Biochemical and mechanical properties of the extracellular matrix (ECM) are known to independently influence cell function. Given the complexity of cellular responses, I hypothesized that the integration of multiple matrix factors as opposed to their individual contribution is key to understanding and controlling cell function. The objective of this thesis was to systematically investigate matrix biochemical and mechanical regulation of multipotent stromal cell (MSC) osteogenesis. First, I demonstrated that substrate stiffness-dependent MSC spreading, proliferation and osteogenic response were differentially regulated by matrix protein type (collagen I vs. fibronectin) and concentration. Second, I developed and characterized a matrix microarray platform that enabled the efficient screening of multiple matrix-derived cues (substrate stiffness, ECM type and density). I implemented the matrix microarray platform together with parametric regression models to elucidate novel matrix interactions in directing mouse MSC osteogenic and adipogenic differentiation. Third, I extended the screening study to examine matrix-dependent human MSC osteogenesis. Non-parametric regression models were used to provide a nuanced description of the multi-factorial matrix regulation in MSC osteogenesis. The response surfaces revealed a biphasic relationship between osteogenesis and substrate stiffness, with the exact location and magnitude of the optimum contingent on matrix composition. Guided by the screening results and perturbation to key cytoskeletal regulators, I identified a novel pathway involving Cdc42 in matrix mechanical and biochemical regulation of MSC osteogenesis. Surprisingly, Cdc42 mediated stiffness-dependent MSC osteogenesis independent of ROCK, suggestive of a contractility-independent mechanism in matrix rigidity signal transduction. In summary, the integration of cell-based arrays and statistical modeling has enabled the systematic investigation of complex cell-matrix interactions. This generalizable approach is readily adaptable to other cellular contexts, complementing hypothesis-driven strategies to facilitate non-intuitive mechanistic discovery. Moreover, the improved understanding of matrix-dependent MSC function also has practical relevance to the development of biomaterials for tissue engineering applications.

Biochemical and mechanical properties of the extracellular matrix (ECM) are known to independently influence cell function. Given the complexity of cellular responses, I hypothesized that the integration of multiple matrix factors as opposed to their individual contribution is key to understanding and controlling cell function. The objective of this thesis was to systematically investigate matrix biochemical and mechanical regulation of multipotent stromal cell (MSC) osteogenesis. First, I demonstrated that substrate stiffness-dependent MSC spreading, proliferation and osteogenic response were differentially regulated by matrix protein type (collagen I vs. fibronectin) and concentration. Second, I developed and characterized a matrix microarray platform that enabled the efficient screening of multiple matrix-derived cues (substrate stiffness, ECM type and density). I implemented the matrix microarray platform together with parametric regression models to elucidate novel matrix interactions in directing mouse MSC osteogenic and adipogenic differentiation. Third, I extended the screening study to examine matrix-dependent human MSC osteogenesis. Non-parametric regression models were used to provide a nuanced description of the multi-factorial matrix regulation in MSC osteogenesis. The response surfaces revealed a biphasic relationship between osteogenesis and substrate stiffness, with the exact location and magnitude of the optimum contingent on matrix composition. Guided by the screening results and perturbation to key cytoskeletal regulators, I identified a novel pathway involving Cdc42 in matrix mechanical and biochemical regulation of MSC osteogenesis. Surprisingly, Cdc42 mediated stiffness-dependent MSC osteogenesis independent of ROCK, suggestive of a contractility-independent mechanism in matrix rigidity signal transduction. In summary, the integration of cell-based arrays and statistical modeling has enabled the systematic investigation of complex cell-matrix interactions. This generalizable approach is readily adaptable to other cellular contexts, complementing hypothesis-driven strategies to facilitate non-intuitive mechanistic discovery. Moreover, the improved understanding of matrix-dependent MSC function also has practical relevance to the development of biomaterials for tissue engineering applications.

The purpose of this study was to identify conditioning strategies for mesenchymal stromal cells (MSC) which optimize cellular immunosuppressive potency. By identifying new treatment strategies and previously unidentified small molecules capable of stimulating MSC we hope to pave the way tailoring licensed MSC phenotypes to be used in a specific disease state, rather than a one size fits all package. We sought to determine how MSC act in response to a changing immune response or environmental condition. MSC are exquisitely sensitive to changes in their environmental conditions and we show that cellular transcriptome and secretome changes are conditionally responsive to their inflammatory stimulus. One of the main subjects of analysis here is the observations of how these cellular profiles evolve over time in the presence of an inflammatory environment. Similarly, this study observes how MSC behavior changes after an inflammatory event has been resolved to address, in part, the plasticity of MSC licensing and the ability of MSC to rapidly recall a previous immunosuppressive state upon secondary challenge with an inflammatory stimulus. Data was obtained from in vitro experiments with human bone marrow derived MSC and donor human peripheral blood mononuclear cells (PBMC), while in vivo data was obtained using C57BL6/J mice. Overall this research demonstrated that MSC potency can be bolstered by small molecule and drug treatment conditioning, and that certain disease conditions may be more effectively paired with specific MSC conditioning strategies to improve their therapeutic effectiveness.

Infections of fracture fixation implants and synovial structures are a primary cause of complications, increased treatment costs, and mortality in people and horses. Treatment failure is often due to biofilms that are communities of bacteria that are adhered to a surface or to each other and are surrounded in a self-secreted extracellular matrix. The biofilm matrix protects the indwelling bacteria from being killed by antibiotics and the immune system. Biofilms also stimulate chronic inflammation and tissue destruction, including peri-implant osteolysis and subsequent implant failure and chondromalacia with subsequent osteoarthritis. In horses, the resulting lameness, reduced athletic potential, and poor quality of life may necessitate euthanasia. Equine bone marrow-derived mesenchymal stromal cells (MSC) reduce inflammation and promote healing in musculoskeletal injuries and have recently been discovered to have antimicrobial properties. Equine MSC kill planktonic (free-floating) bacteria and prevent biofilm establishment in laboratory models. MSC from mice and people also promote the transition from acute inflammation to tissue regeneration (resolution of inflammation) by secretion of specialized pro-resolving lipid mediators (SPM). Whether equine MSC can disrupt established biofilms of orthopedic pathogens and modulate the inflammatory response to orthopedic biofilms is unknown. Using a novel biofilm-MSC co-culture model, our objectives were two-fold. We investigated whether MSC alone or with amikacin sulfate, an antibiotic used to treat equine orthopedic infections, could reduce biomass, pellicle size, and live bacteria of biofilms of orthopedic infectious agents S. aureus and E. coli. Next, we investigated whether MSC could modulate immune response to S. aureus biofilms by reducing secretion of pro-inflammatory cytokines by peripheral blood mononuclear cells (PBMC) and by secreting SPM. MSC demonstrated partial ability to reduce biofilms but performed differently on S. aureus versus E. coli biofilms. Co-culture of biofilms with MSC significantly reduced pellicle area of biofilms of both bacteria, reduced biomass of S. aureus biofilms, and killed live S. aureus bacteria. MSC combined with amikacin also significantly reduced S. aureus biomass to a greater extent compared to amikacin alone. The resolution in detecting differences between groups for E. coli was diminished because of high variation between biofilms treated with MSC between different donors and between control biofilms between experiments. Using the same experimental system, culture of S. aureus biofilms with MSC in the transwell inserts and PBMC in the bottom wells significantly reduced biofilm size compared to untreated biofilms. Co-culture of MSC and PBMC with S. aureus biofilms also significantly increased detection of multiple SPM on lipid chromatography-mass spectrometry compared to MSC or PBMC cultures alone. Using a commercial equine multiplex bead ELISA, multiple inflammatory cytokines and chemokines were increased when S. aureus biofilms were cultured with MSC and PBMC; however, these were not different from untreated biofilms. Our results indicate that the utility of MSC in combating orthopedic biofilm infections lies in their ability to disrupt the biofilm matrix and promote inflammation resolution. These findings support continued investigation into and optimization of the anti-biofilm mechanisms of MSC. / Doctor of Philosophy / Biofilms are coating layers made by bacteria to protect them from being killed by antibiotics or the immune system. Biofilms result in untreatable infection, chronic inflammation and tissue destruction in people and horses with bone and joint infections. The resulting complications, including pain, reduced mobility, and poor quality of life, may result in horses being euthanized. Equine bone marrow-derived mesenchymal stromal cells (MSC) kill free floating bacteria in laboratory models and reduce inflammation in orthopedic injuries. Whether MSC can disrupt formed biofilms and reduce inflammation resulting from biofilm infections is unknown. Using a laboratory model, our objectives were to determine: 1) whether MSC alone or with an antibiotic used to treat orthopedic infections in horses can disrupt biofilms and kill indwelling live bacteria of orthopedic infectious agents S. aureus and E. coli, and 2) whether MSC can modify the immune response to S. aureus biofilms. MSC demonstrated some biofilm reducing ability but performed differently on S. aureus versus E. coli biofilms. Specifically, MSC reduced the size of biofilms of both bacteria, reduced the coating layer of S. aureus biofilms alone and to a greater extent when combined with the antibiotic, and killed live S. aureus bacteria. Using the same system, culture of MSC with S. aureus biofilms and peripheral blood mononuclear cells (PBMC), a type of white blood cell, reduced biofilm size compared to controls. The addition of MSC and PBMC to S. aureus biofilms also increased detection of fatty acid-derived signals that promote resolution of inflammation, compared to controls. Multiple inflammatory cytokines and chemokines were increased with culture of MSC and PBMC with S. aureus biofilms but were not different from untreated biofilms. These results indicate that MSC may be useful to combat biofilm infections by breaking down the coating layer of biofilms and by promoting resolution of inflammation. Taken together, our results support continued investigation into the potential of MSC as a treatment for orthopedic biofilm infections. The potential of MSC to simultaneously break down biofilms and mitigate inflammation in orthopedic infections would improve cure rates and overall outcomes for horses and people afflicted with orthopedic biofilm infections.

Keratins are a family of proteins found within human hair, skin and nails, as well as a broad variety of animal tissue. Prior research suggests hydrogel constructs of keratin and keratin derivatives exhibit several mechanical and biological properties that support their use for tissue engineering and regenerative medicine applications. Microparticle formulations of these hydrogels are an intriguing delivery vehicle for drugs and cellular payloads for tissue engineering purposes due to the ability to exploit size, surface area, loading potential and importantly, non-invasive delivery (i.e. injection) of cells and biologics. Here we examine the water-in-oil emulsion synthesis procedure to produce keratin microparticles using an oxidized keratin derivative, keratose (KOS). Analyses of particle size, microstructure, and other characterization techniques were performed. Drug loading characteristics, release kinetics, and feasibility of use in two different microparticles was subsequently investigated, first using a model-drug and later testing an antibiotic payload on bacterial cultures to validate antibacterial applications. A suspension culture technique was developed to load bone marrow-derived mesenchymyal stromal cells (BM-MSCs), testing the capacity to maintain viability and express key protein-based factors in cell growth and development. Finally, we tested the in vitro effects of cell-loaded microparticles on the L6 skeletal muscle cell line to determine potentially beneficial outcomes for skeletal muscle tissue regeneration. Largely spherical particles with a porous internal structure were obtained, displaying hydrogel properties and forming viscoelastic gels with small differences between synthesis components (solvents, crosslinkers), generating tailorable properties. The uniquely fibrous microstructure of KOS particles may lend them to applications in rapid drug release or other payload delivery wherein a high level of biocompatibility is desired. Data showed an ability to inhibit bacterial growth in the emulsion-generated system, and thereby demonstrated the potential for a keratin-based microparticle construct to be used in wound healing applications. Dense cell populations were loaded onto particles. Particles maintained cell viability, even after freeze-thaw cycling, and provided a material substrate that supported cell attachment through the formation of focal adhesions. Finally, in vitro studies show that both KOS and BM-MSCs support varying aspects of skeletal muscle development, with combinatorial treatments of cell-loaded particles conferring the greatest growth responses. / Doctor of Philosophy / Keratins and keratin hydrogels may exhibit several properties that support their use for tissue engineering and regenerative medicine applications. Microparticle formulations of these hydrogels are an intriguing delivery vehicle for payloads for tissue engineering purposes. Here we examine the water-in-oil emulsion synthesis procedure to produce keratin microparticles that were analyzed based on drug loading characteristics. A suspension culture technique was developed to load bone marrow-derived mesenchymyal stromal cells (BM-MSCs). Finally, we tested these products to determine potentially beneficial outcomes for skeletal muscle tissue regeneration. Particles with a porous structure were obtained. The microstructure of these particles may lend them to applications in drug release or other payload delivery. Data showed an ability to load and unload specific drug payloads. Dense cell populations were loaded onto particles. Finally, studies show that both keratin and BM-MSCs support skeletal muscle development, with combinatorial treatments of cell-loaded particles conferring the greatest growth responses.

Les douleurs viscérales chroniques font partie des effets secondaires des patients traités pour des cancers de la zone pelvienne. Ces douleurs peuvent affecter grandement la qualité de vie de ces patients. Dans les cas les plus graves, il n'existe pas de traitement analgésique efficace. Ainsi le développement de nouvelles stratégies thérapeutiques efficaces constitue un enjeu majeur. Au sein de notre laboratoire, le potentiel réparateur et immuno-modulateur des cellules stromales mésenchymateuses (CSM) a déjà été démontré dans un modèle d'ulcération colorectale radio-induite chez le rat. Dans ce contexte, l'objectif de ma thèse était d'évaluer d'abord la pertinence de l'utilisation de ce modèle pour l'étude de l'hypersensibilité viscérale persistante radio-induite, puis, le bénéfice thérapeutique de l'utilisation des CSM comme agent antinociceptif. Nous avons dans un premier temps démontré, dans ce modèle, le maintien au cours du temps d'une hypersensibilité viscérale associée à une sensibilisation centrale persistante après irradiation, validant ainsi le modèle. Nous avons ensuite montré l'implication des mastocytes (MC) et suggéré l'implication du neuromédiateur NO dans les mécanismes de la sensibilisation périphérique sous-tendant une telle hypersensibilité. Nous avons enfin mis en évidence que le traitement par des CSM permettait la réduction de l'hypersensibilité viscérale radio-induite persistante. La capacité des CSM à moduler l'activation des MC et/ou leurs interactions avec les fibres nerveuses pourrait être impliquée dans leur action antinociceptive. En conclusion, ce travail a permis d'élargir le spectre d'action thérapeutique des CSM dans notre modèle d'étude. / Patients who undergo pelvic radiotherapy may develop significant incidence of undesirable chronic gastrointestinal complications resulting from radiation-induced damages around the tumour. Chronic visceral pain is one of the radiation-induced side effects that greatly affects the quality of life of “cancer survivors”. The lack of effective analgesic treatment highlights the importance of novel and effective therapeutic strategies. In our laboratory, mesenchymal stromal cell (MSC) based approach showed beneficial immunomodulatory and regenerative effects in a rat model of irreversible radiation-induced colonic ulcers. The goal of my work was to assess the relevance of this model to study radiation-induced visceral persistent hypersensitivity and its modulation by MSC treatment. We first demonstrated that this model is associated with long-lasting visceral hypersensitivity and central neuronal sensitization. In this context we showed then that mast cells (MC) are involved in the mechanism of peripheral sensitization. Moreover, we suggested the implication of the neuromediator NO• in the pathophysiology of persistent radiation-induced visceral hypersensitivity. We also suggested that MSC treatment reversed radiation-induced hypersensitivity by a mechanism that in part may involve the modulation of MC activation and/or the decrease in the number of MC and nerve fiber interactions. In addition, MSC treatment reduced the percentage of nitrinergic neurons, increased after irradiation, and restored colonic muscular contractibility. Such processes may promote the therapeutic benefit of MSC observed in our study. In conclusion, this work provided new insights on the therapeutic benefit of MSC in our study model and a new argument in favour of their use in a future clinical trial to cure abdominopelvic radiotherapy side effects.

An emerging non-pharmacological treatment for bone degenerative diseases is whole body vibration (WBV), a mechanical signal composed of low-magnitude, high-frequency (LMHF) vibrations that when applied to bone, have osteogenic and anti-resorptive effects. Currently, the cellular and molecular mechanism underlying the effect of WBV on bone is unclear. In this study, we investigated the response of osteocytes, the putative mechanosensor in bone, under LMHF vibration. As bone cells differentiate from mesenchymal stromal cells (MSCs), we also studied the osteogenic differentiation of rat MSCs in the presence of vibration loading. We found that vibrated osteocytes show gene and protein expression changes suggestive of an anti-osteoclastogenic response, and secrete soluble factors that inhibit osteoclast formation and activity. In contrast, rat MSCs showed moderate to no response to LMHF vibration during osteogenic differentiation. Our data suggest that in vivo effects of LMHF vibration are mediated through mechanosensing and biochemical responses by osteocytes.

An emerging non-pharmacological treatment for bone degenerative diseases is whole body vibration (WBV), a mechanical signal composed of low-magnitude, high-frequency (LMHF) vibrations that when applied to bone, have osteogenic and anti-resorptive effects. Currently, the cellular and molecular mechanism underlying the effect of WBV on bone is unclear. In this study, we investigated the response of osteocytes, the putative mechanosensor in bone, under LMHF vibration. As bone cells differentiate from mesenchymal stromal cells (MSCs), we also studied the osteogenic differentiation of rat MSCs in the presence of vibration loading. We found that vibrated osteocytes show gene and protein expression changes suggestive of an anti-osteoclastogenic response, and secrete soluble factors that inhibit osteoclast formation and activity. In contrast, rat MSCs showed moderate to no response to LMHF vibration during osteogenic differentiation. Our data suggest that in vivo effects of LMHF vibration are mediated through mechanosensing and biochemical responses by osteocytes.

Understanding the molecular networks that maintain articular cartilage and regulate chondrogenic differentiation of mesenchymal stromal cells (MSCs) are important prerequisites for the improvement of cartilage repair strategies. The first study within this thesis demonstrates that equine cord blood-derived MSCs induced towards a chondrogenic phenotype showed significantly increased miR-140 expression from day 0 to day 14, which was accompanied by decreased expression of previously identified miR-140 targets; ADAMTS-5 and CXCL12. The second study shows that in vitro chondrogenesis on fibronectin coated-PTFE inserts results in more homogeneous hyaline-like cartilage with an increased number of differentiated cells compared with pellet cultures. Finally, the expression of miR-140, miR-9, miR-155 and miR-146a was investigated in an in vitro model of osteoarthritis and suggests a possible role for miR-146a. These results suggest that microRNAs may be useful for directing or enhancing eCB-MSC chondrogenic differentiation and for developing novel biomarker panels of in vivo joint health. / Danish Agency for Science, Technology and Innovation; Equine Guelph; Grayson-Jockey Club Research Foundation; BioE.

Abstract Breast cancer, the most common cancer in women in Finland; its prognosis varies from very good to poor. During the last two decades, mesenchymal stromal cells, carcinoma-associated fibroblasts and normal fibroblasts of the breast have been investigated in the context of breast carcinomas because of their presence in the tumor microenvironment. It has been shown that the properties of the non-malignant tumor compartment possess prognostic value. The effects that these three stromal cell types have on cancer progression have been studied, but their exact mechanisms remain still largely unknown. This experimental work was conducted in order to investigate whether the three cell types of mesenchymal origin influence breast cancer cell proliferation in vitro and tumor growth in vivo. Functional and structural differences between the stromal cell types were investigated using multiple methods. A total of 19 primary human bone marrow derived mesenchymal stromal cell lines, and six paired primary fibroblast and carcinoma-associated fibroblast lines of the breast were used in the study. Co-cultures of labeled stromal cells and breast cancer cell lines MDA-MB-231, M-4A4 and NM-2C5 were performed and the proliferation properties of each cell line were assessed. An orthotopic murine breast cancer model was established by injecting NM-2C5 cancer cells in the mammary fat pads of athymic mice either alone or with the three stromal cell types, and tumor growth and histology were analysed. Mesenchymal stromal cells increased the proliferation of breast cancer cell lines NM-2C5 and MDA-MB-231, and carcinoma-associated fibroblasts increased the proliferation of NM-2C5 cells in vitro. The effect was due to both soluble factors and direct cell-cell contact. In the in vivo experiments, the mesenchymal stromal cells inhibited and the fibroblasts enhanced the growth of breast cancer tumors. Histological analysis of the tumors revealed differences in the invasiveness, necrosis and amount of connective tissue. Differences in the expression of CD105 and CD54 were observed between tumors with mesenchymal stromal cells or fibroblasts. Carcinoma-associated fibroblasts differed from mesenchymal stromal cells in their expressions of CD105 and CD54. The fibroblast subtypes differed at the gene expression level in immunological, developmental and extracellular matrix related pathways. / Tiivistelmä Rintasyöpä on Suomessa naisten yleisin syöpä, ja sen ennuste vaihtelee erittäin hyvästä huonoon. Viime vuosikymmenten aikana mesenkymaalisia stroomasoluja, rinnan kasvainsidekudossoluja ja tavallisia sidekudossoluja on tutkittu rintasyövän yhteydessä johtuen kyseisten solujen läsnäolosta syövän mikroympäristössä. Syöpäkudoksen hyvänlaatuisen solukon ominaisuuksilla on osoitettu olevan ennusteellista arvoa, ja kolmen edellä mainitun strooman solutyypin vaikutuksia rintasyövän etenemiseen on tutkittu, mutta tarkat vaikutusmekanismit ovat vielä laajalti tuntemattomat. Tutkimuksen tarkoituksena oli tutkia edellä mainittujen solutyyppien vaikutusta rintasyöpäsolujen lisääntymiseen soluviljelmässä ja syöpäkasvaimen kasvuun koe-eläinmallissa. Lisäksi strooman solujen rakenteellisia ja toiminnallisia eroavaisuuksia tutkittiin molekyylibiologisilla menetelmillä. Tutkimuksessa käytettiin 19:ää luuytimen mesenkymaalista stroomasolulinjaa sekä kuutta rinnan kasvainsidekudossolu–sidekudossolu paria. Leimattuja strooman soluja viljeltiin yhteisviljelmissä rintasyöpäsolulinjojen MDA-MB-231, M-4A4 ja NM-2C5 kanssa, ja kunkin solutyypin lisääntymistä mitattiin. Ortotooppisessa rintasyövän hiirimallissa immuunipuutteisen hiiren rinnan ihonalaisrasvaan injisoitiin NM-2C5-rintasyöpäsoluja yksinään ja yhdessä strooman solujen kanssa, ja kasvainten kasvua ja histologiaa analysoitiin. Mesenkymaaliset stroomasolut kiihdyttivät NM-2C5- ja MDA-MB-231-rintasyöpälinjojen ja kasvainsidekudossolut NM-2C5-solujen lisääntymistä soluviljelmässä. Vaikutuksen aiheuttivat sekä liukoiset tekijät että suora solujen välinen vuorovaikutus. Eläinmallissa mesenkymaaliset stroomasolut hillitsivät mutta sidekudossolut lisäsivät rintasyöpäkasvaimen kasvua. Histologisissa analyyseissä paljastui eroavaisuuksia tuumorien paikallisessa invaasiossa, kudoskuolion määrässä ja sidekudoksen määrässä. Mesenkymaalisia stroomasoluja ja kasvainsidekudossoluja sisältävien kasvainten välillä esiintyi eroja CD105- ja CD54-pinta-antigeenien määrässä. Kasvainsidekudossolut erosivat pintarakenteiltaan mesenkymaalisista stroomasoluista CD105:n ja CD54:n ilmentämisessä. Sidekudossolut ja kasvainsidekudossolut erosivat toisistaan geenien ilmentämisen tasolla immunologisten, kehityksellisten ja soluväliaineeseen liittyvien geenipolkujen osalta.