized human cell line to conduct most of the remaining intracellular signaling experiments because these cells were easier to maintain and their behavior was more consistent compared with ppMSC. To investigate hTMSC responses to tethered EGF, we defined three experimental conditions. The first condition (control) was the unmodified comb copolymer surface, which served as a control substrate to study MSC behavior on the inert biomaterial (i.e., signals generated by adhesion) under EGF-free conditions. Second was the tEGF substrate, obtained by covalently attaching murine EGF to the control substrate to the free ends of PEO side chains. The tethering chemistry yields an average surface ligand density of ~5,000 EGF molecules/µm2, theoretically sufficient to saturate endogenous EGF receptors (estimated to be fewer than 100 EGFR per µm2 cell surface area for cells expressing EGFR at ~10,000 EGFR per cell [18]) and provides an average interligand spacing that should allow homodimerization of tEGF-bound EGFR. We have previously shown that attachment of long PEO tethers to the N terminus of murine EGF does not alter the EGFR-mediated internalization rate constant, and thus presumably the (soluble) affinity of PEO-modified EGF is comparable to native EGF; furthermore, presentation in tethered format greatly increases the local concentration in the vicinity of the cell surface, driving high receptor occupancy [22]. Last, to distinguish cellular responses specifically conferred by the tEGF substrate from those induced by EGF generally, we plated hTMSC on the control substrate in the presence of saturating murine soluble EGF (sEGF) (100 ng/ml). The sEGF condition thus acted as a more stringent comparison for characterizing MSC behavior unique to tEGF substrates.

The tEGF is covalently linked to the substratum; this is postulated not only to prevent ligand-receptor internalization but also to compartmentalize EGFR signaling to the cell surface [23]. As survival and proliferation signals come from both surface and internalized EGFR-induced cascades, it was unclear how tEGF would modify these cascades. To examine the bioactivity of tEGF, we investigated ERK phosphorylation as a recognized intracellular signal downstream of EGFR that also effects transcriptional changes. At 8 hours, ERK phosphorylation in the sEGF group was only slightly elevated compared with the control group, consistent with the rapid adaptation of ERK signaling observed previously in hTMSC treated with EGF on culture-grade polystyrene (Fig. 1A). sEGF-induced ERK phosphorylation continued to decline, such that it was indistinguishable from the control by 24 hours (p < .05; Fig. 2). Strikingly, ERK phosphorylation in the tEGF group was twofold higher than either the control or sEGF group, and this elevated signaling was sustained for at least 24 hours (p < .05; Fig. 2). Thus, the tEGF signaled in a more persistent manner.


Figure 2
View larger version (35K):
[in this window]
[in a new window]

  		Figure 2. ERK phosphorylation in human telomerase reverse transcriptase-immortalized human MSC (hTMSC) was stronger and more sustained with tEGF compared with sEGF. Serum-starved hTMSC were plated at a density of 25 x 103 cells per cm2 on the control substrates with or without sEGF (100 ng/ml) and the tEGF substrates, and cells were lysed at the indicated times. pERK was normalized by densitometry relative to total ERK. Error bars represent SEM of quadruplet independent lysates. *, p < .05 in comparison to the respective control group and sEGF group. Results are representative of three biological replicates. Abbreviations: ERK, extracellular-regulated kinase; hr(s), hour(s); pERK, phospho-extracellular-regulated kinase; sEGF, soluble epidermal growth factor; tEGF, tethered epidermal growth factor.

 
Interestingly, the activation of EGF signaling on tEGF substrate appeared to be pathway specific, since Akt phosphorylation was similar across the three experimental groups (data not shown). Significant Akt phosphorylation in the control group indicated that the Akt signaling in hTMSC might be a nonspecific response to plating and binding through adhesion receptors. Alternatively, Akt could be downregulated by mechanisms distinct from that of ERK [42]. Nevertheless, the clear and persistent ERK activation indicated that tEGF was functional as a bioactive ligand for hTMSC.

MSC Morphology on Polymeric Substrates Modified with tEGF
Tethering EGF to the substratum should enhance plasma membrane signaling, with pronounced effects on cell attachment and spreading, as this cellular process proceeds mainly from periplasma membrane epigenetic events. When hTMSC were plated under control, tEGF, and sEGF conditions, we observed morphological differences among the groups, starting as early as 8 hours after plating and persisting for at least 24–48 hours. On the control substrate, hTMSC attached but remained rounded with little spreading after 24 hours of incubation (Fig. 3A). By contrast, hTMSC on tEGF substrate exhibited a clear adhesive phenotype, evidenced by the presence of lamellipodia and increased cell area (Fig. 3C). Importantly, addition of sEGF to control substrates could not produce the same spreading observed on the tEGF substrate (Fig. 3B). When the average cell area was quantified for the three conditions, we found that the tEGF condition caused a three- to fourfold increase in spreading compared with either the control or sEGF groups (p < .05; Fig. 3D). Thus, EGF caused increased cell spreading of hTMSC when covalently tethered to the substrate to restrict signaling to the cell surface. Therefore, in parallel with the increased ERK signaling, tEGF specifically potentiated hTMSC adhesion.


Figure 3
View larger version (74K):
[in this window]
[in a new window]

  		Figure 3. tEGF causes spreading of human MSC on polymeric substrates. Light microscopy images of human telomerase reverse transcriptase-immortalized human MSC plated and maintained in serum-free medium were taken at a magnification of x10 at 24 hours on control substrates (without sEGF [A]; with sEGF [B]) and tethered EGF substrates (C). Scale bar = 100 µM. (D): Average cell spreading area was obtained by tracing individual cell area on triplicate substrates. Error bars indicate SEM of 60 individual cells. *, p < .05 in comparison with the control and sEGF groups. Results are representative of three biological replicates. Light microscopy images of primary human MSC plated in serum-containing medium for 4 hours and maintained in serum-free medium thereafter were taken at a magnification of x10 at 5 hours (E, F, G) or 24 hours (H, I, J) after plating on control substrates (without sEGF [E, H]; with sEGF [F, I]) and tethered EGF substrates (G, J). Scale bar = 100 µM. Results are representative of three biological replicates. Abbreviations: sEGF, soluble epidermal growth factor; tEGF, tethered epidermal growth factor.

 
The presentation mode of EGF also influenced the morphology of low-passage phMSC. phMSC were allowed to attach to the synthetic polymer substrates for 4 hours in the presence of serum, which contains fibronectin, vitronectin, and other molecules that are known to promote cell attachment and spreading. Images acquired 1 hour after changing to serum-free medium (5 hours after plating) show that although phMSC had begun to spread under all conditions, they were far more extensively spread on tEGF compared with control substrates (Fig. 3E–3G). Specifically, most phMSC on control substrates in the absence of EGF exhibited numerous filipodia and lamellipodia emanating from a rounded central cell body (Fig. 3E). In the presence of soluble EGF, far fewer cells appeared attached to control substrates, and soluble EGF seemed to slightly inhibit spreading of attached phMSC (Fig. 3F) compared with behavior of cells on control substrates in the absence of EGF (Fig. 3E). In contrast, most cells on tEGF were already extremely spread with flattened cell bodies at 5 hours (Fig. 3G). The trends observed at 5 hours were accentuated by 24 hours: phMSC on control substrates were moderately spread and highly elongated in the absence of EGF (Fig. 3H); phMSC were fewer in number and appeared more spindly on control substrates in the presence of soluble EGF (Fig. 3I); and phMSC on tEGF substrates, although still elongated, appeared flatter and more spread (Fig. 3J) than under other conditions. Thus, presentation of EGF in a mode that prevents internalization promoted attachment and spreading of phMSC, whereas soluble EGF appeared inhibitory compared with control.

Cellular Signaling Events Controlling tEGF-Induced MSC Spreading
Although both cell spreading and sustained ERK phosphorylation were apparent on tEGF substrates, it was possible that ERK phosphorylation was not directly induced by tEGF binding to EGFR. Furthermore, it was unclear whether EGFR and ERK signaling were required for the observed differences in morphology. To determine whether ERK activation was directly induced by tEGF and involved in cell spreading, we used various EGF pathway inhibitors to probe the upstream signaling requirements for ERK phosphorylation (Fig. 4). Specifically, we selected an EGF-blocking antibody, C225 [43], to inhibit the binding of EGF to EGFR, the EGFR tyrosine kinase inhibitors AG1478 [44] and PD153035 [45] to suppress the activation of EGFR, and the MEK1/2 inhibitor U0126 [46] to specifically inhibit the MEK-ERK signaling pathway. Through these inhibitors, we were able to examine the dependence of tEGF-induced morphology on EGF binding, EGFR tyrosine kinase activity, and MEK-ERK signaling.


Figure 4
View larger version (111K):
[in this window]
[in a new window]

  		Figure 4. MSC spreading on tEGF was blocked by inhibitors of epidermal growth factor receptor (EGFR) signaling. Serum-starved human telomerase reverse transcriptase-immortalized human MSC were plated on control and tEGF substrates at a density of 25 x 103 cells per cm2 (light microscopy images at a magnification of x10 taken at 8 hours [A, B]). Inhibitors include an anti-epidermal growth factor-blocking antibody (10 µg/ml C225), EGFR tyrosine kinase inhibitors (1 µM AG1478, 1 µM PD153035), and a mitogen-activated protein kinase kinase inhibitor (10 µM U0126) ([D, E, F, G], respectively). DMSO was included as a carrier control (C) for AG1478, PD153035, and U0126. (H): Average cell spreading quantified from (A–G). Error bars indicate SEM of 80 individual cells. *, p < .01 in comparison with the control group as well as the inhibitor groups. Results are representative of three biological replicates. (I): Western blot for pERK and total ERK confirming efficacy of the inhibitors used. Abbreviations: AG, AG1478; DMSO, dimethyl sulfoxide; ERK, extracellular-regulated kinase; PD15, PD153035; pERK, phospho-extracellular-regulated kinase; tEGF, tethered epidermal growth factor.

 
Phase-contrast images of serum-starved hTMSC on tEGF substrates 8 hours after plating showed early increases in spreading, with noticeably more prominent and extended lamellipodia compared with the control group (Fig. 4A, 4B). hTMSC on tEGF substrates showed no reduction of cell spreading when a fixed amount of 0.1% vol/vol dimethyl sulfoxide (DMSO) was added to the medium to account for any possible effects of the inhibitor diluent (Fig. 4C). The quantified average cell spreading area was approximately twofold increased on tEGF (with or without DMSO) compared with the control (p < .01; Fig. 4H). By contrast, cell spreading was reverted by all of the EGFR-pathway inhibitors at 8 hours (Fig. 4D–4G). As expected, the upstream inhibitors (i.e., PD153035 and C225) were more effective at blocking cell spreading. C225 and PD153035 reduced spreading by more than twofold compared with the corresponding tEGF group (p < .01; Fig. 4H), and the same rounded morphology persisted for at least 24 hours. AG1478 and U0126 also significantly reduced spreading This demonstrated that EGFR binding and tyrosine kinase signaling were required for tEGF-induced cell spreading.

The effectiveness of the MEK inhibitor U0126 in blocking the hTMSC morphology raised the possibility that all of the inhibitors were perturbing MEK-ERK as a common mediator of tEGF-induced cell spreading. To examine this, we prepared lysates from each of the conditions in Figure 4A–4G and measured ERK phosphorylation by Western blotting. At 8 hours, ERK phosphorylation from the tEGF groups (with and without carrier control) was elevated in comparison with the control (Fig. 4I). The EGFR-blocking antibody (C225) effectively suppressed downstream ERK activation in the hTMSC, indicating that ERK phosphorylation in the tEGF group was initiated by EGF receptor binding. The EGF receptor kinase inhibitors (PD153035 and AG1478) and the MEK inhibitor (U0216) also blocked ERK activation at 8 hours to a similar extent. Since perturbation of either the initial tEGF-EGFR binding event or the MEK relay point was sufficient to block ERK activation (Fig. 4I) as well as cell spreading (Fig. 4A–4G), we concluded that sustained signaling via ERK (Fig. 2) was responsible for tEGF-induced spreading.

MSC Death Induced by TNF-Family Cytokines
The differential cell spreading on the tEGF substrates provided the first evidence that new MSC responses could be achieved by tethering EGF to the culture substrate. Although cell morphology itself does not directly translate to improved cellular function in vivo for scaffolds bearing tEGF, we reasoned that the increased cell spreading might indicate the existence of other tEGF-induced phenotypes that are relevant for in vivo performance of a tissue engineering scaffolds such as those used for bone grafts. Such a cell response would be important during wound healing when proinflammatory cytokines are prevalent during the acute inflammatory phase when healing initiates [47]. These cytokines activate immune cells to help clear dying and infected tissue, but they can also cause collateral damage to reparative cells at the implant site. We therefore asked whether such cytokines could stimulate cell death in hTMSC. If so, then tEGF on the surface of a bone graft scaffold might help these progenitors resist death and thus promote tissue regeneration.

We screened FasL [24], TRAIL [25], TNF [26], IFN-{gamma} [27], and IL-1 [28] as recognized proinflammatory cytokines that can induce cell death. FasL and TRAIL produced the most apparent apoptotic morphology in hTMSC on tissue-culture polystyrene, as observed by increased cell detachment and the formation of apoptotic bodies (data not shown). We thus focused on FasL and TRAIL and also included another member of the TNF superfamily, RankL, which has been implicated in bone remodeling but is not generally regarded as a prodeath ligand [48–50]. When hTMSC death was quantified by positive staining with PI, we found that both FasL and TRAIL induced cell death significantly, in the presence of the protein synthesis inhibitor and cell stressor cycloheximide (CHX) (p < .05; Fig. 5A). FasL was approximately twice as potent as TRAIL. For many TNF family members, transcriptional pathways are activated to synthesize prosurvival proteins, which resist apoptosis in cells [51, 52]; accordingly, FasL- and TRAIL-induced cell death was dramatically reduced without CHX on the highly adhesive tissue culture polystyrene surfaces. RankL, with or without CHX, did not significantly influence cell death, as expected. These results implicated FasL and TRAIL as relevant prodeath ligands for hTMSC.


Figure 5
View larger version (46K):
[in this window]
[in a new window]

  		Figure 5. FasL and TRAIL induced cell death in hTMSC, hCTP, and phMSC. (A): PI+ stained population percentage quantified in hTMSC. Serum-starved hTMSC on tissue culture polystyrene at subconfluence were stimulated with various tumor necrosis factor-family cytokines in the presence or absence of the protein synthesis inhibitor CHX. Cells were stained with PI at 24 hours and analyzed via fluorescence-activated cell sorting. (B): PI+ stained population percentage quantified in hCTP. Error bars represent SEM of triplicate independent cell samples. *, p < .05 in comparison with the control. (C): PI+ stained population percentage quantified in phMSC. phMSC were plated in serum-containing medium on untreated or FN-coated polymer substrates, medium was changed to serum-free medium at 5 hours, and FasL was added at 6 hours. Cells were harvested and stained with PI at 24 hours. *, p < .05 in comparison with the control. (D): Differential interference contrast (DIC) images of FasL-treated phMSC (on FN) at time of harvest. (E): DIC images of untreated phMSC on FN at time of harvest. Abbreviations: CHX, cycloheximide; FasL, Fas ligand; FN, fibronectin; hCTP, human primary connective tissue progenitors; hTMSC, human telomerase reverse transcriptase-immortalized human MSC; phMSC, primary human MSC; PI, propidium iodide; TRAIL, tumor necrosis factor-related apoptosis-inducing ligand.

 
To determine whether the sensitivity to FasL and TRAIL could be extended beyond the immortalized hTMSC line, we replicated the experiment in a primary hCTP population (as described in Materials and Methods). As with the hTMSC, both FasL and TRAIL, but not RankL, induced cell death significantly above the control level in hCTP (p < .05; Fig. 5B). Furthermore, with hCTP it was not necessary to suppress protein synthesis to observe the cytokine-stimulated cell death, although the levels of cell death increased when CHX was added (p < .05; supplemental online Fig. 2). These data show that the two TNF-family cytokines, FasL and TRAIL, induce cell death in human bone progenitor cells, particularly for hTMSC when protein synthesis is suppressed, and that FasL is a more potent death stimulus than TRAIL (Fig. 5A).

Finally, we confirmed that FasL induced cell death in primary human cells conforming to conventional definitions of mesenchymal stem cells by treating phMSC obtained from a commercial source with FasL. Even in the absence of the protein synthesis inhibitor CHX, or in the presence of the adhesion-promoting protein fibronectin on the substrate, FasL induced cell death significantly above controls in phMSC (p < .05; Fig. 5C). Values for percentage of cell death as assessed by quantifying the PI+ population were consistent with visual observations of cell morphology in FasL-treated and untreated samples (Fig. 5D, 5E).

hTMSC Resistance to FasL-Induced Cell Death on tEGF
Having found that TNF-family cytokines can drive hTMSC death, we next investigated whether tEGF provided any protection. To be in line with the duration of acute inflammation, we extended the hTMSC death experiment to 48 hours after plating on the polymeric substrates. In the absence of cytokines, we observed a subtle decrease in cell death in the tEGF group compared with both the control and sEGF groups (p < .05; Fig. 6A, left). When challenged with FasL 8 hours after plating, cell death in both the control and sEGF groups increased substantially, but the tEGF group remained two- to threefold lower (p < .05). The observation of significant cell death in the absence of CHX likely arises from the difference in adhesion signals for cells on moderately adhesive comb copolymers (Fig. 6) compared with cells on highly adhesive tissue culture polystyrene (Fig. 5). Importantly, the tEGF group had a significantly lower PI+ population than the sEGF group (p < .05; Fig. 6A, center), indicating that tEGF bears a unique protective function to hTMSC that could not be recapitulated by sEGF. When the less potent death stimulus TRAIL was used (Fig. 5A), hTMSC death was lower overall, and no significant difference was seen among the groups (Fig. 6A, right). These results indicate a protective function of tEGF on hTMSC when challenged with the prodeath stimulus FasL.


Figure 6
View larger version (21K):
[in this window]
[in a new window]

  		Figure 6. tEGF conferred resistance to FasL-induced death via epidermal growth factor receptor (EGFR) and extracellular-regulated kinase. (A): Serum-starved human telomerase reverse transcriptase-immortalized human MSC (hTMSC) were plated on control substrates, with and without sEGF, and tEGF substrates. Prodeath stimuli were added at 8 hours. Cells were stained with PI at 24 hours and analyzed via fluorescence-activated cell sorting (FACS). Error bars represent SEM of triplicate independent cell samples. (B): Primary human MSC were plated in serum-containing media for 5 hours and then exchanged to serum-free media. FasL was added at 6 hours. Cells were stained with PI at 24 hours and analyzed via FACS. Error bars represent SEM of triplicate independent cell samples. *, p < .05; #, p = .06 in comparison with the control. (C): Serum-starved hTMSC were plated on control substrates, with and without sEGF, and tEGF substrates. Mitogen-activated protein kinase kinase inhibitor U0126 (1 µM) was added at 7 hours and allowed 1 hour of incubation before FasL was added at 8 hours. Cells were stained with PI at 24 hours and analyzed via FACS. Error bars represent SEM of quadruplet sets of triplicate independent cells samples. *, p < .05 in comparison with the control. (D): Serum-starved hTMSC were plated on control or tEGF substrates. EGFR kinase inhibitor PD153035 (1 µM) was added at 7 hours and allowed 1 hour of incubation before FasL was added at 8 hours. Cells were stained with PI at 24 hours and analyzed via FACS. Error bars represent SEM of triplicate independent cells samples. *, p < .05; **, p < .1 in comparison with the corresponding control. +, p < .01 in comparison to FasL without inhibitor for the same substrate conditions. Abbreviations: FasL, Fas ligand; PI, propidium iodide; sEGF, soluble epidermal growth factor; tEGF, tethered epidermal growth factor; TRAIL, tumor necrosis factor-related apoptosis-inducing ligand.

 
We next examined the % PI response of phMSC to FasL in the presence of soluble and tethered EGF to determine whether the protective effects of tEGF extended to primary cells. Even in the absence of FasL, soluble EGF induced phMSC death significantly more than controls (p < .05; Fig. 6B, left), a finding consistent with light microscopy observations (described above) that soluble EGF appeared to inhibit cell attachment and spreading of phMSC. Similar to the findings with hTMSC, tEGF was slightly protective against the modest level of cell death in phMSC observed in the absence of FasL (p < .05; Fig. 6B, left). In phMSC, % PI+ values for FasL plus tEGF were comparable to control untreated cells, indicating that tEGF offers almost complete protection against FasL-mediated death in these cells (Fig. 6B, right).

It was likely that the tEGF-mediated protection to FasL-induced cell death was occurring via differential signaling from the tethered EGF ligand (Fig. 2). To probe the intracellular basis of the antideath function provided by tEGF, we first used the MEK inhibitor U0126, since this small molecule suppressed both ERK activation and tEGF-induced cell spreading (Fig. 4G–4I). When U0126 was added to hTMSC, we found that the PI+ population on tEGF substrates was raised by threefold, to a level comparable to the control and sEGF groups, indicating that the suppression of ERK activation was sufficient to abolish the antideath function by tEGF (Fig. 6C). Interestingly, direct inhibition of EGF receptor signaling using the EGF receptor kinase inhibitor PD153035 1 hour before addition of FasL resulted in a doubling of the % PI+ population for cells on tEGF (p < .05; Fig. 6D), but it also significantly increased the % PI+ population for FasL-challenged cells under control conditions (p < .01; Fig. 6D). Addition of a monoclonal anti-EGF receptor antibody at the same time as FasL also significantly increased cell death in for both tEGF and control conditions (supplemental online Fig. 3; increases in % PI+ for cells on tEGF were not as dramatic with blocking antibody inhibition as with PD153035 inhibition, perhaps because the dissociation of tEGF-EGF receptor complexes at the cell-substrate interface, required for action of the blocking antibody, is slow). The increase in cell death in the controls suggests the possibility that an EGF receptor autocrine loop may be partially protecting cells against FasL in the absence of exogenous EGF. EGFR autocrine loops are operative in many cell types under basal conditions and can be stimulated by proinflammatory cytokines in both epithelial and mesenchymal cells [34, 53].

Whereas MEK inhibition completely abolished the protective effect of tEGF compared with controls (Fig. 6C), inhibition of EGF receptor signaling did not completely abolish the protective effect of tEGF compared with controls when cells were challenged with the prodeath stimulus FasL (Fig. 6D). For both the control and tEGF conditions, the % PI+ population increased about twofold in the presence of PD153035 and FasL (Fig. 6D), thus maintaining a significant difference between the two conditions. FasL was added 8 hours after the initial switch to various EGF conditions (inhibitor PD153035 was added 7 hours after medium switch), and during this interval, tEGF may initiate prosurvival adhesion signals or autocrine loops that persist during the FasL stimulation period. Integrins and other adhesion receptors signal through ERK [54, 55], as do many prosurvival autocrine loops [34]. Taken together, we conclude that both spreading and anti-death responses mediated by tEGF occured predominantly via potentiated ERK signaling in hTMSC.


   DISCUSSION
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
Disclosure of Potential...
 Acknowledgments
 References
 
In this study, we found that tEGF induced elevated ERK signaling in MSC to cause increased cell spreading and increased resistance to FasL-mediated cell death. Importantly, these responses were unique to tEGF, because saturating concentrations of sEGF did not reproduce the same morphological or functional phenotypes. Our group has reported elsewhere that tEGF inhibits both EGFR internalization and degradation compared with sEGF. EGF tethering thus prevents two recognized mechanisms for downregulating EGFR signaling [56, 57]. Together with the high growth factor concentrations enforced locally by tEGF [22], these mechanisms allow for potentiated and sustained EGFR signaling via ERK, at least in MSC.

ERK is activated by growth factors to affect a diverse array of cellular events, including cell proliferation, migration, and survival [13]. Recent evidence suggests that EGF-stimulated ERK directly phosphorylates the focal adhesion protein vinexin to regulate cell spreading [58], and EGF covalently linked to beads stimulates actin polymerization at the point of bead application [59]. Furthermore, ERK signaling can antagonize several death regulators, including caspase-9 [32], Bad [60], and Bim [61]. Therefore, the implication of ERK in spreading and survival of tEGF-stimulated hTMSC is consistent with published mechanisms.

It is interesting to consider whether the tethering of other EGFR ligands would induce cellular responses similar to tEGF. There are several EGF-family growth factors that bind to EGFR, including transforming growth factor-{alpha} (TGF-{alpha}), amphiregulin, betacellulin, epiregulin, and heparin-binding EGF-like growth factor [62]. Of interest, many of these ligands can induce signaling through EGFR when presented in juxtacrine mode (although this may require partial proteolytic processing of the proforms). These ligands have different binding affinities, receptor-dimerization partners, and trafficking patterns, which diversify intracellular signaling responses and cellular phenotypes [15, 63–65]. For EGF-family ligands such as TGF-{alpha}, we expect that tethered ligand may elicit responses that are more similar to EGF than the soluble ligand, because cells would be unable to distinguish these ligands based on their trafficking. By contrast, tethering a ligand such as heregulin might magnify its differences from EGF because of sustained signaling from distinct ErbB heterodimers [65]. Tethering of other EGF-family or growth-factor ligands may promote healing and tissue regeneration for other therapeutic applications.

EGF-like repeats are contained in various ECM molecules, including tenascin, laminin, thrombospodin, and versican [66–68]. Several of these insoluble ECM ligands induce EGFR-dependent cell adhesion and signaling in vitro [23, 69]. The specific responsiveness of hTMSC to tEGF that is described here suggests that constituent EGF-like repeats may act as endogenous tethered ligands for MSC. Indeed, one such EGF-like repeat-containing protein, laminin 5, was recently identified in bone and was found to be expressed by human MSC and to stimulate their osteogenic differentiation through activation of ERK [69]. Although these results have been attributed to the integrin-binding motifs of laminin 5 (rather than the EGF repeats), taken together with the results reported here, they suggest that the observed difference between soluble EGF and tethered EGF may reflect the ability of MSC to discriminate between naturally occurring molecules.

We further explored MSC physiology by perturbing these cells with cytokines relevant to inflammation, as these might lead to MSC death during the initial phases of wound repair [24, 25]. FasL, in particular, is expressed both by osteoclasts [31], which participate in bone remodeling, and by activated lymphocytes [70, 71], which are recruited during inflammation. FasL is also expressed by CD34+ bone marrow cells [72] and thus may be present when marrow is included in bone grafts [11, 39]. Our results in both immortalized and primary human cells indicated that FasL is a potent prodeath factor for MSC; these results are consistent with a previous report that anti-Fas antibodies induce apoptosis in human osteoblasts [73]. Cell adhesion is generally recognized to inhibit cell death [74], as underscored by the greater sensitivity of the immortalized hTMSC to prodeath stimuli on the moderately adhesive comb polymer substrates compared with highly adhesive tissue culture plastic. On highly adhesive substrates, primary connective tissue progenitors and phMSC were more sensitive than immortalized cells to FasL-induced cell death, as death was induced in the absence of CHX cotreatment in both types of primary cells. The greater resistance of hTERT-immortalized MSC compared with primary cells is consistent with emerging data that hTERT protects against apoptosis [75–77].

Importantly, FasL-induced cell death was significantly reduced by adhesion to tEGF surfaces. Specifically, we showed that tEGF-induced sustained ERK phosphorylation provided a critical antideath signal to MSC stimulated with FasL. These data agree with various studies implicating ERK in survival [78], as well as specific reports showing that ERK can block FasL-mediated apoptosis [79]. These data are also in agreement with a previous report that primary murine MSC transfected with EGFR show enhanced resistance to FasL-mediated apoptosis [80]. The involvement of FasL in MSC biology is likely to be physiologically important, as is the possible antagonism between FasL and EGF-like repeat-containing ECM proteins in the wound-healing milieu.

Most tissue-engineering efforts to date have focused on modifying polymer surfaces with adhesion ligands or fragments (e.g., fibronectin, arginine-glycine-aspartate, laminin) [81, 82] and delivering growth factors through slow release of soluble, diffusible molecules [83]. The potential for improving efficacy and enhancing biological function of growth factors by delivering them in a form tethered to the scaffold or matrix is now emerging as a new paradigm based on promising data from several systems [22, 84–86]. EGF is a recognized soluble mitogen [18] but has only limited effects on hTMSC that are plated on polymer surfaces. By tethering EGF, we increased and sustained ERK activation to promote hTMSC spreading and survival.

The introduction of an implant device initiates an array of tissue responses beginning with an acute inflammatory phase within minutes to days after implantation, followed by a chronic inflammatory phase. Inflammation serves to eliminate any infectious causes of injury, to remove dead cells and debris, and to initiate healing that would restore tissue structure. For an implant, however, many inflammatory functions are deleterious, causing foreign-body reactions and premature implant failure. In particular, inflammation may damage progenitor cells while repopulating an implant surface [87]. A crucial challenge for inductive biomaterials is therefore to promote the regenerative functions of progenitors such as MSC despite inflammation.

The tEGF surface modification we describe provides two key features favoring MSC repopulation. First, the tethering of EGF promoted hTMSC spreading. Progenitors such as MSC are found locally in the marrow, periosteum, and surrounding tissues. The value of tEGF modification would thus be to retain the MSC that interact with the implant surface. Second, tEGF provides a clear prosurvival stimulus to hTMSC. In the current study, we found that tEGF protected MSC for at least 40 hours from the types of apoptotic stimuli that are present during inflammation. This suggests that MSC retained on an implant surface will be more likely to survive against an inflammatory insult.

Although not investigated explicitly here, we speculate that tEGF should augment other MSC functions that are important for bone healing. Previously, we demonstrated that EGF promotes MSC proliferation while preserving pluripotency [18]. EGF can also collaborate with other factors to promote osteogenic differentiation of hTMSC [19]. All of these reports involved sEGF, so it remains unclear whether tEGF would act as an equivalent or more potent stimulus for proliferation or differentiation or both. However, both proliferation [88] and osteogenic differentiation [19, 69, 89] depend on ERK signaling, which is stronger and more sustained following tEGF stimulation. Our studies have focused on the modification of polymers spun on glass coverslips. This was essential for the basic characterization of how MSC respond to EGF tethering. Clearly, the next step toward actual implants would involve tEGF modifications to more realistic polymer structures, such as three-dimensional scaffolds [90], and we are currently adapting our protocols to such formats. Nevertheless, the success of bioglasses (BioGran [91] and PerioGlas [92]) to treat small- to medium-sized bony defects such as periodontal lesions suggests a more immediate clinical application. The tEGF surface-modification protocol could easily be adapted for bioglass to improve cell survival of progenitors in the grafted surroundings. Overall, we expect that the presentation of inductive MSC ligands through tethering will synergize with advances in materials to provide viable therapies for large bony defects.


   DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosure of Potential...
Acknowledgments
 References
 
The authors indicate no potential conflicts of interest.


   ACKNOWLEDGMENTS
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosure of Potential...
 Acknowledgments
References
 
We thank Dr. J. Toguchida (Kyoto University, Kyoto, Japan) for providing the immortalized human MSC, Dr. H. Abukawa from the Vacanti Laboratory (Massachusetts General Hospital, Boston, MA) for harvesting the primary porcine MSC, Dr. G. Muschler's laboratory (Cleveland Clinic, Cleveland, OH) for supplying the primary hCTP, and Prof. A. Mayes (Massachusetts Institute of Technology, Cambridge, MA) for guidance on the polymer chemistry. We also thank Ikuo Taniguchi (Massachusetts Institute of Technology) for preparing and characterizing activated polymers and Nicholas Marcantonio (Massachusetts Institute of Technology) for preparing polymer substrates for some experiments. This work was supported by NIH Grants R01-GM59870, R01-AR42997, and U54-GM064346 and by a the Harvard School of Dental Medicine NIDCR Training Grant.


   REFERENCES
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosure of Potential...
 Acknowledgments
 References
 

  1. Friedenstein AJ, Piatetzky S II, Petrakova KV. Osteogenesis in transplants of bone marrow cells. J Embryol Exp Morphol 1966;16:381–390.[Medline]

  2. Pittenger MF, Mackay AM, Beck SC et al. Multilineage potential of adult human mesenchymal stem cells. Science 1999;284:143–147.[Abstract/Free Full Text]

  3. Bruder SP, Kurth AA, Shea M et al. Bone regeneration by implantation of purified, culture-expanded human mesenchymal stem cells. J Orthop Res 1998;16:155–162.[CrossRef][Medline]

  4. Caplan AI. Review: Mesenchymal stem cells: Cell-based reconstructive therapy in orthopedics. Tissue Eng 2005;11:1198–1211.[CrossRef][Medline]

  5. Miyahara Y, Nagaya N, Kataoka M et al. Monolayered mesenchymal stem cells repair scarred myocardium after myocardial infarction. Nat Med 2006;12:459–465.[CrossRef][Medline]

  6. Pittenger MF, Martin BJ. Mesenchymal stem cells and their potential as cardiac therapeutics. Circ Res 2004;95:9–20.[Abstract/Free Full Text]

  7. Le Blanc K, Rasmusson I, Sundberg B et al. Treatment of severe acute graft-versus-host disease with third party haploidentical mesenchymal stem cells. Lancet 2004;363:1439–1441.[CrossRef][Medline]

  8. Barry FP, Murphy JM. Mesenchymal stem cells: Clinical applications and biological characterization. Int J Biochem Cell Biol 2004;36:568–584.[CrossRef][Medline]

  9. Meuleman N, Tondreau T, Delforge A et al. Human marrow mesenchymal stem cell culture: Serum-free medium allows better expansion than classical alpha-MEM medium. Eur J Haematol 2006;76:309–316.[CrossRef][Medline]

  10. Quarto R, Mastrogiacomo M, Cancedda R et al. Repair of large bone defects with the use of autologous bone marrow stromal cells. N Engl J Med 2001;344:385–386.[Free Full Text]

  11. Muschler GF, Nitto H, Matsukura Y et al. Spine fusion using cell matrix composites enriched in bone marrow-derived cells. Clin Orthop Relat Res 2003;407:102–118.

  12. Mastrogiacomo M, Muraglia A, Komlev V et al. Tissue engineering of bone: Search for a better scaffold. Orthod Craniofac Res 2005;8:277–284.[CrossRef][Medline]

  13. Roux PP, Blenis J. ERK and p38 MAPK-activated protein kinases: A family of protein kinases with diverse biological functions. Microbiol Mol Biol Rev 2004;68:320–344.[Abstract/Free Full Text]

  14. Kennedy SG, Wagner AJ, Conzen SD et al. The PI 3-kinase/Akt signaling pathway delivers an anti-apoptotic signal. Genes Dev 1997;11:701–713.[Abstract/Free Full Text]

  15. Wells A. EGF receptor. Int J Biochem Cell Biol 1999;31:637–643.[CrossRef][Medline]

  16. Lembach KJ. Induction of human fibroblast proliferation by epidermal growth factor (EGF): Enhancement by an EGF-binding arginine esterase and by ascorbate. Proc Natl Acad Sci U S A 1976;73:183–187.[Abstract/Free Full Text]

  17. Satomura K, Derubeis AR, Fedarko NS et al. Receptor tyrosine kinase expression in human bone marrow stromal cells. J Cell Physiol 1998;177:426–438.[CrossRef][Medline]

  18. Tamama K, Fan VH, Griffith LG et al. Epidermal growth factor as a candidate for ex vivo expansion of bone marrow-derived mesenchymal stem cells. STEM CELLS 2006;24:686–695.[Abstract/Free Full Text]

  19. Kratchmarova I, Blagoev B, Haack-Sorensen M et al. Mechanism of divergent growth factor effects in mesenchymal stem cell differentiation. Science 2005;308:1472–1477.[Abstract/Free Full Text]

  20. Krampera M, Pasini A, Rigo A et al. HB-EGF/HER-1 signaling in bone marrow mesenchymal stem cells: Inducing cell expansion and reversibly preventing multilineage differentiation. Blood 2005;106:59–66.[Abstract/Free Full Text]

  21. Gronthos S, Simmons PJ. The growth factor requirements of STRO-1-positive human bone marrow stromal precursors under serum-deprived conditions in vitro. Blood 1995;85:929–940.[Abstract/Free Full Text]

  22. Kuhl PR, Griffith-Cima LG. Tethered epidermal growth factor as a paradigm for growth factor-induced stimulation from the solid phase. Nat Med 1996;2:1022–1027.[CrossRef][Medline]

  23. Swindle CS, Tran KT, Johnson TD et al. Epidermal growth factor (EGF)-like repeats of human tenascin-C as ligands for EGF receptor. J Cell Biol 2001;154:459–468.[Abstract/Free Full Text]

  24. Grossman WJ, Verbsky JW, Barchet W et al. Human T regulatory cells can use the perforin pathway to cause autologous target cell death. Immunity 2004;21:589–601.[CrossRef][Medline]

  25. Wiley SR, Schooley K, Smolak PJ et al. Identification and characterization of a new member of the TNF family that induces apoptosis. Immunity 1995;3:673–682.[CrossRef][Medline]

  26. Laster SM, Wood JG, Gooding LR. Tumor necrosis factor can induce both apoptic and necrotic forms of cell lysis. J Immunol 1988;141:2629–2634.[Abstract]

  27. Boehm U, Klamp T, Groot M, Howard JC. Cellular responses to interferon-gamma. Annu Rev Immunol 1997;15:749–795.[CrossRef][Medline]

  28. Dinarello CA. Biologic basis for interleukin-1 in disease. Blood 1996;87:2095–2147.[Abstract/Free Full Text]

  29. Suda T, Okazaki T, Naito Y et al. Expression of the Fas ligand in cells of T cell lineage. J Immunol 1995;154:3806–3813.[Abstract]

  30. Miwa K, Asano M, Horai R et al. Caspase 1-independent IL-1beta release and inflammation induced by the apoptosis inducer Fas ligand. Nat Med 1998;4:1287–1292.[CrossRef][Medline]

  31. Park H, Jung YK, Park OJ et al. Interaction of Fas ligand and Fas expressed on osteoclast precursors increases osteoclastogenesis. J Immunol 2005;175:7193–7201.[Abstract/Free Full Text]

  32. Allan LA, Morrice N, Brady S et al. Inhibition of caspase-9 through phosphorylation at Thr 125 by ERK MAPK. Nat Cell Biol 2003;5:647–654.[CrossRef][Medline]

  33. Gaudet S, Janes KA, Albeck JG et al. A compendium of signals and responses triggered by prodeath and prosurvival cytokines. Mol Cell Proteomics 2005;4:1569–1590.[Abstract/Free Full Text]

  34. Janes KA, Gaudet S, Albeck JG et al. The response of human epithelial cells to TNF involves an inducible autocrine cascade. Cell 2006;124:1225–1239.[CrossRef][Medline]

  35. Irvine DJ, Mayes AM, Griffith LG. Nanoscale clustering of RGD peptides at surfaces using Comb polymers. 1. Synthesis and characterization of Comb thin films. Biomacromolecules 2001;2:85–94.[CrossRef][Medline]

  36. Okamoto T, Aoyama T, Nakayama T et al. Clonal heterogeneity in differentiation potential of immortalized human mesenchymal stem cells. Biochem Biophys Res Commun 2002;295:354–361.[CrossRef][Medline]

  37. Nakagawa K, Abukawa H, Shin MY et al. Osteoclastogenesis on tissue-engineered bone. Tissue Eng 2004;10:93–100.[CrossRef][Medline]

  38. Muschler GF, Nitto H, Boehm CA et al. Age- and gender-related changes in the cellularity of human bone marrow and the prevalence of osteoblastic progenitors. J Orthop Res 2001;19:117–125.[CrossRef][Medline]

  39. Majors AK, Boehm CA, Nitto H et al. Characterization of human bone marrow stromal cells with respect to osteoblastic differentiation. J Orthop Res 1997;15:546–557.[CrossRef][Medline]

  40. Avruch J. Insulin signal transduction through protein kinase cascades. Mol Cell Biochem 1998;182:31–48.[CrossRef][Medline]

  41. Sullivan TP, Eaglstein WH, Davis SC et al. The pig as a model for human wound healing. Wound Repair Regen 2001;9:66–76.[CrossRef]