Articular cartilage is an extraordinary tissue, allowing frictionless movements of articulated joints, and acting as a load-bearing cushion to protect joints from damage. Breakdown of articular cartilage may result in crippling diseases such as osteoarthritis (OA) and, since articular cartilage has a limited repair capacity, a greater understanding of the mechanisms of joint homeostasis and its response to injury are of great clinical need. In this project the hypothesis that endogenous mesenchymal stem cells (MSCs) may contribute to the healing process of a full-thickness articular cartilage defect was investigated by combining a mouse model of joint surface injury and repair with a nucleoside analogue labelling scheme in DBA/1 mice. Following injury, proliferative responses of nucleoside analogue-retaining cells were detected between 4 and 12 days post injury (dpi) in both the bone marrow and the synovial membrane of the knee joint. Phenotypic analysis of these label-retaining cells using immunofluorescence staining revealed an MSC-compatible phenotype (CD44+, CD105+, CD146+, PDGFRα+ and p75NGFR+), with differences observed between the two tissues in expression of CD105 and CD146. The response of the label-retaining cells to the injury was associated with early activation of Notch signalling (4 dpi), followed by BMP signalling at 8 dpi and TGF-β at 12 dpi. Conversely, canonical Wnt signalling, which was active in uninjured knee joints and in injured knee joints up to 8 dpi, was attenuated at 12 dpi. The contribution of nerve growth factor (NGF), known as a pain mediator in OA, to the repair process was then investigated in vitro. NGF was released by both cartilage explants and femoral head cultures following injury. Using a Transwell-based cell migration assay, NGF was revealed to have a chemotactic effect on human bone marrow derived MSCs, but not synovial membrane derived MSCs. High-density micromass cultures also revealed NGF had a potent stimulatory effect on the chondrogenic differentiation of mesenchymal cells. The data presented here demonstrate a contribution of endogenous MSCs to the repair of articular cartilage in vivo and suggest a possible new therapeutic strategy: stimulation of in vivo recruitment of MSCs by modulating signalling pathways activated during the healing process. Furthermore, a novel role for NGF as a factor involved in migration and the chondrogenic differentiation of MSCs is suggested.

Mesenchymal stem cells (MSCs) have the potential to differentiate into various cell types, possess potent immunomodulatory properties and can influence various functions of immune cells. Since the immunomodulatory properties of MSCs can be modified by cytokines, we compered the effect of unstimulated MSCs and MSCs pretreated with interleukin (IL)-1, interferon (IFN)- , transforming growth factor (TGF)- and IL-10 on the development of regulatory T cells (Treg) and T helper 17 (Th17) cells in vitro and on the inflammatory environment in the eye. MSCs can produce significant levels of TGF- and IL-6. These cytokines represent the key factors that reciprocally regulate the development of naive T cells into Treg and Th17 cells. Unstimulated MSCs produce TGF- , but not IL-6, and the production of TGF- can be further enhanced by IL-10 or TGF- . In the presence of IL-1, MSCs secrete significant levels of IL-6, in addition to spontaneous production of TGF- . MSC producing TGF- induced preferentially expression of Foxp3 and activation of Treg lymphocytes, whereas MSCs supernatants containing TGF- together with IL-6 supported ROR t expression and development of Th17 cells. We demonstrated that MSCs and their products effectively control the development of Tregs and Th17 cells in a population of...

Current treatment options are limited for patients with myocardial infarction or heart failure. Cellular cardiomyoplasty is a promising therapeutic strategy being investigated as a potential treatment, which aims to deliver exogenous cells to the infarcted heart, for the purpose of restoring healthy myocardial mass and mechanical cardiac function. While several cell types have been studied for this application, only bone marrow cells and human mesenchymal stem cells (hMSCs) have been shown to be safe and effective for improving cardiac function in clinical trials. In both human and animal studies, the delivery of hMSCs to infarcted myocardium decreased inflammatory response, promoted cardiomyocyte survival, and improved cardiac functional indices. While the benefits of using hMSCs as a cell therapy for cardiac repair are encouraging, the desired expectation of cardiomyoplasty is to increase cardiomyocyte content that will contribute to active cardiac mechanical function. Delivered cells may increase myocyte content by several different mechanisms such as differentiating to a cardiomyocyte lineage, secreting paracrine factors that increase native stem cell differentiation, or secreting factors that increase native myocyte proliferation. Considerable work suggests that hMSCs can differentiate towards a cardiomyocyte lineage based on measured milestones such as cardiac-specific marker expression, sarcomere formation, ion current propagation, and gap junction formation. However, current methods for cardiac differentiation of hMSCs have significant limitations. Current differentiation techniques are complicated and tedious, signaling pathways and mechanisms are largely unknown, and only a small percentage of hMSCs appear to exhibit cardiogenic traits. In this body of work, we developed a simple strategy to initiate cardiac differentiation of hMSCs in vitro. Incorporating environmental cues typically found in a myocardial infarct (e.g. decreased oxygen tension and increased concentrations of cell-signaling factors), our novel in vitro conditioning regimen combines reduced-O2 culture and hepatocyte growth factor (HGF) treatment. Reduced-O2 culturing of hMSCs has shown to enhance differentiation, tissue formation, and the release of cardioprotective signaling factors. HGF is a pleiotropic cytokine involved in several biological processes including developmental cardiomyogenesis, through its interaction with the tyrosine kinase receptor c-Met. We hypothesize that applying a combined conditioning treatment of reduced-O2 and HGF to hMSCs in vitro will enhance cardiac-specific gene and protein expression. Additionally, the transplantation of conditioned hMSCs into an in vivo infarct model will result in differentiation of delivered hMSCs and improved cardiac mechanical function. In testing our hypothesis, we show that reduced-O2 culturing can enhance hMSC growth kinetics and total c-Met expression. Combining reduced-O2 culturing with HGF treatment, hMSCs can be conditioned to express cardiac-specific genes and proteins in vitro. Using small-molecule inhibitors to target specific effector proteins in a proposed HGF/c-Met signaling pathway, treated reduced-O2/HGF hMSCs show a decrease in cardiac gene expression. When implanted into rat infarcts in vivo, reduced-O2/HGF conditioned hMSCs increase regional cardiac mechanics within the infarct region at 1 week and 1 month. Further analysis from the in vivo study showed a significant increase in the retention of reduced-O2/HGF conditioned hMSCs. Immunohistochemistry showed that some of the reduced-O2/HGF conditioned hMSCs express cardiac-specific proteins in vivo. These results suggest that a combined regimen of reduced-O2 and HGF conditioning increases cardiac-specific marker expression in hMSCs in vitro. In addition, the implantation of reduced-O2/HGF conditioned hMSCs into an infarct significantly improves cardiac function, with contributing factors of improved cell retention and possible increases in myocyte content. Overall, we developed a simple in vitro conditioning regimen to improve cardiac differentiation capabilities in hMSCs, in order to enhance the outcomes of using hMSCs as a cell therapy for the diseased heart.

Cell therapy, including human mesenchymal stem cell (hMSC) therapy, has the potential to treat different pathologies, including myocardial infarctions (heart attacks). Biological sutures composed of fibrin have been shown to effectively deliver hMSCs to infarcted hearts. However, hMSCs rapidly degrade fibrin making cell seeding and delivery time sensitive. To delay the degradation process, we propose using aprotinin, a proteolytic enzyme inhibitor that has been shown to slow fibrinolysis. This project investigated the effects of aprotinin on hMSCs and suture integrity. Viability of hMSCs incubated with aprotinin, examined using a LIVE/DEAD stain, was similar to controls. No differences in proliferation, as determined by Ki-67 presence, and were observed. hMSCs incubated in aprotinin differentiated into adipocytes, osteocytes, and chondrocytes, confirming multipotency. CyQuant assays were used to determine the number of cells adhered to fibrin sutures. The number of adhered cells was increased through aprotinin supplementation at Days 2, 3, and 5 time points. To examine the effect of aprotinin on suture integrity, sutures were loaded to failure to determine ultimate tensile strength (UTS) and modulus (E). Sutures exposed to aprotinin had higher UTS and E when compared to sutures exposed to standard growth media. Degradation of fibrin was quantified using an ELISA to quantify fibrin degradation products (FDP) and by measuring suture diameter. Fibrin sutures incubated in aprotinin had larger diameters and less FDP compared to the controls, confirming decreased fibrinolysis. These data suggest that aprotinin can reduce degradation of biological sutures, providing a novel method for extending the implantation window and increasing the number of cells delivered for hMSC seeded biological sutures.

Mesenchymal stem cells (MSCs) are promising candidates for treating diverse inflammatory disorders due to their capacity to be immunosuppressive. This phenotype is not present at baseline but develops in response to instructive cues. To date, clinical trials use cells grown in basic culture conditions, anticipating the cells will acquire a useful phenotype in response to in vivo cues. This strategy has failed to produce any FDA approved therapies, based on inconsistent efficacy. This thesis explores whether priming MSCs prior to administration can lead to a more uniformly therapeutic phenotype, and it details the design of an optimal in vitro priming regimen. Because interferon gamma (IFN-γ) is known to induce an anti-inflammatory state in MSCs, hypoxia can confer survival benefits, and both cues coexist in known situations of immune tolerance, we hypothesized dual IFN-γ/hypoxia priming would yield a superior immunosuppressive MSC therapy. We show that priming MSCs with hypoxia or IFN-γ alone improves their ability to inhibit T-cells in vitro, but combining these cues results in additive improvements. We next characterize the proteomic and metabolomic changes MSCs undergo when exposed to single or dual IFN-γ/hypoxia priming. While IFN-γ induces MSCs to suppress inflammation and fibrosis, hypoxia leads to cell adaptations to low oxygen, including upregulation of proteins involved in anaerobic metabolism, autophagy, angiogenesis, and cell migration. Dual priming results in additive effects, with many instances of synergy. Finally, we show initial evidence that dual primed MSCs are better able to inhibit disease progression in a mouse model of acute graft-vs-host disease (GvHD).

Human umbilical cord perivascular cells (HUCPVCs) derived from regions surrounding the umbilical cord vessels represent an attractive cell source for cellular therapies, given their proliferative potential and the accessibility of donor material compared with human bone marrow-derived mesenchymal stem cells (hBM-MSCs). However, these cells remain poorly characterised. Using flow cytometry, HUCPVCs were shown to express conventional MSC markers CD29, CD44, CD73, CD90, CD105, CD146, CD166 and integrins alpha1 to -5, alphaV, alphaVβ3, alphaVβ5, β1 and β3, but not CD14, CD34, CD45, STRO-1 or integrin alphaVβ6. HUCPVC marker profiles were consistent between three donors and at different passage numbers. Immunostaining for smooth muscle cell (SMC) markers; alpha-SMA, SM22alpha and smoothelin revealed that HUCPVCs shared expression of these markers with SMCs. However, in comparison with SMCs, HUCPVCs deposited more extensive fibronectin-rich matrices. When compared with hBM-MSCs, HUCPVCs differentiated along adipogenic and osteogenic lineages more slowly and did not progress to terminal phenotypes. mRNA expression of recently identified mesenchymal progenitor cell markers, ROR2, EPHA2, PLXNA2, CDH13 and CD9, was confirmed in HUCPVCs from two donors. In addition, all these markers (except EPHA2) were detected in the umbilical cord vessel wall cells of three donors, confirming their expression in both cultured HUCPVCs and cells of the primary tissue. To determine the roles of these markers in HUCPVCs, they were depleted individually using siRNA. Knockdown (KD) efficiencies of 90-97% were achieved. CD9 KD cells appeared elongated compared to cells treated with control siRNA, and these cells along with ROR2, EPHA2 and PLXNA2 KD cells exhibited larger cell areas than controls. All KD cells also showed decreased proliferative potential by day 6 compared with control siRNA or lipofectamine treated cells. A decrease in total β1 integrins was detected in the CD9 KD cells. Up-regulation of ROR2 and PLXNA2 mRNA expression was detected in HUCPVCs from two donors, when they underwent osteogenic differentiation. ROR2 and PLXNA2 knockdown resulted in increases in PLXNA2 and ROR2 mRNAs respectively, when cells were cultured in osteogenic medium compared with basal conditions. In addition, each individual knockdown revealed that the KD cells showed trends in increasing RUNX2 mRNA expression after 13-16 days in osteogenic medium. These data suggest that ROR2 and PLXNA2 may co-operate in promoting an osteogenic phenotype. Culturing HUCPVCs on non-mineralised BVSMC-derived matrices had very little impact on their differentiation status. In contrast, when HUCPVCs were cultured on mineralised BVSMC-derived matrices in osteogenic medium, their ability to further deposit mineralised matrix was enhanced by 7 days; no accompanying changes in RUNX2, ROR2 or PLXNA2 mRNA expression were detected. Taken together, early up-regulation of RUNX2, ROR2 and PLXNA2 appears to be important in driving osteogenic differentiation in HUPCVCs, whilst subsequent down-regulation of these markers may be required for mineralisation to occur. HUCPVCs express ROR2, PLXNA2, CDH13 and CD9 in vitro and in situ; these markers have distinct roles in regulating cell proliferation, shape and differentiation which may be regulated via changes in β1 integrins. It is not known why HUCPVCs might differentiate along adipogenic and osteogenic lineages more incompletely than hBM-MSCs. Further comparative characterisation of HUCPVCs and hBM-MSCs is a prerequisite for exploiting their vast clinical potential.

Osteoarthritis (OA) is a disease that has significant individual, social, and economic impact worldwide. Although many etiologies lead to the eventual development of OA, one potentially treatable cause is the acute articular cartilage (AC) injury. These injuries are common and have a poor inherent healing capacity, leading to the formation of OA. In an effort to repair AC injuries several treatment strategies have been developed but none have proven completely successful. Studies examining AC tissue-engineering strategies have suggested that those with the most potential for success involve the introduction of autogenous or allogenous cells to the site of injury. These strategies are designed to encourage creation of a matrix with the appropriate characteristics of normal AC. However, development of a completely successful repair method has proven difficult because the biomechanical properties of normal AC are not easy to replicate, a cell source with the appropriate functional characteristics has not been optimized, and the problem of effective incorporation of a repair construct into the host tissue remains unresolved. In an effort to more fully understand the cartilage repair process, this work first focused on the development and utilization of an in vitro human explant model of AC to study the ability of seeded human chondrocytes to integrate into an AC defect. Further work elucidated the gene expression patterns of cultured adult human chondrocytes and human mesenchymal stem cell (MSC)-derived chondrocytes. Results from this work determined that cultured human chondrocytes were able to adhere to articular cartilage defects in a viable in vitro explant model and produce a matrix containing collagen type II. However, further work with the in vitro expanded chondrocytes revealed that these cells have increased expression of collagen type I which promotes the formation of a less durable fibrocartilagenous tissue. This unfavorable expression persisted despite placing the chondrocytes in an environment favoring a chondrocytic phenotype. Further work with MSC-derived chondrocytes demonstrated a similar and unfavorable production of collagen type I. This work represented an important first step towards a treatment for acute AC lesions but it is clear that further work to optimize the culture microenvironment is still required. / Experimental Surgery

Amyotrophic lateral sclerosis (ALS) is a progressive, lethal, degenerative disorder of the CNS. The hallmark of this disease is the premature and selective death of upper and lower motor neurons (MNs) in the brain and spinal cord, leading to fatal paralysis. Although the archetypal vision of neurotoxicity in neurodegenerative diseases is based on the idea that a specific neuronal population is particularly vulnerable to a cumulative toxic event (protein aggregation, mitochondria dysfunction, compromised axonal transport etc…), experimental evidence illustrate that ALS possibly does not arise strictly from damage within MNs. There is now convincing data supporting a non-cell autonomous mechanism in which neurodegeneration is influenced by the toxicity of non-neuronal cells in the vicinity of neurons such as astrocytes and microglia. Considering the accumulation of data implicating astrocytes in the pathogenesis of ALS (loss of GLT-1, secretion of toxic factor, enhanced inflammation, etc…), approaches aiming at replacing astrocytes at site of lesions constitute promising therapeutic strategies. Rapid progresses in the characterization of adult stem cell biology have generated considerable enthusiasm for the development of therapeutic strategies for CNS insults. Several observations support the hypothesis that stem cells may display a valuable influence on diseased host tissues by exerting a protective “chaperone” effect to neurons after differentiation in glial cells. Hence, we decided to study the neuroprotective potential of adult mesenchymal stem cells (MSCs) in ALS. In contrast to neural stem cells (NSCs) which localization in the central nervous system complicates their isolation, MSCs are easily isolated from the bone marrow. The relevance of using on MSCs in stem cell therapies of neurodegenerative disorders is also justified by their capacity to (trans)differentiate into neural cells. For this purpose, we exposed MSCs to growth factors involved in the astroglial differentiation of NSCs. The differentiation of MSCs was characterized by the acquisition of astrocyte morphology in addition to an increased expression of gene related to NSCs (nestin) and astrocytes (glutamine synthetase). The astroglial differentiation of MSCs is associated with the acquisition of a glial-like specific regulation of the production of GDNF, a potent neurotrophic factor for neurons. Then, we characterized the glutamate uptake in differentiated MSCs, a critical function of astrocytes. Our data demonstrate that the differentiation of MSCs is associated with an increased expression of the high affinity glutamate transporter, GLT-1. Thus, our in vitro results confirm the astrocytic differentiation potential of MSCs and we decided to use then in stem cell therapy of ALS. Indeed, we demonstrated that mechanism of stem cell recruitment is present in the spinal cord during the development of the disease by the secretion of stem cell factor (SCF). We injected MSCs derived from healthy animals into the cerebrospinal fluid of a transgenic rat model of familial ALS (expressing a mutated form of the human superoxide dismutase-1, hSOD1G93A) at disease onset. MSCs were found to infiltrate the nervous parenchyma and migrate substantially into the ventral grey matter by interacting with the SCF. At the site of lesion, MSCs differentiated massively into astrocytes around MNs. The intrathecal delivery of MSCs preserved motor functions and extended the survival of hSOD1G93A rats. Investigation of the lumbar spinal cord 35 days after graft demonstrated that the generation of healthy astrocytes from MSCs decreased motor neuron loss. However, this beneficial effect is not related to a decreased excitotoxicity by the rescue of GLT-1 expression but rather a decreased inflammation around MNs. Together, the data presented in this thesis highlight the protective capacity of adult MSC-derived astrocytes in the treatment of ALS.

The conventional treatments for craniofacial bone defects currently are unsatisfactory due to several drawbacks. Replacement of lost bone by autografts typically causes donor site morbidity while allografts, xenografts, and demineralized bone matrix all have a chance of disease transmission. Current synthetic implants placed within the defect site generally lack osseointegration and biodegradability. There are several methods of generating a hybrid extracellular matrix (ECM) and synthetic material construct. These include coating the synthetic material scaffold with collagen and calcium phosphate, incorporating acellular biological tissue within the scaffold material, and using cells to generate an ECM coating on the synthetic material scaffold. The research performed for this thesis developed and characterized mesenchymal stem cell (MSC)-generated extracellular matrix poly(ε-caprolactone) constructs (PCL/ECM) for the replacement of bone tissue. The osteogenic potential of the PCL/ECM constructs was explored by culturing i) MSCs and ii) whole marrow cells combined with MSCs onto the construct with or without the osteogenic differentiation supplement, dexamethasone. It was established that the osteogenic differentiation of MSCs seeded onto ECM-containing constructs was maintained even in the absence of dexamethasone and that the co-culture of MSCs and whole bone marrow cells without dexamethasone and ECM enhances the proliferation of a cell population (or populations) present in the whole bone marrow. The osteogenicity of the constructs encouraged the characterization of the protein and mineral composition of the ECM coating on the PCL/ECM constructs. Characterization revealed that at short culture durations the MSCs used to generate the ECM deposited cellular adhesion proteins that are a prerequisite protein network for further bone formation. At the later culture durations, it was determined that the ECM was composed of collagen 1, hydroxyapatite, matrix remodeling proteins, and regulatory proteins. The prior studies on the PCL/ECM constructs persuaded exploration of the effect of various devitalization and demineralization processes on the retention of the ECM components within and the osteogenicity of the PCL/ECM constructs. Analysis demonstrated that the freeze-thaw technique is a milder method of devitalization of cell-generated ECM constructs as compared to other methods, but it reduced the osteogenicity of the constructs. In addition, it was elucidated that void spaces in the surface of the constructs are important for allowing access of MSCs into the interior of the constructs.

Bone grafts are used to treat more than 300,000 fracture patients yearly, as well as patients with congenital defects, bone tumors, and those undergoing spinal fusion. Given the established limitations of autograft and allograft bone, there is a substantial need for bone graft substitutes. Tissue engineering strategies employing the addition of osteogenic cells and/or osteoinductive factors to porous scaffolds represent a promising alternative to traditional bone grafts. While many bone defects are in load-bearing sites, very little is known about the response of bone grafts and their substitutes to mechanical loading, despite vast documentation on the ability of normal bone to adapt to its mechanical environment. The goal of this research was to quantify the effects of controlled in vivo mechanical stimulation on bone graft repair and bone graft substitutes and identify the local stress/strain environment associated with load-induced changes in bone formation. The global hypothesis that cyclic in vivo mechanical loading improves mineralized matrix formation within bone grafts and bone graft substitutes was addressed in this work using orthotopic and ectopic models specifically designed to facilitate modeling of local stresses and strains. In the first study, a bone defect repair model utilizing an orthotopic implant capable of supplying a controlled mechanical stimulus to a trabecular allograft showed a significant reduction in new bone formation with controlled in vivo mechanical loading. Although the reason remains unclear, loading conditions may not have been ideal for increased bone formation or potential micromotion may have influenced the results. A second study demonstrated for the first time that controlled in vivo mechanical stimulation enhances mineralized matrix production on a mesenchymal stem cell-seeded polymeric construct using a novel subcutaneous implant system. In addition, the local stresses and strains associated with this adaptive response were predicted. The novel subcutaneous implant represents technology which may be adapted for the preparation of tissue-engineered bone constructs, capitalizing on the benefits of mechanical loading and a vascularized in vivo environment. Such an approach may produce larger, stronger, and more homogeneous constructs than could be developed in a static culture system subject to diffusional limitations.