Internal ribosome entry site (IRES)-mediated translation is critical for the cell’s ability to respond to stress. Understanding how RNA binding proteins (IRES trans-acting factors; ITAFs) regulate IRESes is crucial to elucidating the mechanism of alternative translation initiation. Furthermore, determining how these ITAFs are regulated is central to understanding their functions in diseased states. I have identified the tumour suppressor programmed cell death 4 (PDCD4) as a novel ITAF of the XIAP and Bcl-xL IRES elements. I demonstrate that under normal conditions, PDCD4 acts to inhibit translation from these IRES elements by preventing formation of the 48S translation initiation complex. Furthermore, I show that in response to treatment with the pro-survival fibroblast growthfactor-2 (FGF-2), S6 kinase 2 (S6K2) phosphorylates PDCD4 leading to its degradation and the subsequent de-repression of XIAP and Bcl-xL translation. Importantly, I demonstrate the clinical significance of this regulation in glioblastoma multiforme (GBM) tumours where the loss of PDCD4 expression correlates with an increase in Bcl-xL protein and poor patient outcome. Additionally, re-expression of PDCD4 down-regulates Bcl-xL and decreases cell viability, and direct inhibition of Bcl-xL by a small molecule antagonist ABT-737 sensitizes GBM cells to the chemotherapeutic doxorubicin. Finally, I demonstrate that PDCD4 can be regulated at multiple levels. Importantly, I identify the RNA binding protein HuR as a regulator of microRNA (miR) -21 induced silencing of PDCD4. I show that HuR can bind the PDCD4 3'UTR and prevent miR-21 binding, and that a loss of PDCD4 expression following H2O2 treatment is mediated via miR-21. These results provide novel insight into the role of PDCD4 as a tumour suppressor and highlight the importance of ITAFs in cancer progression.

Submitted by Fabio Sobreira Campos da Costa (fabio.sobreira@ufpe.br) on 2017-08-04T13:56:59Z No. of bitstreams: 2 license_rdf: 811 bytes, checksum: e39d27027a6cc9cb039ad269a5db8e34 (MD5) Tese de doutorado - Mauro Gomes..pdf: 11151423 bytes, checksum: 8078c54c11b179c8f72bccc64042575f (MD5) / Made available in DSpace on 2017-08-04T13:56:59Z (GMT). No. of bitstreams: 2 license_rdf: 811 bytes, checksum: e39d27027a6cc9cb039ad269a5db8e34 (MD5) Tese de doutorado - Mauro Gomes..pdf: 11151423 bytes, checksum: 8078c54c11b179c8f72bccc64042575f (MD5) Previous issue date: 2016-08-26 / CAPES / Utilizando a estratégia de hibridação molecular, neste trabalho foi construído uma quimioteca de heterocíclicos furânicos, triazólicos e quinolínicos contendo o núcleo naftoquinona, reunindo características estruturais de compostos bioativos distintos, originando assim, moléculas híbridas com amplo potencial farmacológico. Inicialmente, foram sintetizados seis derivados 2-acetoxi-3-alquinill-1,4naftoquinonas via reação de acoplamento Sonogashira entre o 2-acetoxi-3-iodo-1,4naftoquinona e diversos alquinos terminais funcionalizados com rendimentos que variaram de 40-73%, que posteriormente foram submetidos a uma heterocilização intramolecular, formando os derivados furanonaftoquinonas com rendimentos bons entre 72-85%. Os derivados 2-acetoxi-3-alquinil-1,4-naftoquinonas e os furanonaftoquinonas foram submetidos à avaliação do potencial citotóxico em três linhagens de células de glioblastomas, GBMO2, GBM95 e A172, apresentando, no geral, resultados satisfatórios para inibição do crescimento celular. Os compostos 2acetoxi-3-feniletinil-1,4-naftoquinona, 2-acetoxi-3-(4-metoxilfeniletinil)-1,4naftoquinona e 2-acetoxi-3-(4-metilfeniletinil)-1,4-naftoquinona se destacaram dentre as substâncias analisadas por apresentarem menor CI50 para as três linhagens celulares de glioblastomas testadas, resultados estes significativos para dar continuidade nos estudos de citotoxicidade. Em seguida, foi desenvolvida uma nova rota sintética para obtenção por “click chemistry” de novos compostos aminoalquiltriazóis naftoquinônicos através das reações de cicloadição 1,3-dipolar entre 2azidoalquilamino-1,4-naftoquinonas e diversos alquinos terminais, sendo sintetizados vinte novos derivados 2-[(1H-1,2,3-triazol-1-il)alquilamino]-1,4naftoquinonas com rendimentos entre 70-97%. Estes heterocíclicos triazólicos foram avaliados frente às linhagens tumorais HEp-2 (carcinoma de laringe humana), NCIH292 (carcinoma mucoepidermoide de pulmão humano), HT-29 (adenocarcinoma de colón humano), MCF-7 (câncer de mama humano) e HL-60 (leucemia promielocitica aguda). Os compostos 2-[2-(4-propil-1H-1,2,3-triazol-1-il)etilamino]-1,4-naftoquinona e 2-{3-[4-(2-hidroxibutan-2-il)-1H-1,2,3-triazol-1-il]propilamino}-1,4-naftoquinona exibiram citotoxicidade moderada frente às linhagens HL-60, HL-60 e MCF-7, respectivamente, demonstrando ação inibitória seletiva. Por fim, foram sintetizados vinte quatro derivados 6-alquilamino-5,8-quinolinoquinonas a partir de uma direta aminação nucleofílica do 7-bromo-5,8-quinolinoquinona com aminas primárias e secundárias, sendo desenvolvida uma nova estratégia sintética para a obtenção dos compostos 6-alquilamino-5,8-quinolinoquinonas a partir de aminas primárias. Estes compostos são promissores candidatos para desenvolvimento de novas drogas antitumorais. / By using the strategy of molecular hybridization, in this work it was built a chemical library with furan, triazole and quinoline heterocyclic compounds, containing the naphthoquinone nucleus, gathering structural characteristics of distinct bioactives resulting in hybrid molecules with large pharmacological potencial. Initially, six 2acetoxy-3-alkynyl-1,4-naphthoquinone derivatives were synthesized through Sonogashira cross coupling reaction involving 2-acetoxy-3-iodo-1,4-naphthoquinone and several functionalized terminal alkynes in 40-73% yields, which later were submitted to intramolecular heterocyclization, forming the furan derivatives in 72-85% yields. The 2-acetoxy-3-alkynyl-1,4-naphthoquinone and furanonaphthoquinone derivatives were submitted to cytotoxic screening against three glioblastoma cell lines GBMO2, GBM95 e A172, resulting in satisfactory results to the inhibition of cellular growth. The 2-acetoxy-3-phenylethynyl-1,4-naphthoquinone, 2-acetoxy-3-(4methoxyphenylethynyl)-1,4-naphthoquinone and 2-acetoxy-3-(4methylphenylethynyl)-1,4-naphthoquinone stood out among the substances analyzed by their lower IC50 for the three cell lines tested glioblastomas. These results are significant to continue in the cytotoxicity studies. Then, a new synthetic route by “click chemistry” was developed to obtain new aminoalkyl-triazoles naphthoquinone compounds through the reactions of 1,3-dipolar cycloaddition between 2azidoalkylamino-1,4-napthoquinones and several terminal alkynes. A serie of twenty 2-[(1H-1,2,3-triazole-1-yl)alkylamino]-1,4-naphthoquinones derivatives were synthetized in 70-97% yields. These triazole heterocyclics were tested against the tumor cell lines HEp-2 (human laryngeal carcinoma), NCI-H292 (human mucoepidermoid lung carcinoma), HT-29 (human colon adenocarcinoma), MCF-7 (human breast cancer) and HL-60 (human promyelocytic leukemia). The compounds 2-[2-(4-propyl-1H-1,2,3-triazole-1-yl)ethylamino]-1,4-naphthoquinone and 2-{3-[4-(2hydroxybut-2-yl)-1H-1,2,3-triazole-1-yl]propylamino}-1,4-naphthoquinone showed moderated cytotoxicity against HL-60, HL-60 e MCF-7, showing a selective inhibition profile. Finally, twenty four 6-alkylamino-5,8-quinolinequinones were obtained by direct nucleophilic amination of 7-bromo-5,8-quinolinequinone with primary and secondary alkylamines, providing a new synthetic strategy to the acquisition of 6alkylamino-5,8-quinolinequinones compounds from primary amines. These compounds are promising candidates for the development of new antitumor drugs.

abstract: Glioblastoma Multiforme (GBM) is a grade IV astrocytoma and the most aggressive form of cancer that begins within the brain. The two-year average survival rate of GBM in the United States of America is 25%, and it has a higher incidence in individuals within the ages of 45 - 60 years. GBM Tumor formation can either begin as normal brain cells or develop from an existing low-grade astrocytoma and are housed by the perivascular niche in the brain microenvironment. This niche allows for the persistence of a population of cells known as glioma stem cells (GSC) by supplying optimum growth conditions that build chemoresistance and cause recurrence of the tumor within two to five years of treatment. It has therefore become imperative to understand the role of the perivascular niche on GSCs through in vitro modelling in order to improve the efficiency of therapeutic treatment and increase the survival rate of patients with GBM. In this study, a unique three dimensional (3D) microfluidic platform that permitted the study of intercellular interactions between three different cell types in the perivascular niche of the brain was developed and utilized for the first time. Specifically, human endothelial cells were embedded in a fibrin matrix and introduced into the vascular layer of the microfluidic platform. After spontaneous formation of a vascular layer, Normal Human Astrocytes and Patient derived GSC were embedded in a Matrigel® matrix and incorporated in the stroma and tumor regions of the microfluidic device respectively. Using the established platform, migration, proliferation and stemness of GSCs studies were conducted. The findings obtained indicate that astrocytes in the perivascular niche significantly increase the migratory and proliferative properties of GSCs in the tumor microenvironment, consistent with previous in vivo findings. The novel GBM tumor microenvironment developed herein, could be utilized for further in-depth cellular and molecular level studies to dissect the influence of individual factors within the tumor niche on GSCs biology, and could serve as a model for developing targeted therapies. / Dissertation/Thesis / Masters Thesis Biomedical Engineering 2020

L’hypoxie est une caractéristique majeure des glioblastomes (GB). Elle est la cause principale de la résistance aux traitements observée dans ces tumeurs. Les conséquences de la baisse en oxygène au niveau tumoral, sont médiées par les facteurs de transcription Hypoxia Inducible Factors (HIF). Ces facteurs sont des protéines hétérodimériques HIF-α/HIF-1β responsables de la transcription de nombreux gènes cibles. Certaines de ces cibles participent à la progression tumorale et à la mise en place d’un phénotype agressif. L’une des cibles de l’isoforme HIF-2α, est la βIII-tubuline (βIII-t). Cette protéine, qui compose les microtubules, est décrite comme surexprimée dans les gliomes de haut grade, comme les GB. Ces travaux de thèse s’intéressent au rôle de la βIII-t dans la progression tumorale ainsi qu’au développement de stratégies permettant d’inhiber l’expression de HIF-2α. Les résultats obtenus montrent que la βIII-t a une importance centrale dans le développement tumoral. En effet, les tumeurs issues de l’implantation de cellules humaines de GB invalidées pour la βIII-t, se développent significativement moins vite comparées aux tumeurs contrôles. In vitro, nous avons montré que cette protéine est impliquée dans la prolifération, la migration et l’invasion cellulaires. En revanche, nous n’avons pas pu confirmer que la βIII-t est impliquée dans la résistance aux traitements (chimio- ou radiothérapeutiques). Les deux composés testés comme inhibiteurs de HIF-2α (SR2933 et PT2385) ont montré des résultats prometteurs sur la βIII-t, gène cible spécifique de HIF-2α. Cependant, malgré les stratégies développées, nous n’avons pas pu évaluer l’efficacité directe de ces deux composés sur l’hétérodimérisation de HIF-2α avec HIF-1β. / Hypoxia is a major feature of glioblastoma (GB). It is the main cause of the resistance to treatments observed in these tumors. The consequences of the decrease in oxygen at the tumor level, is mediated by the Hypoxia Inducible Factors (HIF). These transcription factors are heterodimeric proteins HIF-α/HIF-1β responsible for the transcription of many target genes. Some of these targets are responsible for setting up an aggressive phenotype and in tumor progression. One of the targets of the HIF-2α isoform is βIII-tubulin (βIII-t). This protein, which is a constituent of microtubules, is described as overexpressed in high grade gliomas, such as GB. In the present thesis, we examined the role of βIII-t in tumor progression and we developed strategies to inhibit the expression of HIF-2α. Our results show that βIII-t is of central importance in tumor development. Indeed, tumors resulting from the implantation of GB human cells invalidated for βIII-t, developed significantly less rapidly compared to control tumors. In vitro, we have shown that this protein is involved in cell proliferation, migration and invasion. However, we could not confirm that βIII-t is involved in resistance to treatments (chemotherapeutic or radiotherapeutic).The two compounds tested as inhibitors of HIF-2α (SR2933 and PT2385) showed promising results on βIII-t, a specific target gene of HIF-2α. However, despite the different tested strategies, we could not evaluate the direct inhibitory action of these two compounds on the heterodimerization of HIF-2α with HIF-1β.

Les thérapies ciblées utilisent des agents thérapeutiques qui interfèrent specifiquement avec les molécules nécessaires pour la croissance et la progression tumorale. Les chimiothérapies classiques sont toxiques pour les cellules qui se divisent rapidement du à leur interaction avec les cellules en division. Le premier but des thérapies ciblées est de combattre plus précisement les cellules cancereuses et ainsi éviter les effets indesirables.La thérapie anti-cancereuse utilisant les anticorps a été développé depuis environ 15 ans et est actuellement une des plus efficaces des thérapies ciblées. Dans certains cas des anticorps monoclonaux sont conjugués avec des isotopes radioactifs ou des toxines pour permettre une délivrance ciblée de ces derniers dans les cellules cancereuses. De plus en plus, pour remplacer les anticorps, les thérapies ciblées utilisent des peptides ou des acides nucleiques comme agent ciblant.Dans ce travail, nous avons utilisé des stratégies diverses de ciblage pour permettre l’internalisation de substance toxique (une toxine ou son gène) specifiquement dans les cellules cancereuses. Nous avons travaillé sur le modèle du glioblastome.Notre groupe a publié un article décrivant l’utilisation du domaine PDZ d’une proteine hCASK permettant la liaison a un biomarqueur surexprimé dans les cellules cancereuses. En effet, ce domaine PDZ est capable de se lier à la partie C-terminale de la proteine CD98. Le produit de fusion hCASK-PDZ a été génétiquement lié a une toxine (la saporine). Nous avons démontré une activité in vitro évidente de ce conjugué dans les cellules de glioblastome.Dans cette étude, nous avons utilisé d’autres agents ciblant une autre proteine surexprimée à la surface des cellules cancereuses: la nucléoline. Dans ce contexte, un aptamer et un pseudopetide se liant specifiquement à cette dernière ont été developpé et étudié.L’aptamer AS1411 (Antisoma, UK) est un agent ciblant la nucleolin approuvé par la FDA (food and drug administration). Il se lie à la nucleoline et est internalisé, perturbant ainsi l’interaction de cette dernière avec ses partenaires inhibant ainsi proliferation cellulaire.En parallèle, notre groupe a developpé un pseudopeptide, antagoniste de la nucleoline (Nucant, N6L). Il a été montré qu’il inhibe drastiquement la croissance tumorale dans le cancer du sein en induisant l’apoptose et il est actuellement en préparation pour une phase II d’essai clinique (IPP-204106). Nous avons démontré l’effet anti-proliferatif du N6L in vitro sur des cellules primaires de glioblastome.La surexpression de la nucleoline à la surface des cellules de glioblastomes couplée à la specificité de l’aptamere et du N6L pour cette deniere nous ont amené à vouloir augmenter leur efficacité d’action en les liant à une toxine. Le gène codant pour la saporine (proteine inhibitrice des ribosomes ) ou la saporine elle-meme a donc été lié à l’aptamere et au N6L.Nous avons donc dans ce travail etudié l’activité cytotoxique de l’aptamere et du N6L liés à la saporine. Les résultats obtenus sont evalués pour des futurs publications.Toutes les approches ciblées décrites, en dépit de certains problèmes, semblent prometteuses et nécessitent d'autres recherches, mais confirment que l'exploitation de cibles pour fournir des substances toxiques est l'avenir de la thérapie pour les formes cancéreuses difficiles à battre avec les thérapies conventionnelles. / Targeted cancer therapies are drugs designed to interfere with specific molecules necessary for tumor growth and progression. Traditional cytotoxic chemotherapies usually kill rapidly dividing cells in the body by interfering with cell division. A primary goal of targeted therapies is to fight cancer cells with more precision and potentially fewer side effects.Antibody-based therapy for cancer has become established over the past 15 years and is now one of the most successful and important targeted strategies. In some cases, monoclonal antibodies are conjugated to radio-isotopes or toxins (immunotoxin) to allow specific delivery of these cytotoxic agents to the intended cancer cell target. Furthermore thargeted therapies may be based also on the use of targeting molecues other than antibodies, such as peptides, growth factors, and also nucleic acids.Indeed, in this work we studyed a multi targeting strategy to deliver toxic substances (protein toxin or its gene) to cancer cells (glioblastoma).Our group published a paper describing the use of PDZ protein domain of hCASK (serine kinase calcium/calmodulin-dependent of MAGUK family) and to exploit the ability of this protein to bind to the C-terminus of hCD98 in the extracellular space. CD98 is an interesting target because it is overexpressed in different types of tumors (Giansanti F., 2015). hCASK-PDZ was genetically fused to the toxin saporin and this chimeric toxin proved to be active on glioblastoma cells in vitro.Other cell killing agents were designed to recognize and bind specifically nucleolin (NCL). This multifunctional protein is overexpressed on the surface of activated endothelial and tumor cells. In this context, compounds targeting NCL, such an aptamer, and a multivalent pseudopeptide, have been developed and investigated for cancer therapy.The aptamer against NCL, NCL-APT also known as AS1411 (Antisoma, UK), is a US Food and Drug Administration (FDA)-approved NCL targeting agent. It binds to NCL on the cell surface, preferentially gets internalized, and inhibits cancer cell growth sparing normal cells (Bates PJ, 2009).In parallel, our group, recently developed a multivalent synthetic pseudopeptide N6L that selectively binds to nucleolin (Destouches D., 2011). N6L strongly inhibits breast cancer growth by inducing apoptosis of tumor cells and is currently in preparation for phase II clinical trials (IPP-204106). We demonstrated the anti-proliferative effect of N6L on human glioblastoma cells in primary culture prepared form post-surgical specimens (Benedetti E, 2015).The overexpression of NCL on glioblastoma cell surface and the recognized selectivity of AS1411 and N6L prompted us to study a way to increase the efficiency of these ligands binding them Saporin coding gene or the protein toxin Saporin-S6, a type 1 RIP (Ribosome-Inactivating Protein) widely studied because of its potential therapeutic application in a variety of human diseases as toxic moiety of a conjugate.The characterization of the toxic activity of AS1411 linked to saporin gene (APT-SAP) and of NCL linked to saporin protein (SAP-N6L) is therefore described. Both these researches are under evaluation for publication.All the described thargeted approaches, nothwithstanding some problems, look promising and need further research, but confirm the fact that exploiting targets to deliver toxic substances is the future of therapy for cancer forms that are difficult to beat with conventional therapies.

Un cadre statistique bayésien a été créé dans le cadre de cette thèse pour le développement de modèles cliniques basés sur une approche d’apprentissage continu dans laquelle de nouvelles données peuvent être ajoutées. L’objectif des modèles est de prévoir les effets de la radiothérapie à partir de preuves cliniques. Des concepts d’apprentissage machine ont été utilisés pour résoudre le cadre bayésien. Les modèles développés concernent un cancer du cerveau agressif appelé glioblastome. Les données médicales comprennent une base de données d’environ 90 patients souffrant de glioblastome ; la base de données contient des images médicales et des entrées de données telles que l’âge, le sexe, etc. Des modèles de prévision neurologique ont été construits pour illustrer le type de modèles qui sont obtenus avec la méthodologie. Des modèles de récidive du glioblastome, sous la forme de modèles linéaires généralisés (GLM) et de modèles d’arbres de décision, ont été développés pour explorer la possibilité de prédire l’emplacement de la récidive à l’aide de l’imagerie préradiothérapie. Faute d’une prédiction suffisamment forte obtenue par les modèles arborescents, nous avons décidé de développer des outils de représentation visuelle. Ces outils permettent d’observer directement les valeurs d’intensité des images médicales concernant les lieux de récidive et de non-récurrence. Dans l’ensemble, le cadre élaboré pour la modélisation des données cliniques en radiothérapie fournit une base solide pour l’élaboration de modèles plus complexes. / A Bayesian statistics framework was created in this thesis work for developing clinical based models in a continuous learning approach in which new data can be added. The objective of the models is to forecast radiation therapy effects based on clinical evidence. Machine learning concepts were used for solving the Bayesian framework. The models developed concern an aggressive brain cancer called glioblastoma. The medical data comprises a database of about 90 patients suffering glioblastoma; the database contains medical images and data entries such as age, gender, etc. Neurologic grade predictions models were constructed for illustrating the type of models that can be build with the methodology. Glioblastoma recurrence models, in the form of Generalized Linear Models (GLM) and decision tree models, were developed to explore the possibility of predicting the recurrence location using pre-radiation treatment imaging. Following, due to the lack of a sufficiently strong prediction obtained by the tree models, we decided to develop visual representation tools to directly observe the medical image intensity values concerning the recurrence and non-recurrence locations. Overall, the framework developed for modeling of radiation therapy clinical data provides a solid foundation for more complex models to be developed.

Les glioblastomes (GB) sont des tumeurs cérébrales invasives, résistantes et qui récidivent systématiquement malgré un traitement associant chirurgie, radio- et chimiothérapie. Ces tumeurs de très mauvais pronostic, se caractérisent par une survie médiane de 15 mois. L'agressivité des GB est notamment due à la présence d'une sous-population de cellules souches (GSC). Les GSC sont caractérisées par leurs capacités d'auto-renouvellement, d'expression de différents marqueurs souches, de multipotence et de tumorigenèse. Elles sont fortement impliquées dans la résistance et la récidive tumorale et leur ciblage pourrait améliorer le traitement des GB. Au vu de la littérature et de résultats obtenus au laboratoire, l'intégrine ß8 est apparue comme une nouvelle cible potentielle de ces GSC. Cette intégrine est décrite comme possédant un rôle majeur dans la survie et l'auto-renouvellement des cellules progénitrices neurales saines et sa surexpression est associée à une diminution de la survie des patients. Nous avons alors émis l'hypothèse que l'intégrine ß8 pourrait être impliquée dans le maintien de l'état souche des GSC. J'ai démontré au cours de ma thèse que cette intégrine est surexprimée dans des primocultures de GSC issues de résections de GB ainsi que dans des coupes de tumeurs de patients. De plus, j'ai mis en évidence que ß8 est associée à l'état souche et à des fonctionnalités propres à ces cellules, notamment leur auto-renouvellement, leur viabilité, leur migration et leur radiorésistance. En inhibant sélectivement ß8 dans nos primocultures de GSC par si/shRNA, j'ai en effet observé in vitro une diminution de la formation de neurosphères et de la migration cellulaire ainsi qu'une augmentation de la différentiation et de la mort cellulaire, cette dernière étant potentialisée après irradiation. Enfin, in vivo, j'ai mis en évidence que l'inhibition de ß8 se traduit par une diminution de la tumorigenèse et une augmentation de la survie des souris. En conclusion, mes résultats de thèse permettent d'identifier l'intégrine ß8 comme une protéine membranaire nécessaire au maintien de l'état souche dans les GSC mais surtout comme une potentielle cible thérapeutique radiosensibilisante dans les GB. / Glioblastomas (GB) are invasive, resistant and recurrent brain tumors (median overall survival of 15 months) despite standard treatment including surgical resection, radio- and chemotherapy. This tumor aggressiveness could partly be explained by the presence into the tumor of Glioblastoma-Stem like Cells (GSC), characterized by their ability to self-renew, their higher expression of specific GSC markers, their multipotent aptitude and their high tumorigenic potential. They are strongly involved in tumor resistance and recurrence and their targeting could improve GB treatment. Regarding current literature but also transcriptomic results obtained in our lab, a specific ß8 integrin emerged as a potential selective target in GSC. We then hypothesized that ß8 integrin could be involved in stemness maintenance of GSC. I first demonstrated, during my doctoral thesis, that ß8 is overexpressed in primocultures of GSC isolated from patients resections and also in human GB samples. Moreover, I showed that this integrin could be associated with stemness and features unique to these cells, including self-renewal ability, viability, migration and radioresistance. Indeed, the selective inhibition of ß8 in GSC by si/shRNA resulted in vitro in a decrease of neurosphere formation and migration, associated with an increase of differentiation patterns and cell death, this one being potentiated after irradiation. Finally, in vivo, I showed that ß8 inhibition decreased tumorigenesis and increased mice survival. In conclusion, my doctoral results allow to identify ß8 integrin as a membrane protein essential for stemness maintenance of GSC but mostly as a new potential radiosensitizing therapeutic target for GB.

Charles University Faculty of Pharmacy in Hradec Králové Department of Biochemical Sciences Candidate: Lucia Tomášková Supervisor: prof. PharmDr. Tomáš Šimůnek, Ph.D. Title of diploma thesis: Study of exosomes as a drug delivery system in the treatment of glioblastoma Central nervous system disorders are among the most serious diseases affecting humans. They affect not only the patient's life, but also his/her surroundings. Therefore, their therapy, whether at the level of complete cure or alleviation of accompanying symptoms, is a challenge for scientific research. In our research, we focused on glioblastoma multiforme, a brain cancer not yet treatable. The main drawback in therapy is overcoming the blood-brain barrier. Exosomes, such as the body's natural nano-vesicles, have been shown to be a suitable system for delivering drugs to brain tissue. Our research has shown that by a suitable method we are able to obtain sufficient quality exosomes from macrophage and fill them very efficiently with antitumor agents paclitaxel, doxorubicin and temozolomide, while the delivered substances show higher efficacy and fewer side effects than the free form.

Background Glioblastoma (GBM) is a tumor of the central nervous system. After surgical removal and standard therapy, recurrence of tumors is observed within 6–9 months because of the high migratory behavior and the infiltrative growth of cells. Here, we investigated whether carnosine (β-alanine-l-histidine), which has an inhibitory effect on glioblastoma proliferation, may on the opposite promote invasion as proposed by the so-called “go-or-grow concept”. Methods Cell viability of nine patient derived primary (isocitrate dehydrogenase wildtype; IDH1R132H non mutant) glioblastoma cell cultures and of eleven patient derived fibroblast cultures was determined by measuring ATP in cell lysates and dehydrogenase activity after incubation with 0, 50 or 75 mM carnosine for 48 h. Using the glioblastoma cell line T98G, patient derived glioblastoma cells and fibroblasts, a co-culture model was developed using 12 well plates and cloning rings, placing glioblastoma cells inside and fibroblasts outside the ring. After cultivation in the presence of carnosine, the number of colonies and the size of the tumor cell occupied area were determined. Results In 48 h single cultures of fibroblasts and tumor cells, 50 and 75 mM carnosine reduced ATP in cell lysates and dehydrogenase activity when compared to the corresponding untreated control cells. Co-culture experiments revealed that after 4 week exposure to carnosine the number of T98G tumor cell colonies within the fibroblast layer and the area occupied by tumor cells was reduced with increasing concentrations of carnosine. Although primary cultured tumor cells did not form colonies in the absence of carnosine, they were eliminated from the co-culture by cell death and did not build colonies under the influence of carnosine, whereas fibroblasts survived and were healthy. Conclusions Our results demonstrate that the anti-proliferative effect of carnosine is not accompanied by an induction of cell migration. Instead, the dipeptide is able to prevent colony formation and selectively eliminates tumor cells in a co-culture with fibroblasts.:1 Introduction ........................................................................2 1.1 Glioblastoma ........................................................................................................... 2 1.1.1 Taxonomy, epidemiology and general features of GBM ......................... 2 1.1.2 GBM subtypes and molecular diagnostic .................................................. 3 1.1.3 Therapy ........................................................................................................... 4 1.1.4 GBM cell migration and invasion ................................................................. 5 1.2 Carnosine ................................................................................................................ 6 1.2.1 Chemistry, Biology, Distribution .................................................................. 7 1.2.2 Carnosine homeostasis ................................................................................ 7 1.2.3 Physiological functions.................................................................................. 8 1.2.4 Therapeutic potential ..................................................................................... 9 1.2.5 Carnosine and cancer ................................................................................... 9 1.3 Objective of the study.......................................................................................... 10 2 Publication .......................................................................12 2.1 General information ............................................................................................. 12 2.2 Carnosine selectively inhibits migration of IDH-wildtype glioblastoma cells in a co-culture model with fibroblasts ........................................................................... 13 3 Summary ..........................................................................23 4 References .......................................................................27 5 Appendix ..........................................................................34 5.1 Supplemental material ........................................................................................ 34 5.2 Author’s contribution............................................................................................ 36 5.3 Erklärung über die eigenständige Abfassung der Arbeit ............................... 37 5.4 Curriculum vitae ................................................................................................... 38 5.5 List of publications ............................................................................................... 39 5.6 Acknowledgements ............................................................................................. 41

As one of several imidazole-containing dipeptides, carnosine is found primarily in the skeletal muscle, the brain, the olfactory bulb and the kidneys of mammals, fishes and birds. The enzyme Carnosine Synthase 1 regulates its synthesis and the two enzymes responsible for the dipeptide’s cleavage into its constituent amino acids are known as serum carnosinase (CN1) and tissue carnosinase (CN2). The amino acid L-histidine is supposed to be mainly responsible for the dipeptides physiological properties based on its imidazole moiety. Among the physiological properties ascribed to the dipeptide are its ability to scavenge reactive oxygen species and to protect against advanced glycation end products and lipid peroxidation. Furthermore, the biogenic dipeptide regulates intracellular calcium homeostasis, acts as a pH buffer and as a metal ion chelator. Based on these primary functions, the dipeptide supports mitochondrial activity and diminishes proteotoxicity. Current studies mainly consider these benefits in muscle tissue and refer to cardiovascular and neurodegenerative diseases. In 1986, Nagai and Suda first revealed tumor growth inhibition after using carnosine in a sarcoma mouse model. Later, Holliday and McFarland confirmed these observations in HeLa cells in vitro. Afterwards, Renner et al. demonstrated an anti- proliferative effect of carnosine on human glioblastoma cells. Unfortunately, the dipeptide’s exact molecular mechanisms on tumor cells are still not entirely understood. Another unresolved question is, whether the dipeptide itself is required for the anti- neoplastic effect or whether L-histidine with its imidazole moiety is sufficient and has to be released from carnosine by cleavage of the dipeptide. In order to get a better insight into these questions we investigated the response of glioblastoma cells to L-histidine and carnosine in primary cell cultures and cell lines derived from glioblastoma. Glioblastoma multiforme represents the most common and malignant primary brain tumor. Significant risk factors are still unknown. At diagnosis, the median age is 64 years and the disease is usually found in a progressed stage. Histopathologically, glioblastoma is characterized by necrosis and pronounced mitotic activity in slightly differentiated cells. Accordingly, the tumor shows rapid progression, aggressive invasiveness and, morphological variety. Since 2005, standard of care against glioblastoma follows the STUPP-protocol, which comprises microsurgery, adjuvant chemotherapy with temozolomide and radiotherapy. Nevertheless, it remains one of the most treatment-refractory intracranial tumors; the median over survival after standard treatment is only 14.6 months. Experiments by Letzien et al. demonstrated that L-histidine mimics the anti-neoplastic effect of carnosine in three glioblastoma cell lines investigated. In addition, the amino acid also increased expression of pyruvate dehydrogenase kinase 4 (PDK4) mRNA expression. These observations pointed towards the possibility that carnosine could just be a vehicle, delivering L-histidine to target cells, and that release of the imidazole-containing amino acid is required for the observed effects. In order to investigate whether the effects observed in cell lines are of general significance, cells from ten glioblastoma cell lines and 21 primary glioblastoma cell cultures derived from surgically removed tumors were incubated in a medium containing different concentrations of either carnosine or L-histidine. Cell viability assays measuring the amount of ATP in cell lysates and dehydrogenase activity in living cells were performed. Both substances induced a significant loss of viability. In fact, L- histidine appeared to be even more effective than carnosine, at the same concentration. Next, we investigated whether the enzymes known to be able to cleave carnosine into amino acids are expressed in the cell cultures. Using RT-qPCR, the expression of the mRNA encoding the two enzymes serum carnosinase (CN1, extracellular) and cytosolic or tissue carnosinase (CN2, intracellular) were analyzed in all 31 glioblastoma cell cultures.The experiments revealed high expression of mRNA encoding CN2 in all cultures, whereas expression of CN1 mRNA (gene: CNDP1) was only slightly detectable. Immunoblot performed with ten cell lines revealed that CN2 protein was also present in all cell lines investigated. Therefore, it had to be assumed, that carnosine may be cleaved inside the cells. In a next series of experiments, we investigated whether inhibition of CN2 by the dipeptidase-inhibitor bestatin (ubenimex) does attenuate the effect of carnosine on tumor cell proliferation. Therefore, cell viability was analyzed in the presence of carnosine and in the absence or presence of different concentrations of bestatin. Aside from a general effect of bestatin on cell viability, especially at higher concentrations, no attenuation of carnosine’s antineoplastic effect was observed in the two cell lines investigated. Therefore, we concluded that cleavage of the dipeptide does not seem to be a prerequisite for its effect on tumor cell viability. As we could not rule out that other unknown dipeptidases aside from CN2 may cleave carnosine, we finally measured the intracellular abundances of cells incubated in the absence or presence of carnosine. Therefore, cells from ten cell lines and from five primary cultures were incubated in the absence and presence of either L-histidine or carnosine, and their extracts were subjected to high performance liquid chromatography (HPLC-MS) after derivatization. Although the intracellular abundances of L-histidine of cells incubated in the presence of carnosine clearly demonstrated that the dipeptide is cleaved inside the cells, no correlation between the intracellular amount of L-histidine and the response of cells with regard to viability was observed. Furthermore, the abundance of L-histidine in cells incubated in the presence of 50 mM carnosine was considerably lower, compared to that of cells incubated in the presence of 25 mM L-histidine. As both conditions resulted in a comparable loss of viability, this strongly indicates that cleavage of the dipeptide is not required for its anti-tumor effect and may even be not very efficient. In conclusion, we could confirm that cleavage of carnosine does occur in glioblastoma cells, although this does not raise the intracellular abundance of L-histidine when compared to cells incubated in the presence of the free amino acid. More importantly, cleavage is not required in order to deploy carnosine’s antineoplastic effect. In addition, it appears to be very likely that the imidazole-moiety whether bound or not bound to another amino acid may be sufficient for a therapeutic response. These observations raise a number of interesting questions that should be investigated considering exploiting the antineoplastic effect described for a potential therapeutic use. First of all, the simple question has to be asked, whether it would be sufficient to use L- histidine as an antitumor drug. In that case one has to ask whether sufficient concentrations of L-histidine could be achieved at the side of the tumor when the amino acid is supplemented. Given the fact that it is a proteinogenic amino acid one may suggest, that it is rapidly taken up by other cells. On the other hand, this may also be the case for carnosine. In addition, carnosine is rapidly cleaved by the presence of CN1 in serum. Whether this is in fact a problem is difficult to answer as there are different reports of small clinical trials where carnosine was able to attenuate cognitive impairments after oral supplementation. In addition, the recently identified CN1 inhibitor carnostatine could possibly be supplemented together with carnosine. Another consideration would be to identify other imidazole containing compounds that are no substrate of CN1. However, as it appears that the imidazole-moiety needs to enter the cells the question is, whether other compounds could be transported across the cell membrane. With regard to treatment of brain tumors one should also keep in mind that, aside from the fact, that the blood-brain-barrier is impaired in glioblastoma, it may still be limiting sufficient delivery. At this point, it is also interesting to note that no side effects of carnosine aside from a rarely appearing dysesthesia, are known. However, given the fact that the outcome of current treatment of glioblastoma is still disappointing it appears to be worth to further investigate carnosine’s antineoplastic effect. As the primary targets of the dipeptide are also still widely unknown, the observation that the imidazole moiety is the main effector may help to further elucidate the mechanisms responsible for the antineoplastic effect. At this point it is also interesting to note that the recently discovered benzimidazolinum Gboxin, which also contains an imidazole moiety, exhibits antitumor activity in glioblastoma cells, most likely by irreversibly compromising oxygen consumption. In this case, an elevated proton gradient and a lower pH in cancer cell mitochondria appear to be responsible for the inhibition of oxidative phosphorylation.:1. List of Abbreviations 3 2. Introduction 5 2.1. Glioblastoma 5 Risk factors 5 Localization and histopathology of glioblastoma 5 Molecular pathology 6 Clinic 6 Prognosis and treatment 6 2.2. Carnosine 7 Occurrence 7 Enzymes and transporters 7 Functions 9 Carnosine and cancer 9 Carnosine and its possible application for therapy 10 2.3. Histidine and other histidine-containing compounds 11 L-histidine and naturally occurring dipeptides 11 Physiological functions of L-histidine 11 L-histidine in health and disease 12 L-histidine as a precursor of other metabolites 12 2.4. Objectives of the study 14 3. Publication 15 3.1. General informations 15 3.2. Carnosine’s inhibitory effect on glioblastoma cell growth is independent of its cleavage 16 3.3. Supplemental materials 29 4. Summary 35 5. References 39 6. Appendix 47 6.1. Declaration of independent work 47 6.2. Statement of the own contribution 48 6.3. Acknowledgements 51