Coordination between mitosis and apico-basal polarity in tissue organization and tumorigenesis

! Coordination between mitosis and apico-basal polarity in tissue organization and tumorigenesis Cátia Alexandra Carvalho Mendes Dissertação de Mestrado em Bioquímica Universidade do Porto Faculdade de Ciências Instituto de Ciências Biomédicas Abel Salazar 2012 ! 1! Cátia Alexandra Carvalho Mendes Coordination between mitosis and apico-basal polarity in tissue organization and tumorigenesis Dissertação de Candidatura ao grau de Mestre em Bioquímica da Universidade do Porto Orientador: Doutor Claudio E. Sunkel Categoria: Professor Catedrático Afiliação: Grupo de Genética Molecular da Mitose, Instituto de Biologia Molecular e Celular (IBMC), Instituto de Ciências Biomédicas de Abel Salazar (ICBAS), Universidade do Porto Co-orientador: Doutor Eurico Morais de Sá Categoria: Investigador Auxiliar Afiliação: Instituto de Biologia Molecular e Celular 2012 ! 2! Acknowledgments/Agradecimentos Agradeço primeiramente ao Professor Claudio Sunkel por mais uma vez me ter dado a oportunidade de trabalhar no laboratório, pelo incentivo e pelo entusiasmo demonstrado a cada nova pequena descoberta. Muito Obrigada. Ao Sensei Miyagi, um obrigada por todos os sábios conhecimentos transmitidos, pela paciência (que eu sei que foi precisa muita) e pela disponibilidade. Ao meu orientador por ter sido tantas vezes compreensivo, por me ter incentivado a ser o mais autónoma possível, e me ter dado liberdade para o tentar ser. Ao Eurico Sá essencialmente pela boa disposição. É uma das tuas melhores peculiaridades, isso e conseguires fazer-me sentir de consciência pesada com um simples “tu é que sabes”. Obrigada, sem ti nunca tinha chegado lá (não desfazendo o Sensei Miyagi) =) Á Tália Figueiredo, que no fundo continuou a ser a minha “mãe científica”, pelo carinho e pela disponibilidade para cada pergunta menos conseguida da minha parte (o eufemismo é uma das minhas figuras de estilo favoritas). Sinceramente, obrigada! Agradeço a todos os restantes membros do GMM por toda a ajuda prestada, pela disponibilidade e partilha, mas principalmente pelo bom ambiente criado no laboratório! Obrigada a todas as pessoas com quem partilhei a sala das moscas (especificamente o grupo de Biologia do Desenvolvimento) por transformarem as longas permanências em agradáveis passagens. Aos amigos de Mestrado que me acompanharam nesta jornada, que remaram comigo e que ouviram sempre todas as queixas mais infundadas. Obrigada por ampararem os “pré-lab meetings” e por todos os momentos de descontração que convosco encontrei...Não poderia contudo deixar de te mencionar a título especial, tal e qual como mereces! Obrigada Elísio... por todas as conversas, pelas noitadas, pelo apoio mas sobretudo pela confiança em mim. Andreia, Elísio, Jéssica, Pedro e Rita, em especial a vocês os cinco... um Obrigada! À Ana e à Inês um obrigada pelas pausas para a “descompensação dos nervos”! :) Ao Ricardo. Obrigada por teres influenciado a minha primeira escolha, revelou-se uma das melhores opções que tomei. Obrigada pelo abraço, pela presença, pela crítica, acho que até pelas zangas! Por seres tantas vezes, ainda que injustamente, o meu saco de boxe! Lembrar-me-ei sempre da importância que tiveste até aqui, porque neste caminho tu foste essencial! Obrigada pela sincera amizade que é tantas vezes um refúgio... Espero que nunca sejamos alguém que costumávamos conhecer. Sinceramente, Obrigada! À minha eterna colega de casa, Joana, que acompanhaste 4 anos e meio da minha vida... Foste o abraço, o colo, a voz e tudo o que era preciso quando era preciso! Orgulho, é o que sinto em ti... e em nós! Obrigada por toda a compreensão. Juliana, obrigada. Estiveste lá, mas eu quero sobretudo que estejas cá, que fiques. Obrigada por todas as conversas e por todas as tuas calinadas que tanto me fizeram rir. Só mesmo tu! :) Ao meu André... :) Sim, é verdade, vou-te agradecer pelas “chapadas”, pelo “escreve”, pelas “facadas”... pelas inúmeras conversas deprimentes, por me fazeres chorar, por me fazeres sentir culpada pelas minhas atitudes, pela dureza das tuas palavras... E sei que agradecer-te por isto poderá soar estranho, mas sei que tal como eu, tu sabes onde conseguiste chegar com tudo isto... Ou não fossemos nós farinha do mesmo saco! Sabes bem (demais) o quão importante és para mim, e por isso sabes exactamente onde me ! 3! “acertar para magoar”. E no fundo, lá bem no fundo, é isso que adoro em ti... Obrigada por estes 5 anos, obrigada por estas 5 semanas. “Sorri. Está um belo dia para isso” :D E ainda mais difícil do que escrever para a Brigada do Reumático, é sem dúvida escrever para ti... Agradecer-te? Gostava muito, mas é impossível... Tu és só uma, mas foste a minha família toda! Um obrigada nunca será suficiente por todas as vezes que me impediste de desistir, pela companhia de todas as noites até não aguentares, por disponibilizares o teu ombro esquerdo, o direito e até os joelhos para que simplesmente não desabasse tudo nos momentos de desespero! Wordless foi o que tudo aquilo que fizeste por mim me deixou. Isso e a certeza de que te “levo” comigo, seja para onde for... Porque lá no fundo, eu vou fazer parte da tua quinta, e tu, do meu laboratório! Ana, “Be strong. Be brave. And have no fears” because I will always believe in you * Acabaste de me trazer chá ao quarto e com um sorriso disseste: “Já misturei um bocadinho de água fria, mas cuidado não te queimes!”. Olhaste para o ecrã do computador e viste-o vazio. E eu sei exactamente o que pensaste: ”Escreve, eu estou de olho em ti!” e sais-te de mansinho... Estava precisamente a tentar encontrar palavras que conseguissem agradecer o quanto vocês fizeram por mim... Daí a folha estar vazia.. Porque estava difícil... Mas a tua visita tornou tudo claro... Nenhuma frase feita conseguiria algum dia descrever a vossa importância na minha vida. Bastaria a simples descrição de alguns momentos nossos, como o de cima, para que vocês percebessem que aquilo que nós somos... é simplesmente uma família. A minha família! Não vou cair no ridículo de utilizar a palavra “Obrigada”, porque isso seria tão menosprezante como não escrever que vos amo com todos os órgãos que vocês já me tiraram de tanta facada que levei... Não vou cair no ridículo de tentar escrever sobre cada um individualmente... Porque só isso teria mais páginas que esta tese... Não vou cair no ridículo de escrever que nunca nos afastaremos... Porque, infelizmente, isso é uma realidade... Vou sim cair no ridículo de escrever... Que sem vocês, eu nunca teria conseguido estar onde estou hoje... Porque vocês são parte da minha força, da minha vontade, da minha coragem para chegar até aqui! São tão unicamente parte do que eu sou! Porque é com vocês que rio, é com vocês que esqueço, é com vocês que volto a relembrar o que realmente é importante... Com vocês cresci, com vocês aprendi aquilo que não ensinam nos livros... Mas principalmente... Eu vivi... Com vocês eu simplesmente esqueço que existem problemas, responsabilidades, um sem fim de obrigações! Com vocês, viver, torna-se a mais simples de todas as tarefas... A vocês devo aquilo que provavelmente tenho de mais precioso: um chão seguro no qual me sento sempre que alguma ferida se lembra de reabrir... E sem julgamentos ou repreensões, vocês abraçam-me e dizem “Vai ficar tudo bem”. E eu acredito porque são vocês que o dizem... E eu acredito em vocês! E acredito sobretudo em nós... Porque todos juntos... Somos, simplesmente, uns perfeitos imperfeitos! Isto, é para vocês... Para a Brigada do Reumático... Porque sabem que mais? “Diz ao teu pai que lhe mande um beijo, foi o que ela disse” :’D Ao meu Tio, que tornou tudo simples e sobretudo, possível. Obrigada pela confiança! ! 4! Abbreviations AB: apico-basal AurA: Aurora A AJ: Adherens junctions APC/C: Anaphase promoting complex/ cyclosome aPKC: atypical protein kinase C Arm: Armadillo Baz: Bazooka C. elegans: Caenorhabditis elegans Cdk: Cyclin dependent kinase CENP: Centromere-associated Protein Crb: Crumbs DAPI: 4’,6’- diamidino-2-phenylindole DNA: Deoxyribonucleic acid GFP: Green fluorescence protein GMC: Ganglion Mother Cell hsp70: Heat-shock promoter 70 Lgl: Lethal giant larvae MDCK: Madin-Darby Canine Kidney Mis12: Minichromosome instability 12 par: partitioning defective RNAi: RNA interference SAC: Spindle assembly checkpoint SOP: sensory organ precursor cells S2: Drosophila Schneider 2 cell line TSGs: Tumour Supressor Genes UAS: Upstream Activating Sequence ! 5! Index Acknowledgments/Agradecimentos .................................................................................... 2 Abbreviations ....................................................................................................................... 4 Abstract ............................................................................................................................... 7 Resumo ............................................................................................................................... 9 CHAPTER 1 – INTRODUCTION .......................................................................................... 11 1. Cell division ................................................................................................................... 11 2. Cell polarity .................................................................................................................... 16 2.1 Models of epithelial tissue in Drosophila ..................................................................... 17 2.1.1. Follicular epithelium ................................................................................................. 17 2.1.2. Eye/antennal imaginal discs .................................................................................... 19 2.2. Establishment of epithelial apico-basal polarity in Drosophila .................................... 20 2.3. Interactions within the apical Par-6-aPKC-Par-3 complex .......................................... 21 2.4. Apical Polarity regulation ............................................................................................ 21 2.5. Interactions within the Scribble Complex and mutually inhibitory interactions with apical proteins ............................................................................................................ 22 3. Neuroblast polarity ......................................................................................................... 24 4. Aurora A: A mitotic kinase with a role in cell polarity ..................................................... 26 5. The role of cell polarity and chromosomal instability in tumorigenesis .......................... 27 6. Main goals ..................................................................................................................... 29 CHAPTER 2 – MATERIALS AND METHODS ........................................................................ 30 2.1 The GAL4-UAS system ............................................................................................... 31 2.2 The FLP/FRT system .................................................................................................. 32 2.3 The FRT system .......................................................................................................... 33 2.4 Immunofluorescence in Drosophila ovaries ................................................................. 33 2.5 Immunofluorescence on Drosophila eye/antennal imaginal discs ............................... 34 2.6 Clonal analyzes ........................................................................................................... 34 2.7 Drug-induced treatments ............................................................................................. 34 2.8 Primary and Secondary antibodies .............................................................................. 35 2.9 Imaging ........................................................................................................................ 35 CHAPTER 3 – RESULTS .................................................................................................... 36 Cooperation between cell division and apico-basal polarity defects in tumorigenesis induction .................................................................................................................... 36 3. Using the eye/antennal imaginal discs .......................................................................... 36 3.1 Introduction .................................................................................................................. 36 3.2 scribRNAi causes a reduction in the size of the Drosophila eye ..................................... 37 3.3 Cytokinesis and Aurora A altered levels have an interaction with polarity defects to induce eye outgrowth ................................................................................................. 38 3.3.1. Cooperation between cytokinesis and polarity defects in tumorigenesis does not depend on centrosome amplification ......................................................................... 41 3.4 Anillin and Scribble defective levels causes dramatic morphologic changes in the eye/antennal imaginal discs ....................................................................................... 45 3.5 Aurora A overexpression induces tumorigenesis by promoting photoreceptor differentiation failure .................................................................................................. 47 4. Using the follicular epithelium from Drosophila ovaries ................................................. 50 4.1 Background ................................................................................................................. 50 4.2 dlgRNAi causes tissue disorganization in the follicular epithelium ................................. 51 4.3 Anillin depletion and Aurora A overexpression enhance the intermediate phenotype presented by dlgRNAi ................................................................................................... 53 5. Dissecting the role of AurA in the follicular epithelium ................................................... 57 ! 6! 5.1 Background ................................................................................................................. 57 5.2 Aurora A overexpression delays anaphase onset ....................................................... 57 5.3 Low levels of Aurora A do not cause any defects in apico-basal polarity .................... 59 5.4 Aurora A kinase activity is not required to maintain apico-basal polarity in epithelial cells ............................................................................................................................ 61 5.5 Par-6 phosphorylation on Ser34 is dispensable for apico-basal polarity ..................... 62 5.6 Lgl dynamics during mitosis on the follicular epithelium .............................................. 63 5.7 Using BazS980A to address the dynamic behaviour of apical proteins during mitosis in epithelial cells ............................................................................................................ 65 CHAPTER 4 – DISCUSSION AND CONCLUSION .................................................................. 68 6.1 Cytokinesis failures can drive tumorigenesis in a polarity defective background ........ 68 6.2 Aurora A is not required to the establishment of apico-basal polarity in epithelial cells ................................................................................................................................... 70 6.3 Possible role of Lgl in driving mitotic depolarization of epithelial cells ......................... 72 Appendix ............................................................................................................................ 75 References.........................................................................................................................84 ! 7! Abstract The ability to asymmetrically distribute components within a single cell, termed cell polarity, is an essential feature of most cell types. Loss of cell polarity and chromosome instability are hallmarks of cancer, however these defects alone cannot induce tumorigenesis. We aim to understand if mitotic defects cooperate with loss of cell polarity in tumorigenesis. We screened for genes involved in different aspects of cell division, whose simultanous inactivation with polarity genes results in tumorous phenotypes. Using the eye imaginal disc and the follicle epithelium as models of epithelial tissue, we observed that defects in cytokinesis have the ability to induce the overgrowth of disorganized tissue in a polarity defective background. Establishing cell polarity relies on the evolutionary conserved apical Par Complex, comprised by Par-3, Par-6 and aPKC proteins. In neuroblasts, polarization only occurs during division to determine the asymmetric segregation of determinants of the daughter cell fate. In contrast, epithelial cells are polarized along the apico-basal (AB) axis during interphase to localize functions and adhesive properties at distinct cortical domains. However, during mitosis epithelial cells depolarize to accommodate cell shape changes associated with division. Aurora A kinase (AurA) has an essential role in neuroblast polarization during mitosis. It phosphorylates Par-6, participating in aPKC activation and the consequent phosphorylation of Lgl. This cascade of events allows the exclusion of Lgl from the cell cortex, which determines the basal positioning of cell fate determinants. To understand how epithelial cells link cell polarity and mitotic events, we addressed the role of aurA in AB polarity in the follicular epithelium. Mosaic mutant clones of an aurA kinase dead allele revealed that aurA kinase activity is not required to maintain epithelial AB polarity. The Lgl complex controls epithelial polarization by inhibiting the Par Complex activity on the basolateral cortex. Given that Lgl exits from the cell cortex during neuroblast division, we tested if it exhibited the same dynamic behaviour in epithelial cell division. In fact, similarly to the neuroblast, Lgl exits from the cortex to the cytoplasm prior than Nuclear Envelope Breakdown. Importantly, exit from the cortex might depend on phosphorylation, as a triple serine mutant version of Lgl is maintained at the cell cortex during division. The mutated sequence includes the aPKC phosphorylation site on Lgl, suggesting that aPKC could participate in Lgl exclusion from the basolateral cortex. However, when comparing depolarization timings between Lgl protein and apical polarity markers in epithelial cells, we observed that Lgl exit from the cell cortex begins before aPKC complex depolarization. Therefore our data supports that the cascade of events determining changes in polarity during mitosis is inverted in epithelial cells. ! 8! ! 15! Upon SAC inactivation, anaphase proceeds and the cell exits from mitosis (reviewed in [1]; [27, 28]; [29]) . “The Kinetochore is therefore a highly complex machine that does not merely bind and affects the dynamic behaviour of attached microtubules, but is involved in quality-control mechanisms that detect and correct defective or nonproductive kinetochore-microtubule interactions” (reviewed in [30]). Figure 1.3 - KMN Network and Aurora B controling kinetochore-microtubule attachments.!! The CPC component Aurora B is the responsible for sensing incorrect kinetochoremicrotubule attachments, phosphorylating the Ndc80 complex and allowing the reestablishment of bipolar attachments. Adapted from ([1]). ! Figure 1.4 - Sister chromatid separation after activation of the APC/CCdc20 complex . Correct kinetochore-MT attachments inactivate SAC signaling, that in turn allows the binding of Cdc20 to the APC/C. The activation of APC/CCdc20 leads to the polyubiquitylation of securin and cyclin B, targeting them for destruction. Securin destruction releases separase that cleaves the cohesion complex, allowing anaphase completion. Adapted from department of Genetics, University of Bayreuth, Germany! ! 16! 2. Cell polarity Multicellular organisms are though to result from the necessity to adapt to new environmental conditions and to utilize resources that a single-cell organism is unable to use. The cells within a multicellular organism are dependent on each other, such that each cell type is specialized in different and essential functions. These specialized cells constitute distinct tissues that work together within an organism. The epithelial tissue, shields us against pathogens, enables selective absorption and it also actively participates in the overall body secretion. In order to provide these functions, the epithelial tissue must form a selective barrier, so that signals between cells can be transmitted but harm signals from the external environment could be prevented to reach the organism. Epithelial cells developed a strategy to fulfil this propose: the formation of connections between cells. Cell-cell connections are a key feature of epithelial tissue, holding cells together in an organized monolayer that offers protection, and simultaneously allows transcellular transport. The Adherens Junctions are a type of connection only found in epithelium, considered by many as the closest junction in the living world. Besides its extracellular functions, Adherens Junctions participate in the definition of distinct domains within an epithelial cell by acting as an intracellular barrier to macromolecules, regulating cell polarity. Cell polarity refers to all the asymmetry of macromolecules distribution that leads to differential localization of functions into particular regions of a given cell. Cell polarization is a highly complex process: coordination between microtubule and actin remodelling, vesicle trafficking and cell-cell junctions’ establishment must be orchestrated to enable cells to carry out their functions properly. The importance of cell polarity is illustrated by several examples that rely on it: an axon must be polarized in order to transmit information to another neuron or an effector organ ([31]); leukocytes migrate towards a stimulus in a polarized-dependent manner ([32]); the establishment of asymmetric axes on vertebrate animals; the development of a fertilized egg or the growth of an unicellular organism ([33]). In conclusion, a cell from a multicellular organism must be specialized in a particular function, and in many cases to gain this specialization, that cell must become polarized to asymmetrically distribute its components. For instance, in epithelial cells, junctions must be formed at precise locations along the apical-basal axis. ! 17! 2.1 Models of epithelial tissue in Drosophila As previously mentioned, cell-cell connections and a remarkable polarization are common features of epithelial tissues. Giving their essentiality to the proper function of the tissue, these characteristics were conserved among several organisms, being Drosophila and mammals a well-known example of distant related eukaryotes where these features were retained. In spite of sharing common epithelial characteristics, vertebrates and invertebrates display some differences in what epithelial domains organization is concerned. Both eukaryotic classes share the existence of the Adherens Junctions. However while vertebrates have a Tight Junction above the previous referred one, in the invertebrate world such structure does not exist. Invertebrates have a Septate Junction that links epithelial cells just below the Adherens Junctions (Fig. 1.5). Additionally, there is high conservation among the signalling pathways that enables the formation of these features. Drosophila can therefore be used as a model to provide insights of how higher eukaryotes regulate their signalling pathways, (reviewed in [34]). Throughout this work two different epithelial models from Drosophila were used: the ovarian follicular epithelium and the eye/antennal imaginal discs. Description of the principal features of these models will be done in the following section. Figure 1.5 - Vertebrates and invertebrates epithelial domains organization. Both classes have the well-known Adherens junctions, however vertebrates have Tight Junctions instead of Septate Junctions that are presented by invertebrates. 2.1.1. Follicular epithelium Each female has a pair of ovaries that are composed of a variable number of ovarioles (12 to 20). Each ovariole contains the germarium in the most anterior part, followed by progressively old egg chambers ([35]). The germarium accommodates the ! 18! germline stem cells that divide asymmetrically to give rise to other stem cell and a differentiated one, known as cystoblast. By undergoing four rounds of synchronous mitosis without complete cytoplasmic division, the cystoblast originates an oocyte and 15 nurse cells that remain connected by ring canals and therefore share the same cytoplasm. At the same time, somatic follicle cells surround the 16-syncytial germline cells, forming an epithelial monolayer, the follicular epithelium (reviewed in[36];[37]). The epithelial cells within the follicular epithelium show an unequal distribution of its proteins, being organized into apical and baso-lateral domains and so defining a striking apico-basal polarity (Fig. 1.6). As so this epithelial model can be used to understand a variety of biological aspects, as stem cell regulation or cell polarization. Moreover it offers the possibility of easily use molecular and genetic tools to manipulate gene expression, analysing the effects of the loss of a particular gene expression through division. Figure 1.6 - Schematic representation of an ovariole and its specialized structures. The ovariole is composed of an anterior germarium that originates the posterior older egg chambers, ending in the formation of a mature egg. Surrounding each egg chamber is an epithelial monolayer called the follicular epithelium. Within this epithelium, the cells exhibit an apico-basal polarity, being the apical domain in intimate contact with the germline and the basal domain in contact with the extracellular matrix. Adapted from ([38]). ! 19! 2.1.2. Eye/antennal imaginal discs The adult Drosophila compound eye is another epithelial tissue commonly used to different studies. Primordially originated from the eye portion of the eye/antennal imaginal disc, the adult eye is constituted by more than 700 precisely arranged single unit eyes known as ommatidia. Tightly coordination between cell signalling, proliferation and cell death must be achieved in order to position and construct ommatidia correctly, enabling fly’s vision. During Drosophila third instar larval stage, proliferation generates about 10,000 cells in the eye, setting the base where differentiation shall take place (reviewed in [39]). Proliferation ends when a physical constriction – the morphogenetic furrow (MF) – appears close to the posterior part of the eye/antennal disc. This morphogenetic furrow defines a barrier between proliferation - anterior to the morphogenetic furrow - and differentiation – posterior to this physical invagination. Within the furrow, cells arrest in G1-phase of the cell cycle. While it moves to the most anterior part of the eye portion of the disc, it allows the beginning of cell fate determination in its posterior side (reviewed in [40]). The differentiated cells will assemble in cores of 8 photoreceptors, cone cells and pigment cells, constituting each individual ommatidium (Fig. 1.7). As fly viability does not depend on the existence of an eye, the effect of specific genetic manipulations in the eye imaginal disc can be scored in the adult eye, which is particularly useful to test the role of essential genes. ! ! ! Figure 1.7 - Schematic representation of the eye/antennal imaginal disc in the third instar larvae stage. The morphogenetic furrow constriction initiates in the posterior part of the eye portion of the imaginal disc leaving a differentiation track as it moves to the anterior part of the eye disc. Adapted from ([41]). ! 20! 2.2. Establishment of epithelial apico-basal polarity in Drosophila Besides generating a huge number of cells to sustain the viability of a multicellular organism, these cells must acquire distinct and specific functions. So, different groups of cells must become specialized in a particular function to ensure that the organism takes the greatest advantage it can from the environmental resources. Cell specialization requires changes in the morphologic structure of the cell and in the distribution of its molecular contents. The asymmetric distribution of components within a cell that enable it to carry out distinct functions by distinct domains is known as cell polarity. Epithelial cells are segmented into four main specialized regions due to polarization: (1) an apical domain that orchestrates communication between the environment and epithelial cells; (2) specialized structures that promote connection of epithelial cells – adherens junction in Drosophila and tight junctions on vertebrates; (3) a lateral domain where Septate junctions are formed; (4) a basal domain that provides communication with the extracellular matrix (reviewed in [42]). But which are the mechanisms that define cell polarization? In addition to my birth, key mediators of cell polarity were identified in C.elegans during the eighties. Ken Kemphues and Jim Priess aimed to identify proteins that interfere with the partitioning of C. elegans’ embryo. During their genetic screen, six proteins, the Partition defective (PAR) proteins, were found to be involved in asymmetric cell division and distribution of proteins and RNAs essential for cell specification ([43]). This was the first insight that these proteins could coordinate an intracellular polarity pathway. The six identified PAR proteins have distinct biochemistry properties: PAR-1 and PAR-4 are serine-threonine kinases while PAR-5 is a 14-3-3 protein recruited to serines and threonines after phosphorylation. PAR-2 has a characteristic RING finger domain whilst PAR-3 and PAR-6 have PDZ domains (reviewed in [44]). Upon these findings in C.elegans the question was obvious: Do Par proteins also have related functions in other organisms? After cloning the six par genes, its sequences revealed that they were evolutionary conserved. Bazooka, a gene required in Drosophila cellular polarization, was found to be similar to PAR-3 ([45]). Posteriorly, in mammals, it was found that a PAR-3 homolog could bind to an atypical protein kinase C (aPKC) that would be sequentially identified in C. elegans as part of the group of proteins that causes defects in the partitioning of C.elegans’ embryo ([46],[47]). PAR-3, PAR-6 and aPKC not only co-localize on the anterior part of the C. elegans zygote, but were also shown to form a complex in multiple systems, suggesting that they form a functional unit controlling cell polarization ([43];[48];[49]). ! 21! 2.3. Interactions within the apical Par-6-aPKC-Par-3 Complex It is now known that Par-6, aPKC and Par-3 form a complex involved in cell polarization in multiple biological contexts. By acting as a scaffold protein with multiple PDZ (Postsynaptic density 95, Discs large, Zonula occludens 1) domains, Par-3 (Bazooka in Drosophila) – sets the base for the association of Par-6 and aPKC, through self-association via its oligomerization CR1 domain in its N-terminal region. ([46];[50];[51];[52];[49];[53]). Par-6 also acts as a scaffold protein, binding to Par-3 through its C-terminal PDZ domain and to aPKC through its N-terminal Phox Bem1 (PB1) domain. In the middle of its sequence, Par-6 has a semi-CRIB (Cdc42 Rac Interacting Binding) motif that enables it to bind to GTP-bound Cdc42. It is known that when aPKC - the active component of the complex – interacts with Par-6, its kinase activity is suppressed. However the binding of GTP-bound Cdc42 to Par-6 semi-CRIB motif relieves Par-6 suppression of aPKC kinase activity. Thus, aPKC is correctly localized via PAR proteins, while its kinase activity is modulated via GTPase activity (Cdc42) (review in[54];[55];[52]). Mutation in any of the PAR Complex components leads to defects in the establishment of polarity, showing the importance of the interactions within the complex for cell polarization (reviewed in[34]). 2.4. Apical Polarity regulation Adherens junctions (AJs) provide a cue to the establishment of different domains within an epithelial cell. They allow the communication between neighbouring cells while they generate an adhesive belt – the zonula adherens – that avoids the free passage of molecules from the external environment, improving this way epithelial selectivity. Its main constituent is a transmembrane protein, called E-cadherin, whose intracellular domain links to the cytoskeleton through β-catenin (Armadillo in Drosophila) that also has a role in AJs formation. By localizing themselves close to the apical region of the cell, they act like a barrier that is involved in the definition of the apical and the basolateral domains in Drosophila (reviewed in[56];[57]). The formation of distinct domains relies on the existence of three different complexes: the above-mentioned Par Complex, the Scribble Complex, and also on the Crumbs Complex. This last complex is composed by 3 proteins: Crumbs (Crb) that is a transmembrane protein which localizes in the apical domain of a polarized cell; Stardust (Sdt) which binds to the intracellular region of Crb through its PDZ domains; and dPATJ that also interacts with Crb through its multiple PZ domains ([58];[59];[60],[61]). These ! 22! proteins have an essential role in establishing the apical domain, once that mutations in this complex result in loss of the apical identity. However they will not be further described because they were not studied on this work. The Scribble Complex will be described below. Par-3, Par-6 and aPKC have been assumed to work as a complex to regulate polarity in many biological systems. However in epithelial cells, Baz (Par-3) was found to localize at the level of Adherens Junctions, being positioned below the Par-6 and aPKC proteins. ([62];[63];[64]). Consistent with this, Baz interacts with Armadillo and the Nectinlike protein Echinoid in Drosophila, and with JAM1-3 and Nectin in mammals, all components of Adherens and Tight Junctions, respectively, localizing in a more subapical region than the other members of the complex ([65];[66];[67]). In fact Bazooka have an essential role in the definition of the apical-lateral border by setting out the correct position of the AJs in Drosophila and the Tight Junctions in mammalian cells ([62]; [68];[69]). In the most apical region, Par-6, aPKC and Baz bind transiently to each other. However, Bazooka is phosphorylated on its serine 980 by aPKC. This action destabilizes the complex since phosphorylated Bazooka cannot bind to aPKC, disrupting Baz-aPKC interaction. ([46];[52];[49];[70]). Yet, Baz could still be maintained in the tripartite complex via its interaction with Par-6’s PDZ domain. Nevertheless, the epithelial specific Crumbs Complex competes with Baz for the same PDZ domain of Par-6, breaking the interaction with Par-6 and excluding Baz from the apical domain ([70]). This exclusion leads to its localization on the subapical region where Baz defines the position of the Adherens Junctions. Evidences supporting this mechanism came from the fact that a nonphosphorylatable form of Bazooka – BazS980A, co-localizes with aPKC suggesting that phosphorylation by aPKC is necessary for the apical exclusion of Bazooka ([70]). 2.5. Interactions within the Scribble Complex and mutually inhibitory interactions with apical proteins Drosophila provided the basis for the discovery of the Scribble Complex. Scribble was found in a screen with the purpose to identify maternal mutations involved in the disruption of cell adhesion and polarity. scrib mutations caused disorganization in the epithelial monolayer of Drosophila embryos ([71]). The Scribble protein belongs to the LAP (LRR and PDZ) protein family because it has 16 leucine-rich repeats (LRR) at its Nterminal region and four PDZ domains in the opposite region ([71]). Posteriorly, two other genes, Lethal giant larvae (Lgl) and Discs large (Dlg), were shown to present a similar phenotype when mutated, being all three known as tumour suppressor genes as they ! 23! cause neoplastic growth in larvae imaginal discs ([72]). Dlg is a 102-kDa protein that possess 3 PDZ domains, a Src homology 3 (SH3) motif and a guanylate kinase-like (GUK) domain, placing this protein as a member of MAGUK (membrane-associated with guanylate kinase domain) family. Lgl is a 130-kDa protein having short motifs (40 amino acids) of WD (Trp (W)-Asp (D)) repeats ([73];[74]). Due to their similar phenotype when mutated and their interdependent co-localization in multiple biological contexts, Scribble, Lgl and Dlg have been considered as a functional complex that regulates basal polarization on epithelial cells ([72];[73]). It was shown that Scribble indirectly interacts with Dlg in Drosophila neuromuscular synapses by both binding to GUK-holder protein (GUKH) ([75]). Furthermore, in mammalian epithelial cells, it was indicated that Scribble and Lgl2 had a physical interaction ([76]). None of these interactions have been identified in Drosophila epithelial cells so far. Despite of this, it is commonly accepted that these proteins function in a complex that orchestrate the basolateral polarity of epithelial cells. In fact, mutations in components of Scribble Complex result in the lateral extension of the apical domain and lateral disruption of the Adherens Junctions ([71]). As evidence, Ecadherin and Armadillo, both markers of the Adherens Junctions, are spread through plasma membrane in scrib mutants. It is therefore suggested that Scribble Complex participates in the exclusion of the apical proteins from the basolateral domain ([71]). Both apical and basal complexes are involved in the correct positioning of AJs, once that Scribble Complex is essential to avoid AJs spreading through the lateral membrane, whereas the apical complex is required to exclude AJs from the apical domain. On the other hand, Armadillo prevents Dlg extension through the apical domain during follicular epithelium formation ([57]). AJs localization is then controlled by a mutually inhibitory mechanisms that comprise both apical and basolateral proteins. In fact, several lines of evidence suggest that apical and basal complexes restrict each other activity: (1) Loss of Scribble, Dlg or Lgl results in the extension of the apical proteins through the basolateral domain ([77]); (2) Both Lgl1 and Lgl2 bind Par-6 and suppress aPKC kinase activity in mammalian epithelial cells ([78];[79];[80]); (3) Lgl activity is restricted to the lateral cortex in an aPKC-phosphorylation dependent manner ([81]); (4) Crumbs overexpression has a similar phenotype as scrib mutant whilst crb mutant phenotype can be partially rescued by Scribble reduced levels ([77]); (5) it was found that Lgl2 and aPKC/Crb3 have an opposing function during Xenopus blastomeres polarization ([82]). All these data suggest that the separation between apical and basal domains requires the antagonism between the basal and the apical complexes that represses each other ectopic activity (Fig. 1.8). ! 24! Figure 1.8 - Apico-basal interactions on the definition of epithelial cell polarity. Representation of the organization of the polarized domains within an epithelial cell. The main repressible interactions between Adherens Junctions and the apical a basal domains are represented by the inhibitory black curves. 3. Neuroblast polarity While apico-basal polarity enables epithelial cells to organize specific functions within different cellular domains during interphase, the neuroblast makes use of the polarity machinery during mitosis to asymmetrically divide, creating two cells with different fates. Diversity is thus generated due to an asymmetric distribution of proteins that specifies two distinct regions: a self-renewal region, which will allow the maintenance of the neuroblast population, and a differentiation region that will lead to the formation of a differentiated cell, called Ganglion Mother Cell (GMC). This GMC will posteriorly divide originating two postmitotic cells that will become neurons or glial cells ([83]) (Fig. 1.9). Figure 1.9 - Schematic representation of neuroblast division. Neuroblasts exhibit an apico-basal polarity during mitosis. After asymmetric division, two different cells are formed: a differentiated ganglion mother cell and a self-renewal neuroblast. Adapted from ([84]). ! 31! 2.1 The GAL4-UAS system The GAL4-UAS system derived from yeast is a biochemical tool widely used to study the specific expression of a certain gene ([137]). It is composed by two main elements: the GAL4 driver and the Upstream Activating Sequence (UAS). The GAL4 protein is expressed under the control of a tissue-specific endogenous promoter. In turn the UAS sequence drives the expression of our transgene of interest, as the RNAi transgenes used into this study. After appropriate fly crossing and progeny selection, we have flies that carry concomitantly the GAL4 activator and the UAS-transgene as shown in Fig. 2.1. The GAL4 protein will therefore bind to the UAS-sequence, activating the expression of the gene that is under the control of the UAS. Therefore expression of our transgene of interest only occurs in the tissues that specifically express the GAL4 protein. Adaptations were made to this initially system. GAL80ts is a temperature-sensitive protein that binds and represses GAL4 protein activity. As so, by placing the GAL80ts under the control of the same enhancer of GAL4 promoting their concomitant expression, we are able to control the exact time of gene expression by placing the flies in a restrictive temperature that disrupts the GAL80ts protein. Figure 2.1 - Schematic representation of a crossing that enables the expression of a specific gene using the GAL4-UAS system. Addapted from ([138]). ! 32! 2.2 The FLP/FRT system The GAL4 system can be used in combination with other genetic tools to direct the expression of a certain gene to a restrict population of cell within a tissue. One of those tools is the “FLPout” system ([139]). The expression of the GAL4 protein is interrupted by the presence of a cassette flanked by Flipase recombination target (FRT) sequences, regardless the activation of its promoter. Temperature is the key controller of this system. By placing the flies at a restrictive temperature, the Flipase (FLP) protein expression is induced, once that it is under the control of a heat-shock promoter. Upon its expression, the Flipase protein will promote the recombination within the FLP recombination target (FRT) specific sites, promoting the excision of the cassette. After this excision, the GAL4 gene can be transcribed, allowing further activation of UAS sequences. Additionally to the UAS-transgene, this strategy concomitantly expresses a UAS-GFP transgene that positively marks the cells that are expressing the UAS transgenes, thereby creating GFP clones (Fig. 2.2). Figure 2.2 - Schematic representation of the combination of the FRT/FLP tool with GAL-UAS system. Actin5C is an ubiquitous promoter frequently used to drive the expression of the GAL4 protein. Adapted from ([140]). ! 33! 2.3 The FRT system In this strategy, the FRT sites are placed at identical positions within the chromosome arms of non-sister chromatids of the homologs ([141]). Flipase expression is induced by heat-shock, by placing the flies at a restrictive temperature. FLP will then recombine the FRT specific sites, which allows the exchange of the chromosome arms. If this recombination occurs after S-phase, two subsequent different populations of cells will be created as could be visualized in Fig. 3. One cell population will inherit both mutated alleles, becoming homozygous to the mutation. The other clonal population will be inherited both GFP genes, becoming wild-type relatively to the previous mutated allele. The wild-type clones are marked by GFP expression, while the mutated clones are marked by the absence of GFP, allowing clonal identification (Fig. 2.3). Figure 2.3 - Schematic representation of the FRT system operation during mitosis. 2.4 Immunofluorescence in Drosophila ovaries Drosophila females were fed with yeast and when necessary placed at 29ºC to enhance the expression of the genes of interest. Ovaries were dissected in PBT (PBS + ! 34! 0.2% Tween (Sigma-Aldrich)) and fixed for 20 minutes with a solution of 4% paraformaldehyde (Electron Microscopy Sciences) in PBT. Upon fixation the ovaries were washed 3x10 minutes with PBT, blocked with PBT-10% (PBT + 10% BSA) for 1 hour and incubated with the primary antibody in PBT-1% (PBT + 1% BSA) overnight at room temperature. 4x30 minutes washes with PBT+1% followed the removable of the primary antibody. Then the ovaries were incubated with the secondary antibody in PBT 0,1% (PBT + 0,1% BSA) during 2 hours. After the removal of the secondary antibody, the ovaries were washed 3x10 minutes with PBT and mounted in vectashield with DAPI (Vector Laboratories, Inc. Burlingame, CA84010). To perform the phalloidin (Molecular Probes) stainings, the ovaries were fixed for 30 minutes in the same solution and phalloidin was added during the fixation procedure. The subsequent procedure was the same. 2.5 Immunofluorescence on Drosophila eye/antennal imaginal discs Larvae were picked at the third instar stage and the discs dissected in PBS (1x). Posterior fixation in 4% formaldehyde in PBS last for 20 minutes, upon which 3x10 minutes washes in PBT’ (PBS+0,1%Triton (Sigma-Aldrich)) were performed. The primary antibodies were incubated in PBT’ for 2 hours at room temperature or overnight at 4ºC. 3x30 minutes washes preceded the incubation with the secondary antibodies in PBT’ during 1 hour and 30 minutes at room temperature or overnight at 4ºC. The discs were washed 3x10 minutes after removal of the secondary antibody and then vectashield with DAPI was added. The discs were mounted in 50% glycerol in PBS. 2.6 Clonal analyzes In order to generate clones within the follicular epithelium that would allow the expression of specific alleles, female flies were placed at 37ºC for two hours during 3 days. This increase in temperature will allow the expression of the Flipase protein both in FRT and FRT/FLP systems, thereby generating the clones. 2.7 Drug-induced treatments To induce an increase in the quantity of mitotic cells within the follicular epithelium, Drosophila females starved for 12 hours, upon which were fed with yeast in 500 µL of 20 µM colchicine (Sigma-Aldrich) during 17-to-19 hours. Ovaries were then dissected following the same procedure previously described. ! 35! 2.8 Primary and Secondary antibodies The primary antibodies used were: anti-aPKC (rabbit polyclonal;1/500) from Santa Cruz Biotechnology; anti-Arm (mouse monoclonal; 1/100) from Developmental Studies Hybridoma Bank (DSHB); anti-Discs large (mouse monoclonal;1/100) from DSHB; antiElav (rat; 1/100) from DSHB; anti-p-Baz (rabbit polyclonal from Eurico Morais-de-Sá et al., 2010); and pH3 (rabbit polyclonal; 1/500) from Upstate. The secondary antibodies used were Alexa 488, Alexa 594, Alexa 647 from mouse, rabbit and rat (Molecular Probes). 2.9 Imaging For time lapse imaging, ovaries were dissected into carbon oil and imaging at 25ºC using a spinning disk confocal microscope (Andor Revolution XD) with an electron multiplying charge-coupled device camera (iXonEM+; Andor) and a CSU-22 unit (Yokogawa) based on an inverted microscope (IX81; Olympus). A 63x objective was used and imaging was performed using two laser lines - 488 and 561 nm - for the excitation of GFP and RFP respectively. Z stacks were acquired every 30 seconds to all movies. Acquisition parameters, as exposure time or steps, were controlled by iQ software (Andor). Image processing and movie assembly was processed using Fiji. To perform imaging of fixed tissues we used an inverted laser scanning confocal microscope (Leica TCS SP5 II) using a 40X water objective and the LAS 2.6 software. To the imaging four lasers were used: a 405nm Diode laser; the 488 nm Argon laser; the 561 nm DPSS; and the 633 nm HeNe. Images were preocessed using Fiji. ! ! 36! CHAPTER 3 – RESULTS Cooperation between cell division and apico-basal polarity defects in tumorigenesis induction 3. Using the eye/antennal imaginal discs 3.1 Introduction Drosophila has been used as a model to address the underlying mechanisms of cancer development and progression once that a high conservation of signaling pathways between Drosophila and mammals is observed (reviewed in [34]). In turn, epithelial cancers are the most common in humans (reviewed in [142]), which means that these highly polarized cells are susceptible to give rise to tumors. Furthermore, several studies related mutations in polarity proteins to mammalian tumor development (reviewed [143]). Besides loss of cell polarity, mitotic defects are also related to tumorigenesis as referred in the introductory chapter (reviewed in [130]). Tumor development relies in the acquisition of sequential defects that allow the development of each cancer hallmark. Thus, we aimed to determine if polarity and mitotic defects could cooperate in tumorigenesis, using the follicular epithelium and eye/antennal imaginal discs from Drosophila as working models. Scribble (Scrib), Discs-large (Dlg) and Lethal giant larvae (Lgl) had been described as tumour suppressor genes in Drosophila, being also involved in epithelial cell polarity regulation ([73]). Studies reported that ectopic cell proliferation posterior to the morphogenetic furrow is observed in scrib mutants eye discs ([144]). Through interactions with the oncogenic Ras-Raf pathway, loss of scrib leads to the downregulation of the Hippo signaling pathway that is required to control tissue overgrowth ([145]; [144]). Furthermore, it was shown that scrib mutant cells are outcompeted by wild-type cells, inducing apoptosis in neighboring mutant cells. Therefore substantial overgrowth is only observed when all the tissue is mutated ([146]). The eye imaginal discs are widely used in cancer research. As previously referred, the eye/antennal imaginal discs develop in a short period of time. Thus we can observe the effect of interactions between altered levels of proteins on the tissue size and organization just in a few days. In addition, as we can induce errors on distinct mitotic processes, several divisions might be required to allow us to see the global effect on tissue organization, which is possible due to the high proliferation that occurs in the eye imaginal disc. ! 37! To carry out this study we induced the expression of a transgene expressing a Scribble inverted repeat (scribRNAi) that downregulates scrib levels by RNA interference, using the GAL4-UAS system. A specific driver - the eyeless promoter - drive the expression of the GAL4 protein, that in turn binds specifically to the UAS sequence upstream of the scribRNAi transgene activating its expression specifically on the eye imaginal disc. To search for interactions, several RNAis and protein overexpression transgenes that affect different mitotic processes were added to the system. The list of the different proteins that were under study is presented below (Table 3.1). It is important to note that the eyeless promoter was recombined with the Green Fluorescence Protein (GFP) allowing us to label the cells that express the different transgenes. Table 3.1 List of the interacting proteins sorted by the mitotic process that they are affecting. General Regulators Cytokinesis Centrosome related Kinetochores related SAC Aurora A Anillin Abnormalspindle CENP-C Bub3 Aurora B Fascetto Centrosomin Kmn1 BubR1 Orbit Pavarotti SAK Mis12 Mad1 Cdc2 Peanut Sas-4 Nuf2 Mad2 Polo-like kinase Septin1 Mitch UAS-Aurora A Septin2 UAS-Polo Tumbleweed SMC1 SMC5 3.2 scribRNAi causes a reduction in the size of the Drosophila eye Initially, we tested two RNAi constructs directed to the polarity proteins Scrib and Dlg, both driven by the eyeless promoter and raised at 25ºC. Although previous reports suggested a similar phenotype after Scribble and Dlg depletion, we did not observe that with the transgenes used, possibly due to different levels of the knockdown efficiency had a similar effect. While dlgRNAi had no effect on the organization of the eye, expression of scribRNAi leads to smaller eyes with a reduced number of ommatidia when compared to the ! 38! wild-type (Fig. 3.1). This reduced eye size is probably due to cell death induced by normal cells that did not have a sufficient Scribble depletion. As the RNAi construct does not produce complete depletion of Scribble protein, this method generated a reduction in the efficiency of the polarity machinery. This weak phenotype allows therefore the identification of interactors that induce increased defects in eye organization. Figure 3.1 - scribRNAi reduces the normal Drosphila eye size. 3.3 Cytokinesis and Aurora A altered levels have an interaction with polarity defects to induce eye outgrowth A fly stock containing both the driver and scribRNAi was constructed, which was then crossed with the several transgenes to identify potential interactors. Thus, flies expressing both scribRNAi and the modulator of the levels of the interacting genes could be chosen as indicated by the following scheme (Fig. 3.2). In addition, we also choose the sibling progeny that had no scribRNAi as internal controls to check if the altered levels of the tested interactors could produce phenotypes on their own. The phenotypes presented by the progeny were then classified into the categories that are shown in Figure 3.3. Drosophila’s head is essential to the hatching process. We observed that not all the progeny were viable revealing that the crucial role of some proteins impairs cell survival, leading to partial or complete loss of the head. In fact, we observed the existence of headless dead flies in their pupae. The table below summarizes our data relatively to progeny lethality (Table 3.2). Previous studies suggested that differences exist between males and females replicative times during S-phase, showing that differences between males and females are beyond the sexual determinants ([147]). Indeed, we also verified ! 39! that the expressivity of the phenotypes for the same transgene were different in males and females. In this screen, males seem to be more susceptible to tumor development than females since the same interaction could result in no phenotype in females while males show a strong phenotype. Therefore, we quantified males and females phenotypes individually. Figure 3.2 - Crossing scheme for the eye screen for interactions between scribRNAi and mitotic defects. Table 3.2 Progeny lethality in males and females. Semi-lethal phenotype includes the progeny that have 5 or less viable adults expressing the indicated RNAi transgenes per cross. The other tested transgenes were all viable with more than 5 flies [10 eyes] per cross. Differences were found between males and females in what concerns to the lethality of certain transgenes. Males Females Lethal Semi-lethal Lethal Semi-lethal orbit aurora B aurora B mis12 cdc2 bub3 orbit smc1 cenp-c mad1 cdc2 tumbleweed kmn1 mis12 cenp-c peanut mitch kmn1 polo pavarotti peanut smc1 sas-4 polo tumbleweed ! 40! Figure 3.3 Representative pictures of each category used to score the eye phenotype. For a more precise quantification, an increase in the number of ommatidia maintaining the normal structure of the eye was classified as overgrowth, whereas eye masses popping out of the eye were classified as outgrowth. As the main interest of this study was to test the development of tumor like structures due to depletion of interacting genes, we present a graph of the outgrowth phenotype observed for each genetic background (See Fig. 3.4, Graph. 3.1 and 3.2). A table with all scored phenotypes is presented in the appendix. In general, we observed that depletion of proteins with functions on cytokinesis lead to outgrowth phenotypes when Scribble protein was simultaneously depleted. In addition, we also found a high percentage of outgrowths when we disturbed Aurora A protein levels in a polarity defective background. ! 47! 3.5 Aurora A overexpression induces tumorigenesis by promoting photoreceptor differentiation failure Aurora A has several reported functions including involvement in centrosome maturation and separation, establishing of spindle bipolarity and a role in the organization of apical polarity in Drosophila neuroblasts ([102];![120]; [98]). As a consequence, its absence causes many defects, such as monopolar spindles due to its role on centrosome separation and consequently the production of tetraploid cells ([113]). As previously referred, tetraploidy is frequently found in tumors, a fact that might provide insights on how Aurora A causes the outgrowth phenotype previously observed (see graphics 3.1 and 3.2). Additionally Aurora A overexpression was also implied in the generation of multinucleated cells and aberrant mitosis in other organisms ([155]; [156]). Aurora A altered levels could therefore lead to tumor development. Indeed, our observations suggested that when the polarity machinery is weakened, both aurARNAi and UAS-AurA causes tumorigenesis. Therefore, we also performed immunofluorescence analyzes in aurARNAi and UAS-AurA transgenic discs, in the presence and absence of scribRNAi, using both Elav and pH3 antibodies. aurARNAi per se did not result in significant alterations in the organization of the disc when compared with the wild-type, but some photoreceptor disorganization could be seen (Fig. 3.7A). However, when in a defective polarity background, aurARNAi eye imaginal discs show an increased number of cells labeled with pH3, indicating that some cells that should be already differentiated are undergoing mitosis (Fig. 3.7B). In the case of UAS-AurA, it shows no differences in comparison with the wild-type cells (Fig. 3.7C). Yet, when scribRNAi is added to the system, the normal morphology of the discs is lost. It should be noted that we observed a correlation between high expression of the transgenes (as indicated by the higher levels of GFP) and absence of Elav labeling, indicating that the high levels of AurA block cell differentiation (Fig. 3.7D). Furthermore we observe multiple layers of cells, which could be an indicative of cell overproliferation, explaining the observed outgrowth phenotypes (Fig. 3.7E). In addition, pH3 labeling shows random proliferating cells throughout the entire eye/antennal imaginal disc suggesting that the MF is not correctly formed (Fig. 3.8). ! 48! Figure 3.6 Simultaneous anillin and scribble downregulation leads to reduced photoreceptor differentiation. Third instar larvae eye discs with the most posterior part at the bottom of the figure in all discs. DAPI labels the DNA, anti-Elav the differentiated cells and pH3 the cells that are undergoing mitosis. The white bar marks the morphogenetic furrow. ! 49! ! Figure 3.7 UAS-AurA causes differentiation failure. Third instar larvae eye discs with the most posterior part at the bottom of the figure in all discs. DAPI labels the DNA, anti-Elav the differentiated cells and pH3 the cells that are undergoing mitosis. GFP indicates the level of expression of the transgenes. The white bar marks the morphogenetic furrow. ! 50! Figure 3.8 UAS-AurA, scribRNAi causes random proliferation through the entire imaginal disc. In the previous present UAS-AurA transgenic discs, anti-pH3 was also used to label cells that were undergoing mitosis. Proliferation is observed randomly distributed within the imaginal disc. DAPI marks the DNA while Elav labels the differentiated cells. 4. Using the follicular epithelium from Drosophila ovaries ! ! 4.1 Background Drosophila oogenesis is a complex process that involves the establishment of anterior-posterior and dorsal-ventral axis of Drosophila by localizing RNAs specifically within the oocyte ([157]; [158]). As described in the introductory chapter, the ovaries are divided in functional “subunits”, the ovarioles, within each the egg is produced. Each ovariole contains multiple egg chambers at different stages of oogenesis. Each egg chamber develops along 14 stages, classified as early (1 to 6), mid (7 to 10) or late (11 to 14) stages of oogenesis (reviewed in [36]). The early stages of Drosophila oogenesis have particular importance to this work, since it is within these stages that the follicle cells chamber are undergoing mitosis. Follicle cells cease dividing at the end of the sixth stage. Thus, it possible to perform live imaging studies of division in follicular epithelial cells only until stage 6 of oogenesis. Using another epithelial system, the follicular epithelium, we could validate the previously found interactions in the initial eye screen. In comparison with the eye imaginal discs, the follicular epithelium offers us several advantages, including the possibility of observing invasion of the tumour-like cells, which is not possible in the eye imaginal disc. ! 51! Since the epithelial cells are in intimate contact with the germline, when mutations leading to the disorganization of the epitelial monolayer occur, we could observe epithelial cells streaming within the germline environment. Furthermore, imaging of the apico-basal polarity is straightforward in the follicular epithelium, which is compatible with the use of a variety of genetic and molecular tools. Thus, it is a good epithelial model to understand the underlying mechanisms of tumour development, such as factors that can result in tissue disorganization. 4.2 dlgRNAi causes tissue disorganization in the follicular epithelium To start the screen, the phenotypes present by RNAi depletion of the polarity proteins Scribble and Dlg were addressed. Both RNAis transgenes were driven by the actin promoter, however as the loss of TSGs leads to lethality, the flies were raised at 18ºC using the GAL4/UAS system coupled with the repressor of GAL4, the GAL80ts protein. To induce the expression of the RNAi transgenes, flies were placed during 2 days at 29ºC, so that GAL80 repression was blocked. While scribRNAi presented a weak phenotype, showing just two or more individual cells out of the monolayer, dlgRNAi shows groups of cells forming a double layered and disorganized tissue. This was classified as an intermediate phenotype (Fig. 4.1). This phenotype is adequate for the propose of the study since that enables us test if for interactions that can increase the aggressiveness of the phenotype. To confirm that the RNAi against Dlg is being effective, we used the Flpout system to make clones within the follicular epithelium where the dlgRNAi construct driven by the actin promoter is exclusively expressed. The clones expressing the RNAi transgene are marked by the concomitant expression of the GFP protein. As can be seen in Fig. 4.2, Dlg is highly down regulated in clonal cells when compared with the wild-type cells marked by the absence of GFP, showing that the dlgRNAi construct is depleting the Dlg protein. ! 52! Figure 4.1 Weak and intermediate phenotype present by scribRNAi and dlgRNAi, respectively. After two days at 29ºC to inactivate GAL80ts supressor, scribRNAi and dlgRNAi present a weak and intermediate phenotype, respectively. DAPI marks DNA, while phalloidin labels the actin. Figure 4.2 dlgRNAi depletes efficiently the endogenous Dlg protein. The yellow lines mark the clones were the anti-Dlg antibody detects very low amounts of protein. This can be compared with the wild-type cells marked by GFP absence. ! 53! 4.3 Anillin depletion and Aurora A overexpression enhance the intermediate phenotype presented by dlgRNAi The same proteins involved in the mitotic processes mentioned in the eye/antennal imaginal discs screen were also tested in the follicular epithelium (Table 3.1). The dlgRNAi construct was driven by the actin promoter using the system GAL4/UAS system in combination with its repressor, the GAL80ts protein that enables flies to born. A stock containing the actin promoter and dlgRNAi was constructed and the progeny was chosen as indicated in Fig. 4.3. The phenotypes observed were classified into 4 main categories. The “None” phenotype means that no disruption of the epithelial monolayer was observed. The “Weak” phenotype was assigned when one or a few individual cells were seen away from the monolayer. “Intermediate” phenotype corresponds to egg chambers where a group of cells form a second layer. “Strong” phenotype is observed when more than two layers of epithelial cells are seen invading the germline. Examples of the observed phenotypes are shown in Fig. 4.4. Figure 4.3 Schematic representation of the progeny selection for the follicular epithelium screen. The flies were raised at 18ºC once that expression of the RNAi construct in the whole organism is lethal. Upon progeny selection, the flies were placed for four days at 29ºC to disrupt the GAL4 inhibitor - GAL80ts. This screen revealed that depletion of a few mitotic genes is occasionally able to enhance the phenotype of dlgRNAi (Fig. 4.5). However, as these results are still preliminary, we will just focus on the protein interactions that were shown to give a strong phenotype in the eye/antennal imaginal disc screen. Therefore, Aurora A and Anillin were the proteins chosen to proceed to quantification. Flies with RNAi constructs against these proteins show no tumour-like phenotypes per se. The same was tested in relation to UAS-AurA and no invasion was observed. As observed in Fig. 4.6, only two of the tested conditions - ! 54! UAS-AurA2 and anillinRNAi - seem to enhance the tissue disorganization of the polarity defective background tested (see Fig. 4.4). Double anillinRNAi ,dlgRNAi resulted in the strongest phenotype observed in the follicle epithelium, resulting frequently in multiple layers of cells within the germline (Fig. 4.6). ! ! Figure 4.4 Representation of the three types of phenotype that were classified in the follicle epithelium. We show the effect of the combination of defective levels of both Anillin and Dlg as an example of a strong phenotype. Yellow arrows show the localization of invasions as well as the bar. DAPI is labelling the DNA in red. ! ! On the other hand, depletion of AurA does not seem to interfere with the phenotype of dlgRNAi. As can be seen in Graph 4.1, aurARNAi, dlgRNAi have a similar penetrance of the intermediate phenotype relative to dlgRNAi alone. In turn, expression of UASAurA leads to a larger frequency of intermediate phenotypes in the presence of dlgRNAi, comparing with dlgRNAi per se. Thus, overexpression of Aurora A seems to be inducing higher levels of epithelial cell invasion when the follicular epithelium presents defects in the polarity machinery. anillinRNAi, dlgRNAi revealed the presence of several strongly affected ovarioles, confirming that also in this ovarian epithelial context, cytokinesis cooperates with defects in epithelial polarity to induce tissue disorganization, which is a characteristic of tumor malignancy (Graph. 4.1). ! ! 55! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! Figure 4.5 Examples of the strongest phenotypes observed after alteration of the levels of a mitotic protein in a polarity defective background. The yellow lines are indicating the most affected areas. Phalloidin is labelling the actin filaments while DAPI marks the DNA. ! ! Figure 4.6 Examples of the observed phenotypes of anillinRNAi and Aurora A overexpression when combined with dlgRNAi. When Aurora A is overexpressed in a polarity defective background, we observe an increase in the levels of intermediate phenotypes. In turn, anillinRNAi , dlgRNAi shown an average percentage of strong phenotypes in the total quantification. ! 56! ! Graph 4.1 Quantification of the invasion phenotype per ovariole. anillinRNAi gave the strongest interaction with dlgRNAi while Aurora A defective levels seems to have no interaction with defective levels of Dlg. ! 63! Figure 5.5 - Par-6 phosphorylation is dispensable for the maintenance of apicobasal polarity in follicle cells. The non-phosphorylatable form of Par-6 rescues mislocalization of the apical proteins in par-6 mutants. 5.6 Lgl dynamics during mitosis on the follicular epithelium The conservation among the proteins that govern cell polarity is well known. Lgl exclusion from the posterior cortex in SOP cells is essential for cell fate determination since that it exclusion allows the association of Bazooka with the PAR Complex and thereby Numb phosphorylation ([98]). Our results suggested an independence of Aurora A kinase activity in epithelial polarization. However, given the conservation of the proteins among different organisms, it is important to determine if the underlying mechanism of Lgl dynamics is also essential for epithelial cell polarization, given that Aurora A kinase activity is not. The behaviour of Lgl in SOP cells during mitosis is well known: Lgl starts to be excluded from the cortex in early prophase, remaining in the cytoplasm during mitosis ([98]). A form of Lgl that is non-phosphorylatable by aPKC - the Lgl3A – remains localized in the cortex during cell division, suggesting that Lgl exits from the cortex in a phosphorylation-dependent manner ([98]). Accordingly, we wanted to know how Lgl behaves during mitosis in epithelial cells. We followed cell division using His-GFP to mark DNA, together with Lgl-RFP. We concluded that Lgl starts to exit from the cell cortex ! 64! around 21 minutes prior to Nuclear Envelope Breakdown (NEBD) remaining in the cytoplasm during division and returning to the cortex after cytokinesis (Fig. 5.6 and Graph. 5.2). Figure 5.6 - LglWT-GFP behaviour during mitosis in epithelial cells. Frames were taken from movies that show Lgl:RFP dynamics during mitosis. His:GFP enables us to determine the exact moment of anaphase onset. We took this time as a start point to quantify when Lgl exits and then returns to the cortex. Graph 5.2 - Lgl and BazS980A dynamics during mitosis. Lgl starts to exit from the cell cortex 21.3 ± 2 min (N=6) prior than anaphase onset and returns 8.4 ± 2 min (N=6) after anaphase onset. In turn, BazS980A starts to depolarize in average 8.1 ± 3 min (N=4) prior than anaphase onset while it fully returns to the apical mesh in 10.7 ± 2 min (N=6) after AO. This results shows that BazS980A depolarization starts after Lgl cortical exclusion. Accordingly, to determine whether Lgl cortical exclusion is phosphorylation dependent, we performed immunostaining in cells from flies with LglWT-GFP and Lgl3AGFP using anti-pH3 to mark the cells that were undergoing mitosis. Our observations ! 65! show that LglWT-GFP localizes at the cell cortex during interphase, however, in cells that were labelled with pH3, LglWT-GFP was diffused in the cytoplasm (Fig. 5.7A). In turn, the non-phosphorylatable form (Lgl3A-GFP) remained in the cortex during interphase and mitosis (Fig. 5.7B). These results suggest that during mitosis, Lgl has a similar behaviour in SOP cells and in epithelial cells, being excluded from the cortex in an aPKC phosphorylation dependent manner. Figure 5.7. Lgl is excluded from the cortex in a phosphorylation dependent manner. The non-phosphorylatable form of Lgl remains cortically during all cell cycle (B) when comparing to the wild-type (A). Anti-pH3 labels cells that are undergoing mitosis. Frames from movies of Lgl3AGFP show that this mutant variant remains in the cortex during mitosis (C). 5.7 Using BazS980A to address the dynamic behaviour of apical proteins during mitosis in epithelial cells After characterizing the dynamics of Lgl during mitosis, we asked how would the apical domain behave. Given that Lgl cortical exclusion occurs in a phosphorylation- ! 66! dependent manner, lateral exclusion of Lgl would be expected to be preceded by a reorganization of apical proteins. In order to determine the localization of apical proteins during mitosis, we used colchicine to increase the percentage of mitotic cells within the follicular epithelium. We performed immunostainings using flies that expressed Par-6GFP, to address the localization of the apical complex during mitosis. As can be seen in Fig. 5.8, the nucleus of the follicle cells are labelled with pH3, indicating that many cells are undergoing mitosis. Both the anti-aPKC antibody and the endogenous Par-6-GFP revealed that the apical proteins present different localizations in the cells that were labelled with pH3. pH3 marks the phosphorylation of the Histone 3 since the beginning of mitosis – prophase – which means that all the cells that are in prophase or the subsequent phases are labelled with pH3. Since, we observe that in some cells, indicated by the yellow arrow, aPKC and Par-6 are distributed essentially through the cytoplasm while in others (marked by the green arrow) aPKC and Par-6 remain in the apical mesh. This result suggests that the apical complex has a specific time during mitosis within which depolarization starts. Figure 5.8 - Apical Complex localization in a mitotic enriched follicular epithelium. Females were fed with colchicine, which blocks microtubules depolymerisation, thereby arresting cells in metaphase. pH3 staining (red) shows cells that are undergoing mitosis. The Par-6-GFP (green) construct and the anti-aPKC antibody (red) show the subcellular localizations of the apical complex. To characterize the dynamic of depolarization of the apical complex during mitosis, we performed live imaging. In the absence of a functional fluorescent tagged aPKC protein, we used the non-phosphorylatable form of Bazooka (BazS980A-GFP) to mark the localization of apical proteins throughout mitosis. Since it cannot be phosphorylated by aPKC, BazS980A is not excluded from the apical complex. It was surprising to find that BazS980A starts to extend through the cortex around 8 minutes before NEBD (Graph. 5.2). ! 67! Furthermore, complete re-localization of BazS980A to the cortex occurs after cytokinesis around the same time that LglWT-GFP relocalizes to the cell cortex (Fig. 5.9). Differences were seen between Par-6-GFP and BazS980A localization during mitosis. While Par-6-GFP presents a cytoplasmic localization, BazS980A extends through the cell cortex. Figure 5.9 - BazS980A-GFP extends through the cortex during mitosis. Frames from movies of BazS980A-GFP (that is shown in red) showing its cortical extension occuring after LglWT cortical exclusion. The apical-basal axis orientation is shown to each figure. This extension occurs after LglWT cortical exclusion. BazS980A dynamic times during mitosis are shown in Graph. 5.2. ! 68! CHAPTER 4 – DISCUSSION AND CONCLUSION Several mitotic defects have been associated with the induction of cellular transformation presumably by promoting aneuploidy, a chromosomal defective condition that is frequently found in tumors ([130]). On the other hand, loss of cell polarity was also associated with cancer progression since it causes tissue disorganization and cell overproliferation ([162]). These observations provide the basis for the main aim of this work: to search for mitotic interactors that would enhance tumorigenesis in a polarity defective background. To accomplish this, we used two different epithelial models from Drosophila: the eye/antennal imaginal discs and the ovarian follicular epithelium. As described in the results chapter, two proteins, Anillin and Aurora A, were identified as possible interactors that severely enhanced the tumor-like phenotype of these cells when cell polarity is compromised. 6.1 Cytokinesis failures can drive tumorigenesis in a polarity defective background Anillin is a scaffold protein, whose depletion results in cytokinesis failure both in Drosophila and humans, resulting in tetraploidy ([10]). The expression of anillinRNAi within a polarity defective background in both epithelial models revealed an enhancement of the tumor-like phenotype when compared with the respective controls. This result suggests that cytokinesis failure cooperates with polarity defects to induce tumorigenesis. However, incomplete cytokinesis also leads to the accumulation of centrosomes within a single cell. The increasing number of centrosomes could generate multipolar spindles at the time of division, resulting in the formation of aneuploidy cells (reviewed in [130]). Therefore, anillin depletion could be enhancing tumorigenesis due to the formation of multipolar spindles and not due to its role in cytokinesis. To address which was the driving force that was enhancing tumorigenesis after anillin depletion, we depleted sas-4 in the context of scribble and anillin depletions, providing an acentrosomal background to these cells. Our observations suggest that anillin depletion induces tumorigenesis independently of centrosome accumulation, since the reduction of the levels of sas-4, and consequently the absence of centrosomes, was irrelevant to the enhancement of the tumor-like phenotype. However, the levels of protein depletion caused by using sas4RNAi need to be verified to ensure that centrosome duplication is indeed impaired. Furthermore, studies using sas4 null mutants associated with the deficiency for the sas4 locus would be required in order ! 69! to ensure that centrosomes are completely absent from the epithelial tissue. Nevertheless, our results indicate that, polyploidy is most likely the cause underlying the enhancement of tumor-like phenotypes produced by anillin RNAi. Consistent with this, the direct tumorigenic potential of tetraploid cells has been previously documented. Induction of cytokinesis impairment into p53-/- mouse mammary epithelial cells (MMECs) leads to the formation of tetraploid cells ([131]). Carcinogenic experiments in vitro have shown that tetraploid cells were able to be transformed and to induce malignant pathologies after removal of the carcinogen. Consistently, other studies provided evidences that although extra centrosomes can induce tumorigenesis in specific fly tissues, they do not always lead to aneuploidy and genetic instability, given that extra centrosomes appear to coalesce, allowing the formation of mostly bipolar spindles in these abnormal cells ([163]). Although mutations on Aurora A gene seem to be insufficient to initiate the malignant pathology, Aurora A has long been associated with tumour formation in different organisms ([164]; [165]; [156]). The reduction of Aurora A levels causes defects on the mitotic spindle due to the impairment of centrosome maturation and separation, leading to the formation of tetraploid cells. In turn these tetraploid cells are frequently found in cancers suggesting that tetraploidy promotes genomic instability, which is a wellknown hallmark of cancer (reviewed in[159]). Consistently, we observed that the expression of aurARNAi induced outgrowth phenotypes within the Drosophila eye. However, the observation of larvae imaginal discs did not allow us to conclude what happens at the cellular level, since the morphology of the disc was not significantly altered when compared with the wild-type, regardless an increase in cell proliferation throughout the whole disc. Further studies will aim to elucidate if downregulation of AurA resulted in tetraploidy in the larvae imaginal discs. On the other hand, the expression of aurARNAi in the follicle cells whose polarization is compromised, did not result in a significant increase of invasion into the germline, although some intermediate phenotypes are observed. One possible explanation is that the depletion by expression of RNAi constructs is variable. Additionally, in Drosophila early stages of oogenesis, the expression of the RNAi constructs is low and epithelial cells divide just until stages 6/7. Therefore, a possible explanation for these results relies on the weak depletion of the protein levels by aurARNAi with the few divisions of the epithelial cells that did not allow the accumulation of errors and consequently cell transformation and invasion. The imaginal discs are dissected around 6/7 days after egg laying but the Drosophila eye can only be seen after the fly hatches, which means about 10 days after fertilization. As the RNAis have a variable expression and as efficient protein depletion takes several days to be achieved, it is possible that the effect seen in the adult eye is stronger than in the imaginal discs. Furthermore, as the drivers expressing the RNAis constructs in both epithelial models are ! 70! different, the level of depletion of Aurora A achieved within both systems is different. Moreover, the expression of the RNAi construct in the eye is longer than in the follicular epithelium due to the large development period of the eye. This may explain why we observe that Aurora A cooperates with compromised polarity in the eye but not in the follicular epithelium. During both screens we also noticed that UAS-AurA caused an enhancement of the tumour-like phenotype in cooperation with defects in the polarity machinery. Although differences between males and females were significant in the eye, the overexpression of Aurora A lead to a substantial outgrowth in association with a high percentage of overgrowth phenotypes in the adult flies eyes. Further analyses of the larvae imaginal discs have shown that overexpression of Aurora A induced a failure in photoreceptor differentiation along with a random cell proliferation throughout all disc regardless MF constriction. Likewise, overexpression of this mitotic kinase regulator promoted an increase in the intermediate phenotype on the follicular epithelium when compared with the polarity defective control. However, this phenotype was not as aggressive as in the eye. This could be explained by fewer divisions of the epithelial monolayer, which decrease the accumulation of errors that would be required to allow cell transformation. It has been described that Aurora A overexpression induces cytokinesis failures and p53 inactivation and degradation ([118]; [166]). By promoting cytokinesis failure, Aurora A promotes the formation of tetraploid cells that would normally arrest in the next G1-phase ([167]; [168]; [169]; [170]). However, by destabilizing p53 through phosphorylation, the p53-dependent checkpoint that prevent tetraploid cell division is weakened, enabling tetraploid cells to divide and proliferate ([155]; reviewed in [160]). Therefore, the underlying mechanism by which UAS-AurA cooperates with polarity defects probably relies on the ability of Aurora A (when overexpressed) to promote cytokinesis failure resulting in tetraploidy. Phosphorylation and consequently degradation of p53 induced by Aurora A is also observed at physiological levels, however this effect is increased upon Aurora A overexpression. This may explain why cells depleted of Aurora A have a less consistent tumour-like phenotype, since the p53-checkpoint is not weakened as a result of decreased levels of Aurora A phosphorylation ([166]). 6.2 Aurora A is not required to the establishment of apicobasal polarity in epithelial cells Studies performed in Drosophila neuroblasts have shown that Aurora A kinase activity is essential to differentially segregate Numb into the GMC. Aurora A phosphorylation on Par-6 relieves aPKC suppression, and therefore activates it. aPKC ! 71! activation is essential to neuroblast polarization during its asymmetric division to phosphorylate Numb, allowing its specific localization into the GMC and therefore cell differentiation ([98]). Our results show that Aurora A cooperates with polarity defects to induce tumorigenesis, suggesting that it might also have a role in the establishment of epithelial polarity. Therefore, we wanted to determine whether Aurora A is also involved in epithelial polarization. We started by expressing a specific RNAi construct against Aurora A, thereby decreasing its levels in the follicular epithelium. As no defects of the apical-basal polarity were observed, we used a hypomorphic allele combined with the deficiency for the locus of Aurora A to decrease even further its protein levels. Our results suggested that decreased levels of Aurora A had no influence on the establishment of epithelial apicobasal polarity. However, it was still possible that incomplete depletion of Aurora A by the RNAi transgene was enough to support the normal functions of the mitotic kinase. Therefore, we used a kinase dead allele to address if its kinase activity was required for the maintenance of epithelial polarity. Immunostainings using phospho-specific antibodies revealed that Aurora A kinase activity was not required for aPKC apical localization or activity, as shown by the wild-type localization of anti-p-Baz antibody within the GFP clones. Therefore, our results suggested that Aurora A does not have a major role in the establishment of follicle cell polarity. Several hypotheses could be raised to explain why Aurora A activity is dispensable for follicle cell polarity. Looking at aPKC kinase activity essentiality along the establishment of apico-basal polarity during epithelial interphase, it is reasonable to propose that at some point its kinase activity would be activated to carry out aPKC specific functions. This activation could be performed by a kinase other than Aurora A, thereby maintaining part of the mechanism identified in neuroblasts. Other possibility could rely on activation by Cdc42 during interphase, since it was described in mammals that although Par-6 is the responsible for aPKC localization, Cdc42 modulates it kinase activity ([171]; [55]; [52]). Finally, aPKC could be constitutively active throughout the cell cycle and thereby any activation would be dispensable. Nevertheless, follicle cells seem to undergo depolarization during mitosis, and therefore it is possible that the activity of aPKC is dispensable at this stage. Given that the apical and basolateral domains are depolarized during mitosis, aPKC activity may not be needed to restrain their locations during interphase. ! 72! 6.3 Possible role of Lgl in driving mitotic depolarization of epithelial cells To understand if the mechanisms triggered by Aurora A phosphorylation in neuroblasts/SOPs were also conserved in epithelial cells, we analysed the dynamic behaviour of Lgl during mitosis. As in SOP cells, Lgl is cortically excluded during mitosis in a phosphorylation-dependent manner as a form of Lgl (Lgl3A-GFP) that is nonphosphorylatable by aPKC remains cortical during all cell cycle. In SOP cells, Aurora A phosphorylation induces aPKC activation whose kinase activity is responsible for Lgl exclusion from the cortex ([98]). However, despite our results demonstrating that Aurora A function is not required in epithelial cells, aPKC still phosphorylates and excludes Lgl from the apical domain during interphase, whereas it is known that Lgl is able to repress aPKC activity if stably bound to it ([81]; [79]; [96]). Therefore, we wanted to investigate how the apical domain behaves during mitosis in epithelial cells. It has been previously found in the lab [Eurico Morais de Sá and Claudio Sunkel, unpublished data] that the proteins Par6 and aPKC lose their polarized apical distribution during mitosis. Increasing the frequency of mitotic follicle cells (using colchicine) in flies that carry a Par-6-GFP construct, we were able to see that epithelial cells labelled with pH3 had different subcellular localizations of both anti-aPKC and Par-6-GFP, thereby suggesting a dynamic depolarization during mitosis. Therefore we used BazS980A as a surrogate marker of the apical complex to determine in detail the dynamics of the apical complex. Our data supports that the apical complex starts to extend throughout the cortex after Lgl cortical exclusion. However, while Par-6-GFP shows a cytoplasmic localization, BazS980A depolarization was restricted to the extension through the cortex. BazS980A cortical localization could partially reflect its ability to bind Adherens Junctions components. Given that Par-6-GFP is expressed at endogenous levels, its localization is more reliable in reflecting the localization of the apical Complex. Previous results in the lab show that the depolarizing times in Par-6-GFP expressing cells, are nevertheless similar to the ones obtained with BazS980A. A timeline showing and overview of the respective timings is shown in Fig 6. ! 79! Table 5. Relative percentage of the frequency of phenotypes shown in male eyes expressing scribRNAi. RNAi/O.e% WT S-eye M-eye R-eye D-eye T-eye O-eye N-eye anillin 11,1 22,2 0 0 11, 44,4 5,6 5,6 asp 0 100 0 0 0 0 0 0 aurA 2,9 17,6 0 5,9 0 52,9 20,6 0 aurA’ 22,7 27,3 0 2,3 0 13,6 34,1 0 bub3 0 51,9 0 9,3 9,3 24,1 5,6 0 bubRI 10,4 77,1 0 0 4,2 8,3 0 0 cnn 0 100 0 0 0 0 0 0 feo 0 32,7 0 7,7 0 40,4 19,2 0 mad1 75 25 0 0 0 0 0 0 mad2 16,7 56,7 0 20 6,7 0 0 0 mis12 0 100 0 0 0 0 0 0 mitch 50 0 20 0 20 10 0 0 nuf2 20 70 0 0 0 10 0 0 pavi 0 0 0 0 0 25 25 50 sak 0 84,2 0 2,6 0 13,2 0 0 sas-4 100 0 0 0 0 0 0 0 sept1 5,3 94,7 0 0 0 0 0 0 sept2 44,4 16,7 0 0 8,3 30,6 0 0 smc1 0 0 0 0 0 0 0 0 smc5 38,9 13,9 0 16,7 0 30,6 0 0 UAS-AurA2 11,9 31,0 4,8 9,5 2,4 38,1 2,4 0 UAS-AurA3 0 21,4 0 7,1 0 35,7 35,7 0 UAS-Polo 25,7 67,5 0 0 1,4 0 5,7 0 ! 80! Table 6. Relative percentage of the frequency of phenotypes shown in female eyes expressing scribRNAi. RNAi/O.e% WT S-eye M-eye R-eye D-eye T-eye O-eye N-eye anillin 27,0 31,1 0 8,1 6,8 24,3 1,4 1,4 asp 0 90,9 0 0 0 9,1 0 0 aurA 9,4 18,9 0 18,9 0 30,2 22,6 0 aurA’ 25,8 60,6 0 3,0 0 3,0 7,6 0 bub3 12,5 50 7,5 15 2,5 12,5 0 0 bubRI 43,6 44,9 0 0 10,3 1,3 0 0 cnn 0 91,7 0 8,3 0 0 0 0 feo 14,3 53,6 0 19,6 0 0 12,5 0 mad1 27,5 67,5 0 2,5 0 2,5 0 0 mad2 20,6 79,4 0 0 0 0 0 0 mis12 25 33,3 16,7 8,3 0 0 8,3 8,3 mitch 39,3 21,4 0 14,3 14,3 7,1 0 3,6 nuf2 7,1 9,5 4,8 2,4 0 0 2,4 2,4 pavi 0 23,7 7,9 13,2 13,2 31,6 2,6 7,9 sak 0 84,5 0 6,9 0 8,6 0 0 sas-4 45,8 4,2 0 16,7 16,7 16,7 0 0 sept1 25 75 0 0 0 0 0 0 sept2 21,4 35,7 3,6 3,6 3,6 25 7,1 0 smc1 0 25 0 0 0 62,5 12,5 0 smc5 41,0 55,7 0 3,3 0 0 0 0 UAS-AurA2 35,4 49,2 0 0 7,7 6,2 1,5 0 UAS-AurA3 32,1 46,4 0 10,7 0 3,6 7,1 0 UAS-Polo 37,1 50 0 5,7 0 7,1 0 0 ! 81! Table 7. Relative percentage of the frequency of phenotypes shown in male eyes in the absence of scribRNAi. RNAi/O.e% WT S-eye M-eye R-eye D-eye T-eye O-eye N-eye anillin 43,0 29,0 3,2 17,2 7,5 0 0 0 asp 11,8 88,2 0 0 0 0 0 0 aurA 42,2 54,7 0 3,1 0 0 0 0 aurA’ 84,4 15,6 0 0 0 0 0 0 bub3 29,4 70,6 0 0 0 0 0 0 bubRI 100 0 0 0 0 0 0 0 cnn 34,1 53,7 0 4,9 0 7,3 0 0 feo 44,3 48,4 0 4,1 0 0 3,1 0 mad1 75 25 0 0 0 0 0 0 mad2 100 0 0 0 0 0 0 0 mis12 0 16,7 33,3 0 0 0 0 50 mitch 18,2 38,6 0 15,9 18,2 9,1 0 0 nuf2 100 0 0 0 0 0 0 0 pavi 0 0 0 0 0 0 0 100 sak 17,4 82,6 0 0 0 0 0 0 sas-4 95,6 4,4 0 0 0 0 0 0 sept1 68,4 31,6 0 0 0 0 0 0 sept2 64,9 33,8 0 1,4 0 0 0 0 smc1 77,8 22,2 0 0 0 0 0 0 smc5 72,7 27,3 0 0 0 0 0 0 UAS-AurA2 57,8 40 0 0 1,1 1,1 0 0 UAS-AurA3 58,8 26,5 0 0 0 0 14,7 0 UAS-Polo 48,3 51,7 0 0 0 0 0 0 ! 82! Table 8. Relative percentage of the frequency of phenotypes shown in female eyes in the absence of scribRNAi. RNAi/O.e% WT S-eye M-eye R-eye D-eye T-eye O-eye N-eye anillin 74,5 11,5 0,6 9,7 3,6 0 0 0 asp 21,1 78,9 0 0 0 0 0 0 aurA 87,5 10,6 0 1,9 0 0 0 0 aurA’ 100 0 0 0 0 0 0 0 bub3 19,2 79,4 0 1,4 0 0 0 0 bubRI 96,6 3,4 0 0 0 0 0 0 cnn 77,6 22,4 0 0 0 0 0 0 feo 68,1 23,6 0 0,7 0 5,6 2,1 0 mad1 75 25 0 0 0 0 0 0 mad2 100 0 0 0 0 0 0 0 mis12 7,7 34,6 30,8 3,8 11,5 0 0 11,5 mitch 48,5 26,4 5,9 4,4 10,3 2,9 0 1,4 nuf2 86,7 13,3 0 0 0 0 0 0 pavi 100 0 0 0 0 0 0 0 sak 22,2 77,8 0 0 0 0 0 0 sas-4 100 0 0 0 0 0 0 0 sept1 78,1 21,9 0 0 0 0 0 0 sept2 100 0 0 0 0 0 0 0 smc1 50 44,4 0 2,8 2,8 0 0 0 smc5 58,5 39,0 0 0 0 2,4 0 0 UAS-AurA2 56,9 43,1 0 0 0 0 0 0 UAS-AurA3 59,1 40,9 0 0 0 0 0 0 UAS-Polo 93,0 7,0 0 0 0 0 0 0 ! 83! Table 9. Frequency of phenotypes shown in male eyes. Table 10. Frequency of phenotypes shown in female eyes Table 11. Relative percentage of the frequency of phenotypes shown in male eyes RNAi WT Seye Meye Reye Deye Teye Oeye Neye N anillin 15 4 3 14 0 2 0 4 42 sas-4 22 20 0 0 0 0 0 0 42 anillin+aas-4 11 9 8 80 18 7 3 0 136 scrib+anillin 4 0 4 12 2 7 2 0 31 scrib+sas-4 12 30 0 0 2 1 3 0 48 scrib+anillin+sas-4 29 19 3 23 18 20 7 2 121 RNAi WT Seye Meye Reye Deye Teye Oeye Neye N anillin 47 7 5 10 3 4 1 0 77 sas-4 39 9 0 0 0 0 0 0 48 anillin+aas-4 14 7 2 106 16 3 7 1 156 scrib+anillin 20 6 1 18 3 9 4 1 62 scrib+sas-4 22 19 0 1 0 2 2 0 46 scrib+anillin+sas-4 14 12 6 55 26 23 16 0 152 RNAi WT Seye Meye Reye Deye Teye Oeye Neye N anillin 35,7 9,5 7,1 33,3 0 4,8 0 9,5 100 sas-4 52,4 47,6 0 0 0 0 0 0 100 anillin+aas-4 8,1 6,6 5,9 58,8 13,2 5,1 2,2 0 100 scrib+anillin 12,9 0 12,9 38,7 6,5 22,6 6,5 0 100 scrib+sas-4 25 62,5 0 0 4,2 2,1 6,25 0 100 scrib+anillin+sas-4 24,0 15,7 2,5 19,0 14,9 16,5 5,8 1,7 100 ! 84! Table 12. Relative percentage of the frequency of phenotypes shown in female eyes ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! RNAi WT Seye Meye Reye Deye Teye Oeye Neye N anillin 61,0 9,1 6,5 13,0 3,9 5,2 1,3 0 100 sas-4 81,3 18,7 0 0 0 0 0 0 100 anillin+aas-4 9,0 4,5 1,3 68,0 10,3 1,9 4,5 0,6 100 scrib+anillin 32,3 9,7 1,6 29,0 4,8 14,5 6,5 1,6 100 scrib+sas-4 47,8 41,3 0 2,2 0 4,3 4,3 0 100 scrib+anillin+sas-4 9,2 7,9 3,9 36,2 17,1 15,1 10,5 0 100 ! 85! References ! 1. Cheeseman, I.M. and A. Desai, Molecular architecture of the kinetochoremicrotubule interface. Nature reviews. Molecular cell biology, 2008. 9(1): p. 33-46. 2. Alberts, B., et al., Molecular Biology of the Cell. 5th ed2008, New York: Garland Science. 1539. 3. Lodish, H., et al., Molecular Cell Biology. 6th ed. Vol. 1v. 2008: Basingstoke: W. H. Freeman. 4. Morgan, D.O., The cell Cycle: Principles of Control2007, London: New Science Press; Oxford University Press. 297. 5. Mitchison, T.J. and E.D. Salmon, Mitosis: a history of division. Nature cell biology, 2001. 3(1): p. E17-21. 6. Musacchio, A. and E.D. Salmon, The spindle-assembly checkpoint in space and time. Nature reviews. Molecular cell biology, 2007. 8(5): p. 379-93. 7. Mohan, K., P.A. Iglesias, and D.N. Robinson, Separation anxiety: Stress, tension and cytokinesis. Experimental cell research, 2012. 318(12): p. 1428-34. 8. Lacroix, B. and A.S. Maddox, Cytokinesis, ploidy and aneuploidy. The Journal of pathology, 2012. 226(2): p. 338-51. 9. Zhang, D., A. Vjestica, and S. Oliferenko, The cortical ER network limits the permissive zone for actomyosin ring assembly. Current biology : CB, 2010. 20(11): p. 1029-34. 10. Zhao, W.M. and G. Fang, MgcRacGAP controls the assembly of the contractile ring and the initiation of cytokinesis. Proceedings of the National Academy of Sciences of the United States of America, 2005. 102(37): p. 13158-63. 11. Carmena, M. and W.C. Earnshaw, The cellular geography of aurora kinases. Nature reviews. Molecular cell biology, 2003. 4(11): p. 842-54. 12. Przewloka, M.R. and D.M. Glover, The kinetochore and the centromere: a working long distance relationship. Annual review of genetics, 2009. 43: p. 439-65. 13. Santaguida, S. and A. Musacchio, The life and miracles of kinetochores. The EMBO journal, 2009. 28(17): p. 2511-31. 14. Cheeseman, I.M., et al., The conserved KMN network constitutes the core microtubule-binding site of the kinetochore. Cell, 2006. 127(5): p. 983-97. 15. Ciferri, C., et al., Architecture of the human ndc80-hec1 complex, a critical constituent of the outer kinetochore. The Journal of biological chemistry, 2005. 280(32): p. 29088-95. 16. Ciferri, C., et al., Implications for kinetochore-microtubule attachment from the structure of an engineered Ndc80 complex. Cell, 2008. 133(3): p. 427-39. 17. Wei, R.R., P.K. Sorger, and S.C. Harrison, Molecular organization of the Ndc80 complex, an essential kinetochore component. Proc Natl Acad Sci USA, 2005. 102(15): p. 5363-7. 18. DeLuca, J.G., et al., Kinetochore microtubule dynamics and attachment stability are regulated by Hec1. Cell, 2006. 127(5): p. 969-82. 19. Sandall, S., et al., A Bir1-Sli15 complex connects centromeres to microtubules and is required to sense kinetochore tension. Cell, 2006. 127(6): p. 1179-91. 21. Wells, W.A., The spindle-assembly checkpoint: aiming for a perfect mitosis, every time. Trends Cell Biol, 1996. 6(6): p. 228-34. 22. Rieder, C.L., et al., Anaphase onset in vertebrate somatic cells is controlled by a checkpoint that monitors sister kinetochore attachment to the spindle. J Cell Biol, 1994. 127(5): p. 1301-10. 23. Li, X. and R.B. Nicklas, Mitotic forces control a cell-cycle checkpoint. Nature, 1995. 373(6515): p. 630-2. 24. Hoyt, M.A., L. Trotis, and B.T. Roberts, S. cerevisiae genes required for cell cycle arrest in response to loss of microtubule function. Cell, 1991. 66: p. 507-517. ! 86! 25. Li, R. and A.W. Murray, Feedback control of mitosis in budding yeast. Cell, 1991. 66(3): p. 519-31. 26. Morrow, C.J., Bub1 and aurora B cooperate to maintain BubR1-mediated inhibition of APC/CCdc20. Journal of Cell Science, 2005. 118(16): p. 3639-3652. 27. Hwang, L.H., et al., Budding yeast Cdc20: a target of the spindle checkpoint. Science, 1998. 279(5353): p. 1041-4. 28. Kim, S.H., et al., Fission yeast Slp1: an effector of the Mad2-dependent spindle checkpoint. Science, 1998. 279(5353): p. 1045-7. 29. Musacchio, A. and E.D. Salmon, The spindle-assembly checkpoint in space and time. Nat Rev Mol Cell Biol, 2007. 8(5): p. 379-393. 30. Chan, G.K., S.-T. Liu, and T.J. Yen, Kinetochore structure and function. Trends Cell Biol, 2005. 15(11): p. 589-98. 31. Arimura, N. and K. Kaibuchi, Neuronal polarity: from extracellular signals to intracellular mechanisms. Nature reviews. Neuroscience, 2007. 8(3): p. 194-205. 32. Krummel, M.F. and I. Macara, Maintenance and modulation of T cell polarity. Nature immunology, 2006. 7(11): p. 1143-9. 33. Pruyne, D., et al., Mechanisms of polarized growth and organelle segregation in yeast. Annual review of cell and developmental biology, 2004. 20: p. 559-91. 34. Dow, L.E. and P.O. Humbert, Polarity regulators and the control of epithelial architecture, cell migration, and tumorigenesis. International review of cytology, 2007. 262: p. 253-302. 35. Margolis, J. and A. Spradling, Identification and behavior of epithelial stem cells in the Drosophila ovary. Development, 1995. 121(11): p. 3797-807. 36. Roth, S. and J.A. Lynch, Symmetry breaking during Drosophila oogenesis. Cold Spring Harbor perspectives in biology, 2009. 1(2): p. a001891. 37. Kirilly, D. and T. Xie, The Drosophila ovary: an active stem cell community. Cell research, 2007. 17(1): p. 15-25. 38. Frydman, H.M. and A.C. Spradling, The receptor-like tyrosine phosphatase lar is required for epithelial planar polarity and for axis determination within drosophila ovarian follicles. Development, 2001. 128(16): p. 3209-20. 39. Silies, M., et al., The eye imaginal disc as a model to study the coordination of neuronal and glial development. Fly, 2010. 4(1): p. 71-9. 40. Cagan, R., Principles of Drosophila eye differentiation. Current topics in developmental biology, 2009. 89: p. 115-35. 41. Kumar, J.P., Signalling pathways in Drosophila and vertebrate retinal development. Nature reviews. Genetics, 2001. 2(11): p. 846-57. 42. St Johnston, D. and J. Ahringer, Cell polarity in eggs and epithelia: parallels and diversity. Cell, 2010. 141(5): p. 757-74. 43. Watts, J.L., et al., par-6, a gene involved in the establishment of asymmetry in early C. elegans embryos, mediates the asymmetric localization of PAR-3. Development, 1996. 122(10): p. 3133-40. 44. Goldstein, B. and I.G. Macara, The PAR proteins: fundamental players in animal cell polarization. Developmental cell, 2007. 13(5): p. 609-22. 45. Kuchinke, U., F. Grawe, and E. Knust, Control of spindle orientation in Drosophila by the Par-3-related PDZ-domain protein Bazooka. Current biology : CB, 1998. 8(25): p. 1357-65. 46. Izumi, Y., et al., An atypical PKC directly associates and colocalizes at the epithelial tight junction with ASIP, a mammalian homologue of Caenorhabditis elegans polarity protein PAR-3. The Journal of cell biology, 1998. 143(1): p. 95106. 47. Tabuse, Y., et al., Atypical protein kinase C cooperates with PAR-3 to establish embryonic polarity in Caenorhabditis elegans. Development, 1998. 125(18): p. 3607-14. ! 87! 48. Hung, T.J. and K.J. Kemphues, PAR-6 is a conserved PDZ domain-containing protein that colocalizes with PAR-3 in Caenorhabditis elegans embryos. Development, 1999. 126(1): p. 127-35. 49. Lin, D., et al., A mammalian PAR-3-PAR-6 complex implicated in Cdc42/Rac1 and aPKC signalling and cell polarity. Nature cell biology, 2000. 2(8): p. 540-7. 50. Schober, M., M. Schaefer, and J.A. Knoblich, Bazooka recruits Inscuteable to orient asymmetric cell divisions in Drosophila neuroblasts. Nature, 1999. 402(6761): p. 548-51. 51. Wodarz, A., et al., Bazooka provides an apical cue for Inscuteable localization in Drosophila neuroblasts. Nature, 1999. 402(6761): p. 544-7. 52. Joberty, G., et al., The cell-polarity protein Par6 links Par3 and atypical protein kinase C to Cdc42. Nature cell biology, 2000. 2(8): p. 531-9. 53. Petronczki, M. and J.A. Knoblich, DmPAR-6 directs epithelial polarity and asymmetric cell division of neuroblasts in Drosophila. Nature cell biology, 2001. 3(1): p. 43-9. 54. Aranda, V., M.E. Nolan, and S.K. Muthuswamy, Par complex in cancer: a regulator of normal cell polarity joins the dark side. Oncogene, 2008. 27(55): p. 6878-87. 55. Etienne-Manneville, S., Cdc42--the centre of polarity. Journal of cell science, 2004. 117(Pt 8): p. 1291-300. 56. Tepass, U., et al., Epithelial cell polarity and cell junctions in Drosophila. Annual review of genetics, 2001. 35: p. 747-84. 57. Franz, A. and V. Riechmann, Stepwise polarisation of the Drosophila follicular epithelium. Developmental biology, 2010. 338(2): p. 136-47. 58. Bhat, M.A., et al., Discs Lost, a novel multi-PDZ domain protein, establishes and maintains epithelial polarity. Cell, 1999. 96(6): p. 833-45. 59. Muller, H.A. and E. Wieschaus, armadillo, bazooka, and stardust are critical for early stages in formation of the zonula adherens and maintenance of the polarized blastoderm epithelium in Drosophila. The Journal of cell biology, 1996. 134(1): p. 149-63. 60. Bachmann, A., et al., Drosophila Stardust is a partner of Crumbs in the control of epithelial cell polarity. Nature, 2001. 414(6864): p. 638-43. 61. Knust, E., U. Tepass, and A. Wodarz, crumbs and stardust, two genes of Drosophila required for the development of epithelial cell polarity. Development, 1993: p. 261-8. 62. Harris, T.J. and M. Peifer, The positioning and segregation of apical cues during epithelial polarity establishment in Drosophila. The Journal of cell biology, 2005. 170(5): p. 813-23. 63. Martin-Belmonte, F., et al., PTEN-mediated apical segregation of phosphoinositides controls epithelial morphogenesis through Cdc42. Cell, 2007. 128(2): p. 383-97. 64. Satohisa, S., et al., Behavior of tight-junction, adherens-junction and cell polarity proteins during HNF-4alpha-induced epithelial polarization. Experimental cell research, 2005. 310(1): p. 66-78. 65. Ebnet, K., et al., The cell polarity protein ASIP/PAR-3 directly associates with junctional adhesion molecule (JAM). The EMBO journal, 2001. 20(14): p. 3738-48. 66. Itoh, M., et al., Junctional adhesion molecule (JAM) binds to PAR-3: a possible mechanism for the recruitment of PAR-3 to tight junctions. The Journal of cell biology, 2001. 154(3): p. 491-7. 67. Takekuni, K., et al., Direct binding of cell polarity protein PAR-3 to cell-cell adhesion molecule nectin at neuroepithelial cells of developing mouse. The Journal of biological chemistry, 2003. 278(8): p. 5497-500. 68. McGill, M.A., R.F. McKinley, and T.J. Harris, Independent cadherin-catenin and Bazooka clusters interact to assemble adherens junctions. The Journal of cell biology, 2009. 185(5): p. 787-96. ! 88! 69. Chen, X. and I.G. Macara, Par-3 controls tight junction assembly through the Rac exchange factor Tiam1. Nature cell biology, 2005. 7(3): p. 262-9. 70. Morais-de-Sa, E., V. Mirouse, and D. St Johnston, aPKC phosphorylation of Bazooka defines the apical/lateral border in Drosophila epithelial cells. Cell, 2010. 141(3): p. 509-23. 71. Bilder, D. and N. Perrimon, Localization of apical epithelial determinants by the basolateral PDZ protein Scribble. Nature, 2000. 403(6770): p. 676-80. 72. Bilder, D., M. Li, and N. Perrimon, Cooperative regulation of cell polarity and growth by Drosophila tumor suppressors. Science, 2000. 289(5476): p. 113-6. 73. Bilder, D., Epithelial polarity and proliferation control: links from the Drosophila neoplastic tumor suppressors. Genes & development, 2004. 18(16): p. 1909-25. 74. Humbert, P., S. Russell, and H. Richardson, Dlg, Scribble and Lgl in cell polarity, cell proliferation and cancer. BioEssays : news and reviews in molecular, cellular and developmental biology, 2003. 25(6): p. 542-53. 75. Mathew, D., et al., Recruitment of scribble to the synaptic scaffolding complex requires GUK-holder, a novel DLG binding protein. Current biology : CB, 2002. 12(7): p. 531-9. 76. Kallay, L.M., et al., Scribble associates with two polarity proteins, Lgl2 and Vangl2, via distinct molecular domains. Journal of cellular biochemistry, 2006. 99(2): p. 647-64. 77. Bilder, D., M. Schober, and N. Perrimon, Integrated activity of PDZ protein complexes regulates epithelial polarity. Nature cell biology, 2003. 5(1): p. 53-8. 78. Plant, P.J., et al., A polarity complex of mPar-6 and atypical PKC binds, phosphorylates and regulates mammalian Lgl. Nature cell biology, 2003. 5(4): p. 301-8. 79. Yamanaka, T., et al., Mammalian Lgl forms a protein complex with PAR-6 and aPKC independently of PAR-3 to regulate epithelial cell polarity. Current biology : CB, 2003. 13(9): p. 734-43. 80. Yamanaka, T., et al., Lgl mediates apical domain disassembly by suppressing the PAR-3-aPKC-PAR-6 complex to orient apical membrane polarity. Journal of cell science, 2006. 119(Pt 10): p. 2107-18. 81. Betschinger, J., K. Mechtler, and J.A. Knoblich, The Par complex directs asymmetric cell division by phosphorylating the cytoskeletal protein Lgl. Nature, 2003. 422(6929): p. 326-30. 82. Chalmers, A.D., et al., aPKC, Crumbs3 and Lgl2 control apicobasal polarity in early vertebrate development. Development, 2005. 132(5): p. 977-86. 83. Betschinger, J. and J.A. Knoblich, Dare to be different: asymmetric cell division in Drosophila, C. elegans and vertebrates. Current biology : CB, 2004. 14(16): p. R674-85. 84. Prehoda, K.E., Polarization of Drosophila neuroblasts during asymmetric division. Cold Spring Harbor perspectives in biology, 2009. 1(2): p. a001388. 85. Mayer, B., et al., Quantitative analysis of protein dynamics during asymmetric cell division. Current biology : CB, 2005. 15(20): p. 1847-54. 86. Rhyu, M.S., L.Y. Jan, and Y.N. Jan, Asymmetric distribution of numb protein during division of the sensory organ precursor cell confers distinct fates to daughter cells. Cell, 1994. 76(3): p. 477-91. 87. Doe, C.Q., et al., The prospero gene specifies cell fates in the Drosophila central nervous system. Cell, 1991. 65(3): p. 451-64. 88. Chu-Lagraff, Q., et al., The prospero gene encodes a divergent homeodomain protein that controls neuronal identity in Drosophila. Development, 1991. Suppl 2: p. 79-85. 89. Spana, E.P. and C.Q. Doe, The prospero transcription factor is asymmetrically localized to the cell cortex during neuroblast mitosis in Drosophila. Development, 1995. 121(10): p. 3187-95.