A major challenge in cell-based tissue engineering is achieving survival of delivered cells upon implantation in vivo. Developing methods to maintain delivered cell viability between times of implantation and construct vascularization could prove beneficial in a variety of cell-based tissue engineering applications. Platelet lysate (PL), a mixture rich in growth factors such as PDGF, b-FGF, and VEGF, has been shown to effectively increase the proliferative capacity of hMSCs. This study investigates utilization of PL for the development of cell-based bone tissue engineering strategies. In particular, this work aims to elucidate whether PL incorporation can be used to enhance the quantity of viable stem cells within a hydrogel construct by examining cells from different sources: human mesenchymal stem cells (hMSCs) and amniotic fluid-derived stem (hAFS) cells. While many studies have reported beneficial effects of PL on hMSC growth in a 2D culture, the impact of PL gel in a three-dimensional hydrogel environment or on hAFS cell proliferation has not been investigated. We hypothesized that incorporating PL gel would maintain the viability of hydrogel-embedded hMSCs and hAFS cells through a 1-week culture period in vitro. hMSCs and hAFS cells were co-transduced for constitutive expression of GFP and firefly luciferase reporter genes. Cells were then seeded in 2% RGD-alginate hydrogels either with (MSC-PL+, AFS-PL+ groups) or without (MSC-PL-, AFS-PL- groups) 15% PL gel (a gift from Emory University). Cell-embedded hydrogels were then injected into nanofiber mesh tubes (150uL / mesh) and monitored for bioluminescent signal (indicative of live cell number) over the course of 1 week in vitro. For both cell types, incorporation of PL significantly increased bioluminescent imaging (BLI) counts. For the hMSC groups, MSC-PL+ and MSC-PL-, the highest BLI counts were on Day 1. Furthermore, MSC-PL+ counts were higher in comparison to control (MSC-PL-) on Day 1. For hAFS cell constructs, the control group (AFS-PL-) did not show an increase in BLI at any timepoint. AFS-PL+ counts were seen to peak at both Day 1 and Day 7. Microscopy images of the hydrogels at day 7 showed the hAFS cell groups containing high-density regions of live cells whereas the hMSC groups did not. BLI results suggest better maintenance of stem cell viability as a result of PL addition to the hydrogel construct. Furthermore, the impact of PL incorporation appeared to be cell type dependent. While PL addition was shown to promote hAFS cell viability through 1 week, this effect on survival was observed for the hMSC group only at the Day 1 timepoint. The presence of high density regions throughout the hAFS cell hydrogels may be due to the enhanced proliferative capacity of hAFS cells. Future work will use a proliferation assay and microscopy analysis at additional timepoints to investigate how PL gel incorporation impacts live cell number. Additionally, these studies aim to determine whether PL-induced changes in stem cell viability are due to alterations in the hydrogel structure or biochemical properties.

The incidence of Barrett’s esophagus with progression to adenocarcinoma is dramatically increasing and esophageal cancer has become the world’s sixth leading cause of cancer death. Overall morbidity and mortality rates associated with esophagectomy are substantial. Esophageal resection has been challenged as a necessary treatment for high grade dysplasia (HGD) and intramucosal cancer since lymph node involvement is unlikely (<5%), but, early lesions have the potential to be lethal and are only curable if completely removed. Screening has increased the number of patients detected with early stage disease and interest in less invasive endoscopic treatments has grown in parallel with early diagnosis.Unfortunately currently available treatments have significant limitations. Photodynamic therapy (PDT) is being abandoned due to photosensitivity-related side effects, recurrent disease, uncontrolled depth of ablation, and stricture formation. Early results of radiofrequency ablation have been encouraging, but just as with PDT there is no specimen available for histopathologic examination and the depth of ablation is limited to 500µm preventing its use for treatment of invasive cancer. Although recent data suggest that radiofrequency ablation may reduce the occurrence of subsquamous metaplasia when compared to prevalent pre-treatment cases, patients are committed to a lifetime of screening procedures and subsequent interventions. The use of endomucosal resection (EMR) has generated excellent results in treatment of high grade dysplasia and early adenocarcinoma with an extrapolated 5 year survival rate of 98% in selected patients. Currently used cap or snare EMR techniques limit resection size to 20mm, requiring piecemeal resections of larger lesions compromising histological assessment of radial resection margins. Long lesions require a stepwise approach to prevent stricture requiring multiple procedures and exhaustive follow-up.Esophageal resection is the standard treatment for Barrett’s esophagus with HGD and invasive malignancy. Despite a significant reduction in mortality reported by experienced centers, esophagectomy is associated with substantial morbidity rates approaching 50%. As a result, there has been an impetus to move towards esophageal preservation in patients with intramucosal neoplastic lesions in which lymphatic involvement is unlikely. A successful strategy that would allow for aggressive resection of full circumference mucosa with minimal risk of stricture would assure a quantum leap forward in the treatment of esophageal pathology. Biologic scaffolds composed of extracellular matrix (ECM) have been extensively studied in the context of regenerative medicine for their ability to modify the default healing response in numerous tissues including the esophagus. In a pre-clinical model, critically sized, full circumferential defects could be repaired with minimal stricture formation if adjacent autologous muscle tissue was placed in direct apposition to the ECM scaffold at the time of surgery. A follow-up pre-clinical study of esophageal transection mimicking the “gastric pull-up” procedure showed restoration of mature epithelium and islands of muscle that bridged the gap between native muscle tissues on either side of the surgical transection. Most recently, the use ECM was extended to a pre-clinical model of EMR such as that proposed for treatment of Barrett’s disease. A full-circumferential EMR of the cervical esophagus was performed endoscopically for a length of over a 5 cm. After two months, the injury treated with ECM was completely covered with a submucosal layer and stratified epithelium, and showed less stenosis than controls without ECM. Based upon these findings, five patients with stage1 neoplasia were treated with aggressive EMR and placement of an ECM scaffold material in the place of the ressected mucosa. The successful results of this clinical translation with 3 year follow up will be reported.

Strategies to overcome inherently poor elastin synthesis and hierarchical elastic matrix assembly by adult smooth muscle cells (SMCs) are critical to engineer compliant and functional small-diameter conduits for bypass grafting. This is especially important since intact elastic fibers regulate vascular SMC behavior via mechanotransduction. To overcome these limitations we propose to (a) recruit a potentially more elastogenic source of progenitor-derived, autologous SMCs in vivo within a rat peritoneal cavity, (b) employ 3-D electrospun scaffolds to align the cells and cell-generated matrix, and (c) functionalize the scaffolds with biological factors – i.e., hyaluronan oligomers (HA-o) – to provide a further elastogenic impetus to recruited cells. In this study we determined the role of pore size and electrospun fiber diameter on recruited macrophage phenotype (e.g. M1 and M2) and matrix synthesis. Conduits were electrospun from solutions containing 14, 17, and 22% w/w of poly(ε-caprolactone) (PCL) to produce mean fiber diameters of 0.62±0.4, 1.4±0.7, and 1.9±0.8 μm, respectively. The conduits were incorporated in non-adhesive PTFE pouches with different pore areas (0.04 and 0.36 mm2) and implanted for 14 days in the peritoneal cavity. The tensile moduli of the conduits in saline at 37°C, both before and after implantation, were similar to that of a human coronary artery. At 14 days, a highly cellular fibrous capsule had formed around the conduit containing aligned and elongated cells, which infiltrated into the mesh. Cellularity increased with fiber size (p=0.05) and pouch pore size (p = 0.02), as determined using one-way ANOVA with Tukey-Kramer comparisons. TEM suggested cells within the mesh/capsule to be macrophages and fibroblasts and those on the capsule surface, mesothelial cells. The cells produced significant amounts of collagen, HA, and fibrin. The elongated cells in the capsule stained for α-smooth muscle actin (α-SMA) but not SM22α, suggesting a myofibroblastic phenotype. Other cells expressed the macrophage marker CD68. Elastic matrix deposition was limited; thus, in ongoing studies we are evaluating elastogenic factor (HA-o) tethered conduits electrospun from a 25/75 % w/w collagen/PCL blend. Separately, harvested peritoneal grafts were autologously implanted in rat infrarenal aortae for 6 weeks. The grafts remodeled with the cells at the periphery exhibiting increases in collagen content and decreases in α-SMA expression, similar to an adventitia.  Also, there was significant intimal hyperplasia. Our study to date shows that myofibroblastic cells derived from peritoneal cell recruits are present within our conduits and that further remodeling by these cells occur following intra-aortal grafting. With graft inversion to provide a luminal mesothelium, tethering elastogenic factors, and longer implantation times in the peritoneal cavity, functional and patent vascular grafts can be potentially created.

Regenerative dentistry seeks to reverse morbidities that arise from genetic, traumatic, or infectious pathologies of the dentition; such maladies affect one in every three humans at birth and almost all people in a lifetime. As we turn to cell-based therapies for repair of missing teeth and damaged oral tissues, we look to developmental biology to understand how nature grew and renewed these tissues in the first place. What we know about tooth development or odontogenesis has been almost exclusively derived from the mouse model. While we have learned a great deal from this system, in contrast to humans and most other vertebrates, the mouse lacks a successional or replacement set of teeth. Therefore, we know relatively little about natural mechanisms of dental replacement. To discover how nature perfected the growth and renewal of teeth in an adult organism, we study the de-novo continuously replacing dentitions of Lake Malawi cichlid fishes. We first employed immunohistochemistry to describe organogenesis of the continually replacing cichlid dentition and discovered an epithelial down-growth, analogous to a successional lamina, with a proliferation bias that initiates the replacement cycle. Next, we identified sets of co-expressed genes active during de novo tooth replacement and tooth morphogenesis. Of note, we found two distinct cell populations expressing putative stem cell markers: one residing in the outer oral epithelium and the other within the intramedullary dental epithelium. Additionally, we identified signaling centers of gene expression associated with the differentiation of replacement teeth into complex cuspid forms. Finally, we manipulated targeted gene pathways (Bmp, Fgf, Wnt, Hh, Notch) with small molecules and demonstrated effects on both tooth replacement and tooth shape. Our data suggest a model wherein the process of tooth replacement is linked to tooth morphogenesis by a common set of signaling pathways. Our dissection of the molecular mechanics of vertebrate tooth replacement coupled with complex shape pinpoints aspects of odontogenesis that might be 're-evolved' by dental bioengineers to solve problems in regenerative medicine.

Electrospun non-woven structures have the potential to form bioresorbable vascular grafts that promote tissue regeneration in situ as they degrade and are replaced by autologous tissue. Current bioresorbable grafts lack appropriate regeneration potential since they do not have optimal architecture, and their fabrication must be altered by the manipulation of process parameters, especially enhancing porosity. We describe here an air-impedance process where the solid mandrel is replaced with a porous mandrel that has pressurized air exiting the pores to impede fiber deposition. The mandrel design, in terms of air flow rate, pore size, and pore distribution, allows for control over fiber deposition and scaffold porosity, giving greater cell penetration without a detrimental loss of mechanical properties or structural integrity. Air-flow pressure in a perforated mandrel (0.75 mm diameter holes laterally spaced 2 mm center to center, a center to center longitudinal distance of 1.5 mm, 6 mm diameter, and 8 cm in length) was increased over a range from 0-400 kPa during the electrospinning process. Scaffolds were first characterized for fiber diameter, pore diameter, and permeability. Results displayed a significantly larger pore size for both 50 and 100 kPa grafts compared to solid mandrel (control). This was confirmed with a significantly higher permeability for both 50 and 100 kPa compared to the control. Mechanical evaluation exhibited a slight decrease in burst strength properties for all air-flow grafts when compared to solid mandrel; however, these values were still physiologically relevant for an artery. Cell seeding was performed using human dermal fibroblasts over a 7 day period, and demonstrated a significant increase in cellular penetration of the graft wall for both static and pressure seeding. At day 7, 50 kPa, with an average wall thickness of 400 microns, displayed the highest overall value with a cellular penetration depth of 235 microns, while 100 kPa demonstrated an average value of 156 microns. As future studies are performed, we will investigate differences in mandrel pore size and pore spacing, where each are optimized using computational fluid dynamics to determine the proper air-flow rate for each design. Additionally, the authors of this work are investigating several other tissue applications other than vascular tissue, including dermal, bone, muscle, ligament, and nerve repair.

Improved refrigerated storage of natural and engineered blood vessels is needed for both research purposes and arterial bypass surgery. It is well documented that culture media and physiological salt solutions provide poor hypothermic preservation of blood vessel function resulting in significant loss within hours. We have been investigating the potential of employing sub-zero temperatures, without freezing, and insect-derived antifreeze proteins (AFP) to develop hypothermic storage solution formulations that provide longer short-term storage of functional vessels. Several recombinant AFPs were previously tested and Dendroides canadensis-derived AFP-10 was selected as the best with respect to retention of venous physiology. The purpose of the work presented here was to determine the optimal AFP-10 concentration and the duration that venous physiological functions were retained. We also investigated whether or not AFP-10 benefited arterial preservation at sub-zero temperatures. Recombinant insect AFP-10 was expressed in E. coli, purified and formulated in combination with 0.5M glycerol and 0.150M trehalose in Unisol™. Vascular tissue rings derived from rabbit jugular veins and carotid arteries were stored in this solution between -7 and -10°C for 6 or 12 days. Fresh untreated controls and storage solution controls without AFP-10 were also performed. Smooth muscle and endothelial cell physiology was studied employing drugs that induce either smooth muscle contraction (histamine and bradykinin or norepinephrine and phenylephrine), endothelial dependent (acetylcholine or calcium ionophore A21837) or independent (sodium nitroprusside) smooth muscle relaxation in an organ bath test system. The data was expressed as the mean grams tension ± 1 standard error, either for contractions or relaxations, of ≥7 samples from ≥2 donor rabbits. Veins stored with AFP-10 demonstrated AFP-10 concentration dependent preservation of smooth muscle functions in vessel rings stored for 6 days, endothelial mediated relaxations were observed in all storage groups but less than fresh controls (p<0.05). The best preservation, similar to untreated fresh controls, was observed at 1.5mg/mL. On day 12 of storage the contractions were reduced to ~50% of fresh controls (p<0.05) and decreased levels of endothelial-mediated relaxation were still present. Direct sodium nitroprusside mediated relaxation was retained at fresh control values. In contrast, there was no effect of AFP-10 on artery physiology after 6 or 12 days of storage. However, the storage solution control group without AFP-10 demonstrated similar physiological functions at both storage time points compared with fresh untreated controls. In conclusion, venous functions were preserved for 6 days in the presence of 1.5mg/mL AFP-10. In contrast, preservation of carotid artery functions did not require the presence of AFP-10. Control solution stored arteries retained physiological functions for 12 days. It is anticipated that the AFP-10 supplemented solution will provide at least 6 days of refrigerated storage of either veins or arteries. This hypothermic storage solution will be of value for short-term storage or transport of blood vessels used for physiology studies during drug development and as cell sources in tissue engineering. It may also be used for short-term storage of cellular tissue engineered vessels. Supported by NIH Grant #R44DK081233.

Temporomandibular joint (TMJ) disorders affect more than 10 million American patients per year. For approximately 50,000-100,000 of these patients, the only available treatment is removal of the joint meniscus. While these patients often report reduction of joint pain and increased joint mobility, absence of the meniscus results in significant wear of the articulating surfaces, leading to multiple costly corrective surgeries and a potential need for a full joint prosthetic. Currently, no effective alternatives exist for reconstruction of degenerative, non-reparable TMJ menisci following removal, and meniscectomy without replacement remains the “gold-standard”. In the present study, a device consisting of powdered urinary bladder extracellular matrix (UBM) encapsulated within sheets of the same was investigated as a scaffold for temporomandibular joint (TMJ) meniscus reconstruction. The device was designed to provide a resorbable interpositional “pillow” and a surrounding anchoring site. Five dogs were subjected to unilateral resection of the native meniscus and replacement with a UBM device. Briefly, following excision of the native meniscus, the conformable “pillow” was positioned between the mandibular condyle and temporal fossa, and the device was anchored to the temporal bone and adjacent musculature. Necropsies were performed at 3, 4, 8, 12, and 24 weeks to evaluate remodeling of the implanted device. Ten additional dogs were subjected to bilateral resection of the native menisci and replacement on one side with a UBM device, leaving the contralateral side devoid of a meniscal substitute. Macroscopically, the UBM-ECM implants were remodeled rapidly and were indistinguishable from newly deposited host tissue at all time points. Microscopically, remodeling was characterized by a dense infiltration of predominantly CD68+ mononuclear cells and smooth muscle actin-positive fibroblast-like cells at early time points changing with time to a sparse population of smooth muscle actin-negative spindle-shaped cells resembling those of the native TMJ meniscus. Furthermore, the remodeling process showed deposition of predominantly type I collagen, the density and organization of which resembled that of the native meniscus by the 24-week time point. Ingrowth of calsequestrin-positive skeletal muscle tissue was also observed at the periphery of the remodeled UBM device and was similar to that found at the peripheral attachment site of the native meniscus to the surrounding soft tissues. Biomechanical testing showed that the mechanical properties of the remodeled UBM device were similar to those of the native meniscus at 24 weeks post-implantation. No adverse changes in the articulating surfaces of the condyle or fossa were observed in UBM-implanted joints. Results of this study suggest that the UBM device may represent an effective, off-the-shelf interpositional material while also serving as an inductive template for reconstruction of the TMJ meniscus.

In response to vascular injury the wound healing response is activated and stimulates the coagulation cascade leading to localized deposition of the provisional matrix fibrin.  Though fibrin serves as a successful hemostatic in most injury situations this system often fails in trauma situations leading to major hemorrhaging as a result of dilution of critical factors required for fibrin polymerization.  In this study we aim to exploit natural fibrin polymerization towards the development of bio-synthetic hybrid polymer systems with superior hemostatic and provisional matrix properties.  We have designed hydrogel microparticles (microgels) that can be finely tuned with respect to their self-assembly characteristics and upon triggering by biological cues, such as contact with deposited fibrin, rapidly form strong, swelling highly interdigitated networks.  In order to impart fibrin specificity to the microgels, humanized synthetic single chain variable fragment (scFv) antibodies with high affinity for fibrin were identified and coupled to the microgels.  ScFv antibodies with high affinity for fibrin were identified using two phagemid libraries in biopanning assays against fibrin.  Clones from each library were isolated, amplified, and induced to produce only the scFv fragments. Using fibrinogen and fibrin-based ELISAs, we identified 10 clones per library that displayed high affinity to fibrin and low affinity to fibrinogen.  The specificity and binding affinity of each clone to fibrin was evaluated with surface plasmon resonance and the clone found to have the highest affinity for fibrin was conjugated to the microgels using EDC/NHS coupling.  The scFv-conjugated microgels represent a novel hemostatic system that utilizes the bodies’ native clotting cascade while enhancing its overall effectiveness by concentrating clotting factors and generating compressive force through triggered swelling.  Because of the potential integration of microgels with the resulting blood clot, this material could additionally assist in repair and regeneration of the damaged tissue.

Cell delivery for regenerative medicine is often achieved using a matrix, scaffold, template, or vehicle.  It is becoming increasingly common to functionalize such materials with various biomolecular components to provide carefully tuned combinations of microenvironmental cues.  However, the design rules necessary for properly interfacing such complexly biofunctional materials with the immune system remain unclear.  Our group’s interest is in developing self-assembling biomaterials from peptides and proteins for cell culture and delivery.  We have previously developed the fibrillizing peptide Q11 as a modular domain for predictably assembling different biofunctional peptides within chemically defined hydrogels useful for the culture and encapsulation of several cell types.  However, we have also found that the immunogenicity of fibrillized Q11 peptides varies significantly, depending on the ligand or epitope attached to Q11. In the work reported here, we investigated the mechanisms by which these materials engage adaptive immune processes, and whether such mechanistic insight could be used to modulate, attenuate, or ablate the materials’ immunogenicity. 

We synthesized a series of self-assembling peptides with the fibrillizing domain Q11 at their C-termini and various N-terminal functional components, including RGD (RGD-Q11), the OT-II antigen from ovalbumin (OVA-Q11), and a short epitope peptide from P. falciparum (Pfal-Q11), at their N-termini.  Each peptide self-assembled into nanofibers about 15nm wide and several hundred nanometers long in physiological buffers, and they formed hydrogels at millimolar concentrations.  OVA-Q11 and Pfal-Q11 raised strong and durable antibody responses in C57BL/6 mice lasting at least one year, whereas no detectable antibodies were raised against Q11 or RGD-Q11.  Using T cell adoptive transfer experiments in mice and transgenic mice lacking functional T cell receptors, we found that these strong antibody responses were entirely dependent on CD4+ T cells, suggesting a route for avoiding immunogenicity.  We also investigated the immunogenicity of these peptides in transgenic mice having non-functional Toll-like receptor-2 (TLR-2), TLR-4, TLR-5, MyD88, and NALP3.  Q11-peptide immunogenicity was strongly dependent on MyD88, but not on any of the Toll-like receptors investigated, nor on NALP3, a component of inflammasome signaling.  Thus, after surveying several candidate pathways, it was concluded that proper epitope restriction and T cell help were required and were the most appropriate targets for modulating the peptides’ immunogenicity. Indeed, by deleting amino acid regions in the OVA-Q11 peptide known to be recognized by T cells, immunogenicity could be ablated.  These findings will facilitate the design of highly oligomerized and multifunctional biomaterials for regenerative medicine, where immunocompatibility is a significant concern. 

In this study we show that pharmacological inhibition of S1P3 via VPC01091 significantly increases mobilization of BMSCs into peripheral blood resulting in accelerated bone repair in rat cranial defects. Additionally, MSCs pretreated with FTY720 exhibit increased migration towards SDF-1, a CXCR4+ ligand and critical component of the bone marrow niche. These findings advocate the significant role of S1P3 in stem cell chemotaxis. We show that treating animals with both FTY720 coated allografts locally, and VPC01091 systematically is beneficial if controlled temporally. We propose that S1P3 receptor antagonists aids in the mobilization of MSCs, while agonists of the same receptor are critical for stem cell recruitment. Thus, suggesting the presence of a push-pull mechanism that is dictated by S1P receptor specific small molecules. Transwell assays were conducted on MSCs treated with FTY720 to assess migration towards SDF-1. 5mm cranial defects were made in 36 nine weeks old Sprague Daley rats, which were divided into 4 groups (n=9). The rats were treated with a systemic dose of 1 mpk VPC01091, FTY720 coated semi-circular allograft, FTY720 coated semi-circular allograft + a systemic dose of 1 mpk VPC01091 or left untreated. VPC01091 was given the day after surgery and 3 weeks post surgery. Hemavet (Drew Scientifics) was used to measure the concentrations of blood cells at days 0, week 1 and week 2 (n=6) (data not shown). The amount of bone regeneration was measured bi-weekly with microCT imaging (n=3-9). Flow cytometry was performed on the tissue harvested from the defect sites at week 3 (n= 3), and from peripheral blood at week 6 (n=3). Monoclonal antibodies rat CD45, CD11b, CD54, CD90 were used in both cases. Mason’s Trichrome and H&E staining were done for done for all groups (n=3). Treatment with systemic VPC01091 resulted in substantial bi-weekly increase in bone regeneration compared to the empty defect controls. This group also showed an increase in the % of CD54 and CD90 positive cells (rats MSC markers) in the defect region at week 3 and in the blood at week 6. Animals treated with FTY720 allografts showed a temporal response to VPC01091. Initially, they showed lesser bone growth compared to just FTY720 treatment, but the trend reversed after 4 weeks. These results indicate that a systemic treatment with VPC01091 will significantly accelerate bone regeneration in the absence of any local implant. However, the effectiveness of locally released FTY720 to promote healing requires recruitment of BMSCs via S1P3, suggesting that the time of systemic delivery of a S1P3 anatgonist is crucial for the body to engage in this push-pull mechanism of endogenous stem cells. This manifests in the fact that the rate of increase in bone volume at later time points is the highest for the group treated with both FTY720 allograft and VPC01091. The presence of an increased number of MSCs both in the blood, and defect region tissue denotes that the cells required for bone healing are being mobilized into the blood, and recruited to the defect site as late as 6 weeks after injury. Thus, this study shows that the rate of bone growth in large defects can be controlled by a combination of S1P receptor specific small molecules in a time dependent manner.

Cell-based therapy has been viewed as a promising alternative to organ transplantation, but cell transplantation aimed at organ repair is not always possible. We propose the lymph node as a thriving environment for normal cell engraftment and growth. Direct injection of hepatocytes into a single lymph node generates enough ectopic liver mass to rescue mice with lethal metabolic disease. Furthermore, thymuses transplanted into a lymph node of athymic nude mice provide a functional immune system capable of rejecting tumor growth. Finally, pancreatic islets injected into the lymph node of diabetic mice increase survival and improve glucose levels. Collectively, these results validate a novel practical approach of targeting lymph nodes to restore, maintain or improve tissue and organ functions.

Cardiovascular disease accounts for approximately 36% of deaths in Europe and the United States, with myocardial infarction and heart failure representing 80% of cardiovascular deaths. Heart failure is a clinical syndrome caused by ventricular remodelling due to injury to the myocardium that results in reduced cardiac output and a reduction in the heart’s ability to meet functional requirements. There is no cure for this progressive condition. It is proposed that miRNA based therapeutics can be implemented to target specific genetic pathways to reduce the onset of hypertrophy in heart failure and may be of significant clinical relevance in modulating the negative myocardial remodelling seen after injury. This project aims to assess the efficacy of ultrasound mediated microbubble delivery of miRNA mimics to cardiomyocytes to enhance the therapeutic potential of miRNA modifiers in heart failure. Cationic lipid microbubbles were prepared and characterized pre- and post-complexation with pGFP or miRNA mimics. The HL-1 cardiomyocyte cell line was transfected initially with a pGFP-lipid microbubble complex using defined ultrasound parameters.  A transfection efficiency of 76.53±2.24% was achieved using this regime in comparison to 29.2±2.84% & 38.53±6.26% using liposome or commercially available lipids respectively. Using these optimal parameters fluorescently tagged miRNA-mimics were delivered to HL-1 cells with an uptake efficiency of 29.56±3.44%. The miRNA therapeutic formulation was then adapted to target some of the basic pathophysiology to be corrected in heart failure, particularly the reduction of hypertrophied mechanically inefficient cardiomyocytes by using targeted delivery of miRNA-133 mimics. miR-133 expression has been shown to be down-regulated by 50% in diseased ventricles in human samples obtained from heart failure patients and may be a target to reverse the hypertrophic phenotype. The delivery of the therapeutic microRNA – miR-133 using the ultrasound responsive microbubble regime to endothelin induced hypertrophic cardiomyocytes resulted in a significant reduction in the hypertrophic phenotype to baseline levels. This data demonstrates the potential of ultrasound responsive microbubble mediated delivery of miRNA mimics to cardiomyocytes. It has the potential to act as a novel delivery platform for miRNAs which may enhance the efficacy of such therapies in the future.
 
 

This talk will describe the platform technologies under development in the Duvall lab that harness “smart” polymer functionalities for drug delivery in regenerative medicine applications. First, a technological platform will be introduced that enables temporally-controlled, efficient delivery of siRNA from an injectable polyester urethane tissue engineering scaffold. This porous, biodegradable, and biocompatible platform has been shown to provide an effective tissue template and to also enable tunable release of bioactive siRNA carriers that can mediate local gene-specific silencing. Recent progress will be discussed from experiments to validate this system in vivo and on application of this technology for regenerative purposes in the setting of wound repair. The Duvall lab is also pursuing delivery systems for intracellular delivery of a therapeutic peptide to inhibit intimal hyperplasia of transplanted vascular grafts. A “smart” nano-micelle platform for efficient, intracellular delivery of this therapeutic peptide for improving graft function will be introduced. Finally, new technologies will be described that have been designed for cell-demanded and environmentally-targeted drug release triggered by the increased levels of reactive oxygen species in damaged and diseased tissues.

ESCs and iPSCs have untapped potential for the development of cell-based therapies to replace damaged ventricular myocardium and impact the enormous clinical need for treating injured or failing hearts. While robust protocols exist to differentiate stem cells into beating cardiomyocytes, these cells are functionally immature, and currently offer little clinical benefit. The ideal myocardial tissue fabrication strategy would follow the naturally rapid self-organization process contained in embryonic development, where the simultaneous differentiation of ventricular, atrial, and conduction system lineages occur. Here we test the hypothesis that developmentally based genetic and microenvironmental cues will enhance cardiac specific lineage restriction and chamber specific fate in engineered tissues. First, we discovered that conditional expression of the transcription factor Gata5 is sufficient to dramatically enhance, as much as 1000-fold, expression of cardiac progenitor markers in embryoid body (EB) cultures. Gata5 directs cardiac fate in the absence of serum or previously used cardiogenic growth factors. 100% of the Gata5-induced EB derivatives display extensive beating clusters that stain for cardiac markers. Expression of Gata5 directs development of atrial and ventricular cells at least 250-fold compared to controls. The addition of cardiac-inducing growth factors at any point in the assay, or in any combination, provides no benefit over simply inducing the expression of Gata5. This suggests that Gata5 functions downstream of known procardiac signaling programs and is therefore sufficient to direct cardiac specification and generation, providing a platform to screen for factors that promote differentiation and maturation of defined fates. We are currently testing the effects of 3D culture conditions, including collagen type 1 and various hydrogel constructs, on promoting differentiation and maturation. In addition, given that cells in the cardiovascular system are subjected to mechanical forces, this platform is ideal for testing the role of shear stress or cyclic strain on phenotype outcome. Preliminary analysis demonstrates changes in the profile of the cardiac differentiation markers for cells seeded onto different culture conditions. Finally, we have developed methods to decellularize embryonic myocardial tissues, using embryonic chick as a model system. We tested the effects of detergents SDS, Triton-X100, and CHAPS on cell and matrix removal in three different stages of heart development (embryonic day 4, 7, and 10), using histology, DNA and BCA protein assays. We show that embryonic myocardium was at least 90% decellularized. Cells seeded on these novel material scaffolds adhere and migrate within the matrix. Chick extracellular matrix is a promising biomaterial source for directing stem cell fate in 3D because it is inexpensive, naturally non-toxic and non-immunogenic.

Combat related blast injuries have resulted in service members sustaining multiple extremity injuries and amputations. Improvements in body armor, faster transport to treatment facilities and improved resuscitative techniques has translated into increased patient survivability and more service members with complex wounds. Following initial injury treatment, patients are transported by the aeromedical evacuation system for definitive management stateside. Greater than 70% of the injuries sustained during these conflicts involved the musculoskeletal system and cause the majority of disability. Massive skin and soft tissue loss are common challenges in the management of multiple extremity-injured patients. Studies suggest that significant soft tissue loss is one of the most frequently encountered injuries. Furthermore, soft tissue losses can directly lead to a compromise in the maintenance of a functional limb. Currently, the standard treatments for extremity injuries with massive tissue loss include a combination of therapies: Limb amputation, limb shortening to assist in stump closure, free tissue transfers, pedicle flaps, local flaps or split thickness skin grafting. However, these procedures may result in decreased functional outcomes, significant donor site morbidities, and non-durable surface areas prone to erosive wear with prosthetic use. Moreover, as a number of our wounded warriors have multiple limb injuries and amputations, the common accepted donor sites for autologous tissues are becoming increasingly limited. The recent introduction of regenerative medicine has provided an important role in dealing with soft tissue injuries in the war wounded. Extracellular matrices offer numerous benefits to these patients including: decreased wound or wear related breakdown with prosthetic use, a viable dermal substitute layer, thereby converting full thickness wounds into partial thickness defects, and a protective layer underneath the overlying epidermis, thus reducing adherence to underlying anatomic structures such as muscle, tendon, or bone. This purpose of this report is to describe our current applications of regenerative medicine in the management and treatment of complex injuries sustained by service members injured in the wars in Afghanistan and Iraq. Through case reports and clinical images, it details the main outcome measures including a description of the mechanism and location of injuries, number of surgical interventions, regenerative medicine application employed and clinical outcomes. It illustrates that regenerative medicine techniques are effective in managing complex combat-related injuries. Lessons learned can be adapted and transferred to civilian care. The views and opinions expressed herein are those of the author and do not necessarily reflect those of the United States Navy or US Government.

The outcome of stem cell therapy for the heart will inevitably depend on inflammatory signals and the type of inflammatory cells at the site of injury. Following myocardial infarction, tissue repair is mediated by the recruitment of monocytes and their subsequent differentiation into macrophages. Recent findings have revealed the dynamic changes in the presence of polarized macrophages with pro-inflammatory (M1) and anti-inflammatory (M2) properties during the early (acute) and late (chronic) stages of cardiac ischemia. The presence of macrophages following myocardial infarction is clearly bimodal, with a strong M1 response at the onset of inflammation followed by a dominated M2 response. Mesenchymal stem cells (MSCs) delivered into the injured myocardium, alone or within a cardiac patch, are subjected to polarized macrophages and the inflammatory milieu. The present study investigated how M1 or M2 environments affect the survival and function of MSCs in vitro. A screening platform was developed to enable parallel co-cultures of MSCs and polarized macrophages with and without direct cell contact. Human MSCs were cultured with combined M1 and M2 cytokines: IL-1β, IL-6, TNF-α, and IFN-γ (for M1), and IL10, TGF-β1, TGF-β3, and VEGF (for M2). Three different concentrations of each cytokine were tested: 12.5, 25, and 50 ng/ml. After 48 hours, hMSCs were lysed and DNA quantified. For indirect and direct co-cultures, human THP-1 monocytes and primary human monocytes were differentiated and polarized into M1 and M2 macrophages by the addition of LPS/IFN- and IL-4/IL-13 respectively. The effect of macrophage polarization on the growth and function of mesenchymal stem cells was determined by measuring cell growth and differentiation potential. The presence of M1 cytokines decreased cell numbers when compared to controls, while the presence of M2 cytokines resulted in either no change or increase in cell numbers. IL-1β and IFN-γ showed a significant reduction in the number of cells while TNF-α, VEGF, and TGF-β3 showed an increase in cell number. Overall, the M1 cytokines inhibited the growth of hMSCs in vitro in a dose dependent manner, whereas M2 cytokines supported the growth of hMSCS in vitro. In either indirect or direct co-culture, hMSCs showed better survival in the presence of M2 as compared to M1 macrophages. The present study found a consistent increase in the number of MSCs during direct co-culture with M2 macrophages over M1 macrophages. Similar results were obtained using indirect (transwell) co-culture of human MSCs and polarized macrophages suggesting a supportive paracrine effect of M2s on MSC growth.

Hereditary diseases such as muscular dystrophy, cystic fibrosis, sickle cell disease, and hemophilia currently have no cure and can only be managed by attempts to alleviate the symptoms. For decades, the field of gene therapy has promised a cure to these diseases. However technical hurdles regarding the safe and efficient delivery of therapeutic genes to cells and patients have limited this approach. To address these limitations, we are engineering synthetic nucleases to alter the DNA sequence at specific genomic locations. These nucleases are based on artificial DNA-binding proteins that can be programmed to target any desired DNA sequence. The nucleases cleave the target DNA sequence and activate endogenous cellular DNA repair pathways to facilitate gene editing by non-homologous end-joining (NHEJ) or high-fidelity homology directed repair (HDR). With this technology it is possible to edit the human genome and correct hereditary mutations that cause degenerative diseases. Additionally, therapeutic transgenes can be integrated into “safe harbor” sites in the human genome to avoid adverse effects of random gene addition to the cellular genome, such as insertional mutagenesis, and misregulation of the transgene. These genome editing approaches circumvent several of the obstacles that have hindered the gene therapy field. We are developing this genome editing technology as a therapeutic approach to Duchenne Muscular Dystrophy (DMD). DMD is a recessive, X-linked hereditary disease caused by mutations in the gene encoding dystrophin, an essential musculoskeletal protein. We have designed and assembled a panel of nucleases against sequences in or around exon 51 of the human dystrophin gene, which is a hotspot for DMD mutations. Nucleases were screened by surrogate reporter assays of gene editing activity. Several nucleases were further shown to drive gene editing at the endogenous locus in human cells isolated from DMD patients. The efficiency of genetic modification was approximately 5-10% of treated cells. This includes the creation of random small insertions and deletions by NHEJ, the targeted deletion of mutated sequences by NHEJ, or the precise insertion of short 6-19 base pair tags by HDR. The modifications to the dystrophin gene that we have created are predicted to correct functional dystrophin expression, and current work is investigating proper protein expression and phenotype of these corrected patient’s cells. In other experiments, we have used engineered nucleases to direct the chromosomal integration of an engineered minidystrophin expression cassette, which is currently in clinical trials, to a safe harbor locus in mouse and human cells. Collectively, this work represents a new and promising approach to correcting inherited degenerative diseases.

Myostatin is a TGF-β superfamily member known to play an important role in regulating muscle growth and development. Myostatin binds the activin type IIB receptor (ActRIIB, or Acvr2B), activating Smad2/3 signaling and regulating the expression of several myogenic genes. We have found that mice lacking myostatin show increased Sox-5 expression and increased fracture callus size following fibula osteotomy, suggesting that myostatin might play a role in chondrogenesis. Indeed, we have also reported that myostatin suppresses chondrogenesis and chondrocyte proliferation in vitro. Recent studies suggest that myostatin can activate the p38 MAPK, Erk1/2, and Wnt pathways in skeletal muscle, and it has also been shown that TGF-β signaling activates the Wnt/β-catenin signaling pathway and stimulates Sox9 expression during chondrogenesis. We used an in vitro cartilage aggregate culture system to determine the effects of myostatin treatment (100 ng/ml for six days) on wnt signaling pathway activation. Our Wnt array expression screen showed that Dkk1 expression was increased by 150% (P<0.01) in aggregates treated with myostatin compared to PBS control. Dkk1 inhibits the Wnt signaling pathway by binding to and antagonizing LRP5/6, forming a ternary complex with LRP6 that induces its rapid endocytosis and removal from the plasma membrane. GSK3A, a component of the GSK3/Axin/APC proteosomal complex responsible for ubiquitination of β-catenin to prevent its accumulation in the cytoplasm and translocation into the nucleus, was upregulated by 136% (P<.05) in the presence of myostatin. Frizzled 3, a primary Wnt/β-catenin receptor, was downregulated by 120% (P<0.01) with myostatin treatment. Fzd3 is a transmembrane protein that binds Wnt ligands including Wnt5a, and Wnt 5a is known to be a molecule that plays a key role in chondrogenesis. The downregulation of this receptor in the presence of myostatin during TGF-β1-induced chondrogenesis further supports the hypothesis that inhibition of Wnt-mediated signaling is one possible mechanism by which myostatin exerts its negative effects on the chondrogenic differentiation of BMSCs. Together, these data provide additional evidence that myostatin can alter wnt signaling in a variety of different cell types of mesenchymal origin. Funding for this research was provided by the Office of Naval Research, the Congressionally Directed Medical Research Programs, and the National Institutes of Health.

Tendons are connective tissues that transmit loads between bone and muscle. We are developing anisotropic collagen-GAG (CG) scaffolds containing aligned tracks of ellipsoidal pores for regenerative repair of tendon and its insertion into bone. For this work we hypothesized that coincident patterns of soluble factors and ECM proteins/minerals can regulate MSC lineage specification towards, and long-term transcriptomic and functional stability of distinct tendinous, osseous, and fibrocartilagenous (interface) phenotypes. Here we describe a new class of anisotropic CG biomaterial for tendon repair and further detail the use of overlapping patterns of biomolecules and/or mineral content to improve scaffold bioactivity. Homologous series of anisotropic CG scaffolds that display an aligned 3D pore microstructure can be fabricated with sufficient mechanical competence to mimic elements of native tendon. We investigated the influence of scaffold pore size and relative density on its biophysical properties (mechanics, permeability) and resultant cell bioactivity. Anisotropic scaffolds with large (230 vs. 50, 150 µm) pores support increased tenocyte (primary equine) alignment, proliferation, and soluble collagen synthesis; increased and sustained expression of tenocyte genes (tenascin-C, scleraxis, COMP, decorin); and increased soluble collagen synthesis during in vitro culture compared to 2D culture and isotropic scaffold controls. Scaffold relative density significantly influences this response; when relative density was decreased to where tenocyte-mediated contraction could contract the scaffold and destroy its anisotropy, cell alignment and tenocyte-specific gene expression profiles were rapidly lost. Conversely, more dense scaffolds prevent contraction and support long-term tenocyte alignment, transcriptomic stability, and soluble collagen synthesis. Tenocyte bioactivity is further linked to biomolecule supplementation. While soluble supplementation and insoluble sequestration of PDGF and IGF-1 both drive proliferation in dose-dependent manners, tenocyte-specific gene expression is lost; conversely bFGF and GDF-5 maintain tenocyte-specific gene expression without proliferation. We have demonstrated that combinations of proliferative and phenotype factors, namely IGF-1 and GDF-5, can drive simultaneous tenocyte proliferation and function. For tendon-bone junction repair, we have integrated the anisotropic CG scaffold with a mineralized CG (CGCaP) scaffold using a previously developed synthesis process. This layered scaffold contains distinct tendinous (anisotropic CG) and osseous (isotropic CGCaP) compartments and a continuous interface; human MSCs cultured in growth media in this layered scaffold show divergent osteoblast vs. tenocyte differentiation profiles, suggesting a path towards a functional interfacial tissue. Ongoing work is exploring the use of sulfated and non-sulfated GAG content to modulate transient biomolecule sequestration and release as well as the use of tensile stimulation to improve tenocyte and interface bioactivity.

Cell therapies are gaining more and more evidence for efficacy in medicine, and most of them are being delivered to the patient through tubular instruments such as needles and catheters. Not withstanding sheer and fluidic forces that may influence the cells, it is important to consider the cell interactions directly with the materials that compose the delivery devices. In fact, there is a whole other level of device compatibility, deemed cytocompatibility, that is often overlooked and in need of better definition and standardization. Within the delivery devices, cells may adhere to the surfaces and be lost, may get stimulated to die by a cytotoxin, or may simply get activated or pushed down an undesired differentiation pathway. To detect these failure modes, factors such as which cell types or populations of cells to use for testing, what exposure times are relevant, and what specific functions to measure all must be considered. We have chosen to examine a few aspects of cytocompatibility, including cell counts, metabolic assays, and toxicity studies, by comparing common catheter materials to various controls. We would like to stimulate the dialogue of how standardization of these tests can be achieved and what is necessary and reasonable.

Bone Marrow Stem Cell (BMSC) differentiation is regulated through the cytokine SDF-1. Interaction with its receptor CXCR4 directs the migration of stem cells in injury repair. In vitro pre-conditioning with SDF-1 prior to transplantation enhances cell engraftment and new tissue formation; however, it does not allow for direct modulation of SDF-1 in vivo. Tetracycline-dependent regulatory systems provide for tight control of transgene expression in vitro and in vivo. SDF-1α is the most abundant splice variant, however, SDF-1β is twice as potent and has a longer half-life due to heparin sulfate binding properties. Recently, we described the development of a Tet-Off-SDF-1β transgenic stem cell line conditionally over-expressing SDF-1β. The current study evaluated the regenerative potential of this novel cell line using in vitro (e.g. osteogenic assay, signaling pathway assessment, gene expression qRT-PCR, cell migration) and in vivo analyses. In vitro osteogenic differentiation of Tet-Off-SDF-1β BMSCs was assessed using ELISA (osteocalcin; OCN), Alizarin red S (ARS) staining, and qRT-PCR analysis (e.g., OCN, Runx2, Col1) after culturing cells in differentiation medium ±BMP-2 for 21 d. Doxycycline (Dox) was added to suppress SDF-1β transgene expression in controls. For in vivo studies, lethally irradiated 6-month-old C57BL/6 male mice were given tibial transplants (left: 9.24x105 cells/70 µl; right: saline for vehicle control). Control mice had access to Dox in drinking water +5% glucose. After 4 wks, animals were euthanized and tibias collected for BMD analysis and 3-D histomorphometry with µCT (e.g., BV/TV, Tb.N).ELISA of Tet-Off-SDF-1β BMSCs (-Dox) revealed increased OCN production relative to controls. SDF-1β transgene expression accelerated and significantly enhanced osteogenic differentiation (±BMP-2) compared to controls, detected by ARS staining (p<0.001). qRT-PCR confirmed that SDF-1β transgene expression enhanced BMP-2-induced upregulation of osteogenic markers (e.g., OCN: 17.2-fold/-Dox; 6.5-fold/+Dox; p<0.05). Tibial µCT analyses showed significant augmentation of new bone formation in experimental and control transplants relative to vehicle controls (p<0.0001). SDF-1β transgene expression resulted in even greater bone formation compared to controls (e.g., BV/TV: 46.0/-Dox; 28.5/+Dox; p<0.05).In vitro and in vivo analyses indicate a critical role of SDF-1β in BMP-2-induced osteogenic differentiation and high regenerative potential of our novel Tet-Off-SDF-1β transgenic stem cell line.

Chronic kidney disease (CKD) is a global public health concern involving progressive loss in renal function over a period of months or years. Disease progression typically leads to dialysis and eventually a kidney transplant. An urgent need exists for new treatments to restore renal function thereby delaying or eliminating the need for dialysis and transplant. Regenerative medicine approaches are a logical next step in addressing this issue. Tengion’s unique integrated regenerative medicine technology platform has generated products that catalyze regeneration of tissues and organs. This study reports on the development of Neo-Kidney Augment (NKA) prototypes using selected regenerative renal cells (SRC) and natural biomaterials for catalyzing kidney tissue regeneration in nephrectomy models of CKD in two animal species. SRC have previously been shown to delay disease progression in rodent nephrectomy models of CKD. SRC are obtained from enzymatic digestion of a kidney biopsy and density gradient separation of cells. Natural biomaterials tested included gelatin based hydrogels and hyaluronic acid. In vivo response to NKA implantation was evaluated by microinjection into the kidney parenchyma of a 5/6 nephrectomized female Lewis rat model of CKD. Treated rats were monitored for renal specific, whole animal, and terminal histological parameters for 3 months post implantation. A modified canine surgical nephrectomy model was used for evaluation in the large animal dog model of CKD. Study animals underwent a 2-step surgical nephrectomy procedure that first removed poles and additional cortical tissue from one kidney, followed by a brief recuperation period, and then complete removal of the remaining intact kidney. Within 14 to 15 weeks post-nephrectomy, using IRIS standards for staging of canine CKD, the model resulted in a late Stage I/early Stage II CKD status. SRC and biomaterials were delivered to nephrectomized animals via injection targeting the corticomedullary junction of the remnant kidney at onset of disease (fifteen weeks post-nephrectomy). Dogs were followed through approximately one-year post-nephrectomy period by monitoring renal specific, whole animal, and terminal histological parameters. We show that gelatin-based hydrogels are biocompatible with rat kidney tissue, producing minimal inflammatory and fibrotic responses, and facilitating neo-vascularization when delivered into the kidney. Implanted NKA prototypes elicited regeneration of renal structures to replace the area occupied by biomaterial in vivo. Histological evaluation in the dogs (47-week post implantation) revealed that the cell biomaterial prototype was well tolerated. In conclusion, our observations suggest that selected renal cells and gelatin based natural biomaterials can be used for neo-kidney tissue regeneration in chronic kidney disease.

Building a clinically relevant sized tissue or organ using cells requires maintenance of viable cells until host vasculature is established and integrated into the implanted engineered constructs. However, delay in vasculogenesis results in premature cell death due to the inadequate supply of oxygen and nutrients. Several strategies have been proposed to overcome this challenge; however, none has demonstrated clinical relevance to date. One potential solution is to develop methods to maintain cell viability over a long-term by downregulating cellular metabolism until host vascularization is established. Adenosine, a nucleoside which functions as an energy transferring molecule, is reported to increase during hypoxia and functions as a modulator of ion-channel arrest. This results in a decrease in ATP consumption thus, oxygen demand. In this study, we attempted to promote cell survival under hypoxic conditions by exploiting this property of adenosine.
 
500 µL of a cell suspension (C2C12 cells, murine myoblasts) in high glucose DMEM containing 10% FBS, 500 U/mL penicillin and 500 µg/mL streptomycin was placed in each well of a 48-well culture plate at a density of 1052 cells/cm2. Cells were incubated for 24 hr. in normoxic conditions (21% O2, 37°C) prior to placement in a hypoxic chamber. At day 0, the plates designated as the hypoxic group were transferred to the hypoxic chamber (0.1% O2). This group was incubated for up to 7 days under hypoxia receiving daily doses of adenosine (0, 0.25, 1 and 5 mM), then placed back into normoxic conditions without additional supply of adenosine. Media to be used under hypoxia was deoxygenized in the hypoxic chamber 24 hr. prior to its use. The metabolic activity of viable cells at each pre-determined time point was assessed using an MTS assay, which measures mitochondrial activity of cells.
 
The metabolic activity of cells grown in normoxic conditions increased linearly with respect to time. Hypoxic cells not treated with adenosine showed a similar pattern of increasing metabolic activity for 7 days under hypoxia, but this resulted in eventual cell death. However, when treated with adenosine, cells under hypoxic conditions maintained a steady state of metabolic activity and these cells resumed their normal metabolic activity instantly when they were returned to normoxic conditions and the adenosine was removed at 7 days. As the dose of adenosine increased from 0 to 10 mM, an escalation of steady hypometabolic state was maintained under hypoxic conditions, and the cells were able to resume their normal metabolic activity after 7 days. This demonstrated that the effects of adenosine on cellular metabolic activity are dose dependent.
 
In this study we demonstrate the novel concept that cell viability can be maintained by downregulating cellular metabolism under hypoxic conditions. Application of adenosine to cells under hypoxic conditions prolongs survival by decreasing the metabolic activity to a steady hypometabolic state, thus reducing O2 demand. This concept represents a novel method for increasing cellular survival in tissue-engineered constructs during vasculogenesis.

 Multipotency of stem cells makes them an attractive, alternative source of cells from which functional T cells can be generated by in vitro differentiation.  Together with soluble factors, two key insoluble signals presented in the thymus by the thymic cells play crucial roles in generating functional T cells: (a) Delta-like ligands(DLL, Notch ligands)-Notch receptor signaling and (b) Major Histocompatibility Complex (MHC)-T cell receptor (TcR) signaling.  Most in vitro T cell differentiation efforts to date focus on mimicking these signals in two dimensional coculture with stromal cells expressing signaling molecules.  Thus, there is a need for a stroma-free system that supports T cell differentiation in a more controlled manner for therapeutic and clinical purposes.  Previous studies show that when Notch ligands are presented on a biomaterial surface, direct contact with stromal cells is not required to drive differentiation of hematopoietic stem cells into early T cells.  To further differentiate these early T cells into antigen-specific, cytotoxic CD8+ T cells, MHC/T-cell receptor (TcR) signaling is required.  Our data demonstrates that LCMV GP34-specific CD8+ T cells can be generated from mouse ES-cell derived early T cells when cultured with GP34 peptide-loaded MHC Class I tetramers or on GP34 peptide-loaded OP9-DL1 monolayer.  These MHC tetramers have been shown to provide graded differentiated signals in a concentration-dependent manner.   In addition, we have demonstrated that human CD34+CD38- cord-blood (CB) derived hematopoietic stem cells (HSCs) can be differentiated into early T cells using immobilized Notch ligands.  These early T cells give rise to CMV-specific CD8+T cells after incubation with CMV-pp65 HLA tetramers or peptide.  CD8+CMV+ T cells co-cultured with CMV peptide-loaded target cells in a CTL assay secreted cytokines and up-regulated surface expression of a cytotoxic marker, indicating functionality.To investigate a possible mechanism for in vitro positive selection with MHC/HLA tetramers, we performed TCR Vb repertoire analysis.  CMV-specific CD8+ T were found to be polyclonal, with little variation among CB samples, suggesting that TCR editing could be involved in positive selection.     
We are now translating these two signaling pathways into 2-dimensional and 3-dimensional biomimetic, soft material-based systems by fabricating flat hydrogel surfaces or porous hydrogel scaffolds that can present immobilized DLL or MHC molecules to seeded cells. A polyacrylamide-based gel of different material stiffness was modified to immobilize DLLs at varying, controllable density.  Mouse hematopoietic stem cells seeded on these gels of varying material stiffness and different DLL-densities differentiated into early T cells.  Three dimensional, inverse opal porous scaffold from poly(ethylene glycol) functionalized with binding moieties such as protein A or streptavidin were fabricated.  We were able to control and vary the density of Notch ligands and MHC molecules on these scaffolds, as well as induce Notch signaling of varying strengths in cells seeded within the scaffolds.  This system allows for the controlled presentation of ligands essential for T cell differentiation in a stromal cell-free system.  We hypothesize that the generated 2D and 3D soft microenvironments will induce Notch-DLL and MHC-TcR signaling a more biomimetic way, and provide an engineered system in which complete differentiation of functional T cells can be performed.

The sensitivity of stem cells to environmental perturbations has prompted many studies which aim to characterize the influence of mechanical factors on stem cell morphogenesis and differentiation. Hydrodynamic cultures, often employed for large scale bioprocessing applications, impart complex fluid shear and transport profiles, and influence cell fate as a result of changes in the mixing conditions. However, previous studies of hydrodynamic cultures have been limited in the ability to distinguish confounding factors that may affect differentiation, including modulation of embryoid body size in response to changes in the hydrodynamic environment. The objective of this study was to establish a method for systematically controlling EB formation and size, in order to understand the effects of hydrodynamic environments on ESC differentiation, while in parallel characterizing transport within the EB microenvironment to inform the use of convective flow formats for the directed differentiation of ESCs. Prior to introduction into hydrodynamic cultures, EBs were be formed via centrifugation of murine ESCs (D3 line) into 400µm diameter PDMS micro-wells (AggrewellTM) with approximately 1000 cells per well. After 24 hours of formation, pre-formed EBs were transferred into rotary orbital suspension cultures at a range of mixing frequencies (25-65 rpm). The impact of mixing conditions on pluripotent and differentiated cell phenotypes were analyzed using an ESC PCR array in combination with real time PCR (Oct4, Nkx2.5, AFP, Pax-6) and flow cytometry (AFP-GFP cell line). To assess the transport characteristics of the EB microenvironment, a custom-designed microfluidic bioreactor was employed to maintain EBs either statically or under convection (20µL/min) in media containing a fluorescently labeled dextran (10 – 70 kDa) and visualized by confocal microscopy. Micro-well aggregation produced large, homogeneous yields of EBs of defined size after 24 hours of formation, which were subsequently transferred into bulk suspension cultures. The hydrodynamic conditions imposed in rotary orbital suspension at 45 rpm and 65 rpm prevented agglomeration of individual EBs, resulting in similar yields and EB sizes over 7 days of differentiation. Analysis of the effects of hydrodynamics on ESC phenotype indicated that, despite few subtle changes in temporal and spatial differentiation, EBs largely undergo similar morphological changes and exhibit similar differentiation profiles over 14 days of differentiation. Changes in transport within the three-dimensional structure indicated that, after 2 days of differentiation, the concentration of dextran 50 µm from the exterior of EBs was similar to the exterior (0.88 ± 0.04; normalized to 1 at the exterior), whereas EBs from day 8 of differentiation exhibited statistically decreased concentrations at the same distance into the EB (0.43 ± 0.05), indicating that transport limitations also arise during EB differentiation. However, convective transport at 20 µL/min within day 8 EBs led to increased dextran concentration (0.71 ± 0.02), which was significantly increased compared to static diffusion in day 8 EBs. These results indicate that controlling EB size, using a combination of micro-well formation and hydrodynamics, may be amenable to standardization of ESC cultures, which enables further characterization of the EB microenvironment. The transport limitations that arise within EBs highlight important principles for bioreactor design, whereby modulation convective flow profiles may be an important technique for the development of directed ESC differentiation protocols.

Stem cell transplantation offers enormous potential for the treatment of spinal cord injuries and neurodegenerative diseases such as Parkinson’s disease and Alzheimer’s disease. Although the regulatory mechanisms involved in stem cell differentiation into the three germ layers have been broadly identified, successful clinical translation of this therapeutic approach is contingent on understanding the critical role of extracellular matrix microenvironment on stem cell fate, post-transplantation. In this work, we investigate the synergistic role of extracellular matrix proteins and signaling molecules on embryonic stem cell (ESC) differentiation into various neural (motor neurons, dopaminergic neurons) and glial (astrocytes, oligodendrocytes) lineages.
Mouse embryoid bodies were cultured for 2 weeks in 3D matrix networks made of collagen-1, Matrigel, hyaluronic acid or peptide gels (lab-synthesized), in the presence or absence of retinoic acid (RA; 0.5-10 microM) and sonic hedgehog (SHH; 100-500 nM). The stiffness of these networks was varied by modulating the matrix concentration in these gels. Embryoid bodies cultured directly on protein-coated 2D surfaces under the above conditions served as reference controls. The effect of growth factor dosage, matrix composition, topology and rigidity, on stem cell differentiation and neurite outgrowth were quantified using flow cytometry, immunolabeling and image analysis. The mechanical properties of the 3D networks were measured using atomic force microscopy with a spherical indentor. The key role of integrin receptors in mediating protein-ESC signaling mechanism was quantified using ELISA. 
Our data suggest that neural differentiation was optimal at 1 microM RA addition, compared to 0.5 or 10 microM doses (p < 0.01), under all the culture conditions studied. In general, 3D Matrigel and collagen-1 networks dramatically enhanced neural differentiation and neurite outgrowth in the presence of 1 microM RA, compared to other 3D networks or protein-coated 2D surfaces (p < 0.001 for all the cases). It was observed that neural differentiation in 3D networks was directly proportional to matrix stiffness, while astrocyte differentiation under similar conditions was inversely modulated by matrix stiffness (p < 0.01). ELISA analysis suggests that the above observed effects of 3D Matrigel on embryonic stem cell differentiation and neurite outgrowth were induced specifically by alpha6 and beta1 integrins. While 1 microM RA and 3D collagen-1 (2 mg/mL) synergistically promoted motor neuron differentiation, supplementation of 3D Matrigel (as-received) and 300 nM SHH had a synergistic effect on dopaminergic neuron differentiation from mouse ESCs in vitro. These results could serve as the basis for scalable culture systems that produce specific neuronal cell types on demand for clinical transplantation applications.

 A major obstacle to three-dimensional tissue engineering is the incorporation of a functional vascular network to diffuse oxygen and nutrients to the cells.  In a previous study, we demonstrated that implantation of multiple cell sheets and the intrinsic vascularization on rat subcutaneous muscle could overcome this obstacle.  However, in this approach tissue thickness is still a limitation making clinical application difficult.  Since in vitro cell sheet stacking is not invasive we fabricated a microvascular bed that harnessed the natural in vivo biological revascularization and used it in an in vitro bioreactor system to stack cell sheets. An engineered microvascular bed to promote angiogenesis was made from the femoral arteriovenous of a rat. A silicone chamber was constructed to incorporate the arteriovenous, but isolated the microvascular bed from the surrounding tissue using 2% atelocollagen gel as the internal support for neovascularization.  The gaps between the chamber and vessels were sealed completely with biocompatible adhesive silicone.Two weeks after implantation into the rat, the in vivo chamber produced new capillaries from the arteriovenous. However, this early vascular network was not sufficient for in vivo perfusion, so a rat neonatal cardiac cell sheet (2.4×106 cells), which included endothelial cell networks engrafted to the microvascular bed to accelerate vascular formation in the construct, and the chamber was then resealed for another two weeks in the original location.  Four weeks after the first implantation, we observed a well organized microvasculature and clear cardiac pulsation in the engrafted cell sheet after the chamber was reopened.  A cross sectional view of the HE staining shows that the microvascular network in the transplanted cell sheet contained red cells, and indicated a substantial vascular connection between the microvascular bed and implanted cell sheet in the isolated chamber.  Before removing the chamber from the rat, we ligated the distal arteriovenous and then we injected the central artery with saline, it drained successfully from the central vein.  This result indicated the establishment of a working vascular network between the arteries, veins and scaffold to supply blood to the engrafted cell sheet.  The full extent of the vascular networks was also confirmed by fluorescence microscopy by injecting fluorescent microparticles.  This construct can be easily transferred from the rat to an in vitro bioreactor by resecting only the arteriovenous.
 

Neural regeneration within the central nervous system (CNS) is a critical unmet challenge as brain and CNS disorders continue to be the leading cause of disability nationwide. Common tissue engineering goals seek to customize cell-biomaterial interactions and guide cell behavior. Here we have developed a material that is both cell instructive and cell responsive, creating a dynamic interplay between cells and their engineered extracellular matrix (ECM). Using recombinant protein technology, we engineered a family of elastin-like protein hydrogels with multiple independently tunable properties. With these materials we can investigate individual and synergistic effects of elastic modulus, degradation rate, and cell adhesivity on cell behavior in a tunable microenvironment. Unlike natural matrices, the concentration of adhesive ligands such as the fibronectin-derived sequence RGD can be precisely controlled without altering the mechanical properties of the hydrogel. When dorsal root ganglion neurons were cultured in 3D in these gels, RGD promoted neuron-specific growth, limited astrocyte proliferation, and more than doubled the rate of neurite extension. Crosslinking density was tuned to create scaffolds with elastic moduli from 1-100 kPa which also impacted DRG growth independently of adhesive site concentration.
We have designed the proteins with a second set of bioactive sequences that specifically respond to changes in cell phenotype. By incorporating cell-mediated degradable subunits into the elastin-like proteins, we are able to mimic the natural remodeling of the ECM. Neural stem cells (NSCs) undergoing differentiation may change their production of the protease urokinase plasminogen activator (uPA), which has previously been found at the growth cones of extending neurites. We engineered multiple uPA target sites with different degradation kinetics into the elastin-like protein to allow neural cell-mediated control of the scaffold degradation dynamics. This strategy was also used to enhance the functionality of the polymer by controlling delivery of multiple molecules with distinct release kinetics. One molecule was tethered to the matrix via a fast-degrading uPA-responsive sequence and fully released in 48 hours, while another molecule tethered by a more slowly degrading uPA-responsive sequence was continually released for more than 240 hours.
These crosslinked scaffolds are useful for directing the growth and differentiation of multiple cell types including clinically relevant NSCs. Adult murine NSCs were capable of proliferation and differentiation into neurons and glia when seeded on top of RGD-containing scaffolds. These tunable scaffolds are responsive to neural cells which may be able to specifically self-modulate the release of multiple bioactive factors while undergoing differentiation. This work demonstrates the versatility and responsiveness of our modularly-designed protein hydrogels for neural cell culture and encourages continued development as a biomaterial tissue construct for treating spinal cord injury.

We analyzed the therapeutic effect of the transplantation of the human embryonic stem cell (NIH Code: MB01)-derived neuronal precursor (hES-NP) cell and post-ischemic exercise in rats with the middle cerebral artery (MCA) infarct model. A cortical infarct was induced in 20 adult Sprague Dawley rats by occlusion and reperfusion of the MCA. The rats were divided into four groups: hES-NP cell transplantation and exercise, transplantation only, exercise only, and Sham-operated with no exercise. In the cell-transplanted group, hES-NP cells were transplanted by stereotactic inoculation into the ipsilateral basal ganglia 7 days after infarct.We evaluated the clinical recovery of deficit, the size of infarct and the survival, migration, and differentiation of the transplanted cells. The transplanted hES-NP cells survived robustly in the ischemic brains 3 weeks post transplant. The majority of migrating cells in the ischemic rats had a neuronal phenotype. The clinical scores of all of the experimental groups were better than those of the Sham-operated group. Whereas the exercise-only group showed continuous clinical improvement, the celltransplanted groups manifested less improvement than the exercise-only group. Moreover, the cell-transplanted groups did not differ in clinical improvement according to postinfarct-exercise or not. The infarct size was significantly reduced in both the cell-transplanted groups and the postischemic exercise group, compared with the Sham-operated group; however, the reduction of infarct size was most prominent in the exercise-only group. In our study, the inoculated site of the basal ganglia showed some damage induced by inoculation, such as loss of neuroglial cells, reactive gliosis and microcalcification, which was found in the Sham-operated group as well, and yet no inoculation-site injury has ever been reported. Our study revealed that stem cell transplantation can have a positive effect on behavioral recovery and reduction of infarct size, but the effect shown was no better than the effect of the exercise, which finding reconfirmed the importance of post-infarct rehabilitation. In addition, it was found that cell inoculation should be replaced by a noninvasive procedure.

Our laboratory has developed and optimized a hollow fiber-based bioreactor with the ability to culture adipose tissue reproducibly in a three-dimensional system. Long-term culture of adipocytes in two-dimensional, flat flasks is not plausible. Hollow fiber-based bioreactors enable the growth and expansion of adipocytes, resulting in adipose tissue. In order to further differentiate human adult healthy and diabetic adipose-derived stem cells (ASCs) into mature adipocytes and create 3D adipose tissue in vitro, we have used a 3D, hollow fiber-based bioreactor system.  We hypothesize that a 3D hollow fiber bioreactor will result in the long term (e.g., 3 month culture) of human adipocytes ex vivo, providing metabolically active tissue that serves as an experimental model for restorative medicine applications such as fat grafting and high throughput screening of diabetic drug therapies. Human adipose-derived stem cells (ASCs) derived from the discarded superficial abdominal subcutaneous adipose tissue of a patient with Type II diabetes (female, BMI 27.4~32, 40~60 years old) The ASCs were expanded in 2D culture then inoculated into a hollow fiber-based bioreactor developed by our group. The perfusion-based system allows cells to differentiate in the presence of integrated oxygenation and counter-current flow for nutrient and gas exchange via interwoven artificial hollow fiber membranes. Daily measurements were taken to ensure pH, glucose and lactate levels within the system. Metabolic activity of the adipose tissue generated from expanded ASCs within the 3D hollow fiber-based bioreactors was assessed with TNF-α and recovered by a 4-hour insulin treatment period. While the diabetic ASCs were in a three-week expansion period, there was notably lower glucose uptake and lactate production compared to a bioreactor culture of non-diabetic ASCs.  Once the cells were differentiated into adipocytes, no difference was observed in the metabolic activity of the diabetic tissue compared to our previously studied non-diabetic adipose tissue. In a TNF-α/insulin stimulation test, the adipocytes experienced a decrease in insulin-stimulated glucose uptake when TNF-α was introduced to the system, followed by a "recovery" when insulin was injected into the bioreactor. After 4 hrs, the glucose uptake returned to baseline. To evaluate mature adipocytes in the bioreactor after 9 wks, immunohistochemical as well as gene expression were analyses performed on the tissue and bioreactor fibers. Hollow fiber-based bioreactors provide the ability to rapidly expand ASCs obtained from a diabetic patient to very high cell densities and allow mature adipocyte differentiation of diabetic ASCs, and for the first time to our knowledge, generating adipose tissue ex vivo for up to 9 wks. Reproducible metabolic activity of the adipose tissue in the bioreactor was demonstrated, rendering our model potentially useful for regenerative medicine applications.

Rapid in-situ capturing of endothelial progenitor cells (EPCs) on blood-contacting surfaces under arterial flow should exert nonthrombogenic potential similar to native endothelial tissues. In this study, first, based on ex vivo cellular potentials including adhesion, proliferation and differentiation characteristics on various surface-bound candidate ligands, the best-suited pair of target cell receptor, which is exclusively expressed on endothelial lineage cells, and surface-bound ligand is defined. Second, the arterial flow resistance and prolonged activation of intracellular signaling transduction pathways of EPCs adhered on the defined ligand-bound substrate are determined. Lastly, in-situ capture of EPCs and full endothelialization on stents and artificial grafts, both of which blood-contacting surfaces are bound with defined ligand, is evidenced in porcein model.
Methods include 1) covalent bonding of molecules [vascular endothelial growth factor (VEGF) and two VEGF receptor antibodies and Tie-1 and -2 antibodies] on thin-layered vinyl alcohol-copolymer. 2) Culture of human mononuclear cells on these protein-bound substrates and histocytochemical analyses, 3) determination of hydrodynamic shear stress dependence of adhered EPCs and endothelial cells (ECs) by radial flow chamber technique, 4) implantation of stent (deployed diameter; ca. 3 mm) in coronaries and electrospun artificial graft (inner diameter; ca. 3 mm) in carotid arteries.
Results showed: 1)proteins were covalently bound to the copolymer surface via activation of hydroxyl group. 2) Among molecules examined, only VEGF exhibited high adhesion and proliferation characteristics similar to those of fibronectin, and a quite high differentiation potential (express of surface markers specific for endothelial lineage cell) with culture time. In addition, day-order continuous activation of intracellular signaling transduction pathways (phosphorylation of VEGF receptor, FAK, ERK and Akt) was observed for ECs adhered on VEGF-bound substrate. Once adhered, high detachment resistance to laminar flow was observed under arterial shear stress. 3) Implantation study in porcine models using ultrasonic-atomized spray-coated stents and custom-design electrospun artificial grafts, both of which are surface-architectured with bound VEGF at high density, was conducted. The results showed that cells expressing VEGF receptor adhered on blood-contacting surface, followed by complete endothelialization similar to native endothelial tissue was observed within one or two weeks for stents.
This study concludes that VEGF is best-suited for selective capture of EPCs while avoiding adhesion of a vast numbers of adherent blood cells, which is based on high adhesion, proliferation and induced differentiation potentials, and large retention potential of adhered cells against arterial flow stress. Rapid capture of EPCs and complete endothelialization observed in vivo, and simple but reliable surface architecture devices promise to exert non-thrombogenic potential to devices.
 
 
 

Demineralized bone matrix (DBM) has been used extensively as an osteoinductive material for a variety of clinical applications intended to promote bone growth and regeneration. However, the osteoinductive potency of DBM can vary significantly from lot to lot due to the fact that it is derived from allograft sources and therefore is inherently subject to donor-to-donor variability, in addition to also being limited by donor availability. As an alternative, acellular matrices derived from osteogenic cell sources differentiated under controlled conditions in vitro could serve as a robust, reproducible and scalable platform to derive novel osteoinductive materials. In this study, we examined the in vivo osteoinductivity of devitalized aggregates of differentiated mouse embryonic stem cells (referred to as “embryoid bodies”, EBs) that had been directed towards the osteogenic lineage in vitro. Uniform populations of EBs were created by formation in PDMS microwell templates and maintained for up to 14 days in suspension culture using rotary orbital suspension culture. Treatment of EBs with β-glycerophosphate (βGP) starting at day 5 of differentiation promoted the expression of osteogenic transcription factors, such as Runx2, and ECM molecules, including osteopontin, osteocalcin and bone sialoprotein, coincident with increasing calcium phosphate mineralization content within EBs. Devitalized EB material (EBM) treated with or without βGP (7.5 mg) for up to 10 days of differentiation was combined with inactive DBM (7.5 mg) and implanted into the hindlimbs of nude mice; DBM (15 mg) and heat-inactivated DBM (15 mg) alone served as positive and negative controls respectively. Osteogenic EB-derived materials (EBM) exhibited potent osteoinductive activity, based on ectopic mineralization visualized and quantified by X-ray and CT analyses, and histomorphometric quantification of new bone formation. The in vivo results demonstrate that the embryonic-like microenvironment created by osteogenic EBs can stimulate significant new bone formation in adult animals. Moreover, these results establish the proof-of-principle that acellular biomaterials derived from stem cell sources could serve as a novel approach to develop tissue-specific regenerative molecular therapies.

Despite advances in our understanding of the complex course of events involved in neuronal regeneration, bridging long gaps still remains a clinical challenge. The current clinical gold standard for repairing larger nerve deficits is nerve autografts. Even though using polymeric scaffold offers a compelling alternative to autograft, it has not yet been able to match or exceed the autograft performance.
Several approaches to accelerate regeneration have been developed including the use of neurotrophic factors, cell transplantation technologies as well as supporting regeneration via physical and topological cues. While these approaches have merit, the overall success is limited. It appears that the primary reason for this limitation is that the focus of these approaches are mostly on one cell type of the regeneration process (e.g. axons), while successful nerve regeneration results from multiple cellular and molecular events. Therefore targeting cell types which influence the regenerative pathways upstream that have a direct impact on axons and other critical cells should yield a more significant regenerative outcome.
Macrophages are one the most abundant and phenotypically diverse cells that are present at the site of nerve injury and involved in different aspect of regeneration. Here, we report the effect of modulation of macrophage phenotype on Schwann cell (SC) migration and axonal regeneration in vitro and in vivo. The phenotype of macrophages was regulated by employing two well-known cytokines IFN-γ or IL-4 polarizing them toward pro-inflammatory (M1) or pro-healing (M2a and M2c) phenotypes, respectively.
First, we show that pro-healing macrophages accelerate the SC migration up to 3 times in vitro. Then the effect of IFN-γ and IL-4 on modulating the phenotype of macrophages and subsequently their effect on the SC infiltration and axonal growth in a critically sized rat sciatic nerve gap model (15 mm) were studied in vivo. Our study demonstrates that initial polarization of macrophages in the scaffold to M2a and M2c phenotype results in higher SC infiltration as well as up to 50 times more axonal growth in comparison to the control.
Our results demonstrate that more than the degree of macrophage presence, their specific phenotype plays a major role in the PNS regeneration. In fact, we defined a new term, regenerative index, based on the ratio of (M2a+M2c)/M1 macrophages and found a direct correlation between this index and the rate of axonal growth.
To the best of our knowledge, this study is the first to attempt to modulate the axonal growth by biasing macrophages to a pro-healing phenotype in the peripheral nervous system. This approach by minimizing the need for sophisticated delivery vehicle(s), demonstrates that short and early stimulation of macrophages can significantly influence the longer-term regenerative outcomes.  This approach exploits the endogenous regenerative pathways to promote healing in otherwise non-healing, critically sized nerve gaps.
 

Large-scale wounds resulting in volumetric muscle loss, such as those seen in trauma injuries and surgical rhabdomyosarcoma resections result in impaired muscle function and strength. In such severe cases, where the native tissue architecture is destroyed or lost, the natural regenerative capacity of skeletal muscle is diminished and wounds heal by non-functional scar formation. We have previously shown that a scaffold system composed of fibrin microthreads can provide an efficient delivery system for cell-based therapies aimed at treating volumetric muscle loss in the tibialis anterior (TA) muscle of the mouse. Fibrin microthreads seeded with adult human cells derived from skeletal muscle and cultured using a simple yet novel culture system were implanted into a SCID mouse model of volumetric muscle loss. Cell microthread implants resulted in reduced the overall deposition of collagen containing scar tissue present in the wound bed and promoted in-growth of new muscle tissue from the wound margins. The cell culture system which consists of reduced ambient oxygen concentration, FGF2 supplementation, and modified substrate extends the in vitro life-span of primary human cells as well as preserves expression of stem cell markers compared to convention myoblast culture systems. In long term culture of primary adult human skeletal muscle derived cells, this novel culture system suppresses spontaneous differentiation of myoblasts and preserves expression of nestin, which is an early marker of regenerating skeletal muscle. Following implantation, cells cultured using these conditions contributed to the nascent host tissue architecture by forming skeletal muscle fibers, connective tissue, and PAX7-positive satellite cells. Furthermore, improved recovery of total muscle strength and reduced overall wound size were observed in cell/microthread implanted animals at up to 4 months post implantation. Taken together, these results suggest that fibrin microthreads represent a suitable platform for skeletal muscle cell therapy and, when coupled to an improved cell culture system, has the potential to improve the healing outcome for cases of volumetric muscle loss.

Skeletal muscle (SM) is the most abundant tissue of the body, and has the capacity to regenerate after an injury or diseases such as muscular dystrophy.SM is anatomically and functionally organized in myofibers, syncytial tubular cells. Between the basal laminae that surround each fiber and their plasma membranes, within a specific Extra Cellular Matrix (ECM), are the muscle stem cells, or satellite cells (SCs), in a quiescent state. SCs can undergo self-renewal and differentiation in the regeneration process after the activation from quiescence. Effectively mimicking in vivo niches for tissue engineering is a critical research and clinical tool to create new regenerative therapies. Although most in vitro work is done in two dimensions, it does not yet reflect an adequate in vivo environment of the SM niche. We approached the problem applying a reverse engineering strategy, biomimicking the skeletal myofiber as the critical environment for regeneration associated with the SC. In our design, we took into account bio-topographical and physical cues of the myofiber: the geometry, the ECM organization, the mechanical properties, to attempt a reconstruction of an artificial muscle fiber niche (AMFN). The goal was to implant SCs on the AMFN to achieve a culture of quiescent SCs for transplantation purposes. We designed the engineering of the AMFN employing an extrusion process of polymerized Collagen, one of the major components of the ECM of SM. The Collagen scaffold, defined for size and mechanical properties, was then sequentially coated with different layers of SC niche specific proteins. We first characterized the AMFN using SEM imaging and AFM scanning. Our scaffold displayed mechanical and structural similarities to a myofiber. SCs implanted on the AMFN were able to proliferate, align in parallel structures and eventually differentiate into millimeter long parallel functional myotubes, fully repopulating the scaffold. Moreover the self renewal potential and the number of quiescent SCs were improved compared to plastic culture conditions.To track in vivo the SC engraftment, we employed recombinant SCs expressing bioreportes. SCs implanted in AMFN resulted in improved engraftment. Notably, for human clinical application, it is important to preserve the “stemness” of isolated SCs from biopsies. It is indeed critical to manage very small numbers of SCs for in vitro manipulation and in vivo transplantation. However, the culture conditions in use show a poor efficiency as the SCs expand into myoblasts, losing their regenerative potential. To evaluate the efficiency of the AMFN to successfully engraft a small number of SCs (10-100), we followed the regeneration process in vivo by means of non invasive bioluminensce assays. Our data show that a small number of SCs on AMFN is sufficient to support a successful engraftment. Our results therefore present a 3D physical niche based on ECM as a controlled environment to study muscle stem cell biology and for tissue repair engineering.

The concept of using synthetic scaffolds that have not been pre-seeded with cells for tissue engineering in vivo is relatively new. The expectation for such scaffolds upon implantation is that they will heal in a fibrotic, avascular manner.  We have demonstrated that if the pore size is precisely controlled (all pores are approximately 40 microns in diameter and interconnected) , a non-fibrotic, vascularized healing outcome is noted. Such reconstructive healing has been observed in skin, bone, sclera, heart and vaginal wall and with poly(2-hydroxyethyl methacrylate), silicone rubber, and polyurethane scaffolds. We have previously reported that such healing is associated with a larger fraction of macrophages in the M2 phenotype compared to non-porous control implants. Here we show that the macrophages within the pore structure are predominantly in the M1 phenotype and those external (but in close proximity) to the implant are in the M2 phenotype. To ensure reliability of the measurement, two cell-surface markers for M1 and two cell surface markers for M2 were measured. This result contrasts with healing results using decellularized submucosal tissues where the good healing observed is associated with an implant heavily infused with M2 phenotype macrophages. Aspects of the healing reaction and associated pathways that might be involved will be elaborated upon in the talk.

Vascular tissue engineering research has led to the successful clinical use of tissue engineered blood vessels (TEBV) for pediatric surgery, and fistulas for adult dialysis patients.  Engineered vascular tissue may also be useful as in vitro models for drug discovery and basic research.  However, TEBVs that have been used clinically are comprised of either non-vascular cells (e.g., dermal fibroblasts) or exogenous scaffold materials, and therefore may not fully recapitulate native vascular tissue structure and function.  The goal of our work is to develop and characterize an entirely cell-based (“scaffold-less”) approach to create engineered tissue from vascular cells by directed cellular self-assembly.  We previously reported that rat aortic smooth muscle cells (rSMCs) seeded into custom ring-shaped agarose wells self-assembled into cohesive cellular rings within 24 hours (Gwyther et al., Cells Tissues Organs, 2011; Gwyther et al., JoVE, e3366, 2011).  After 8 days in static culture, self-assembled cell rings (2, 4 or 6mm i.d.) exhibited ultimate tensile strengths of 100-500kPa.  In addition to their utility for assessing tissue biomechanics, we also demonstrated that cell rings made of self-assembled rSMCs (500,000 cells per 2mm ring sample, cultured statically for 7 days) could fuse to form tissue tubes.
The objectives of the present study were to evaluate the effects of cell ring culture duration on the kinetics of tissue tube fusion, and to determine whether the cellular self-assembly approach could be applied to create tissue tubes from primary human SMCs (hSMCs).  Briefly, rSMCs (500,000 cells/ring) or hSMCs (750,000 cells/ring; coronary artery hSMCs, Lonza) were seeded into ring-shaped agarose wells (2mm i.d.).   Rat SMC rings were harvested after 1, 3, 5 or 7 days, loaded onto 1.9mm i.d. silicone tubing mandrels, pressed end-to-end to place rings in close contact, and cultured for 7 days (n=3 rings per sample; n=7-10 samples per time point).  Contact angles between adjacent rings were measured each day from phase contrast images of tissue tube samples.  To compare the fusion potential of hSMCs to rSMCs, cell rings harvested after 1 day were cultured as tubes on silicone mandrels for 7 days (n=3 rings per tube; n=4 tubes per group).  Within 7 days of “fusion culture”, all groups successfully formed tissue tubes which could be removed from the mandrels.  Tissue rings cultured for 1, 3 or 5 days fused at a significantly faster rate than rings cultured for 7 days prior to fusion culture.   Ring fusion was also successfully observed in tissue tubes generated from one-day-old hSMC rings, although fusion occurred at a slower rate than rSMC rings.  Histological analysis revealed evidence of cell ring fusion in all tissue tube samples, although ring margins were still visible in samples generated from cell rings pre-cultured for 3, 5, or 7 days.  In contrast, individual cell rings were virtually indistinguishable in tissue tubes created from one-day-old rSMC or hSMC rings.  These results demonstrate the feasibility of applying a modular approach to generating “scaffold-less” tissue tubes by fusing self-assembled cell rings made from human or rat SMCs. 

Cardiovascular progenitor cells (CPCs) have been identified within the developing mouse heart by expression of the transcription factors Nkx2.5 and Islet 1 (Isl1). Detailed study of endogenous CPCs has been limited due to the lack of specific cell surface markers needed to isolate them and the absence of suitable conditions to expand them in vitro. We sought to identify specific cell surface markers to identify endogenous embryonic CPCs and recapitulate and functionally validate these CPC markers in induced pluripotent stem cell (iPSCs)-derived Isl1+ progenitors. As well, we developed conditions that would allow propagation and characterization of endogenous and iPSC-derived Isl1+ CPCs and protocols for their clonal expansion in vitro and transplantation in vivo. Microarray analysis of the transcriptome of Isl1+/Flk1+ mouse CPCs identified a panel of surface markers expressed on Isl1+ CPCs. The combination of Flt1+/Flt4+ best identified and facilitated purification of Isl1+ CPCs from embryonic hearts as well as differentiating iPSCs. Selective inhibition of p300-dependent β-catenin signals with small molecule IQ-1 allowed in vitro CPC clonal expansion while maintaining their phenotype and multipotency. Endogenous mouse and iPSC-derived Flt1+/Flt4+ CPCs differentiated into all three cardiovascular lineages. Flt1+/Flt4+ CPCs transplanted into left ventricles demonstrated robust engraftment and differentiation into mature adult cardiomyocytes in vivo. Here, we have demonstrated that the combination of cell surface markers Flt1 and Flt4 can specifically identify endogenous and iPSC-derived Isl1+ CPCs with trilineage cardiovascular potential in vitro and robust engraftment and differentiation into cardiomyocytes in vivo post transplantation.

Regeneration of lost or damaged bone due to trauma or congenital diseases is a challenge frequently faced by reconstructive surgeons. Mandibular bone regeneration is especially important in order to restore function and pre-injury facial appearance. The particular challenges of mandibular reconstruction are inherent in the complex anatomy and function of the bone. Common reconstructive techniques involving cadaveric bone grafts, inert artificial implants, and free tissue transfer have limited effectiveness and carry excessive risks of failure, rejection, and donor site complications. Polymer nanofiber scaffolds have been used successfully for the sustained local delivery of both pro-osteogenic drugs and multipotent mesenchymal stem cells to the defect sites. Mesenchymal stem cells are known to be immune-privileged and have the potential to improve osteogenesis through either differentiation into new cell types or regulation of other cell types via release of cytokines. FTY720, a selective agonist of S1P1 and S1P3 receptors, has been shown to promote neovascularization and osseous tissue in-growth into critical sized defects in both rat cranial and long bone models. It has also been shown to modulate inflammation and host immune response, leading to better graft incorporation. In this study, we show that bone regeneration in a critical size mandibular defect can be improved by the combination treatment of FTY720 loaded PCL/PLAGA nanofibers and primary isolated ASCs. The effect of FTY720 loaded nanofibers on vascularization was measured in a dorsal window chamber model in mice over a period of 14 days. The effect of FTY720 loaded nanofibers on bone regeneration in mandibular defects was evaluated in a rat model. Primary ASCs were isolated from fat bodies of rats ubiquitously expressing GFP and confirmed by flow cytometry. 5mm mandibular defects were made in 35 nine weeks old Sprague Daley rats, which were divided into 5 groups (n=7). PCL/PLAGA nanofiber scaffolds were loaded with FTY720 and/or ASCs. The rats were treated with an empty scaffold, scaffold loaded with FTY720, scaffold loaded with ASCs, scaffold loaded with FTY720 and ASCS, or left untreated. The amount of bone regeneration was measured at weeks 0, 3, 5, 7, 10 and 12 using microCT (n=3-7). The amount of vascularization was measured at weeks 3 and 12 using Microfil enhanced imaging. Mason’s Trichrome and H&E staining were done for all groups at weeks 3 and 12 (n=3). FTY720 loaded nanofibers caused an increase in total volume fraction of blood vessels at the end of 2 weeks, suggesting that it acts as a vascularizing agent. A combination of loaded nanofibers and ASCs treatment on critical size mandibular defect resulted in substantial bi-weekly increase in bone regeneration compared to ASCs or FTY720 nanofibers alone. Though animals treated with only FTY720 scaffolds caused more bone growth than empty scaffolds, this difference was not significant at earlier time points. The presence of GFP+ ASCs allows for measuring the host vs. donor contribution to new bone formation. These results indicate that the local delivery of FTY720 and ASCs will significantly accelerate bone regeneration by promoting neovascularization, osteogenesis, and increased graft incorporation into the defect site. Comparison with FY720 and ASC controls suggests that both factors have an important and perhaps synergistic effect.

Biomaterials development provides a unique opportunity to recapitulate features of cellular microenvironments in culture.  Using a modular design strategy, we synthesized a family of protein-engineered biomaterials where we exert molecular-level control over the component modules through application of recombinant protein technology.  This allows us to incorporate specific and critical cues that recapitulate in vivo muscle tissue conditions in order to create an in vitro model of skeletal muscle tissue, which will be instructive for muscle tissue engineering.  In this study, we used our biomaterials to tune cell-adhesive ligand density using the well-known extended RGD sequence derived from fibronectin, a key component of the basement membranes of muscle fibers in vivo.  We were able to tune ligand density from 930 to 9300 RGD ligands/micron^2.  Independently of ligand density, we also varied the topographical features of the material, creating ridged channels ranging from 20 to 200 microns in width in order to recapitulate the organization of muscle fibers in vivo. We therefore decoupled the effects of biochemical and structural cues in achieving aligned, striated muscle tissue using primary human skeletal muscle myoblasts (hMBs) that have been isolated from tissue biopsies. We utilized our biomaterial system to identify conditions for the alignment and functional maturation of hMBs in order to recreate the stem cell niche for adult muscle stem cells (MuSCs), which typically reside atop mature muscle fibers. For our biomaterials, we found that topographical spacing of 20 microns and a cell-adhesive ligand density of 9,300 RGD ligands/micron^2 most significantly enhanced primary hMB alignment, elongation, and spread cell area.  We were also able to differentiate human myoblasts into aligned, elongated, and multinucleated myotubes ranging hundreds of microns in length.  Interestingly, an intermediate topographical spacing of 50 microns was required in order to optimize hMB differentiation into myotubes with the desired controlled directionality and organization, indicating that optimal myoblast conditions may not necessarily translate into the best conditions for myotube differentiation.  Finally, we were also able to electrically stimulate these primary human myotubes so as to induce myotube contraction, thus demonstrating functional maturation of human muscle fibers in an in vitro environment. We are currently using this engineered in vitro platform to study the niche requirements of adult MuSC, which are known to be highly sensitive to their biochemical and biomechanical microenvironment. Further, as the utilized biomaterial is biodegradable, biocompatible, and implantable, this work provides the foundation for future regenerative medicine applications aimed at replacing aged, injured, or diseased skeletal tissue.

Insulin-dependent diabetes is one of the major health problems worldwide and there is a great deal of interest in developing a potential cure by transplantation of islet cells isolated from a donor pancreas. A critical component of this approach is the availability of sufficient high quality islets to reverse diabetes in the patient. Moreover, the potential for xenogeneic transplantation to relieve the demand on an inadequate supply of human pancreases will also be dependent upon the efficiency of techniques for isolating islets from the source pancreases. Porcine islets are favored for xenotransplantation for a number of compelling reasons but mature pigs (>2yrs) present logistic and economic challenges and young pigs (3-6 months) have not yet proved to be an adequate source.Recent findings for the plasticity of immature porcine islets in culture demand a greater understanding of their biology in vitro and in vivo. In this study, islets were isolated by conventional collagenase (Roche MTF) digestion of 7 juvenile pancreases (3m; 25kg Yorkshire pigs) and characterized using a battery of tests at various time points during culture in silicone membrane flasks. Islet biology assessment included oxygen utilization, insulin secretion (static and perifusion), histopathology and in vivo function using the diabetic nude-mouse bioassay. Islet yields (1823±158 IEQ/g) comprised a high proportion (>90%) of small islets (<100µm), and purity after optiprep density gradient purification was 64±6%. Morphologically, islets appeared as “grape-like” clusters on day-0, and loosely disaggregated at day-1. Further culture showed a transition of the tissue to more condensed aggregated structures comprising both exocrine and endocrine cells by day-6. Histopathology confirmed both insulin and glucagon staining of the cultured aggregates and the islet grafts excised after 30days following transplantation in diabetic immunodeficient mice. Nuclear staining (Ki67) confirmed mitotic activity consistent with the observed plasticity of these structures. Metabolic integrity was demonstrated by oxygen consumption rates = 178±20 nmol/min/ng DNA and physiological function was intact by both static and dynamic glucose stimulation. In vivo function was confirmed by sustained (>30d) normalization of blood glucose in >50% (8/17) transplants. Preparation and culture of juvenile porcine islets as a readily available source for islet transplantation requires specialized conditions. These immature islets undergo apparent plasticity in culture involving aggregation with contaminating exocrine tissue to form fully functional “pseudopancreas” structures. Further development of this new method of culturing immature porcine islets is expected to generate small pancreatic tissue-derived organoids as a therapeutic product from juvenile pigs for xenotransplantation and diabetes research.

The therapeutic potency of delivered mesenchymal stem cells (MSCs) in tissue engineering applications may be improved by priming cells toward a differentiated state prior to implantation. Mimicking native ECM interactions, the electrostatic attraction between negatively-charged glycosaminoglycans (GAGs such as heparin and chondroitin sulfate) and positively-charged proteins may act to locally sequester factors present in the culture media to further enhance stem cell differentiation in response to soluble factors prior to cell transplantation for musculoskeletal disorders. In a first set of experiments, we demonstrated that human MSCs encapsulated in PEG-based hydrogels with higher amounts of heparin and cocultured with osteoblasts exhibited an over 36-fold increase in alkaline phosphatase activity and 13-fold increase in calcium accumulation by day 21, compared to MSCs cocultured with MSCs at the same heparin content. Moreover, hydrogels with higher amounts of heparin and cocultured with osteoblasts exhibited enhanced mineralization on the edges, indicating that heparin may be important in sequestering osteoblast-secreted soluble factors, particularly on the surfaces of hydrogels. In a separate, but related, set of studies, we desulfated the GAG chondrotin sulfate (CS) and modified both the CS and desulfated chondrotin (Ch) to form hydrogels. Disaccharide analysis indicated that the CS molecules have approximately twice the negative charge compared to Ch. In release studies with PEG-based hydrogels containing either 50 wt% CS or 50 wt% Ch, 50% Ch hydrogels exhibited significantly greater release, suggesting that TGF-B1 retention is dependent on the presence of sulfate groups. After encapsulation of human MSCs within these gels, MSCs in desulfated Ch hydrogels significantly upregulated gene expression of the cartilaginous ECM molecules collagen II and aggrecan on days 7, 14, and 21, compared to CS, and upregulated the cartilaginous transcription factor SOX9 on day 7 only when cultured with exogenous TGF-B1 in the media. In addition, pericellular aggrecan deposition was more obvious in Ch than CS hydrogels after immunostaining, indicating that, surprisingly, Ch materials may promote chondrogenic differentiation over CS in the presence of TGF-B1. Taken together, these results suggest that GAG-based biomaterials may be an exciting means to locally sequester soluble factors and enhance MSC differentiation, and that, through chemical removal of the sulfate groups, the scaffolds’ interactions with growth factors may be tuned to promote optimal MSC response for specific culture conditions. Acknowledgements: Complex Carbohydrate Research Center at U. Georgia, NSF Graduate Research Fellowship to JJL, NSF CAREER Award to JST

Bone regeneration is enhanced through the interaction of a biodegradable extracellular matrix combined with a competent cell source and growth factors. Gene-activated matrices (GAMs) have shown potential in localised gene delivery resulting in bone tissue regeneration. Recently, we have developed a novel nano-hydroxyapatite (nHA) synthesis technique and have incorporated the nHA particles into collagen-based scaffolds engineered for bone repair to yield a coll-nHA scaffold. In this study, the possibility of using this coll-nHA scaffold as a GAM is investigated. Proof of concept experiments were performed using reporter genes and two non-viral gene delivery vectors to optimize transfection parameters in mesenchymal stem cells (MSCs) in both monolayer culture and MSCs seeded on GAMs. The potential of combining two vectors (which cannot be identified due to I.P. issues) containing an angiogenic and an osteogenic gene with coll-nHA scaffolds to produce a scaffold which mimics the natural bone healing response is explored.
 
The transfection parameters of two different non-viral gene delivery vectors were optimised using green fluorescent protein (GFP) and luciferase in monolayer MSCs. Fluorescence microscopy was used to analyse GFP-transfected rMSCs, while a LumiFlex GLuc Assay kit was used to quantify luciferase expression using each delivery vector. % transfection efficiency was assessed using flow cytometry. Monolayer osteogenesis and Matrigel™ assays were performed with each vector and gene to determine their suitability for therapeutic gene delivery. Coll-nHA scaffolds with 100% nHA relative to collagen were prepared and soak-loaded with various vector-pDNA complexes as follows: Vector 1 both genes; Vector 2 both genes; Vector 1 Gene 1-Vector 2 Gene2 (Mix GAM). These GAMs were then seeded with MSCs and cultured in media containing osteogenic supplements to determine the most effective GAM. Following 14 days in culture, the GAMs were analysed by calcium quantification, µCT and alizarin red staining.
 
The monolayer study enabled optimisation of the transfection parameters using both delivery vectors. Flow cytometry was performed on vector 1/vector 2-GFP rMSCs to determine their respective transfection efficiencies. Vector 1 was observed to be more suitable for delivery of the osteogenic gene while Vector 2 was apt for gene 2. The mix GAM exhibited significantly superior therapeutic osteogenic potential when analysed using µCT, calcium quantification and histology compared to the individual delivery vectors with both genes and non-transfected cell controls.
 
   
 
This research has demonstrated the potential of using novel collagen-nHA scaffolds as gene activated matrices for therapeutic gene therapy. As GAMs allow for a more controllable, sustained and localised delivery of DNA and improved distribution of the overexpressed proteins by the transfected cells, the two optimal gene carriers (with alternate associated gene transfer profiles), were combined within the GAM. In summary, our coll-nHA mix GAMs have demonstrated superior osteogenic capabilities compared to non-transfected control groups and single delivery vector dual gene groups. We have shown that this collagen-nHA scaffold can be used as a platform for the delivery of multiple genes and gene vectors. Incorporating two different vectors allows us to efficiently harness the synergistic potential of amalgamating two therapeutic genes to successfully mimic natural bone healing.
 
 
 

Tissue engineering provides a means to create small diameter arterial grafts that will not fail like synthetic materials. To date, arterial grafts developed using the cell sheet method and a synthetic polymer-based approach have been successfully implanted. However, a completely-biological graft that can be produced in just two months, possessing circumferential alignment/tensile anisotropy, that can be successfully implanted into the arterial system has not yet been reported. We recently demonstrated the ability to create implantable arterial grafts using fibrin gel seeded with human dermal fibroblasts followed by conditioning in a pulsed-flow-stretch (PFS) bioreactor (Syedain et al. 2011). Here, we evaluated the in vivo performance of decellularized engineered allografts in a sheep model. Engineered grafts (4 mm ID, 2-3 cm long, 0.4 mm thickness) were fabricated from ovine dermal fibroblasts using the PFS bioreactor and decellularized using sequential treatment with SDS, Triton-X, and DNase, which had little effect on graft properties. The total graft culture time was 7 weeks. The burst pressure of the decellularized grafts exceeded 4000 mm Hg and had the same compliance as the ovine femoral artery. No cells were visible with histological staining and DNA content was less than 10% of the untreated grafts. Grafts were implanted interpositionally in the femoral or carotid artery of 4 sheep (n=6), in some cases the contralateral position being used for a graft or sham control, where the native artery segment being sutured back into place. Anticoagulation therapy was used for the duration, but no immunosuppression was used. At 8 weeks, all grafts were patent and showed no evidence of dilatation or mineralization. Mid-graft lumen diameter was unchanged. Lumen diameter at the ends was 15% smaller for the grafts at 8 weeks based on echocardiography, with a similar trend for the controls. A neointima was evident near the ends of the grafts. Extensive recellularization occurred, with most cells expressing SMA. No elastin deposition was evident. Endothelialization was complete at the ends of the grafts and partial mid-graft. Complementary in vitro studies with a parallel plate flow chamber also indicate excellent shear resistance at physiological shear stresses of pre-seeded blood outgrowth endothelial cells and mesenchymal stem cells. These studies indicate that our completely biological grafts, which are cultured only 7 weeks in vitro, possessing circumferential alignment/tensile anisotropy, can be implanted into the arterial circulation without dilation or mineralization and minor intimal hyperplasia, and with favorable remodeling.

Extracellular matrix (ECM) scaffolds derived from decellularized tissues are increasingly being used for the repair and reconstruction of injured tissues. One important, yet unexplored, variable is the age of the animal from which the ECM is prepared. The objectives of this study were to investigate 1) the structural, mechanical and bioactive properties of small intestinal submucosa (SIS) ECM, harvested from pigs whose only differing characteristic was age, and 2) investigate the effects of these structural and biological differences on the ability of SIS-ECM to promote constructive tissue remodeling in a rat body wall defect model. SIS-ECM was prepared from pathogen-free pigs at 3, 12, 26 or >52 weeks of age. These pigs were genetically similar and were raised in identical husbandry conditions. The SIS-ECM was characterized by physical properties, mechanical strength, degradation rate, and composition including growth factor (VEGF and bFGF) and glycosaminoglycan content. The bioactivity of ECM degradation products was measured by cell proliferation and migration assays. Four-layer multilaminate constructs of each aged SIS-ECM were implanted in a partial thickness rat abdominal wall defect model to determine their effect on constructive tissue remodeling. Evaluation of constructive remodeling was performed at 14, 28, 120 and 180 days post surgery and involved histologic assessment of host immune response, skeletal muscle, blood vessels and nerve formation and uniaxial tensile testing of the remodeled tissue. Mechanical strength, scaffold thickness and degradation resistance all increased with animal age while glycosaminoglycan content decreased. Growth factor content was highest in SIS-ECM from animals between 12 and 26 weeks of age with significantly lower levels of VEGF and bFGF in SIS-ECM from 3 and >52 week old animals. SIS-ECM from 12-week old animals was the most potent source of chemotactic degradation products while SIS-ECM from >52 week old animals promoted the greatest cell proliferation. In vivo, increasing SIS-ECM age was associated with an increasingly pro-inflammatory host response and less constructive tissue remodeling. Consequently, SIS-ECM from 3 and 12 week animals promoted the greatest amounts of new skeletal muscle tissue while SIS-ECM from >52-week old animals remodeled into dense collagenous tissue. All SIS-ECM samples showed evidence of angiogenesis, but in 3 and 12 week samples nerve fibers were also seen in close proximity to the new skeletal muscle. All the remodeled tissues had similar mechanical properties to native tissue although SIS-ECM from 3-week old animals remodeled into tissue that was more resistant to mechanical failure. It is clear that the age of the source animal makes an important difference to the constructive tissue remodeling outcome. This difference in remodeling is likely the result of changes in physical properties and compositional makeup of the ECM that occur with increasing age. Source animal age should be an important consideration in the development of regeneration strategies that utilize biologic scaffolds.