Skeletal muscle injuries and muscle degenerative diseases pose significant challenges to the healthcare. Surgical interventions are restricted due to tissue availability, donor site morbidity and alterations to tissue biomechanics. Current cell-based therapies are hindered by low survival and long-term engraftment for the transplanted cells due to the lack of appropriate supportive microenvironment (cell niche) in the injured muscle. Therefore, there is a critical need for developing strategies that provide cellular and structural support in the regeneration of functional muscle. In the present work, a bioengineered cell niche mimicking the native skeletal muscle microenvironment has been developed for skeletal muscle regenerative engineering. It is hypothesized that the bioengineered scaffolds with appropriate structural and cell instructive properties will support myoblast alignment and function, as well as promote the myogenic responses in clinically relevant skeletal muscle injuries. The current work utilized a three-pronged approach to design biomaterial scaffolds to aid in skeletal muscle regeneration. In the first part, aligned poly(lactide-co-glycolide) (PLGA) fiber scaffolds mimicking the oriented muscle fiber microenvironment with fiber diameters of 335±154 nm (nanoscale), 1352±225 nm (microscale) and 3013±531 nm (microscale) were fabricated and characterized. Myoblasts were found to respond to fiber diameter as observed from the differences in cell alignment, cell elongation, cell spreading area, proliferation and differentiation. <i>In vivo</i> study demonstrated the potential of using microscale fiber scaffolds to improve myogenic potential in the <i>mdx</i> mouse model. In the second part, we designed, synthesized, and characterized an implantable glycosaminoglycan-based composite hydrogel consisting of hyaluronic acid, chondroitin sulfate and polyethylene glycol (HA-CS) with tailored structural and mechanical properties for skeletal muscle regeneration applications. We demonstrated that HA-CS hydrogels provided a suitable microenvironment for <i>in vitro</i> myoblast proliferation and differentiation. Furthermore, <i>in vivo</i> studies using a volumetric muscle loss model in the mouse quadriceps showed that HA-CS hydrogels integrated with the surrounding host tissue and facilitated <i>de novo</i> myofiber generation, angiogenesis, nerve innervation and minimized scar tissue formation. In the third part, we investigated the effects of PC12 secreted signaling factors in modulating C2C12 myoblast behavior. We showed that PC12 conditioned media modulated myoblast proliferation and differentiation in both 2D culture and 3D aligned electrospun fiber scaffold system in a dose dependent manner. We also demonstrated the biomimetic HA-CS hydrogel system enabled 3D encapsulation of PC12 cells secreting signaling factors and promoted survival and proliferation of myoblasts in co-culture. Further proteomics analysis identified a total of 2088 protein/peptides from the secretome of the encapsulated PC12 cells and revealed the biological role and overlapping functions of nerve secreted proteins for skeletal muscle regeneration, potentially through regulating myoblast behavior, nerve function, and angiogenesis. These set of experiments not only provide critical insight on exploiting the interactions between muscle cells and their microenvironment, but they also open new avenues for developing advanced bioengineered scaffolds for regenerative engineering of skeletal muscle tissues.<br>

La réparation endovasculaire (EVAR) est une technique minimalement invasive permettant de traiter l’anévrisme de l’aorte abdominale (AAA) par l’entremise d’un stent- graft (SG). L’utilisation d’EVAR est actuellement limitée par de fréquentes complications liées à une guérison inadéquate autour de l’implant. Ce manque de guérison est principalement dû au type de recouvrement polymérique des SG, au milieu pro-apoptotique des AAA et à l’accès réduit aux nutriments et à l’oxygène après EVAR. L’objectif de cette thèse consistait à concevoir un revêtement bioactif permettant d’inhiber l’apoptose et stimuler la croissance des cellules musculaires lisses vasculaires (CMLV), pour ainsi favoriser la guérison des tissus vasculaires autour des SG. La chondroïtine-4-sulfate (CS) a d’abord été choisie, car elle a été identifiée comme un médiateur important de la réparation vasculaire. Il a été démontré que la CS en solution influence directement la résistance à l’apoptose des CMLV, en plus de favoriser la différenciation myofibroblastique chez les fibroblastes. Dans le cadre de ce projet, un premier revêtement à base de CS et de collagène a été créé. Bien que le revêtement permettait d’induire une résistance à l’apoptose chez les CMLV, il se désintégrait trop rapidement dans des conditions aqueuses. Une nouvelle méthodologie a donc été adaptée afin de greffer la CS directement sur des surfaces aminées, à l’aide d’un système utilisant un carbodiimide. Dans le but d’accroître la croissance des CMLV à la surface des revêtements, le facteur de croissance de l’épiderme (EGF) a ensuite été sélectionné. En plus de ses propriétés mitogéniques et chimiotactiques, l’EGF stimule la production d’éléments de la matrice extracellulaire, comme le collagène et la fibronectine. De plus, l’activation du récepteur de l’EGF inhibe également l’apoptose des CMLV. L’EGF a donc été greffé sur la CS. Le revêtement de CS+EGF a démontré une bonne uniformité et bioactivité sur des surfaces de verre aminé. iii iv Dans une 3ème étape, afin de permettre de transposer ce revêtement bioactif sur des implants, plusieurs méthodes permettant de créer des groupements d’amines primaires sur les biomatériaux polymériques comme le PET ou le ePTFE ont été étudiées. La polymérisation par plasma a été choisie pour créer le revêtement CS+EGF à la surface de PET. Une fois de plus, celui-ci a permis d’inhiber l’apoptose des CMLV, dans des conditions pro-apoptotiques, et de favoriser la croissance des cellules. Le revêtement de CS et d’EGF, déposé sur des surfaces aminées, possède des caractéristiques biologiques intéressantes et semble donc prometteur pour favoriser une meilleure guérison autour des SG. / Endovascular aneurysm repair (EVAR) is a minimally invasive technique performed to treat abdominal aortic aneurysm (AAA) through the use of a stent-graft (SG). The usage of EVAR is presently limited by postoperative complications related to an incomplete healing of the surrounding tissues. The materials currently used in SG, the pro- apoptotic phathophysiology of AAA and the limited access to nutrients and oxygen, all limit the wound healing process and proper tissue ingrowth around the implant. The main objective of this thesis was to create of a bioactive coating inhibiting cell apoptosis and increasing vascular smooth muscle cells (VSMC) growth, to promote healing of the vascular tissues surrounding SG. Chondroitin sulfate (CS) was chosen since recent findings have shown that this polysaccharide triggers key mechanisms involved in vascular repair. CS in solution was shown to inhibit apoptosis of VSMC, as well as stimulate myofibroblast differentiation. A coating of CS and collagen was first created for the purpose of this work. Although the coating was shown to increase cell resistance to apoptosis with VSMC, it was not stable enough, since it rapidly disintegrated in aqueous solutions. A new methodology was thus proposed, where CS was grafted right on aminated surfaces, through carbodiimide chemistry. Epidermal growth factor (EGF) was then chosen to increase VSMC growth on the coatings. EGF is a known mitogenic and chemotactif growth factor for VSMC. It also stimulates the production of extracellular matrix elements, such as collagen and fibronectin. The activation of EGF receptor (EGFR) also triggers various cell signalling pathways modulating VSMC resistance to apoptosis. EGF was thus grafted on CS. CS+EGF coating on aminated glassed slides was shown to be uniform and bioactive. Finally, several methodologies to produce primary amines on polymeric biomaterials, such as PET and ePTFE, were studied in order to eventually transfer the v vi coating on implants. Plasma polymerization was chosen to create the CS+EGF coating. Once again the coating was shown to decrease VSMC apoptosis, in apoptotic conditions, and favour cell growth. Overall, the CS and EGF coating on aminated surfaces possesses interesting biological features and is a promising avenue to stimulate vascular healing around SG.

Racional: A epilepsia do lobo temporal está comumente associada à farmacorresistência e tem a esclerose hipocampal como achado neuropatológico em mais da metade dos casos. Histologicamente, a esclerose hipocampal está associada à perda neuronal diferencial e gliose, além de alterações nos níveis de moléculas associadas à homeostase da água tecidual, como a aquaporina 4 e a molécula de matriz sulfato de condroitina. Em imagens de ressonância nuclear magnética, a esclerose é caracterizada por redução de volume em sequências ponderadas em T1, aumento de sinal e tempo de relaxamento em sequências ponderadas em T2 e redução na transferência de magnetização. Justificativa e Objetivos: Uma vez que tanto o sinal T2 quando a transferência de magnetização são dependentes da água tecidual, nosso objetivo é avaliar, na formação hipocampal de pacientes com epilepsia do lobo temporal, as correlações entre populações celulares e moléculas ligadas à homeostase da água e as imagens ponderadas em T2 e transferência de magnetização. Visamos ainda definir, na formação hipocampal de indivíduos sem alterações neuropatológicas, o volume de cada um dos subcampos hipocampais. Metodologia: Pacientes com epilepsia do lobo temporal farmacorresistente (ELT, n = 43), bem como voluntários sadios (controle radiológico, CH, n = 20), foram submetidos a exames de ressonância magnética em máquina de 3T para mensuração da volumetria hipocampal, tempo de relaxamento T2 e transferência de magnetização hipocampal (exames in vivo). Após o tratamento cirúrgico para o controle das crises, os hipocampos dos pacientes com ELT foram fixados por 8 dias e submetidos aos exames ex vivo em máquina de 3T para cálculo do tempo de relaxamento T2 de cada subcampo hipocampal. Hipocampos controle (Controle historadiológico, CHR, n = 14), foram obtidos de autópsias de pacientes sem histórico ante-mortem de doença neurológica ou presença de patologia no exame do encéfalo pos mortem. Ambos os grupos controle foram pareados para idade em relação ao grupo ELT. Alguns dos casos CHR (n = 6) foram também submetidos à imagem 3D T2 em máquina de 4,7T para cálculo de volumetria dos subcampos hipocampais. Após emblocamento em parafina, secções coronais hipocampais dos casos CHR e ELT foram submetidas às técnicas de histoquímica básica Hematoxilina e Eosina e Luxol Fast Blue, e às imuno-histoquímicas para avaliação das populações neuronais (NeuN), astrócitos reativos (GFAP), micróglias ativadas (HLA-DR) e para a expressão de aquaporina 4 (AQP4) e níveis de sulfato de condroitina (CS-56). Para a comparação entre os grupos, foram realizados testes t para dados paramétricos e Mann-Whitney para dados não-paramétricos. Testes de correlação foram empregados para análise da associação entre as avaliações histológicas e os exames de ressonância magnética. Resultados: Pacientes com ELT apresentaram menor volume hipocampal, maior tempo de relaxamento T2 e menor transferência de magnetização no exame in vivo, quando comparados com o CR. O exame ex vivo para a volumetria dos subcampos hipocampais em casos do grupo CHR indicou que a fascia dentata, a região CA1 e o subículo correspondem à 85 % do volume hipocampal total. Quanto ao tempo de relaxamento T2 ex vivo, foi observado aumento em todos os subcampos hipocampais do grupo ELT, à exceção da fascia dentata, quando comparados ao CHR. A avaliação da densidade neuronal indicou redução significativa em todos os subcampos dos casos ELT, à exceção do subículo, quando comparados ao CHR. Em relação aos valores do grupo CHR, foi observada astrogliose em quase todos subcampos da formação hipocampal (a exceção da zona subgranular e do hilo) e microgliose em todos os subcampos (exceto pelo subículo) dos casos com ELT. Pacientes com ELT apresentaram redução na expressão de aquaporina 4 perivascular em todos os subcampos do hipocampo, comparados ao CHR. Aumento nos níveis de sulfato de condroitina foi observado em todos os subcampos da formação hipocampal, à exceção da camada granular, nos pacientes com ELT. O volume hipocampal e a transferência de magnetização in vivo dos pacientes com ELT correlacionaram-se tanto com a população neuronal como com os níveis de sulfato de condroitina, enquanto que o tempo de relaxamento in vivo correlacionou-se com a população astroglial e os níveis de sulfato de condroitina. O exame ex vivo corroborou a correlação entre a população glial e o tempo de relaxamento observado nos pacientes com ELT. A diferença entre o tempo de relaxamento in vivo e ex vivo correlacionou-se tanto com a difusibilidade da água no tecido como com os níveis de sulfato de condroitina. Conclusões: Nossos dados indicam correlação entre a patologia hipocampal e as imagens de ressonância nuclear magnética, sendo que a maior qualidade das imagens ex vivo permitiu uma avaliação mais direta entre o sinal de ressonância e a patologia, indicando importância da população celular e matriz extracelular para o volume hipocampal e a transferência de magnetização, e da astrogliose para o tempo de relaxamento T2. Finalmente, nossos dados mostraram que CA1, subículo e fascia dentata tem grande participação no volume hipocampal, sendo que alterações nestas regiões tem um papel mais relevante nas alterações observadas na ressonância magnética, como indicado por nossas correlações. / Rationale: Drug resistant temporal lobe epilepsy is often associated with hippocampal sclerosis. Histological evaluation reveals differential neuronal loss, gliosis and changes in molecules associated with water homeostasis, such as aquaporin 4 and chondroitin sulfate. Magnetic resonance imaging in these cases often reveals hippocampal atrophy, increased T2 signal and T2 relaxation and reduced magnetization transfer ratio in the hippocampus. Aims: Once both T2 signal and magnetization transfer are affected by tissue water, our goal was to evaluate, in the hippocampus of drug-resistant temporal lobe epilepsy patients who underwent surgery for seizure control, the associations between cellular populations, aquaporin 4 and chondroitin sulfate with T2 relaxation time and magnetization transfer. Additionally, we intended to measure the individual volume of each hippocampal subfield in hippocampus from patients without neurological disease. Methods: Patients with drug-resistant temporal lobe epilepsy (TLE, n = 43) and age-matched health volunteers (radiological control, RC, n = 20) were submitted to magnetic resonance in a 3T machine for hippocampal volumetry measure, T2 relaxation and magnetization transfer (in vivo examination). After surgical treatment for seizure control, hippocampi from the TLE patients were fixed in formalin for 8 days and then submitted to ex vivo imaging in 3T for relaxation time of every hippocampal subfield. Control hippocampi were obtained from autopsies of age-matched patients without ante mortem history of neurological disease or post mortem neurological pathology, and underwent the same ex vivo imaging (histo-radiological control, HRC, n = 14). Six cases from the HRC underwent 3D T2 imaging in a 4.7T machine, in order to measure the volumes of the hippocampal subfields. Paraffin embedded hippocampal sections from TLE and HRC were submitted to Hematoxilin-Eosin and Luxol Fast Blue histochemistries, and to immunohistochemistries for the evaluation of neurons (NeuN), reactive astrocytes (GFAP), activated microglia (HLA-DR), for aquaporin 4 (AQP4) and for chondroitin sulfate (CS-56). Students t-test or Mann-Whitneys test were performed for comparison between groups, and correlation tests were performed for the comparison between histological and magnetic resonance measures. Results: Patients with TLE presented reduced hippocampal volume, increased T2 relaxation time and reduced magnetization transfer, when compared to RC. The ex vivo volumetry of the hippocampal subfields revealed that fascia dentata, CA1 and subiculum together correspond to 85 % of the total hippocampal volume. Ex vivo relaxation time, as the in vivo, were increased in the subfields of TLE patients, when compared to HRC. Compared to HRC, TLE patients presented neuron loss and microgliosis in all hippocampal subfields but the subiculum, and astrogliosis in all hippocampal subfields but the subgranule zone and the hilus. Reduced perivascular aquaporin 4 was observed in all hippocampal subfields of TLE patients, and increased chondroitin sulfate was observed in all hippocampal subfields, with the exception of granule cell layer, of TLE patients, when compared to HRC. In TLE, both in vivo hippocampal volume and magnetization transfer correlated with the levels of chondroitin sulfate and the neuronal population, whereas the in vivo relaxation time correlated with the astroglial population and the levels of chondroitin sulfate. Ex vivo relaxation time also correlated with the astroglial population in TLE patients. The difference between in vivo and ex vivo relaxation values correlated with water difusibility and the levels of chondroitin sulfate. Conclusion: Our data indicate the importance of neuron population and extracellular matrix to both hippocampal volume and magnetization transfer, and of the reactive astrocytes for T2 relaxation. Ex vivo relaxation time allowed a more detailed evaluation, and indicated more robust correlations between reactive astrocytes and T2 relaxation. Finally, Our data indicated that CA1, the subiculum and fascia dentata are the major contributors to hippocampal volume, so changes in these subfields most likely will affect magnetic resonance imaging.

OBJETIVOS: Avaliar o efeito de sulfato de glicosamina e sulfato de condroitina sobre a nocicepção e o dano da cartilagem articular em um modelo de osteoartrite experimental em ratos. MÉTODOS: Osteoartrite (OA) foi induzida em ratos Wistar machos por transecção do ligamento cruzado anterior (TLCA) do joelho direito. Um grupo falso-operado (SHAM) foi utilizado como controle. Animais OA foram tratados v.o. desde 7 dias antes, até o sacrifício, 70 dias após a TLCA, com sulfato de glicosamina 500 mg/kg (Glu), a combinação sulfato de glicosamina 500 mg/kg e sulfato de condroitina 400 mg/kg (GluChon), ou salina (OANT). Um grupo controle positivo recebeu meloxicam 6 mg/kg s.c. A dor articular foi avaliada pelo teste de incapacitação articular para ratos. Os animais foram sacrificados em diferentes períodos (7,14,28 e 70 dias) após a TLCA. A gravidade das lesões histopatológicas foi graduada nos fêmures após coloração por H&E e azul de toluidina através do escore da OARSI. A quantidade de glicosaminoglicanos (GAGs) extraídos da cartilagem articular dos côndilos femorais foi medida após eletroforese em gel de agarose. O tamanho molecular dos GAGs foi medido após eletroforese em gel de poliacrilamida. A liberação de NO no líquido sinovial foi medida 7 dias após TLCA. RESULTADOS: GluChon reduziu significativamente a dor articular nesse modelo (p<0,01). Glu também reduziu a dor articular, mas não se alcançou significância estatística. Os animais OA apresentaram um aumento significativo dos GAGs na cartilagem articular aos 70 dias após TLCA (77,68±3,38 µg/mg) quando comparados aos animais SHAM (53,46±4,58 µg/mg). O tamanho molecular dos GAGs foi significativamente maior nos animais OA que nos animais SHAM 70 dias após a cirurgia (p<0,01). GluChon preveniu tanto o aumento da quantidade de GAGs (54,42±5,39 µg/mg), quanto de seu tamanho molecular (p<0,05), e estes resultados acompanharam-se de melhora significativa da lesão histopatológica (p<0,05). Os resultados obtidos com Glu foram semelhantes, porém menos evidentes, e não alcançaram significância estatística. Glu aumentou significativamente a quantidade de NO na cavidade articular 7 dias após a TLCA. CONCLUSÕES: GluChon foi antinociceptivo e reduziu significativamente a incapacitação articular. A lesão da cartilagem articular no modelo de TLCA em ratos se acompanha de aumento da concentração e do tamanho molecular dos GAGs, e o tratamento com GluChon previne essas alterações e reduz o dano histopatológico. GluChon foi mais eficaz do que Glu tanto na redução da dor quanto da lesão da cartilagem articular, sugerindo benefício da combinação sobre o uso isolado de sulfato de glicosamina. / OBJECTIVES: Evaluate the antinociceptive and chondroprotective effects of glucosamine sulfate and chondroitin sulfate. METHODS: Male Wistar rats underwent anterior cruciate ligament transection (ACLT) or sham operation of the right knee. Animals were treated p.o. with glucosamine sulfate (Glu) 500 mg/kg, the combination of glucosamine sulfate 500 mg/kg and chondroitin sulfate 400 mg/kg (GluChon), or saline, since 7 days prior to surgery until the sacrifice 70 days after ACLT. A positive control group received meloxicam 6 mg/kg s.c. for antinociceptive evaluation comparisons. Joint pain was evaluated with rat incapacitation test. Animals were sacrificed 7,14,28 or 70 days after ACLT. The severity of histopathologic lesions was evaluated on femoral condyles after hematoxylin-eosin or toluidine blue staining with OARSI grading and staging system. Cartilage extracted glycosaminoglycans (GAGs) concentration was assessed after agarose gel electrophoresis. GAGs\' molecular weight was evaluated after polyacrylamide gel electrophoresis. NO release in synovial fluids was assessed 7 days after ACLT. RESULTS: GluChon reduced joint pain (p<0.01). Glu also reduced joint pain, but results did not reach statistical significance. A significant increase in articular cartilage\'s GAGs concentration was observed in OA animals (77.68±3.38 µg/mg) as compared to sham (53.46±4.58 µg/mg). OA animals also had significantly higher molecular weight GAGs than sham (p<0.01). GluChon prevented both GAG concentration and molecular weight elevations in OA animals, and that was associated with significant less cartilage damage assessed by histopathologic examination (p<0.05). Glu effects were less evident, and did not reach statistical significance. Glu was associated with significant higher concentrations of NO in synovial fluid. CONCLUSIONS: GluChon was antinociceptive on the ACLT model in rats. Articular cartilage damage was associated with increased amounts of GAG with higher molecular weight, and the prevention of these alterations with GluChon treatment was associated with less histopathologic damage. GluChon was more efficient than Glu for both pain and articular cartilage damage reduction, suggesting that combined treatment is better than glucosamine sulfate alone.

OBJETIVOS: Avaliar o efeito de sulfato de glicosamina e sulfato de condroitina sobre a nocicepção e o dano da cartilagem articular em um modelo de osteoartrite experimental em ratos. MÉTODOS: Osteoartrite (OA) foi induzida em ratos Wistar machos por transecção do ligamento cruzado anterior (TLCA) do joelho direito. Um grupo falso-operado (SHAM) foi utilizado como controle. Animais OA foram tratados v.o. desde 7 dias antes, até o sacrifício, 70 dias após a TLCA, com sulfato de glicosamina 500 mg/kg (Glu), a combinação sulfato de glicosamina 500 mg/kg e sulfato de condroitina 400 mg/kg (GluChon), ou salina (OANT). Um grupo controle positivo recebeu meloxicam 6 mg/kg s.c. A dor articular foi avaliada pelo teste de incapacitação articular para ratos. Os animais foram sacrificados em diferentes períodos (7,14,28 e 70 dias) após a TLCA. A gravidade das lesões histopatológicas foi graduada nos fêmures após coloração por H&E e azul de toluidina através do escore da OARSI. A quantidade de glicosaminoglicanos (GAGs) extraídos da cartilagem articular dos côndilos femorais foi medida após eletroforese em gel de agarose. O tamanho molecular dos GAGs foi medido após eletroforese em gel de poliacrilamida. A liberação de NO no líquido sinovial foi medida 7 dias após TLCA. RESULTADOS: GluChon reduziu significativamente a dor articular nesse modelo (p<0,01). Glu também reduziu a dor articular, mas não se alcançou significância estatística. Os animais OA apresentaram um aumento significativo dos GAGs na cartilagem articular aos 70 dias após TLCA (77,68±3,38 µg/mg) quando comparados aos animais SHAM (53,46±4,58 µg/mg). O tamanho molecular dos GAGs foi significativamente maior nos animais OA que nos animais SHAM 70 dias após a cirurgia (p<0,01). GluChon preveniu tanto o aumento da quantidade de GAGs (54,42±5,39 µg/mg), quanto de seu tamanho molecular (p<0,05), e estes resultados acompanharam-se de melhora significativa da lesão histopatológica (p<0,05). Os resultados obtidos com Glu foram semelhantes, porém menos evidentes, e não alcançaram significância estatística. Glu aumentou significativamente a quantidade de NO na cavidade articular 7 dias após a TLCA. CONCLUSÕES: GluChon foi antinociceptivo e reduziu significativamente a incapacitação articular. A lesão da cartilagem articular no modelo de TLCA em ratos se acompanha de aumento da concentração e do tamanho molecular dos GAGs, e o tratamento com GluChon previne essas alterações e reduz o dano histopatológico. GluChon foi mais eficaz do que Glu tanto na redução da dor quanto da lesão da cartilagem articular, sugerindo benefício da combinação sobre o uso isolado de sulfato de glicosamina. / OBJECTIVES: Evaluate the antinociceptive and chondroprotective effects of glucosamine sulfate and chondroitin sulfate. METHODS: Male Wistar rats underwent anterior cruciate ligament transection (ACLT) or sham operation of the right knee. Animals were treated p.o. with glucosamine sulfate (Glu) 500 mg/kg, the combination of glucosamine sulfate 500 mg/kg and chondroitin sulfate 400 mg/kg (GluChon), or saline, since 7 days prior to surgery until the sacrifice 70 days after ACLT. A positive control group received meloxicam 6 mg/kg s.c. for antinociceptive evaluation comparisons. Joint pain was evaluated with rat incapacitation test. Animals were sacrificed 7,14,28 or 70 days after ACLT. The severity of histopathologic lesions was evaluated on femoral condyles after hematoxylin-eosin or toluidine blue staining with OARSI grading and staging system. Cartilage extracted glycosaminoglycans (GAGs) concentration was assessed after agarose gel electrophoresis. GAGs\' molecular weight was evaluated after polyacrylamide gel electrophoresis. NO release in synovial fluids was assessed 7 days after ACLT. RESULTS: GluChon reduced joint pain (p<0.01). Glu also reduced joint pain, but results did not reach statistical significance. A significant increase in articular cartilage\'s GAGs concentration was observed in OA animals (77.68±3.38 µg/mg) as compared to sham (53.46±4.58 µg/mg). OA animals also had significantly higher molecular weight GAGs than sham (p<0.01). GluChon prevented both GAG concentration and molecular weight elevations in OA animals, and that was associated with significant less cartilage damage assessed by histopathologic examination (p<0.05). Glu effects were less evident, and did not reach statistical significance. Glu was associated with significant higher concentrations of NO in synovial fluid. CONCLUSIONS: GluChon was antinociceptive on the ACLT model in rats. Articular cartilage damage was associated with increased amounts of GAG with higher molecular weight, and the prevention of these alterations with GluChon treatment was associated with less histopathologic damage. GluChon was more efficient than Glu for both pain and articular cartilage damage reduction, suggesting that combined treatment is better than glucosamine sulfate alone.

<p>Osteoarthritis (OA) is a degenerative joint disease characterized by cartilage damage and loss in the joints that affects approximately 27 million adults in the US. Tissue that is damaged by OA is a major health concern since cartilage tissue has a limited ability to self-repair due to the lack of vasculature in cartilage and low cell content. Tissue engineering efforts aim towards the development of cartilage repair strategies that mimic articular cartilage and are able to halt the progression of the disease as well as restore cartilage to its normal function.</p><p>This study harnesses the biological activity of collagen type II, present in articular cartilage, and the superior mechanical properties of collagen type I by characterizing gels made of collagen type I and II blends (1:0, 3:1, 1:1, 1:3, and 0:1). The collagen blend hydrogels were able to incorporate both types of collagen and retain chondroitin sulfate (CS) and hyaluronic acid (HA). Cryoscanning electron microscopy images showed that the 3:1 ratio of collagen type I to type II gels had a lower void space percentage (36.4%) than the 1:1 gels (46.5%) and the complex modulus was larger for the 3:1 gels (G*=5.0 Pa) compared to the 1:1 gels (G*=1.2 Pa). The 3:1 blend consistently formed gels with superior mechanical properties compared to the other blends and has the potential to be implemented as a scaffold for articular cartilage engineering.</p> <p>Following the work done to characterize the collagen scaffolds, we studied whether an aggrecan mimic, CS-GAHb, composed of CS and HA binding peptides, GAH, and not its separate components, is able to prevent glycosaminoglycan (GAG) and collagen release when incorporated into chondrocyte-embedded collagen gels. Bovine chondrocytes were cultured and embedded in collagen type I scaffolds with CS, GAH, CS and GAH, or CS-GAHb molecules. Gels composed of 3:1 collagen type I and II with CS or CS-GAHb were also studied. The results obtained showed CS-GAHb is able to decrease GAG and collagen release and increase GAG retention in the gels. CS-GAHb also stimulated cytokine production during the initial days of scaffold culture. However, the addition of CS-GAHb into the chondrocyte-embedded collagen scaffolds did not affect ECM protein expression in the gels. The incorporation of collagen type II into the collagen type I scaffolds did not significantly affect GAG and cytokine production and ECM protein synthesis, but did increase collagen release. The results suggest the complex interaction between CS-GAHb, the chondrocytes, and the gel matrix make these scaffolds promising constructs for articular cartilage repair.</p> <p>Finally, we used Dunkin Hartley guinea pigs, a commonly used animal model of osteoarthritis, to determine if high frequency ultrasound can ensure intra-articular injections of the aggrecan mimic are accurately positioned in the knee joint. A high-resolution small animal ultrasound system with a 40 MHz transducer was used for image-guided injections. We assessed our ability to visualize important anatomical landmarks, the needle, and anatomical changes due to the injection. From the ultrasound images, we were able to visualize clearly the movement of anatomical landmarks in 75% of the injections. The majority of these showed separation of the fat pad (67.1%), suggesting the injections were correctly delivered in the joint space. The results demonstrate this image-guided technique can be used to visualize the location of an intra-articular injection in the joints of guinea pigs and we are able to effectively inject the aggrecan mimic into knee joints.</p><p>All of the work presented here suggests that the addition of the aggrecan mimic to collagen I and collagen I and II scaffolds has shown that this type of construct could be useful for treating cartilage damage in the future.</p>

Glycosaminoglycans are heterogeneous polysaccharides that mediate important biological functions. There has been considerable interest in deciphering the precise GAG sequences that are responsible for protein interactions. In fact, several GAG oligosaccharides have been discovered to date as targeting proteins with higher level of specificity. Yet, it has been difficult to develop GAG oligosaccharides as drugs. One of the key reasons for this state of art is that GAG synthesis is extremely challenging and is highly structure-specific. Thus, much of the biology and pharmacology of GAG remains unknown and unexploited to date. An alternative approach is to prepare GAG oligosaccharides using enzymatic depolymerization of polymeric GAGs. GAG lyases, including heparinases and chondritinases represent powerful tools that can theoretically generate multiple oligosaccharides in parallel. However, it is difficult to implement such procedures with high consistency. Moreover, GAG lyases can digest GAGs down to disaccharides. A priori, non-polymeric GAGs, or alternatively GAG oligosaccharides containing 4 to 10 residues, would be expected to function better as therapeutic agents because they would be more homogeneous and less non-specific than their polymeric precursors. Thus, we reasoned that immobilization of these enzymes may engineer altered biopolymer processing, which may afford longer oligosaccharides in higher proportions and greater consistency. Heparinase-I and chondroitinase ABC were immobilized on CNBr-activated Sepharose and compared with the free form of the enzyme. Immobilized GAG lyases retained high efficiency of depolymerization over a wide range of pH, temperature and reusability. Most importantly, the immobilized enzyme was found to produce larger proportions of oligosaccharides longer than di- and tetra-saccharides as compared to lyases in the free form. A two dimensional separation involves size exclusion chromatography followed by reversed phase ion-pairing ultra performance liquid chromatography coupled to electrospray ionization mass spectrometry was employed to separate and characterize oligosaccharide structures. We have identified 40 heparin oligosaccharides, including regular and rare structures ranging from dp4 to dp10 and 39 chondroitin sulfate oligosaccharides in high homogeneity and significant yields. Overall, this technology is likely to offer a simple and cost effective route to preparation of larger amounts of sequences that can be expected to bind and modulate protein function.

La réparation endovasculaire (EVAR) est une technique minimalement invasive permettant de traiter l’anévrisme de l’aorte abdominale (AAA) par l’entremise d’un stent- graft (SG). L’utilisation d’EVAR est actuellement limitée par de fréquentes complications liées à une guérison inadéquate autour de l’implant. Ce manque de guérison est principalement dû au type de recouvrement polymérique des SG, au milieu pro-apoptotique des AAA et à l’accès réduit aux nutriments et à l’oxygène après EVAR. L’objectif de cette thèse consistait à concevoir un revêtement bioactif permettant d’inhiber l’apoptose et stimuler la croissance des cellules musculaires lisses vasculaires (CMLV), pour ainsi favoriser la guérison des tissus vasculaires autour des SG. La chondroïtine-4-sulfate (CS) a d’abord été choisie, car elle a été identifiée comme un médiateur important de la réparation vasculaire. Il a été démontré que la CS en solution influence directement la résistance à l’apoptose des CMLV, en plus de favoriser la différenciation myofibroblastique chez les fibroblastes. Dans le cadre de ce projet, un premier revêtement à base de CS et de collagène a été créé. Bien que le revêtement permettait d’induire une résistance à l’apoptose chez les CMLV, il se désintégrait trop rapidement dans des conditions aqueuses. Une nouvelle méthodologie a donc été adaptée afin de greffer la CS directement sur des surfaces aminées, à l’aide d’un système utilisant un carbodiimide. Dans le but d’accroître la croissance des CMLV à la surface des revêtements, le facteur de croissance de l’épiderme (EGF) a ensuite été sélectionné. En plus de ses propriétés mitogéniques et chimiotactiques, l’EGF stimule la production d’éléments de la matrice extracellulaire, comme le collagène et la fibronectine. De plus, l’activation du récepteur de l’EGF inhibe également l’apoptose des CMLV. L’EGF a donc été greffé sur la CS. Le revêtement de CS+EGF a démontré une bonne uniformité et bioactivité sur des surfaces de verre aminé. iii iv Dans une 3ème étape, afin de permettre de transposer ce revêtement bioactif sur des implants, plusieurs méthodes permettant de créer des groupements d’amines primaires sur les biomatériaux polymériques comme le PET ou le ePTFE ont été étudiées. La polymérisation par plasma a été choisie pour créer le revêtement CS+EGF à la surface de PET. Une fois de plus, celui-ci a permis d’inhiber l’apoptose des CMLV, dans des conditions pro-apoptotiques, et de favoriser la croissance des cellules. Le revêtement de CS et d’EGF, déposé sur des surfaces aminées, possède des caractéristiques biologiques intéressantes et semble donc prometteur pour favoriser une meilleure guérison autour des SG. / Endovascular aneurysm repair (EVAR) is a minimally invasive technique performed to treat abdominal aortic aneurysm (AAA) through the use of a stent-graft (SG). The usage of EVAR is presently limited by postoperative complications related to an incomplete healing of the surrounding tissues. The materials currently used in SG, the pro- apoptotic phathophysiology of AAA and the limited access to nutrients and oxygen, all limit the wound healing process and proper tissue ingrowth around the implant. The main objective of this thesis was to create of a bioactive coating inhibiting cell apoptosis and increasing vascular smooth muscle cells (VSMC) growth, to promote healing of the vascular tissues surrounding SG. Chondroitin sulfate (CS) was chosen since recent findings have shown that this polysaccharide triggers key mechanisms involved in vascular repair. CS in solution was shown to inhibit apoptosis of VSMC, as well as stimulate myofibroblast differentiation. A coating of CS and collagen was first created for the purpose of this work. Although the coating was shown to increase cell resistance to apoptosis with VSMC, it was not stable enough, since it rapidly disintegrated in aqueous solutions. A new methodology was thus proposed, where CS was grafted right on aminated surfaces, through carbodiimide chemistry. Epidermal growth factor (EGF) was then chosen to increase VSMC growth on the coatings. EGF is a known mitogenic and chemotactif growth factor for VSMC. It also stimulates the production of extracellular matrix elements, such as collagen and fibronectin. The activation of EGF receptor (EGFR) also triggers various cell signalling pathways modulating VSMC resistance to apoptosis. EGF was thus grafted on CS. CS+EGF coating on aminated glassed slides was shown to be uniform and bioactive. Finally, several methodologies to produce primary amines on polymeric biomaterials, such as PET and ePTFE, were studied in order to eventually transfer the v vi coating on implants. Plasma polymerization was chosen to create the CS+EGF coating. Once again the coating was shown to decrease VSMC apoptosis, in apoptotic conditions, and favour cell growth. Overall, the CS and EGF coating on aminated surfaces possesses interesting biological features and is a promising avenue to stimulate vascular healing around SG.

Submitted by (lucia.rodrigues@ufrpe.br) on 2016-06-01T16:45:23Z No. of bitstreams: 1 Fernanda Katharine de Souza Lins Borba.pdf: 4059966 bytes, checksum: cce20c494a8d5b3926e4508d0ff50750 (MD5) / Made available in DSpace on 2016-06-01T16:45:23Z (GMT). No. of bitstreams: 1 Fernanda Katharine de Souza Lins Borba.pdf: 4059966 bytes, checksum: cce20c494a8d5b3926e4508d0ff50750 (MD5) Previous issue date: 2012-02-29 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / Chondroitin Sulfate (CS) and Glucosamine Sulfate (GS) are functional constituents of vertebrate tissues. GS is an amino sugar and CS is part of the glucosaminoglycans group (GAGs). Studies have suggested CS and GS to have anti-inflammatory properties, however it has also been shown that these compounds promote scarring and proliferation of fibroblasts, which express molecules important for blood vessel growth (angiogenesis). This study was aimed at evaluating the effects of CS and GS on in vitro models regarding cell viability (cytotoxicity - MTT), proliferation (BrdU incorporation) and differentiation (tubulogenesis in Matrigel support) on human umbilical vein endothelial cells (HUVEC line). In vivo angiogenesis was also evaluated in (1) extraembryonic membranes of Gallus domesticus (number of chorioallantoic vessels - CAM assay and vitelinic YSM assay; and fractal geometry analysis); (2) and subcutaneous tissue of adult mice (Mus muscullus) by hemoglobin quantification (Spectroscopy) in Gelfoam implants. In the HUVEC assay, both CS and GS (1-3000 g/mL) displayed partial cytotoxic effect (~50% viability), but only in the highest tested concentrations (3000 and 1000 g/mL). It was observed that CS (3 g/mL), but not GS, promoted proliferation and tubulogenesis of HUVEC in 40% (P < 0.05) and 64% (P < 0.05), respectively, relative to control (RPMI-1640 medium). These effects did not significantly differ from the respective 28% and 53% promoted by the well known angiogenic growth factor FGF-2 (50 ng/mL). In the in vivo vasculoangiogenesis YSM assay on 2 to 4-day old embryos, GS (0.001-0.1mg/disk) and, to a lesser extent, CS (0.030-0.1mg/disk) increased the amount of vessels relative to control (P < 0.05). The effects of administration of CS and GS (0.1mg/disk) did not differ from what was observed in groups treated with 50 ng/mL FGF2. In the CAM angiogenesis assay on 6 to 8-days old embryos, again both CS and GS increased the amount of vessels relative to control, but only in concentrations as high as 2.0 mg/disk. This effect was no different from what was observed in groups treated with 50 ng/mL FGF2. The pro-angiogenic effects of CS (2 mg/disk) in embryonary angiogenesis were confirmed in the advanced angiogenesis of mice: only the group treated with CS (2 mg/implant) displayed a significant increase in the amount of blood vessels, expressed as hemoglobin content (0.52 ± 0.08g/dL), relative to control (vehicle; PBS; 0.20 ± 0.07 g/dL). This pro-angiogenic effect was no different than that of FGF2 (0.53 ± 0.1g/dL). The in vitro and in vivo results indicate the pro-angiogenic properties of CS and GS. However, CS (GAG) was the more effective compound in the tests performed. As a constituent of proteoglycans, it is suggested that CS exerts its effects by interacting with FGF and other angiogenic factors in the extracellular matrix, stabilizing the receptor, and thus positively modulating the pro-angiogenic signal in endothelial cells. While the cellular mechanisms underlying CS and GS activity demand more specific research, there is an evident potential therapeutic use for both compounds in clinical situations, such as those related to vascular discrepancy. / Sulfato de glucosamina (SG) e Sulfato de condroitina (SC) são constituintes funcionais dos tecidos de vertebrados. O SG é um aminoaçúcar e o SC integra o grupo das glicosaminoglicanas (GAG). Estudos apontam propriedades antiinflamatórias do SC e SG, e demonstram ainda que essas substâncias promovem a cicatrização e a proliferação de fibroblastos, os quais expressam moléculas que atuam na formação de vasos sanguíneos (angiogênese). Os objetivos deste estudo foram avaliar a ação do SC e SG em modelos in vitro sobre a viabilidade (citotoxicidade pelo MTT), proliferação (incorporação por BrdU) e diferenciação (tubulogênese em suporte matrigel) na linhagem de células endoteliais de veia umbilical humana (HUVEC). Também se investigou a angiogênese in vivo: (1) em membranas anexas de embriões de Gallus domesticus (número de vasos corioalantóides - ensaio da CAM, e vitelínicos – ensaio da YSM; e análise por geometria fractal); (2) e no tecido subcutâneo de camundongos adultos por meio de quantificação da hemoglobina em implantes de Gelfoam. No ensaio com HUVEC, SC e SG (1-3000 g/mL) exerceram efeito citotóxico parcial (~50% de viabilidade), e somente nas respectivas maiores concentrações (3000 e 1000 g/mL). Verificou-se que o SC (3 g/mL), mas não o SG, estimulou a proliferação e a tubulogênese de HUVEC em 40% (p < 0,05) e em 64% (p < 0,05) respectivamente, em relação ao controle (meio RPMI-1640). Estes efeitos não diferiram estatisticamente dos 28% e 53%, respectivamente, promovidos pelo bem conhecido fator de crescimento angiogênico FGF-2 (50 ng/mL). No ensaio de vasculo-angiogênese na YSM de embriões de 2-4 dias de idade o SG (0,001-0,1mg/disco) principalmente, e o SC (0,030-0,1mg/disco) aumentaram o número de vasos em relação ao grupo controle (p < 0,05). Os efeitos da administração de SC e SG (0,1 mg/disco) não diferiram do observado no grupo tratado com 50 ng/mL de FGF-2. No ensaio de angiogênese na CAM de embriões de 6-8 dias de idade, ambos, SC e SG também elevaram o número de vasos em relação ao controle na concentração elevada de 2,0 mg/disco. Este efeito também não diferiu do observado no grupo exposto a 50 ng/mL de FGF- 2. O efeito pró-angiogênico do SC (2 mg/disco) na angiogênese embrionária foi confirmado na angiogênese avançada de camundongos adultos. Apenas o grupo que recebeu SC (2 mg/implante) mostrou um aumento significativo de vasos sanguíneos, expresso como conteúdo de hemoglobina (0,52 ± 0,08g/dL), comparado ao controle (veículo; PBS; 0,20 ± 0,07 g/dL). Este efeito pró-angiogênico não diferiu do obtido com FGF2 (0,53 ± 0.1g/dL). Os resultados in vitro e in vivo demonstram as propriedades pró-angiogênicas do SC e SG, contudo o SC (GAG) foi o mais efetivo nos ensaios. Como um constituinte de proteoglicanas, o SC sugere exercer seus efeitos pela interação com o FGF e outros fatores angiogênicos na matriz extracelular, estabilizando-os nos receptores e modulando assim, positivamente, o sinal pró-angiogênico nas células endoteliais. Embora mecanismos celulares subjacentes à atividade de SC e SG demandem mais estudos, evidencia-se um potencial papel terapêutico das duas substâncias em situações clínicas relacionadas à defasagem vascular.

Racional: A epilepsia do lobo temporal está comumente associada à farmacorresistência e tem a esclerose hipocampal como achado neuropatológico em mais da metade dos casos. Histologicamente, a esclerose hipocampal está associada à perda neuronal diferencial e gliose, além de alterações nos níveis de moléculas associadas à homeostase da água tecidual, como a aquaporina 4 e a molécula de matriz sulfato de condroitina. Em imagens de ressonância nuclear magnética, a esclerose é caracterizada por redução de volume em sequências ponderadas em T1, aumento de sinal e tempo de relaxamento em sequências ponderadas em T2 e redução na transferência de magnetização. Justificativa e Objetivos: Uma vez que tanto o sinal T2 quando a transferência de magnetização são dependentes da água tecidual, nosso objetivo é avaliar, na formação hipocampal de pacientes com epilepsia do lobo temporal, as correlações entre populações celulares e moléculas ligadas à homeostase da água e as imagens ponderadas em T2 e transferência de magnetização. Visamos ainda definir, na formação hipocampal de indivíduos sem alterações neuropatológicas, o volume de cada um dos subcampos hipocampais. Metodologia: Pacientes com epilepsia do lobo temporal farmacorresistente (ELT, n = 43), bem como voluntários sadios (controle radiológico, CH, n = 20), foram submetidos a exames de ressonância magnética em máquina de 3T para mensuração da volumetria hipocampal, tempo de relaxamento T2 e transferência de magnetização hipocampal (exames in vivo). Após o tratamento cirúrgico para o controle das crises, os hipocampos dos pacientes com ELT foram fixados por 8 dias e submetidos aos exames ex vivo em máquina de 3T para cálculo do tempo de relaxamento T2 de cada subcampo hipocampal. Hipocampos controle (Controle historadiológico, CHR, n = 14), foram obtidos de autópsias de pacientes sem histórico ante-mortem de doença neurológica ou presença de patologia no exame do encéfalo pos mortem. Ambos os grupos controle foram pareados para idade em relação ao grupo ELT. Alguns dos casos CHR (n = 6) foram também submetidos à imagem 3D T2 em máquina de 4,7T para cálculo de volumetria dos subcampos hipocampais. Após emblocamento em parafina, secções coronais hipocampais dos casos CHR e ELT foram submetidas às técnicas de histoquímica básica Hematoxilina e Eosina e Luxol Fast Blue, e às imuno-histoquímicas para avaliação das populações neuronais (NeuN), astrócitos reativos (GFAP), micróglias ativadas (HLA-DR) e para a expressão de aquaporina 4 (AQP4) e níveis de sulfato de condroitina (CS-56). Para a comparação entre os grupos, foram realizados testes t para dados paramétricos e Mann-Whitney para dados não-paramétricos. Testes de correlação foram empregados para análise da associação entre as avaliações histológicas e os exames de ressonância magnética. Resultados: Pacientes com ELT apresentaram menor volume hipocampal, maior tempo de relaxamento T2 e menor transferência de magnetização no exame in vivo, quando comparados com o CR. O exame ex vivo para a volumetria dos subcampos hipocampais em casos do grupo CHR indicou que a fascia dentata, a região CA1 e o subículo correspondem à 85 % do volume hipocampal total. Quanto ao tempo de relaxamento T2 ex vivo, foi observado aumento em todos os subcampos hipocampais do grupo ELT, à exceção da fascia dentata, quando comparados ao CHR. A avaliação da densidade neuronal indicou redução significativa em todos os subcampos dos casos ELT, à exceção do subículo, quando comparados ao CHR. Em relação aos valores do grupo CHR, foi observada astrogliose em quase todos subcampos da formação hipocampal (a exceção da zona subgranular e do hilo) e microgliose em todos os subcampos (exceto pelo subículo) dos casos com ELT. Pacientes com ELT apresentaram redução na expressão de aquaporina 4 perivascular em todos os subcampos do hipocampo, comparados ao CHR. Aumento nos níveis de sulfato de condroitina foi observado em todos os subcampos da formação hipocampal, à exceção da camada granular, nos pacientes com ELT. O volume hipocampal e a transferência de magnetização in vivo dos pacientes com ELT correlacionaram-se tanto com a população neuronal como com os níveis de sulfato de condroitina, enquanto que o tempo de relaxamento in vivo correlacionou-se com a população astroglial e os níveis de sulfato de condroitina. O exame ex vivo corroborou a correlação entre a população glial e o tempo de relaxamento observado nos pacientes com ELT. A diferença entre o tempo de relaxamento in vivo e ex vivo correlacionou-se tanto com a difusibilidade da água no tecido como com os níveis de sulfato de condroitina. Conclusões: Nossos dados indicam correlação entre a patologia hipocampal e as imagens de ressonância nuclear magnética, sendo que a maior qualidade das imagens ex vivo permitiu uma avaliação mais direta entre o sinal de ressonância e a patologia, indicando importância da população celular e matriz extracelular para o volume hipocampal e a transferência de magnetização, e da astrogliose para o tempo de relaxamento T2. Finalmente, nossos dados mostraram que CA1, subículo e fascia dentata tem grande participação no volume hipocampal, sendo que alterações nestas regiões tem um papel mais relevante nas alterações observadas na ressonância magnética, como indicado por nossas correlações. / Rationale: Drug resistant temporal lobe epilepsy is often associated with hippocampal sclerosis. Histological evaluation reveals differential neuronal loss, gliosis and changes in molecules associated with water homeostasis, such as aquaporin 4 and chondroitin sulfate. Magnetic resonance imaging in these cases often reveals hippocampal atrophy, increased T2 signal and T2 relaxation and reduced magnetization transfer ratio in the hippocampus. Aims: Once both T2 signal and magnetization transfer are affected by tissue water, our goal was to evaluate, in the hippocampus of drug-resistant temporal lobe epilepsy patients who underwent surgery for seizure control, the associations between cellular populations, aquaporin 4 and chondroitin sulfate with T2 relaxation time and magnetization transfer. Additionally, we intended to measure the individual volume of each hippocampal subfield in hippocampus from patients without neurological disease. Methods: Patients with drug-resistant temporal lobe epilepsy (TLE, n = 43) and age-matched health volunteers (radiological control, RC, n = 20) were submitted to magnetic resonance in a 3T machine for hippocampal volumetry measure, T2 relaxation and magnetization transfer (in vivo examination). After surgical treatment for seizure control, hippocampi from the TLE patients were fixed in formalin for 8 days and then submitted to ex vivo imaging in 3T for relaxation time of every hippocampal subfield. Control hippocampi were obtained from autopsies of age-matched patients without ante mortem history of neurological disease or post mortem neurological pathology, and underwent the same ex vivo imaging (histo-radiological control, HRC, n = 14). Six cases from the HRC underwent 3D T2 imaging in a 4.7T machine, in order to measure the volumes of the hippocampal subfields. Paraffin embedded hippocampal sections from TLE and HRC were submitted to Hematoxilin-Eosin and Luxol Fast Blue histochemistries, and to immunohistochemistries for the evaluation of neurons (NeuN), reactive astrocytes (GFAP), activated microglia (HLA-DR), for aquaporin 4 (AQP4) and for chondroitin sulfate (CS-56). Students t-test or Mann-Whitneys test were performed for comparison between groups, and correlation tests were performed for the comparison between histological and magnetic resonance measures. Results: Patients with TLE presented reduced hippocampal volume, increased T2 relaxation time and reduced magnetization transfer, when compared to RC. The ex vivo volumetry of the hippocampal subfields revealed that fascia dentata, CA1 and subiculum together correspond to 85 % of the total hippocampal volume. Ex vivo relaxation time, as the in vivo, were increased in the subfields of TLE patients, when compared to HRC. Compared to HRC, TLE patients presented neuron loss and microgliosis in all hippocampal subfields but the subiculum, and astrogliosis in all hippocampal subfields but the subgranule zone and the hilus. Reduced perivascular aquaporin 4 was observed in all hippocampal subfields of TLE patients, and increased chondroitin sulfate was observed in all hippocampal subfields, with the exception of granule cell layer, of TLE patients, when compared to HRC. In TLE, both in vivo hippocampal volume and magnetization transfer correlated with the levels of chondroitin sulfate and the neuronal population, whereas the in vivo relaxation time correlated with the astroglial population and the levels of chondroitin sulfate. Ex vivo relaxation time also correlated with the astroglial population in TLE patients. The difference between in vivo and ex vivo relaxation values correlated with water difusibility and the levels of chondroitin sulfate. Conclusion: Our data indicate the importance of neuron population and extracellular matrix to both hippocampal volume and magnetization transfer, and of the reactive astrocytes for T2 relaxation. Ex vivo relaxation time allowed a more detailed evaluation, and indicated more robust correlations between reactive astrocytes and T2 relaxation. Finally, Our data indicated that CA1, the subiculum and fascia dentata are the major contributors to hippocampal volume, so changes in these subfields most likely will affect magnetic resonance imaging.