An approach to molecular genetics of thyroid cancer: from novel mutations to a zebrafish model

ANA INÊS LOURENÇO DE ALMEIDA AN APPROACH TO MOLECULAR GENETICS OF THYROID CANCER: FROM NOVEL MUTATIONS TO A ZEBRAFISH MODEL Tese de candidatura ao grau de Doutor em Patologia e Genética Molecular submetida ao Instituto de Ciências Biomédicas Abel Salazar da Universidade do Porto Orientador – Doutora Ana Paula Soares Dias Ferreira Categoria – Professora Auxiliar, Faculdade de Medicina da Universidade do Porto e Coordenadora do Grupo Cancer Signaling & Metabolism, IPATIMUP/Instituto de Investigação e Inovação em Saúde Afiliação – Faculdade de Medicina da Universidade do Porto e IPATIMUP/Instituto de Investigação e Inovação em Saúde Coorientador – Miguel Godinho Ferreira Categoria – Coordenador do Grupo Telomere and Genome Stability, Instituto Gulbenkian de Ciência Afiliação – Instituto Gulbenkian de Ciência Financial Support The candidate was supported by a PhD fellowship (SFRH/BD/79135/2011) from Fundação para a Ciência e Tecnologia (FCT). There are many people in my life And then there’s you, To my Dad, to my Mum Acknowledgments Thank you to every one of you that came along on my five-year journey. Believe me, I haven’t forgotten each one of you. Yes, you! Thank you Paula. Thank you Miguel. Thank you Professor Sobrinho Simões. A BIG THANK YOU to my Dad and my Mum. Prefácio Eu acredito que cada pessoa tem a sua própria filosofia. Na minha filosofia há uma grande disposição para absorver a experiência da vida e com ela apreciar o mundo. Grande parte da minha satisfação pessoal provém das minhas opções profissionais. Há cinco anos atrás lancei um desafio. Candidatei-me a um programa doutoral. E tracei o meu caminho. Ao longo dele encontrei fraquezas, encontrei esperanças. E fui guardando tudo o que vi, tudo o que fiz. A minha tese reflecte o meu percurso destes cinco anos. Comecei o meu projecto de doutoramento no grupo Cancer Biology no IPATIMUP, o qual reportou uma elevada prevalência de mutações BRAF em carcinomas papilares da tiróide esporádicos e linhas celulares derivadas destes carcinomas. A procura de factores adicionais que explicassem a tumorigénese do cancro da tiróide potencialmente relacionadas com as mutações do BRAF e a descoberta de mutações no promotor da telomerase permitiram ao grupo explorar estas últimas mutações nas várias séries existentes no banco de tumores. Entretanto, enquanto outros elementos do grupo procuravam perceber mecanisticamente o efeito das mutações no promotor da telomerase, eu foquei-me no desenvolvimento de um modelo animal que permitisse inicialmente compreender o efeito de alterações frequentemente encontradas no cancro da tiróide (mutações nos genes BRAF e p53) e futuramente estudar factores adicionais que agora se sabem terem um papel relevante na tumorigénese tais como as mutações no promotor da telomerase. Por esta altura foi estabelecida uma colaboração com o grupo Telomeres and Genome Stability no IGC passando de uma ciência translacional para uma ciência básica. O desenvolvimento de um modelo para o estudo do efeito das mutações nos genes BRAF e p53 em peixe-zebra foi o fruto dessa colaboração. Deixarei um legado de ferramentas que muitos poderão usufruir. Publications Ao abrigo do disposto do nº 2, alínea a) do artigo 31º do Decreto-Lei n.º 115/2013 de 7 de Agosto fazem parte integrante desta tese de doutoramento os seguintes trabalhos já publicados ou submetidos para publicação: Artigo I - Vinagre J, Almeida A, Pópulo H, Batista R, Lyra J, Pinto V, Coelho R, Celestino R, Prazeres H, Lima L, Melo M, da Rocha AG, Preto A, Castro P, Castro L, Pardal F, Lopes JM, Santos LL, Reis RM, Cameselle-Teijeiro J, Sobrinho-Simões M, Lima J, Máximo V, Soares P. 2013. Frequency of TERT promoter mutations in human cancers. Nat Commun. 4:2185. Artigo II – Almeida A, Sobrinho-Simões M, Ferreira MG, Soares P. (Submetido). O seguinte capítulo de livro não faz parte do corpo principal de resultados desta tese, mas é parte integrante da mesma, tendo sido utilizado na sua Introdução e Discussão. Appendix I - Almeida AL, Boaventura P, Soares P, Clinical Management of Thyroid Cancer: Etiopathogenic factors of thyroid cancer, Pages 46-62, Future Medicine. 2013 i Table of contents Abbreviations iv Abstract ix Resumo xi Chapter I – Introduction 1 I.1 Thyroid gland 2 I.1.1 Thyroid physiology 2 I.1.2 Thyroid disorders 3 I.1.2.1 Goiter 4 I.1.2.2 Neoplasias 5 Papillary thyroid cancer 7 Molecular genetics of papillary thyroid cancer 9 Signaling pathways altered in papillary thyroid carcinomas 19 I.2 Zebrafish as a model system 23 I.2.1 Zebrafish in Cancer Research 23 I.2.2 Studies on zebrafish thyroid physiology and function 25 I.3 Aims 28 Chapter II – Material & Methods 30 II.1 Plasmid cloning 31 II.1.1 p5E-tg promoter plasmid 31 II.1.2 pME - mCherry and p3E - polyA plasmids 31 II.1.3 pME - mCherry - T2A - BRAF WT and pME - mCherry - T2A - BRAF V600E plasmids 32 II.1.4 pTol2A2-tg:mCherry–pA and pTol2CG2-tg:mCherry–pA plasmids 32 II.1.5 pTol2CG2-tg:mCherry-T2ABRAF V600E –pA plasmids 33 II.1.6 pTol2CG2-tg:loxP-CFP-loxP-mCherry–pA and pTol2CG2tg:loxP-CFP-loxP-mCherry-T2A-BRAF V600E –pA plasmids 34 II.1.7 pCS2-CMV:mCherry-pA plasmid 36 ii II.1.8 pCS2-CMV:mCherry-T2A-BRAF WT -pA and pCS2CMV:mCherry-T2A-BRAF V600E -pA plasmids 36 II.1.9 Summary of the plasmids generated 36 II.2 Cloning - auxiliary techniques 38 II .2.1 PCR 38 II.2.2 DNA sequencing 38 II.2.3 DNA quantification 38 II.2.4 Restriction digestion and ligation 39 II.2.5 Isolation of DNA by agarose gel electrophoresis 39 II.2.6 Plasmid transformation in competent E. Coli 39 II.2.7 Plasmid growth in solid and liquid cultures 40 II.2.8 Plasmid purification 40 II.3 Capped mRNA synthesis 40 II.3.1 Capped transposase mRNA synthesis 40 II.3.2 Capped mCherry mRNA, mCherry - T2A - BRAF WT mRNA and mCherry-T2A-BRAF V600E mRNA synthesis 41 II.4 Microinjections 41 II.4.1 DNA plasmid microinjections 41 II.4.2 Capped mRNA microinjections 42 II.5 Transgenesis 42 II.5.1 WT lines 43 II.5.2 tp53 M214K lines 44 II.6 Fish strains and husbandry 45 II.7 Screening and Imaging 46 II.8 Fin clip and gDNA extraction 46 II.9 Genotyping 46 II.10 Histopathology 47 II.11 Measurement of standard length in larvae 47 II.12 Measurement of body mass index (BMI) 48 II.13 Measurement of thyroid volume in adult fish 48 II.14 Preparation of embryo lysates 49 II.15 Dissection of thyroid tissue in adult zebrafish 49 II.16 Preparation of tissue lysates 49 II.17 Immunoblotting 50 II.18 Heatshock and drug treatment 51 II.19 Statistical analysis 51 ix Abstract Thyroid diseases are extremely frequent and are most often of benign nature. Thyroid cancer is the most common endocrine malignancy in humans and the majority of tumors harbor genetic alterations such as the BRAF mutation. Most of this mutation is an activating mutation in the kinase domain of the BRAF. In sporadic papillary thyroid carcinomas (PTCs), BRAF gene mutations are found in 29%-83% of all cases and almost never co-exist with RAS mutations or RET (RET/PTC) and NTRK1 rearrangements. As a result of BRAF mutation, the MAPK pathway is activated and cellular processes such as proliferation, survival, motility and invasion are promoted. Recently, mutations in the telomerase reverse transcriptase (TERT) promoter have been described in thyroid cancer and considered one of the possible mechanisms that underlies TERT reexpression in several types of human tumors including those of the thyroid. In the first part of the thesis, I describe the study regarding TERT promoter mutations in which I was deeply involved. Our study highlighted the presence of recurrent somatic mutations in the TERT promoter in cancers of the central nervous system (43%), bladder (59%), thyroid (follicular cell-derived, 10%) and skin (melanoma, 29%). Concerning thyroid cancer, the presence of TERT promoter mutations was found to be significantly associated with higher TERT mRNA expression and with older age of the patients. We concluded that TERT promoter mutations are relatively frequent in several specific types of human cancer and that such mutations may enhance expression of telomerase. Over the years, thyroid cancer has been studied using mice models. These models have provided evidence showing that thyroid-specific expression of BRAF V600E induced goiter as well as invasive PTCs which progress to poorly differentiated carcinoma closely recapitulating some human thyroid tumor phenotypes. Very successful, mice models of thyroid cancer are being used to explore molecular mechanisms involved in thyroid tumorigenesis. Mice models can also be used to monitor tumors and to perform drug screening in the setting of thyroid cancer but such tasks remain time-consuming. In the second part of the thesis, I developed a thyroid-targeted BRAF V600E - expressing transgenic zebrafish and evaluated the thyroid tissue phenotypes x during all stages of development up to 12 months of age. I observed that thyroidspecific expression of BRAF V600E induced abnormal thyroid morphogenesis early in life that developed later on into hyperplasia by ~2-3 months of age and colloid goiter by 12 months of age. BRAF V600E –expressing cells disclosed upregulation of proliferation, concomitant with MAPK pathway activation, and there was promotion of apoptosis. High levels of p53 suggested that this protein may be restraining progression to malignancy. Loss of WT p53 using a tp53 M214K zebrafish prevented impairment of thyroid morphogenesis induced by BRAF V600E and surprisingly no evidence of thyroid hyperplasia, goiter and/or neoplasia was detected in those animals up to 12 months of age. tp53 M214K BRAF V600E –expressing cells were low proliferative, consistent with downregulation of the MAPK pathway; suppression of apoptosis was also observed. In conclusion, my work showed that TERT promoter mutations are relatively frequent in specific types of human cancer, including those of the thyroid, and that may enhance telomerase expression. Also, thyroid-specific expression of BRAF V600E induces hyperplasia and colloid goiter in zebrafish and together with the absence of WT p53, BRAF V600E was not able to develop thyroid cancer. This data provides evidence that BRAF activation is sufficient for thyroid cell transformation and that BRAF and p53 pathways must interact genetically in zebrafish thyroid. xi Resumo As doenças na tiróide são extremamente frequentes e são geralmente de natureza benigna. O cancro da tiróide é o tumor endócrino maligno mais comum em humanos e a maioria destes tumores possui alterações genéticas tais como a mutação no gene BRAF. A maioria das mutações do BRAF activam o domínio de cinase da proteína BRAF. Em carcinomas papilares da tiróide (CPT) esporádicos, as mutações no gene BRAF são encontradas em 29% a 83% do total de casos e quase nunca co-existem com as mutações do RAS e rearranjos do RET (RET/PTC) e NTRK1. Como resultado das mutações do BRAF, a via de sinalização das MAP cinases é activada e processos celulares, tais como proliferação, sobrevivência, motilidade e invasão, são promovidos. Recentemente, foram descritas mutações no promotor da telomerase transcriptase reversa (TERT) e estas são consideradas um dos possível mecanismos de reexpressão da telomerase em vários tipos de cancro humano incluindo os da tiróide. Na primeira parte da tese eu descrevo o estudo relativo às mutações no promotor da TERT no qual eu estive envolvida. O nosso estudo realçou a presença de mutações somáticas recorrentes no promotor da TERT em tumores como os do sistema nervoso central (43%), da bexiga (59%), da tiróide com origem nas células foliculares (10%) e da pele (melanoma) (29%). Relativamente ao cancro da tiróide, foi encontrada uma associação significativa entre a presença de mutações no promotor da TERT e níveis elevados de expressão de mRNA e também uma associação com pacientes mais velhos. Concluímos que as mutações no promotor da TERT são relativamente frequentes em determinados tipos de cancro humanos e que estas mutações podem aumentar a expressão da TERT. Ao longo dos anos, o cancro da tiróide tem sido estudado usando modelos de ratinho. Estes modelos deram evidências que demonstram que a expressão específica de BRAF V600E induziu bócio bem como CPT invasivos que progrediram para carcinomas pouco diferenciados recapitulando alguns dos fenótipos dos tumores de tiróide humanos. Os modelos de ratinho ainda são usados para explorar mecanismos moleculares envolvidos na tumorigénese da tiróide. Estes modelos podem também ser usados para monitorizar tumores e para realizar ensaios de drogas no contexto do cancro da tiróide mas estas tarefas são morosas. xii Na segunda parte do tese eu desenvolvi uma linha transgénica com expressão específica de BRAF V600E na tiróide de peixe-zebra e avaliei os fenótipos na tiróide durante todos os estádios de desenvolvimento e até aos doze meses de idade. Observei que a expressão específica de BRAF V600E na tiróide induziu uma morfogénese anormal deste tecido em estádios iniciais que se desenvolvem em hiperplasia aos 2-3 meses de idade e bócio colóide ao fim de 12 meses. Células da tiróide que expressavam BRAF V600E tinham uma sobreregulação da proliferação, concomitante com a activação da via de sinalização das MAP cinases, e foi observada indução da apoptose. Nível elevados de p53 sugerem que esta proteína pode ter contido a progressão para malignidade. Perda da proteína selvagem de p53, usando uma linha homozigota para a mutação M214K do tp53, preveniu morfogénese anormal da tiróide induzida pelo BRAF V600E e surpreendentemente não foram encontradas evidências de hiperplasia, bócio e/ou carcinomas em peixes até doze meses de idade. Células da tiróide que expressavam BRAF V600E eram pouco proliferativas, consistente com desregulação da via de sinalização das MAP cinases, e foi observada supressão da apoptose. Em conclusão, o meu trabalho demonstrou que as mutações no promotor da TERT são relativamente frequentes em determinados tipos de cancro humanos, incluindo os da tiróide, e podem aumentar a expressão da TERT. A expressão específica de BRAF V600E induziu hiperplasia e bócio colóide em peixe-zebra e que em conjunto com a ausência da proteína p53, o BRAF V600E não foi capaz de induzir cancro na tiróide. Estas observações demonstram evidências de que a activação do BRAF é suficiente para a transformação de células da tiróide e que as vias de sinalização da qual fazem parte o BRAF e o p53 devem interagir geneticamente na tiróide do peixe-zebra. xiii 1 Chapter I Introduction 2 THE HUMAN ENDOCRINE SYSTEM comprises part of the body’s communication system connecting the brain to the organs which in turn control the metabolism, growth and reproduction. Tight control of the system is possible through complex feedback mechanisms that maintains homeostasis. Any disruption to an endocrine gland or to the feedback mechanisms may result in endocrine disturbance. Ultimately, cancer may either contribute or be the outcome of such disturbance. I.1 Thyroid gland The human thyroid gland is a butterfly-shaped gland located on the trachea and comprises two lobes connected by an isthmus (VanPutte et al., 2010). The gland is highly vascularized and is one of the largest endocrine glands in the human body (VanPutte et al., 2010). The thyroid gland comprises numerous and varying sized follicles consisting of a thin-layer of cuboidal epithelial cells and a central lumen. The lumen is filled with a homogeneous protein-rich colloid named thyroglobulin which is essential to thyroid hormone (TH) synthesis (Manson et al., 1973; VanPutte et al., 2010). Two hormones are produced in the thyroid gland by the follicular cells in response to thyroid-stimulating hormone (TSH) released from the pituitary: triiodothyronine (T3) and thyroxine (T4). These hormones have an effect on all body systems at all stages of life regulating the basal metabolic rate and tissue growth and maturation (Manson et al., 1973; Kumar et al., 2005; VanPutte et al., 2010). Parafollicular cells secreting calcitonin are also found in clusters surrounding the follicles and in the connective tissue (VanPutte et al., 2010). I.1.1 Thyroid physiology Thyroid follicles are the factory and the storage of THs. The presence of TSH is indispensable for the synthesis and secretion of THs as well as an adequate iodide nutrition. The first step in the synthesis of THs is the uptake of iodide by sodium-iodine symporter (NIS) which is converted to iodine and then condensed onto tyrosine residues from the backbone of thyroglobulin, a protein produced inside the thyroid cells. The newly formed iodothyroglobulin can be either mono-iodinated or diiodinated. When coupled, two di-iodotyrosine molecules result in the formation of 3 T4 whereas a di-iodotyrosine coupled with a mono-iodotyrosine results in T3. Although the T3 is more biologically active than the T4, the production of T3 occurs preferentially outside the thyroid gland by peripherical conversion from T4. THs are stored inside the thyroid follicles composing the majority of the colloid material. T3 and T4 are released by proteolysis from the thyroid to the bloodstream where they bind to TH binding proteins namely the thyroxin binding globulin (TBG) (Figure 1) (Kumar et al., 2005; Brix et al., 2011). Figure 1. TH synthesis. TSH signaling via the TSH receptor controls TH synthesis. NIS at the basolateral membrane takes up iodide from the blood. Iodide is organified in the tyrosyl residues of tg in a reaction catalyzed by thyroid peroxidase (TPO). T 3 , and T 4 are stored in colloid until they are released into the blood. I.1.2 Thyroid disorders Thyroid disorders can range from an enlarged thyroid gland that does not need treatment to thyroid cancer. The most common thyroid problems include goiter and benign thyroid nodules. Also relatively frequent, and clinically more evident, is the abnormal production of TH that can be classified into two groups: hyperthyroidism and hypothyroidism (Manson et al., 1973; VanPutte et al., 2010). 4 In hyperthyroidism, the thyroid gland is overactive producing high levels of thyroid hormones and speeding up the metabolism. Graves’ disease also known as toxic diffuse goiter is the most common cause of hyperthyroidism (Manson et al., 1973; VanPutte et al., 2010). In hypothyroidism, the thyroid gland is underactive producing inadequate levels of thyroid hormones and slowing down the metabolism. Hashimoto’s thyroiditis, congenital hypothyroidism and irradiation are some of the causes of hypothyroidism (Manson et al., 1973; VanPutte et al., 2010). The proper treatment of hyperthyroidism and hypothyroidism depends on the symptoms of the disease and the aetiology. In hyperthyroidism, treatments include thiouracils or thioamides, radioiodine therapy, thyroidectomy, radioactive iodine and/or beta blockers. In hypothyroidism, levothyroxine is a hormone replacement used for treatment (VanPutte et al., 2010). I.1.2.1 Goiter A goiter is an enlarged thyroid gland, it can be either diffuse or nodular and it may extend into the retrosternal space with or without substantial anterior enlargement (Kumar et al., 2005; Lam et al., 2014). A deficiency in iodine intake or in TH synthesis leads to an increased TSH production which in turn sustains increased cellularity and hyperplasia of the thyroid gland as an attempt to normalize the levels of the TH. TH deficiency can be due to defects on hormone synthesis, iodine deficiency and goitrogens (Figure 2). Also, a goiter may appear as a result of the stimulation of the thyroid gland by thyroid stimulating hormone receptor (TshR) agonists such as TSH receptor antibodies, pituitary resistance to thyroid hormone, adenomas of the hypothalamus or pituitary gland and human chorionic gonadotropin-producing tumours (Kumar et al., 2005; Lam et al., 2014). 5 Figure 2. Relation between hypothalamus-pituitary-thyroid (HPT) axis and human goiter development. Iodine deficiency and/or goitrogens disrupt the TH synthesis in many ways. Upon insufficient TH levels, the negative feedback inhibition is lost resulting in increased secretion of tropic hormones (TRH and TSH). High TSH levels stimulate thyroid cells promoting goiter. Also, TSH receptor agonists can ilicitly mimick the biological activity of TSH. Small benign euthyroid goiters do not require any treatment, however their size may be reduced with levothyroxine suppressive therapy. Large and complicated goiters usually require surgical and/or radiation treatment followed by TH replacement (Kumar et al., 2005; Lam et al., 2014). I.1.2.2 Neoplasias Thyroid cancer is the 16 th most common cancer worldwide and accounts for approximately 2% of total of human malignancies with around 298.000 new cases diagnosed in 2012 worldwide (latest data reported) (Nikiforov, 2012; Ferlay et al., 2013). Incidence rates of thyroid cancer are highest in Northern America and other developed countries and lowest in western Africa but this is partly due to data quality and reflects different prevalence of risk factors and screening and 12 mice developed PTC that closely recapitulated the phenotype in humans. Strikingly, treatment of these mice with a MEK inhibitor reduced thyroid size, restored the production of THs and inhibited tumorigenesis (Charles et al., 2011). More evidence of BRAF V600E involvement in mice thyroid tumorigenesis was observed when BRAF V600E expression was induced in follicular thyroid cells in a doxycyclineinducible manner. As early as one week after doxycycline treatment, development of highly penetrant and high-grade PTCs with poorly differentiated features and a reversible activation of the MAPK pathway were observed. Upon doxycycline withdrawal, follicular architecture was reestablished but a second induction not only resulted in hypothyroidism but also reduced thyroid-specific genes expression (Chakravarty et al., 2011). It is reasonable to assume that BRAF V600E is an early event in thyroid tumorigenesis due to a high prevalence of BRAF V600E in papillary microcarcinomas and development of tumors with histological features of human PTCs induced by expression of BRAF V600E in transgenic mice in the absence of any other genetic alterations (Park et al., 2010). Intriguingly, it has been showed that some human PTCs have intratumor heterogeneity of the BRAF genotype as there are two distinct cell populations either with the wild-type (WT) or the BRAF V600E . This may suggest that the clonal occurrence of BRAF mutation is a rare event, occuring only in a subpopulation of cells, and BRAF mutations are rather a late subclonal event in PTCs (Guerra et al., 2012). Also, genome-wide allelotyping and BRAF mutation analysis of foci in multifocal human PTCs showed that BRAF V600E mutation is an early event during clonal evolution in most but not all cases. In fact, BRAF V600E is not always present in all tumor foci which suggests that other genetic factors in the primary tumor clone may have triggered neoplastic transformation (Jovanovic et al., 2008). Nonetheless, in most mice studies BRAF V600E is expressed in all thyroid cells very early in life (fetal or during the first month) and BRAF-induced suppression of thyroid function led to TSH elevation which in turn promoted thyroid tumorigenesis. One particular study generated a model in which BRAF V600E expression was temporally and spatially restricted so that it can recapitulates the human sporadic PTC that usually arises postnatally from follicular cells under physiological serum TSH concentrations. With this approach, thyroid carcinomas under normal TSH levels were not found (Shimamura et al., 2013). This shows that 13 the timing of BRAF activation may be the key to determine cell transformation as the induction of an oncogene in poorly dividing cells, such as the follicular thyroid cells, during adulthood may not trigger tumorigenesis (Shimamura et al., 2013). These data questions whether BRAF V600E initiates thyroid tumorigenesis or BRAF V600E is a consequence of tumor development and not a driver mutation. An alternative scenario is that BRAF V600E does initiate the formation of a PTC however as secondary genetic alterations and/or epigenetic changes take over to maintain tumor sustainability, BRAF V600E is no longer selected and/or important for tumor maintenance (Xing, 2012). RET/PTC and NTRK1 rearrangements The RET gene is a member of the cadherin superfamily and encodes one of the first TK receptors that were found to have a role in human cancer (Phay et al., 2010). RET ligands belong to the glial cell-derived neurotrophic factor family and, when bound to RET co-receptors (GFRα-1), brings together two RET molecules leading to the autophosphorylation of the intracellular tyrosine portion (Manie et al., 2001; Airaksinen et al., 2002; Trovisco et al., 2007). In turn, there is the recruitment and binding of adaptor proteins and subsequent activation of signaling pathways such as the MAPK pathway which are able to control cell proliferation, differentiation, motility and survival (Manie et al., 2001; Airaksinen et al., 2002; Trovisco et al., 2007). NTRK1 gene encodes a member of the neurotrophic TK receptor family. Upon neurotrophin binding, this membrane-bound receptor auto-phosphorylates and activates other members of the MAPK signaling pathway leading to cell differentiation (Teng et al., 2004; Trovisco et al., 2007). Rearrangements of RET and NTRK1 usually involve the fusion with heterologous genes resulting in RET/PTC and NTRK1 chimeric transcripts. The chimeric proteins have an aberrant and persistent activation of their TK domains (Trovisco et al., 2007). Somatic rearrangements of RET gene are found in 3% and up to 60% of sporadic PTCs and lead to a de novo expression of the TK on RET domain in the cytoplasm of follicular thyroid cells (Nikiforov, 2002; Santoro et al., 2002; Frasca et al., 2008). The most common RET rearrangements are the RET/PTC1 and RET/PTC3. RET/PTC1 is by far the most prevalent type comprising 60-70% of all the 14 rearrangements however RET/PTC3 is the most frequent rearrangement found early after radiation exposure. RET/PTC1 and RET/PTC3 are paracentric rearrangements with CCDC6 (coiled-coil domain containing 6) and NCOA4 (nuclear receptor coactivator 4) genes, respectively (Trovisco et al., 2007). Another rearrangement, RET/PTC2, involves reciprocal translocations with the protein kinase, cAMP-dependent, regulatory, type I, alpha (PRKAR1A) gene. RET gene can still be involved in other rearrangements but it is mainly associated with radiation (Trovisco et al., 2007). RET/PTC fusions leave intact the TK domain of the RET receptor enabling the protein to induce activation of signaling cascades including MAPK and phosphatidylinositol-4,5-bisphosphate 3-kinase (PI3K)-V-Akt murine thymoma viral oncogene homolog (AKT) pathways (Kuroda et al., 2003; Knauf et al., 2009). Rearrangements of NTRK1 gene are rare and are found in less than 10% of sporadic PTCs (Trovisco et al., 2007). RAS mutations The RAS gene encodes for a family of related proteins that stay at the center of a cascade of molecular interactions. Most proteins are activated by RAS upon phosphorylation as Ras switches between its “on” and “off” state. Usually, RAS binds to guanosine-5'-diphosphate (GDP) but upon a receptor activation, GDP is expelled allowing guanosine-5'-triphosphate (GTP) to bind. In turn, GTP causes a subtle rearrangement of the RAS protein ultimately leading to the activation of MAPK signaling pathway. As GTP is hydrolyzed to GDP, RAS turns itself “off”, selflimiting its activity (Lodish et al., 2000). RAS mutations lead to a loss of the GTPase activity in RAS protein in such a way that it locks RAS in a constitutively active GTP-bound state which potentiate uncontrolled proliferative signals (Lodish et al., 2000). In PTCs, RAS gene mutations are more frequently found in the follicular variant of PTC (Zhu et al., 2003, Giordano et al., 2005, Frasca et al., 2008). The prevalence of RAS mutations in PTCs ranges from 0% to 16% and neuroblastoma RAS viral (v-ras) oncogene homolog (NRAS) gene is the most predominantly mutated namely on codon 61 (Zhu et al., 2003; Trovisco et al., 2007). PAX8-PPAR gamma rearrangements The paired box 8 (PAX8) gene encodes for a member of the paired box family of transcription factors involved in follicular 15 thyroid cell development and expression of thyroid-specific genes (Kimura, 2011). Peroxisome proliferator-activated receptor gamma (PPARG) gene encodes for a peroxisome proliferator-activated receptor that regulates the expression of target genes involved in cell proliferation, differentiation and immune and inflammatory responses (Kroll et al., 2000). PAX8-PPAR gamma rearrangements are tipically found in follicular thyroid adenomas, FTCs and on the follicular variant of PTCs, in the latter with a prevalence up to 38% (Castro et al., 2006). PTEN and PIK3CA mutations The phosphatase and tensin homolog (PTEN) gene encodes for a phosphatidylinositol-3,4,5-triphosphate 3-phosphatase which regulates dephosphorylation of phosphoinositide substrates thereby negatively regulating the PI3K-AKT signaling pathway (Sun et al., 1999; Hou et al., 2007). PTEN mutations are found in 1-2% of PTCs and mutations in this tumor suppressor gene activate the PI3K-AKT pathway (Hou et al., 2007). Phosphatidylinositol-4,5-bisphosphate 3-kinase, catalytic subunit alpha (PIK3CA) encodes for a catalytic subunit that uses ATP to phosphorylate phosphatidylinositol-4-monophosphate and phosphatidylinositol-4,5diphosphate (Samuels et al., 2004). PIK3CA mutations have also a very low prevalence in PTCs (1-3%) and most are found in the helical and kinase domains of the p110α protein producing variants that are independent of the regulatory subunits and are capable of inducing cell proliferation, invasiveness and resistance to apoptosis (Gymnopoulos et al., 2007; Santarpia et al., 2010). PIK3CA copy gain prevalence goes up to 14% (Wang et al., 2007; Hou et al., 2007). Either mutations in the PIK3CA gene or increase in copy number result in the gain of function reflected by high or constitutive activation of the PI3K activity and promotion of tumorigenesis (Santarpia et al., 2010). p53 mutations The TP53 gene encodes for a protein that maintains genome integrity by binding specifically to a DNA consensus sequence to induce growth inhibitory genes or nonspecifically to damaged sites leading to DNA repair or apoptosis (Liu et al., 2001). The levels of p53 in normal cells are very low but upon p53 activation in response to environmental challenges such as cellular stress, p53 protein is accumulated and stabilized (Liu et al., 2001). p53 is capable of arresting the cell cycle at G1, G2 or in S-phase by inducing p21 which in turn blocks the 16 cycling-dependent kinases (CDKs) responsible for checkpoint regulation and progression of the cell cycle (Bai et al., 2006). This allows time to repair damaged DNA or induce cell death (Bai et al., 2006). p53 is also an activator of the MDM2 proto-oncogene, E3 ubiquitin protein ligase (MDM2) gene which negatively autoregulates p53 maintaining low levels of the p53 protein in normal cells (Bai et al., 2006). At variance with other human tumors, TP53 mutations are not frequent in thyroid cancer (only 10%) and most have been documented in anaplastic carcinomas. Indeed, well-differentiated thyroid cancers do not harbor mutations in TP53 suggesting a role on thyroid cancer progression to poorly differentiated and aggressive phenotypes (Morita et al., 2008). Of note, studies on thyroid tumor samples revealed an accumulation of p53 in poorly differentiated and anaplastic forms but also well-differentiated tumors in the absence of any p53 mutation suggesting that p53 inactivation may result from loss of interaction with MDM2 (Soares et al., 1994). In thyroid cancer, TP53 mutations are commonly found at codons 213 and 238 and in anaplastic carcinomas are also found at codons 248 and 273 (Bai et al., 2006). Modeling thyroid cancer in mice have shown that acquired mutations drive tumor progression and in general BRAF V600E is sufficient to initiate PTCs. Already demonstrated is evidence that p53 constrains progression from papillary to anaplastic thyroid carcinoma (Preto et al., 2004). By generating a thyroid-specific Cre recombinase-estrogen receptor (CreER) transgenic mouse and using a Creregulated BRAF V600E and a conditional Trp53, it was found that p53 loss does enable progression to aggressive anaplastic thyroid cancer but additional events may be required for full anaplastic conversion (McFadden et al., 2014). Telomerase promoter mutations Telomerase is a ribonucleoprotein complex that adds telomeres repeats sequences to the ends of telomeric DNA. The protein component has reverse transcriptase (TERT) activity while the RNA component serves as a template for the telomere repeat (Capezzone et al., 2009; Hanahan et al., 2011). Telomerase is active in the vast majority of human cancer cells (80-90%) enabling their replicative immortality (Figure 4) (Hanahan et al., 2011) and sporadic thyroid carcinomas are no exception. In PTCs, telomerase activity measured by TRAP assay 17 ranged from 20% to 87.5% (Capezzone et al., 2009) but the high levels reported in some studies may be due to the presence of lymphocytic infiltration coexisting with the neoplasia (Brousset et al., 1997; Umbricht et al., 1997; Saji et al., 1999). Of note, telomerase activity in normal thyroid tissue is almost absent (Capezzone et al., 2009). Figure 4. Regulation of telomere length by telomerase. (A) In normal somatic cells, telomerase is absent and the telomere repeat sequence (light blue boxes) is lost everytime the cell divides. After many cell divisions, telomeres reach a critically short length triggering senescence and cessation of proliferation. (B) In cancer cells, reexpression of the telomerase expression bypasses senescence and telomere length is compatible with proliferation. Telomere elongation sustains cancer. Telomerase activity is considered the result of clonal selection as telomeres become critically shortened (Skvortzov et al., 2009). It was shown that genetic mechanisms that promote telomerase reactivation in human tumors include TERT alternative splicing, TERT gene amplification and TERT promoter mutations (Figure 5) (Skvortzov et al., 2009). As recently reported two recurrent, non-overlapping somatic mutations on chromosome 5, -124C>T (C228T) and -146C>T (C250T) (where -1 is the base just upstream of the ATG translation start site of the TERT 18 gene), in the TERT promoter are very frequent in sporadic melanoma (Huang et al., 2013, Horn et al., 2013). These mutations were subsequently found in several tumors including thyroid cancer lines and either well-differentiated or poorly and anaplastic thyroid carcinomas (Vinagre et al., 2014). TERT promoter mutations create an eleven-base nucleotide stretch 5’- CCCCTTCCGGG3’ which contains a new consensus binding site GGAA, generating de novo consensus binding motifs for E-twenty-six (erythroblast transformationspecific –ETStranscription factors). It was shown that it increases the transcriptional activity of the telomerase promoter by two-to-six-fold in human cancer cell lines (Huang et al., 2013; Horn et al., 2013; Liu et al., 2013). This is consistent with a mildly enhanced TERT expression with respect to normal tissues (Muzza et al., 2015). One of the transcription factors found to be recruited specifically to the mutant promoter is the multimeric GA-binding protein (GABP) (Bell et al., 2015). TERT promoter mutations are found in 8% to 22% of PTCs, being the -124C>T mutation more prevalent (Liu et al., 2013; Landa et al., 2013; Vinagre et al., 2013; Melo et al., 2014; Liu et al., 2014; Wang et al., 2014; Muzza et al., 2015). Noticeably, 33% of PTCs with distant metastasis harbored TERT promoter mutations (Melo et al., 2014; Xing et al, 2014; Gandolfi et al., 2015). It was proposed that TERT promoter mutations may be more common in cancers derived from terminally differentiated cells which have a low self-renewing capacity. Indeed, follicular thyroid cells have a very low mitotic rate postnatally (proliferative rate lower than 0.1 in adults). Also, well-differentiated PTCs are usually indolent lesions with low rate of growth. A significant rate of TERT promoter mutations in those carcinomas suggests that thyroid cancer cells may benefit from this mechanism to maintain telomerase lengthening (Saad et al., 2006; Killela et al., 2013). Moreover, it was found a significant overrepresentation of TERT promoter mutations in thyroid tumors harboring alterations in BRAF or RAS genes. TERT promoter mutations seemed to be more frequent in BRAF V600E that in BRAF WT -PTCs (Liu et al., 2013; Melo et al., 2014; Xing et al., 2014). In differentiated thyroid carcinomas, TERT promoter mutations are associated with older age at diagnosis, tumor size and male gender and are correlated with a 19 reduced progression free survival and overall survival (Vinagre et al., 2013; Xing et al., 2014; Melo et al., 2014). This comes as no surprise that TERT promoter mutations are associated with older age at diagnosis which is consistent with a progressive shortening of telomeres in follicular cells during lifetime. When these cells acquire genetic alterations, such as the BRAF V600E mutation, and begin to reply to the oncogenic stimuli by inducing proliferation, very short telomeres also trigger telomere dysfunction which may be compensated by telomerase reactivation via TERT promoter mutations (LondoñoVallejo, 2008; Muzza et al., 2015). Furthermore, BRAF mutations leading to MAPK activation can be conceived as inducers of expression of members of the ETS transcription factor family. Having de novo consensus binding sites for ETS factors promoted by TERT promoter mutations, the lifespan of BRAF-driven clones is extended. Further accumulation of additional genetic defects is promoted which in turn allows progression to advanced tumor stages (Pratilas et al., 2009; Huang et al., 2013; Horn et al., 2013; Liu et al., 2013). These hypothesis could explain why well differentiated PTCs harboring BRAF mutations are more likely to harbor TERT promoter mutations than those PTCs without BRAF mutations. It is also consistent with an enrichment of TERT promoter mutations in poorly differentiated thyroid carcinomas and anaplastic thyroid carcinomas which have partially and completely lost differentiation, respectively, and are the most aggressive thyroid carcinomas (Landa et al., 2013; Liu et al., 2013). Of note, TERT promoter mutations must be only one of the several mechanisms that illegitimately activate telomerase in human cancer. Signaling pathways altered in papillary thyroid carcinomas MAPK signaling pathway The MAPK pathway comprises evolutionarily conserved kinase modules that join extracellular signals to the machinery responsible for cell growth, proliferation, differentiation, migration and apoptosis. One of the groups of MAPK characterized in mammals is the extracellular signal-regulated kinase (ERK)1/2 (Dhillon et al., 2007). Constitutive activation of the MAPK signaling pathway is a frequent event in human cancers particularly in PTCs and melanomas (Xing, 2013). 20 For most cancers, constitutive activation of ERK signaling is established by sustained autocrine or paracrine production of activating ligands, overexpression and activating mutations of the TK receptor and activating mutations in RAS and BRAF (Dhillon et al., 2007; Knauf et al., 2009). In thyroid cancers, besides RAS and BRAF mutations, RET-PTC, NTRK1 and ALK mutations also mediate tumorigenesis via the MAPK pathway (Xing, 2013). In this pathway, ligand-mediated activation of TK receptor promote RAS GTPase conversion which recruits and activates RAF kinases to the plasma membrane. ERK1 and ERK2 are activated upon phosphorylation by MEK1 and MEK2 which are themselves activated when phosphorylated by RAF proteins. Activated ERKs phosphorylate cytoplasmic and nuclear targets including kinases, phosphatases, transcription factors and cytoskeletal proteins (Figure 5) (Dhillon et al., 2007). The effect of ERK signaling activation is consistent to the cellular processes that itself regulates. Sustained ERK signaling promotes phosphorylation and stabilization of FBJ murine osteosarcoma viral oncogene homolog (Fos), Jun protooncogene (Jun), V-Myc avian myelocytomatosis viral oncogene homolog (Myc) and ETS-Related (Erg-1) genes and also cyclin D1 thereby promoting cell-cycle entry and can repress genes responsible for inhibiting proliferation (Yamamoto et al., 2006). On the other hand, high levels of ERKs can induce cell-cycle arrest by expression of CDK-inhibitor proteins such as p21 and p27 that must be counteracted by elevated RhO signaling or activation of AKT so that cells continue to proliferate (Dhillon et al., 2007). PI3K-AKT signaling pathway PI3Ks represent a family of kinases that phosphorylate the 3’-hydroxyl group of phosphatidylinositol inositides that are activated by many TK receptors. Class I of PI3Ks consists of heterodimers of regulatory (p85) and catalytic (p110) subunits. p110α and p110β subunits have an important role in tumorigenesis as RAS function is mediated by its interaction with the RAS-binding site present in those subunits. Also, activation of TK receptor by extracellular signals is itself sufficient to activate p110 subunits which in turn phosphorylates phosphatidylinositol (PtdIns)-3,4-P2 producing PtdIns-3,4,5-P3 leading to the recruitment of AKT to the cytosolic membrane. AKT is then phosphorylated and activated by PDK1 resulting in the phosphorylation of downstream effector such as the mammalian target of rapamycin (mTOR) (Figure 21 5). This result in a broad cascade of signaling responsible for cell growth and proliferation, glucose uptake, migration and apoptosis resistance (Saji et al., 2010). Constitutive activation of the PI3K-regulated signaling pathway is relevant in a wide variety of human tumors including thyroid cancer. This is particularly valid in Cowden’s syndrome, that present thyroid carcinomas, which is characterized by mutations in the PTEN gene that encodes a phosphatase that dephosphorylate PtdIns-3,4,5-P3 and thereby negatively regulates the PI3K-AKT pathway. Additionally, RAS mutations, RET/PTC rearrangements and PIK3CA and AKT1 mutations are further evidences that PI3K signaling pathway has a fundamental role in thyroid tumorigenesis (Saji et al., 2010, Xing, 2010). mTOR signaling pathway mTOR is a serine/threonine kinase that belongs to the phosphoinositide 3-kinase (PI3K)-related kinase family and functions as a regulator of cell growth-related processes. mTOR can form two distinct complexes with other proteins, mTOR complex 1 (mTORC1) and complex 2 (mTORC2), that have different upstream inputs and downstream outputs. Regulation of mTOR by growth factors occurs through the PI3K/AKT pathway which is counteracted by PTEN. There is evidence of overactivation of AKT/mTOR pathway in PTCs when compared to other differentiated thyroid carcinomas and correlation with BRAF V600E mutation which could be explained by BRAF-induced phosphorylation of tumor suppressor liver kinase B1 (LKB1) Ser428, a main upstream kinase of AMP-activated protein kinase (Faustino et al., 2012). 28 I.3 Aims Not so long ago cancer was seen as arising owing to the accumulation of mutations in critical genes that altered the normal programme of cell proliferation, differentiation and death. But the impact of such mutations, including those in the BRAF gene, is only part of the cancer paradigm, indeed many other mechanisms mediate tumorigenesis. One of the most novel mechanism is related to TERT promoter mutations. TERT promoter mutations were initially found at high frequency in human melanoma but not in nevi and are believed to illicitly contribute to telomerase reactivation in melanoma as in other cancers. I was involved in a study to investigate whether TERT promoter mutations were present in cancer types other than melanomas: thyroid, kidney, bladder, gastrointestinal stromal tumors (GIST), adrenal medulla and central nervous system (CNS) tumors. The frequency of TERT promoter mutations was assessed in the above-mentioned cancers and contributed to further studies to understand why those mutations are important in some cancers but not in others and how do they relate with additional genetic alterations. Virtually every breakthrough in the study of human diseases has been the direct result of using animal models. Although many alternative animal-replacing methods are also used, living systems remain a necessity in research. In thyroid cancer studies, mice models have provided enough evidence that thyroid-specific expression of BRAF V600E induced goiter and invasive PTCs which progress to poorly differentiated carcinomas that closely recapitulated the human PTC phenotypes. Those mice models are still used nowadays to disclose new mechanisms of thyroid tumorigenesis, however tumor imaging and translational studies namely drug screening are difficult and time-consuming. In the last years, zebrafish has proved to be a good addition to animal models of human cancer not only because they develop cancer spontaneously, after mutagen exposure or through transgenesis but also because tumors that arise resemble those of the humans at the histological, genetic and genomic levels. Taking advantage of this versatile animal model, I developed a novel transgenic line that expressed BRAF V600E specifically in thyroid cells of zebrafish to understand the effect of BRAF V600E in this tissue and validate a model useful to study thyroid disease. Upregulation of p53 was found to play an important role in the BRAF V600E -induced phenotype in that line. Also, human and mice data showed that p53 loss is essential 29 in tumor progression. Therefore, I developed a transgenic line that expressed BRAF V600E specifically in thyroid cells of tp53 M214K zebrafish to understand whether the absence of WT p53 would promote cancer or at least exacerbate the BRAF V600E effect on thyroid. Finally, to closely recapitulate the timing of BRAF activation in sporadic thyroid carcinomas, I developed a transgenic model that allowed temporally restricted expression of BRAF V600E in thyroid cells of zebrafish. 30 Chapter II Material & Methods 31 The material and methods from the “Frequency of TERT promoter mutations in human cancers” work are exclusively described in chapter III.1. The material and methods from the studies in zebrafish are described in detail in this chapter and briefly in chapter III.2 II.1 Plasmid cloning II.1.1 p5Etg promoter plasmid The (-2041; -1) region, being -1 the position immediately upstream of the ATG sequence of the zebrafish thyroglobulin (tg) gene, was amplified by polymerase chain reaction (PCR) from bacterial artificial chromosome (BAC) genomic clone DKEY-97I18 BAC, GenBank CR855311.15 (SourceBioScience) using the following primers: forward primer 5’-CAGCTGGTACTCTAAATGTGAGAAA-3’ and reverse primer 5’-TGTTTAAAAGGGACGATGTAGC-3’. The fragment was initially cloned into a pCR™-BluntII-TOPO® plasmid (Thermo Fisher Scientific Inc.). The insert was then excised using restriction sites from the backbone plasmid and cloned into the KpnI (Thermo Fisher Scientific Inc.) and XhoI (Thermo Fisher Scientific Inc.) restriction sites on the multicloning site (MCS) of a p5E-MCS plasmid, containing attL4 and attR1 recombination sites and provided by the Tol2Kit (Kwan et al., 2007). Ligation product was transformed into One Shot® TOP10 Chemically Competent Cells (Thermo Fisher Scientific Inc.). The full-length construct was sequenced for accuracy. II.1.2 pME-mCherry and p3E-polyA plasmids pME-mCherry is a middle entry (ME) clone containing attL1 and attL2 recombination sites. p3E-polyA is a 3’ entry clone containing simian vacuolating virus 40 (SV40) late polyA signal and attR2 and attL3 recombination sites. Both entry clones were provided by the Tol2Kit (Kwan et al., 2007). pME-mCherry and p3E-polyA plasmids were transformed into One Shot® TOP10 Chemically Competent Cells (Thermo Fisher Scientific Inc.). The full-length constructs were sequenced for accuracy. 32 II.1.3 pME-mCherry-T2A-BRAF WT and pME-mCherry-T2A-BRAF V600E plasmids Full-length human B-type Raf kinase (BRAF) WT and BRAF V600E coding sequences were amplified by PCR from expression vectors previously generated in my lab: pCMVBRAF WT and pCMV-BRAF V600E , respectively (Faustino et al., 2012) using the following primers: forward primer 5’-CCGGCCCTATGGCGGCGCTGAGC-3’ and reverse primer 5’- GTTTCCTGTCCACTGATGATATCG-3’. mCherry coding sequence was amplified by PCR from pME-mCherry (Kwan et al., 2007) using the following primers: forward primer 5’-GGGCCCCCCCTCGAGGGCCGCCACCATGGTG-3’ and reverse primer 5’-GAGAATCCCGGCCCTATGGCGGC-3’. T2A sequence (5’- GAGGGCAGAGGAAGTCTTCTAACATGCGGTGACGTGGAGGAGAATCCCGGCCCT-3’) was synthetized as an oligomer of 54bp and amplified by PCR. All PCR products generated were purified individually and assembled in frame into a previously linearized pME-MCS (containing attL1 and attL2 recombination sites) using the GeneArt® Seamless Assembly Kit (Thermo Fisher Scientific) and according to the following order: (1 st ) mCherry - (2 nd ) T2A – (3 rd ) BRAF WT or BRAF V600E . Ligation product was transformed into One Shot® TOP10 Chemically Competent Cells (Thermo Fisher Scientific Inc.). The full-length constructs were sequenced for accuracy. II.1.4 pTol2A2tg :mCherry–pA and pTol2CG2tg :mCherry–pA plasmids p5E-tg promoter, pME-mCherry and p3E-polyA were assembled into a pDestTol2pA2, containing attR4 and attR3 recombination sites and provided by the Tol2Kit (Kwan et al., 2007), using a Gateway LR Clonase II Enzyme Mix (Thermo Fisher Scientific Inc.) and according to the manufacturer’s instructions. Ligation product was transformed into One Shot® TOP10 Chemically Competent Cells (Thermo Fisher Scientific Inc.). The full-length pTol2A2-tg:mCherry–pA construct was sequenced for accuracy. Same protocol was performed to generate the pTol2CG2-tg:mCherry–pA, however the three entry clones were assembled into a pDestTol2CG2, containing the cmlc2:EGFP cassette and the attR4 and attR3 recombination sites, provided by the Tol2Kit (Kwan et al., 2007) (Figure 1). The full-length pTol2CG2-tg:mCherry–pA construct was sequenced for accuracy. 33 Figure 1. Gateway cloning strategy to generate the pTol2CG2tg :mCherry-pA plasmid. Schematic of the three-part LR recombination reaction used to generate pTol2CG2tg:mCherry-pA. Entry clones (p5E-tg, pME-mCh and p3E-pA) containing attL sites and pDestTol2CG2 destination plasmid containing attLR sites recombined when LR clonase was added to the reaction. II.1.5 pTol2CG2tg :mCherry-T2ABRAF V600E –pA plasmids p5E-tg promoter, pME-mCherry-T2A-BRAF V600E and p3E-polyA were assembled into a pDestTol2CG2, containing the cardiac myosin light chain (cmlc2): enhanced green fluorescent protein (EGFP) cassette and containing attR4 and attR3 recombination sites, provided by the Tol2Kit (Kwan et al., 2007) using a Gateway LR Clonase II Enzyme Mix (Life Technologies) and according to the manufacturer’s instructions (Figure 2). Ligation product was transformed into One Shot® TOP10 Chemically Competent Cells (Thermo Fisher Scientific Inc.). The full-length pTol2CG2-tg:mCherry-T2A-BRAF V600E –pA construct was sequenced for accuracy. 34 Figure 2. Gateway cloning strategy to generate the pTol2CG2tg :mCherry-T2ABRAF V600E -pA plasmid. Schematic of the three-part LR recombination reaction used to generate pTol2CG2-tg:mCherry-T2A-BRAF V600E –pA. Entry clones (p5E-tg, pME-mCh-T2ABRAF V600E and p3E-pA) containing attL sites and pDestTol2CG2 destination plasmid containing attLR sites recombined when LR clonase was present in the reaction. II.1.6 pTol2CG2tg :loxP-CFP-loxP-mCherry–pA and pTol2CG2tg :loxPCFP-loxP-mCherry-T2A-BRAF V600E –pA plasmids pTol2CG2-tg:mCherry-pA was cut with SalI (Thermo Fisher Scientific Inc.), gel purified and dephosphorylated with Shrimp Alkaline Phosphatase (SAP) (Thermo Fisher Scientific Inc.). loxP-cyan fluorescent protein (CFP)-loxP was amplified from a donor plasmid generated in the lab (unpublished plasmid) using the following primers: forward primer 5’- CAAGCTTATCGATACCGCCCGGGATAACTTCGTATAATGTATGCTATACGAAGTTATC TTG-3’ and reverse primer 5’-CTAGAGAAGCTGAGGACAGGGATCCGGCCGGCC-3’. These primers were engineered for recombination deoxyribonucleic acid (DNA) assembly and the PCR product was gel purified. Plasmid and insert were ligated using the Gibson Assembly Cloning Kit (New England Biolabs) and ligation product 35 was transformed into One Shot® TOP10 Chemically Competent Cells (Thermo Fisher Scientific Inc.). The full-length pTol2CG2-tg:loxP-CFP-loxP-mCherry-pA construct was sequenced for accuracy. Same protocol was performed to generate the pTol2CG2-tg:loxP-CFP-loxP-mCherry-T2A-BRAF V600E –pA however pTol2CG2tg:mCherry-T2A-BRAF V600E –pA was used as a backbone. The full-length pTol2CG2tg:loxP-CFP-loxP-mCherry-T2A-BRAF V600E –pA construct was sequenced for accuracy. Figure 3. Cloning strategy to generate conditional constructs. (A) pTol2CG2-tg:loxPCFP-loxP-mCherry–pA and (B) pTol2CG2-tg:loxP-CFP-loxP-mCherry-T2A-BRAF V600E –pA plasmids. The loxP-CFP-loxP cassette was inserted into previously generated constructs. Scissor represent SalI cut. 36 II.1.7 pCS2CMV :mCherry-pA plasmid pTol2-tg:mCherry–pA and pCS2-CMV plasmid were excised with BamI (Thermo Fisher Scientific Inc.). mCherry insert was gel purified and linearized pCS2-CMV was dephosphorylated with SAP (Thermo Fisher Scientific Inc.). Plasmid and insert were ligated using T4 ligase (Thermo Fisher Scientific Inc.) and ligation product was transformed into One Shot® TOP10 Chemically Competent Cells (Thermo Fisher Scientific Inc.) (Figure 3). The full-length pCS2-CMV:mCherry-pA construct was sequenced for accuracy. II.1.8 pCS2CMV :mCherry-T2A-BRAF WT -pA and pCS2CMV :mCherryT2A-BRAF V600E -pA plasmids pME-mCherry-T2A-BRAF WT and pCS2-CMV plasmid were excised with XbaI (Thermo Fisher Scientific Inc.) and XhoI (Thermo Fisher Scientific Inc.). mCherryT2A-BRAF WT insert was gel purified and linearized pCS2-CMV was dephosphorylated with SAP (Thermo Fisher Scientific Inc.). Plasmid and insert were ligated using T4 ligase (Life Technologies) and ligation product was transformed into One Shot® TOP10 Chemically Competent Cells (Thermo Fisher Scientific Inc.). The full-length pCS2-CMV:mCherry-T2A-BRAF WT -pA construct was sequenced for accuracy. Same protocol was performed to generate the pCS2-CMV:mCherry-T2ABRAF V600E -pA however pME-mCherry-T2A-BRAF V600E was used to excise the mCherry-T2A-BRAF V600E insert (Figure 4). The full-length pCS2-CMV:mCherry-T2ABRAF V600E -pA construct was sequenced for accuracy. II.1.9 Summary of the plasmids generated Table 1 summarizes the plasmids generated in this work. 37 Name Insert Size 5’ entry clones (kan amycin resistant) p5E-tg promoter 2.0 kb tg prom oter 4843 Middle entry clones ( kan amycin resistant) pME - mCherry mCh 3261 pME - mCherry - T2A - BRAF WT mCh and BRAF WT separated by T2A sequence 5818 pME - mCherry - T2A - BRAF V600E mCh and BRAF V600E separated by T2A sequence 5818 3’ entry clone ( kan amycin resistant) p3E - polyA SV40 late polyA signal 2838 Destination vector (amp icillin resistant) pTol2-tg:mCherry-pA tg promoter and mCh reporter* † 7385 pTol2CG2 - tg:mCherry-pA tg promoter and mCh reporter * † 9298 pTol2CG2 - tg:mCherry-T2ABRAF V600E -pA tg promoter, mCh reporter and BRAF V600E separated by T2A sequence* † 11855 pTol2CG2-tg:loxPCFP-loxP-mCherry–pA tg promoter, CFP flanked by lox P inverted sequences and mCh reporter* † 10604 pTol2CG2-tg:loxPCFP-loxP-mCherryT2A-BRAF V600E –pA tg promoter, CFP flanked by lox P inverted sequences and mCh reporter and BRAF V600E separated by T2A sequence* † 13161 Plasmids for in vitro RNA transcription pCS2-CMV: mCherrypA CMV/SP6 cassett e with mCh reporter * 4934 pCS2-CMV: mCherryT2A-BRAF WT -pA CMV/SP6 cassette with mCh reporter and and BRAF WT separated by T2A sequence* 7203 pCS2-CMV: mCherryT2A-BRAF V600E -pA CMV/SP6 cassette with mCh reporter and and BRAF V600E separated by T2A sequence* 7203 Table 1. List of plasmids generated for transgenesis or in vitro RNA transcription. * with SV 40 polyA; † flanked by Tol2 inverted repeats 44 tg( tg :mCh-T2A-BRAF V600E ; cmlc2 :EGFP) line Embryos from AB strain were injected with pTol2CG2-tg:mCherry-T2A-BRAF V600E -pA plasmid to generate mosaic tg(tg:mCh-T2A-BRAF V600E ;cmlc2:EGFP) fish. Founder fish were crossed with WT fish to generate the tg(tg:mCh-T2A-BRAF V600E ;cmlc2:EGFP) line. This line was selected for mCh and EGFP reporters. tg( tg :mCh ; tg :mCh-T2A-BRAF V600E ; cmlc2 :EGFP) line tg:mCh fish were crossed with tg(tg:mCh-T2A-BRAF V600E ;cmlc2:EGFP) fish to produce the tg(tg:mCh; tg:mChT2A-BRAF V600E ;cmlc2:EGFP) line. This line was selected for mCh and EGFP reporters. tg( hsp70l :mCh-T2A-CreER T2 ; tg :loxP-CFP-loxP-mCh; cmlc2 :EGFP) line tg(hsp70l:mCh-T2A-CreER T2 ) fish were used to provide embryos to inject the pTol2CG2-tg:loxP-CFP-loxP-mCherry–pA plasmid and generate mosaic tg(hsp70l:mCh-T2A-CreER T2 ; tg:loxP-CFP-loxP-mCh; cmlc2:EGFP) fish. Founder fish were crossed with WT fish if they carried the hsp70l:mCh-T2A-CreER T2 allele or tg(hsp70i:mCh-T2A-CreER T2 ) fish if they did not. This line was selected for CFP and EGFP reporters and the hsp70l:mCh-T2A-CreER T2 allele. tg( hsp70l :mCh-T2A-CreER T2 ; tg :loxP-CFP-loxP-mCh T2A-BRAF V600E ; cmlc2 :EGFP) line tg(hsp70l:mCh-T2A-CreER T2 ) fish were used to provide embryos to inject the pTol2CG2-tg:loxP-CFP-loxP-mCherry-T2A-BRAF V600E –pA plasmid and generate mosaic tg(hsp70l:mCh-T2A-CreER T2 ; tg:loxP-CFP-loxP-mCh T2ABRAF V600E ; cmlc2:EGFP) fish. Founder fish were crossed with WT fish if they carried hsp70l:mCh-T2A-CreER T2 allele or tg(hsp70i:mCh-T2A-CreER T2 ) fish if they did not. This line was selected for CFP and EGFP reporters and the hsp70l:mCh-T2A-CreER T2 allele. II.5.2 tp53 M214K lines tp53 M214K tg( tg :mCh) line tp53 M214K fish were crossed with tg(tg:mCh) fish to produce heterozygous tp53 M214K tg(tg:mCh). This line was crossed again with tp53 M214K fish to produce homozygous tp53 M214K tg(tg:mCh) fish. All lines were 45 selected for mCh reporter. tp53 M214K mutation was confirmed by genotyping the fin of selected fish. tp53 M214K tg( tg :mCh; cmlc2 :EGFP) line tp53 M214K fish were crossed with tg(tg:mCh;cmlc2:EGFP) fish to produce heterozygous tp53 M214K tg(tg:mCh;cmlc2:EGFP). This line was crossed again with tp53 M214K fish to produce homozygous tp53 M214K tg(tg:mCh;cmlc2:EGFP) fish. All lines were selected for mCh and EGFP reporters. tp53 M214K mutation was confirmed by genotyping the fin of selected fish. tp53 M214K tg( tg :mCh-T2A-BRAF V600E ; cmlc2 :EGFP) line tp53 M214K fish were crossed with tg(tg:mCh-T2A-BRAF V600E ;cmlc2:EGFP) fish to produce a heterozygous tp53 M214K tg(tg:mCh-T2A-BRAF V600E ;cmlc2:EGFP) line. This line was crossed again with tp53 M214K fish to produce a homozygous tp53 M214K tg(tg:mCh-T2ABRAF V600E ;cmlc2:EGFP) line. All lines were selected for mCh and EGFP reporters. tp53 M214K mutation was confirmed by genotyping the fin of selected fish. tp53 M214K tg( tg :mCh ; tg :mCh-T2A-BRAF V600E ; cmlc2 :EGFP) line tp53 M214K tg(tg:mCh) fish were crossed with tp53 M214K tg(tg:mCh-T2A-BRAF V600E ;cmlc2:EGFP) fish to produce the tp53 M214K tg(tg:mCh; tg:mCh-T2A-BRAF V600E ;cmlc2:EGFP) line. This line was selected for mCh and EGFP reporters. tp53 M214K mutation was confirmed by genotyping the fin of selected fish. II.6 Fish strains and husbandry AB wild-type fish was the zebrafish strain used throughout the work. Embryos and larvae <7dpf/dpi were collected from pair mating or outcrosses and kept in embryo medium (5.0mM NaCl, 0.17mM KCl, 0.33mM CaCl, 0.33mM MgSO 4 , 0.05% methylene blue, pH 7.4) at 28ºC on a 14 hour light/10 hour dark cycle. Zebrafish ≥7dpf/7dpi were kept in a recirculating system at 28ºC on a 14 hour light/10 hour dark cycle according to Westerfield et al., 2000. Fish were kept in standard densities (10-12 fish/liter of water). 46 II.7 Screening and Imaging The transgenic lines generated expressed mCherry, EGFP and/or CFP reporters. Screening of embryos, larvae and adult fish was performed on a Zeiss Stereo LUMAR stereoscope. Images were acquired on the Zeiss Stereo Lumar.V12 stereoscope equipped with a Hamamatsy Orca-ER CCD Camera, controlled with the MicroManager v1.14 software. Bright field (BF) images were acquired using the Lumar filter LP420 and fluorescence images were acquired with Lumar filters for Texas Red®, EGFP and CFP. II.8 Fin clip and gDNA extraction Fish were fasten, prior to anesthesia, in tricaine methane sulfonate (MS222) (168ug/L, Sigma). The fin was clipped with a sharp scalpel at a point not greater than halfway between the tip of the fin and the point where the scales end. Fish were then immediately transferred to a container with fresh system water and monitored until they recovered and regained swimming ability. Fin tissue was placed in an eppendorf tube containing 100μL of 50mM NaOH. The sample was heated at 95ºC for 15 minutes or until the tissue was noticeably friable. The sample was cooled to 4ºC and 1/10 th volume of 1M Tris-HCl, pH 8.0 was added to neutralize the basic solution. A centrifugation to pellet the debris was performed at maximum speed for 10 minutes at room temperature. The supernatant containing the genomic DNA was collected to a new tube and stored at 4ºC or longer at -20ºC. II.9 Genotyping tp53 M214K mutation was genotyped using primers to amplify the loss-of-function (LOF) point mutation on exon 7 of the zebrafish tp53 gene (homozygous fish were selected if T>A was in both alleles; heterozygous fish were selected if T>A was in one allele) as described by Berghmans et al., 2005. The following primers were used: forward primer 5’-CACAAGTGTCCTGTTATCGAT-3’ and reverse primer 5’- CATGGCAAGGCAACTGAACTGT-3’. Conditional lines were also genotyped for the hsp70l:mCh-T2A-CreER T2 allele using primers to amplify the Cre recombinase (Cre) coding sequence. The 47 tg(hsp70l:mCherry-T2A-CreER T2 ) line (Hans et al., 2011) does not have a transgenesis marker and progeny from this line can result in WT or transgenic fish. It was only possible to genotype adults. The following primers were used: forward primer 5’-GCATTTCTGGGGATTGCTTA-3’ and reverse primer 5’- CCCGGCAAAACAGGTAGTTA-3’. II.10 Histopathology 7dpf larvae were euthanized by submersion in ice water (5 parts of ice/1 part of water, 0-4ºC) for at least twenty minutes to ensure death by hypoxia. Larvae ≥8dpf were euthanized by an overdose of MS222 (200mg/L, Sigma) by prolonged immersion. Larvae 7-14dpf were fixed in formaldehyde (Sigma-Aldrich) for one day and larvae ≥15dpf were fixed in formaldehyde (Sigma-Aldrich) for two days and then decalcified in 0.5M ethylenediaminetetraacetic acid (EDTA) for one or two more days before paraffin embedding and sectioning. 3μm longitudinal sections representative of the whole larvae or representative of the adult head were performed. Hematoxylin-Eosin (HE) staining were performed according to standard techniques by the HistopathUnit at Instituto Gulbenkian de Ciência (IGC). Images were acquired on a Leica DM LB2 upright microscope equipped with an IDS color CCD camera using the following objectives: 10x 0.25NA and 40x 0.75NA and using the uEye Cockpit software (Imaging Development Systems GmbH, Germany). II.11 Measurement of standard length in larvae Parichy et al., 2009 proposed standard length (SL), defined by the distance from the snout to the caudal peduncle, as a readout of developmental stage (Parichy et al., 2009). SL was measured at 7dpf in larvae anesthetized in tricaine methane sulfonate (MS222 168ug/L, Sigma). Images were acquired on the Zeiss Stereo LUMAR stereoscope equipped with a Hamamatsy Orca-ER CCD Camera and EGFP and Texas Red® filtersets, controlled with the MicroManager v1.14 software. Measurements were performed using the LAS AF Lite software (Leica Microsystems). 48 II.12 Measurement of body mass index (BMI) Body weight (kg) and body length (m) of adult fish were measured at 3 months post fertilization (mpf) and 12mpf. Fish were fasten prior to anesthesia in MS222 (168ug/L, Sigma). Body length was measured from the head to the end of the caudal fin. Body weight and body length were used to calculate the BMI according to the formula: BMI = (body weight kg)/(body length m 2 ). II.13 Measurement of thyroid volume in adult fish Paraffin blocks were exhaustively sectioned into 3-μm-thick sections covering the whole thyroid of the adult fish as previously described. Thyroid middle sections were determined according to the distance between the first and the last sections displaying thyroid tissue and the ventral aorta position. 5 longitudinal sections of each fish were chosen (middle section, 1/3 and 2/3 of total thyroid sections starting from the middle section) (Figure 5). Images were acquired on a Leica DM LB2 upright microscope equipped with an IDS color CCD camera using the following objectives: 10x 0.25NA and 20x 0.5NA and using the uEye Cockpit software (Imaging Development Systems GmbH, Germany). To measure the follicle area, measurements were performed on ImageJ2 (National Institutes of Health, USA) and for each fish total follicle area was calculated by the sum of all individual areas. To calculate a representative total thyroid volume, total follicle area and thickness were multiplied. Figure 5. Measurement of thyroid volume in adult fish using five 3-μm-thick sections that would be representative of the volume. Pink circles outlined by purple dashes represent thyroid follicles and grey boxes represent 3μm-tissue sections. 49 II.14 Preparation of embryo lysates Embryos from mRNA experiments were manually dechorionated and deyolked at 4hpf/4hpi (n=50) as described by Link et al., 2006. Embryo lysates were prepared by adding 100μL of radioimmunoprecipitation assay buffer (RIPA) Buffer (10 mM Tris-Cl pH 8.0, 1 mM EDTA, 0.5 mM EGTA, 1% Triton X-100, 0.1% sodium deoxycholate, 0.1% SDS, 140 mM NaCl and 1 mM PhenylMethylSulfonyl FluoridePMSF) supplemented with 1x phosphatase and 1x protease inhibitors (ThermoFisher Scientific). Embryos were mashed with a cell pestle for 30 seconds on ice. Protein extracts were incubated for 30 minutes on ice and centrifuged at 13.000rpm for 10 minutes at 4ºC. The supernatant was collected to a new eppendorf tube and stored at -80ºC prior to use. II.15 Dissection of thyroid tissue in adult zebrafish Adult zebrafish was euthanized by an overdose of MS222 (200mg/L, Sigma, MI, USA) by prolonged immersion, rinse in water and then dried on a paper towel. Thyroid tissue was dissected with the help of a Zeiss Stereo LUMAR stereoscope and immediately transferred to an eppendorf placed on dry ice until preparation of tissue lysates. II.16 Preparation of tissue lysates Tissue lysates were prepared by adding 100μL of RIPA Buffer (10 mM Tris-Cl pH 8.0, 1 mM EDTA, 0.5 mM EGTA, 1% Triton X-100, 0.1% sodium deoxycholate, 0.1% SDS, 140 mM NaCl and 1 mM PMSF) supplemented with 1x phosphatase and 1x protease inhibitors (ThermoFisher Scientific). Dissected tissue was mashed with a cell pestle for 45 seconds on ice. Protein extracts were incubated on ice for 30 minutes and centrifuged at 13000rpm for 10 minutes at 4ºC. The supernatant was collected to a new eppendorf tube and stored at -80ºC prior to use. 50 II.17 Immunoblotting Total protein extracts from tissue or embryo lysates were thawed and prepared by adding an equal volume of 2x Laemmli sample buffer (100 mM Tris-Cl pH 6.8, 2% SDS and 20% glycerol supplemented with 1/10 volume of B-mercaptoethanol). Samples were heated at 95ºC for 5 minutes and immediately run in 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS PAGE) gels. After proper resolution, samples were transferred to HyBond-P PVDF membranes (GE Healthcare) previously activated with methanol. Following the transfer, the membranes were stained with Ponceau Red to confirm protein loading. Membranes were blocked with 5% non-fat milk in 1x tris-buffered saline (TBS)-t20 for 60 minutes in slow shaking. Then, membranes were probed overnight at 4ºC with primary antibodies. The following primary antibodies and dilutions were used: Living Colors® DsRed Polyclonal Antibody (dilution 1:1000 in 5% non-fat milk in 1x TBS-t20, Clontech Lab), Raf-B Antibody - C19 sc:166, (dilution 1:500 in 5% nonfat milk in 1x TBS-t20, Santa Cruz Biotechnology), phospho p44/42 MAPK (ERK1/2) (Thr202/Tyr 204) Antibody (dilution 1:1000 in 5% non-fat milk in 1x TBS-t20, Cell Signaling), PCNA Antibody FL-261 (dilution 1:1000 in 5% non-fat milk in 1x TBSt20, Santa Cruz Biotechnology), CDKN2a/p16 (F12) Antibody (dilution 1:1000 in 5% non-fat milk in 1x TBS-t20, Santa Cruz Biotechnology), p53 Antibody (dilution 1:500 in 5% non-fat milk in 1x TBS-t20, Anaspec), Anti-active caspase 3 Antibody (dilution 1:500 in 5% non-fat milk in 1x TBS-t20, Abcam), Anti-pan-Akt (Phospho T308) Antibody (dilution 1:1000 in 5% non-fat milk in 1x TBS-t20, Abcam), Phospho-Akt (Ser 473) (D9E) XP® Antibody (dilution 1:1000 in 5% non-fat milk in 1x TBS-t20, Cell Signaling) and Phospho-p38 MAPKinase (Thr180/Tyr182) Antibody (dilution 1:1000 in 5% non-fat milk in 1x TBS-t20, Cell Signaling). Overnight probing was followed by washes in 1x TBS-t20 and by incubation for 13 hours at room temperature with species-specific horseradish peroxidase (HRP)- conjugated antibodies. The following secondary antibodies and dilutions were used: goat anti-mouse IgG-HRP Antibody (dilution 1:1000 or 1:2000 in 5% nonfat milk in 1x TBS-t20, Santa Cruz Biotechnology) and goat anti-rabbit IgG-HRP Antibody (dilution 1:1000/1:2000 in 5% non-fat milk in 1x TBS-t20, Santa Cruz Biotechnology, USA). Membranes were washed in 1x TBS-t20 and developed manually using the Pierce™ ECL Plus Western Blotting Substrate detection system (ThermoFisher Scientific). Protein expression levels from western blot band densities were quantified on ImageJ2 and normalized for tubulin. 51 II.18 Heatshock and drug treatment 2 ½ mpf fish were heat-shocked in a 37ºC water bath for 30 minutes and immediately a drug treatment was carried out. 4-hydroxyl-tamoxifen (4-OHT) (H7904, Sigma) was dissolved at 10mM in 100% ethanol and stored at -20ºC (protected from light). Subsequent dilutions to prepare working solutions were made immediately before use, in embryo medium, to 2.5μM. Controls were treated with an equivalent amount of 100% ethanol diluted in embryo medium. Fish received treatment at 28ºC for 5 hours in the dark followed by a recovery period in fresh water for 12 hours. II.19 Statistical analysis Statistical analysis was performed using GraphPad Prism version 6.0 (GraphPad Software Inc, San Diego, CA) using T-Student test and Mann-Whitney U test. Results were considered statistically significant when p<0.05 (*p<0.05; **p<0.01; ***p<0.001 and ****p<0.0001) and not statistically significant when p≥0.05. 52 Chapter III Results 53 III.1 Frequency of TERT promoter mutations in human cancers SUPPLEMENTARY INFORMATION FOR JVINAGRE - NCOMMS-13-03296-A Supplementary Figure S1 - The panel illustrates the three mutations identified in this study and their location in chromosome 5 in relation to the TERT ATG translation start site. Supplementary Figure S2 - Bar plot indicating the number of cases with TERT mutations in BRAF-mutated and BRAF-wild-type conventional papillary thyroid carcinomas (cPTC) and melanomas. For both tumour types, the presence of TERT mutations is significantly associated with the presence of BRAF mutations (asterisks, Fisher’s exact test two-sided). Supplementary Table S1 - Genetic alterations in thyroid and melanoma cell lines Origin Cell line BRAF RAS RET GNAQ TERT Thyroid XTC-1 WT WT ND - WT HTH74 V600E WT ND - -124G>A K1 V600E WT ND - -124G>A BCPAP V600E WT ND - -124/-125GG>AA C643 WT HRASG13E ND - -124G>A 8505C V600E WT ND - -146G>A TPC-1 WT WT RET/PTC1 - -124G>A T238 WT TT C634W WT MZ-CRC-1 M918T WT Skin melanoma A375 V600E WT WT -146G>A BLM WT WT WT -146G>A G361 V600E WT WT -138/-139GG>AA Mewo WT WT WT -146G>A Uveal melanoma 92.1 WT WT Q209L WT OMM1 WT WT WT WT OMM2.3 WT WT Q209L WT Mel 202 WT WT Q209L WT Mel 270 WT WT Q209P WT Mel 285 WT WT WT WT *All the cell lines are deposited in the cell line bank from the IPATIMUP and were authenticated using DNA profile analysis, obtained with the PowerPlex 16 system (Promega, Madison, USA), according to ATCC and HSRRB available DNA profiles. Supplementary Table S2 - Genetic alterations in cell lines Origin Derivation Cell line TERT Colorectal Colon adenocarcinoma COLO-205 WT Colon adenocarcinoma Caco-2 WT Colon adenocarcinoma SW 480 WT Colon adenocarcinoma HT29 WT Colon adenocarcinoma HCT-15 WT Colon adenocarcinoma HCT-116 WT Colon adenocarcinoma SW48 WT Colon adenocarcinoma Co-115 WT Gastric Gastric adenocarcinoma AGS WT Gastric adenocarcinoma MKN28 WT Well-differentiated gastric carcinoma NCI-N87 WT Poorly-differentiated gastric carcinoma SNU-638 WT Poorly-differentiated gastric carcinoma MKN-45 WT Poorly-differentiated gastric carcinoma Metastasis KATO III WT Diffuse Gastric Carcinoma IPA220A WT Diffuse Gastric Carcinoma GP202 WT Breast Breast adenocarcinoma MDA-MB-231 -124G>A Breast adenocarcinoma MCF-7AZ WT Breast adenocarcinoma SKBR3 WT Breast adenocarcinoma MCF-7/6 WT Breast adenocarcinoma BT-549 -146G>A Breast adenocarcinoma BT-20 WT Breast adenocarcinoma MDA-MB-468 WT Breast adenocarcinoma SUM149 WT Kidney CCRCC Caki-2 -124G>A CCRCC 786-O -124G>A Leukemia Acute promyelocytic leukemia NB-4 WT Acute promyelocytic leukemia HL-60 WT Peripheral B Lymphocytes Lymphoblastoid cell lines GM20770 WT Lymphoblastoid cell lines GM20890 WT Lymphoblastoid cell lines GM11840 WT Lymphoblastoid cell lines GM19782 B WT Lymphoblastoid cell lines GM20515 WT Lymphoblastoid cell lines GM19777 WT Lymphoblastoid cell lines GM20797 WT Lymphoblastoid cell lines GM10847 B WT Lung Lung Adenocarcinoma A549 WT CNS Glioblastoma SF-767 WT *All the cell lines are deposited in the cell line bank from the IPATIMUP and were authenticated using DNA profile analysis, obtained with the PowerPlex 16 system (Promega, Madison, USA), according to ATCC and HSRRB available DNA profiles. Supplementary Table S3 – Summary of clinico-pathological and genetic data in 56 skin melanoma Case Code Melanoma subtype Age Gender Thickness (mm) BRAF NRAS TERT MEL1 Superficial spreading 74 M 1.20 Pos MEL2 Superficial spreading 36 F 0.95 MEL3 Superficial spreading NA F 0.50 V600E MEL4 Superficial spreading NA M 1.40 -146 G>A MEL5 Superficial spreading 46 F 3.10 MEL6 Nodular 52 F 0.90 MEL7 Nodular 59 M 3.20 V600E MEL8 Lentigo maligna 67 F 0.50 V600E MEL9 Nodular 54 F 16.00 MEL10 Nodular 33 M 19.50 V600E MEL11 Superficial spreading 57 F 3.65 -146 G>A MEL12 Lentigo maligna 79 F 0 V600E -146 G>A MEL13 Acral lentiginous 94 F 2.10 MEL14 Acral lentiginous 50 F 1.00 MEL15 Superficial spreading 49 M 10.00 Pos -124 G>A MEL16 Superficial spreading 55 M 2.30 MEL17 Nodular 82 F 3.50 V600E MEL18 Nodular 35 F 2.50 V600E -124 G>A MEL19 Superficial spreading 58 M 7.70 MEL20 Nodular 42 F 1.70 V600E -124 G>A MEL21 Nodular 61 F 4.40 -124 G>A MEL22 Acral lentiginous 69 M 4.00 MEL23 Superficial spreading 41 M 0 V600E MEL24 Acral lentiginous 72 F 0.70 MEL25 Superficial spreading . M 0.50 MEL26 Nodular 76 M 6.00 MEL27 Superficial spreading 66 M 0 -124 G>A MEL28 Superficial spreading 55 F 1.78 MEL29 Acral lentiginous 67 F 4.00 MEL30 Nodular 63 M 5.00 MEL31 Acral lentiginous 60 M 5.40 V600E Pos -124 G>A MEL32 Superficial spreading NA M 0.50 V600E -146 G>A MEL33 Acral lentiginous 71 M 5.20 MEL34 Superficial spreading 75 F 70.00 V600E -124 G>A MEL35 Nodular 69 F 7.00 V600E -146 G>A MEL36 Acral lentiginous 9 F 3.25 MEL37 Superficial spreading 39 F 2.30 MEL38 Acral lentiginous 68 M 4.50 MEL39 Acral lentiginous NA M 2.70 MEL40 Superficial spreading NA F 1.20 V600E -146 G>A MEL41 Superficial spreading 59 F 3.30 V600E MEL42 Acral lentiginous NA F 4.00 MEL43 Lentigo maligna NA M 5.90 Pos Supplementary Table S4 - Summary of clinico-pathological and genetic data in 118 gliomas Case Code Age Gender Diagnosis (Grade) TERT GLI1 4 F Pilocytic Astrocytoma (I) GLI2 27 M Pilocytic Astrocytoma (I) GLI3 22 M Pilocytic Astrocytoma (I) GLI4 31 F Pilocytic Astrocytoma (I) GLI5 27 M Pilocytic Astrocytoma (I) -124 G>A GLI6 18 M Pilocytic Astrocytoma (I) GLI7 30 F Pilocytic Astrocytoma (I) GLI8 12 F Pilocytic Astrocytoma (I) GLI9 NA NA Pilocytic Astrocytoma (I) GLI10 NA NA Pilocytic Astrocytoma (I) GLI11 NA NA Pilocytic Astrocytoma (I) GLI12 NA NA Pilocytic Astrocytoma (I) GLI13 NA NA Pilocytic Astrocytoma (I) GLI14 18 F Diffuse Astrocytoma (II) GLI15 NA NA Diffuse Astrocytoma (II) -124 G>A GLI16 38 M Diffuse Astrocytoma (II) GLI17 22 M Diffuse Astrocytoma (II) GLI18 50 M Diffuse Astrocytoma (II) GLI19 31 F Diffuse Astrocytoma (II) GLI20 30 M Diffuse Astrocytoma (II) GLI21 59 M Diffuse Astrocytoma (II) GLI22 NA NA Diffuse Astrocytoma (II) GLI23 53 F Diffuse Astrocytoma (II) GLI24 20 M Diffuse Astrocytoma (II) GLI25 69 M Diffuse Astrocytoma (II) -146 G>A GLI26 42 F Diffuse Astrocytoma (II) MEL44 Lentigo maligna NA F 0.50 MEL45 Acral lentiginous 68 F 2.20 V600E MEL46 Nodular 71 F 7.80 MEL47 Superficial spreading 59 M 1.20 MEL48 Superficial spreading NA M 0.35 MEL49 Melanoma nodular 42 M 20.00 V600E -124 G>A MEL50 Superficial spreading 61 F 4.30 V600E MEL51 Nodular 79 M 22.00 Pos -146 G>A MEL52 Lentigo maligna 61 F 3.40 MEL53 Acral lentiginous 75 M 6.10 MEL54 Superficial spreading 49 F 2.90 V600E -146 G>A MEL55 Acral lentiginous NA M 4.40 MEL56 Lentigo maligna 75 M 0.30 GLI27 27 F Diffuse Astrocytoma (II) GLI28 60 F Diffuse Astrocytoma (II) -124 G>A GLI29 NA NA Diffuse Astrocytoma (II) GLI30 NA NA Diffuse Astrocytoma (II) GLI31 NA NA Diffuse Astrocytoma (II) GLI32 NA NA Diffuse Astrocytoma (II) GLI33 NA NA Diffuse Astrocytoma (II) GLI34 72 F Oligodendroglioma (II) -124 G>A GLI35 9 M Oligodendroglioma (II) GLI36 33 M Oligodendroglioma (II) GLI37 42 F Oligodendroglioma (II) -124 G>A GLI38 46 F Oligodendroglioma (II) GLI39 45 F Oligodendroglioma (II) -124 G>A GLI40 43 M Oligodendroglioma (II) GLI41 53 M Oligodendroglioma (II) -146 G>A GLI42 53 F Oligodendroglioma (II) -124 G>A GLI43 54 F Oligodendroglioma (II) -124 G>A GLI44 43 F Oligodendroglioma (II) -124 G>A GLI45 34 M Oligodendroglioma (II) -146 G>A GLI46 40 M Oligodendroglioma (II) GLI47 45 F Oligodendroglioma (II) -124 G>A GLI48 NA NA Oligodendroglioma (II) GLI49 NA NA Oligodendroglioma (II) GLI50 NA NA Oligodendroglioma (II) GLI51 NA NA Oligodendroglioma (II) GLI52 NA NA Oligodendroglioma (II) GLI53 NA NA Oligodendroglioma (II) GLI54 NA NA Oligodendroglioma (II) GLI55 NA NA Oligodendroglioma (II) -124 G>A GLI56 65 M Anaplastic Oligodendroglioma (III) -124 G>A GLI57 47 F Anaplastic Oligodendroglioma (III) -124 G>A GLI58 65 F Anaplastic Oligodendroglioma (III) GLI59 50 M Anaplastic Oligodendroglioma (III) -146 G>A GLI60 36 F Anaplastic Oligodendroglioma (III) -124 G>A GLI61 53 M Anaplastic Oligodendroglioma (III) -146 G>A GLI62 65 M Anaplastic Oligodendroglioma (III) -146 G>A GLI63 47 M Anaplastic Oligodendroglioma (III) -124 G>A GLI64 50 M Anaplastic Oligodendroglioma (III) GLI65 70 F Anaplastic Oligodendroglioma (III) GLI66 40 M Anaplastic Oligodendroglioma (III) -124 G>A GLI67 67 F Anaplastic Oligodendroglioma (III) -124 G>A GLI68 64 F Anaplastic Oligodendroglioma (III) -146 G>A GLI69 44 M Anaplastic Oligodendroglioma (III) GLI70 42 F Anaplastic Oligodendroglioma (III) -124 G>A GLI71 68 F Anaplastic Oligodendroglioma (III) -124 G>A GLI72 NA NA Anaplastic Oligodendroglioma (III) GLI73 NA NA Anaplastic Oligodendroglioma (III) GLI74 NA NA Anaplastic Oligodendroglioma (III) GLI75 NA NA Anaplastic Oligodendroglioma (III) -124 G>A GLI76 NA NA Anaplastic Oligodendroglioma (III) GLI77 NA NA Anaplastic Oligodendroglioma (III) GLI78 NA NA Anaplastic Oligodendroglioma (III) GLI79 NA NA Anaplastic Oligodendroglioma (III) GLI80 56 M Glioblastoma (IV) -146 G>A GLI81 27 F Glioblastoma (IV) GLI82 54 F Glioblastoma (IV) -124 G>A GLI83 46 F Glioblastoma (IV) GLI84 77 M Glioblastoma (IV) -124 G>A GLI85 67 F Glioblastoma (IV) -146 G>A GLI86 50 F Glioblastoma (IV) -146 G>A GLI87 73 M Glioblastoma (IV) -146 G>A GLI88 57 M Glioblastoma (IV) -124 G>A GLI89 28 M Glioblastoma (IV) GLI90 51 F Glioblastoma (IV) -124 G>A GLI91 69 M Glioblastoma (IV) -146 G>A GLI92 53 M Glioblastoma (IV) GLI93 56 F Glioblastoma (IV) -124 G>A GLI94 66 M Glioblastoma (IV) -124 G>A GLI95 68 F Glioblastoma (IV) GLI96 58 F Glioblastoma (IV) GLI97 73 M Glioblastoma (IV) -124 G>A GLI98 61 F Glioblastoma (IV) GLI99 66 F Glioblastoma (IV) -124 G>A GLI100 60 F Glioblastoma (IV) -124 G>A GLI101 54 M Glioblastoma (IV) -124 G>A GLI102 66 F Glioblastoma (IV) -124 G>A GLI103 79 M Glioblastoma (IV) -124 G>A GLI104 31 M Glioblastoma (IV) GLI105 NA NA Glioblastoma (IV) -124 G>A GLI106 NA NA Glioblastoma (IV) GLI107 NA NA Glioblastoma (IV) -124 G>A GLI108 NA NA Glioblastoma (IV) -124 G>A GLI109 NA NA Glioblastoma (IV) -124 G>A GLI110 NA NA Glioblastoma (IV) -124 G>A GLI111 NA NA Glioblastoma (IV) GLI112 NA NA Glioblastoma (IV) GLI113 NA NA Glioblastoma (IV) -124 G>A GLI114 NA NA Glioblastoma (IV) GLI115 NA NA Glioblastoma (IV) -146 G>A GLI116 NA NA Glioblastoma (IV) GLI117 NA NA Glioblastoma (IV) GLI118 NA NA Glioblastoma (IV) Supplementary Table S5 – Comparison of the clinico-pathological features of tumours with and without TERT mutations Tumour type Clinical features All patients TERT wild type TERT mutated p value Thyroid* (n=263) Mean age at diagnosis, yr (±SD) Gender Male Female Mean tumour size, cm 48±17 61 (24%) 189 (76%) 3.5±3.4 46±16 50 (22%) 174 (78%) 3.4±3.4 62±13 11(42%) 15 (58%) 4.6±2.7 <0,0001 0,03 0,008 CNS (n=118) Mean age at diagnosis, yr (±SD) Gender Male Female 47±18 42 (49%) 43 (51%) 38±17 21 (51%) 20 (49%) 57±12 20 (47%) 23 (53%) <0,0001 NS Skin melanoma (n=56) Mean age at diagnosis, yr (±SD) Gender Male Female 60±16 26 (46%) 30 (54%) 60±14 19 (48%) 21 (52%) 63±12 7 (44%) 9 (56%) NS NS Bladder (n=82) Mean age at diagnosis, yr (±SD) Gender Male Female 63±10 66 (80%) 16 (20%) 65±12 24 (70%) 10 (30%) 63±12 42 (88%) 6 (12%) NS NS The numbers in parentheses indicate percentages. In some cases information regarding age, gender or tumour size was missing. *Only follicular cell-derived thyroid carcinomas: FTC, PTC, PDTC and ATC. Supplementary Table S6 – Clinico-pathological characteristics and TERT promoter mutation in the 399 thyroid samples analysed Histotypes No. Of cases Gender ratio (♀:♂) Mean age (years±SE) Tumour size (cm±SE) BRAFV600E mutation TERT n (%) Normal Thyroid 27 3,5:1 47,0 ± 14,8 NA 0 0 (0,0) Lymphocytic Thyroiditis 9 9:0 44,6 ± 15,7 NA 0 0 (0,0) Nodular Goiter 12 1,75:1 56,2 ± 13,4 NA 0 0 (0,0) Follicular Thyroid Adenoma Conventional 52 7:1 43,8 ± 12,1 3,5 ± 1,6 0 0 (0,0) Oncocytic Variant 8 7:0 47,0 ± 14,0 2,7 ± 1,2 0 0 (0,0) Follicular Thyroid Carcinoma Conventional 36 2,5:1 53,7 ± 14,7 4,5 ± 2,8 0 9 (25,0) Oncocytic Variant 28 5,5:1 51,4 ± 20,2 4,6 ± 3,6 0 0 (0,0) Papillary thyroid carcinoma Conventional 110 3,1:1 43,1 ± 16,2 2,6 ± 1,6 61 12 (10,9) Follicular Variant 39 3,9:1 45,8 ±14,7 2,6 ± 1,2 2 0 (0,0) Microcarcinoma 5 5:0 46,0 ±14,4 0,4 ± 0,4 1 0 (0,0) Mucoepidermoid Carcinoma 4 2:0 43,5 ± 33,2 NA 0 1 (25,0) Oncocytic Variant of PTC Conventional Pattern 4 4:0 45,3 ±11,1 2,8 ± 1,8 4 0 (0,0) Follicular Pattern 3 2:1 45,3 ±22,5 5,2 ± 3,9 1 0 (0,0) Warthin-like PTC 4 4:0 55,5 ±11,2 2,0 ± 0,8 4 0 (0,0) PDTC 14 1:1 57,0 ± 17,4 6,7 ± 2,8 1 2 (14,2) Anaplastic 16 1,3:1 62,1 ± 12,8 6,9 ± 3,1 5 2 (12,5) MTC 28 2,3:1 46,4 ± 16,7 3,0 ± 0,9 0 0 (0,0) Supplementary Table S7 - Summary of clinico-pathological and genetic data in 351 tumour and tumour like thyroid lesions Case Code Age Gender Diagnosis RAS BRAF TERT THY1 50 F FTA THY2 50 F FTA THY3 59 M FTA THY4 28 F FTA THY5 45 F FTA THY6 51 F FTA THY7 37 F FTA THY8 53 F FTA THY9 57 F FTA THY10 66 F FTA THY11 63 F FTA THY12 40 F FTA THY13 34 F FTA THY14 65 F FTA THY15 35 F FTA THY16 34 F FTA THY17 18 F FTA THY18 37 F FTA THY19 43 F FTA THY20 53 F FTA THY21 45 F FTA THY22 33 F FTA THY23 34 F FTA THY24 46 M FTA THY25 53 F FTA THY26 65 F FTA THY27 29 F FTA THY28 34 F FTA THY29 34 F FTA THY30 NA NA FTA THY31 NA NA FTA THY32 56 F FTA THY33 33 F FTA THY34 47 M FTA THY35 30 F FTA THY36 55 M FTA THY37 NA F FTA THY308 60 M Anaplastic V600E -124 G>A THY309 72 M Anaplastic THY310 NA NA Anaplastic V600E THY311 74 F Anaplastic V600E THY312 52 F Anaplastic THY313 55 M Anaplastic THY314 NA NA Anaplastic THY315 NA F Anaplastic THY316 56 F Anaplastic THY317 59 F Anaplastic THY318 75 F Anaplastic V600E THY319 82 F Anaplastic THY320 69 M Anaplastic V600E THY321 39 M Anaplastic THY322 70 M Anaplastic THY323 44 F Anaplastic -146 G>A THY324 63 M MTC THY325 NA NA MTC THY326 NA NA MTC THY327 32 F MTC THY328 NA NA MTC THY329 NA NA MTC THY330 NA NA MTC THY331 NA NA MTC THY332 NA NA MTC THY333 NA NA MTC THY334 68 F MTC THY335 NA NA MTC THY336 37 F MTC THY337 NA NA MTC THY338 71 F MTC THY339 40 M MTC THY340 NA NA MTC THY341 56 F MTC THY342 NA NA MTC THY343 38 F MTC THY344 NA NA MTC THY345 NA NA MTC THY346 23 M MTC THY347 NA NA MTC THY348 36 F MTC THY349 NA NA MTC THY350 NA NA MTC THY351 NA NA MTC Supplementary Table S8 – Comparison of the clinico-pathological features of conventional PTC with and without TERT mutations Tumour type Clinical features All patients TERT wild type TERT mutated p value Conventional Papillary Thyroid Carcinoma (n=110) Mean age at diagnosis, yr (±SD) Gender Male Female Mean tumour size, cm LN metastasis (positive/total) BRAF mutation (positive/total) NRAS mutation (positive/total) 43±16 26 (25%) 80 (75%) 2,6±1,6 67 (73%) 61 (57%) 4(7%) 41±15 18 (19%) 76 (81%) 2,4±1,4 56 (68%) 49 (50%) 4 (10%) 63±14 8 (67%) 4 (33%) 4,3±2,2 11 (100%) 12 (100%) 0 (0%) <0,0001 0,0012 0,0004 0,03 0,001 NS Supplementary Table S9 - Summary of clinico-pathological and genetic data in 82 bladder cancers Case Code Age Gender Stage Grade TERT BC1 66 M Ta Low Grade -146 G>A BC2 NA M T1 Low Grade BC3 80 M Ta Low Grade -146 G>A BC4 70 M Ta Low Grade BC5 67 M Ta Low Grade -124 G>A BC6 80 M Ta Low Grade -124 G>A BC7 65 M Ta Low Grade -124 G>A BC8 40 M Ta Low Grade -124 G>A BC9 36 M T1 Low Grade -124 G>A BC10 64 M Ta Low Grade BC11 65 M Ta Low Grade -124 G>A BC12 47 F Ta Low Grade BC13 54 M Ta Low Grade -124 G>A BC14 41 M T1 Low Grade BC15 50 M T1 Low Grade -124 G>A BC16 67 M Ta Low Grade -124 G>A BC17 97 M Ta Low Grade -146 G>A BC18 40 F Ta Low Grade -124 G>A BC19 60 M Ta Low Grade BC20 70 M T1 Low Grade -146 G>A BC21 46 F Ta Low Grade BC22 61 F Ta High Grade -146 G>A BC23 NA M T1 High Grade -124 G>A BC24 59 M T1 High Grade -124 G>A BC25 77 M T1 High Grade -146 G>A BC26 65 M T1 High Grade -124 G>A BC27 83 F T1 High Grade -124 G>A BC28 64 M T1 High Grade BC29 NA M NA High Grade -124 G>A BC30 NA M NA High Grade -146 G>A BC31 67 M T1 High Grade -124 G>A BC32 NA M Ta High Grade BC33 NA M T1 High Grade BC34 NA M T1 High Grade BC35 NA M T1 High Grade -146 G>A BC36 74 M T1 High Grade -124 G>A BC37 75 M Ta High Grade BC38 76 F T1 High Grade BC39 45 M Ta High Grade BC40 NA M T1 High Grade BC41 76 F T1 High Grade BC42 NA M T1 High Grade -124 G>A BC43 72 F Ta High Grade BC44 NA F T1 High Grade BC45 67 M T1 High Grade -124 G>A BC46 NA M T1 High Grade -146 G>A BC47 NA M T1 High Grade -146 G>A BC48 51 M T1 High Grade -146 G>A BC49 50 M Ta High Grade -124 G>A BC50 74 M T1 High Grade -124 G>A BC51 70 F T1 High Grade -124 G>A BC52 50 M T1 High Grade -146 G>A BC53 68 M T1 High Grade -124 G>A BC54 56 M T1 High Grade BC55 53 M T1 High Grade BC56 60 F T1 High Grade BC57 74 M T1 High Grade BC58 73 M T1 High Grade BC59 60 F T1 High Grade BC60 64 F T1 High Grade BC61 74 M T1 High Grade BC62 51 M T1 High Grade BC63 63 M T1 High Grade -124 G>A BC64 57 M Ta High Grade -146 G>A BC65 67 F Ta High Grade -146 G>A BC66 58 M Ta High Grade -124 G>A BC67 66 M T1 High Grade -124 G>A BC68 64 M T1 High Grade BC69 70 M T1 High Grade -146 G>A BC70 68 M Ta High Grade -146 G>A BC71 64 M T1 High Grade BC72 59 F T1 High Grade -124 G>A BC73 63 M T1 High Grade -124 G>A BC74 66 M Ta High Grade BC75 76 M T1 High Grade -124 G>A BC76 71 M T1 High Grade BC77 61 M T1 High Grade -124 G>A BC78 76 M T1 High Grade -124 G>A BC79 61 M T1 High Grade BC80 68 F T1 High Grade BC81 72 M T1 High Grade BC82 71 M T1 High Grade -146 G>A Supplementary Table S10 - Summary of clinico-pathological and genetic data in 26 kidney cancers CCRC – Clear cell renal carcinoma; CromRCC – chromophobe renal cell cancer; PRCC – Papillary renal cell carcinoma; Case code Diagnosis Age Gender Nuclear grade Staging TERT RCC1 CCRCC 68 M RCC2 CCRCC 44 M RCC3 CCRCC 49 M RCC4 CCRCC 49 M 3 pT1a Nx Mx RCC5 CCRCC 57 M 2 pT1a Nx Mx RCC6 CCRCC 48 F 2 pT1a Nx Mx RCC7 CCRCC 48 F 2 pT1a Nx Mx RCC8 CCRCC 68 M 2 pT3a Nx Mx RCC9 CCRCC 50 M 4 pT3b Nx M0 RCC10 CCRCC 50 M 4 pT3b Nx M1 RCC11 CCRCC 62 M 3 pT2 Nx Mx RCC12 CCRCC 62 M 3 pT2 Nx Mx RCC13 CromRCC 59 M RCC14 CromRCC 77 F 2 pT2 Nx Mx RCC15 CromRCC 77 F 2 pT2 Nx Mx RCC16 CromRCC 72 M 4 pT2 Nx Mx RCC17 PRCC NA M RCC18 PRCC 61 M RCC19 PRCC 26 F 2 pT2 N2 MX RCC20 PRCC 78 M 2 pT1b Nx Mx RCC21 PRCC 34 M 1 pT1a Nx Mx RCC22 PRCC 60 M pT1b Nx Mx RCC23 PRCC 60 M pT1b Nx Mx RCC24 PRCC 71 M RCC25 PRCC 71 M RCC26 PRCC 75 M 3 pT1a Nx Mx Supplementary Table S11 - Summary of clinico-pathological and genetic data in 17 phaeochromocytoma Case code Age Gender Tumor size (cm) MEN/NF/VHL TERT PHEO1 51 M 4,5 NA PHEO2 56 M 7 NA PHEO3 54 F 4 NA PHEO4 46 F 7 NA PHEO5 53 F 2,5 NA PHEO6 42 M 8,5 NA PHEO7 20 F 1,5 MEN2 PHEO8 55 F 3,5 NA PHEO9 19 F 7 NA PHEO10 56 M 4,5 NA PHEO11 62 M 6 NA PHEO12 45 M 7 NA PHEO13 73 F 5 NA PHEO14 57 M 3,5 NA PHEO15 61 M 1,7 NA PHEO16 20 F 6 MEN2 PHEO17 46 F 10 NF1 Supplementary Table S12 - Summary of clinico-pathological and genetic data in 36 gastrointestinal stromal tumour Case code Age Sex Tumor Size (cm) KIT PDGFRA TERT GIST1 80 F exon 11 N567K, L576V, 568del575 GIST2 59 M 1,2 exon 11 V560D GIST3 52 M 9,5 exon 11 69997:T>A GIST4 67 M 17 exon 11 W557F delK558 GIST5 58 F 3,5 exon 11 556del557 GIST6 54 M 19 exon 11 V559D GIST7 79 F 7 exon 11 del557-558 GIST8 46 M 5,5 exon 11 del 550-557, ins550-551 GIST9 54 M 3,1 exon 11 552del557, K550Q, P551R GIST10 76 F 6,7 exon 18 D842V GIST11 55 F 6,5 exon 11 del560V GIST12 43 M 14,5 exon 11 564del578 GIST13 73 F 0,8 exon 11 delV559 GIST14 62 M 2,5 exon 11 del_K558 GIST15 38 M 3,5 exon 11 V560E GIST16 88 F 23 exon 11 L576P GIST17 81 F exon 11 L576P GIST18 78 F exon 12 583del586 GIST19 40 F 4 exon 9 502-503dup GIST20 63 F 9 exon 9 502-503dup GIST21 75 M 8 exon 11 Q575H, P577T, delL576 GIST22 59 F exon 11 552del570 GIST23 78 F 24 exon 11 551del553 GIST24 58 M 13,5 exon 11 557del558 GIST25 76 M 14 exon 9 502-503dup GIST26 20 F 2 GIST27 77 F 6 GIST28 74 M 3 GIST29 58 F 3,5 GIST30 72 F 8 GIST31 62 M 6,5 GIST32 73 M 6 GIST33 66 F 14 GIST34 63 F 6,5 GIST35 55 M 4 GIST36 82 F 7 83 III.2 Targeted expression of BRAF V600E in thyroid cells of transgenic zebrafish induces hyperplasia reverted by loss of WT p53 Abstract The BRAF V600E mutation is the most common genetic alteration in papillary thyroid carcinomas (PTCs) and are found in 29%-83% of all cases. PTCs harboring B-type Raf kinase (BRAF) V600E are often invasive and this mutation is also present in more advanced stages of the disease such as in poorly differentiated and anaplastic carcinomas arising from PTCs. BRAF V600E kinase activate the mitogen-activated protein kinases (MAPK) pathway, promoting cellular processes such as proliferation, survival, motility and invasion. To explore the role of BRAF V600E in thyroid cancer pathogenesis, I targeted its expression to thyroid cells of transgenic zebrafish under a thyroid-specific promoter. Fish showed impairment of normal thyroid morphogenesis at early stages of development and they seemed to compensate for BRAF V600E -induced thyroid dysfunction by developing hyperplasia and goiter but not neoplasia. Also, activation of BRAF V600E expression in thyroid cells of transgenic zebrafish during adulthood resulted in a hyperplasic phenotype similar to when BRAF V600E was expressed shortly after birth. In BRAF V600E -expressing animals, upregulation of p53 was observed suggesting a protective mechanism to prevent cancer. Thus, I targeted BRAF V600E expression to thyroid cells of tp53 M214K zebrafish in order to overcome p53 blockage and progression to cancer. Suprisingly, BRAF V600E - expressing tp53 M214K fish developed a normal thyroid until adulthood. In summary, thyroid-specific expression of BRAF V600E induced goiter but not neoplasia which was prevented by loss of WT p53 protein. Introduction To create functional follicles, thyroid cells must aggregate, polarize and establish selective permeability barriers between the lumen and the outer compartments. Cellular adhesion, intracellular trafficking, specialized cell-cell junction assembly 84 and precise regulated morphogenetic cell movements must all be coordinated (Yap et al., 1997). Disturbances in these cellular processes can be mediated by external factors such as goitrogens or internal factors such as hormonal imbalance or genetic alterations. Ultimately, these factors have important implications in thyroid diseases, mostly characterized by disturbed follicular architecture. PTCs are characterized by nonoverlapping genetic alterations in more than 70% of the cases. These alterations include rearrangements in the tyrosine kinase (TK) receptors Rearranged-during-transfection (RET), Neurotrophic Tyrosine Kinase, Receptor, Type 1 (NTRK1), Anaplastic Lymphoma Receptor Tyrosine Kinase (ALK) and A Kinase (PRKA) Anchor Protein 9/B-type Raf kinase (AKAP9/BRAF) and also point mutations in Rat Sarcoma Viral Oncogene Homolog (RAS) and BRAF genes. Ultimately, most cases will lead to an aberrant activation of the RAS-RAF-MEK-ERK kinase pathway (DeLellis et al., 2004; Kumar et al., 2005; Nikiforov, 2012). Up to 90% of the BRAF gene mutations consist in the BRAF V600E (Garnett et al., 2004; Frasca et al., 2008). The initial evidence that BRAF V600E is required for the cell proliferation, transformation and tumorigenicity of follicular thyroid cells was demonstrated in a xenograft mice model (Liu et al., 2007) even though many studies have supported the role of BRAF mutation in tumor initiation. Targeted expression of the BRAF V600E in thyroid cells was induced in transgenic FVB/N mice using a bovine thyroglobulin promoter. This study showed that BRAF induced thyroid dysfunction which was compensated by increased levels of thyroid-stimulating hormone (TSH) and goiter development. Also, multifocal tumors involving both lobes of the thyroid gland with mixed papillary and follicular growth pattern were observed in 12 and 22-week-old mice. PTCs presented the classical architecture, tall-cell features and a high potential for invasiveness. Tumors from one of the transgenic lines progressed into poorly differentiated carcinomas (Knauf et al., 2005). PTCs were also observed in a thyrocyte-specific knock-in of BRAF V600E in mice but with a very short latency and complete penetrance by 3 weeks. When this model was crossed with a thyroid stimulating hormone receptor (TshR) knockout mice to genetically ablate TSH signaling, thyroid growth was reduced and low-grade PTCs were observed but only at 9 weeks of age (Franco et al., 2011). More evidence of BRAF V600E involvement in mice thyroid tumorigenesis was perceived when BRAF V600E expression was induced in follicular thyroid cells in a 91 Figure 3 . tg( tg :mCh) line. (A) 14dpf tg(tg:mCh) larvae revealed (B) mCherry protein expression specifically in thyroid cells. Larvae is oriented with anterior to the left. Scale bar: 250μM. At this point, EGFP reporter expression in the heart was used to select animals from the BRAF V600E -expressing line. By crossing the tg(tg:mCh) with the tg(tg:mCh-T2ABRAF V600E ) line, a tg(tg:mCh;tg:mCh-T2ABRAF V600E ) line was generated in order to have a thyroid-specific reporter in the BRAF V600E -expressing line. Embryos showing mCherry expression in the thyroid given by the Tol2tg:mCh transgene and EGFP expression in the heart given by the Tol2CG2tg:mCh-T2A-BRAF V600E transgene were grown to adulthood. Because the tg(tg:mCh-T2ABRAF V600E ) line was not expressing mCherry, an experiment to determine the protein translation efficiency from mCh-T2A-BRAF V600E was performed in order to exclude possible problems with the 2A cleavage. A pCS2+ expression plasmid was chosen because it contains a strong enhancer/promoter (simian CMV) followed by a polylinker and a SV40 late polyadenylation site. Also, a SP6 promoter is present in the 5’ untranslated region of the mRNA from the sCMV promoter which allows in vitro RNA synthesis of sequences cloned into the polylinker. mCherry coding sequence was used as a control and mCh-T2A-BRAF WT was used as a control of T2A efficiency and BRAF toxicity, because BRAF wt proteins do not have an effect on the development of zebrafish embryos (Anastasaki et al., 2012). Capped mRNA was in vitro transcribed from the constructs generated: pCS2-CMV:mCh, pCS2CMV:mCh-T2A-BRAF WT and pCS2-CMV:mCh-T2A-BRAF V600E (Figure 4A). 92 After injecting capped mCh-T2A-BRAF V600E mRNA in WT embryos, a severe developmental arrest was found at around 4hpi preventing embryos to further develop beyond the blastula stage (Figure 4B). Malformations were found in the majority of injected embryos (90.43%) (Figure 4C) and survival at 24 hours post injection (hpi) was very low (34.8%) (Figure 4D). By fluorescence microscopy, a polka-dot pattern was observed confirming mCherry expression in the cell (Figure 4B). The phenotype produced was believed to be due to expression of mCherry and BRAF V600E because no phenotypes were observed in non-injected WT embryos (Figure 4B). WT embryos injected with capped mCh or mCh-T2ABRAF WT mRNAs showed no developmental arrest (Figure 4B), a very low frequency of malformations (0.76% and 0.64%, respectively) (Figure 4C) and survival at 24hpi (85.8% and 75.5%, respectively) not significantly different from non-injected WT embryos (Figure 4D). By fluorescence microscopy, it was observed an ubiquitous expression of mCherry in WT embryos injected with capped mCh or mCh-T2A-BRAF WT mRNAs (Figure 4B) which was consistent with the observations reported by Anastasaki et al., 2012. In order to confirm if BRAF V600E was indeed inducing a phenotype, a western blot of embryo lysates was performed by probing with antibodies to mCherry and human BRAF. Blots confirmed the presence of both mCherry and BRAF proteins in embryos injected with mCh-T2A-BRAF V600E mRNA and BRAF probing was similar to those injected with the WT form of BRAF (Figure 4E; E’ and E’’). 93 94 95 Figure 4. mRNA experiment in WT embryos. (A) Schematic representation of the pCS2CMV:mCh, pCS2-CMV:mCh-T2A-BRAF WT and pCS2-CMV:mCh-T2A-BRAF V600E constructs (top to bottom) containing the SP6 promoter to drive in vitro RNA synthesis. (B) Representative images of 24hpf non-injected and 24hpi injected WT embryos with one of the three mRNAs: mCh, mCh-T2A-BRAF WT or mCh-T2ABRAF V600E . On the left are shown BF images; on the right images acquired with a Texas Red® filter. Scale bar: 300μm. (C) Percentage of developmental defects at 24hpf or 24hpi (statistical analysis only shown for relevant comparisons; p<0.001). (D) Survival at 24hpf or 24hpi (statistical analysis only shown for relevant comparisons; p<0.001). (E) Western blot analysis of lysates from embryos at 4hpf or 4hpi. 50 embryos were pooled for each condition. Membranes were blotted with mCh and BRAF antibodies. Ponceau S stain showed equal amount of protein in all lanes. (E’) mCherry and (E’’) BRAF protein levels were quantified and normalized to the amount of protein. Values are shown in bar graphs and represent three independent experiments (statistical analysis only shown for relevant comparisons; p<0.001). To rule out a possible activation of p53-mediated apoptosis due to off-targeting effect of the mRNA injected (Robu et al., 2007), an mRNA experiment was conducted in tp53 M214K embryos. tp53 M214K embryos lack apoptosis and cell-cycle arrest responses to deoxyribonucleic acid (DNA) damage because the tp53mediated cell-cycle control is absent (Berghmans et al., 2005). 96 After injecting capped mCh-T2A-BRAF V600E mRNA in tp53 M214K embryos, a lower developmental arrest was found in embryos at around 4hpi (Figure 5A) when compared to tp53 M214K embryos. Malformations were found in injected embryos (39.40%) (Figure 5B) and survival was low (33.00%) (Figure 5C). By fluorescence microscopy, ubiquitous expression of mCherry was observed (Figure 5A). In contrast, tp53 M214K embryos injected with capped mCh or mCh-T2ABRAF WT mRNAs showed no developmental arrest (Figure 5A), very low frequency of malformations (1.00% and 0.28%, respectively) (Figure 5B) and survival at 24hpi (92.60% and 85.50%, respectively) (Figure 5C) similar to non-injected tp53 M214K embryos (0.00% and 98% for malformation frequency and survival, respectively) (Figure 5B and 5C). Because a less severe phenotype was observed in BRAF V600E -expressing tp53 M214K embryos, a western blot of embryo lysates was conducted to determine whether the phenotype would be due to a decrease of BRAF expression. Blots confirmed the presence of both mCherry and BRAF proteins in tp53 M214K embryos injected with mCh-T2A-BRAF V600E mRNA but BRAF protein levels were significantly lower than those found for BRAF WT -expressing tp53 M214K embryos (Figure 5D; D’ and D’’) (p<0.001) or even BRAF V600E -expressing WT embryos (Figure 4E’’). 97 98 Figure 5. mRNA experiment in tp53 M214K embryos. (A) Representative images of 24hpf non-injected and 24hpi tp53 M214K embryos injected with one of the three mRNAs: mCh, mCh-T2A-BRAF WT or mCh-T2A-BRAF V600E . On the left are shown BF images; on the right 99 images acquired with a Texas Red® filter. Scale bar: 300μm. (B) Percentage of developmental defects at 24hpf or 24hpi (statistical analysis only shown for relevant comparisons; p<0.001). (C) Survival at 24hpf or 24hpi (statistical analysis only shown for relevant comparisons; p<0.001). (D) Western blot analysis of lysates from embryos at 4hpf or 4hpi. 50 embryos were pooled for each condition. Membranes were blotted with mCh and BRAF antibodies. Ponceau S stain showed equal amount of protein in all lanes. (D’) mCherry and (D’’) BRAF protein levels were quantified and normalized to the amount of protein. Values are shown in bar graphs and represent three independent experiments (statistical analysis only shown for relevant comparisons; p<0.001). Taken together, mRNA experiments showed that BRAF V600E -induced phenotype in zebrafish embryos was conditioned by p53 and the presence of a WT p53 protein triggered protective mechanisms (possibly apoptosis and/or cell cycle arrest) against the oncogene. The mRNA experiments also confirmed an efficient translation of two independent proteins from mCh-T2A-BRAF V600E mRNA. This suggested that, independently of the absence of mCherry expression in the tg(tg:mCh-T2A-BRAF V600E ) fish, the mutant form of BRAF was able to be translated. To confirm BRAF expression in the stable tg(tg:mCh-T2A-BRAF V600E ) line, western blots of protein extracts from thyroid tissues of three month-old fish (Figure 6A) using a BRAF antibody raised against a peptide of human origin were performed. It was confirmed an increase in the levels of BRAF when compared to tg(tg:mCh) (Figure 6B). Also, a robust increase in the levels of phosphoExtracellular Regulated Kinase (pERK), a downstream of BRAF, confirmed activation of the MAPK pathway by BRAF V600E (Figure 6B). This suggested that the BRAF protein was indeed functional in thyroid cells of the transgenic line. 100 Figure 6. BRAF and pERK expression in tg( tg :mCh-T2A-BRAF V600E ) fish. (A) Schematic representation of the thyroid tissue dissected from mCh-expressing transgenic zebrafish to produce tissue lysates. All follicles were dissected and separated from non-mCh expressing tissues. (B) Western blot analysis of lysates from thyroid tissues of tg(tg:mCh) (first lane) and tg(tg:mCh-T2A-BRAF V600E ) (second lane) fish at 3 months of age. Three fish were pooled for each line. Membranes were blotted with BRAF and pERKs antibodies. Ponceau S showed equal amount of protein in both lanes. 107 Paradoxically, BRAF V600E was also found to induce senescence and apoptosis by blocking proliferation (Wajapeyee et al., 2008). In BRAF V600E -expressing transgenic animals it was be expected the opposite observation as proliferation was being induced. p16, an important mediator in blocking cell progression from G1 to S phase, was evaluated to exclude BRAF V600E -induced senescence as a restraining factor to malignancy. p16 protein levels were downregulated in tg(tg:mCh-T2A-BRAF V600E ) fish (p>0.05) (Figure 8F and F’) suggesting that BRAF V600E did not induce senescence. Activated caspase 3, an important protein in the execution-phase of cell apoptosis, was also evaluated to exclude BRAF V600E -induced apoptosis as a mechanism to clear genetically unstable cells. Suprisingly, activated caspase 3 protein levels were increased in tg(tg:mCh-T2A-BRAF V600E ) fish (p<0.0001) (Figure 8F and F’). To confirm if caspase activation was induced by p53 as a response to oncogenic stress, p53 protein levels were assessed. Increased levels of p53 protein were found in tg(tg:mCh-T2A-BRAF V600E ) fish (p<0.0001) (Figure 8F and F’) confirming that apoptosis was triggered in BRAF V600E -expressing thyroid cells. To determine whether p53 activation was mediated indirectly by BRAF V600E through BRAF V600E -mediated p38 activation, levels of p38 were assessed. p38 protein levels in tg(tg:mCh-T2A-BRAF V600E ) fish were not different from the controls (p>0.05) (Figure 8F and F’). 108 109 110 Figure 8. Characterization of tg( tg :mCh-T2A-BRAF V600E ) juveniles. (A) Representative phenotypes of tg(tg:mCh) and tg(tg:mCh-T2A-BRAF V600E ) at 3 months of age. Scale bar: 5mm. (B) BMI measured from tg(tg:mCh) (n=39) and tg(tg:mCh-T2A-BRAF V600E ) (n=39) (p<0.0001). (C) HE stain of longitudinal sections representative of the lower jaw at 2and 3mpf. Scale bar: 250μm. C1’’ and C2’’ are higher magnifications of the follicles. Scale bar: 62μm. (D) Incidence of hyperplasia at 3mpf. Values correspond to the percentage of fish with hyperplasia confirmed by histology (ten fish were studied from each line) (p<0.0001). (E) Representative thyroid volume assessed in tissue sections. Values correspond to the volume of each fish studied; ten in total for each line (p<0.05). (F) Western blot analysis of lysates from thyroid tissue at 3mpf. Three fish were pooled for each line. Membranes were blotted with relevant antibodies. Ponceau S is shown in Figure 6. (F’) Protein levels were quantified and normalized to the amount of protein. Values are shown in bar graphs and represent three independent experiments (statistical analysis only shown for relevant comparisons; p<0.05, p<0.01 and p<0.0001). tg(tg:mCh-T2A-BRAF V600E ) fish were monitored until they reach 12 months of age. At this stage, the differences in body weight and body length proved not to be transient (Figure 9A) as reflected by a 11% decrease in body mass when compared to tg(tg:mCh) fish (p<0.001) (Figure 9B). At this stage, it was observed a pattern of mCherry expression that suggested enlarged follicles so histological examination was performed to confirm such hypothesis. tg(tg:mCh) fish displayed thyroid follicles scattered predominantly in the connective tissue near the ventral aorta (Figure 9C: C1 and C1’). Each follicle was round or oval and was lined by a single-layered epithelium; the center of the follicle was filled with homogenous colloid (Figure 9C: C1 and C1’). tg(tg:mChT2A-BRAF V600E ) fish displayed signs indicating colloid goiter which involved the lower jaw. Follicles were predominantly large or very large close to each other with varying shapes (Figure 9C: C2 and C2’). Small follicles were also observed. Follicles were lined by basophilic and cuboidal epithelial cells and follicles with excessive colloid lined by a flattened layer of epithelial cells (Figure 9C: C2 and C2’). These follicles were well-differentiated and in some cases due to their size, follicles even constricted the ventral aorta. Taken together, none of the tg(tg:mCh) (0/20) but most of the tg(tg:mCh-T2A-BRAF V600E ) fish (18/20) displayed colloid goiter as confirmed by histology (goiter incidence at 12mpf: p<0.0001) (Figure 9D). 111 When examining tissue sections of tg(tg:mCh-T2A-BRAF V600E ) fish, many exhibited follicles that were markedly larger (Figure 9C: C2 and C2’). In order to determine the extension of the goiterous thyroid, 3-μm-thick sections covering the whole thyroid of the adult fish were selected to determine thyroid volume. On average, tg(tg:mCh-T2A-BRAF V600E ) fish displayed a 6-fold increase in thyroid volume when compared to controls (p<0.001) (Figure 9E). Although colloid goiter has a good prognosis and survival in humans, thyroid dysfunction without a proper follow-up can result in complications. In tg(tg:mChT2A-BRAF V600E ) fish there was never an experimental interference in the endocrine homeostasis besides the expression of BRAF oncogene by thyroid cells and fish were subjected to the same husbandry conditions as age-matched controls. To verify whether a goiterous state could have influenced the lifespan, all fish from this work that were not used for histological examination were included to determine survival rates. Survival up to 12mpf was similar to that observed for the tg(tg:mCh) fish (p>0.05) (Figure 9F). 112 113 Figure 9. Characterization of tg( tg :mCh-T2A-BRAF V600E ) adults. (A) Representative phenotypes of tg(tg:mCh) and tg(tg:mCh-T2A-BRAF V600E ) at 12 months of age. Scale bar: 7.5mm. (B) BMI measured from tg(tg:mCh) (n=25) and tg(tg:mCh-T2A-BRAF V600E ) (n=28) (p<0.001). (C) HE stain of longitudinal sections representative of the lower jaw at 12mpf. Scale bar: 250μm. C1’ and C2’ are higher magnifications of the follicles. Scale bar: 62μm. (D) Incidence of goiter at 12mpf. Values correspond to the percentage of fish with goiter confirmed by histology (twenty fish were studied from each line) (p<0.0001). (E) Representative thyroid volume assessed in tissue sections. Values correspond to the volume of each fish studied; twenty in total for each line (p<0.001). (F) Survival curves for tg(tg:mCh) (n=25) and tg(tg:mCh-T2A-BRAF V600E ) (n=25) up to 12mpf (p>0.05). Values are percentages of fish alive for a certain timepoint. 114 Mutating tp53 prevents BRAF V600E -induced thyroid dysfunction in transgenic fish tg(tg:mCh-T2A-BRAF V600E ) fish displayed an increase in the protein levels of WT p53 by 3 months of age concomitant with thyroid hyperplasia suggesting that p53 was restraining tumor progression. In order to facilitate cancer development, a tp53 M214K zebrafish line was used. tp53 M214K -homozygous zebrafish are viable and develop tumors by 16 months of age that histologically resemble malignant peripheral neural shealth tumors not seen in humans and mice with germline mutations of TP53 (Berghmans et al., 2005; Storer et al., 2010). Of note, zebrafish p53 is highly similar to mammalian p53 in both structure and function sharing 48% homology in amino acid sequence to human p53 (Cheng et al., 1997). p53 is ubiquitously expressed early in zebrafish development (1hpf), at the pharyngula stage (24-48hpf) is predominantly expressed in the head and by 48hpf the protein levels are barely detectable (Cheng et al., 1997; Lee et al., 2008). tp53 M214K missense mutation affects the tp53 protein at an aminoacid position that is orthologous to TP53 mutations found in human cancer cells. The orthologous human codon for the tp53 M214K mutation is in exon 7, methionine-246, which was found to be mutated in 124 different human tumors, 8 of which exhibit the same amino acid change (Olivier et al., 2002; Berghmans et al., 2005). This codon is also positioned between other known mutation hot spots in the DNA-binding domain (DBD) of the human TP53 gene at codons 245, 248, and 249 of exon 7 (Olivier et al., 2002; Berghmans et al., 2005). tp53 M214K mutated protein is not able to activate transcription through the p21 response element and behaves in a dominant-negative manner, inhibiting the ability of the WT tp53 to activate downstream effectors (Berghmans et al., 2005). Homozygous tp53 M214K fish were identified from an incross of heterozygous fish. The formed were crossed with tg(tg:mCh) fish to produce heterozygous tp53 M214K tg(tg:mCh) or with tg(tg:mCh-T2A-BRAF V600E ) to produce heterozygous tp53 M214K tg(tg:mCh-T2A-BRAF V600E ) fish. Heterozygous fish were incrossed and the homozygous progeny also harboring the transgene were used for all further experiments. These clutches exhibited Mendelian segregation for the tp53 M214K mutation and embryonic development of the mutant was not affected as described by Berghmans et al., 2005. 115 When analyzing live tp53 M214K tg(tg:mCh-T2A-BRAF V600E ) embryos, a robust thyroidspecific reporter signal became detectable at around 34hpf and it was maintained throughout thyroid morphogenesis similar to tp53 M214K tg(tg:mCh) embryos. To study whether BRAF V600E would interfere with thyroid morphogenesis in the absence of WT p53 protein in tp53 M214K tg(tg:mCh-T2A-BRAF V600E ), mCherry expression was evaluated from 60-72hpf and up to 7dpf. In that line it was observed the organization of follicular-like structures in the pharyngeal region consistent with normal thyroid morphogenesis similarly to tp53 M214K tg(tg:mCh). Throughout the screening it was obvious that homozygous BRAF V600E -expressing larvae developed normally and were indistinguishable from homozygous controls in terms of viability and fertility (Figure 10A). When assessing the percentage of larvae displaying normal thyroid morphogenesis, 84% (137/163) of tp53 M214K tg(tg:mCh;tg:mCh-T2A-BRAF V600E ) and 90% (87/96) of controls showed signs indicating no impairment of thyroid morphogenesis (p>0.05) (Figure 10B). The number of follicles were quantified in order to confirm that thyroid morphogenesis was indeed normal. In tp53 M214K tg(tg:mCh;tg:mCh-T2A-BRAF V600E ) larvae the number of follicles (2-5 follicles) was similar to the number found in tp53 M214K tg(tg:mCh) (2 to 6 follicles) (p>0.05) (Figure 10C). A similar growth between both lines was verified when measuring SL at 7dpf (p>0.05) (Figure 10D). By conventional fluorescence microscopy, it was verified that thyroid was developing normally in tp53 M214K tg(tg:mCh;tg:mCh-T2A-BRAF V600E ) larvae. However, it was important to clarify whether thyroid cells were organized similarly to tp53 M214K controls so histological examination of the lower jaw was performed from 7-, 14- , 21and 28-days-old larvae. tp53 M214K tg(tg:mCh) revealed thyroid follicles located predominantly in the connective tissue of the lower jaw. The individual follicles were round or oval and were lined by a single-layered cuboidal epithelium; the center of the follicle was filled with colloid and vesicles were seen (Figure 10E). tp53 M214K tg(tg:mCh-T2A-BRAF V600E ) larvae displayed the same histology features described above (Figure 10E). The survival in tp53 M214K tg(tg:mCh-T2A-BRAF V600E ) was similar to the controls which was consistent with normal thyroid growth during the larval stage (p>0.05) (Figure 10F). 116 123 volume when compared to controls (p<0.001) (Figure 12E), still lower that the volume measured in tg(tg:mCh-T2A-BRAF V600E ) for the same age. To evaluate survival, all fish from this work that were not used for histological examination were included. Survival up to 12mpf was similar to that observed for controls (p>0.05) (Figure 12F). 124 Figure 12. Characterization of tp53 M214K tg( tg :mCh-T2A-BRAF V600E ) adults. (A) Representative phenotypes of tp53 M214K tg(tg:mCh) and tp53 M214K tg(tg:mCh-T2A-BRAF V600E ) at 12 months of age. Scale bar: 7.5mm. (B) BMI measured from tp53 M214K tg(tg:mCh) (n=25) and tp53 M214K tg(tg:mCh-T2A-BRAF V600E ) (n=28) (p>0.05). (C) HE stain of longitudinal sections representative of the lower jaw at 12mpf. Scale bar: 250μm. C1’ and C2’ are higher magnifications of the follicles. Scale bar: 62μm. (D) Incidence of goiter at 12mpf. Values are percentages of fish with goiter confirmed by histology (eleven and nine fish were studied from BRAF V600E and control line, respectively) (p>0.05). (E) Representative thyroid volume assessed by tissue sections. Values correspond to the volume of each fish studied; eleven and nine fish were studied from BRAF V600E and control line, respectively (p<0.001). (F) 125 Survival curves for tp53 M214K tg(tg:mCh) (n=25) and tp53 M214K tg(tg:mCh-T2A-BRAF V600E ) (n=25) up to 12mpf (p>0.05). Values are percentages of fish alive for a certain timepoint. Reduced BRAF V600E expression in tp53 M214K fish may explain absence of BRAF-induced phenotype It was observed an abnormal thyroid morphogenesis in tg(tg:mCh-T2A-BRAF V600E ) larvae by 7dpf and accelerated growth that was not seen in age-matched tp53 M214K tg(tg:mCh-T2A-BRAF V600E ) fish (p<0.0001) (Figure 13A). tg(tg:mCh-T2A-BRAF V600E ) fish developed thyroid hyperplasia by 3 months of age and colloid goiter was observed at 12 months of age as well as a delay in growth (BMI at 3mpf: p<0.01 and at 12mpf: p<0.01), concomitant with BRAF V600E expression in thyroid cells (Figure 13B). Hyperplasia and reduced growth were not observed in age-matched tp53 M214K tg(tg:mCh-T2A-BRAF V600E ) fish (Figure 13B). This suggested that in the absence of WT p53, the BRAF V600E -induced phenotypes in fish were not sustained. In one hand, BRAF V600E expression at 3 months of age in tg(tg:mCh-T2A-BRAF V600E ) fish is consistent with activation of MAPK and PI3K-AKT pathways through pERKs and pATKs, respectively (Figure 13D). This suggested a role of these pathways in the hyperplasia observed. On the other hand, upregulation of p53 and activated caspase 3 (Figure 13D) suggested that p53 put a brake in BRAF-expressing cells. When compared to the tp53 M214K tg(tg:mCh-T2A-BRAF V600E ) fish, BRAF V600E expression is reduced concomitant with low levels of pERKs and pATKs (Figure 13D). This suggested that even in the absence of the brake that would favor progression to malignancy, the downregulation of important pathways have contributed to produce no evidences of hyperplasia or cancer until 12 months of age in the BRAF V600E -expressing tp53 M214K line. 126 127 Figure 13. Comparison between BRAF V600E -expressing tp53 wt and tp53 M214K lines. (A) SL measured at 7dpf from tp53 wt tg(tg:mCh-T2A-BRAF V600E ) (n=31) and tp53 M214K tg(tg:mChT2A-BRAF V600E ) (n=29) (p<0.0001). (B) BMI measured at 3 months of age from tp53 wt tg(tg:mCh-T2A-BRAF V600E ) (n=39) and tp53 M214K tg(tg:mCh-T2A-BRAF V600E ) (n=30) (p<0.01). (C) BMI measured at 12 months of age from tp53 wt tg(tg:mCh-T2A-BRAF V600E ) (n=27) and tp53 M214K tg(tg:mCh-T2A-BRAF V600E ) (n=18) (p<0.01). (D) Western blot analysis of lysates from thyroid tissue at 3mpf. Values are shown in bar graphs and represent three independent experiments, each corresponding to a pool of three fish (statistical analysis only shown for relevant comparisons; p<0.001 and p<0.0001). Conditional targeted expression of BRAF V600E in thyroid cells of transgenic zebrafish (preliminary data) Expressing BRAF V600E in thyroid cell lines and in the thyroid of mice models have shown contradictory data. Mitsutake et al., 2005 showed that in vitro conditional BRAF V600E expression failed to transform rat differentiated thyroid PCCL3 cells and induce apoptosis in parallel with increased DNA synthesis, dedifferentiation and chromosomal instability (Mitsutake et al., 2005). Moreover, Vizioli et al., 2011 demonstrated that primary cultures of thyroid transfected with BRAF V600E promote oncogene-induced senescence but not oncogenic transformation (Vizioli et al., 2011). In mice, Shimamura et al., 2013 described that conditional BRAF V600E expression did not induce tumor formation and engineered postnatal expression of BRAF V600E in a small number of thyroid cells did not initiate tumorigenesis in thyroid (Shimamura et al., 2013) despite of mice modeled to thyroid-specific expression of BRAF V600E developed goiter and invasive PTCs early in life (Knauf et al., 2005; Mercer et al., 2005; Franco et al., 2011). In tg(tg:mCh-T2A-BRAF V600E ) fish, a very early impairment of thyroid morphogenesis and development of hyperplasia was observed by 2 months of age. In order to model sporadic thyroid cancer, a transgenic model was developed in which BRAF V600E was expressed late in life and controlled in a very efficient manner under physiological serum TSH concentrations. 128 In the past decade, the Cre-loxP technology, when combined with inducible systems, has allowed a controlled spatial and/or temporal expression of mutations from tumor suppressor genes and oncogenes. The tg(hsp70l:mCherry-T2A-CreER T2 ) transgenic line is a non-leaky conditional line that has a bicistronic mRNA coding for mCherry and CreER T2 separated by a viral T2A peptide sequence under the control of the zebrafish heat shock cognate 70kd protein, like (hsp70l) promoter (Hans et al., 2011). In the absence of tamoxifen (TAM), Cre-loxP-site-specific recombination does not occur whereas in the presence of TAM and after heat shock full recombination can be achieved. A construct, in which BRAF V600E expression would be normally suppressed by the presence of a floxed gene but induced when this gene was rearranged by Cre, was developed. To facilitate cloning, the following previous constructs were used: Tol2CG2tg:mCh and Tol2CG2tg:mCh-T2A-BRAF V600E . The loxP-CFP-loxP cassette was cloned downstream of the tg promoter and upstream of the mCh or mCh-T2A-BRAF V600E (Figure 14A and 14B, respectively). CFP reporter was chosen because it would allow to monitor the recombination process in thyroid cells upon heat shock and TAM treatment. Tol2CG2tg:loxP-CFP-loxP-mCh or Tol2CG2tg:loxP-CFP-loxP-mCh-T2A-BRAF V600E was injected with capped transposase mRNA into one-cell-stage tg(hsp70l:mCherry-T2A-CreER T2 ) embryos. By fluorescence microscopy, mosaic CFP expression was observed in thyroid tissue of F0 animals injected with Tol2CG2tg:loxP-CFP-loxP-mCh or Tol2CG2tg:loxP-CFP-loxP-mCh-T2A-BRAF V600E . This was the confirmation that the construct was functional and tg promoter was sufficient to drive expression of CFP specifically to the thyroid cells. Embryos presenting mosaic CFP or EGFP expression in the thyroid or the heart, respectively were grown to adulthood and F0 founders were identified by a specific-reporter signal exclusively in those tissue of their progeny. Stable transgenic lines were established using F0 founders. A robust CFP expression was confirmed in thyroid cells of both F0-hsp70l:mCherry-T2A-CreER T2 ;tg:loxP-CFPloxP-mCh and F0-hsp70l:mCherry-T2A-CreER T2 ;tg:loxP-CFP-loxP-mCh-T2ABRAF V600E lines. 129 When analyzing live embryos from the lines above-mentioned, a robust thyroid specific reporter signal became detectable at around 34hpf and it was maintained throughout thyroid growth similar to tg(tg:mCh) embryos. A preliminary experiment was conducted by inducing Cre-loxP-mediated recombination in tg(hsp70l:mCherry-T2A-CreER T2 ;tg:loxP-CFP-loxP-mCh-T2ABRAF V600E ) fish at 2 ½ months of age. This age was chosen not only because recombination during larval stage would not portrayed the sporadic cancer but also because fish would handle better the anesthesia, performed in order to monitor the reporter expression. The recombination was induced by heat shock followed by exposure to the active metabolite of TAM, 4-Hydroxyl-Tamoxifen (4-OHT). Recombination efficiency was assessed by a gradual loss of CFP expression and gradual gain of mCherry expression, which was ubiquitous in the initial hours after the heat shock and specifically to the thyroid cells in the days that followed (Figure 14C). 130 Figure 14. Conditional constructs (A) Schematic representation of the (A) Tol2tg:loxPCFP-loxP-mCh and (B) Tol2tg:loxP-CFP-loxP-mCh-T2A-BRAF V600E constructs containing Tol2 elements and the cmlc2:EGFP-pA cassette. (C) Schematic representation of the Cremediated recombination of the hsp70l:mCh-T2A-CreER T allele in the blue-to-red reporter tg:loxP-CFP-loxP-mCh/mCh-T2A-BRAF V600E line in the presence of heat and 4-OHT exposure and final transgenes kept in the genome. Transgenic embryos carrying the CFP reporter and the tg(hsp70l:mCh-T2ACreER T2 ) allele displayed a strong CFP signal until 2 ½ months of age. Heat shock and exposure to 4-OHT led shortly after to a strong ubiquitous mCherry expression 131 indicating successful induction of CreER T2 expression. Fish were monitored and CFP expression was completely lost during the first weeks after induction. This confirmed the successful recombination event in most or all thyroid cells. mCherry expression was never detected; observation that corroborated with the absent mCherry expression in tg(tg:mCh-T2A-BRAF V600E ) fish. Without having an expressing reporter that would help to monitor thyroid morphology, fish were grown until they reached 4 months of age and then histological examination was performed. tg(hsp70l:mCherry-T2A-CreER T2 ;tg:loxPCFP-loxP-mCh-T2A-BRAF V600E ) fish, in which recombination was not induced, displayed thyroid follicles scattered predominantly in the connective tissue near the ventral aorta (Figure 15 A). Follicles were round to oval and were lined by a single-layered epithelium confining a very homogenous colloid (Figure 15 A’). Fish in which recombination was induced showed follicular hyperplasia in the lower jaw (Figure 15B). The follicles were either big or very small and were very close. The epithelium lining the follicles was variably basophilic cuboidal to columnar (Figure 15B’). There was a marked inflammatory infiltrate likely to be of lymphocytic origin surrounding the follicles and inside the blood vessels nearby (Figure 15: B’’ and B’’’). Because the hyperplasia described resembled that found in tg(tg:mCh-T2ABRAF V600E ) fish, it suggested that expression of BRAF V600E was induced successfully. 132 Figure 15 . Cre-mediated recombination of the hsp70l :mCh-T2A-CreER T2 allele in the blue-to-red reporter tg( tg: loxP-CFP-loxP-mCh-T2A-BRAF V600E ) line in the absence and presence of heat and 4-OHT exposure. (A) Scheme of the transgene in the absence of recombination and HE stain of longitudinal sections representative of the lower jaw. Scale bar: 250μm. (B) Higher magnification of the follicles. Scale bar: 62μm. B: Scheme of the