Rozprawy doktorskie na temat „Glioma Cell Lines”

The World Health Organization classifies glioblastoma (GBM) as a type IV astrocytoma, making it one of the most fatal tumors that exists. Despite the advances in chemotherapy, surgery, and radiation treatments that improve a patients length of survival, the overall trajectory of the disease remains unchanged. It has been shown that GBM cells produce significant levels of prostaglandins, including prostaglandin D2 (PGD2). PGD2 possesses pro- and anti-tumorigenic properties. Hence, a more complete understanding of PGD2 activity in GBM could yield more effective treatments against GBM. Through techniques like RT-PCR, immunohistochemistry, and HPLC tandem mass spectrometry, we were able to confirm the presence of the PGD2 synthesis in GBM cell lines. We treated GBM cell lines with various concentrations of exogenous PGD2 over 72 hours and observed its effects on cell count, apoptosis, mitosis and viability. Our results suggest that PGD2 possesses contradictory functions in GBM depending on concentration (mM PGD2 vs. nM PGD2) and receptor activation.
A Organização Mundial de Saúde classifica glioblastoma (GBM) como um astrocitoma tipo IV, fazendo uns dos tumores mais fatais que existe. A pesar dos avanços em quimioterapia, cirurgia e radioterapia que melhoram a longevidade de sobrevivência, a trajetória geral da doença permanece imutável. Tem sido demonstrado que células de GBM produzem níveis significativos de prostaglandinas, incluindo prostaglandina D2 (PGD2). PGD2 possui propriedades pro- e anti-tumorigenicos. Então, um entendimento mais completo da atividade de PGD2 em GBM pode gerar tratamentos mais efetivos. Através de técnicas como RT-PCR, imunohistoquimicas e HPLC espectrometria de massa em tandem, conseguimos confirmar a presença da síntese de PGD2 em linhagens de GBM. Tratamos linhagens de GBM com concentrações variáveis de PGD2 exógeno durante 72 horas e observamos seus efeitos na contagem de células, apoptose, mitose e viabilidade. Nossos resultados sugerem que PGD2 possui funções opostas em GBM dependendo em concentração (mM PGD2 vs. nM PGD2) e ativação de receptores.

Il glioblastoma multiforme rappresenta la forma più frequente e maligna fra i tumori cerebrali primari. Nonostante un approccio terapeutico multimodale, che include resezione chirurgica, radio e chemioterapia, la prognosi è generalmente infausta. La maggior parte dei pazienti muore dopo soli 12-15 mesi dalla diagnosi e la percentuale di recidiva è pari all’80%. Crescenti evidenze suggeriscono che una delle principali cause della comparsa di recidive e del fallimento delle attuali strategie terapeutiche è la presenza, all’interno del tumore, di una sottopopolazione di cellule tumorali con caratteristiche staminali, chiamate glioma stem cells (GSCs). Le GSCs sono caratterizzate da un aumentato self-renewal, come dimostrato dall’espressione di tipici marcatori di staminalità, quali CD133 e Nestina, da un’elevata invasività, dalla capacità di differenziare in diversi lineages neurali e da una spiccata chemo e radio resistenza. Una tipica caratteristica delle GSCs è anche l’elevata instabilità cromosomica (CIN). Infatti, tali cellule, presentano numerose alterazioni numeriche e strutturali, delezioni, amplificazioni e perdite di eterozigosità. Diverse alterazioni sono state indicate come responsabili dell’instabilità cromosomica, tra cui difetti in geni codificanti per proteine coinvolte nel macchinario mitotico, come le Aurora chinasi, rendendole un possibile e promettente target terapeutico. La mia tesi si propone, perciò, di indagare due aspetti principali di quest’area di ricerca, con l’obiettivo di individuare nuove possibili strategie terapeutiche GSC-targeted, necessarie per una completa eradicazione del GBM. Nella prima parte del mio progetto ho indagato l’effetto di Danusertib, un inibitore delle Aurora chinasi, su 5 linee di cellule staminali tumorali isolate da glioblastoma, precedentemente caratterizzate nel nostro laboratorio da un punto di vista citogenomico ed epigenomico. I risultati ottenuti hanno mostrato una risposta eterogenea delle diverse linee cellulari all’inibitore: alcune di esse sono risultate maggiormente sensibili, mostrando evidenti alterazioni della morfologia, un considerevole aumento della ploidia e la presenza di un fenotipo senescente associato ad una riduzione del potenziale clonogenico e della proliferazione cellulare. Tuttavia un risultato particolarmente interessante è che anche le linee cellulari più resistenti dopo diversi rounds di esposizione all’inibitore diventano maggiormente sensibili, suggerendo la presenza di una sorta di soglia della plodia superata la quale le cellule vengono indotte alla senescenza. Nella seconda parte della mia tesi ho presentato alcuni dati preliminari ottenuti nel laboratorio del Dr. Hochegger (Genome Damage and Stability Center, University of Sussex, Brighton, UK), dove ho trascorso sei mesi durante il dottorato. Durante questo periodo ho preso parte ad un progetto mirato a sviluppare alcune linee di GSCs in cui i marcatori di staminalità CD133 e Nestina sono taggati con proteine fluorescenti attraverso la tecnica CrispR/Cas9. L’obiettivo è quello di utilizzare tali linee cellulari modificate geneticamente per studiare attraverso tecniche di live imaging la modalità di divisione cellulare messa in atto dalle GSCs, attraverso la quale esse sarebbero in grado, non solo di mantenere in pool di cellule staminali tumorali, ma anche di generare cellule più differenziate che costituiscono la maggior parte della massa tumorale. La comprensione di tale modalità di divisione cellulare potrebbe favorire l’individuazione di nuove strategie terapeutiche, attraverso l’individuazione di nuovi target coinvolti nel controllo della divisione asimmetrica delle GSCs nel tessuto cerebrale.
Glioblastoma is the most common primary malignant brain tumour in the adult population. Despite multimodality treatment with surgery, radiotherapy and chemotherapy, outcomes are very poor, with less than 15% of patients alive after two years. Increasing evidence suggests that Glioma stem cells (GSCs) are likely to play an important role in the biology of this disease and are involved in treatment resistance and tumour recurrence following standard therapy. GSCs are characterized by enhanced self-renewal, highlighted by the expression of stem cell markers, such as CD133 and Nestin, elevated invasive behaviour, chemo and radiotherapy resistance, and the ability to generate multi-lineage progenities. A typical feature of GSCs is also the elevated chromosomal instability (CIN): they are characterized by various numerical and structural aberrations, deletions, amplification and loss of heterozygosity. A variety of alterations have been proposed as being responsible for CIN, including defects in genes involved in the regulation of the mitotic machinery, such as the Aurora Kinases, making them a promising therapeutic target for GSCs depletion. My thesis address two main aspects of this research area, aiming at the identification of new GSCs-targeted therapeutic strategies for GBM complete eradication. In the first part of my project I investigated the effect of Danusertib, a pan-Aurora kinases inhibitor on 5 GSC lines isolated from glioblastoma patients, previously characterized in our laboratory from a cytogenomic and epigenomic point of view. Results showed that response to Danusertib exposure was heterogeneous among GSC lines. Some of them were more sensitive to subtle changes in Aurora kinases activity, which result in huge morphological alterations, a rapid increase in polyploidy and subsequently in senescence, with a consistent reduction in clonogenic survival and proliferation. Interestingly I also observed that the more resistant cell lines showed an increase in ploidy and senescence after repeated rounds of Danusertib exposure, suggesting that there could be the presence of an intolerable ploidy threshold that leads cells to senescence. In the second part of my thesis I presented some preliminary results I achieved in Dr Hochegger’s lab (Genome Damage and Stability Center, University of Sussex, Brighton, UK), where I took part in a project aimed on setting up CrispR/Cas9 mediated GFP or RFP-tagged CD133 (PROM1 gene) and Nestin (NES gene) glioma stem cell lines in order to look, with live cell imaging techniques, for signs for asymmetric cell division, by which a single GSC would be able to both maintain a pool of self-renewing stem cells and produce differential progeny, using live cell imaging. The biological significance of asymmetric or symmetric division modes is not yet fully understood, but improved understanding of this phenomenon may lead to the development of preventative treatments or improved therapeutic options for brain tumour patients through the identification of novel targets that are involved in the control of asymmetric cell division in human brain tissue.

Thesis (M.S.)--Southern Illinois University Carbondale, 2007.
"Department of Molecular Biology, Microbiology and Biochemistry." Includes bibliographical references (leaves 109-126). Also available online.

The main purpose of my research project was to investigate the role of two extracellular matrix-degrading enzymes, heparanase (HPSE) and membrane-type 1 matrix metalloproteinase (MT1-MMP), in pediatric gliomas. I spent the first two years of my PhD program in Dott. Maurizio Onisto’s laboratory (University of Padua). Then I continued my work at New York University School of Medicine, under the supervision of Prof. Paolo Mignatti, whose experimental work focuses on the molecular mechanisms of proteolysis-independent signaling by MT1-MMP and its physiological inhibitor, tissue inhibitor of metalloproteinase-2 (TIMP-2). Gliomas, the most common primary brain tumors, comprise a heterogeneous group of neoplasms that originate from glial cells. Despite recent advances in the management of these tumors, children affected by gliomas, particularly the more aggressive forms, have a poor prognosis. Gliomas can diffusely penetrate throughout the brain, even though they remain localized in this organ. One of the most important events during glioma cell invasion is extracellular matrix (ECM) degradation, a complex mechanism that involves both glycosidic and proteolytic enzymes. HPSE is an endo-β-D-glucuronidase secreted in the ECM, where it cleaves the heparan sulfate side chains of both soluble and membrane-bound proteoglycans. MT1-MMP, a cell membrane-bound proteinase with an extracellular catalytic domain and a short cytoplasmic tail, has been implicated in the proteolytic degradation of extracellular and transmembrane proteins. High levels of HPSE and MT1-MMP are present in a variety of aggressive malignancies, a finding that highlights their important role in cancer invasion and metastasis. In this study we characterized pediatric glioma cell lines derived from different types of gliomas: two glioblastoma multiforme, one anaplastic astrocytoma, one diffuse astrocytoma and one pilocytic astrocytoma. In addition, we used a human breast adenocarcinoma cell line to examine the role of MT1-MMP, because these cells do not express this proteinase and thus represent an ideal model for the regulation of its expression. The data reported here show that MT1-MMP controls activation of intracellular signaling by fibroblast growth factor-2 (FGF-2) and FGF-2 binding to the breast adenocarcinoma cells. We found no clear correlation between HPSE, MT1-MMP or FGF-2 expression and the aggressiveness of the pediatric astrocytoma cells. Gene silencing of HPSE in a pediatric glioblastoma cell line does not affect vascular endothelial growth factor (VEGF) expression or cell proliferation, but upregulates matrix metalloproteinase-2 (MMP-2) and MT1-MMP expression. Moreover, ERK1/2 activation by FGF-2 does not correlate with MT1-MMP expression and is modified by an MMP inhibitor in these pediatric glioma cells. Finally, TIMP-2 controls ERK1/2 activation in all glioma cells. Taken together, the results show that MT1-MMP does not have the same effects in breast carcinoma and pediatric glioma cells, indicating a different and more complex control mechanism of intracellular signaling. This initial characterization of these unique pediatric astrocytoma cell lines provides new insights into the knowledge of this poorly studied group of tumors.
L’obiettivo principale del mio progetto di ricerca è stato analizzare il ruolo di due enzimi che degradano la matrice extracellulare, l’“heparanase” (HPSE) e la “membrane-type 1 matrix metalloproteinase” (MT1-MMP), nei gliomi pediatrici. Ho trascorso i primi due anni di Dottorato nel laboratorio del Dott. Maurizio Onisto (Università di Padova). Ho poi continuato il mio lavoro presso la New York University School of Medicine, sotto la supervisione del Prof. Paolo Mignatti, il cui lavoro sperimentale è focalizzato sull’approfondimento dei meccanismi molecolari alla base dell’attivazione del segnale intracellulare da parte di MT1-MMP e del suo inibitore fisiologico, il “tissue inhibitor of metalloproteinases-2” (TIMP-2). I gliomi, i più comuni tumori cerebrali primari, comprendono un gruppo eterogeneo di neoplasie che originano dalle cellule gliali. Nonostante i recenti progressi raggiunti nel trattamento e nel controllo di tali tumori, la prognosi dei bambini affetti da glioma, ed in particolare dalle sue forme più aggressive, rimane tuttora infausta. Pur essendo confinati nell’organo nel quale originano, i gliomi possono invadere tutte le aree del cervello. Uno degli eventi più importanti che caratterizzano l’invasività dei gliomi è costituito dalla degradazione della matrice extracellulare, un complesso meccanismo che coinvolge enzimi sia glicosidici sia proteolitici. HPSE è una endo-β-D-glucuronidasi secreta nella matrice extracellulare, nella quale taglia le catene di eparan solfato dei proteoglicani solubili e legati alla membrana. MT1-MMP, una proteasi legata alla membrana e composta da un dominio catalitico extracellulare e da una piccola coda citoplasmatica, è coinvolta nella degradazione proteolitica di proteine extracellulari e di membrana. Elevati livelli di HPSE e MT1-MMP sono stati riscontrati in numerosi tipi di tumore e tale evidenza sottolinea il ruolo chiave che essi svolgono nell’invasività tumorale e nella formazione di metastasi. In questo studio sono state caratterizzate cinque linee cellulari di glioma pediatrico derivanti da diversi tipi di glioma: due glioblastomi multiformi, un astrocitoma anaplastico, un astrocitoma diffuso ed un astrocitoma pilocitico. Con lo scopo iniziale di esaminare il ruolo di MT1-MMP nell’attivazione del segnale indotto dall’FGF-2, è stata inoltre utilizzata una linea cellulare di carcinoma mammario, la quale non esprime MT1-MMP e perciò rappresenta un modello ideale per studiare la regolazione della sua espressione. I dati riportati mostrano che, nelle cellule di carcinoma mammario, MT1-MMP regola l’attivazione del segnale intracellulare da parte del “fibroblast growth factor-2” (FGF-2) e controlla il legame di questo fattore di crescita alla superficie delle cellule. Nelle cellule di astrocitoma pediatrico non è stata identificata alcuna chiara correlazione tra espressione di HPSE, MT1-MMP o FGF-2 ed aggressività tumorale. I risultati inoltre dimostrano che il silenziamento genico di HPSE in una linea cellulare di glioblastoma pediatrico non influenza l’espressione del “vascular endothelial growth factor” (VEGF) o la proliferazione cellulare, ma determina la sovraespressione della “matrix metalloproteinase-2” (MMP-2) e di MT1-MMP. Inoltre, nelle cellule di glioma, l’attivazione di ERK1/2 da parte di FGF-2 non correla con l’espressione di MT1-MMP e risulta modificata dal trattamento con un inibitore di MMP. Infine, in tutte le cellule di glioma, anche TIMP-2 regola l’attivazione del segnale intracellulare. In conclusione, i risultati ottenuti mostrano che MT1-MMP non ha gli effetti nelle cellule di carcinoma mammario e di glioma pediatrico, indicando l’esistenza di un differente e più complesso meccanismo di controllo del segnale intracellulare. La caratterizzazione delle linee cellulari di astrocitoma pediatrico presentata in questa tesi offre una più completa conoscenza di questo gruppo di tumori ancora poco studiati.

BACKGROUND. Glioblastoma multiforme (GBM) is the most common and malignant type of glioma and it is characterized by extensive heterogeneity, both at the cellular and molecular level. The poor prognosis and the lack of an effective treatment are due to the presence of a small sub-population of cells with stem-like properties, termed glioma stem cells (GSCs). At the genomic level, heterogeneity is characterized by multiple levels of alterations, including cytogenetic, genomic and epigenomic alterations. Drug sensitivity is an additional level of GSC complexity and heterogeneity and the identification of an effective treatment for GBM depends on the depletion of the GSC pool. Valproic acid (VPA) is a histone deacetylase inhibitor and so it can be used for an epigenetic therapy for cancer. Besides, a differentiation inducing ability of VPA on cancer cells was demonstrated. Paclitaxel (PTX) is a conventional chemotherapeutic agent and in the last years it was shown to be a potential therapeutic drug for gliomas. AIMS AND PROJECT DESIGN. Six GSC lines were studied, as they represent a valuable tool for the investigation of cytogenomic and epigenomic landscapes of GBM, in order to unravel specific molecular pathways, involved in the stem-like counterpart. Drug sensitivity profiles were assessed, evaluating cell viability and cytomorphological parameters (mitotic index, ploidy and polymorphic nuclei), after VPA and PTX administration. The reliability of differentiation and epigenetic therapy through the use of VPA was further investigated by morphological and molecular epigenomic analysis, investigating the DNA methylation status. RESULTS AND DISCUSSION. Several shared cytogenetic and genomic alterations linked to GBM pathogenesis were found among the GSC lines. Specifically, polysomy of chromosome 7, loss of chromosome 10, CDKN2A and CDKN2B deletions are aberrations related to highly relevant pathways in GBM tumorigenesis. Moreover, a minimal deleted region at 1p36.31 was common among the six GSC lines, including CAMTA1 gene, a putative tumor suppressor gene, specific for cancer stem-like cells. Disregulated cytogenenomic pathways in GSCs were preferentially linked to the control of stem cell proliferation, invasion, cellular development and differentiation. The evaluation of the methylation profiles of GSC lines revealed aberrant methylation of developmental genes, which are targeted by Polycomb Repressive Complex 2 in embryonic stem cells and involved in cellular development and nervous system differentiation, evidencing a specific impairment of these processes in cancer stem-like cells. VPA is able to begin a differentiation process in GSCs, as demonstrated by the study of methylation changes caused by VPA, through the methylation of pathways which are involved in self-renewal maintenance, such as Wnt/β-catenin, and several cancer-related mechanisms. Anyway, terminal differentiation was impaired, due to an intrinsic characteristic of cancer cells endowed with stem like properties. GSCs viability was severely affected by dual drug treatment, combining VPA and PTX: VPA caused an initial differentiation, enabling PTX to induce cell death of downstream cells in tumor hierarchy. Thus, a dual approach with drugs affecting different features of malignancy could be a successful approach to GBM treatment. CONCLUSIONS. A multi-level study for the evaluation of cytogenomic and epigenomic landscapes of GSCs is an effective approach for the identification of molecular pathways, specifically de-regulated in stem-like cells, giving an outstanding contribution in the identification of key mechanisms sustaining self-renewal. GSC lines are a valuable tool to evaluate the potentiality of new therapeutical approaches, which should be able to overwhelm the stem-like related counterpart. VPA and PTX combined treatment was found to fulfill the therapeutical potential of VPA and might be a successful approach to unlock the self-renewal loop, typical of GSCs and affect their growth.

Previous work demonstrated that influx of (2-chloroethyl)-1-sarcosinamide-1-nitrosourea (SarCNU) may be altered from the known influx of clinically available chloroethylnitrosoureas which occurs through passive diffusion. The objective of this thesis was to determine the exact mode of influx of SarCNU into human glioma cell lines SK-MG-1 and SKI-1 which are sensitive and resistant to SarCNU, respectively. The influx of SarCNU into SK-MG-1 cells was shown to be mediated by a saturable, energy and sodium independent epinephrine sensitive carrier system. Analysis of influx of SarCNU into SKI-1 cells demonstrated a technically non-saturable mechanism of entry consistent with passive diffusion. Steady-state accumulation of SarCNU was demonstrated to be greater in SK-MG-1 cells versus SKI-1 cells at 37$ sp circ$C whereas there was similar accumulation at 22$ sp circ$C. Differences in steady-state accumulation were not attributable to altered metabolism or efflux. Increased accumulation of SarCNU in SK-MG-1 cells at 37$ sp circ$C was identified to be a consequence of an increased initial rate of influx at 37$ sp circ$C in SK-MG-1 cells versus SKI-1 cells with no significant difference at 22$ sp circ$C. Analysis of chloroethylnitrosourea sensitivity revealed that SKI-1 cells were 3 fold resistant versus SK-MC-1 to SarCNU at 37$ sp circ$C but only 2 fold resistant at 22$ sp circ$C with no temperature shift effect on the 2 fold level of resistance to BCNU. A more detailed analysis of the SarCNU carrier involving the analysis of norepinephrine influx revealed that SarCNU influx into SK-MG-1 cells occurs through the extraneuronal catecholamine uptake$ sb2$ transporter which is not detectable in SKI-1 cells. This is the first direct demonstration of the presence of the uptake$ sb2$ transporter in a human glioma cell line. These findings suggest that increased sensitivity of SK-MG-1 cells to SarCNU is secondary to enhanced accumulation of SarCNU mediated via the uptake$ sb2$ t

Interferons (IFNs) are a class of structurally related proteins first discovered to be produced by virus-infected cells. By now, several other inducing agents have been described. IFNs exert multiple effects on cells exemplified by the establishment of an antiviral state, inhibition of cell proliferation and alteration of different immune reactions. In the present thesis the inhibition of cellular growth concentrated on effects in the early cell cycle have been studied. The human glioma cell line 251 MG was found to be blocked in the S phase of the cell cycle upon addition of IFN both to exponentially growing and growth-factor depleted, synchronized cells. Thymidine kinase and DNA-polymerase activities were reduced in parallel with the S phase effect. 2-5 oligo Anucleotides transfected into glioma cells lead to inhibition of cell growth, exponentially growing cells being blocked in the S phase as during IFN treatment. In contrast, synchronized, restimulated cells were blocked in the cellcycle phase where they resided at the time of transfection. As 2-5 oligo A synthetase activity was induced in the middle of the Gl phase, these results might indicate that the kinetics of expression of oligonucleotides after IFN additiondetermines the type of cell cycle block obtained in differenttumor cells. IFN inhibited preferentially proteins originating from newly synthesized mRNA in Sw 3T3 cells, c-mvc did not seem to be included among these proteins. In both cell systems c-myc expression was unaltered after IFN treatment. In clone T1 selected from the the Sw 3T3 cell line , c-mvc expression was uncoupled to growth and seemed to be growth factor independent. The change in c-myc expression in clone T1 compared to SW 3T3 cells did not render the cells sensitive to IFN. Hence, c-myc regulation does not seem to be the mechanism by which IFN regulates cell growth in this system. The proliferation marker KI-67 antigen was shown not to be causatively involved in growth inhibition of IFN. The reduced levels of the antigen was proposed to be a secondary effect caused by the G0/G1 arrest.

Diss. (sammanfattning) Umeå : Umeå universitet, 1991, härtill 6 uppsatser

digitalisering@umu

O desenvolvimento e invasão de tumores cerebrais primários são diretamente influenciados pela matriz extracelular. Considerando que lisil oxidase (LOX) e os demais membros da família das lisil oxidases (LOXL1, LOXL2, LOXL3 e LOXL4) apresentam complexidade tanto estrutural quanto funcional e estão envolvidos em processos biológicos vitais, como motilidade celular, sinalização celular e regulação gênica, a desregulação da expressão destas proteínas pode levar à gênese e progressão tumoral. O presente trabalho teve como objetivos avaliar os níveis de expressão dos genes que codificam todos os membros da família das lisil oxidases em astrocitomas de diferentes graus de malignidade e correlacionar com a expressão de BMP1 e HIF1A, mutação de IDH1 e tempo de sobrevida total dos pacientes. Adicionalmente, as expressões das proteínas codificadas por estes genes foram também realizadas, além de um estudo funcional in vitro do papel de LOX em astrocitomas. A análise da expressão dos genes foi realizada por PCR quantitativa em tempo real numa série de 153 astrocitomas e 22 amostras de tecido cerebral não neoplásico. A expressão proteica foi conduzida por imuno-histoquímica em amostras de astrocitomas. O silenciamento da expressão de LOX foi realizado em linhagens celulares de glioblastoma humano U87MG e A172 transfectadas com o siRNA para os ensaios funcionais. A expressão de todos os genes (LOX, LOXL1, LOXL2, LOXL3, LOXL4, BMP1 e HIF1A) aumentou com o grau de malignidade dos astrocitomas, com maiores níveis nos casos de glioblastoma. Foi encontrada uma correlação positiva nos valores de expressão dos genes principalmente nos glioblastomas. Somente a expressão de LOXL3 teve um impacto na sobrevida dos casos com GBM. Pacientes com maior expressão apresentaram maior sobrevida em relação aos com menor expressão de LOXL3. Os casos de astrocitoma grau II com mutação de IDH1 apresentaram menor expressão de LOXL1 e de LOXL4 quando comparados com os casos sem mutação. Os casos de GBM com mutação de IDH1, por sua vez, apresentaram menores níveis de expressão de LOX e de LOXL1 do que os casos sem IDH1 mutado. Os níveis de expressão das proteínas da família das lisil oxidases também estavam maiores nas amostras de glioblastoma, com localização nuclear e citoplamastica das células, além de marcação do endotélio. Interessantemente, um caso de glioblastoma com mutação de IDH1 apresentou menor expressão de LOX, inclusive nas células endoteliais. Nas análises funcionais, o silenciamento de LOX por siRNA e o tratamento com o inibidor BAPN das linhagens celulares U87MG e A172 afetaram a capacidade de migração. Além sito, a menor expressão de LOX afetou a capacidade de invasão e crescimento independente de ancoragem das células. Em conjunto, esses resultados corroboram o papel de LOX em processo importantes da tumorigênese dos astrocitomas. Adicionalmente, a expressão de LOX é influenciada pelo status de mutação de IDH1. Portanto, este trabalho fornece novas informações para as possíveis intervenções terapêuticas para o tratamento dos pacientes com astrocitomas
The development and invasion of primary brain tumors are directly influenced by the extracellular matrix. Considering that lysyl oxidase (LOX) and other lysyl oxidase family members (LOXL1, LOXL2, LOXL3 e LOXL4) have both structural and functional complexity and that they are involved in vital biological processes such as cell motility, cell signaling and gene regulation, a deregulation of these proteins can lead to the genesis and tumor progression. This study aimed to evaluate the expression levels of genes that code for the lysyl oxidase family members in astrocytomas of different malignant grades and to correlate to the expression of BMP1 and HIF1A, IDH1 mutation and overall patients\' survival. Moreover, protein expression coded by these genes was also analyzed, besides an in vitro functional study of LOX role in astrocytomas. Gene expression analysis was performed by quantitative real-time PCR in a series of 153 astrocytomas and 22 samples of non-neoplastic brain. Protein expression was analyzed by immunohistochemistry in astrocytoma samples. LOX knockdown was performed in cells of human glioblastoma U87MG and A172 transfected with siRNA. Expression levels of all genes (LOX, LOXL1, LOXL2, LOXL3, LOXL4, BMP1 e HIF1A) increased with the malignant grade of astrocytomas, glioblastomas presenting the higher levels. Positive correlations of gene expression values were observed specially in glioblastomas. Only LOXL3 expression impacted in the overall survival of glioblastoma cases. Patients with higher expression presented longer survival time than those with lower LOXL3 expression. Astrocytoma grade II cases with IDH1 mutation presented lower LOXL1 and LOXL4 expression when compared to those cases with wild type IDH1. On the other hand, GBM cases with IDH1-mutated presented lower LOX and LOXL1 expression than GBM cases without IDH1 mutation. Protein expression levels of lysyl oxidase family members were also higher in glioblastoma samples, with both nuclear and cytoplasmic localization, and also endothelium staining. Interestingly, a glioblastoma case with IDH1-mutated had lower LOX expression, including endothelial cells. For functional analysis, LOX knockdown by siRNA and treatment with inhibitor BAPN of U87MG and A172 cell lines affected migration behavior. Furthermore, lower LOX expression affected invasion capacity and anchorage independent growth. Altogether, these results corroborate LOX role in important processes of astrocytoma tumorigenesis. Additionally, LOX expression is influenced by IDH1 mutational status in glioblastomas. Therefore, our work provides new insights for possible therapeutic interventions for patients with astrocytomas

Les neurones propriocepteurs sont nécessaires au contrôle du mouvement et à la locomotion. Ils connectent les fuseaux musculaires et les tendons aux motoneurones de la moelle épinière pour informer le système nerveux central de l’état d’élongation et de contraction des muscles. Leurs corps cellulaires sont localisés dans les ganglions rachidiens dorsaux (GRD), où ils sont intimement entourés de cellules gliales GFAP-positives appelées cellules satellites gliales (CSG). Comme les astrocytes du système nerveux central, les CSG expriment à leur surface des récepteurs couplés aux protéines Gq (Gq RCPG) qui peuvent être activés par les neurotransmetteurs libérés par les corps cellulaires de neurones sensoriels du GRD. Les corps cellulaires des neurones sensoriels expriment aussi un certain nombre de récepteurs et transmetteurs. Ces caractéristiques, ainsi que la proximité physique entre les CSG et les neurones sensoriels a permis d’émettre l’hypothèse que les deux types cellulaires sont capables de communiquer. De récentes données de la littérature suggèrent que les CSG et les neurones sensoriels responsables de la détection de la douleur sont capables de dialoguer. Cependant, à notre connaissance, aucune donnée n’a permis jusqu’à présent de démontrer une interaction entre les CSG et les neurones propriocepteurs. Dans cette étude, nous avons émis l’hypothèse que l’activation des Gq RCPG des CSG permet la modulation de l’activité des propriocepteurs. Pour tester cette hypothèse, nous avons utilisé des approches techniques complémentaires (imagerie calcique bi-photonique, immunohistochimie, biochimie et analyses comportementales) combinées à un outil chemogénétique puissant basé sur la technologie DREADD afin d’activer sélectivement la voie de signalisation Gq RCPG dans les CSG. Nous avons démontré dans une préparation de GRD intacte que les CSG sont capables de moduler l’activité des propriocepteurs via une signalisation purinergique. Pour tester la pertinence de cette communication, nous avons réalisé des expériences de comportement sensorimoteur et mis en évidence que l’activation des cellules gliales GFAP-positives induit des déficits sensorimoteurs. Déterminer si la modulation des propriocepteurs par les CSG affecte la transmission sensorimotrice a de profondes implications pour la compréhension du système sensorimoteur et de ses dérèglements
Proprioceptive neurons (one’s own neurons) are necessary for controlling motor control and locomotion. They arise from muscle spindles and tendons and synapse onto ventral horn motoneurons to deliver information about the length and contraction of muscles. Proprioceptor somata reside within the dorsal root ganglia (DRG) and are tightly enwrapped in a thin sheath of GFAP-expressing glial cells, called satellite glial cells (SGCs). Interestingly, SGCs express a number of Gq protein- coupled receptors (Gq GPCRs), which can be activated by neurotransmitters released by sensory neuron somata. Sensory neuron somata also express a number of receptors and transmitters. Both the expression of receptors and the close contact between SGCs and sensory neurons led to the hypothesis that these two cell types communicate. There is emerging evidence that SGCs and nociceptive sensory neuron (pain-sensing neurons) somata can communicate. Furthermore, to date, there is no study conducted on SGC-proprioceptor interaction. We hypothesized that SGC Gq GPCR signaling induces the release of neuroactive molecules from SGCs, leading to the modulation of proprioceptor activity. The main goal of this project has been to test this hypothesis using complementary technical approaches (2-photon Ca2+ imaging, immunohistochemistry, biochemistry and behavior) combined with a powerful chemogenetic DREADD-based tool to activate SGC Gq GPCR activity. We have demonstrated ex vivo that SGCs modulate proprioceptive neuron activity through a purinergic pathway. In order to test the physiological relevance of this discovery in vivo, we performed sensorimotor behavioral experiments and have shown that activating GFAP-expressing glial cells induces sensorimotor deficits. Determining whether SGC-induced proprioceptor activity has profound implications in the understanding of sensorimotor functions in health and diseases

Le traitement du gliome experimental c6 de rat par des hautes doses d'acide 13-cis retinoique (cra) est responsable du deces de ces animaux en rapport avec une necrose hemorragique de la tumeur. Nous demontrons que le traitement des gliomes c6 in vivo par des doses de 25 mg/kg/j de cra, augmente l'activateur tissulaire du plasminogene (t-pa), qui apparait comme l'espece moleculaire majeure associee a une activite fibrinolytique responsable de l'hemorragie intra-tumorale. La production de t-pa resulte d'un effet direct des retinoides sur la tumeur. La production de t-pa par la tumeur c6 in vivo est liee a la synthese specifique, temps et dose-dependant, de t-pa par les cellules c6. L'expression du gene du t-pa implique la synthese proteique du recepteur de l'acide retinoique (rar). La voie de signalisation de l'ampc agit synergiquement pour induire la synthese du t-pa. L'action du cra precede celle de l'ampc, et l'activite proteine kinase a est necessaire pour une induction optimale de la production de t-pa par le cra. Les cellules c6 lient le plasminogene et sont responsables d'une augmentation de 42 fois de l'efficience catalytique de la generation de plasmine par le t-pa, principalement due a une diminution de 20 fois du km de la reaction. Cette augmentation de l'affinite de l'enzyme pour son substrat est particulierement favorable a la generation de plasmine a la surface cellulaire. Ceci pourrait expliquer l'activite fibrinolytique et les aires de necrose hemorragiques decrites dans ces tumeurs.

La therapie genique par transfert du gene de la thymidine kinase du virus de l'herpes simplex de type 1 (hsv1-tk) suivi d'un traitement avec la prodrogue ganciclovir (gcv) a ete utilisee pour le traitement de divers cancers. L'efficacite de cette therapie est en partie due a l'existence d'un effet de toxicite de voisinage : le traitement au ganciclovir entraine non seulement la mort des cellules exprimant hsv1-tk mais aussi celle des cellules adjacentes non transfectees. Nous avons entrepris une etude in vitro des mecanismes moleculaires de la toxicite de ce systeme enzyme/prodrogue sur des lignees issues de tumeurs cerebrales. Les resultats obtenus montrent que le couple hsv1-tk/gcv declenche deux reactions cellulaires differentes au stress cytotoxique, selon les lignees utilisees. Dans un premier cas, le traitement au ganciclovir entraine un blocage du cycle cellulaire en phase s rapidement suivi d'une mort cellulaire associant des phenomenes d'apoptose et de necrose. Dans le second cas, le couple hsv1-tk/gcv declenche une mort cellulaire tardive et atypique, sans reel arret du cycle cellulaire. Ce phenomene, associe a l'apparition de cellules geantes et polyploides, est appele catastrophe mitotique et pourrait impliquer la proteine p21. Ces deux modes de mort cellulaire semblent independants de la proteine p53 et impliquent une activation de la proteine pro-apoptotique bax. Le gene represseur de la mort cellulaire bcl-2 inhibe en partie le processus apoptotique declenche par le systeme hsv1-tk/gcv ; ce gene pourrait donc jouer un role dans certaines formes de resistance observees au cours de cette etude. Ainsi, cette etude demontre que le phenotype tumoral va conditionner le destin cellulaire apres le traitement et s'avere donc un determinant critique de la sensibilite a cette approche therapeutique.

Gliomas are primary brain tumors in adults that are believed to originate from different types of glial cells. Central nervous system gliomas pose particularly difficult problems because of their tendency towards malignancy, rate of tumor spread, and the lack of effective therapy. Glioblastoma (GBM/ grade IV glioma) is one of the most aggressive types of gliomas. Despite the advancement in treatment modalities including surgery, chemotherapy and radiotherapy, the overall median survival of a GBM patient remains at 14.6 months. This dismal overall picture of survival is further aggravated by the grim quality of life of the GBM patients. Advances in high-throughput technologies have enabled us to mine the molecular pathways contributing to pathogenesis and resistance of GBM. Integrated genomic and epigenomic screens have revealed several molecular markers and has tremendously improved the classification of these heterogeneous tumors. These finding have also given key insights into the various genetic and epigenetic derailments that contribute towards gliomagenesis. Broadly speaking, genomics and transcriptomics for several decades has kept the focus of research on mutation, copy number variation, chromatin remodellers and transcriptional regulation as major determinants of gene expression alterations in cancer. Recent advances in these fields have paved the way to gain a better understanding of the regulation imposed at the post-transcriptional level. Non-coding RNAs and RNA binding proteins (RBPs) are emerging as important post-transcriptional modulators of gene expression. Our study illustrates the various aspects of RBP biology in GBM, where our focus ranges from surfacing the RBP landscape of GBM, delineating the magnitude of regulation of a particular RBP, IMP3 (IGF2 mRNA binding protein 3) and also determine one of the targets of this protein which contributes towards glioma stem-like cell (GSC) maintenance in GBM. The thesis is divided in three work-related chapters: chapter 3, chapter 4 and chapter 5. In chapter 3, we have performed extensive bioinformatics analyses to get an idea about the altered RBPome of GBM and the probable mechanisms leading to this mis-regulation. In chapter 4, we exemplify IMP3 to study the extent of multi-level post-transcriptional regulation imposed by RBPs in glioma cells. Further, in chapter 5, we conclude that IMP3 is a critical RBP for GSC maintenance. Moreover, we establish p65, a subunit of NF-κB pathway as a translationally activated target of IMP3. p65 also acts as a downstream mediator of IMP3 in maintaining glioma stem-like cell survival and proliferation. CHAPTER 3: Elucidation of the genetic and epigenetic landscape alterations in RNA binding proteins in glioblastoma RNA binding proteins (RBPs) are global regulators which participate in various steps of RNA metabolism. Though, recent studies have indicated their importance in diseases including cancer, the molecular mechanisms regulated by these proteins still remains elusive. In this study, we have carried out an integrated bioinformatics analysis of the status of RBPs (n = 1756) in various datasets (n = 11) to identify several genetic and epigenetically altered events among RBPs in GBM. Alterations in RBPs which included mutation, InDels (Insertion and Deletions) and expression level changes, which may lead to aberrant activity of these global regulators were assessed in our analyses. Whole exome sequencing (WES) data from The Cancer Genome Atlas (TCGA) was analysed to identify mutated RBPs in GBM. We identified 13 RBPs to be mutated in minimum of 2% of GBM samples. Further, correlation of mutations in AHNAK predicted poor prognosis. Integrated analyses of transcriptome, Copy Number Variation (CNV) data, DNA methylome and miRnome were carried out to identify differentially regulated RBPs. There were 472 differentially regulated RBPs. Additional analysis revealed that a significant proportion of differential regulation is due to CNV (9%), DNA methylation (9.5%) and miRNA targeting (37%). To obtain insights into transformation and aggressiveness related RBPs in GBM, we analysed transcript levels of RBPs in grade II and grade IV (GBM) samples. Differentially regulated RBPs in grade II astrocytoma when compared to control brain samples, which had similar expression in GBM were identified as transformation related RBPs. On the other hand, RBPs that showed differential regulation only in GBM when compared to grade II and control brain samples were identified as aggressiveness related RBPs. These two sets of genes may be implicated in initial astrocytic transformation and glioma progression respectively. We also compared the transcriptome data of neural stem cells (NSC), glioma stem-like cells (GSC) and differentiated glioma cells (DGC) to identify GSC specific RBPs signature, which was further subjected to survival analysis and gene set enrichment analysis. These analyses led to the identification of a unique set of differentially regulated RBPs (n = 34) specifically in GSCs compared to NSC and DGC. RBP risk score derived from four prognostic RBPs (NOL3, SUCLG1, HERC5 and AFF3) is demonstrated to be an independent poor prognostic indicator in GBM. RBP risk score also stratified GBM patients into low-risk and high-risk groups with a significant survival difference. Gene set enrichment analysis (GSEA) of differentially regulated genes between high-risk and low-risk identified positive enrichment of NF-κB, inflammatory response, epithelial mesenchymal transition and hypoxia pathways in high-risk GBM. Thus, our study provides a comprehensive overview of genetic and epigenetic regulation of RBPs in glioma development and progression. CHAPTER 4: Transcriptome and translatome regulation by IMP3 in glioma cells IMP3 in the previous section was found to be an upregulated and aggressiveness related RBP in GBM. Previous work from our laboratory had established IMP3 as an RBP which contributes to proliferation, migration, invasion, chemoresistance and angiogenesis of glioma cell lines. Moreover, it was shown to enhance IGF2 translation, without altering its transcript levels. This regulation was attributed to the increased activation of pro-survival pathways like PI3K and MAPK in IMP3 overexpressing cells. Enormous volume of literature demonstrates that RBPs are bestowed with the ability to bind multiple targets and regulate their fate at various levels including their degradation, localization and translation. Recent research has laid emphasis on RBPs as global regulators of RNA metabolism and translation. Hence, we were intrigued to understand the effect of IMP3 on transcriptome and translatome of glioma cells. To get insights in IMP3 modulated transcriptome, we performed a microarray based global gene profiling of total cellular RNA of IMP3 silenced U251 cells. We identified 2788 differentially regulated genes at the transcript level. Further, these differentially regulated genes were classified as direct and indirect targets depending on the presence of IMP3 binding sites. We speculate that the RNA stability of these direct targets may be regulated by IMP3. Our observations suggest that, IMP3 may act as a bimodular regulator of RNA stability of the identified direct targets. This implies that few of the targets may be stabilized while others may be destabilized by IMP3 binding. Correlation of the expression of the differentially regulated targets with IMP3 transcript in GBM tumors (TCGA data is used) revealed a list of genes which is more likely to be regulated by IMP3 in GBM tumors. Biological processes enrichment analysis of these direct and correlated targets suggested that IMP3 regulates cell cycle progression by regulating these genes. Next, we were interested in identifying the genes getting differentially regulated exclusively at the level of translation. Polysome analysis was performed on IMP3 silenced cells, and RNA from the pooled heavy polysome fractions was subjected to microarray. Differentially regulated genes in heavy polysome fractions whose expression remain unaltered in total RNA were selected for further investigation. Similar trend of bimodular regulation was also observed at the level of translation. Direct and indirect targets were identified on the basis of presence of IMP3 binding sites. Interestingly, we found that several direct translation targets of IMP3 were associated with apoptosis and cell death related pathways. We have also directed our efforts in unravelling few of the possible mechanisms which may contribute in regulating the indirect targets, at transcriptome and translatome level. Transcription factors (TFs) and RBPs which are direct targets of IMP3, and which may influence the gene expression of indirect targets of IMP3 were identified. Taken together, we have unravelled the IMP3 regulons in glioma cells and their role in cell cycle progression and apoptosis. CHAPTER 5: IMP3 contributes to glioma stem-like cell maintenance and chemoresistance by promoting the translation of RelA/p65 Therapy resistance presents a severe challenge in battling GBM. One of the culprits for chemo- and radioresistance identified in GBM tumors is a small proportion of slow-dividing glioma stem-like cells (GSCs) which are refractory to current treatment modalities. These cells are thus spared by the treatment and then repopulate to give rise to an even more belligerent recurrent tumor. Identification of molecules specifically expressed in GSCs, but not in their normal counterparts (adult human neural stem cells-ahNSC) may provide potential lucrative targets for therapy. Interestingly, using a publically available microarray data for GSC and ahNSC, we found that IMP3 was the most upregulated RBP in GSCs as compared to ahNSC. Our Gene Set Enrichment Analysis results indicated that GSC signature genes were also closely linked to IMP3 expression in GBM tumors. Experiments carried out in IMP3 silenced conditions revealed that IMP3 is required for GSC maintenance and imparts chemoresistance to glioma cells and GSCs. These experiments provide compelling evidence that expression of IMP3 is imperative for GSC survival and their chemoresistance. Owing to the promiscuous binding of RNA binding proteins, we were intrigued to identify a direct target regulated by IMP3 which is also necessary for GSC maintenance by IMP3. We were keen to identify the transcription factors which are regulated by IMP3 at the translation level. Thus, we focussed on the transcription factors harbouring IMP3 binding sites and which may be unregulated at the transcript level by IMP3. Integrated bioinformatic analysis using published datasets revealed RELA (p65) as a potential target which fulfilled all the aforementioned criteria. Ectopic overexpression and silencing of IMP3 in glioma cell lines confirmed that IMP3 increases the NF-κB pathway activity. Moreover, as expected, p65 transcript levels did not change under IMP3 modulated conditions, while significant protein level changes were observed. IMP3 overexpression led to increased p65 protein levels, while its silencing reduced p65 protein levels significantly in the glioma cell lines. Furthermore, the reduced concentration of p65 protein in IMP3 depleted cells was due to decreased translation of p65 transcript in these conditions, with no effect on its protein stability upon IMP3 modulation. We also establish that IMP3 directly binds to three sites present at the 3’UTR of p65 transcript. Furthermore, alleviation of decrease in neurosphere numbers was observed upon exogenous overexpression of p65 in IMP3 silenced GSCs. We also establish IMP3 as a transcriptional target of p65. Taken together, this study establishes p65 as a novel and bonafide target of IMP3 and as a mediator of IMP3 in GSC maintenance. It also underscores the significance of IMP3 as a therapeutic target, which can indirectly be used to target NF-κB pathway in glioblastoma.

Infiltration of myeloid cells in the tumor microenvironment is often associated with enhanced angiogenesis and tumor progression, resulting in poor prognosis in many types of cancer. The polypeptide chemokine PK2 (Bv8, PROK2) has been shown to regulate myeloid cell mobilization from the bone marrow, leading to activation of the angiogenic process, as well as accumulation of macrophages and neutrophils in the tumor site. Neutralizing antibodies against PK2 were shown to display potent anti-tumor efficacy, illustrating the potential of PK2-antagonists as therapeutic agents for the treatment of cancer. In this study we demonstrate the anti-tumor activity of a small molecule PK2 antagonist, PKRA7, in the context of glioblastoma and pancreatic cancer xenograft tumor models. For the highly vascularized glioblastoma, PKRA7 was associated with decreased blood vessel density and increased necrotic areas in the tumor mass. Consistent with the anti-angiogenic activity of PKRA7 in vivo, this compound effectively reduced PK2-induced microvascular endothelial cell branching in vitro. For the poorly vascularized pancreatic cancer, the primary anti-tumor effect of PKRA7 appears to be mediated by the blockage of myeloid cell migration/infiltration. At the molecular level, PKRA7 inhibits PK2-induced expression of certain pro-migratory chemokines and chemokine receptors in macrophages. Combining PKRA7 treatment with standard chemotherapeutic agents resulted in enhanced effects in xenograft models for both types of tumor. Taken together, our results indicate that the anti-tumor activity of PKRA7 can be mediated by two distinct mechanisms that are relevant to the pathological features of the specific type of cancer. This small molecule PK2 antagonist holds the promise to be further developed as an effective agent for combinational cancer therapy.
Dissertation

Indiana University-Purdue University Indianapolis (IUPUI)
After spinal cord injury (SCI), poor axonal regeneration of the central nervous system, which mainly attributed to glial scar and low intrinsic regenerating capacity of severely injured neurons, causes limited functional recovery. Combinatory strategy has been applied to target multiple mechanisms. Schwann cells (SCs) have been explored as promising donors for transplantation to promote axonal regeneration. Among the central neurons, descending propriospinal neurons (DPSN) displayed the impressive regeneration response to SCs graft. Glial cell line-derived neurotrophic factor (GDNF), which receptor is widely expressed in nervous system, possesses the ability to promote neuronal survival, axonal regeneration/sprouting, remyelination, synaptic formation and modulate the glial response. We constructed a novel axonal permissive pathway in rat model of thoracic complete transection injury by grafting SCs over-expressing GDNF (SCs-GDNF) both inside and caudal to the lesion gap. Behavior evaluation and histological analyses have been applied to this study. Our results indicated that tremendous DPSN axons as well as brain stem axons regenerated across the lesion gap back to the caudal spinal cord. In addition to direct promotion on axonal regeneration, GDNF also significantly improved the astroglial environment around the lesion. These regenerations caused motor functional recovery. The dendritic plasticity of axotomized DPSN also contributed to the functional recovery. We applied a G-mutated rabies virus (G-Rabies) co-expressing green fluorescence protein (GFP) to reveal Golgi-like dendritic morphology of DPSNs and its response to axotomy injury and GDNF treatment. We also investigated the neurotransmitters phenotype of FluoroGold (FG) labeled DPSNs. Our results indicated that over 90 percent of FG-labeled DPSNs were glutamatergic neurons. DPSNs in sham animals had a predominantly dorsal-ventral distribution of dendrites. Transection injury resulted in alterations in the dendritic distribution, with dorsal-ventral retraction and lateral-medial extension of dendrites. Treatment with GDNF significantly increased the terminal dendritic length of DPSNs. The density of spine-like structures was increased after injury and treatment with GDNF enhanced this effect.