EXPERIMENTAL Development of Chemotactic Smart Scaffold for Use in Tissue Regeneration Akishige Hokugo, D.D.S., Ph.D. Andrew Li, M.D. Luis A. Segovia, M.D. Anisa Yalom, M.D. Kameron Rezzadeh, B.A. Situo Zhou, M.D. Zheyu Zhang Patricia A. Zuk, Ph.D. Reza Jarrahy, M.D. Los Angeles, Calif.
Background: Regenerative medicine aims to obviate the need for autologous grafting through the use of bioengineered constructs that combine stem cells, growth factors, and biocompatible vehicles. Human mesenchymal stem cells and vascular endothelial growth factor (VEGF) have both shown promise for use in this context, the former because of their pluripotent capacity and the latter because of its chemotactic activity. The authors harnessed the regenerative potential of human mesenchymal stem cells and VEGF to develop a chemotactic scaffold for use in tissue engineering. Methods: Human mesenchymal stem cells were transduced with human VEGF via lentivirus particles to secrete VEGF. The chemotactic activity of the VEGFtransduced stem cells was evaluated via a trans-well assay. Migration through semipermeable membranes was significantly greater in chambers filled with medium conditioned by VEGF-transduced cells. VEGF-transduced cells were then seeded on apatite-coated poly(lactic-co-glycolic acid) scaffolds, thereby creating the Smart Scaffold. To determine in vivo angiogenesis, the Smart Scaffolds were implanted into subcutaneous pockets in the backs of nude mice. Results: Significantly larger numbers of capillaries were observed in the Smart Scaffold compared with control implants on immunohistologic studies. For the chemotactic in vivo study, human mesenchymal stem cells tagged with a fluorescent dye (1,1-dioctadecyl-3,3,3,3-tetramethylindotricarbocyanine iodide) were injected intravenously via tail vein after the subcutaneous implantation of the Smart Scaffolds. In vivo fluorescent imaging revealed that fluorescent dye–tagged human mesenchymal stem cells successfully accumulated within the Smart Scaffolds. Conclusion: These observations suggest that VEGF may play a vital role in the design of clinically relevant tissue regeneration graft substitutes through its angiogenic effects and ability to chemoattract mesenchymal stem cells. (Plast. Reconstr. Surg. 135: 877e, 2015.)
C
omplex wounds that require reconstruction represent a clinical problem that accounts for hundreds of millions of dollars in health care expenditures per year, a cost that does not include the unquantifiable emotional costs to patients who undergo painful reconstructive procedures. Autologous bone grafting is technically challenging, painful for the patient, and not always successful in bone reconstruction. Alloplastic implants are difficult to incorporate into the body, making them prone to infection, rejection, and extrusion. Therefore, bone graft From the Regenerative Bioengineering and Repair Laboratory, Division of Plastic and Reconstructive Surgery, Department of Surgery, David Geffen School of Medicine at University of California, Los Angeles. Received for publication May 14, 2014; accepted October 22, 2014. Copyright © 2015 by the American Society of Plastic Surgeons DOI: 10.1097/PRS.0000000000001199
substitutes that can regenerate mature bone without the need for painful operations would represent a significant achievement in bone reconstructive surgery. Unfortunately, the exact composition of such a substitute remains elusive, although current research suggests that some combination of stem cells, growth factors, and delivery vehicles are necessary in the development of a successful construct, such as the use of osteoconductive materials such as apatite-coated Disclosure: The authors have no commercial associations or financial disclosures that might pose a conflict of interest with any information presented in this article. This work was supported by THE PLASTIC SURGERY FOUNDATION.
www.PRSJournal.com
877e
Plastic and Reconstructive Surgery • May 2015 scaffolds,1,2 osteoinductive growth factors such as bone morphogenetic protein,3 and exogenously delivered stem cells, including human mesenchymal stem cells. Angiogenesis and osteogenesis have been shown to be interrelated,4,5 and vascular endothelial growth factor (VEGF) has demonstrated a potent chemotactic influence on the migration of human mesenchymal stem cells.6 It remains unclear, however, as to how VEGF and its chemotactic and angiogenic activity can be optimally harnessed for use in a bone regeneration construct model. Moreover, it is unclear whether the traditional tissue-engineering paradigm—exogenously expanded, treated, and subsequently implanted stem cells—is suitable in the setting of a complex wound environment. Recruiting stem cells to the site of tissue regeneration, rather than transferring a large number en bloc, addresses some of the shortcomings of this traditional tissue-engineering paradigm. By providing a nidus for tissue regeneration that then recruits circulating stem cells gradually, we believe we can see greater survival of those cells and, ultimately, more efficient tissue regeneration. The central hypothesis driving this study is that circulating stem cells will “home in” on a complex defect as a result of the chemotactic effects of VEGF and will successfully contribute to tissue regeneration. In this study, we attempt to use a tissue-engineering system wherein the substrate for reconstruction is derived from circulating stem cells. To recruit these cells to the defect site, we use VEGF as a chemotactic growth factor. We demonstrate that the continual release of VEGF from the cell-embedded scaffold induces angiogenesis, attracts mesenchymal stem cells to the site of injury, and successfully promotes tissue regeneration with the scaffold.
MATERIALS AND METHODS Transduction Human mesenchymal stem cells were obtained from Lonza Group Ltd. (Basel, Switzerland). Cells were plated in growth medium (MSCBM and SingleQuots; Lonza Group Ltd.). Second- or thirdpassage cell populations were used in this study. Premade lentiviral particles with hVEGF-A gene (LVP388; GenTarget Inc., San Diego, Calif.) were used for hVEGF gene transduction to human mesenchymal stem cells. This lentiviral particle can express the human targets under super-strong cytomegalovirus promoter. A blasticidin–red
878e
fluorescent protein fusion marker under Rous sarcoma virus promoter allows for selection of transduced cells for long-term expression via red fluorescent protein signal. Human mesenchymal stem cells (2.5 × 104 cells) were seeded on a well of six-well plates. After 1 day, lentiviral particles solution (100 μl) was added in a well. Three days after transduction, expression of red fluorescent protein was observed by fluorescent microscope. hVEGF Secretion by Enyzme-Linked Immunosorbent Assay To elucidate hVEGF transduction, secretion levels of hVEGF-A protein from transduced or nontransduced (normal) human mesenchymal stem cells were evaluated by enzyme-linked immunosorbent assay. After 3 days of culture, the medium was collected and evaluated with a hVEGF-A enzyme-linked immunosorbent assay kit (Abnova Corporation, Taipei City, Taiwan) according to the manufacturer’s instructions. The statistical difference was determined by t test. Difference with p < 0.05 was considered significant. In Vitro Transwell Migration Assay To verify the chemotactic capacity of hVEGFtransduced human mesenchymal stem cells on human mesenchymal stem cell migration, an in vitro migration assay was performed using a cell migration assay kit (InnoCyte Calbiochem; EMD Chemicals, Inc., Darmstadt, Germany) that consisted of two-chamber dishes in which the chamber contents are separated by a semipermeable membrane, allowing for exchange of soluble diffusible factors. Human mesenchymal stem cells or VEGF-transduced human mesenchymal stem cells were cultured in control media consisting of basal medium (MSCBM; Lonza Group Ltd.) and growth supplements (SingleQuots; Lonza Group Ltd.). After 3 days of cell culture, culture media were collected and stored at −20°C as conditioned media. Cells (5 × 105 cells/mL) in basal medium (without growth supplements) were seeded on the upper chambers (8-μm pores), and various conditioned media were added to the lower chambers. After 3 hours, cells that had migrated through the membrane were assessed quantitatively by staining with calcein-AM using a fluorescent plate reader (Bio-Rad, Hercules, Calif.). Statistical differences among groups were evaluated by one-way analysis of variance, and post hoc multiple-comparison tests (Tukey-Kramer multiple comparison test) were performed when significance was obtained.
Volume 135, Number 5 • The Chemotactic Smart Scaffold Preparation of the Smart Scaffold To prepare the Smart Scaffold, three-dimensional, apatite-coated poly(lactic-co-glycolic acid) scaffolds were prepared, and then VEGF-transduced human mesenchymal stem cells (1 × 106 cells) were seeded onto the scaffolds. First, apatite-coated substrates were prepared by a wet precipitation method through immersion in 5× concentrated simulated body fluid solutions, as previously described.7,8 Briefly, substrates were initially incubated in a 5× simulated body fluid-1, which consisted of CaCl2, MgCl2·6H2O, NaHCO3, K2HPO4·3H2O, Na2SO4, KCl, and NaCl in distilled deionized water for 24 hours at 37°C. The substrates were then transferred to a 5× simulated body fluid-2 solution that is similar to 5× simulated body fluid-1 but devoid of the crystal growth inhibitors Mg2+ and HCO3−. Incubation in 5× simulated body fluid-2 was performed at 37°C for 48 hours. Apatite-coated substrates were then rinsed gently three times with distilled deionized water and air-dried before further experimentation. Second, poly(lactic-co-glycolic acid) scaffolds were prepared by a solvent casting/porogen leeching method using 85:15 D,L-poly(lactic-coglycolic acid) (inherent viscosity, 0.61 dL/g; LACTEL Absorbable Polymers, Birmingham, Ala.). Briefly, 1.0 g of a 17.5% poly(lactic-co-glycolic acid) in chloroform solution was mixed with 1.5 g of 200- to 300-μm-diameter sucrose particles and pressed between Teflon molds to create 0.5-mmthick polymer sheets. The sheets were then frozen overnight and lyophilized for ~8 hours in a lyophilizer to remove solvent from the scaffolds. After lyophilization, the scaffolds were immersed in three changes of ddH2O for 1 hour each to leech the sucrose particles. After the scaffolds air-dried, a biopsy punch was used to obtain 5-mm-diameter circular scaffolds. The scaffolds were treated with plasma etching with the use of argon glow discharge to improve wetting, followed by sterilization in 70% ethanol for 30 minutes. Residual ethanol was rinsed from the scaffolds using three changes of sterile ddH2O. At this point, a subset of the sterile poly(lactic-coglycolic acid) scaffolds was prepared for apatite coating. For this, scaffolds were immersed in simulated body fluid-1 solution for 24 hours, followed by simulated body fluid-2 for 48 hours. At the end of the apatite coating process, the scaffolds were rinsed three times with sterile ddH2O and allowed to air dry in a laminar flow hood overnight. The cell suspension of VEGF-transduced human mesenchymal stem cells (1 × 106 cells/50
μL) was poured on the three-dimensional, apatite-coated poly(lactic-co-glycolic acid) scaffolds and incubated in 37°C 5% CO2 for 1 hour to prepare the Smart Scaffold. The growth medium was added gently, and the Smart Scaffold was cultured for 7 days to allow the VEGF-transduced human mesenchymal stem cells to grow and proliferate in the scaffold. In Vivo Angiogenesis Evaluation of Smart Scaffold in Mouse Dorsal Subcutaneous Pockets Subcutaneous pockets were made on the backs of nude mice (NC(NCr)-Foxn1, Charles River Laboratory, Mass.). The Smart Scaffold was placed in contact with the dorsal muscle tissue. Apatite-coated scaffolds with unmodified (normal) human mesenchymal stem cells and acellular apatite-coated scaffolds (scaffold only) were also implanted as controls. The overlying skin was closed with suture. At day 14,9,10 implants were removed along with surrounding tissue and fixed. All samples were stained with anti–platelet endothelial cell adhesion molecule-1 primary antibody (Abcam, Cambridge, Mass.) to detect the capillaries and vessels. Platelet endothelial cell adhesion molecule-1–positive vessels with morphologic circumference of one or more endothelial cells were counted. The number of vessels labeled with platelet endothelial cell adhesion molecule-1 was expressed as number of capillaries per section. Statistical differences among groups were evaluated by one-way analysis of variance, and post-hoc multiple-comparison tests (Tukey-Kramer multiple comparison test) were performed when significance was obtained. A p-value of less than 0.05 was considered statistically significant. Measuring the In Vivo Chemotactic Capacities of the Smart Scaffold Subcutaneous pockets were made on the backs of male nude mice. The Smart Scaffolds were placed in contact with the dorsal muscle tissue. Apatite-coated scaffolds seeded with unmodified human mesenchymal stem cells were also implanted. Mice undergoing creation of a subcutaneous pocket without implantation and primary closure (sham group) were prepared as controls. Each mouse was randomly assigned to receive three different scaffold implants. The overlying skin was closed with suture. Human mesenchymal stem cells were then labeled with the lipophilic tracer 1,1-dioctadecyl-3,3,3,3-tetramethylindotricarbocyanine iodide (DiR; Molecular Probes, Eugene, Ore.) for the imaging experiments. This fluorophore is excited at 750 nm and has an
879e
Plastic and Reconstructive Surgery • May 2015 emission peak at 782 nm. One to 2 weeks after implantation, the DiR-labeled human mesenchymal stem cells were injected intravenously into the caudal vein of mice implanted with the various scaffolds implants. In vivo imaging was then performed at 1 day after injection to assess migration of systemically injected, DiR-labeled cells. Mice were imaged under anesthesia using the imaging system (IVIS Lumina II; Caliper Life Sciences, Inc., Runcorn, United Kingdom) with a 1 minute capture, medium binning.
RESULTS VEGF Transduction to Human Mesenchymal Stem Cells From the observation by fluorescent microscope, almost 100 percent of the human mesenchymal stem cells successfully expressed red fluorescent protein (Fig. 1, left). In contrast, control cells expressed none (data not shown). These data indicate that our VEGF transduction rate of human mesenchymal stem cells was quite high. We had used these transduced cells without any sorting and selection for the further experiments. VEGF Secretion by Enzyme-Linked Immunosorbent Assay Figure 1, right, shows VEGF-A level in the medium cultured with transduced and
nontransduced human mesenchymal stem cells. The level of VEGF-A in the medium of transduced cells was significantly greater than that of nontransfected cells, indicating that our human mesenchymal stem cells are capable of overexpressing VEGF. In Vitro Transwell Migration Assay Figure 2 shows the data from the in vitro transwell migration assay. The number of human mesenchymal stem cells that migrated through the membrane toward the lower chamber that contained VEGF-transduced cell media was significantly greater than that in control media. This indicated that the cells preferentially migrated toward the media cultured with VEGF-transfected human mesenchymal stem cells. In Vivo Angiogenesis Evaluation in Mouse Dorsal Subcutaneous Pockets From the gross observation, a significantly larger number of vessels were seen at the “Smart Scaffold” site relative to controls (Fig. 3, above). From the immunohistological sections, significantly larger numbers of platelet endothelial cell adhesion molecule-1–positive vessels with morphologic circumference were observed in the Smart Scaffold. In contrast, a smaller number of smaller capillaries were observed in the human mesenchymal stem cell scaffold. No
Fig. 1. (Left) Human mesenchymal stem cells (hMSCs) were treated with lentiparticles with hVEGF-A gene. After 3 days, the cells were observed by fluorescent microscope. (Left, above) Bright light observation; (left, below) fluorescent light observation; magnification was ×40. (Right) After 3 days, culture media were collected and evaluated for VEGF-A level by enzyme-linked immunosorbent assay. Nontransfected (normal) human mesenchymal stem cells were also cultured and media were evaluated. *p
Background: Regenerative medicine aims to obviate the need for autologous grafting through the use of bioengineered constructs that combine stem cells, growth factors, and biocompatible vehicles. Human mesenchymal stem cells and vascular endothelial growth factor (VEGF) have both shown promise for use in this context, the former because of their pluripotent capacity and the latter because of its chemotactic activity. The authors harnessed the regenerative potential of human mesenchymal stem cells and VEGF to develop a chemotactic scaffold for use in tissue engineering. Methods: Human mesenchymal stem cells were transduced with human VEGF via lentivirus particles to secrete VEGF. The chemotactic activity of the VEGFtransduced stem cells was evaluated via a trans-well assay. Migration through semipermeable membranes was significantly greater in chambers filled with medium conditioned by VEGF-transduced cells. VEGF-transduced cells were then seeded on apatite-coated poly(lactic-co-glycolic acid) scaffolds, thereby creating the Smart Scaffold. To determine in vivo angiogenesis, the Smart Scaffolds were implanted into subcutaneous pockets in the backs of nude mice. Results: Significantly larger numbers of capillaries were observed in the Smart Scaffold compared with control implants on immunohistologic studies. For the chemotactic in vivo study, human mesenchymal stem cells tagged with a fluorescent dye (1,1-dioctadecyl-3,3,3,3-tetramethylindotricarbocyanine iodide) were injected intravenously via tail vein after the subcutaneous implantation of the Smart Scaffolds. In vivo fluorescent imaging revealed that fluorescent dye–tagged human mesenchymal stem cells successfully accumulated within the Smart Scaffolds. Conclusion: These observations suggest that VEGF may play a vital role in the design of clinically relevant tissue regeneration graft substitutes through its angiogenic effects and ability to chemoattract mesenchymal stem cells. (Plast. Reconstr. Surg. 135: 877e, 2015.)
C
omplex wounds that require reconstruction represent a clinical problem that accounts for hundreds of millions of dollars in health care expenditures per year, a cost that does not include the unquantifiable emotional costs to patients who undergo painful reconstructive procedures. Autologous bone grafting is technically challenging, painful for the patient, and not always successful in bone reconstruction. Alloplastic implants are difficult to incorporate into the body, making them prone to infection, rejection, and extrusion. Therefore, bone graft From the Regenerative Bioengineering and Repair Laboratory, Division of Plastic and Reconstructive Surgery, Department of Surgery, David Geffen School of Medicine at University of California, Los Angeles. Received for publication May 14, 2014; accepted October 22, 2014. Copyright © 2015 by the American Society of Plastic Surgeons DOI: 10.1097/PRS.0000000000001199
substitutes that can regenerate mature bone without the need for painful operations would represent a significant achievement in bone reconstructive surgery. Unfortunately, the exact composition of such a substitute remains elusive, although current research suggests that some combination of stem cells, growth factors, and delivery vehicles are necessary in the development of a successful construct, such as the use of osteoconductive materials such as apatite-coated Disclosure: The authors have no commercial associations or financial disclosures that might pose a conflict of interest with any information presented in this article. This work was supported by THE PLASTIC SURGERY FOUNDATION.
www.PRSJournal.com
877e
Plastic and Reconstructive Surgery • May 2015 scaffolds,1,2 osteoinductive growth factors such as bone morphogenetic protein,3 and exogenously delivered stem cells, including human mesenchymal stem cells. Angiogenesis and osteogenesis have been shown to be interrelated,4,5 and vascular endothelial growth factor (VEGF) has demonstrated a potent chemotactic influence on the migration of human mesenchymal stem cells.6 It remains unclear, however, as to how VEGF and its chemotactic and angiogenic activity can be optimally harnessed for use in a bone regeneration construct model. Moreover, it is unclear whether the traditional tissue-engineering paradigm—exogenously expanded, treated, and subsequently implanted stem cells—is suitable in the setting of a complex wound environment. Recruiting stem cells to the site of tissue regeneration, rather than transferring a large number en bloc, addresses some of the shortcomings of this traditional tissue-engineering paradigm. By providing a nidus for tissue regeneration that then recruits circulating stem cells gradually, we believe we can see greater survival of those cells and, ultimately, more efficient tissue regeneration. The central hypothesis driving this study is that circulating stem cells will “home in” on a complex defect as a result of the chemotactic effects of VEGF and will successfully contribute to tissue regeneration. In this study, we attempt to use a tissue-engineering system wherein the substrate for reconstruction is derived from circulating stem cells. To recruit these cells to the defect site, we use VEGF as a chemotactic growth factor. We demonstrate that the continual release of VEGF from the cell-embedded scaffold induces angiogenesis, attracts mesenchymal stem cells to the site of injury, and successfully promotes tissue regeneration with the scaffold.
MATERIALS AND METHODS Transduction Human mesenchymal stem cells were obtained from Lonza Group Ltd. (Basel, Switzerland). Cells were plated in growth medium (MSCBM and SingleQuots; Lonza Group Ltd.). Second- or thirdpassage cell populations were used in this study. Premade lentiviral particles with hVEGF-A gene (LVP388; GenTarget Inc., San Diego, Calif.) were used for hVEGF gene transduction to human mesenchymal stem cells. This lentiviral particle can express the human targets under super-strong cytomegalovirus promoter. A blasticidin–red
878e
fluorescent protein fusion marker under Rous sarcoma virus promoter allows for selection of transduced cells for long-term expression via red fluorescent protein signal. Human mesenchymal stem cells (2.5 × 104 cells) were seeded on a well of six-well plates. After 1 day, lentiviral particles solution (100 μl) was added in a well. Three days after transduction, expression of red fluorescent protein was observed by fluorescent microscope. hVEGF Secretion by Enyzme-Linked Immunosorbent Assay To elucidate hVEGF transduction, secretion levels of hVEGF-A protein from transduced or nontransduced (normal) human mesenchymal stem cells were evaluated by enzyme-linked immunosorbent assay. After 3 days of culture, the medium was collected and evaluated with a hVEGF-A enzyme-linked immunosorbent assay kit (Abnova Corporation, Taipei City, Taiwan) according to the manufacturer’s instructions. The statistical difference was determined by t test. Difference with p < 0.05 was considered significant. In Vitro Transwell Migration Assay To verify the chemotactic capacity of hVEGFtransduced human mesenchymal stem cells on human mesenchymal stem cell migration, an in vitro migration assay was performed using a cell migration assay kit (InnoCyte Calbiochem; EMD Chemicals, Inc., Darmstadt, Germany) that consisted of two-chamber dishes in which the chamber contents are separated by a semipermeable membrane, allowing for exchange of soluble diffusible factors. Human mesenchymal stem cells or VEGF-transduced human mesenchymal stem cells were cultured in control media consisting of basal medium (MSCBM; Lonza Group Ltd.) and growth supplements (SingleQuots; Lonza Group Ltd.). After 3 days of cell culture, culture media were collected and stored at −20°C as conditioned media. Cells (5 × 105 cells/mL) in basal medium (without growth supplements) were seeded on the upper chambers (8-μm pores), and various conditioned media were added to the lower chambers. After 3 hours, cells that had migrated through the membrane were assessed quantitatively by staining with calcein-AM using a fluorescent plate reader (Bio-Rad, Hercules, Calif.). Statistical differences among groups were evaluated by one-way analysis of variance, and post hoc multiple-comparison tests (Tukey-Kramer multiple comparison test) were performed when significance was obtained.
Volume 135, Number 5 • The Chemotactic Smart Scaffold Preparation of the Smart Scaffold To prepare the Smart Scaffold, three-dimensional, apatite-coated poly(lactic-co-glycolic acid) scaffolds were prepared, and then VEGF-transduced human mesenchymal stem cells (1 × 106 cells) were seeded onto the scaffolds. First, apatite-coated substrates were prepared by a wet precipitation method through immersion in 5× concentrated simulated body fluid solutions, as previously described.7,8 Briefly, substrates were initially incubated in a 5× simulated body fluid-1, which consisted of CaCl2, MgCl2·6H2O, NaHCO3, K2HPO4·3H2O, Na2SO4, KCl, and NaCl in distilled deionized water for 24 hours at 37°C. The substrates were then transferred to a 5× simulated body fluid-2 solution that is similar to 5× simulated body fluid-1 but devoid of the crystal growth inhibitors Mg2+ and HCO3−. Incubation in 5× simulated body fluid-2 was performed at 37°C for 48 hours. Apatite-coated substrates were then rinsed gently three times with distilled deionized water and air-dried before further experimentation. Second, poly(lactic-co-glycolic acid) scaffolds were prepared by a solvent casting/porogen leeching method using 85:15 D,L-poly(lactic-coglycolic acid) (inherent viscosity, 0.61 dL/g; LACTEL Absorbable Polymers, Birmingham, Ala.). Briefly, 1.0 g of a 17.5% poly(lactic-co-glycolic acid) in chloroform solution was mixed with 1.5 g of 200- to 300-μm-diameter sucrose particles and pressed between Teflon molds to create 0.5-mmthick polymer sheets. The sheets were then frozen overnight and lyophilized for ~8 hours in a lyophilizer to remove solvent from the scaffolds. After lyophilization, the scaffolds were immersed in three changes of ddH2O for 1 hour each to leech the sucrose particles. After the scaffolds air-dried, a biopsy punch was used to obtain 5-mm-diameter circular scaffolds. The scaffolds were treated with plasma etching with the use of argon glow discharge to improve wetting, followed by sterilization in 70% ethanol for 30 minutes. Residual ethanol was rinsed from the scaffolds using three changes of sterile ddH2O. At this point, a subset of the sterile poly(lactic-coglycolic acid) scaffolds was prepared for apatite coating. For this, scaffolds were immersed in simulated body fluid-1 solution for 24 hours, followed by simulated body fluid-2 for 48 hours. At the end of the apatite coating process, the scaffolds were rinsed three times with sterile ddH2O and allowed to air dry in a laminar flow hood overnight. The cell suspension of VEGF-transduced human mesenchymal stem cells (1 × 106 cells/50
μL) was poured on the three-dimensional, apatite-coated poly(lactic-co-glycolic acid) scaffolds and incubated in 37°C 5% CO2 for 1 hour to prepare the Smart Scaffold. The growth medium was added gently, and the Smart Scaffold was cultured for 7 days to allow the VEGF-transduced human mesenchymal stem cells to grow and proliferate in the scaffold. In Vivo Angiogenesis Evaluation of Smart Scaffold in Mouse Dorsal Subcutaneous Pockets Subcutaneous pockets were made on the backs of nude mice (NC(NCr)-Foxn1, Charles River Laboratory, Mass.). The Smart Scaffold was placed in contact with the dorsal muscle tissue. Apatite-coated scaffolds with unmodified (normal) human mesenchymal stem cells and acellular apatite-coated scaffolds (scaffold only) were also implanted as controls. The overlying skin was closed with suture. At day 14,9,10 implants were removed along with surrounding tissue and fixed. All samples were stained with anti–platelet endothelial cell adhesion molecule-1 primary antibody (Abcam, Cambridge, Mass.) to detect the capillaries and vessels. Platelet endothelial cell adhesion molecule-1–positive vessels with morphologic circumference of one or more endothelial cells were counted. The number of vessels labeled with platelet endothelial cell adhesion molecule-1 was expressed as number of capillaries per section. Statistical differences among groups were evaluated by one-way analysis of variance, and post-hoc multiple-comparison tests (Tukey-Kramer multiple comparison test) were performed when significance was obtained. A p-value of less than 0.05 was considered statistically significant. Measuring the In Vivo Chemotactic Capacities of the Smart Scaffold Subcutaneous pockets were made on the backs of male nude mice. The Smart Scaffolds were placed in contact with the dorsal muscle tissue. Apatite-coated scaffolds seeded with unmodified human mesenchymal stem cells were also implanted. Mice undergoing creation of a subcutaneous pocket without implantation and primary closure (sham group) were prepared as controls. Each mouse was randomly assigned to receive three different scaffold implants. The overlying skin was closed with suture. Human mesenchymal stem cells were then labeled with the lipophilic tracer 1,1-dioctadecyl-3,3,3,3-tetramethylindotricarbocyanine iodide (DiR; Molecular Probes, Eugene, Ore.) for the imaging experiments. This fluorophore is excited at 750 nm and has an
879e
Plastic and Reconstructive Surgery • May 2015 emission peak at 782 nm. One to 2 weeks after implantation, the DiR-labeled human mesenchymal stem cells were injected intravenously into the caudal vein of mice implanted with the various scaffolds implants. In vivo imaging was then performed at 1 day after injection to assess migration of systemically injected, DiR-labeled cells. Mice were imaged under anesthesia using the imaging system (IVIS Lumina II; Caliper Life Sciences, Inc., Runcorn, United Kingdom) with a 1 minute capture, medium binning.
RESULTS VEGF Transduction to Human Mesenchymal Stem Cells From the observation by fluorescent microscope, almost 100 percent of the human mesenchymal stem cells successfully expressed red fluorescent protein (Fig. 1, left). In contrast, control cells expressed none (data not shown). These data indicate that our VEGF transduction rate of human mesenchymal stem cells was quite high. We had used these transduced cells without any sorting and selection for the further experiments. VEGF Secretion by Enzyme-Linked Immunosorbent Assay Figure 1, right, shows VEGF-A level in the medium cultured with transduced and
nontransduced human mesenchymal stem cells. The level of VEGF-A in the medium of transduced cells was significantly greater than that of nontransfected cells, indicating that our human mesenchymal stem cells are capable of overexpressing VEGF. In Vitro Transwell Migration Assay Figure 2 shows the data from the in vitro transwell migration assay. The number of human mesenchymal stem cells that migrated through the membrane toward the lower chamber that contained VEGF-transduced cell media was significantly greater than that in control media. This indicated that the cells preferentially migrated toward the media cultured with VEGF-transfected human mesenchymal stem cells. In Vivo Angiogenesis Evaluation in Mouse Dorsal Subcutaneous Pockets From the gross observation, a significantly larger number of vessels were seen at the “Smart Scaffold” site relative to controls (Fig. 3, above). From the immunohistological sections, significantly larger numbers of platelet endothelial cell adhesion molecule-1–positive vessels with morphologic circumference were observed in the Smart Scaffold. In contrast, a smaller number of smaller capillaries were observed in the human mesenchymal stem cell scaffold. No
Fig. 1. (Left) Human mesenchymal stem cells (hMSCs) were treated with lentiparticles with hVEGF-A gene. After 3 days, the cells were observed by fluorescent microscope. (Left, above) Bright light observation; (left, below) fluorescent light observation; magnification was ×40. (Right) After 3 days, culture media were collected and evaluated for VEGF-A level by enzyme-linked immunosorbent assay. Nontransfected (normal) human mesenchymal stem cells were also cultured and media were evaluated. *p
