The limbal stem cell population is located in the limbal junctional zone between the cornea and the conjunctiva, and is responsible for maintaining the corneal epithelium. Damage to the limbal stem cell population results in a condition known as limbal stem cell deficiency (LSCD), which is characterised by conjunctivalisation of the cornea, visual impairment and persistent irritation. To treat LSCD, an alternative source of human limbal epithelial (HLE) cells must be transplanted back onto the diseased cornea. Limbal tissue grafts have had a moderate degree of success. However, autologous grafts risk damage to the healthy eye, whilst allogeneic grafts are susceptible to immunological rejection. Cultured HLE grafts offer a promising alternative to whole tissue grafts. The production of cultured HLE grafts involves the removal of a small (1-2 mm2) biopsy from the patient’s healthy limbus, followed by ex vivo expansion to produce an epithelial sheet, which is subsequently transplanted onto the damaged corneal surface. However, the production of cultured HLE grafts usually requires the addition of animal-derived products during cell culture. Animal-derived components, such as foetal bovine serum (FBS) and murine 3T3 feeder cells, introduce the patient to potential crossspecies infection and immune responses to xenogeneic antigens. Consequently, the overall aim of this project has been to develop a culture technique free of xenogeneic products for the establishment and propagation of HLE cells. To achieve this aim, alternatives to FBS in the culture medium and 3T3 feeder cells were pursued. A defined serum-free medium (SFM) containing vitronectin (VN), insulin-like growth factor binding protein 3 (IGFBP3), insulin-like growth factor-I (IGF-I), and epidermal growth factor (EGF) was investigated as an alternative to serumsupplemented medium (SSM) for HLE cell culture. Initial studies focused on the effects of these growth factors on HLE cell metabolic activity and migration. Metabolic activity was primarily stimulated by IGF-I and EGF, with the combination of IGF-I and EGF in solution stimulating metabolic activity to a significantly greater extent than the SSM positive control (p = 0.006). HLE cell migration was also effected by combinations of VN, IGFBP3, IGF-I and EGF. Migration was stimulated above the SFM negative control by the combination of IGFBP3 and IGF-I either with or without the addition of EGF. However, the presence of VN was required for optimal migratory responses (p < 0.003). Hepatocyte growth factor (HGF) and keratinocyte growth factor (KGF) were also investigated as additional components to the SFM formulation. HGF significantly stimulated HLE cell metabolic activity and migration (p < 0.02). In contrast, KGF did not significantly stimulate either HLE cell metabolic activity or migration. The addition of either HGF or KGF to the SFM supplemented with VN, IGFBP3, IGF-I and EGF did not significantly enhance the metabolic activity of HLE cells. Therefore, HGF and KGF were no longer pursued as additional components to the SFM formulation. Additional studies were conducted to examine the efficacy of replacing murine 3T3 feeder cells with human ocular stromal cells during HLE cell culture. Initially, stromal cells were isolated from the cornea, limbus and sclera to determine whether there were differences between these stromal cell populations. The results indicated that scleral stromal cells had a significantly larger area and perimeter than either corneal or limbal stromal cells (p < 0.001). Scleral stromal cells were also significantly more rounded than either corneal or limbal stromal cells, as determined by the elliptical factor equation (p < 0.001). Immunocytochemistry also revealed that scleral stromal cells expressed significantly more of the myofibroblast marker ..- smooth muscle actin than either corneal or limbal stromal cells (p < 0.001), and significantly less of the fibroblast/myofibroblast marker Thy-1 than corneal or limbal stromal cells (p < 0.001). Therefore, scleral stromal cells were identified as different in comparison to corneal and limbal stromal cells. Primary HLE cells were cultured with irradiated corneal, limbal and scleral stromal cells. HLE cultures established with either corneal or limbal stromal feeder cells contained more cellular protein (as measured by rhodamine B dye absorbance) than cultures established without feeder cells (p < 0.001). The colony forming efficiency (CFE) of HLE cells established with corneal or limbal stromal feeder cells was also significantly greater than HLE cells established without feeder cells (p < 0.001). In contrast, HLE cultures established with scleral stromal feeder cells contained low levels of cellular protein and had a low CFE, which was not significantly different to the HLE cultures established without feeder cells. Immunocytochemistry indicated that HLE cultures established with scleral feeder cells also showed lower expression of the stem cell markers ABCG2 and C/EBP ... These results suggest that freshly isolated HLE cells can be cultured with irradiated corneal or limbal stromal cells as a replacement for murine 3T3 feeder cells. Finally, the SFM supplemented with VN+IGFBP3+IGF-I+EGF was combined with limbal stromal feeder cells, and examined as a culture technique free of animalderived products. Freshly isolated HLE cells established in SFM supplemented with VN+IGFBP3+IGF-I+EGF and limbal feeder cells contained a similar amount of cellular protein (as measured by crystal violet dye absorbance) when compared to the SSM+3T3 positive control. In addition, the CFE of freshly isolated HLE cells established in VN+IGFBP3+IGF-I+EGF and limbal feeder cells was significantly higher than the SSM+3T3 positive control (p = 0.004). However, a live/dead assay revealed a reduced HLE cell viability in SFM supplemented with VN+IGFBP3+IGFI+ EGF and limbal feeder cells after seven days in culture. In addition, immunocytochemistry demonstrated a lower expression of the stem cell markers ABCG2 and C/EBP .. in the SFM treatment with limbal feeder cells. Therefore, freshly isolated HLE cells can be cultured in SFM supplemented with VN+IGFBP3 +IGF-I+EGF and limbal feeder cells. However, this culture technique is less likely to support the growth of immature limbal stem cells when compared to the SSM+3T3 positive control. Overall, this research has attempted to create a culture system free of animal-derived products for the production of cultured HLE grafts to treat limbal stem cell deficiency. The results show that HLE cells respond to a serum-free medium formulation containing VN+IGFBP3+IGF-I+EGF. In addition, this culture medium can be combined with irradiated stromal cells isolated from the limbus to support HLE culture production. However, the combination of VN+IGFBP3+IGF-I+EGF and limbal feeder cells demonstrated a reduced viability, which indicates that further refinement of the formulation is required. This thesis has also demonstrated differences between stromal cells isolated from the cornea, limbus, and sclera, and has generated knowledge which may impact on the understanding of stromalepithelial regulation.

Optimiser le diagnostic paraclinique prédictif et non invasif du déficit en cellules souches limbiques (DCSL) : in vitro après empreinte cytologique (EC) par la recherche immunohistochimique (IH) de marqueurs pertinents et par reverse-transcription polymerase chain reaction (RT-PCR), technique de haute sensibilité ; in vivo par microscopie confocale (MCIV). Matériel et Méthodes: La preuve diagnostique de DCSL est la présence de cellules conjonctivales au sein de la cornée, la persistance de cellules cornéennes signant un DCSL partiel. L'EC a été standardisée. La spécificité de chacun des marqueurs cornéens ou conjonctivaux sélectionnés a été recherchée sur des tissus normaux avant d'aborder une étude prospective monocentrique, sur 22 yeux de 18 patients cliniquement suspects de DCSL. Les cellules épithéliales de la cornée centrale étaient recueillies par EC puis transférées sur lame. L'expression des marqueurs de différentiation conjonctivale (K7, K13, K19, MUC5AC) et cornéenne (K12) a été recherchée par IH sur les 22 EC et celle de la MUC5AC par RT-PCR sur 4 d'entre elles. Enfin, la cornée et le limbe de 7 patients ont étés analysés par MCIV. La concordance entre les techniques est évaluée. Conclusion: Ces techniques complémentaires dépendent de l'équipement du service et de l'accessibilité à un laboratoire compétent. Un organigramme décisionnel a été établi pour permettre de faire un diagnostic fiable de DCSL, les examens paracliniques étant inutiles dans certains cas. La recherche IH de kératines conjonctivales K7/K13 et/ou la détection de MUC5AC par RT-PCR sur cellules cornéennes recueillies par EC sont des techniques diagnostiques de haute valeur scientifique, tandis que l'IH K12/MUC5AC doit être abandonnée pour manque de sensibilité et celle de K3/K19 pour manque de spécificité. La MCIV, quand elle est réalisable, est une technique hautement sensible qui permet une quantification du degré de sévérité de la maladie avec une forte concordance avec les autres examens / Purpose: to optimize minimally invasive techniques for diagnosing limbal stem cell deficiency: in vitro after impression cytology (IC) by means of immunocytochemical detection of relevant markers and reverse-transcription polymerase chain reaction (RT PCR), an highly sensitive method; in vivo by confocal microscopy (IVCM). Material and Methods: Presence of conjunctival cells in central cornea was diagnosis proof of LSCD, whereas corneal epithelial cells’ remaining traduces partial LSCD. IC was labeled. Corneal or conjunctival specificity of each marker was previously assessed on healthy tissues. In a prospective case control observational study, 22 eyes of 18 patients clinically suspected of LSCD were enrolled. Epithelial cells from central cornea were collected by IC. Conjunctival (K7, K13, K19, MUC5AC) and corneal (K12) differentiation markers were assessed by immunocytochemistry on each of 22 eyes and MUC5AC RT-PCR was assessed for 4 of them. Cornea and limbus of 7 eyes were assessed by IVCM. The inter-examination agreement was determined. Conclusion: These techniques require skilled technicians and laboratory facilities. We propose a decision tree model to provide unfailing LSCD diagnosis, complementary examinations being sometimes useless. Clinical examination can lead to LSCD misdiagnosis. Immunostaining of conjunctival keratins K7 and K13 as well as MUC5AC detection by RT-PCR are highly effective methods whereas MUC5AC/K12 immuno staining are not sensible and both K3 and K19 are not specific. IVCM of great sensitivity if realizable allows quantification of LSCD severity. Combining both methods provides in every case unfailing diagnosis of LSCD and evaluation of the extent of the disease with high agreement

Aims: Limbal stem cell (LSC) deficiency is one of the most challenging ocular surface diseases. The aim of this thesis was to study damaged ocular surface reconstruction. Therefore, a mouse model of limbal transplantation was estab- lished. Furthermore, LSC isolation, transfer of LSCs and bone marrow-derived mesenchymal stem cells (MSCs) on nanofiber scaffolds were studied. Material and methods: Syngeneic, allogeneic and xenogeneic (rat) limbal grafts were transplanted orthotopically into BALB/c mice. Graft survival, immune re- sponse and the effect of monoclonal antibodies (mAb) (anti-CD4 and anti-CD8 cells) were analyzed. Mouse LSCs were separated by Percoll gradient; subse- quently, they were analyzed for the presence of LSC and differentiation corneal epithelial cell markers and characteristics using real-time PCR and flow cytom- etry. Nanofiber scaffolds seeded with LSCs and MSCs were transferred onto the damaged ocular surface in mouse and rabbit models. Cell growth on scaffolds, post-operative inflammatory response and survival of transferred cell were ana- lyzed. Results: Limbal allografts were rejected promptly by the Th1-type of immune response (IL-2, IFN-γ) involving CD4+ cells and nitric oxide produced by macro- phages, contrary to the prevailing Th1 and Th2 immune responses (IL-4, IL-10) in...

Aims: Limbal stem cell (LSC) deficiency is one of the most challenging ocular surface diseases. The aim of this thesis was to study damaged ocular surface reconstruction. Therefore, a mouse model of limbal transplantation was estab- lished. Furthermore, LSC isolation, transfer of LSCs and bone marrow-derived mesenchymal stem cells (MSCs) on nanofiber scaffolds were studied. Material and methods: Syngeneic, allogeneic and xenogeneic (rat) limbal grafts were transplanted orthotopically into BALB/c mice. Graft survival, immune re- sponse and the effect of monoclonal antibodies (mAb) (anti-CD4 and anti-CD8 cells) were analyzed. Mouse LSCs were separated by Percoll gradient; subse- quently, they were analyzed for the presence of LSC and differentiation corneal epithelial cell markers and characteristics using real-time PCR and flow cytom- etry. Nanofiber scaffolds seeded with LSCs and MSCs were transferred onto the damaged ocular surface in mouse and rabbit models. Cell growth on scaffolds, post-operative inflammatory response and survival of transferred cell were ana- lyzed. Results: Limbal allografts were rejected promptly by the Th1-type of immune response (IL-2, IFN-γ) involving CD4+ cells and nitric oxide produced by macro- phages, contrary to the prevailing Th1 and Th2 immune responses (IL-4, IL-10) in...

The damage of ocular surface represents one of the most common causes of decreased quality of vision or even blindness. If the injury is extensive and includes the region of limbus, niche of limbal stem cells (LSC), LSC deficiency occurs and the natural corneal regeneration is stopped. Conjunctival epithelium migrates into the injured area. Neovascularization, local inflammation and corneal opacity occur. Corneal transplantation is an insufficient treatment in such case. If the injury is bilateral, the allogenic limbal graft or LSC transplantation is required. In such cases systemic immunosuppressive drugs with many negative side-effects must be administered. The search for an adequate autologous substitution is important for avoid immunosuppressive medication. Mesenchymal stem cells (MSC) represent a perspective substitution for the reason of their immunomodulatory properties and the capability to differentiate in many cell types. There is possibility to isolate sufficient number of these cells from adipose tissue or bone marrow which are relatively easily accessible. Our goal was to observe local production of cytokines and other molecules which are present in inflammatory reaction after the chemical burn of the murine cornea and after the treatment with stem cells growing on nanofiber scaffold....

Certaines brûlures oculaires graves ou d'autres pathologies oculaires rares peuvent être associées à une perte totale de cellules souches épithéliales de la cornée (DCSL), ce qui conduit à une opacification de la cornée par invasion de la conjonctive. Lorsque la DCSL est totale et bilatérale, le limbe controlatéral n'est pas disponible pour la greffe de limbe autologue ou la culture de cellules souches autologues limbiques et la transplantation allogénique de la cornée est impossible car toujours rejetée en raison de la néovascularisation. Une thérapie innovante testée avec succès au sein de notre laboratoire, en collaboration avec le service d'ophtalmologie des HCL, consiste en une greffe autologue de Feuillet Epithélial (FE) dérivé de Muqueuse Orale (MO). Cette approche permet de restaurer la transparence et autorise, si nécessaire, une greffe cornéenne complémentaire. Cette technique a montré son efficacité lors d'un essai clinique conduit dans notre hôpital mais le dispositif breveté permettant le détachement non enzymatique des feuillets cultivés n'est plus disponible en Europe. Nous avons donc mis au point un nouveau procédé de production de FE dérivant de la MO dont la preuve de concept a été obtenue à partir d'études in-vitro et ex-vivo. En effet, un détachement avec 0,5 mg / mL de collagénase n'endommage pas le FE de MO et les protéines de la membrane basale, et, dans un modèle de stroma porcin exvivo, ces feuillets adhérents au stroma, continuent à se renouveler sous la forme d'épithélium différencié. De plus, dans le cas d'une opacité stromale associée à une DCSL, une greffe secondaire de cornée est nécessaire pour améliorer l'Acuité Visuelle (AV), c'est pourquoi nous avons cherché à développer pour ces patients à haut risque de rejet, un stroma décellularisé. Ce stroma pourrait également répondre à la pénurie de cornées dans les pays en voie de développement grâce à sa conservation longue et facile. La lyophilisation a été combinée à la décellularisation des cornées au SDS à 0,1 %, validée sur les cornées humaines sur le maintien de leur transparence et de l'ultrastructure du stroma associé à l'absence des antigènes HLA-ABC et HLA-DR. Enfin, dans un modèle de Kératoplastie Lamellaire Antérieure Profonde (KLAP) ex-vivo, nous avons montré que les cellules épithéliales et stromales de la cornée humaine receveuse colonise le stroma décellularisé. Ainsi, nos travaux permettent de proposer un traitement des DSCL totales et bilatérales associées ou non à une atteinte du stroma / Some severe ocular burns or other rare ocular pathologies may be associated with a complete loss of corneal epithelial stem cells (LSCD), leading to an opacification of the cornea by invasion of the conjunctiva. When DCSL is total and bilateral, contralateral limbus is not available for autologous limb transplant or limbal autologous stem cell culture, and allogeneic transplantation of the cornea is not an option since neovascularization is constantly responsible for graft rejection. An innovative therapy tested with success in our laboratory, in collaboration with the ophthalmology department of the HCL, consists in an autologous Epithelial Cell Sheet (ECS) graft derived from Oral Mucosa (MO). This approach restores transparency and allows in a second step a complementary corneal graft when necessary. This technique has been shown to be effective in a clinical trial conducted in our hospital, but the patented device for the non-enzymatic detachment of cultivated ECS is no longer available in Europe. We have therefore developed a new method for the production of ECS from OM, the proof of concept of which has been obtained from in-vitro and ex-vivo studies. Indeed, detachment with 0.5 mg / mL of collagenase does not damage ECS from OM and basement membrane proteins, and in an ex-vivo porcine stroma model, these cell sheets adhered on corneal stroma, continued to self-renew and generated a differentiated epithelium. Since stromal opacity associated with DCSL requires a secondary corneal graft to improve Visual Acuity (VA), we also sought to develop for these patients with high rejection risk, a new approach for the generation of decellularized stroma. Such a procedure for the production of decellularized stroma was also aimed at allowing a money-saving and reliable long-term storage for stromal grafts and thus circumventing the shortage of corneas in developing countries. Our process, combining lyophilization with decellularization of the corneas at 0.1 % SDS, was validated on human corneas regarding the maintenance of stroma transparency, the stromal ultrastructure associated with the absence of HLA-ABC and HLA-DR antigens. Finally, in an ex-vivo model of Deep Anterior Lamellar Keratoplasty (DALK), we have shown that the epithelial and stromal cells of the recipient human cornea colonized efficiently the decellularised stroma. Overall, our work makes it possible to propose a treatment for total and bilateral LSCD associated or not with lesions of the stroma