Markers of Bone Turnover In Preclinical Development of Drugs for Skeletal Diseases

Skeletal tissue is constantly remodeled in a process where osteoclasts resorb old bone and osteoblasts form new bone. Balance in bone remodeling is related to age, gender and genetic factors, but also many skeletal diseases, such as osteoporosis and cancer-induced bone metastasis, cause imbalance in bone turnover and lead to decreased bone mass and increased fracture risk. Biochemical markers of bone turnover are surrogates for bone metabolism and may be used as indicators of the balance between bone resorption and formation. They are released during the remodeling process and can be conveniently and reliably measured from blood or urine by immunoassays. Most commonly used bone formation markers include N-terminal propeptides of type I collagen (PINP) and osteocalcin, whereas tartrate-resistant acid phosphatase isoform 5b (TRACP 5b) and C-terminal cross-linked telopeptide of type I collagen (CTX) are common resorption markers. Of these, PINP has been, until recently, the only marker not commercially available for preclinical use. To date, widespread use of bone markers is still limited due to their unclear biological significance, variability, and insufficient evidence of their prognostic value to reflect long term changes. In this study, the feasibility of bone markers as predictors of drug efficacy in preclinical osteoporosis models was elucidated. A non-radioactive PINP immunoassay for preclinical use was characterized and validated. The levels of PINP, N-terminal mid-fragment of osteocalcin, TRACP 5b and CTX were studied in preclinical osteoporosis models and the results were compared with the results obtained by traditional analysis methods such as histology, densitometry and microscopy. Changes in all bone markers at early timepoints correlated strongly with the changes observed in bone mass and bone quality parameters at the end of the study. TRACP 5b correlated strongly with the osteoclast number and CTX correlated with the osteoclast activity in both in vitro and in vivo studies. The concept “resorption index” was applied to the relation of CTX/TRACP 5b to describe the mean osteoclast activity. The index showed more substantial changes than either of the markers alone in the preclinical osteoporosis models used in this study. PINP was strongly associated with bone formation whereas osteocalcin was associated with both bone formation and resorption. These results provide novel insight into the feasibility of PINP, osteocalcin, TRACP 5b and CTX as predictors of drug efficacy in preclinical osteoporosis models. The results support clinical findings which indicate that short-term changes of these markers reflect long-term responses in bone mass and quality. Furthermore, this information may be useful when considering cost-efficient and clinically predictive drug screening and development assays for mining new drug candidates for skeletal diseases.

Preclinical development of cancer vaccines

Triplex cell vaccine is a cancer immunopreventive cell vaccine that can prevent almost completely mammary tumor onset in HER-2/neu transgenic mice. A future translation of cancer immunoprevention from preclinical to clinical studies should take into account several aspects. The work reported in this thesis deals with the study of three of these aspects: vaccine schedule, activity in a therapeutic set-up and second-generation DNA vaccines. An important element in determining human acceptance and compliance of a treatment protocol is the number of vaccinations. In order to improve the vaccination schedule a minimal protocol was searched, i.e. a schedule consisting of a lower number of administrations than standard protocol but with a similar efficacy. A candidate optimal protocol was identified by the use of an in silico model, SimTriplex simulator. The in vivo test of this schedule in HER-2/neu transgenic mice only partially confirmed in silico predictions. This result shows that in silico models have the potential ability to aid in searching of optimal treatment protocols, provided that they will be further tuned on experimental data. As a further result this preclinical study highlighted that kinetic of antibody response plays a major role in determining cancer prevention, leading to the hypothesis of a threshold that must be reached rapidly and maintained lifetime. Early clinical trials would be performed in a therapeutic, rather than preventive, setting. Thus, the activity of Triplex vaccine was investigated against experimental lung metastases in HER-2/neu transgenic mice in order to evaluate if the immunopreventive Triplex vaccine could be effective also against a pre-existing tumor mass. This preclinical model of aggressive metastatic development showed that the vaccine was an efficient treatment also 4 for the cure of micrometastases. However the immune mechanisms activated against tumor mass were not antibody dependent, i.e. different from those preventing the onset of primary mammary carcinoma. DNA vaccines could be more easily used than cellular ones. A second generation of Triplex vaccine based on DNA plasmids was evaluated in an aggressive preclinical model (BALBp53neu female mice) and compared with the preventive ability of cellular Triplex vaccine. It was observed that Triplex DNA vaccine was as effective as Triplex cell vaccine, exploiting a more restricted immune stimulation.

Preclinical development of anti-cancer strategies against human HER-2

HER-2 is a 185 kDa transmembrane receptor tyrosine kinase that belongs to the EGFR family. HER-2 is overexpressed in nearly 25% of human breast cancers and women with this subtype of breast cancer have a worse prognosis and frequently develop metastases. The progressive high number of HER-2-positive breast cancer patients with metastatic spread in the brain (up to half of women) has been attributed to the reduction in mortality, the effectiveness of Trastuzumab in killing metastatic cells in other organs and to its incapability to cross the blood-brain barrier. Apart from full-length-HER-2, a splice variant of HER-2 lacking exon 16 (here referred to as D16) was identified in human HER-2-positive breast cancers. Here, the contribution of HER-2 and D16 to mammary carcinogenesis was investigated in a model transgenic for both genes (F1 model). A dominant role of D16, especially in early stages of tumorigenesis, was suggested and the coexistence of heterogeneous levels of HER-2 and D16 in F1 tumors revealed the undeniable value of F1 strain as preclinical model of HER-2-positive breast cancer, closer resembling the human situation in respect to previous models. The therapeutical efficacy of anti-HER-2 agents, targeting HER-2 receptor (Trastuzumab, Lapatinib, R-LM249) or signaling effectors (Dasatinib, UO126, NVP-BKM120), was investigated in models of local or advanced HER-2-positive breast cancer. In contrast with early studies, data herein collected suggested that the presence of D16 can predict a better response to Trastuzumab and other agents targeting HER-2 receptor or Src activity. Using a multiorgan HER-2-positive metastatic model, the efficacy of NVP-BKM120 (PI3K inhibitor) in blocking the growth of brain metastases and the oncolytic ability of R-LM249 (HER-2-retargeted HSV) to reach and destroy metastatic HER-2-positive cancer cells were shown. Finally, exploiting the definition of “oncoantigen” given to HER-2, the immunopreventive activity of two vaccines on HER-2-positive mammary tumorigenesis was demonstrated.

Metabolomics and its potential in drug development

Metabolomics is the global and unbiased survey of the complement of small molecules (say, <1 kDa) in a biofluid, tissue, organ or organism and measures the end-products of the cellular metabolism of both endogenous and exogenous substrates. Many drug candidates fail during Phase II and III clinical trials at an enormous cost to the pharmaceutical industry in terms of both time lost and of financial resources. The constantly evolving model of drug development now dictates that biomarkers should be employed in preclinical development for the early detection of likely-to-fail candidates. Biomarkers may also be useful in the preselection of patients and through the subclassification of diseases in clinical drug development. Here we show with examples how metabolomics can assist in the preclinical development phases of discovery, pharmacology, toxicology, and ADME. Although not yet established as a clinical trial patient prescreening procedure, metabolomics shows considerable promise in this regard. We can be certain that metabolomics will join genomics and transcriptomics in lubricating the wheels of clinical drug development in the near future.

In vivo imaging of glycol chitosan-based nanogel biodistribution

The preclinical development of nanomedicines raises several challenges and requires a comprehensive characterization. Among them is the evaluation of the biodistribution following systemic administration. In previous work, the biocompatibility and in vitro targeting ability of a glycol chitosan (GC) based nanogel have been validated. In the present study, its biodistribution in the mice is assessed, using near-infrared (NIR) fluorescence imaging as a tool to track the nanogel over time, after intravenous administration. Rapid whole body biodistribution of both Cy5.5 labeled GC nanogel and free polymer is found at early times. It remains widespreadly distributed in the body at least up to 6 h postinjection and its concentration then decreases drastically after 24 h. Nanogel blood circulation half-life lies around 2 h with the free linear GC polymer presenting lower blood clearance rate. After 24 h, the blood NIR fluorescence intensity associated with both samples decreases to insignificant values. NIR imaging of the organs shows that the nanogel had a body clearance time of 48 h, because at this time point a weak signal of NIR fluorescence is observed only in the kidneys. Hereupon it can be concluded that the engineered GC nanogel has a fairly long blood circulation time, suitable for biomedical applications, namely, drug delivery, simultaneously allowing efficient and quick body clearance.

Pregnancy malaria: cryptic disease, apparent solution

Malaria during pregnancy can be severe in non-immune women, but in areas of stable transmission, where women are semi-immune and often asymptomatic during infection, malaria is an insidious cause of disease and death for mothers and their offspring. Sequelae, such as severe anaemia and hypertension in the mother and low birth weight and infant mortality in the offspring, are often not recognised as consequences of infection. Pregnancy malaria, caused by Plasmodium falciparum, is mediated by infected erythrocytes (IEs) that bind to chondroitin sulphate A and are sequestered in the placenta. These parasites have a unique adhesion phenotype and distinct antigenicity, which indicates that novel targets may be required for development of an effective vaccine. Women become resistant to malaria as they acquire antibodies against placental IE, which leads to higher haemoglobin levels and heavier babies. Proteins exported from the placental parasites have been identified, including both variant and conserved antigens, and some of these are in preclinical development for vaccines. A vaccine that prevents P. falciparum malaria in pregnant mothers is feasible and would potentially save hundreds of thousands of lives each year.

Therapeutic nanoparticles in clinics and under clinical evaluation.

This article reviews nanoparticulate-chemotherapeutic systems that have been developed for human therapy, considering the components of the nanoparticles, the therapeutic agents associated with the nanoparticles and the clinical indications these therapeutic nanoparticles have been developed for. In this evaluation we have put into perspective the types of nanomaterials and their therapeutic indications. We have reviewed the nanoparticulate-chemotherapeutic systems that have been published, approved and marketed and that are currently in clinical use. We have also analyzed the nanoparticulate-chemotherapeutic systems that are in clinical trials and under preclinical development.

Investigation of femtosecond laser technology for the fabrication of drug nanocrystals in suspension

La technique du laser femtoseconde (fs) a été précédemment utilisée pour la production de nanoparticules d'or dans un environnement aqueux biologiquement compatible. Au cours de ce travail de maîtrise, cette méthode a été investiguée en vue d'une application pour la fabrication de nanocristaux de médicament en utilisant le paclitaxel comme modèle. Deux procédés distincts de cette technologie à savoir l'ablation et la fragmentation ont été étudiés. L'influence de la puissance du laser, de point de focalisation, et de la durée du traitement sur la distribution de taille des particules obtenues ainsi que leur intégrité chimique a été évaluée. Les paramètres ont ainsi été optimisés pour la fabrication des nanoparticules. L’évaluation morphologique et chimique a été réalisée par microscopie électronique et spectroscopie infrarouge respectivement. L'état cristallin des nanoparticules de paclitaxel a été caractérisé par calorimétrie differentielle et diffraction des rayons X. L'optimisation du procédé de production de nanoparticules par laser fs a permis d'obtenir des nanocristaux de taille moyenne (400 nm, polydispersité ≤ 0,3). Cependant une dégradation non négligeable a été observée. La cristallinité du médicament a été maintenue durant la procédure de réduction de taille, mais le paclitaxel anhydre a été transformé en une forme hydratée. Les résultats de cette étude suggèrent que le laser fs peut générer des nanocristaux de principe actif. Cependant cette technique peut se révéler problématique pour des médicaments sensibles à la dégradation. Grâce à sa facilité d'utilisation et la possibilité de travailler avec des quantités restreintes de produit, le laser fs pourrait représenter une alternative valable pour la production de nanoparticules de médicaments peu solubles lors des phases initiales de développement préclinique. Mots-clés: paclitaxel, nanocristaux, laser femtoseconde, ablation, fragmentation

Semisynthetic aminoglycoside antibiotics : toward biomimetic synthesis, evasion of bacterial resistance and reduced toxicity

Les antibiotiques aminoglycosidiques sont des agents bactéricides de grande valeur et d’efficacité à large spectre contre les pathogènes Gram-positifs et Gram-négatifs, dont plusieurs membres naturels et semisynthétiques sont importants dans l’histoire clinique depuis 1950. Des travaux crystallographiques sur le ribosome, récompensés par le prix Nobel, ont démontré comment leurs diverses structures polyaminées sont adaptées pour cibler une hélice d’ARN dans le centre de codage de la sous-unité 30S du ribosome bactérien. Leur interférence avec l’affinité et la cinétique des étapes de sélection et vérification des tARN induit la synthèse de protéines à basse fidélité, et l’inhibition de la translocation, établissant un cercle vicieux d’accumulation d’antibiotique et de stress sur la membrane. En réponse à ces pressions, les pathogènes bactériens ont évolué et disséminé une panoplie de mécanismes de résistance enzymatiques et d’expulsion : tels que les N acétyltransférases, les O phosphotransférases et les O nucleotidyltransférases qui ciblent les groupements hydroxyle et amino sur le coeur des aminoglycosides; des méthyl-transférases, qui ciblent le site de liaison ribosomale; et des pompes d’expulsion actives pour l’élimination sélective des aminoglycosides, qui sont utilisés par les souches Gram-négatives. Les pathogènes les plus problématiques, qui présentent aujourd’hui une forte résilience envers la majorité des classes d’antibiotiques sur le bord de la pan-résistance ont été nommés des bactéries ESKAPE, une mnémonique pour Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa et Enterobacteriaceae. La distribution globale des souches avec des mécanismes de résistance envers les standards cliniques aminoglycosides, tels que la tobramycine, l’amikacine et la gentamicine, est comprise entre 20 et 60% des isolées cliniques. Ainsi, les aminoglycosides du type 4,6-disubstitués-2-deoxystreptamine sont inadéquats comme thérapies anti-infectieuses à large spectre. Cependant, la famille des aminoglycosides 4,5-disubstitués, incluant la butirosine, la neomycine et la paromomycine, dont la structure plus complexe, pourrait constituter une alternative. Des collègues dans le groupe Hanessian et collaborateurs d’Achaogen Inc. ont démontré que certains analogues de la paraomomycine et neomycine, modifiés par désoxygénation sur les positions 3’ et 4’, et par substitution avec la chaîne N1-α-hydroxy-γ-aminobutyramide (HABA) provenant de la butirosine, pourrait produire des antibiotiques très prometteurs. Le Chapitre 4 de cette dissertation présente la conception et le développement d’une stratégie semi-synthétique pour produire des nouveaux aminoglycosides améliorés du type 4,5 disubstitués, inspiré par des modifications biosynthétiques de la sisomicine, qui frustrent les mécanismes de résistance bactérienne distribuées globalement. Cette voie de synthèse dépend d’une réaction d’hydrogénolyse de type Tsuji catalysée par palladium, d’abord développée sur des modèles monosaccharides puis subséquemment appliquée pour générer un ensemble d’aminoglycosides hybrides entre la neomycine et la sisomicine. Les études structure-activité des divers analogues de cette nouvelle classe ont été évaluées sur une gamme de 26 souches bactériennes exprimant des mécanismes de résistance enzymatique et d’expulsion qui englobe l’ensemble des pathogènes ESKAPE. Deux des antibiotiques hybrides ont une couverture antibacterienne excellente, et cette étude a mis en évidence des candidats prometteurs pour le développement préclinique. La thérapie avec les antibiotiques aminoglycosidiques est toujours associée à une probabilité de complications néphrotoxiques. Le potentiel de toxicité de chaque aminoglycoside peut être largement corrélé avec le nombre de groupements amino et de désoxygénations. Une hypothèse de longue date dans le domaine indique que les interactions principales sont effectuées par des sels des groupements ammonium, donc l’ajustement des paramètres de pKa pourrait provoquer une dissociation plus rapide avec leurs cibles, une clairance plus efficace et globalement des analogues moins néphrotoxiques. Le Chapitre 5 de cette dissertation présente la conception et la synthèse asymétrique de chaînes N1 HABA β substitutées par mono- et bis-fluoration. Des chaînes qui possèdent des γ-N pKa dans l’intervalle entre 10 et 7.5 ont été appliquées sur une neomycine tétra-désoxygénée pour produire des antibiotiques avancés. Malgré la réduction considérable du γ N pKa, le large spectre bactéricide n’a pas été significativement affecté pour les analogues fluorés isosteriques. De plus, des études structure-toxicité évaluées avec une analyse d’apoptose propriétaire d’Achaogen ont démontré que la nouvelle chaîne β,β difluoro-N1-HABA est moins nocive sur un modèle de cellules de rein humain HK2 et elle est prometteuse pour le développement d’antibiotiques du type neomycine avec des propriétés thérapeutiques améliorées. Le chapitre final de cette dissertation présente la proposition et validation d’une synthèse biomimétique par assemblage spontané du aminoglycoside 66-40C, un dimère C2 symétrique bis-imine macrocyclique à 16 membres. La structure proposée du macrocycle a été affinée par spectroscopie nucléaire à un système trans,trans-bis-azadiène anti-parallèle. Des calculs indiquent que l’effet anomérique de la liaison α glycosidique entre les anneaux A et B fournit la pré-organisation pour le monomère 6’ aldéhydo sisomicine et favorise le produit macrocyclique observé. L’assemblage spontané dans l’eau a été étudié par la dimérisation de trois divers analogues et par des expériences d’entre croisement qui ont démontré la généralité et la stabilité du motif macrocyclique de l'aminoglycoside 66-40C.

Nanonization of megestrol acetate by laser fragmentation in aqueous milieu

Faculté de Pharmacie

Antitumor activity of Titanocene Y against freshly explanted human breast tumor cells and in xenografted MCF-7 tumors in mice

Bis-[(p-methoxybenzyl)cyclopentadienyl] titanium dichloride, better known as Titanocene Y, is a newly synthesized transition metal-based anticancer drug. We studied the antitumor activity of Titanocene Y with concentrations of 2.1, 21 and 210 mu mol/l against a freshly explanted human breast cancer, using an in-vitro soft agar cloning system. The sensitivity against Titanocene Y was highly remarkable in the breast cancer tumor in the full concentration range. Titanocene Y showed cell death induction at 2.1 mu mol/l, well comparable to cisplatin, given at a concentration of 1.0 mu mol/l. A further preclinical development of Titanocene Y was warranted and therefore an MCF-7 human breast cancer xenograft nonobese diabetic/severe combined immunodeficient mouse model was used. Titanocene Y was given for 21 days at 30 mg/kg/ day (75% of the maximum tolerable dose of Titanocene Y), which resulted in the reduction of the tumor volume to around one-third, whereas no mouse was lost because of the surprisingly low toxicity of Titanocene Y.

Ruthenium(II) complexes with 2-(diphenylphosphinomethyl)aniline

2-(Diphenylphosphinomethyl)aniline. H2L1, reacts with [RuCl2(PPh3)(3)] to yield the monomeric complexes [RuCl2(H2L1)(PPh3)(CH3CN)], [RuCl2(H2L1)(2)]and the chloro-bridged dimer [(H2L1)(PPh3)Ru(mu-Cl)(2)Ru(PPh3) (H2L1)] depending on the conditions applied. Exclusively the monochelate [RuCl2 (H2L1)(dmso)(2)] is formed during reactions of H2L1 with [RuCl2(dmso)(4)]. H2L1 acts as a neutral, bidentate ligand in all complexes. The products are studied spectroscopically and by X-ray diffraction. (C) 2012 Elsevier Ltd. All rights reserved.

Gamma-secretase modulators and inhibitors in Alzheimer's disease: influence of genetic background on efficacy and mechanism of action

According to the amyloid hypothesis, Alzheimer’s disease (AD) is caused by aberrant production or clearance of the amyloid-β (Aβ) peptides, and in particular of the longer more aggregation-prone Aβ42. The Aβ peptides are generated through successive proteolytic cleavage of the amyloid precursor protein (APP) by the β-site APP cleaving enzyme (BACE) and γ-secretase. γ-secretase produces Aβ peptides with variable C-termini ranging from Aβ34 to Aβ48, presumably by sequential trimming of longer into shorter peptides. γ-secretase is a multiprotein complex consisting of at least four different proteins and the presenilin proteins (PS1 or PS2) contain the catalytic center of the complex. In 2001 several non-steroidal anti-inflammatory drugs were identified as the founding members of a new class of γ-secretase modulators (GSMs) that can selectively reduce production of Aβ42. Concomitantly, these GSMs increase Aβ38 production indicating closely coordinated generation of Aβ42 and Aβ38 and a potential precursor-product relationship between these peptides. GSMs seem to exert their activity by direct modulation of γ-secretase. Support for this hypothesis is drawn from the finding that some PS mutations associated with early-onset familial AD (FAD) can modulate the cellular response to GSMs and to γ-secretase inhibitors (GSIs), which inhibit production of all Aβ peptides and are known to directly interact with PS. A particularly interesting FAD PS mutation is PS1-ΔExon9, a complex deletion mutant that blocks endoproteolysis of PS1 and renders cells completely non-responsive to GSMs. Studies presented in this thesis show that the diminished response of PS1-ΔExon9 to GSMs is mainly caused by its lack of endoproteolytic cleavage. Furthermore, we were able to demonstrate that a reduced response to GSMs and GSIs is not limited to PS1-ΔExon9 but is a common effect of aggressive FAD-associated PS1 mutations. Surprisingly, we also found that while the Aβ42 response to GSMs is almost completely abolished by these PS1 mutations, the accompanying Aβ38 increase was indistinguishable to wild-type PS1. Finally, the reduced response to GSIs was confirmed in a mouse model with transgenic expression of an aggressive FAD-associated PS1 mutation as a highly potent GSI failed to reduce Aβ42 levels in brain of these mice. Taken together, our findings provide clear evidence for independent generation of Aβ42 and Aβ38 peptides, and argue that the sequential cleavage model might be an oversimplification of the molecular mechanism of γ-secretase. Most importantly, our results highlight the significance of genetic background in drug discovery efforts aimed at γ-secretase, and indicate that the use of cellular models with transgenic expression of FAD-associated PS mutations might confound studies of the potency and efficacy of GSMs and GSIs. Therefore, such models should be strictly avoided in the ongoing preclinical development of these promising and potentially disease-modifying therapeutics for AD.

Development of human central nervous system 3D in vitro models for preclinical research

Neurological disorders are a major concern in modern societies, with increasing prevalence mainly related with the higher life expectancy. Most of the current available therapeutic options can only control and ameliorate the patients’ symptoms, often be-coming refractory over time. Therapeutic breakthroughs and advances have been hampered by the lack of accurate central nervous system (CNS) models. The develop-ment of these models allows the study of the disease onset/progression mechanisms and the preclinical evaluation of novel therapeutics. This has traditionally relied on genetically engineered animal models that often diverge considerably from the human phenotype (developmentally, anatomically and physiologically) and 2D in vitro cell models, which fail to recapitulate the characteristics of the target tissue (cell-cell and cell-matrix interactions, cell polarity). The in vitro recapitulation of CNS phenotypic and functional features requires the implementation of advanced culture strategies that enable to mimic the in vivo struc-tural and molecular complexity. Models based on differentiation of human neural stem cells (hNSC) in 3D cultures have great potential as complementary tools in preclinical research, bridging the gap between human clinical studies and animal models. This thesis aimed at the development of novel human 3D in vitro CNS models by integrat-ing agitation-based culture systems and a wide array of characterization tools. Neural differentiation of hNSC as 3D neurospheres was explored in Chapter 2. Here, it was demonstrated that human midbrain-derived neural progenitor cells from fetal origin (hmNPC) can generate complex tissue-like structures containing functional dopaminergic neurons, as well as astrocytes and oligodendrocytes. Chapter 3 focused on the development of cellular characterization assays for cell aggregates based on light-sheet fluorescence imaging systems, which resulted in increased spatial resolu-tion both for fixed samples or live imaging. The applicability of the developed human 3D cell model for preclinical research was explored in Chapter 4, evaluating the poten-tial of a viral vector candidate for gene therapy. The efficacy and safety of helper-dependent CAV-2 (hd-CAV-2) for gene delivery in human neurons was evaluated, demonstrating increased neuronal tropism, efficient transgene expression and minimal toxicity. The potential of human 3D in vitro CNS models to mimic brain functions was further addressed in Chapter 5. Exploring the use of 13C-labeled substrates and Nucle-ar Magnetic Resonance (NMR) spectroscopy tools, neural metabolic signatures were evaluated showing lineage-specific metabolic specialization and establishment of neu-ron-astrocytic shuttles upon differentiation. Chapter 6 focused on transferring the knowledge and strategies described in the previous chapters for the implementation of a scalable and robust process for the 3D differentiation of hNSC derived from human induced pluripotent stem cells (hiPSC). Here, software-controlled perfusion stirred-tank bioreactors were used as technological system to sustain cell aggregation and dif-ferentiation. The work developed in this thesis provides practical and versatile new in vitro ap-proaches to model the human brain. Furthermore, the culture strategies described herein can be further extended to other sources of neural phenotypes, including pa-tient-derived hiPSC. The combination of this 3D culture strategy with the implemented characterization methods represents a powerful complementary tool applicable in the drug discovery, toxicology and disease modeling.

Development and preclinical assessment of a bioartificial pancreas.

Transplantation of insulin secreting cells is regarded as a possible treatment for type 1 diabetes. One major difficulty in this approach is, however, that the transplanted cells are exposed to the patient's inflammatory and autoimmune environment, which originally destroyed their own beta-cells. Therefore, even if a good source of insulin-secreting cells can be identified for transplantation therapy, these cells need to be protected against these destructive influences. The aim of this project was to evaluate, using a clonal mouse beta-cell line, whether genetic engineering of protective genes could be a viable option to allow these cells to survive when transplanted into autoimmune diabetic mice. We demonstrated that transfer of the Bcl-2 anti-apoptotic gene and of several genes specifically interfering with cytokines intracellular signalling pathways, greatly improved resistance of the cells to inflammatory stresses in vitro. We further showed that these modifications did not interfere with the capacity of these cells to correct hyperglycaemia for several months in syngeneic or allogeneic streptozocin-diabetic mice. However, these cells were not protected against autoimmune destruction when transplanted into type 1 diabetic NOD mice. This suggests that in addition to inflammatory attacks by cytokines, autoimmunity very efficiently kills the transplanted cells, indicating that multiple protective mechanisms are required for efficient transplantation of insulin-secreting cells to treat type 1 diabetes.