Tesi sul tema "Human Induced neurons"

La maladie d'Alzheimer (MA) est le principal type de démence et représente un défi majeur pour la santé publique mondiale. Elle se caractérise par un déclin progressif de la cognition, de la mémoire et des fonctions comportementales et touche plus de 55 millions de personnes dans le monde. Au niveau moléculaire, la MA se définit par la présence d'enchevêtrements neurofibrillaires agrégés dans les neurones et par l'accumulation de plaques d'amyloïde-β (Aβ) dans le cerveau. Ces caractéristiques pathologiques sont associées à des altérations de l'activité neuronale, à la perte de synapses, à la gliose et à la neuroinflammation, conduisant à une neurodégénérescence irréversible. L'étiologie et la physiopathologie de la MA impliquent une interaction complexe entre des facteurs génétiques et environnementaux. Les études d'association à l'échelle du génome ont permis d'identifier plusieurs loci porteurs de polymorphismes de nucléotides simples (SNP) associés au risque de maladie d'Alzheimer. Parmi ces loci, celui qui héberge la Protéine Tyrosine Kinase 2β (PTK2B) est mis en évidence dans le présent travail. Ce gène code pour une protéine tyrosine kinase qui est impliquée dans la régulation des canaux ioniques induite par le calcium et dans l'activation de nombreuses voies de signalisation, telles que la MAP kinase. Des variations génétiques non synonymes dans le locus PTK2B ont été associées à un risque accru de maladie d'Alzheimer et on pense qu'elles régulent l'expression de PTK2B. Cependant, les rôles physiologiques et physiopathologiques de la PTK2B ne sont pas entièrement compris. Dans le cerveau humain, l'expression de la PTK2B est principalement observée dans les neurones glutamatergiques. Au cours de la progression de la maladie d'Alzheimer, son expression diminue et peut contribuer aux dysfonctionnements neuronaux observés, tels que l'augmentation de l'excitabilité électrique et les altérations synaptiques. Par conséquent, la compréhension du rôle de la PTK2B dans les neurones humains peut contribuer à révéler les mécanismes des dysfonctionnements neuronaux dans la MA. Dans cette optique, les objectifs de cette thèse sont de découvrir les processus cellulaires et les voies moléculaires régulés par la PTK2B dans les neurones humains. Pour ce faire, nous avons utilisé des cellules souches pluripotentes induites humaines (hiPSC) isogéniques pour générer des neurones exprimant différents niveaux de PTK2B. Ensuite, nous avons utilisé des tests fonctionnels et moléculaires pour étudier les conséquences de l'altération de l'expression de la PTK2B dans un contexte physiologique et dans un contexte similaire à celui de la MA. Nous montrons qu'une réduction de l'expression de PTK2B entraîne une augmentation de la phosphorylation de TAU à divers épitopes associés à la pathologie de la MA, ce qui suggère un rôle central de PTK2B dans la régulation de l'agrégation de TAU. En utilisant la transcriptomique à noyau unique, nous montrons également que l'expression réduite de la PTK2B entraîne des altérations transcriptionnelles spécifiques liées à l'activité électrique neuronale et à la transmission synaptique, principalement dans les neurones glutamatergiques. Les expériences d'imagerie calcique indiquent que la réduction de l'expression de PTK2B contribue à augmenter la fréquence des pointes de calcium sans affecter la synchronisation, ce qui indique une activité électrique neuronale élevée. En outre, les résultats des enregistrements électrophysiologiques effectués à partir de réseaux multi-électrodes (MEA) montrent une activité électrique accrue et des schémas d'éclatement perturbés dans les neurones mutants PTK2B. Dans l'ensemble, ces travaux mettent en lumière l'implication de PTK2B dans les processus cellulaires liés à la maladie d'Alzheimer, en donnant un aperçu des mécanismes moléculaires et des altérations fonctionnelles associés à la dysrégulation de PTK2B dans les cellules neuronales humaines dérivées des iPSCs
Alzheimer's disease (AD) is the main type of dementia and poses a significant global public health challenge. It is characterized by a progressive decline in cognition, memory, and behavioral functions and affects more than 55 million people worldwide. At the molecular level, AD is defined by the presence of aggregated neurofibrillary tangles within neurons and the accumulation of amyloid-β (Aβ) plaques in the brain. These pathological features are associated with alterations in neuronal activity, synapse loss, gliosis, and neuroinflammation, leading to irreversible neurodegeneration. AD etiology and pathophysiology involves a complex interplay between genetic and environmental factors. Genome-Wide Association Studies have identified several loci carrying single nucleotide polymorphisms (SNPs) associated with AD risk. Among these loci, the one harboring the Protein Tyrosine Kinase 2β (PTK2B) is highlighted in the present work. This gene encodes a protein tyrosine kinase that is involved in calcium-induced regulation of ion channels and activation of numerous signaling pathways, such as MAP kinase. Non-synonimous genetic variations in the PTK2B locus have been associated with an increased risk of AD and are thought to regulate PTK2B expression. However, both the physiological and pathophysiological roles of PTK2B are not fully understood. In the human brain, PTK2B expression is mainly observed in glutamatergic neurons. Its expression declines during AD progression and may contribute to neuronal dysfunctions observed in the disease, such as increased electrical excitability and synaptic alterations. Therefore, understanding the role of PTK2B in human neurons may contribute to reveal the mechanisms of neuronal dysfunctions in AD. Considering that, the aims of this thesis are to uncover the cellular processes and molecular pathways regulated by PTK2B in human neurons. To that, we took advantage of isogenic human induced-pluripotent stem cells (hiPSCs) to generate neurons expressing different levels of PTK2B. Next, we employed functional and molecular assays to probe the consequences of altered PTK2B expression both in a physiological and in an AD-like context. We show that reduced PTK2B expression leads to increased TAU phosphorylation at various epitopes associated with AD pathology, suggesting a central role of PTK2B in regulating TAU aggregation. Using single-cell transcriptomics, we also show that reduced PTK2B expression leads to specific transcriptional alterations related to neuronal electrical activity and synaptic transmission mainly in glutamatergic neurons. Calcium imaging experiments indicate that PTK2B downregulation contributes to increased calcium spikes frequency without affecting synchronization, indicating an elevated neuronal electrical activity. Additionally, results from electrophysiological recordings from multi-electrode array (MEA) show increased electrical activity and disrupted bursting patterns in PTK2B mutant neurons. Overall, this work sheds light on the involvement of PTK2B in AD-related cellular processes, providing insights into the molecular mechanisms and functional alterations associated with PTK2B dysregulation in human iPSC-derived neural cells

Parkinson’s disease (PD) is an incurable, chronically progressive neurodegenerative disease leading to premature invalidity and death. The locomotor disability of PD patients is mainly rooted in the gradual and insidious degeneration of dopaminergic neurons (DA) projecting from the midbrain substantia nigra (SN) to the basal ganglia striatum, a pathological process highlighted microscopically by the formation of insoluble cytosolic protein aggregates, known as Lewy bodies and Lewy neurites. The pathogenic mechanisms leading to PD remain poorly understood, arguably owing to the lack of suitable animal and cellular experimental models of the disease. Therefore, there is an urgent need for developing reliable experimental models that recapitulate the key features of PD. The recent development of induced pluripotent stem cell (iPSC) technology has enabled the generation of patient-specific iPSC and their use to model human diseases, although it is currently unclear whether this approach could be useful to successfully model age-related conditions. Importantly, disease modeling using iPSC largely relies on the existence of efficient protocols for the differentiation of disease-relevant cell types. Here, we first developed an efficient protocol for the differentiation of iPSC to authentic midbrain-specific DA neurons with SN properties by forced expression of LMX1A using a lentivirus-mediated gene delivery system. Next, we generated an iPSC-based cellular model of PD that recapitulates key phenotypic features of PD, such as DA neuron loss and α-synuclein accumulation in DA neurons from PD patients. Overall, our results demonstrate that we have developed a valuable tool for elucidating the pathogenic mechanisms leading to PD, as well as an experimental platform for screening new drugs that may prevent or rescue neurodegeneration in PD.
La malaltia de Parkinson (MP) és una malaltia neurodegenerativa incurable que causa invalidesa i mort prematura. Els pacients de la malaltia de Parkinson presenten alteracions motores degudes a una degeneració gradual de les neurones dopaminèrgiques que projecten des de la substància nigra fins a l’estriat. A nivell microscòpic s’observa la presència d’agregats proteics insolubles en el citosol de les neurones coneguts com cossos o neurites de Lewy. Els mecanismes patològics responsables de la MP no es coneixen bé, possiblement a causa de la manca de models animals i cel•lulars adequats. Per tant, existeix una gran necessitat de desenvolupar models experimentals fiables que recapitulin les característiques bàsiques de la MP. El recent desenvolupament de les cèl•lules mare pluripotents induïdes (iPSC) ha permès la generació de iPSC específiques de pacient i el seu ús per modelar malalties humanes, ara bé, no és clar si aquesta estratègia es pot utilitzar per modelar exitosament malalties d’origen tardà, com ara la MP. És important destacar que el modelatge de malalties utilitzant iPSC, es basa, en gran mesura en l'existència de protocols eficients per a la diferenciació de les iPSC cap al tipus cel•lular rellevant per a la malaltia. Durant aquest període, per primera vegada, s’ha desenvolupat un protocol per a l’eficient diferenciació de les iPSC cap a neurones dopaminèrgiques amb les propietats característiques de neurones dopaminèrgiques nigrostriatals, mitjançant l’expressió forçada de LMX1A utilitzant vectors lentivirals. A continuació, s’ha generat un model cel•lular usant iPSC derivades de pacients de MP que recapitula les principals característiques fenotípiques de la malaltia, com ara la pèrdua de neurones dopaminèrgiques i l'acumulació de α-sinucleïna en les neurones dopaminèrgiques. En general, els nostres resultats demostren que hem desenvolupat una eina valuosa per a l’estudi dels mecanismes patològics que condueixen a la MP, així com una nova plataforma pel descobriment de nous fàrmacs encaminats a prevenir o evitar la neurodegeneració.

Questo lavoro di tesi ha avuto lo scopo di studiare lo sviluppo della neocorteccia umana ed i meccanisimi alla base della sua compromissione che risultano nell’insorgenza di patologie del neurosviluppo mediante un’analisi dei profili trascrizionali e della morfologia di neuroni neocorticali umani generati a partire da cellule staminali pluripotenti indotte (iPSCs). Data l’importanza di basarsi su un paradigma di neurogenesi in vitro riproducibile e affidabile nel generare neuroni neocoritcali umani autentici, prima di adottare questo sistema modello per lo studio di patologie del neurosviluppo, nella prima fase di questa ricerca abbiamo eseguito un’ampia caratterizzazione trascrizionale, molecolare e funzionale del protocollo di differenziamento. Le dinamiche trascrizionali che regolano il neurosviluppo in vitro sono state studiate effettuando esperimenti di RNA-sequencing sia a livello di popolazione che di singola cellula. In combinazione con diverse analisi bioinformatiche tra cui l’analisi delle componeti principali (PCA), l’analisi dei geni differenzialemtne espressi e l’analisi WGCNA. L’analisi dei profili trascrizionali è stata accompagnata da un’ampia analisi di d’immunocitochimica che ha permesso di confermare l’identità e lo stadio di sviluppo delle cellule in coltura. Inoltre, la maturità funzionale dei neuroni derivati da iPSCs è stata ulteriormente confermata dalla loro capacità di generare potenziali d’azione, sostenere pattern di scarica complessi e sviluppare attività sinaptica spontanea eccitatoria ed inibitoria. Complessivamente, i risultati ottenuti da questo ampio e diversificato pannello di analisi hanno permesso di stabilitre la riproducibilità del protocollo di differenziamento e la sua competenza nel generare con elevata efficienza principalmente neuroni neocorticali autentici. Successivamente abbiamo applicato questo protocollo di differenziamento neocorticale come sistema modello per studiare due patologie del neurosviluppo dovute alla delezione e duplicazione di una regione comprendente circa 1.5 - 1.8 Mb (megabasi) collacata sul braccio lungo (q) del cormosoma 7 nella banda 11.23. Duplicazioni e delezioni di questa regione sono di particolare interesse in quanto le due sindromi che ne risultano, rispettivamente la sindrome di Willams (WS) e la sindrome da duplicazione 7q11.23 (7q11DUP), presentano fenotipi cognitivi e comportamentali caratterizzati da profili simili e tratti simmetricamente opposti. La frequente comorbidità della sindrome da duplicazione 7q11.23 con altre patologie del neurosviluppo come l’autismo e la schizofrenia in contrasto con la sindrome di Williams che è una sindrome ben caraterizzata non associata ad altre patologie del neurosviluppo, rende lo studio dell’ alterato dosaggio genico del locus 7q11.23 estremamente interessante per identificare con precisione i meccanismi molecolari caratteristici di ciascuna condizione clinica, condivisi da entrabme le sindromi e comuni anche ad altre patologie del neurosviluppo. A questo scopo, abbiamo generato diverse linee di iPS a partire da un ampio gruppo di individui, comprendente individui sani e pazienti affetti dalla sindrome di Williams (WS) e dalla sindrome di duplizazione 7q11.23, che sono poi state differenziate in neuroni neocorticali applicando il protocollo precedentemetne caraterizzato. Confermata l’identità e l’autenticità dei neuroni neocorticali generati da iPSCs, stiamo attualmente identificando i geni ed i meccanismi molecolari disregolati in specifici sottotipi di neuroni che abbiano la maggior rilvenza clinica. Inoltre, l’analisi morfologica dei neuroni neocorticali umani ottenuti da pazienti WS e soggetti sani ha permesso di confermare nell’uomo molte alterazioni morfologiche dei neuroni neocorticali osservate in un modello murino knockout per Dnajc30, un gene ancora funzionalmente non caraterizzato compreso nel locus 7q11.23.
This research project has been aimed to investigate human neocortical development in healthy and diseased subjects by analyzing and comparing the transcriptional profiles and cellular morphologies of human neocortical cells derived from induced pluripotent stem cells (iPSCs). Given the importance to rely on a solid and highly reproducible iPSCs-based differentiation protocol that generates authentic neocortical neurons in vitro with high efficiency before applying it as a model system of human neurodevelopmental disorders, in the first phase of this study we performed a comprehensive transcriptional, cellular and physiological characterization of the in vitro neurodevelopmental paradigm. The transcriptional dynamics regulating in vitro neocortical development have been investigated by performing RNA-sequencing (RNA-seq) at both population and single- cell level in combination with several bioinformatics analyses including principal component analysis (PCA), differential gene expression analysis and weighted gene co-expression network analysis (WGCNA). The transcriptional results were corroborated by the widespread positivity for a selected panel of informative cell-fate and cell-stage specific markers detected through immunocytochemistry and the physiological maturity of our iPSCs-derived neocortical neurons was further confirmed by their ability to generate action potentials, develop complex firing patterns and sustain excitatory and inhibitory spontaneous synaptic activity. Overall, these results fully validated the reproducibility of the differentiation protocol and its efficiency and reliability in generating physiologically mature authentic neocortical neurons. Subsequently, we applied this extensively characterized neocortical differentiation paradigm to model in vitro two human neurodevelopmental disorders caused by symmetrical copy number variations (CNVs) of the Williams-Beuren syndrome chromosome region (WBSCR) located on the long arm (q) of chromosome 7 at position 11.23 (7q.11.23 locus). 7q11.23 CNVs are of special interest as the two disorders resulting from the deletion (Williams syndrome, WS) and duplication (7q.11.23 duplication syndrome, 7q11DUP) of this region exhibit cognitive and behavioral phenotypes marked by both similar features and symmetrically opposite traits. The association of 7q11DUP to complex neurodevelopmental disorders such as autism spectrum disorder and schizophrenia, while WS is a well-characterized syndrome without clear overlap to complex neurodevelopmental disorders make the study of this locus extremely interesting to identify the molecular mechanisms unique to each clinical condition, common to both syndromes and shared with other complex neurodevelopmental disorders. To this aim, we generated several iPSCs lines from a large cohort comprising WS individuals, 7q11DUP patients and healthy subjects and differentiated them into neocortical neurons by applying the previously in-depth characterized protocol. Having assessed the quality of our iPSCs-derived neocortical neurons, we are currently identifying neuronal subtypes specific genes and gene networks having the most statistically significant relationship to these disorders through single cell RNA-sequencing analysis. Furthermore, morphometric analysis of WS and control iPSCs-derived neocortical neurons has confirmed in humans many neuronal morphological abnormalities observed in a mouse knockout for Dnajc30, a previously uncharacterized gene contained in the 7q11.23 locus.

La sclérose latérale amyotrophique (SLA) est une maladie neurodégénérative incurable de l’adulte qui affecte principalement les motoneurones. Les mécanismes conduisant à la mort des motoneurones restent mal connus, notamment du fait de l'hétérogénéité de la maladie et du manque d'accès aux neurones humains affectés. La technologie des cellules souches pluripotentes induites humaines (iPSc) est un outil prometteur pour la modélisation de la SLA, car elle offre la possibilité unique d'obtenir et d’étudier des motoneurones humains.Des clones d’iPSc de deux sujets témoins ont été générés et nous avons comparé plusieurs protocoles afin de mettre au point un protocole efficace de différenciation des iPSc en motoneurones. Les cultures obtenues étaient hétérogènes et contenaient différents types de neurones et des précurseurs neuraux. Afin de pouvoir étudier des mécanismes intrinsèques aux motoneurones dans la SLA, nous avons développé une nouvelle technique pour purifier les motoneurones. Cette technique a consisté à trier les motoneurones par FACS en combinant l'utilisation d'un vecteur lentiviral rapporteur exprimant une protéine fluorescente sous le contrôle d'un promoteur spécifique des motoneurones, et d'un anticorps monoclonal dirigé contre le récepteur aux neurotrophines p75. Cette double sélection a permis l'isolement efficace de motoneurones purs. En parallèle, la technologie iPSc a été utilisée pour établir des modèles cellulaires de la SLA. Des clones de cellules iPS ont été générés à partir d’un patient avec une forme familiale de la SLA présentant une mutation dans le gène TARDBP (codant pour une protéine de liaison à l’ADN, TDP-43) et un patient atteint d’une forme sporadique de SLA. Afin de valider nos modèles, nous avons recherché des phénotypes caractéristiques de la maladie au cours de la différenciation des iPSc : i) la formation d’agrégats cytoplasmiques, ii) des altérations de génération et de survie des motoneurones, iii) des défauts de croissance neuritique
Amyotrophic lateral sclerosis (ALS) is a fatal adult-onset neurodegenerative disorder primarily affecting motor neurons. Mechanisms leading to motor neuron death in ALS are poorly understood mostly because of disease heterogeneity and lack of access to affected cells. The induced pluripotent stem cell (iPSc) technology provides the opportunity to obtain and study human motor neurons and is therefore a promising tool for ALS modeling.IPSc clones from control subjects were generated, and we compared several protocols in order to set up an efficient protocol for iPSc differentiation into motor neurons. The obtained cultures were heterogenous, comprising different neuron subtypes and neural precursors. To allow investigation of intrinsic disease mechanisms in ALS motor neurons, we developed a new technique to purify motor neurons by FACS sorting. By combining the use of a lentiviral vector expressing a fluorescent protein under control of a motoneuron-specific promoter and of a monoclonal antibody directed against the p75 neurotrophin receptor, isolation of exquisitely pure motor neurons was achieved. In parallel, iPSc technology was used to establish cellular models of ALS. IPSc were generated from one patient with familial ALS carrying a mutation in the TARDBP gene (encoding a DNA-binding protein, TDP-43) and one patient with sporadic ALS. To validate our models, we investigated characteristic disease phenotypes during iPSc differentiation, including i) cytoplasmic aggregate formation, ii) motor neuron generation and survival defects, iii) neurite growth alterations

Reprogramming of somatic cells into induced pluripotent stem cells (iPSCs) by overexpression of the transcription factors OCT4, SOX2, KLF4, and c-Myc holds great promise for the development of personalized cell replacement therapies. In an attempt to minimize the risk of chromosomal disruption and to simplify reprogramming, several studies demonstrated that a reduced set of reprogramming factors is sufficient to generate iPSC. We recently showed that a reduction of reprogramming factors in murine cells not only reduces reprogramming efficiency but also may worsen subsequent differentiation. To prove whether this is also true for human cells, we compared the efficiency of neuronal differentiation of iPSC generated from fetal human neural stem cells with either one (OCT4; hiPSC1F-NSC) or two (OCT4, KLF4; hiPSC2F-NSC) reprogramming factors with iPSC produced from human fibroblasts using three (hiPSC3F-FIB) or four reprogramming factors (hiPSC4F-FIB). After four weeks of coculture with PA6 stromal cells, neuronal differentiation of hiPSC1F-NSC and hiPSC2F-NSC was as efficient as iPSC3F-FIB or iPSC4F-FIB. We conclude that a reduction of reprogramming factors in human cells does reduce reprogramming efficiency but does not alter subsequent differentiation into neural lineages. This is of importance for the development of future application of iPSC in cell replacement therapies.

Parkinson's disease (PD) primarily manifests as loss of motor control through the degeneration of nigrostriatal dopaminergic neurons. The microtubule-associated protein tau (MAPT) locus is highly genetically associated with PD, wherein the H1 haplotype confers disease risk and the H2 haplotype is protective. As this haplotype variation does not alter the amino acid sequence, disease risk may be conferred by altered gene expression, either of total MAPT or of specific isoforms, of which there are six in adult human brain. To investigate haplotype-specific control of MAPT expression in the neurons that die in PD, induced pluripotent stem cells (iPSCs) from H1/H2 heterozygous individuals were differentiated into dopaminergic neuronal cultures that expressed all six mature isoforms of MAPT after six months' maturation. A reporter construct using the human tyrosine hydroxylase locus was also generated to identify human dopaminergic neurons in mixed cultures. Haplotype-specific differences in the inclusion of exon 3 and total MAPT were observed in iPSC-derived dopaminergic neuronal cultures and a novel variant in MAPT intron 10 increased the inclusion of exon 10 by two-fold. RNA interference tools were generated to knockdown total MAPT or specific isoforms, wherein knockdown of the 4-repeat isoform of tau protein increased the velocity of axonal transport in iPSC-derived neurons. MAPT knockdown also reduced p62 levels, suggesting an impact of tau on macroautophagy, likely through altered axonal transport. These results demonstrate how variation at a disease susceptibility locus can alter gene expression, thereby impacting on neuronal function.

no
To investigate the underlying nature of the effects of transcranial magnetic stimulation (TMS) on speed perception, we applied repetitive TMS (rTMS) to human V5/MT+ following adaptation to either fast- (20 deg/s) or slow (4 deg/s)-moving grating stimuli. The adapting stimuli induced changes in the perceived speed of a standard reference stimulus moving at 10 deg/s. In the absence of rTMS, adaptation to the slower stimulus led to an increase in perceived speed of the reference, whilst adaptation to the faster stimulus produced a reduction in perceived speed. These induced changes in speed perception can be modelled by a ratio-taking operation of the outputs of two temporally tuned mechanisms that decay exponentially over time. When rTMS was applied to V5/MT+ following adaptation, the perceived speed of the reference stimulus was reduced, irrespective of whether adaptation had been to the faster- or slower-moving stimulus. The fact that rTMS after adaptation always reduces perceived speed, independent of which temporal mechanism has undergone adaptation, suggests that rTMS does not selectively facilitate activity of adapted neurons but instead leads to suppression of neural function. The results highlight the fact that potentially different effects are generated by TMS on adapted neuronal populations depending upon whether or not they are responding to visual stimuli.
BBSRC

La Sclérose Latérale Amyotrophique (SLA) est une maladie neurodégénérative caractérisée par la mort des motoneurones (MNs). Malgré plusieurs hypothèses pouvant expliquer les mécanismes à l’origine de leur mort sélective, l’hétérogénéité de la SLA rend difficile la compréhension des causes exactes de la dégénérescence. Dans ce contexte, les cellules souches pluripotentes induites humaines (iPSC) permettent l’étude des formes familiales de la maladie comme des formes sporadiques. Contrairement à la majorité des travaux publiés à ce jour qui étudient des iPSC de patients porteurs de mutation dans un seul gène de SLA, mon projet a eu pour objectif de comparer plusieurs formes de SLA dans un même contexte expérimental. A partir d’iPSC de patients présentant différentes formes génétiques de SLA (C9ORF72, SOD1, TARDBP), nous avons obtenu des cultures pures de MNs humains. Alors que nous n’avons pas observé de mort des MNs mutants après plusieurs semaines, des études fonctionnelles d’électrophysiologie ont montré une altération tardive de l’excitabilité des MNs en fonction des patients. De façon plus précoce, nous avons observé la présence d’agrégats protéiques communs ou spécifiques aux différentes formes de SLA, avec certaines accumulations localisées au niveau du segment proximal de l’axone, une région importante pour la maintenance de l’identité axonale et le déclenchement des potentiels d’action. Des altérations physiques ou moléculaires ont été mises en évidence au niveau de ce segment dans les MNs mutants, suggérant qu’une perturbation du segment proximal de l’axone pourrait être un évènement très précoce altérant ainsi l’intégrité et la fonctionnalité des MNs de patients
Amyotrophic Lateral Sclerosis (ALS) is a neurodegenerative disorder characterized by motor neurons death (MNs). Despite several hypothesis trying to explain this selective loss, the exact reasons of MNs degeneration remain unidentified mainly due to the disease heterogeneity. In this respect, the use of human induced pluripotent stem cells (iPSC) are opening up opportunities to model not only familial but also sporadic forms of ALS. In comparison to previously published studies, which focus only on one type of ALS mutation, my thesis had the objective to compare in a same experimental context multiple forms of ALS in order to distinguish similarities and discrepancies inherited by the mutation. Using iPSC obtained from genetic forms of ALS patients (C9ORF72, SOD1, TARDBP) as well as control subjects, we generated pure cultures of human MNs. While ALS MNs were not sensitive to death after few weeks of culture, electrophysiological functional studies revealed a patient-dependent late alteration in MNs excitability. Early defects were also reported, with observations of generic and mutation-specific protein aggregates. Interestingly, some accumulations were localized at the axonal initial segment (AIS) region, which is important for maintaining axonal identity and crucial for action potentials’ initiation. Physical and/or molecular alterations were reported at the AIS in ALS MNs, suggesting that AIS perturbation could be an early event in MN degeneration by disruption of ALS patients’ MNs integrity and functionality

Alzheimer's disease is characterized by (1) senile plaques formed of aggregated amyloid peptides of 40 to 42 amino acids (A$ beta sb{1-40}$ and A$ beta sb{1-42}$) which are derived from the metabolism of the amyloid precursor protein, (2) by neurofibrillary tangles involving a dysfunction of the cytoskeleton due to the hyperphosphorylation of the tau protein, and (3) by amyloid-laden cerebral vessels. The goal of this project is to determine if there is a link between the presence of aggregated amyloid peptides and tau hyperphosphorylation.
We incubated the human fetal primary neuron cultures with A$ beta sb{1-40}$ and A$ beta sb{1-42}$ (100nM). The western blots show that serine 202 epitope of tau protein is phosphorylated in a cyclic manner. With the tau phosphorylation assay, we showed that A$ beta sb{1-40}$ and A$ beta sb{1-42}$ activate one or more protein kinase(s) able to phosphorylate tau protein and more specifically the serine 202 epitope.
We conclude that pathological concentrations of aggregated peptides A$ beta sb{1-40}$ and A$ beta sb{1-42}$ activate protein kinases which induce the phosphorylation of tau protein and serine 202 epitope.

The discovery of resetting human somatic cells via introduction of four transcription factors into an embryonic stem cell-like state that enables the generation of any cell type of the human body has revolutionized the field of medical science. The generation of patient-derived iPSCs and the subsequent differentiation into the cells of interest has been, nowadays, widely used as model system for various inherited diseases. The aim of this thesis was to generate iPSCs and to subsequently derive NPCs which can be differentiated into neurons in order to model the two most common forms of the NCLs: LINCL which is caused by mutations within the TPP1 gene, encoding a lysosomal enzyme, and JNCL which is caused by mutations within the CLN3 gene, affecting a lysosomal transmembrane protein. It was shown that patient-derived fibroblasts can be successfully reprogrammed into iPSCs by using retroviral vectors that introduced the four transcription factors POU5F1, SOX2, KLF4 and MYC. The generated iPSCs were subsequently differentiated into expandable NPCs and finally into mature neurons. Phenotype analysis during the different stages, namely pluripotent iPSCs, multipotent NPCs and finally differentiated neurons, revealed a genotype-specific progression of the disease. The earliest events were observed in organelle disruption such as mitochondria, Golgi and ER which preceded the accumulation of subunit c of the mitochondrial ATPase complex that was only apparent in neurons. However, none of these events led to neurodegeneration in vitro. The established disease models recapitulate phenotypes reported in other NCL disease models such as mouse, dog and sheep model systems. More importantly, the hallmark of the NCLs, accumulation of subunit c in neurons, could be reproduced during the course of disease modeling which demonstrates the suitability of the established system. Moreover, the derived expandable NPC populations can be used for further applications in drug screenings. Their robust phenotypes such as low levels of TPP1 activity in LINCL patient-derived NPCs or cytoplasmic vacuoles, containing storage material, observed in CLN3 mutant NPCs, should serve as possible phenotypic read-outs.

Type I interferons (IFNs) are known to cause neuropsychiatric side effects, including cognitive and mood disturbances, through mechanisms still not completely defined. To gain more information about type I IFN neurotoxicity, I investigated whether these cytokines could act directly on neuronal cells and regulate intracellular signaling pathways involved in cell death. In primary cultures of mouse cortical neurons acute exposure to IFN-β induces a marked tyrosine phosphorylation of signal transducer and activator of transcription (STAT) 1 and 3, whereas long-term exposure to the cytokine increased the number of neuronal cells displaying an increased labeling for cleaved caspase 3. In human SH-SY5Y neuroblastoma cells, used as a neuronal cell model, type I IFNs rapidly stimulated the tyrosine phosphorylation of Janus kinase (JAK) and STAT1, STAT3 and STAT5 and this response was antagonized by blockade of type I IFN receptor. Prolonged exposure to IFN-β induced apoptotic cell death accompanied by cytochrome C release, cleavage of caspase-9, -7, -3 and poly-(ADP ribose) polymerase (PARP), DNA fragmentation and p38 mitogenactivated protein kinase (p38 MAPK) phosphorylation. JAK inhibition reduced IFN-β- stimulated Tyk2 and STAT1 phosphorylation, STAT1 transcriptional activity, induction of double-stranded RNA-activated protein kinase (PKR), caspase cleavage and p38 MAPK phosphorylation. PKR induction was associated with enhanced PKR activity and chemical inhibition of PKR reduced IFN-stimulated caspase activation. Moreover, long-term IFN-β treatment led to down-regulation of phosphatidylinositol 3- kinase (PI3K)/protein kinase B (Akt) signaling and IFN-β-induced apoptosis was attenuated in cells expressing constitutively active Akt. The proapoptotic effect of IFNs was markedly attenuated by pharmacological blockade of either glycogen 8 synthase kinase-3 (GSK-3) or p38 MAPK. In retinoic acid-differentiated SH-SY5Y cells, treatment with IFN-β inhibited BDNF-induced regulation of Akt and GSK-3β, activation of extracellular signal-regulated protein kinase 1 and 2 (ERK1/2), phosphorylation of tyrosine hydroxylase, and neuritogenesis. Thus, type I IFNs can directly impair neuronal survival by regulating multiple mechanisms, including activation of the canonical JAK-STAT signaling and the ancillary p38 MAPK pathway and down-regulation of the PI3K/Akt and ERK1/2 signaling systems. The final outcome of these molecular events is the activation of the apoptotic program and impaired neuronal differentiation in response to neurotrophic stimuli. It is proposed that these actions directed on neuronal cells may contribute to the neuropsychiatric disturbances elicited by these cytokines.

Sanfilippo C syndrome is a lysosomal storage disorder that presents an autosomal recessive inheritance pattern and is caused by mutations in the HGSNAT gene, identified in 2006 in the chromosome 8. This gene codes for a lysosomal transmembrane protein, acetyl-CoA α-glucosaminide N-acetyltransferase, which acetylates the terminal glucosamine in the heparan sulfate chain during its degradation, a crucial step previous to the action of the next enzyme of the pathway. Heparan sulfate is a glycosaminoglycan localized in the extracellular matrix being part of proteoglycans and participate in several and important cellular processes. The HGSNAT protein dysfunction promotes the storage of partially degraded heparan sulfate chains inside the lysosomes, causing an alteration in many different cellular processes and affecting especially neurons. This fact promotes the progressive and severe neurodegeneration that appears during childhood as the main phenotypic feature in patients. This thesis represents an important study on the molecular basis of Sanfilippo C syndrome. Firstly, a mutational analysis has been performed, identifying the mutations causing the disease in 15 patients from different origins. A total of 13 different mutations have been found, seven of which were not previously described. The pathogenicity of four missense mutations identified has been proved by measuring the enzyme activity after in vitro expression of the proteins. Also, the pathogenicity of five mutations affecting different conserved splice sites has been demonstrated since they were shown to alter the splicing process. It has been established that two prevalent mutations in Spanish patients accounts for almost the 70% of the total and, using a haplotype analysis, a single origin for each of them has been suggested. Secondly, some therapeutic approaches have been tested, as a first step in the pursuit of an effective therapy that to date does not exist for this disease. The use of modified U1 snRNAs that present a higher complementarity to the mutated splice site sequences than the wild type U1 snRNA has been proved to partially restore the normal splicing process for one of the splicing mutations analyzed. In the case of missense mutations or mutations that result in the loss of some amino acids, this work suggests the possibility to use glucosamine as a chaperone to prevent the incorrect folding of the protein and to facilitate the trafficking process of the protein from the endoplasmic reticulum to the Golgi apparatus. Finally, the use of siRNAs to inhibit important genes in the heparan sulfate synthetic pathway, specifically the EXTL genes, has been suggested as a possible substrate reduction therapy, with the best results obtained on the inhibiton of EXTL2 expression. Finally, during this thesis, a neuronal model for Sanfilippo C syndrome has been obtained. This represents an important progress in the study of this disease since to date, neither cellular nor animal model exists. To achieve this goal, fibroblasts from two different Sanfilippo C patients have been reprogrammed to produce induced pluripotent cells that later have been differentiated to neurons. It has been demonstrated that these neurons present the typical phenotypic features of the disease such as the lack of enzyme activity, the heparan sulfate storage, the increased size and number of lysosomes, an alteration in the autophagy process and an increase in the number of apoptotic cells. Using specific experiments to study the neuronal activity in these cultures, a progressive decrease in the patients’ neurons activity and problems in the maintenance of the developed neuronal networks has been detected. This model will be a good platform to study profoundly the molecular, cellular and brain basis of the disease and to develop and test different therapeutic approaches for Sanfilippo C syndrome in the cellular type most affected in patients.
La síndrome de Sanfilippo és una malaltia monogènica hereditària que presenta una severa i progressiva neurodegeneració que s’inicia durant els primers anys de vida dels pacients. Està causada per mutacions en el gen HGSNAT, identificat l’any 2006 en el cromosoma 8, el qual codifica per l’enzim acetil-CoA α-glucosaminida N-acetiltransferasa, una proteïna de membrana lisosomal. La seva funció és acetilar les glucosamines terminals de les cadenes de heparà sulfat que estan sent degradades. Quan la proteïna està mutada, es promou l’acumulació de cadenes d’heparà sulfat parcialment degradades en els lisosomes, les quals augmenten en nombre i mida, provocant la seva disfunció. Per aquesta raó, la síndrome de Sanfilippo es classifica com a malaltia d’acumulació lisosòmica, en concret com a una mucopolisacaridosi, degut a la naturalesa del substrat acumulat. L’heparà sulfat és un dels glicosaminoglicans, anteriorment coneguts com a mucopolisacàrids, que es troba en la matriu extracel·lular formant part dels proteoglicans. Aquestes molècules participen en diferents i importants funcions cel·lulars com ara la migració i l’adhesió. La desregulació de la seva homeòstasi provoca una disfunció de múltiples processos cel·lulars. Aquesta tesi contribueix de manera important a l’estudi molecular de la malaltia. S’ha portat a terme un anàlisi mutacional i la conseqüent caracterització de les mutacions identificades per tal d’aprofundir en el coneixement de la malaltia. Per altra banda, s’han provat diferents possibles aproximacions terapèutiques com un primer pas en l’obtenció d’una teràpia exitosa per a aquesta devastadora malaltia neurodegenerativa per la qual encara no hi ha un tractament efectiu. Finalment, s’ha generat un model neuronal utilitzant cèl·lules mare pluripotents induïdes. Aquest model serà d’utilitat per estudiar i entendre els processos moleculars i cel·lulars que contribueixen al desenvolupament de la malaltia a nivell de la neurona i representarà una ajuda molt valuosa en la cerca de tractaments efectius.

Les cellules souches pluripotentes induites obtenues par reprogrammation des cellules somatiques primaires ont révolutionné les domaines de la biologie cellulaire et de la modélisation des maladies. Cependant, la modélisation des maladies musculaires squelettiques et neuromusculaires humaines a été entravée par un nombre limité de protocoles pour la génération de fibres musculaires matures avec une organisation sarcolemmale. Par co-différenciation simultanée de hiPSC dans les cellules musculaires et les motoneurones, nous avons développé une nouvelle procédure pour générer des fibres musculaires squelettiques matures multinucléées innervées. La présence des deux types de cellules améliore considérablement la différenciation des myoblastes et donne des fibres musculaires multinucléées fonctionnelles longues de plusieurs millimètres. De plus, cette culture de type organoïde peut être maintenue sur de longues périodes avec une régénération cellulaire autonome grâce à la présence de cellules positives pour PAX7 et à la synthèse de la matrice extracellulaire. Ce protocole applicable aux hiPSC d'individus en bonne santé a été validé dans la dystrophie musculaire de Duchenne, la dystrophie myotonique de type 1, la dystrophie Facio-Scapulo-Humerale et la dystrophie des ceintures de type 2A ouvrant de nouvelles voies pour l'exploration de la différenciation musculaire, la modélisation des maladies et la découverte de médicaments
Induced pluripotent stem cells obtained by reprogramming of primary somatic cells have revolutionized the cell biology and disease modeling fields. However, modeling human skeletal muscle and neuromuscular disorders has been hindered by a limited number of protocols for generation of mature muscle fibers with sarcolemmal organization. Through simultaneous co-differentiation of hiPSC into muscle cells and motor neurons, we developed a novel procedure for generating innervated multinucleated mature skeletal muscle fibers. Presence of both cell types greatly enhances myoblast differentiation and yields mature functional millimeter-long multinucleated muscle fibers. Furthermore, this organoid-like culture can be maintained over long periods of time with autonomous cell regeneration thanks to the presence of PAX7-positive cells and extracellular matrix synthesis. This protocol applicable to hiPSCs from healthy individuals was validated in Duchenne Muscular Dystrophy, Myotonic Dystrophy, Facio-Scapulo-Humeral Dystrophy and type 2A Limb-Girdle Muscular Dystrophy opening new paths for exploration of muscle differentiation, disease modeling and drug discovery

The study of the central nervous system represents a great challenge in the field of neuroscience. For years, various techniques have been developed to study neuronal cells in-vitro as it is difficult to conduct in-vivo experiments due to ethical problems deriving from its anatomical location. Consequently, both in-vivo and in-vitro animal models have been used extensively to gain new insights into basic functioning principles of neuronal tissue and therapeutic approaches for brain diseases. Over time, we have seen that there is a poor correlation between the clinical diagnosis and the underlying pathological mechanisms. In fact, some symptoms that may occur in the patient are not replicated in the animal, making many promising approaches in animal studies not translatable in the clinic. With the advent of human-induced pluripotent stem cells (h-iPSC) several protocols for the generation of human-neuronal cells are becoming available for all laboratories. The importance of this technique lies in the opportunity to develop a human model derived directly from the patient: the patient's in-vitro cells will exhibit the same genetic and epigenetic modifications as the in-vivo cells. This has raised hopes for the generation of engineered brain models that can be coupled to sensors / actuators in order to better investigate their functional properties in-vitro (i.e. brain-on-a-chip). A reliable method for evaluating the functionality of neuronal cultures is the study of the spontaneous electrophysiological activity using microelectrode arrays (MEA). There are numerous studies in the literature that used h-iPSC on MEAs, showing the characterization of neuronal patterns of patient-derived cultures, demonstrating how this platform is valid for disease phenotyping, drug discovery and translational medicine. Although these models helped to shed light on fundamental biological mechanisms, the majority is based on two-dimensional neuronal cultures, which lack some key features to mimic in-vivo behavior. Three-dimensional h-iPSC-derived models possess a microenvironment, tissue architecture and potential to model network activity with greater complexity than two-dimensional models. Depending on the purpose of the study, we can choose different approaches to recreate 3D in-vitro brain, from those that aim to reproduce the trajectories of neurodevelopment (i.e. brain-organoids) to the use of synthetic materials that reproduce the functionalities of the extracellular matrix (ECM) (i.e. scaffold-based) (Chiaradia and Lancaster, 2020, Tang et al., 2006). Although h-iPSC-derived brain models summarize many aspects of network function in the human brain, they are subject to variability and still do not perfectly mimic behavior in-vivo. Therefore, to reach the full potential of this model we need improvements in differentiation methods and bioengineering, making these models engineered and reproducible. The aim of this PhD thesis was to implement different 3D neuronal culture generation methodologies that can be integrated on MEA devices to offer robust engineered platforms for functional studies.

"Alzheimer’s disease (AD) is a neurodegenerative disorder that leads to dementia in adults. The mechanism of neurodegeneration is thought to involve the extracellular production of a highly toxic A-beta peptide that engages cell surface receptors to induce cellular oxidative stress and apoptosis, but the signal transduction pathways that lead to A-beta induced cell death are unknown. We previously showed that a human ependymin neurotrophic peptide mimetic (hEPN-1) can promote cell survival in an in vitro AD model system. This initial observation was extended in this thesis by investigating the mechanism of A-beta induced apoptosis and hEPN-1 induced survival. Immunoblots were used to assay the total cellular levels of specific caspase proteins. The results show that A-beta induced apoptosis uses an extrinsic caspase pathway involving caspases-2 and -3, and that hEPN-1 treatment can reduce those caspase levels. A caspase activity assay showed that A-beta increased caspase-3/7 activity, while hEPN-1 treatment lowered it. Moreover, in vivo studies with AD transgenic mice showed that hEPN-1 treatment increased antioxidative superoxide dismutase levels in brain. Thus, hEPN-1 holds potential as a therapeutic to treat the underlying neurodegenerative cause of AD, not merely its symptoms as with other currently approved AD drugs."

In the case of Parkinson’s disease, a common neurodegenerative disorder, the loss of motor function results from the selective degeneration of dopaminergic neurons (DNs) in the brain. Current treatments focus on pharmacological approaches that lose effectiveness over time and produce unwanted side effects. A more complete concept of rehabilitation to improve on current treatments requires the production of DNs to replace those that have been lost. Although pluripotent stem cells (PSCs) are a promising candidate for the source of these replacement neurons, current protocols for the terminal differentiation of DNs require a complicated cocktail of factors. Recently, a naturally occurring steroid called guggulsterone has been shown to be an effective terminal differentiator of DNs and can simplify the method for the production of such neurons. I therefore investigated the potential of long-term guggulsterone release from drug delivery particles in order to provide a proof of concept for producing DNs in a more economical and effective way. Throughout my study I was able to successfully encapsulate guggulsterone in Poly-ε-caprolactone (PCL)-based microspheres and I showed that the drug was capable of being released over 44 days in vitro. These guggulsterone-releasing microspheres were also successfully incorporated in human induced pluripotent stem cell (hiPSC)-derived neural aggregates (NAs), providing the foundation to continue investigating their effectiveness in producing functional and mature DNs. Together, these data suggest that guggulsterone delivery from microspheres may be a promising approach for improving the production of implantable DNs from hiPSCs.
Graduate

A doença de Gaucher (Gaucher disease – GD) é a doença lisossomal com maior taxa incidência em todo o mundo (Mistry et al., 2017). As doenças de armazenamento lisossomal, são um grupo de doenças metabólicas hereditárias caracterizadas pela redução de atividade ao longo de uma via metabólica, levando à acumulação de um ou mais dos seus substratos, e consequente desregulação celular. GD é causada por níveis reduzidos de atividade da hidrolase β-glucocerebrosidase (GCase), o que pode ocorrer por mutações no centro ativo, alterações no processamento, no ativador ou nos transportadores. A doença transmite-se quase exclusivamente de por transmissão autossómica recessiva de mutações no gene da GCase, GBA1. Ainda que raro, alguns casos de GD causados por mutações no gene PSAP, que codifica para o cofator da GCase, Saposin C (Kang et al., 2018). Até à data, mais de 300 mutações GBA1 foram descritas como causadoras de GD e, apesar da grande maioria poupar o centro ativo da enzima, tendem a alterar drasticamente a estrutura tridimensional, estabilidade e transporte da GCase (Dvir et al., 2003; Manoj Kumar Pandey, Rani, Zhang, Setchell, & Grabowski, 2012). Produção de GCase com uma estrutura tridimensional errada ativa o sistema de controlo de qualidade do reticulo endoplasmático que, através do uso de chaperones do sistema UPR (Unfolded Protein Response), tenta promover uma conformação correta. Após vários ciclos falhados de remodelação molecular, o UPR despoleta apoptose através da via mitocondrial (Bendikov-Bar & Horowitz, 2012; Grabowski, Zimran, & Ida, 2015).Para além da pressão exercida sob o Retículo Endoplasmático (ER), a baixa atividade da GCase leva a uma acumulação intralisossomal de substratos glucosilceramida (GlcCer) e glucosilsesfingosina (Awad et al., 2015; Choi et al., 2011). Apesar desta ser expressa ubiquamente, devido à sua natureza e função fisiológica com altos níveis de remodelação e atividade lisossomal, as células do sistema reticuloendotelial e neuronal são primordialmente afetadas. Nos macrófagos, GlcCer acumula-se em vesículas que ocupam grande parte do citoplasma, formando as características células de Gaucher, identificáveis por histologia, que mantêm um estado de inflamação sistémica constante (Hollak, van Weely, van Oers, & Aerts, 1994; Klein & Futerman, 2013). Neurónios, por outro lado, são especialmente sensíveis a baixos níveis de GCase devido à alta taxa de processamento de gangliósidos por processos associados à sinapse, entrando em apoptose e gerando um ciclo de inflamação e morte neuronal (Grabowski, 2008; Mencarelli & Martinez–Martinez, 2013). O modo como cada mutação ou genótipo afeta estes dois tipos celulares explica a panóplia de sintomas que são geralmente associados a GD e são comumente divididos em sintomas viscerais ou sistémicos que afetam fígado, baço, pulmões e ossos, e sintomas neurológicos, com afeção de áreas específicas do cérebro e sistema nervoso central. Apesar de algumas mutações apresentarem um fenótipo predominantemente sistémico ou neurológico, a vasta maioria das mutações, não está diretamente associada a um padrão sintomático ou severidade. Atualmente, num contexto académico reconhece-se que a dispersão dos sintomas e sua severidade formam um gradiente, das formas neurológicas congénitas mais agudas até formas mais leves de envolvimento somático exclusivamente detetadas numa fase mais tardia de vida ou que não são detetadas de todo (Aureli et al., 2012). No contexto clínico, a sintomatologia é classificada em três tipos de acordo com as necessidades terapêuticas e prognóstico: tipo 1 (não neuropático); tipo 2 (neuropático agudo); e tipo 3 (neuropático crónico). As manifestações sistémicas mais comuns incluem inflamação generalizada, hepatomegalia, esplenomegalia, pancitopenia e osteoporose, e tendem a estar presentes em todos os tipos de GD ao passo que as manifestações neurológicas incluem espasticidade, nistagmo, défice cognitivo e neurodegeneração e manifestamse nos tipos 2 e 3 (Butler, 2001; McNeill et al., 2012).GD tipo 1 é a forma mais frequente da doença, afetando mais de 90% dos pacientes, ao passo que o tipo 2 é a forma mais severa, apresentando um prognóstico muito reservado com uma degeneração neurológica rápida e acentuada, resultando numa esperança média de vida de 9 meses. O tipo 3 encontra-se entre o tipo 1 e 2 em termos de severidade, apresentando tanto sintomas viscerais como neurológicos crónicos (Ozlem Goker-Alpan et al., 2003). Ainda que não haja cura para GD, existem presentemente várias alternativas de tratamento e gestão de sintomas, dependendo da sintomatologia e apresentação fenotípica da doença. O primeiro tratamento desenvolvido foi o transplante de medula óssea (bone marrow transplantation - BMT) através da transfusão de progenitores hematopoiéticos de dadores saudáveis. Se feito correta e precocemente, BMT pode prevenir tanto manifestações viscerais como neurológicas da doença, no entanto este efeito terapêutico é perdido com a idade (Platt & Jeyakumar, 2008). Para além disso BMT não tem efeito nas formas neuropatológicas mais agudas de GD tipo 2 pela rápida progressão e deterioração (Platt & Jeyakumar, 2008; Shawky & Elsayed, 2016). Durante vários anos, a única alternativa a BMT foi, a suplementação endovenosa de GCase, num sistema terapêutico de substituição enzimática (enzyme replacement therapies – ERT). Terapias ERT são ainda a primeira linha de tratamento de GD, com melhorias como a marcação de GCase com resíduos de manose para facilitar a sua captação por macrófagos (Grabowski, 2008; McEachern et al., 2007). No entanto, apesar de aliviar a grande maioria dos sintomas sistémicos, uma vez que GCase não atravessa a barreira hematoencefálica, ERT não tem qualquer efeito nas manifestações neuronais da doença (R Kornfeld & Kornfeld, 1985). A classe de fármacos mais recente na prática clínica, aposta numa diminuição da acumulação de GlcCer através da inibição a montante da via catabólica, numa abordagem denominada terapia de redução de substrato (substrate reduction therapy - SRT). Esta abordagem tem foco na inibição da glucosilceramida sintetase, reduzindo a concentração intracelular dos seus produtos, glucose e ceramida (Rao Vunnam & Radin, 1980). Apesar de apresentarem um potencial terapêutico mais baixo do que ERT para sintomas viscerais, as terapias por SRT, ao serem baseadas em pequenas moléculas glicomiméticas são capazes de atravessar a barreira hematoencefálica e aliviar os sintomas neuronais. Contudo, apesar do baixo custo, os efeitos secundários e o baixo efeito terapêutico relegam SRT para uma segunda linha de tratamento. Inspirados na terapia SRT, uma nova classe de fármacos tem vindo a ganhar a atenção da comunidade científica: chaperones farmacológicas (pharmacological chaperones – PC). Estes compostos atuam como inibidores reversíveis da GCase e, ligando-se ao centro ativo da enzima melhoram a sua conformação, ajudando-a a iludir o Sistema de Controlo de Qualidade do Retículo Endoplasmático, aumentando a quantidade de Gcase e permitindo que exerça a sua função com a capacidade hidrolítica remanescente. Recentemente, investigação e desenvolvimento de novas abordagens para compensar ou suplementar défices de atividade de GCase no sistema nervoso central têm ganho novo fôlego devido a uma série de associações entre mutações em GBA1 e desenvolvimento de patologias neurológicas degenerativas. Apesar de GD tipo 1 ser, por definição, ausente de qualquer manifestação neurológica ao longo dos anos, uma série de casos esporádicos de manifestações semelhantes à sintomatologia clássica da doença de Parkinson (PD) começaram a ser notadas. Apesar de inicialmente episódica, a incidência mais elevada de bradicinesia, demência, défice cognitivo, hiposmia e depressão levaram a que fossem feitos os primeiros estudos genéticos de associação ente PD e GD (Sidransky & Lopez, 2012; Yang, Lee, Lee, Kim, & Lee, 2013). Estudos recentes apontam para uma sobre representação de pacientes PD com mutações GBA1, 5 a 7 vezes superior quando comparado a um grupo controlo, com uma idade média no momento do diagnóstico 4 a 6 anos inferior (Aharon-Peretz , Rosenbaum , & Gershoni-Baruch 2004; Sidransky & Hart, 2012; Toft, Pielsticker, Ross, Aasly, & Farrer, 2006). No entanto, apesar da correlação estatística, o mecanismo subjacente à relação entre GCase e α-sinucleína ainda é pouco claro. O primeiro modelo mecanístico foi desenvolvido em Mazzulli et al em 2011, no qual é proposta a existência de uma relação inversamente proporcional entre a atividade de GCase e os níveis de α-sinucleína, na qual reduzida atividade de GCase resulta em elevados níveis de α-sinucleína, que, por sua vez, diminuem o transporte de GCase para o lisossoma, reduzindo os níveis de hidrólise e aumentando o numero de fibrilas de α-sinucleína (Mazzulli et al., 2011). GCase tornou-se, deste modo, não apenas o alvo terapêutico para GD, mas também para PD (Aflaki et al., 2016). O presente trabalho teve como objetivos: I) Estabelecer um modelo neuronal in vitro a partir de iPSc para os genótipos de GD tipo 2 L444P/L444P (clone A), L444P/P415R (clone B), G325R/C342G (clone C), L444P/G202R (clone D), e controlo (wild type - WT), e corroborar a relação entre a atividade de GCase e os níveis de α-sinucleína; II) testar um grupo de 12 chaperones farmacológicas em neurónios de pacientes GD2 e caracterizá-los quanto ao seu efeito na Gcase, quer na quantidade de proteína, quer na sua atividade; III) testar o efeito de retroalimentação positiva proposto por Mazzulli et al., através da medição dos níveis de α-sinucleína e da relação destes com a atividade de GCase e quantidade de proteína (Mazzulli et al., 2011). Para tal, fibroblastos de 3 genótipos GD tipo 2 (Clones A, B e C) foram reprogramados ao estado de pluripotência através de transdução e expressão dos fatores de reprogramação Oct4, Sox2, Klf4 e CMyc (OSKM) (Takahashi et al., 2007). Um controlo sem mutação GBA1 (WT) e um outro genótipo GD tipo 2 (clone D) previamente reprogramados e testados foram também usados durante este trabalho. O conjunto dos clones iPSc WT, A, B, C e, D foram então diferenciados em neurónios através de um protocolo de diferenciação em camada única adaptado e otimizado por nós. Culturas confluentes de neurónios foram tratadas com um conjunto de chaperones pertencentes a iminoaçucares, piperidinas monocíclicas (MTD131, TMB69, TMB65 e TMB84); piperidinas bicíclicas (MTD106, MG174, MTD132 e RV21); e nortropanos (MG235, CVI62, DW43 e DW45), com os seguintes resultados: I) No modelo estabelecido, os níveis de GCase estão de acordo com o que foi descrito para neurónios GD2, com níveis de atividade hidrolítica inferiores 20% aos observados para o controlo. Simultaneamente, com níveis reduzidos de GCase, estes clones apresentavam níveis aumentados de α-sinucleína. II) Ainda que a maioria dos compostos tenha afetado os níveis de proteína GCase, o efeito foi pouco consistente nos diferentes genótipos, apresentando variações consideráveis consoante as mutações GBA1 pressentes. Não obstante, alguns compostos demonstraram um efeito potencialmente terapêutico num ou mais genótipos: Clone A (L444P/L444P) apresentou um aumento dos níveis proteicos de GCase e redução dos níveis de α-sinucleína na presença dos chaperones TMB69, TMB65, MTD132, RV21, MG235 e CVI62; Clone B (L444P/P415R), aumentou os níveis proteicos de GCase por um fator de 3 quando tratado com TBM69, com concomitante redução de α-sinucleína em 70%. Neste mesmo genótipo, TMB65, reduziu α-sinucleína em 90% apesar de não ter nenhum efeito observável na GCase; para o Clone C (G325R/C342G), o chaperone MTD106 elevou a atividade de GCase e reduziu em 50% os níveis de α-sinucleína ; relativamente ao Clone D (L444P/G202R), apesar de nenhum dos compostos reduzir efetivamente os níveis de α-sinucleína, um aumento da atividade de GCase foi alcançado quando tratado com MTD106. III) Apesar do efeito positivo de alguns chaperones seja na quantidade ou atividade de GCase, como na redução dos níveis de α-sinucleína, esse efeito não é consistente. Esta observação pode ser causada por vários fenómenos como o efeito inibitório do composto, a afinidade e resistência de alguns dos compostos usados aos métodos de extração e desnaturação usados. Novos estudos serão necessários de modo a clarificar a interação entre os diferentes alelos de GD e chaperones de modo a garantir não só um aumento da atividade de GCase, mas também a segurança quanto aos níveis de α- sinucleína.
Gaucher Disease (GD) is the most common lysosomal storage disease. It is caused by mutations in GBA1 which lead to dysfunctional β-glucocerebrosidase (GCase) activity and an accumulation of its substrates, glucosylceramide and glucosylsphingosine. Reticuloendothelial system cells and neurons are especially affected by high glucosylceramide levels, generating multiple symptoms. According to the distribution and severity of the symptoms, and degree of neuronal involvement, GD can be classified as systemic (GD1), severe acute neuropathic (GD2) or chronic neuropathic (GD3). Presently, none of the available therapies are effective at alleviating neuronopathic forms of the disease. Recent studies have pointed to GBA1 mutations as the most frequent genetic risk factor for Parkinson Disease. In 2011, Mazzulli et al. proposed a feedforward mechanistic loop model explaining the inverse correlation between GCase activity and α-syn levels. In the present work, we developed an optimized differentiation protocol to test iPSc derived neurons from four GD2 genotypes (L444P/L444P, L444P/P415R, G325R/C342G and L444P/G202R). Each clone was treated with a set of 12 chaperone compounds regarding their effect on GCase protein levels, and activity. Simultaneously, α-syn levels were measured for each sample to test Mazzulli’s hypothesis. Our results identified some compounds which effectively enhanced GCase and decreased α-syn, which can be considered and explored as novel therapies for GD and PD. Our results indicate that chaperone treatment does not affect GCase levels/activity and the relation with α-syn levels in a way consistent with Mazzulli’s proposed model. This might be due to the multiple variable at play in our experimental system and suggests the need for follow-up studies.
This work was supported by National Portuguese funding through FCT – Fundação para a Ciência e a Tecnologia, scholarship PD/BD/52423/2013 and Genzyme Young Investigator Award 2012, with the title: Development of an induced pluripotent stem cell model of neuronopathic Gaucher’s Disease for investigating mechanisms of pathogenesis and small molecule testing.

博士
中國醫藥大學
藥物化學研究所博士班
97
Oxidative stress is widely implicated in the neuron cell death that is associated with various neurodegenerative disorders such as Parkinson’s disease and Alzheimer’s disease. Ugonin K, a flavonoid isolated from the rhizomes of Helminthostachys zeylanica (L) Hook, possesses potent antioxidant property. In this study, we investigate the neuroprotective effects of ugonin K on hydrogen peroxide-induced apoptosis in SH-SY5Y cells. Incubation of SH-SY5Y cells with H2O2 for 24 h induced cell death measured with MTT assay. Hoechst 33258 staining confirmed that the reduced cell viability by H2O2 was due to apoptosis. In addition, H2O2 increased the expression of 17-kDa cleaved fragment of caspase-3 which could be reversed by pretreatment with ugonin K. Pretreatment with ugonin K attenuated H2O2-induced cell death in a dose-dependent manner. Neuroprotective effect of ugonin K was abolished by ERK and PI3K inhibitors. Pretreatment with JNK kinase and p38 MAPK inhibitors had no effect on ugonin K-mediated protection against H2O2-induced apoptosis. Western blotting showed that ugonin K increased both ERK1/2 and AKT phosphorylation. These results suggest that ugonin K by activation of ERK1/2 and PI3K/AKT signal pathways protects SH-SY5Y cells from H2O2-induced apoptosis. Therefore, the molecular mechanisms of neuroprotective effects of ugonin K may include not only their antioxidant activities but also their interaction with cell signaling cascades leading to cell survival and cell proliferation. Ceramide accumulates in neurons during various disorders associated with acute or chronic neurodegeneation. In the present study, we investigate the neuroprotective effects of furopyrazole derivatives on C2 ceramide-induced cell death in primary cortical neurons. Among the 12 furopyrazoles tested, carbinol derivatives (CLC107, CLC-507, CLC-604 and CLC-609) exhibited strong neuroprotective against C2 ceramide-induced cell death. CLC107 and CLC-507 at concentration of 10 μM was also able to inhibit cell death but by far less extent than those of CLC-604 and CLC-609. Hoechst 33258 staining confirmed that the reduced cell viability by C2 ceramid was due to apoptosis. Pretreatment with CLC-609 attenuated C2 ceramide-induced apoptosis in a dose-dependent manner. Neuroprotective effect of CLC-609 was abolished by ERK inhibitors. These results suggest that CLC-609 by activation of ERK1/2 signal pathways protects primary cortical neurons from C2 ceramide-induced apoptosis. In conclusion, our data demonstrate that ugonin K and furopyrazole derivatives might be an important clue for drug design as neuroprotectants and potential therapeutic agents for against neurodegenerative disease.

博士
國立中興大學
生命科學系所
102
In regular culture conditions with leukemia inhibitory factor (LIF), the majority of mouse embryonic stem cells (mESCs) are maintained in a self-renewal stage; very few mESCs have differentiated morphology. When LIF is withdrawn, mESCs tend to differentiate; this differentiation process can be enhanced by the introduction of exogenous FGF. Here, we show that even in the presence of exogenous FGF1, mESCs can maintain self-renewal and expression of pluripotency markers in the presence of LIF. To elucidate the mechanism in which LIF dominates over FGF1, extracellular signal-regulated kinase 1/2 (Erk1/2) signaling of mESCs cultured in medium containing FGF1 or LIF/FGF1 was examined. The results demonstrate that Erk1/2 was activated by FGF1 in the absence of LIF; however, the FGF1-induced Erk1/2 phosphorylation was suppressed when LIF was introduced. Moreover, FGF1-Erk1/2 down-regulation was inhibited by signal transducer and activator of transcription 3 (Stat3) inhibitor WP1066, suggesting that LIF-induced Stat3 activation plays an important role in FGF1-Erk1/2 inhibition in mESCs. We further demonstrate that the binding affinity of phospho-Erk1/2 and Sprouty2 was increased via Stat3 activation. Binding of phospho-Erk1/2 and Sprouty2 blocks the activation of Erk1/2 signaling, and thus inhibits the downstream differentiation process in mESCs. Our findings demonstrate, for the first time, that LIF-induced Stat3 phosphorylation plays an important role in promoting the binding of phospho-Erk1/2 and Sprouty2, and thus inhibiting FGF1-induced differentiation. The feasibility of reprogramming human cells to induced neurons (iNs) was also explored in this study. Due to their high efficiency of generating functional neurons, the iNs show great potential in cell therapeutic medicine development. Most of nowadays studies use fibroblasts as the cell source for reprogramming. It would be beneficial to broaden the cell source for iNs. Taking the advantage of broadly banked blood cells, it is valuable to develop a blood cell neural reprogramming strategy for disease modeling and personalized iNs. Here, we reported that EBV-negative Burkitt''s lymphoma B cells (BJAB cells) cannot be reprogrammed into iNs through the transfection of miR124, BRN2 and MYT1L (IBM), even with the introduction of SH2B1 (S-IBM), which has been shown to improve the neurite activity/complexity and to accelerate the reprogramming progress during the conversion of fibroblasts to iNs. Our results showed that these BJAB cells, with the IBM or S-IBM neuronal reprogramming factors, revealed no morphological changes during 28 days post-transfection. These cells still maintain B cell characteristics including round-shaped and aggregated morphology. Although this reprogramming strategy is not working in BJAB cells, methods of producing iNs were reviewed and several alternative ways that might help convert blood cells to functional neurons were discussed. The development of combinations suitable for conversion of human cells to iNs could be a great help in the cell therapy field.

碩士
國立中興大學
生命科學系所
102
The accumulation of β-amyloid (Aβ), produced by endoproteolysis of the amyloid precursor protein (APP), results in amyloid plaques formation and is the cytopathological hallmark in the patient’s brain of Alzheimer’s disease (AD). To generate the AD cell model by in vitro differentiation of pluripotent stem cells, initially, we attempt to use APP-transgenic human embryonic stem cells (hESCs) as a platform for studying amyloid plaques associated diseases. Dual Oct4 and α-Tubulin promoters driving the neomycin-2A-green fluorescent protein (NeoGFP) and APP protein, respectively, are cloned into the PiggyBac vector for the establishment of APP-mutants overexpressing hESCs. Expression of APP mutants under α-Tubulin promoter was confirmed in HEK293T cells. Although the efficacy of both promoters was valid in individual vectors, the Oct4 promoter conjugated with α-Tubulin promoter in single PiggyBac vector were dramatically repressed for the expression of NeoGFP. This low expression of NeoGFP hampers the rapid selection and establishment of APP-transgenic ESCs. Alternatively, induced pluripotent stem cells (iPSCs) from Down syndrome patients (T21 iPSCs) have been shown to produce robust neural cells and to faithfully recapitulate the Aβ accumulation in vitro. Using our novel BiSF neural differentiation method, we successfully differentiated the T21 iPSCs into Nestin+ and Sox1+ neuroepithelial precursor cells on day 15 (D15) and mature TuJ1+ neurons on D21. Importantly, significant BTA-1+ Aβ plaques and Aβ42 accumulation were observed in the T21 iPSC-derived neurons on D30, but not in the differentiated normal-karyotype iPSCs. These results emphasize that differentiating neurons from T21 iPSCs can recapitulate the cytopathological features of AD and the BiSF method can rapidly steer the neural differentiation of T21 iPSCs and produce the cell model of AD efficiently.

碩士
國立成功大學
細胞生物及解剖學研究所
91
Neuronal differentiation in the mammalian CNS is driven by multiple events. Mitogen activated protein kinases (MAPKs) have been observed to play roles in neuronal development. It have been demonstrated that hNT-2 cells (NT2 cells), a cell-line derived from human teratocarcinoma, act as be a good model for study neuronal development. Following retinoic acid (RA) treatment, NT-2 cells can differentiate into postmitotic neuronal cells and express several mature neuronal markers. In this thesis, during period of RA induction, phosphorylated form of extracellular signal-regulated protein kinases1/2 (ERK1/2; members of MAPKs) raised and cell aggregation was also observed at the 7th day after RA induction. In addition, co-treatment with RA and MEK1/2 inhibitor, U0126, phospho-ERK1/2 were down-regulated and cell aggregation was also abrogated, while no detectable change in cellular morphology and the expression of neurofilament middle-chain (NF-M) were observed. It implies that ERK1/2 may take participated in the process of RA-induced neuronal differentiation and in charge for cellular aggregation during neuronal development. Furthermore, the expression of N-cadherin, a cell adhesion molecule, is consistent with the alteration of phosphor-ERK1/2 level. These data indicate that ERK1/2 regulate the N-cadherin molecule expression to modulate cell aggregation. The results may provide a thought for further explore the physiological functions in the process of neuronal differentiation.

ABSTRACT Introduction: Mounting evidence supports the beneficial effects of the Mediterranean diet (MD) and the Asian diet in delaying ageing and in preventing age-related dysfunctions, cancer, diabetes and neurodegenerative diseases. The beneficial effects of the MD and Asian diets in reducing age-related dysfunctions, including Alzheimer’s disease (AD), could be the consequence of the presence in specific foods of substantial amounts of specific polyphenols whose beneficial properties include the ability to interfere with amyloid aggregation. Our previous data have highlighted the beneficial effects of oleuropein aglycone (OLE) against protein/peptide aggregation in vitro and in TgCRND8 (Tg) mice, a model of Aβ deposition. Aim: To check in Tg mice 1) the effects of OLE or a mix of polyphenol extracts or hydrohytyrosol (HT) on cognitive functions and neuropathology and 2) the effect of andrographolide administration on cognitive functions and its ability to cross the blood brain barrier (BBB) conjugating it with fluorescent albumin nanoparticles. In addition, in rats we investigated 1) the ability of intravenous (i.v.) administered human fetal cholinergic neurons (hfCNs), isolated from the nucleus basalis of Meynert (NBM) of 12-week old fetuses, to cross the BBB and to improve memory functions of quisqualic acid NBM-lesioned rats and 2) the ability of intravenously administered solid lipid nanoparticles (SLNs) to cross the BBB and to penetrate the central nervous system of adult rats. Methods: In vivo experiments were carried out using: 1) Tg and wild type (wt) mice treated for 8 weeks with a modified low-fat (5.0%) AIN-76A diet (10 g/day/ mouse) as such, supplemented with OLE (50mg/kg of diet) or with a mix of polyphenol extracts (50 mg/kg of diet) found in olive mill waste water or with HT (50 mg/kg of diet), 2) Tg and wt mice injected intraperitoneally (i.p.) for 4 weeks (3 i.p. for week) with andrographolide (ANDRO) (4 mg/kg) conjugated with fluorescent albumin nanoparticles (NPs-ANDRO), 3) male Wistar rats divided into the following groups: Group I: injected into the NBM with 0.5 μL of quisqualic acid and subsequent intravenous (i.v.) administration by the tail vein of 300 μL of 1.5ML hfCNs. Group II: injected into the NBM with 0.5 μL of quisqualic acid. Group III: injected into the NBM with 0.5 μL saline and subsequent i.v. administration by the tail vein of 300 μL of 1.5M hfCNs. Group IV: injected into the NBM with 0.5 μL saline; Group V: un-injected rats and 4) male Wistar rats divided into the following groups Group I: injected i.v. with 200 L of nanoparticles (SLN 28.57 mg/ml plus FITC 1.43 g/ml) and sacrificed after 3h; Group II: injected i.v. with 200 L of nanoparticles and sacrificed after 24h; Group III: injected i.v. with 200 L of nanoparticles and sacrificed after 72h; Group IV: injected i.v. with 200 L of 0.9% of saline. For in vitro experiments were used neuroblastoma cell line (N2a) maintained in a 5.0% CO2 humidified atmosphere at 37°C. Results and Conclusion: Our results show that OLE supplementation induces epigenetic modifications as previously reported, favoring the expression of SIRT1 and reducing that of PARP1 confirming the existence of a functional link between PARP1 and SIRT1. Furthermore, diet supplementation with a mix of polyphenols or HT strongly improved mouse cognitive performances and reduced ß-amyloid deposits. An intense activation of autophagy was found in the cortex of all treated groups. HT administration also affected the inflammatory response in the hippocampal areas, as shown by the reduced astrocytes activation and TNF-α mRNA levels. These results show that diet supplementation with a mix of polyphenols or HT at the same dose as that of pure OLE previously administered ameliorate cognitive functions and indicate that either a mix of polyphenols or HT treatment is useful to treat neurodegenerative diseases. In addition, we found that i.v. administration of hfCNs migrated to the injected NBM and along the needle tract. In the quisqualic acid NBM-injected rats the cognitive impairments were significantly improved by hfCNs. Finally, we found that both SLNs and NPs were able to cross the BBB and that NPs-ANDRO were able to cross the BBB and to improve cognitive functions in Tg mice, supporting that the developed nanocarriers represent new potential brain delivery system to increase the efficacy of andrographolide treatment in neurodegenerative diseases.

Human induced pluripotent stem cells (hiPSCs) have two main properties: pluripotency and self-renewal. Physical cues presented by biomaterial scaffolds can stimulate differentiation of hiPSCs to neurons. In this work, we develop and analyze a mathematical model of aggregate growth and neural differentiation on melt electrospun biomaterial scaffolds. An ordinary differential equation model of population size of each cell state (stem, progenitor, differentiated) was developed based on experimental results and previous literature. Analysis and numerical simulations of the model successfully capture many of the dynamics observed experimentally. Analysis of the model gives optimal parameter sets, that correspond to experimental procedures, to maximize particular populations. The model indicates that a physiologic oxygen level (~5%) increases population sizes compared to atmospheric oxygen levels (~21%). Model analysis also indicates that the optimal scaffold porosity for maximizing aggregate size is approximately 63%. This model allows for the use of mathematical analysis and numerical simulations to determine the key factors controlling cell behavior when seeded on melt electrospun scaffolds.
Graduate

The discovery of resetting human somatic cells via introduction of four transcription factors into an embryonic stem cell-like state that enables the generation of any cell type of the human body has revolutionized the field of medical science. The generation of patient-derived iPSCs and the subsequent differentiation into the cells of interest has been, nowadays, widely used as model system for various inherited diseases. The aim of this thesis was to generate iPSCs and to subsequently derive NPCs which can be differentiated into neurons in order to model the two most common forms of the NCLs: LINCL which is caused by mutations within the TPP1 gene, encoding a lysosomal enzyme, and JNCL which is caused by mutations within the CLN3 gene, affecting a lysosomal transmembrane protein. It was shown that patient-derived fibroblasts can be successfully reprogrammed into iPSCs by using retroviral vectors that introduced the four transcription factors POU5F1, SOX2, KLF4 and MYC. The generated iPSCs were subsequently differentiated into expandable NPCs and finally into mature neurons. Phenotype analysis during the different stages, namely pluripotent iPSCs, multipotent NPCs and finally differentiated neurons, revealed a genotype-specific progression of the disease. The earliest events were observed in organelle disruption such as mitochondria, Golgi and ER which preceded the accumulation of subunit c of the mitochondrial ATPase complex that was only apparent in neurons. However, none of these events led to neurodegeneration in vitro. The established disease models recapitulate phenotypes reported in other NCL disease models such as mouse, dog and sheep model systems. More importantly, the hallmark of the NCLs, accumulation of subunit c in neurons, could be reproduced during the course of disease modeling which demonstrates the suitability of the established system. Moreover, the derived expandable NPC populations can be used for further applications in drug screenings. Their robust phenotypes such as low levels of TPP1 activity in LINCL patient-derived NPCs or cytoplasmic vacuoles, containing storage material, observed in CLN3 mutant NPCs, should serve as possible phenotypic read-outs.

碩士
臺北醫學大學
保健營養學研究所
102
Background： Present studies suggested that lack of mitochondrial biogenesis in Alzheimer’s disease patient’s brain. Quercetin, has been shown to increase mitochondrial biogenesis. Purpose：Our purpose was to exam whether quercetin can increase mitochondrial biogenesis to ameliorate beta amyloid accumulation in H2O2-induced SH-SY5Y neuron cell. Methods: SH-SY5Y cell were treated H2O2 to induce oxidative stress. We pretreated 0, 2.5, 5, 7.5, 10 μM quercetin then to analyze the cellular mitochondrial biogenesis, ATP production and cell apoptosis in H2O2 induced cell. We measured SIRT1, PGC-1α, TFAM, ADAM10, BACE and Aβ levels in SH-SY5Y after quercetin and H2O2 were treated. Results：We found quercetin increased SIRT1, PGC-1α and TFAM level to increase mitochondrial biogenesis. Quercetin treatments also increased ADAM10 level and decrease BACE level to inhibit beta amyloid accumulation in H2O2-induced SH-SY5Y neuron cell. Moreover, decreasing in cell apoptosis was observed in quercetin pretreated SH-SY5Y neuron cell. Conclusion：Querctin can protect cell from H2O2 induced oxidative stress by elevating mitochondrial biogenesis and decreasing beta amyloid accumulation.

碩士
輔仁大學
生命科學系碩士班
97
Alzheimer's disease ( AD ) is a progressive neurodegenerative disorder and characterized with the cerebral deposition of senile plaque which arises from the abnormal accumulation of -amyloid ( A). A is a peptide of 39–43 amino acids from cleavage of the amyloid precursor protein ( APP ). Several evidences indicate that A-activated microglia cells are involved in the neuropathology observed in AD. High levels of inducible nitric oxide synthase ( iNOS ) and nitric oxide ( NO ) have been detected in microglial cells of AD patients and cause neurons death. Epidemiological studies have shown that non-steroidal anti-inflammatory drugs ( NSAIDs ) prevent or delay the onset of AD. To identify potential lead compounds with neuron protection activities, a human microglial cell line, C13NJ, was used as target cells and A-stimulated cells to express iNOS mRNA as an drug-screening model. The following results were obtained: (1) Data of reverse transcription- polymerase chain reaction (RT-PCR) demonstrated that iNOS mRNA could be induced by 20 M A at 2 hr postactivation. (2) Results of Western blotting shown that iNOS proteins expressed in A-activated C13NJ cells at 12 hr postactivation. (3) The data from RT-PCR indicated that there were 3 synthetic compounds (No. 2644, No. 2647, and No. 2648) to downregulate iNOS mRNA expression in C13NJ cells stimulated with A (4) To elucidate Ainduced iNOS mRNA expression in C13NJ cells by which signaling pathways, various inhibitors including phosphoinositide 3-kinase ( PI3K ), NF-κB, and mitogen activated protein kinases ( MAPK ) inhibitors were added into C13NJ cells then iNOS mRNA expression was determined by RT-PCR. The results indicated that p38, JNK, and NF-B inhibitors decreased iNOS gene expression in C13NJ cells stimulated by A. (5) To confirm whether A stimulated NF-B activation, the IB degradation was analyzed with Western blotting. The data shown that IB was significantly degraded at 20 min postactivation. (6) To elucidate whether A. stimulated MAPK activation, the phosphorylation of p38, JNK, and ERK was assayed by Western blotting. The results demonstrated that the phosphorylation of p38, JNK, and ERK could be significantly induced at 20 min postactivation. Based on these results, I suggest that p38, JNK, and NF-B activation are involved in iNOS gene expression in C13NJ cells induced by A. (7) The data of Western blotting indicated that No.2644 and No.2648 increased the levels of IB proteins in A -activated C13NJ cells. (8) Both No.2644 and No.2648 did not affect phosphorylation of JNK and p38 in C13NJ cells induced by A. I suggest that No.2644 and No.2648 inhibit iNOS gene expression in C13NJ cells by modulation of NF-B activation. Furthermore, I set up a neuron protection drug-screening platform in this study. This model can be used to screening more bioactive lead compounds in future.

碩士
國立臺灣海洋大學
食品科學系
106
Alzheimer's disease is a neurodegenerative disease caused by the stacked of β-amyloid peptides. Our laboratory has established the culture medium and condition for the solid-state cultivation of G.lucidum mycelia. This study investigated the protective effect of solid-state fermented crops by G.lucidum against oxidation stress caused by H2O2 and Aβ25-35 in neuroblastoma SH-SY5Y cell line. Various extracts by 10% ethanol extraction using microwave (A), 70oC water extraction (B) and 100oC water extraction followed by ethanol precipitation (C) of G.lucidum fermented product were prepared and the bioactive contents were analyzed. The total glucan and triterpene contents for the 3 extracts were 6.27 – 60.73 g/100 g and 1.43 – 20.30 mg/g, respectively, with the extract C being the highest total glucan content, and the extract B the highest triterpene content. The extract A and B content the more GABA. Median inhibitory concentration for DPPH and ABTS scavenging activity for the all extracts were 6.24 – 10.61 and 8.07 – 11.14 mg/mL, respectively. Pretreatment of extract A significantly reduce the ROS concentration in SH-SY5Y cells. Pretreatment with extracts (10 – 500 μg/mL) followed by H2O2 (150 μM) or Aβ25-35 (10 μM) damage, significantly increased cell viability 6.22-30.64%. In addition, a significant increase in glutathione peroxidase (GPx) was observed in all extracts treated group and GABA standard at 50, 100 μg/mL. A high concentration of G.lucidum extract and 10 μM Aβ25-35 were used to simulate the damage of Alzheimer’s disease cells. Pretreatment with 100 μg/mL extract B could significantly enhance catalase (CAT), superoxide dismutase (SOD) and GPx activity by 1.33, 1.52 and 1.34 times compared to Aβ25-35 control group, respectively. In addition, extract B and C decrease the acetylcholinesterase (AChE) activity and malondialdehyde (MDA) content in neuron cells, respectively. Meanwhile, extract B (100 μg/mL) inhibits cell apoptosis to protect neuron cells from H2O2- and Aβ25-35-induced damage.

博士
長庚大學
生物醫學研究所
97
Collapsin response mediator protein-2 (CRMP-2), a phosphoprotein involved in axonal outgrowth and microtubule dynamics, is aberrantly phosphorylated in Alzheimer disease (AD) brain. Alteration of glycogen synthase kinase-3 (GSK-3) activity is associated with the pathogenesis of AD. CRMP-2 is highly expressed in the developing nervous system. It is believed that CRMP-2 is a neuronal tissue-specific protein, however, its expression in non-neuronal cells has not been investigated clearly. Here I demonstrate that CRMP-2 is not only expressed in neuronal cells but also in non-neuronal cells by RT-PCR and Western blotting analyses. CRMP-2 is differentially expressed in different cell lines and the amount of CRMP-2 protein seems to correlate with the degree of malignancy. In addition, I also identify a novel variant of CRMP-2, the long form of CRMP-2 (CRMP-2L), in human neuronal and non-neuronal cells. Putative human CRMP-2L cDNA sequence was predicted from human genomic database and its expression in human cells was also confirmed. So far as I known, it is the first report that demonstrates the long form of CRMP-2 exist in human cells. I also provide evidence to show that CRMP-2 is one of the major substrates for GSK-3 in pig brain extracts. Both GSK-3 and 3 phosphorylate purified pig brain CRMP-2 and significantly alter it mobility in SDS-gels, resembling the CRMP-2 modification observed in AD brain. Interestingly, this modification can be detected in SK-N-SH neuroblastoma cells treated with a phosphatase inhibitor, okadaic acid (OA), and GSK-3 inhibitors completely block this OA-induced event. Knock-down of both GSK-3 and 3but not either kinase alone, impairs OA-induced modification of CRMP-2. Mutation of Ser-518 or Ser-522 of CRMP-2, which are highly phosphorylated in AD brain, to Ala blocks the OA-induced modification of CRMP-2 in SK-N-SH cells. Ser-522 prephosphorylated by Cdk5 is required for subsequent GSK-3-mediated hyperphosphorylation of CRMP-2 in vitro. However, inhibition of Cdk5 by roscovitine or siRNA pools has little effects on the OA-induced modification of CRMP-2 in cells. Collectively, our results demonstrate for the first time that OA can induce hyperphosphorylation of CRMP-2 in SK-N-SH cells at sites aberrantly phosphorylated in AD brain, and both GSK-3 and 3and Ser-522 kinase(s) are involved in this process.

Tese de mestrado integrado, Engenharia Biomédica e Biofísica (Engenharia Clínica e Instrumentação Médica) Universidade de Lisboa, Faculdade de Ciências, 2020
Neurosurgery has been considered as a treatment or a therapy option for brain lesions with satisfactory outcomes regarding the maximal resection of the lesioned area and the minimal post-surgical neurological dysfunctions by avoiding eloquent areas. For the last two decades, resting-state functional magnetic resonance imaging (rs-fMRI) has emerged as an effective non-invasive neuro-imaging technique that can be used for pre-surgical functional brain mapping at rest. The analysis of these maps can focus on both the local function of specific regions (segregation) and the functional connections between them (integration) for the assessment of lesion-induced changes. The brain network can be characterized resorting to graph theory analysis for the calculation of these segregation and integration properties which is more facilitated on thresholded binarized matrices. These can be obtained by proportional thresholds revealing the top strongest connections that are present in the network. This study intends to analyse and compare the segregation and integration properties of lesioned and non lesioned hemispheric networks from a group of 7 patients with brain tumors and a cavernous malformation. Moreover, it also aims to evaluate the effect of the proportional thresholds in those properties. By using rs-fMRI and graph theory analysis, the network features of lesioned and non lesioned hemispheres were investigated, over a range between 20%-40% (at intervals of 5%) of proportional thresholds. The results reflected more integrated and segregated networks as more connections were included in the networks. The lesioned network revealed higher global integration and local processing at the highest density of 40%. However, the lesioned small-world organization was less optimal when comparing to the non lesioned network. In conclusion, our findings indicated that the lesion-induced perturbations disturbed the functional connectivity of the lesioned hemisphere. Nevertheless, compensatory mechanisms should be accounted for the pre-surgical evaluation of the affected and unaffected brain areas.
A remoção cirúrgica de uma lesão cerebral envolve a resseção da maior área lesada sem comprometer o tecido eloquente e não lesado envolvente. Desta forma, pretende-se minimizar disfunções neurológicas após a cirurgia garantindo a melhor qualidade de vida possível. Como tal, há a necessidade de incluir técnicas de imagem auxiliares à cirurgia que permitam fazer um mapeamento da lesão cerebral, não só ao nível anatómico mas principalmente funcional. Atualmente, existem diversas técnicas de neuro-imagem que atuam de forma não invasiva e que contribuem para o mapeamento funcional do cérebro, num contexto pré-cirúrgico. Destas destaca-se a imagem de ressonância magnética funcional que, tradicionalmente requer que o sujeito execute uma tarefa de modo a extrair as redes neuronais associadas `as regiões ativadas pelo desempenho da tarefa. No entanto, o facto desta técnica apenas possibilitar a extração de redes neuronais singulares somente associadas à tarefa em questão bem como sujeitos com lesões poderem ter dificuldades em realizar a tarefa requerida levou a que nas últimas duas décadas tenha emergido uma nova modalidade - a imagem de ressonância magnética funcional de repouso. Esta distingue-se da anterior no sentido do sujeito não executar qualquer tarefa nem ser submetido a qualquer estímulo. Através da aquisição da actividade cerebral espontânea é possível extrair as redes neuronais relacionadas com a atividade neuronal. Desta forma, a imagem de ressonância magnética funcional de repouso não só supera a limitação da eventual dificuldade ou incapacidade de execução de uma tarefa como também faculta a identificação de múltiplas redes neuronais, ao contrário da ressonância funcional baseada numa tarefa requerida. A extração destas redes neuronais funcionais é conseguida mediante a aplicação de métodos de análise que se focam na localização da função de determinadas regiões cerebrais (segregação) ou na conectividade funcional entre as mesmas (integração). Complementarmente, métodos que englobam tanto a análise da atividade (segregação) como da conectividade (integração) da imagem da ressonância magnética funcional de repouso têm sido combinados com a teoria dos grafos com o objetivo de investigar o cérebro enquanto uma rede neuronal complexa com conexões contínuas e dispersas entre as suas regiões. Ao mesmo tempo também permitem determinar as propriedades da organização funcional cerebral tanto a nível global como local, isto é, em grupos de regiões interligados igualmente denominados de módulos ou comunidades. Para caracterizar a integração e segregação da rede neuronal diversas medidas topológicas associadas à capacidade da rede neuronal partilhar informação entre diferentes regiões podem ser calculadas. O cálculo destas medidas implica a construção da rede neuronal funcional enquanto um conectoma definido por um dado número de regiões cerebrais, designadas de nodos, e pelas conexões funcionais estabelecidas entre as mesmas. O nível de conectividade funcional entre os nodos, também denominado de correlação funcional, é determinado pela computação da correlação entre as séries temporais de cada par de nodos. Posteriormente, estes dados podem ser organizados numa matriz de conectividade cujas entradas representam os pesos da correlação funcional. Contudo, o elevado número de conexões entre as regiões cerebrais dificulta a extração de informação relevante. Neste sentido, a binarização e aplicação de um limiar à matriz de conectividade assegura a redução dessas interações facilitando a determinação das propriedades topológicas da rede neuronal. Limiares que consideram um coeficiente de correlação funcional entre as regiões como o valor limiar para a inclusão das conexões (limiar absoluto) podem ser impostos `a matriz. Por outro lado, ao invés de um limite de correlação, também se pode selecionar uma percentagem das conexões mais fortes a serem incluídas na rede (limiar proporcional ou densidade). Estudos anteriores mostram evidências que limitar as redes neuronais por via de um limite de densidade resulta em medidas topológicas mais estáveis, razão pelo qual os limites proporcionais têm sido mais frequentemente aplicados na sua computação. Ainda assim, não existe um consenso relativamente à escolha ideal do valor do limiar que evita resultados incompletos ou falaciosos. O objetivo deste estudo engloba a análise de redes neuronais relativas aos hemisférios lesados e não lesados para um grupo de sete sujeitos com tumores cerebrais e uma malformação cavernosa. Adicionalmente, para cada hemisfério será calculado um conjunto de medidas de segregação e integração que permite caracterizar a respetiva rede neuronal, seguido de uma comparação entre as mesmas. Estas medidas serão determinadas com base na teoria de grafos aplicação de um conjunto de limiares proporcionais à matriz binarizada, com fundamento nos benefícios explorados acima. Deste modo, em cada rede neuronal, as conexões funcionais entre as suas regiões serão limitadas por um leque de densidades que varia entre 20% e 40%, com intervalos de 5%. Assim, a comparação entre as propriedades hemisféricas será realizada para cada limiar. Além do mais, uma comparação das medidas de grafos entre os diferentes limiares também será conduzida de modo a estudar o efeito dos mesmos na topologia de cada rede neuronal. Em primeiro lugar, os resultados deste estudo demonstraram que tanto o hemisfério lesado como o hemisfério não lesado estão organizados segundo uma rede de pequeno mundo, equilibrando de forma eficaz o processamento local e a integração global. Contudo, o hemisfério lesado mostrou ter uma organização topológica sub-ótima em comparação com o hemisfério não lesado. Os resultados da análise das medidas de integração revelaram que a inclusão de mais conexões nas redes lesadas e não lesadas, através do aumento do limiar proporcional, levou a uma diminuição da distância mínima entre duas regiões, sendo essa diminuição maior no hemisfério não lesado. Complementarmente, a eficiência global associada à comunicação entre as regiões também aumentou com a inclusão de mais conexões nos hemisférios. Desta forma pôde concluir-se que a consideração de mais conexões funcionais nas redes neuronais permitiu comunicações intra hemisféricas mais curtas e consequentemente mais eficientes, numa perspetiva global para as redes neuronais. Para além disso, o hemisfério não lesado revelou uma maior integração global para todas as densidades, exceto para a densidade de 40%. Para esta densidade, a reorganização funcional da rede neuronal lesada mostrou ser mais construtiva permitindo uma eficiência global mais elevada. Ao nível da segregação, os resultados evidenciaram que a escolha de densidades mais elevadas levou a que as redes neuronais fossem mais segregadas. Excluindo as densidades mais elevadas, o nível de conexões locais na rede lesada bem como a sua eficiência de propagação informação local foi menor em comparação com a rede não lesada. Para a densidade de 40%, a transferência de informação local foi mais eficiente no hemisfério lesado. Assim, o hemisfério lesado revelou não só uma maior integração global como também uma especialização local mais eficiente. Em conclusão, as medidas topológicas calculadas pareceram depender da escolha do limiar proporcional que foi aplicado nas redes neuronais. Para as densidades entre 20%-35%, os resultados mostraram que a lesão localizada no hemisfério lesado conduziu a disrupções na estrutura funcional desse mesmo hemisfério (menor integração e segregação). Porém, na densidade de 40%, a reorganização funcional pareceu indicar o estabelecimento de mecanismos e conexões compensatórios suficientes para compensar as perturbações causadas pela presença da lesão nesse hemisfério. De notar que os resultados não mostraram ser totalmente conclusivos quanto ao impacto da lesão na rede não lesada, através de interações funcionais entre os hemisférios. Assim, os resultados deste estudo sugeriram que as perturbações funcionais induzidas pela lesão afetaram a conectividade funcional entre as regiões do hemisfério onde a mesma estava localizada. Como consequência, a escolha de densidades mais elevadas pareceu clarificar conexões e mecanismos de compensação no hemisfério lesado. Deste modo, estas alterações devem ser tidas em conta para a avaliação pré-cirúrgica das áreas cerebrais afetadas e não afetadas.

No
The surface expression and regulated endocytosis of kainate (KA) receptors (KARs) plays a critical role in neuronal function. PKC can modulate KAR trafficking, but the sites of action and molecular consequences have not been fully characterized. Small ubiquitin-like modifier (SUMO) modification of the KAR subunit GluK2 mediates agonist-evoked internalization, but how KAR activation leads to GluK2 SUMOylation is unclear. Here we show that KA stimulation causes rapid phosphorylation of GluK2 by PKC, and that PKC activation increases GluK2 SUMOylation both in vitro and in neurons. The intracellular C-terminal domain of GluK2 contains two predicted PKC phosphorylation sites, S846 and S868, both of which are phosphorylated in response to KA. Phosphomimetic mutagenesis of S868 increased GluK2 SUMOylation, and mutation of S868 to a nonphosphorylatable alanine prevented KA-induced SUMOylation and endocytosis in neurons. Infusion of SUMO-1 dramatically reduced KAR-mediated currents in HEK293 cells expressing WT GluK2 or nonphosphorylatable S846A mutant, but had no effect on currents mediated by the S868A mutant. These data demonstrate that agonist activation of GluK2 promotes PKC-dependent phosphorylation of S846 and S868, but that only S868 phosphorylation is required to enhance GluK2 SUMOylation and promote endocytosis. Thus, direct phosphorylation by PKC and GluK2 SUMOylation are intimately linked in regulating the surface expression and function of GluK2-containing KARs.