Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Identification of echinus and characterization of its role in Drosophila eye development Bosdet, Ian Edward 2008

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


24-ubc_2008_fall_bosdet_ian.pdf [ 24.21MB ]
JSON: 24-1.0066540.json
JSON-LD: 24-1.0066540-ld.json
RDF/XML (Pretty): 24-1.0066540-rdf.xml
RDF/JSON: 24-1.0066540-rdf.json
Turtle: 24-1.0066540-turtle.txt
N-Triples: 24-1.0066540-rdf-ntriples.txt
Original Record: 24-1.0066540-source.json
Full Text

Full Text

Identification of echinus and characterization of its role in Drosophila eye development    by     IAN EDWARD BOSDET  B.Sc., University of British Columbia, 1995      A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY    in    THE FACULTY OF GRADUATE STUDIES  (Genetics)           THE UNIVERSITY OF BRITISH COLUMBIA  (Vancouver)  August, 2008  © Ian Edward Bosdet, 2008  ii ABSTRACT   The precise structure of the adult Drosophila eye results from a coordinated process of cell sorting, differentiation and selective cell death in the retinal epithelium.  Mutations in the gene echinus cause supernumerary pigment cells due to insufficient cell death.  This study reports the identification of echinus and the characterization of its role in Drosophila retinal development.  Using a combination of deletion mapping, gene expression analysis and genomic sequencing, echinus was cloned and several alleles were sequenced.  echinus encodes a ~180kDa protein containing an ubiquitin hydrolase domain at its N-terminus and a polyglutamine tract at its C-terminus.  echinus is expressed in the retina during pupal development and mutants of echinus have decreased levels of apoptosis during several stages of retinal development.  Defects in the cell sorting process that precedes cell death are also observed in echinus loss-of-function mutants and echinus overexpression can cause defects in ommatidial rotation and the morphology of cone cells.  echinus is a positive regulator of DE- cadherin and Enabled accumulation in adherens junctions of retinal epithelial cells.  Genetic interactions were observed between echinus and the genes wingless, enabled and expanded.  An immunofluorescence assay in Drosophila S2 cell cultured demonstrated that Echinus localizes to intracellular vesicles that do not appear to be endocytic in nature, and the C-terminal region of Echinus was shown to be necessary for this association.  A protein interaction screen using an immunoprecipitation and mass spectrometry approach identified interactions between Echinus and the vesicle coat protein Clathrin, the scaffolding protein RACK1 and the casein kinase Iε (Dco).  Co-immunoprecipitation additionally identified an interaction between Echinus and Enabled.  This work has revealed echinus to be an important regulator of cell sorting and adherens junction formation in the developing retina and has identified multiple interactions between echinus and enabled, a regulator of the actin cytoskeleton.  iii TABLE OF CONTENTS ABSTRACT................................................................................................................................ ii TABLE OF CONTENTS ............................................................................................................iii LIST OF TABLES ...................................................................................................................... vi LIST OF FIGURES ................................................................................................................... vii LIST OF ABBREVIATIONS...................................................................................................... ix ACKNOWLEDGEMENTS .......................................................................................................... x DEDICATION .........................................................................................................................xii CHAPTER 1 | INTRODUCTION ............................................................................................. 1 1.1) Overall rationale............................................................................................................... 1 1.2) Drosophila melanogaster as a model organism....................................................................... 2 1.3) Programmed cell death .................................................................................................... 2 1.4) Development of the Drosophila retina................................................................................ 4 1.5) Epithelial adherens junctions......................................................................................... 14 1.6) Ena/VASP proteins: regulators of the actin cytoskeleton............................................. 17 1.7) Ubiquitination................................................................................................................ 17 1.8) echinus............................................................................................................................... 19 1.9) Objectives, specific aims and hypotheses ...................................................................... 20 1.9.1) Specific aim 1) Identify the gene or genomic region responsible for the echinus rough- eye phenotype. ......................................................................................................... 21  iv 1.9.2) Specific aim 2) Characterize the role of echinus in development of the Drosophila retina. .................................................................................................................................. 21 1.9.3) Specific aim 3) Characterize the molecular function of echinus. ................................. 21 CHAPTER 2 | ECHINUS ALLELES ARE ASSOCIATED WITH MUTATIONS IN THE GENE CG2904, WHICH HAS HOMOLOGY TO DEUBIQUITINATING ENZYMES ...................... 22 2.1) Introduction ................................................................................................................... 22 2.2) Materials and Methods ................................................................................................... 25 2.2.1) Drosophila Strains and Genetics .................................................................................... 25 2.2.2) Genomic and Plasmid DNA Purification and PCR ..................................................... 25 2.2.3) DNA Sequencing ....................................................................................................... 28 2.2.4) RNA Isolation and Realtime RT-PCR......................................................................... 28 2.2.5) Computational analysis of protein sequence data...................................................... 30 2.3) Results ............................................................................................................................. 31 2.3.1) The echinus mutation is localized to the 3F2-3F3 region of the X chromosome......... 31 2.3.2) Two transcripts are present on the opposite strand to CG2904 ................................ 34 2.3.3) The gene CG2904 in the echinus candidate region has altered expression levels in some echinus mutants........................................................................................................... 36 2.3.4) Mutations in the predicted gene CG2904 are associated with the echinus phenotype. 39 2.3.5) echinus has multiple splice forms ................................................................................ 42 2.3.6) Features of echinus inferred protein sequence ............................................................. 44 2.4) Discussion ....................................................................................................................... 46 CHAPTER 3 | MUTATIONS IN ECHINUS DISRUPT NORMAL DEVELOPMENT OF THE DROSOPHILA RETINA.............................................................................................................. 49 3.1) Introduction ................................................................................................................... 49 3.2) Materials and Methods ................................................................................................... 52 3.2.1) Drosophila lines and genetics........................................................................................ 52 3.2.2) Pupal retina cell counts ............................................................................................. 52 3.2.3) Immunohistochemistry............................................................................................. 53 3.2.4) TUNEL staining......................................................................................................... 53 3.2.5) RNA in situ staining.................................................................................................... 54 3.3) Results ............................................................................................................................. 55 3.3.1) echinus mutants have excess interommatidial cells...................................................... 55 3.3.2) echinus mutations show decreased developmental apoptosis in the Drosophila retina.... 58 3.3.3) echinus is expressed in all epithelial cells of the developing retina .............................. 60  v 3.3.4) echinus mutations cause defective cell sorting............................................................. 63 3.3.5) echinus is a positive regulator of DE-cadherin accumulation in retinal epithelial cell adherens junctions .................................................................................................... 66 3.3.6) Septate junctions are intact in echinus mutants............................................................ 68 3.3.7) echinus is a positive regulator of Enabled accumulation in the adherens junctions of retinal epithelial cells ................................................................................................ 70 3.3.8) Over-expression of echinus in the retina results in ommatidial rotation defects and alterations in cone cell morphology.......................................................................... 72 3.3.9) Echinus appears within IOCs and in association with the IOC/PPC junction at an early stage of pupal retinal development ........................................................................... 74 3.3.10) echinus mutants have abnormal development of some mechanosensory bristles ...... 76 3.3.11) echinus interacts genetically with enabled, wingless and components of the fat tumor suppressor pathway .................................................................................................. 79 3.4) Discussion ....................................................................................................................... 84 CHAPTER 4 | ECHINUS IS A VESICLE-ASSOCIATED PROTEIN THAT INTERACTS WITH CLATHRIN AND ENABLED.................................................................................................... 96 4.1) Introduction ................................................................................................................... 96 4.2) Materials and Methods ................................................................................................. 100 4.2.1) cDNA cloning into GatewayTM vectors .................................................................... 100 4.2.2) Cell culture and transfection ................................................................................... 100 4.2.3) Immunofluorescence .............................................................................................. 101 4.2.4) Co-immunoprecipitation and MS/MS analysis........................................................ 102 4.2.5) Western blotting ..................................................................................................... 104 4.2.6) Computational analysis of protein sequences.......................................................... 104 4.3) Results ........................................................................................................................... 105 4.3.1) Echinus protein localizes to intracellular vesicles .................................................... 105 4.3.2) The C-terminal region of Echinus is required for localization to vesicles................ 107 4.3.3) Echinus vesicles do not co-localize with internalized dextran or Rab5 ................... 111 4.3.4) Echinus protein interacts with the vesicle coat protein Clathrin, Casein kinase Iε and the protein kinase C scaffolding protein RACK1 ..................................................... 113 4.3.5) Echinus co-localizes and interacts with the Ena/VASP protein Enabled .................. 119 4.4) Discussion ..................................................................................................................... 122 CHAPTER 5 | SUMMARY AND FUTURE RESEARCH......................................................... 131 CHAPTER 6 | REFERENCES ................................................................................................. 136   vi LIST OF TABLES  Table 2.1 | Primer sequences for deletion-mapping amplicons .............................................. 27 Table 2.2 | RT-PCR forward (F) and reverse (R) primer sequences ........................................ 29 Table 3.1 | echinus mutants have excess interommatidial cells.................................................. 57 Table 3.2 | Dorsocentral and scutellar bristle numbers are altered in echinus mutants. ............. 78 Table 3.3 | Mutant alleles that failed to modify the echinus rough-eye phenotype.................... 81 Table 4.1 | Proteins identified to interact with both N- and C-terminal Echinus FLAG......... 115 Table 4.2 | Top scoring proteins identified to interact with either N- or C-terminal Echinus FLAG................................................................................................................... 116 Supplemental Table 1 | Complete list of proteins identified by LC-MS/MS analysis as interactors of Echinus ......................................................................................... 130   vii LIST OF FIGURES  Figure 1.1 | Drosophila imaginal discs.......................................................................................... 5 Figure 1.2 | Pattern formation in the pupal eye. ....................................................................... 9 Figure 1.3 | Cell topologies and geometries in the eye. .......................................................... 10 Figure 1.4 | AJ dynamics in IPCs............................................................................................. 13 Figure 1.5 | Schematic of epithelial cell structure in a late-embryonic/larval epidermal cell. . 15 Figure 1.6 | The cadherin–catenin protein complex. .............................................................. 16 Figure 1.7 | The ubiquitination pathway. ............................................................................... 18 Figure 2.1 | Map of deficiencies in the echinus candidate region. ............................................. 32 Figure 2.2 | Organization of the echinus candidate region. ....................................................... 33 Figure 2.3 | QRT-PCR analysis of the four genes in the echinus candidate region..................... 35 Figure 2.4 | Quantitative RT-PCR analysis of transcripts within the echinus candidate region. . 37 Figure 2.5 | Strand-specific realtime RT-PCR of CG2904 and antisense transcripts. ............... 38 Figure 2.6 | PCR amplification of genomic DNA reveals a 22.5 kb deletion in ecΔ9 and ec19R. .. 40 Figure 2.7 | Location of CG2904 mutations in echinus mutant strains...................................... 41 Figure 2.8 | echinus encodes multiple splice forms. .................................................................. 43 Figure 2.9 | Protein structure of Echinus. ............................................................................... 45 Figure 3.1 | echinus mutant retina have excess interommatidial cells........................................ 56 Figure 3.2 | echinus mutations decrease apoptosis during retinal development. ....................... 59 Figure 3.3 | echinus is expressed in epithelial cells of the developing ommatidia. .................... 61 Figure 3.4 | RNA in situ staining of wild-type retinas. ........................................................... 62 Figure 3.5 | Roughest protein distribution is altered in echinus mutants. ................................. 65 Figure 3.6 | echinus promotes DE-cadherin accumulation in the adherens junction of retinal cells....................................................................................................................... 67  viii Figure 3.7 | Septate junctions appear intact in echinus mutants................................................. 69 Figure 3.8 | echinus regulates association of the actin stimulating protein Enabled with adherens junctions in the developing retina. ....................................................................... 71 Figure 3.9 | Over-expression of echinus causes defects in patterning, ommatidial rotation and cone cell morphology........................................................................................... 73 Figure 3.10 | Echinus protein is found in interommatidial cells. ............................................ 75 Figure 3.11 | The dorsocentral and scutellar bristles develop abnormally in echinus mutants... 77 Figure 3.12 | Schematic representation of genetic crosses employed for interaction testing. . 80 Figure 3.13 | echinus interacts with enabled, wingless and components of the fat tumor suppressor pathway. ............................................................................................. 83 Figure 4.1 | Overview of protein interaction assay by immunoprecipitation followed by LC- MS/MS analysis..................................................................................................... 99 Figure 4.2 | Echinus-FLAG localizes to intracellular vesicles. ................................................ 106 Figure 4.3 | Echinus subclones constructed for subcellular localization analysis................... 108 Figure 4.4 | The Echinus C-terminus is required for localization to vesicles......................... 110 Figure 4.5 | Echinus vesicles do not co-localize with the early endosome marker Rab5 or an endocytosis marker............................................................................................. 112 Figure 4.6 | Myc-Echinus co-immunoprecipitates with Dco-FLAG, RACK1-FLAG and Clathrin- FLAG................................................................................................................... 117 Figure 4.7 | Co-expression of Dco-FLAG, RACK1-FLAG and Clathrin-FLAG with Myc-Echinus in Drosophila S2 cell culture................................................................................... 118 Figure 4.8 | Enabled co-localizes with Echinus-FLAG in Drosophila S2 cells. ........................... 120 Figure 4.9 | Enabled co-immunoprecipitates with Echinus-FLAG......................................... 121 Figure 5.1 | Speculative model of Echinus function.............................................................. 134   ix LIST OF ABBREVIATIONS   AJ  Adherens junction CC  Cone cell DUB  De-ubiquitinating enzyme Ena/VASP Enabled / vasodialator-stimulated protein EVH1  Ena/VASP Homology 1 IOC  Interommatidial cell PCD  Programmed cell death PPC  Primary pigment cell SPC  Secondary pigment cell TPC  Tertiary pigment cell TUNEL  TdT-mediated dUTP-Nick-End Labeling     x ACKNOWLEDGEMENTS  Firstly, I wish to thank Dr. Marco Marra for agreeing to be my supervisor and for your constant support since I arrived in your lab.  You have always challenged me to improve my knowledge, critical thinking and caffeine tolerance, and I am sure I will be the better for it.  I am in your debt for the opportunity you have given me.  My very special thanks to my committee member, Dr. Sharon Gorski, who has been so generous with her time and effort and is such an enjoyable person to work with.  Your support and assistance have been greatly appreciated.  I would also like to thank my committee members Dr. Steven Jones and Dr. Vanessa Auld for your time and guidance throughout my studies.  I would like to thank the other members of the programmed cell death group at the Genome Sciences Centre for making it such a friendly and stimulating environment.  In particular I would like to thank Dr. Suganthi Chittaranjan, Claire Hou and Doug Freeman for all their assistance during my research and for being wonderful people to work with.  I would also like to thank Lindsay Devorkin for her help in generating the echinus subclones.  Thank you also to Jacqueline Schein for your friendship, guidance and patience with me over the last nine years.  Thanks to Dr. J. Copeland and Dr. B. Hay for helpful advice, numerous reagents and their collaboration on the first echinus manuscript.   xi I thank Dr. G.B. Morin and Dr. M. Kuzyk for their collaborations on the protein interaction analysis and Dr. G. Cheng and Carri-Lynn Mead for helpful suggestions.  I also thank Dr. Sam Aparicio for the use of the confocal microscope and John Fe for instrument training and support.  Thank you to Dr. John Nomellini, who gave me my first training in how to be a scientist and who has remained a good friend ever since.  I wish to thank my lovely daughter, Audrey, for making me so happy, reminding me of the important things in life and being so understanding when I didn’t have time to play.  Finally, I thank my beautiful wife, Taryn, for all of your support, patience and sacrifices throughout my studies.  In every way, I could not have done this without you.    xii DEDICATION       To my father, John Bosdet, for the lifetime of love, wisdom and inspiration I received in too few years.  With this I can finally say I have kept my last promise to you.       1 Chapter 1 | Introduction  1.1) Overall rationale The cellular processes of programmed cell death (PCD) and cell adhesion are important both for normal development of multicellular organisms and for the pathology of diseases such as cancer.  The developing Drosophila retina is an ideal experimental system for studying these processes.  It is a complex tissue with many different cell types that arise from a pool of undifferentiated cells and its regular patterning allows the detection and examination of subtle changes in cell numbers and arrangement.   Despite rapid progress in the understanding of PCD in recent years, and the discovery of many cell death effectors, many questions remain unanswered.  The regulation of the signals that initiate cell death during development are poorly characterized, as are the mechanisms by which cell death is directed in a cell type- specific and position-specific manner.  As well, it is becoming increasingly apparent that cell adhesion is a dynamic and highly-regulated process that has important implications for programmed cell death regulation and tissue development. The objective of this study was to identify and characterize echinus, a gene influencing apoptotic cell death in the Drosophila retina, by cloning the corresponding gene and exploring the function of the gene product.   This research has successfully identified echinus and has revealed a role for this gene in the correct development of cell adhesive contacts during retinal development.  Homology of the Echinus protein to ubiquitin hydrolase enzymes and its interaction with the Drosophila Ena/VASP homologue Enabled have also been identified.  This characterization has enhanced our understanding of the mechanisms of PCD and cell adhesion, and the role they play in development of the Drosophila retina.  2 1.2) Drosophila melanogaster as a model organism Drosophila melanogaster, known as the fruit fly, has been used as a model organism for genetic research for over a century due to its many advantages for researchers.  Drosophila has a short generation time of approximately two weeks, facilitating genetic analyses.  Additional advantages include the availablily of large collections of deletions and insertional mutants, and an extensive collection of available stocks containing distinctive alleles useful for genetic analyses (Flybase, 1994; Flybase, 2002).  The complete genome sequence is also available (Adams et al., 2000) and this provides a powerful tool for gene discovery and subsequent molecular analyses.  The UAS/Gal4 system can be utilized for spatial and temporal control over expression of transgenes inserted into the genome (Duffy, 2002) and more recently new genomic resources have become available, such as comprehensive RNAi stock collections (Flockhart et al., 2006), cDNA libraries (Drosophila Genomics Resource Center) and microarrays for gene expression analyses.  Another advantage to using Drosophila as a model system for the development of tissues is the ability to observe the fly’s complex development and morphology and to produce quickly large numbers of mutations and screen them for observable phenotypes in vivo (Lee, 1989; Adams and Sekelsky, 2002; St Johnston, 2002).  A particularly valuable tissue in this regard is the Drosophila retina.  1.3) Programmed cell death Programmed cell death (PCD) is a genetically regulated, evolutionarily conserved process that directs the removal of unwanted, aberrant or damaged cells (Wolff and Ready, 1991b; Vaux and Korsmeyer, 1999; Baehrecke, 2002).  It is activated by the appearance or disappearance of specific cell signals, or by external stimuli (Song and Steller, 1999). Apoptosis is a form of programmed cell death that is characterized by membrane blebbing, chromatin condensation, cell shrinkage, internucleosomal fragmentation and protein cleavage  3 (Kerr et al., 1972; Wolff and Ready, 1991b; Meller et al., 1997; Zakeri and Lockshin, 2002). The resulting cellular remains (‘apoptotic bodies’) are engulfed by phagocytes (Bangs et al., 2000).  The process of apoptosis is critical for correct embryogenesis (Bangs and White, 2000), nervous system development (Tissot and Stocker, 2000), and immune sytem differentiation (Krammer, 2000).  Defects in PCD play an important role in autoimmune and neurodegenerative diseases (Ross, 2002), as well as the uninhibited cell proliferation observed in cancer (Evan and Littlewood, 1998; Baehrecke, 2002). The core death machinery is composed of molecules that mediate cellular death and the regulators of their activity.  It was first described in the nematode worm C. elegans with the discovery of three key proteins that control cell death in the worm: ced-3, ced-4 and ced-9 (Metzstein et al., 1998; Liu and Hengartner, 1999).  The ced-3 gene product is a member of the evolutionarily conserved caspase family of proteases.  The caspases are cysteine proteases that are present in all cells as inactive precursors (zymogens) and are critical for cell death in all metazoans (Alnemri et al., 1996; Thornberry and Lazebnik, 1998).  Caspases effect cell death through the targeted inactivation of critical cellular components such as the nuclear lamina, DNA replication and repair molecules and numerous negative regulators of apoptosis (Thornberry and Lazebnik, 1998).  The ced-4 gene product is homologous to mammalian Apaf1 and is a positive regulator of ced-3 caspase activity.  The Bcl-2-like ced-9 protein acts as a negative regulator of ced-4 and so suppresses cell death.  In Drosophila numerous other molecules contribute to the regulation of caspase activity.  The inhibitors of apoptosis (IAPs) bind directly to caspases and are the only known negative regulators of caspase function (Hay, 2000).  Three additional proteins that mediate apoptosis in Drosophila are Reaper, Grim and Hid.  These proteins, either individually or in combination, are required for essentially all PCD in Drosophila and expression of either Reaper or Grim invariably indicates that cellular death is imminent (Chen et al., 1996; Nordstrom et al., 1996).  These genes are the targets of  4 numerous different signals, including pro-apoptotic stimulation by p53 in response to DNA damage (Nordstrom and Abrams, 2000) or repression by anti-apoptotic EGF/Ras signals (Kurada and White, 1998).  Members of the Bcl-2 family of proteins (Chen and Abrams, 2000; Colussi et al., 2000) are also important in both pro- and anti-apoptotic roles through their interactions with the mitochondria and dark (the Drosophila ced-4 homologue) (Rodriguez et al., 1999; Rodriguez et al., 2002).  Vertebrates display an even greater complexity (Aravind et al., 2001) with respect to the size of the core death machinery gene family members. Much of the PCD machinery is evolutionarily conserved between humans and Drosophila (Meier et al., 2000), and numerous homologues of human PCD genes have been discovered in the fly (Aravind et al., 1999; Potter et al., 2000; Vernooy et al., 2000; Aravind et al., 2001). These include the caspases (e.g. Caspase-9/Dredd) (Thornberry and Lazebnik, 1998; Zimmermann et al., 2002), caspase inhibitors (e.g. IAP/DIAP-1) (Goyal et al., 2000; Martin, 2002; Muro et al., 2002; Palaga and Osborne, 2002) and central regulatory molecules (e.g. Apaf-1/Dark) (Rodriguez et al., 2002).  These similarities enable studies of PCD in the fly to act as models for our general understanding of the process in various organisms, including vertebrates (Abrams, 1999; Potter et al., 2000; Gorski and Marra, 2002; Richardson and Kumar, 2002).  1.4) Development of the Drosophila retina The Drosophila eye originates in the cellular blastoderm of the embryo from a region of basolateral ectoderm that folds to form the eye-antennal disc (Wolff and Ready, 1993). During development, this eye disc (Figure 1.1) undergoes an initial proliferation of cells that is followed by cell differentiation, reorganization and then selective cell death to leave a highly ordered array of 14-cells called an ommatidia (Wolff and Ready, 1993; Baker, 2001).   5 Figure 1.1  Figure 1.1 | Drosophila imaginal discs.  During larval and pupal development the imaginal discs develop from pools of undifferentiated cells into the final complex structures of the adult. [Figure from (Held, 2002).]      6 Cells in the imaginal disc proliferate throughout the embryonic and early larval stages. During the third larval instar, proliferation ceases and a wave of differentiation sweeps across the disc and stimulates the development of approximately 750 ommatidia.   Each ommatidium is composed of 8 photoreceptors, 4 non-neuronal cone cells and 2 optically insulating primary pigment cells, which develop progressively during the late larval and early pupal stages.  By the late pupal stage of development a lattice composed of 9 secondary and tertiary pigment cells and 3 mechanosensory bristles surrounds each ommatidium (Ready et al., 1976; Wolff and Ready, 1993). Cell fate during development of the retina is determined by cell-cell interactions; laser ablation studies have shown that lineage-derived information does not play an important role in determining retinal cell fates (Miller and Cagan, 1998; Baker and Yu, 2001).  In the larval third instar stage the undifferentiated pool of proliferating cells in the eye disc undergo cell cycle arrest at G1 in a progressive manner, causing a characteristic morphogenetic furrow that proceeds anteriorly across the disc (Freeman, 1997; Spencer et al., 1998; Baker and Yu, 2001). Cells posterior to the furrow undergo one additional cell division (termed the second mitotic wave), after which they begin to differentiate and are recruited into the regularly spaced ommatidia (Ready et al., 1976; Freeman, 1997).  Progression of the morphogenetic furrow is driven anteriorly by hedgehog expression in the developing photoreceptors posterior to the furrow (Kumar and Moses, 2001).  Formation of each ommatidium is accomplished by the accretion of cells around a single photoreceptor precursor (the R8 cell), which differentiates due to expression of atonal (ato) in response to hedgehog signaling (Freeman, 1997).  Cells surrounding this precursor are progressively recruited to the ommatidia, undergo differentiation and then contribute to the further development of the ommatidia by assisting in the recruitment of additional nearby cells.  This recruitment is accomplished by the expression of Spitz (Freeman, 1994), a ligand of the Drosophila EGF receptor (EGFR), by all differentiated  7 cells beginning with the R8 photoreceptor (Spencer et al., 1998).  Importantly, cell fate upon activation of EGFR by Spitz is dependent on the developmental stage; near the furrow EGFR activation stimulates cells to become photoreceptors, while in regions further behind the furrow, where cells are developmentally more mature, they become the non-neuronal cone cells (Freeman, 1996).  In pupae, activation of EGFR initially commits cells to a primary pigment cell fate, and later secondary or tertiary pigment cells.  The activity of Spitz is limited to the region immediately surrounding the differentiated cells that express the ligand (a range of approximately 3-4 cell diameters) (Freeman 1994).  Differentiated cells also express Argos , which is an extracellular inhibitor of EGFR and has a larger distance of activity (over 10 cell diameters) (Freeman, 1994; Spencer et al., 1998).  Thus, cells in the immediate proximity of the developing ommatidia are stimulated by the Spitz ligand to differentiate, while more distant cells are inhibited from beginning this process by Argos.  In this way each ommatidium can progressively develop and expand the number of cells it contains while ensuring that a pool of undifferentiated cells (the interommatidial cells or IOCs) surrounds it (Freeman, 1997) (Figure 1.2).  Continued expression of Spitz by the cone cells and primary pigment cells of each ommatidium acts in part as a survival signal that protects nearby IOC cells so they can survive to differentiate into secondary and tertiary pigment cells - ablation of primary pigment cells and cone cells results in widespread death of IOCs (Miller and Cagan, 1998).  Acting in opposition to this survival signal IOCs themselves express Delta, which activates the Notch signaling pathway and promotes PCD in neighboring IOCs; mutations that reduce Notch or Delta activity in young pupae block this PCD (Cagan and Ready, 1989b). The Notch signaling pathway is used iteratively throughout retinal development and its regulation is complex (Lai, 2004; Le et al., 2005; Bray, 2006; Le, 2006).  Among other functions, Notch activity is required for differentiation of PPCs and correct sorting and apoptosis of IOCs (Cagan and Ready, 1989b; Yu et al., 2002) and loss of Notch signaling can  8 result in redistribution of Roughest to the IOC-IOC contacts during the cell sorting process (Gorski et al., 2000; Grzeschik and Knust, 2005).  Of the two known Notch ligands, Delta expression appears to be required for normal eye development (Gorski et al., 2000; Tanenbaum et al., 2000) while Serrate does not (Baonza and Freeman, 2005). Prior to PCD, Roughest is expressed within the pigment cells and localizes specifically to the cell contact boundaries between primary pigment cells and the secondary and tertiary pigment cells (Gorski et al., 2000).  This localization is required for correct cellular organization of the IOC lattice (cell sorting) prior to PCD; mutations in either Notch/Delta or Roughest inhibit this reorganization and block subsequent PCD (Gorski et al., 2000).  Once the IOC cells are correctly organized approximately 1/3 undergo selective PCD to perfect the cellular patterning (Figure 1.2 and 1.3); mutations in the echinus locus block PCD at this step and surviving cells instead differentiate into secondary or tertiary pigment cells (Wolff and Ready, 1991b).  Thus, the spatial regulation of PCD in the retina is accomplished through the interplay between opposing anti-apoptotic (EGFR activation) and pro-apoptotic (Notch activation) signals, as well as appropriate cell-cell contacts mediated by Roughest.  9 Figure 1.2   Figure 1.2 | Pattern formation in the pupal eye. The apical surface of eyes at different stages of pupal life. Gray cells are uncommitted interommatidial cells (IOCs) and coloured cells are determined. (A) At the beginning of pupation, the cone cells (blue and yellow) are embedded within a mosaic of IOCs. (B) After 20 hours of pupation, PP cells (orange) enlarge and start surrounding the cone cells. (C) After 30 hours of pupation, PP cells start contacting each other. (D) After 40 hours of pupation, cone and primary pigment cells (PPCs) enlarge, and IOCs sort into single rows between them. IOCs positioned in certain niches differentiate into secondary pigment cells (SPCs, green) and tertiary pigment cells (TPCs, red). IOCs that are not committed undergo apoptosis. (E) After 60 hours of pupation, pattern formation is complete. [Reprinted from Current Opinion in Genetics and Development, 17, R.W. Carthew, Pattern formation in the Drosophila eye, 309-313, Copyright 2007, with permission from Elsevier.]  10 Figure 1.3   Figure 1.3 | Cell topologies and geometries in the eye. Schematic of an ommatidium at 60 hours of pupal life. A cross-section view (A) is at the level of the adherens junction (AJ), and a side view (B) is equatorial to the midplane. A central group of cone cells are laterally surrounded by the two PP cells. The cone cell group sits over a cluster of eight photoreceptor cells (R) and under the lens (L). A lattice of SP and TP cells, and bristles are indicated. Nuclei in their defined positions are indicated as ovals. The schematics do not contain the small apical and basal processes of photoreceptor and cone cells that project from top to bottom.  PP = primary pigment cell, SP = secondary pigment cell, TP = tertiary pigment cell. [Reprinted from Current Opinion in Genetics and Development, 17, R.W. Carthew, Pattern formation in the Drosophila eye, 309-313, Copyright 2007, with permission from Elsevier.]  11  The fly retina thus undergoes a highly specific process of cell migration (cell sorting) and programmed cell death (PCD) during the larval and pupal stages (Figure 1.2), which sculpts the tissue into its precise arrangement of cells (Cagan and Ready, 1989a; Wolff and Ready, 1991b; Wolff and Ready, 1993) (Figure 1.3).  The structure of the eye makes it particularly useful for the analysis of PCD and cell-cell interactions.  During the mid-pupal stage the ommatidial cells are arranged in a monolayer, which allows much more precise resolution and detection of cell numbers and positions than does the fully developed adult eye with its more complex morphology.  In the adult eye changes in cell death during its development are detectable either as a smaller (or absent) eye due to increased cell death or an eye with a rough appearance due to the presence of additional cells.  Even subtle changes in the number and positioning of cells in the eye can produce visible changes in its morphology, which makes it a sensitive detector of changes in cell death pathways (Wolff and Ready, 1991b; Bonini and Fortini, 1999). Programmed cell death occurs in four stages during eye development.  Initially, a low level of PCD is observed in third-instar larval discs among the undifferentiated cells immediately anterior to the morphogenetic furrow (Wolff and Ready, 1991a).  The significance of this PCD is unknown.  During the early- to mid-pupal stage two rounds of PCD are observed throughout the retina.  The first is regulated by Notch and Wingless signaling and appears prior to rearrangement of the IOCs (Wolff and Ready, 1991b; Cordero et al., 2004).   The next stage of PCD, and the best characterized, occurs approximately eight hours later once cell sorting is complete.  This stage involves the death of 1500-2000 excess IOCs (Cagan and Ready, 1989a; Wolff and Ready, 1991b) in a highly selective manner that is required for correct patterning of the eye.  A recent study has demonstrated positions within the lattice that highly favor induction of apoptosis at this stage, and suggest that Notch and EGFR signaling are required to sensitize cells to an unidentified death signal from a specific  12 subcellular domain within the primary pigment cells (Monserrate and Brachmann, 2007).   A final round of PCD is observed along the periphery of the eye late in development, resulting in the death of cells within incomplete ommatidia at the edges of the cellular lattice (Wolff and Ready, 1991b). The cell death observed during eye development is characteristic of apoptosis.  Dying cells display membrane blebbing, nuclear condensation and cytoskeleton collapse.  Dying cells and the cellular remains may be visualized with the vital dye acridine orange, which can be employed in unfixed tissue to selectively stain dead cells with disrupted membranes (Wolff and Ready, 1991b; Abrams et al., 1993), or with the TUNEL (TdT-mediated dUTP-Nick-End Labeling) apoptosis assay (Chen et al., 1996) that detects DNA fragmentation.  Retinal PCD can be inhibited by ectopic expression of the viral caspase inhibitor P35; IOC cells expressing P35 survive and differentiate into supernumerary pigment cells (Hay et al., 1994). Prior to the second round of PCD during the pupal stage, Roughest is expressed within the pigment cells, and localizes specifically to the cell contact boundaries between 1° pigment cells and the 2°/3° pigment cells (Figure 1.4).  This localization is required for correct cellular organization of the IOC lattice prior to PCD and mutations in either Notch/Delta or roughest inhibit this reorganization and block subsequent PCD (Gorski et al., 2000).  Roughest facilitates this reorganization through heterophilic interactions with Hibris, which is expressed in primary pigment cells (Bao and Cagan, 2005).  DE-cadherin within the cell contacts is also a regulator of Roughest by determining its distribution within the IOC contacts, and recent studies outline a similar role for TGF-β signaling (Cordero et al., 2007) and the CIN85 orthologue Cindr (Johnson et al., 2008).  Thus, recent studies have identified the stability and dynamic regulation of cell contacts as key to the successful reorganization of cells preceding cell death.  13 Figure 1.4   Figure 1.4 | AJ dynamics in IPCs.    (a,b) Four pupal stages (%pd, pupal development) are shown, during which IPCs rearrange and lock into their final pattern. (a) First, AJs are uniform between all cells of the neuroepithelium. When IPCs rearrange, AJs decline and become undetectable between IPCs using markers such as DE-cad. After IPCs are patterned, AJs are again uniform between all cells. A single, vertex-invading IPC is labeled. Note that this cell is specified as a tertiary pigment cell only after it has successfully competed with other IPCs and contacts three primary pigment cells. (b) The lower four panels show the distribution of Roughest and Hibris during IPC rearrangement. Hibris expression is found in cone cells and primary pigment cells but its protein distribution is not known. Roughest distribution in IPCs follows that of AJs in the first three stages shown; it declines at the interfaces between IPCs and remains enriched at the interface between IPCs and primary pigment cells. In contrast to AJs, Roughest does not redistribute into the interfaces between IPCs once the final pattern of IPCs has been locked in. AJ = adherens junction; IPC = interommatidial precursor cell = IOC [Reprinted from Trends in Cell Biology, 17, U. Tepass and K. P. Harris, Adherens junctions in Drosophila retinal morphogenesis, 26-35, Copyright 2007, with permission from Elsevier.]   14 1.5) Epithelial adherens junctions The adherens junction (AJ), also known as the zonula adherens, is an essential structure in polarized epithelial cells. Adherens junctions function in cell adhesion, contribute to maintenance of the apical–basal cell polarity and provide a site for cell-cell signaling (Perez- Moreno et al., 2003; Gumbiner, 2005; Wang and Margolis, 2007).  They are highly organized structures that encircle the apical–lateral boundary of cells and are composed of the transmembrane proteins of the cadherin family (Figure 1.5).  The extracellular domain of cadherin mediates cell adhesion by forming homophilic trans-dimers in a Ca2+-dependent manner (Gumbiner, 2005) (Figure 1.6).  The cytoplasmic domain of cadherins is associated with cytoplasmic proteins and the actin cytoskeleton (Mege et al., 2006; Weis and Nelson, 2006).  The primary binding partners of cadherin are β-catenin (Armadillo) and p120catenin. After association with cadherin, β-catenin then binds α-catenin, which mediates interaction with the actin cytoskeleton in a poorly-understood manner (Gates and Peifer, 2005; Miyoshi and Takai, 2008).  The integrity and dynamic regulation of these junctions are vital for correct development of the Drosophila retina (Tepass and Harris, 2007), and part of this process involves a dynamic regulation of the cytoskeleton and its association with the cadherin complex (Mege et al., 2006).     15 Figure 1.5   Figure 1.5 | Schematic of epithelial cell structure in a late-embryonic/larval epidermal cell. Subdomains of the plasma membrane and cellular junctions are indicated to the left and the distribution some junction-associated proteins are listed to the right. Abbreviations: AJ, adherens junction; Crb, Crumbs; DE-cad, DE-cadherin; GJ, gap junction; MZ, marginal zone; Nrx IV, Neurexin IV; SJ, septate junction. [Figure and text adapted from (Tepass et al., 2001).]       16 Figure 1.6  Figure 1.6 | The cadherin–catenin protein complex. Cadherin in the outer membrane is found as a parallel or cis homodimer. The extracellular region of classic cadherins consists of five cadherin-type repeats (EC domains; extracellular cadherin domains) that are bound together by Ca2+ ions (yellow circles) to form stiff, rod-like proteins. The core universal- catenin complex consists of p120 catenin, bound to the juxtamembrane region, and β-catenin, bound to the distal region, which in turn binds α-catenin. In a less well understood way, α- catenin binds to actin and actin-binding proteins, such as α-actinin or formin-1. Cadherins and catenins have also been found to interact with many other proteins, especially signaling proteins, in various cell types or circumstances.  Ena/VASP proteins such as Enabled (Ena) are often found in close association to the cadherin-catenin complex, possibly through an association with α-catenin.  [Figure and text adapted from (Gumbiner, 2005).]   17 1.6) Ena/VASP proteins: regulators of the actin cytoskeleton The Ena/VASP (Enabled / vasodialator-stimulated protein) proteins, of which the Drosophila protein Enabled was the first to be described, are a conserved family of proteins found in vertebrates and invertebrates (Krause et al., 2003).  A common structural feature of these proteins is the EVH1 domain, which binds to proline-rich (F/LPPPP) motifs.  The molecular function of these proteins is to stimulate extension of actin filaments at barbed ends by disrupting the function of actin capping proteins (Kwiatkowski et al., 2003; Pasic et al., 2008).  Proteins containing EVH1-domains have been implicated in many different cellular processes including axon guidance, phagocytosis, cell-cell adhesion and intracellular movement of the pathogen Listeria monocytogenes.  Ena/VASP proteins are often found at the edge of lamellipodia where they stimulate extension of unbranched actin fibres and help to promote motility.  In mammalian cells, a major function of the Enabled homologue, along with its regulator Abelson, is to regulate the stability of adherens junctions, and a similar role of Enabled in Drosophila has been proposed (Grevengoed et al., 2003).  1.7) Ubiquitination Ubiquitin is a 76 amino-acid peptide that can be conjugated to proteins as a post- translational modification.  Initially the process of ubiquitination was thought to have a single purpose – the degradation of unwanted proteins by the proteasome (Hochstrasser, 1996; Pickart, 2001).  More recently this process has been associated with several cellular functions including apoptosis, protein transport, activation of protein kinases, DNA repair and chromatin modification (Hicke and Dunn, 2003; Soboleva and Baker, 2004; Sun and Chen, 2004).  Ligation of the ubiquitin monomer to a substrate is accomplished by a series of enzymes (Figure 1.7).  18 Figure 1.7   Figure 1.7 | The ubiquitination pathway. Free ubiquitin (Ub) is activated in an ATP- dependent manner with the formation of a thiol-ester linkage between E1 and the carboxyl terminus of ubiquitin. Ubiquitin is transferred to one of a number of different E2s. E2s associate with E3s, which might or might not have substrate already bound. For HECT domain E3s, ubiquitin is next transferred to the active-site cysteine of the HECT domain followed by transfer to substrate (S) (as shown) or to a substrate-bound multi-ubiquitin chain. For RING E3s, current evidence indicates that ubiquitin might be transferred directly from the E2 to the substrate. [Reprinted by permission from Macmillan Publishers Ltd: Nature Reviews Molecular Cell Biology, 2, 169-178, A.M. Weissman, Themes and variations on ubiquitylation, copyright 2001.].   19 The first two enzymes, E1 and E2, activate the ubiquitin monomer and bind to it in preparation for transfer to the substrate.  These enzymes are relatively general in nature as it is the E3 enzyme that specifies the substrate and promotes transfer of the activated ubiquitin peptide.  E3 enzymes utilize one of two known ubiquitin ligase catalytic domains - the RING finger or the HECT domain.  Ubiquitin itself can be a target for ligation, resulting in a polymer of ubiquitin peptides forming on a substrate (“polyubiquitination”).  It is these polyubiquitin chains that are thought to target proteins for degradation by the proteasome. Monoubiquitination of a substrate, however, is often associated with protein-protein interactions and regulation of protein localization, and a number of ubiquitin-interacting domains have been described (Andersen et al., 2005; Hurley et al., 2006).  Similar to phosphorylation, modification of proteins by ubiquitin can be reversible via de-ubiquitinating proteases that remove the peptide from substrates (Hu et al., 2002; Soboleva and Baker, 2004; Nijman et al., 2005).   As a method for targeting a protein to a specific location this reversible quality of ubiquitination gives it a distinct advantage over signals incorporated into the primary amino acid sequence; proteins can be targeted in a much more dynamic way.  1.8) echinus The echinus locus was initially defined (Patterson 1932) as the result of an X-ray generated mutation that caused a rough-eye phenotype.  Later, a spontaneous mutation led to the isolation of another allele (ec1), which has been more extensively studied (Wolff and Ready, 1991b; Reiter et al., 1996). The rough-eye phenotype is the result of insufficient IOC death following cellular reorganization; echinus mutants occasionally display ommatidia with a third primary pigment cell, or a fifth cone cell, and surplus secondary and tertiary pigment cells are common (Wolff and Ready 1991).  These additional cells give the adult eye a mildly rough and disorganized appearance, although the absolute number of ommatidia in each eye  20 remains normal.  Other phenotypes can also be attributed to echinus, such as changes to wing and body shape as well as alterations in bristle patterns (Sturtevant, 1970).   These additional phenotypes suggest that its role is not limited to retinal development. During development, once the IOC cells have correctly sorted approximately 1/3 undergo PCD to perfect the cellular patterning; mutations in the echinus locus block PCD at this step, and surviving cells instead differentiate into secondary and tertiary pigment cells  (Wolff and Ready, 1991b). The block in PCD observed in roughest mutants suggests that echinus acts downstream of roughest, and that its role in cell death is dependant upon correct cellular organization (Reiter et al., 1996). Mutations in echinus may act to enhance cell survival signals or block death signals.  It is possible that mutations in echinus may block cell death by promoting or strengthening survival signaling from the primary pigment cells (Cagan and Ready, 1989b; Yu et al., 2002).  Understanding echinus function may provide insight into mechanisms by which extracellular signals are used to determine cell fate.  1.9) Objectives, specific aims and hypotheses The objectives of this study were to identify the gene echinus and characterize its function in Drosophila retinal development.  Mutation of echinus has previously been shown to reduce the level of programmed cell death in the developing Drosophila retina, suggesting that this gene plays an important role in this process.  The Drosophila retina is used extensively as a model system for studying programmed cell death, cell differentiation and cell-cell communication.  Identifying echinus and characterizing its function will lead to a better understanding of the development of this complex tissue.  In addition, since defective programmed cell death is an important contributor to the development and progression of diseases such as cancer, a better understanding of the role of echinus in this process may also  21 provide insights into these diseases.   To achieve these objectives, the following specific aims and hypotheses were developed:  1.9.1) Specific aim 1) Identify the gene or genomic region responsible for the echinus rough-eye phenotype.  Mapping, positional cloning techniques and expression analyses will be employed to locate the gene, or genomic region, corresponding to the echinus mutation. Hypothesis:  echinus is a novel gene involved in programmed cell death, or represents a novel mutation and associated function for a previously characterized gene.  1.9.2) Specific aim 2) Characterize the role of echinus in development of the Drosophila retina.   The effect of echinus mutations on apoptosis and tissue structure during development of the retina will be examined using immunofluorescence.  A search for genetic interactions between echinus and known regulators of PCD and retinal development will also be conducted. Hypothesis: The phenotypic effects of echinus mutations on the development of the Drosophila retina, and echinus genetic interactions with known regulators of retinal development, will help define the role that echinus plays in retinal development and mechanism by which it regulates PCD.  1.9.3) Specific aim 3) Characterize the molecular function of echinus.    Proteins interacting with Echinus will be isolated by immunoprecipitation and identified by mass spectrometry. The protein sequence of Echinus will be examined for possible functional domains or similarity to proteins in Drosophila and other species.  Hypothesis: The echinus gene product contains functional domains that physically interact with other proteins that will provide insight into Echinus function and its role in eye development and PCD.  22 Chapter 2 | Echinus alleles are associated with mutations in the gene CG2904, which has homology to deubiquitinating enzymes 1  2.1) Introduction The first phase of this study was to identify the gene responsible for the echinus phenotype.  Mutations in echinus result in abnormal development of the retina and give the adult eye a disordered and rough appearance.  The molecular function of echinus is unknown but it has been shown to reduce the amount of programmed cell death during the pupal stage of eye development (Wolff and Ready, 1991b).  The Drosophila retina is used extensively as a model system for studying the development of complex tissues (Larson et al., 2008), and programmed cell death is an essential part of eye development (Brachmann and Cagan, 2003; Monserrate and Brachmann, 2007) and the pathology of diseases such as cancer (Igney and Krammer, 2002; Vidal and Cagan, 2006).  Thus, it is hypothesized that the identification and characterization of echinus will further our understanding of Drosophila eye development and expand our understanding of the cellular processes responsible for programmed cell death. Echinus has previously been mapped genetically to a region on the distal end of the X chromosome (1:5.5) and more precisely by deficiency mapping to the cytological bands 3E8- EF1.  The previously isolated alleles of echinus (ec1, ec64d) have the characteristics of loss-of- function mutations since they are not complemented by genomic deficiencies.  Previous to this study, two separate and independent attempts to identify the echinus locus were made.   Kramer   1 A portion of the data in this chapter has been published.  The characterization of ec1, ec3c3, ec56 and ecΔ9 mutations is contained in: Copeland, J. M., Bosdet, I., Freeman, J. D., Guo, M., Gorski, S. M. and Hay, B. A. (2007). echinus, required for interommatidial cell sorting and cell death in the Drosophila pupal retina, encodes a protein with homology to ubiquitin-specific proteases. BMC Dev Biol 7, 82.  23 et al. (personal communication) employed insertional mutagenesis with a transposable P element, followed by a screen for rough eye mutants.  They recovered an insertional mutant (roX119R) that displayed both the adult and pupal echinus phenotype, and precise excision of the P element restored wild-type appearance of both pupal and adult eyes.  An inverse-PCR based rescue method was used to clone, and obtain sequence from, the insertion site and these data were used to identify a cDNA that corresponded to the disrupted gene.  The isolated cDNA corresponded to roX1, a male-specific non-coding RNA transcript (Meller et al., 1997). Attempts to complement the echinus mutant using roX1 cDNA clones failed (Kramer et al., unpublished data).  Large genomic fragments (>11kb) spanning the roX1 region were also unable to complement the ec1 mutation.  Imprecise excision of the inserted P element created a roX1 deletion mutant (roX1Δ9) that retained the echinus phenotype, although this mutant could also not be complemented with either roX1 cDNA clones or genomic fragments (Kramer et al., unpublished data).  A similar finding was obtained by Judd et al. (personal communication and unpublished data) who identified, in a strain carrying the spontaneous ec1 allele, a copia insertional element that mapped near the location of echinus as predicted by cytological and genetic mapping.  Sequence flanking the insertion site of the copia element was used to probe a cDNA library, and this also identified roX1 as the disrupted gene.  However, the roX1 cDNA clone also did not complement ec1 (Judd et al.,  unpublished data). The Drosophila genome has been mapped and sequenced to a high level of quality and contiguity (Adams et al., 2000; Celniker et al., 2002), providing a valuable resource for novel gene discovery and genomic analyses.   A large number of gene expression studies, including EST and cDNA sequencing, microarray experiments, SAGE analyses and protein interaction data have provided a wealth of data regarding the structure and expression of known and novel genes (Flybase, 2002; Matthews et al., 2005; Breitkreutz et al., 2008). In addition, advances in the computational analysis of genomic sequence data have permitted rapid and  24 accurate prediction of a large number of previously unknown genes.  These data provided an essential starting point for the identification of echinus and a detailed characterization of multiple alleles was undertaken to determine the gene responsible for the rough eye phenotype in echinus mutants.    25 2.2) Materials and Methods 2.2.1) Drosophila Strains and Genetics Drosophila strains were maintained under standard laboratory conditions on yeast-extract and cornmeal agar medium at approximately 21°C unless otherwise specified.  The wild type strain used for all experiments is OreR (Bloomington).  The alleles ec1 and ec64d (Bloomington) have been described previously. The ec1 allele (Sturtevant, 1970) was isolated as a spontaneous mutation and its precise nature had not previously been characterized.  The ec64d allele (Lefevre and Wilkins, 1966) was produced using X-ray mutagenesis that created an insertional transposition of a region of the X chromosome (bands 3C2-3E8) into chromosome 2 (band 37D).    The ec3c3 allele was generated by EMS mutagenesis as previously described (Morris et al., 1999).  The ec19R allele resulted from a P-element insertion within the gene roX1 and the ecΔ9 deletion allele resulted from imprecise excision of this P-element and was predicted by Southern blot analysis to be approximately 3kb in size (H. Kramer, unpublished).  The roX1ex6 strain carries a deletion 1.4kb in size that removes the transcriptional start and 5’ region of the RNA gene roX1 (Kelley et al., 1999).  I confirmed, using complementation tests, that ec3c3, ec19R and ecΔ9 are alleles of the same gene.  2.2.2) Genomic and Plasmid DNA Purification and PCR For genomic DNA isolation, whole pupae were removed from the puparium at 24hrs APF at 25°C, placed in DNAzol reagent (Invitrogen), homogenized and centrifuged at 14,000g for 15 minutes to remove insoluble matter.  DNA isolation was performed immediately thereafter as per the manufacturer’s instructions.  cDNA plasmid DNA (BDGP) was obtained from E. coli grown in LB medium supplemented with the appropriate antibiotic and isolated via  26 a standard alkaline lysis methodology (Sambrook et al., 1989). Primers (IDT) were designed using Primer3 (Untergasser et al., 2007) (Table 2.1).  DNA was quantitated using a spectrophotometer and amplified using Platinum Taq or Platinum Pfx polymerase (Invitrogen) as per the manufacturer’s recommended protocol.  Products were separated via horizontal gel electrophoresis (Owl Separation Systems) using 1% Ultrapure Agarose (Invitrogen) and TBE buffer, stained with SYBR-Green (Invitrogen) and visualized on a Fluorimager 595 scanner (Molecular Dynamics).   27 Table 2.1 del-amplicon1 F 5’-GCTCCATTAGCTTGGACCTG-3’ del-amplicon1 R 5’-AACATGGCACTGTGGAAATG-3’ del-amplicon2 F 5’-TTTTGTTGCCGCTGTAATTG-3’ del-amplicon2 R 5’-CAGAGCACAAGGGGCTTATC-3’ del-amplicon3 F 5’-TACAGGGTGCTTCGTCTGTG-3’ del-amplicon3 R 5’-ACCACCTAGCTGCGATCTTC-3’ del-amplicon4 F 5’-TGAAAATCGTCGCACTGAAG-3’ del-amplicon4 R 5’-AAAGGTAAAGATCGGCAACG-3’ del-amplicon5 F 5’-GCGCTCTAAAGGTGAGTTCG-3’ del-amplicon5 R 5’-AGTGGCCCAGGACTACAGG-3’  Table 2.1 | Primer sequences for deletion-mapping amplicons   28 2.2.3) DNA Sequencing The coding portions and flanking DNA of CG2904 were amplified using PCR with Platinum Taq polymerase (Invitrogen), sequenced using the ABI Big Dye terminator cycle sequencing kit (v3.1) at 1/16th or 1/24th standard dilution and analyzed on 3100, 3700 and 3730XL sequencers (Applied Biosystems). Sequence data were assembled using the Phred/Phrap (Ewing, 1998a; Ewing 1998b) software package and the resulting sequence contigs were visualized using Consed (Gordon, 1998) and compared to the published Drosophila genomic sequence. Both strands were completely sequenced for each exon and any ambiguities or mutations were re-sequenced.  cDNA clones were sequenced using M13forward and M13reverse primers corresponding to sites in the vector flanking the insert. End sequences were aligned to the genome using BLAST (Altschul et al., 1990) (  2.2.4) RNA Isolation and Realtime RT-PCR Pupae were removed from the puparium at 26hrs APF at 25°C, placed in Trizol reagent (Invitrogen), homogenized and centrifuged at 14,000g for 15 minutes. RNA isolation from Trizol-solubilized tissues was performed as per manufacturer’s instructions.  Reactions used SYBR Green One-Step PCR reagent (ABI) and contained 50ng of DNase-treated total RNA, and 0.1uM of each primer in a total volume of 15uL.  Strand-specific reactions were performed using a modified two-step procedure with a single primer added during the reverse transcriptase reaction and the second added prior to the PCR amplification stage.  Primers (IDT) to transcripts within the echinus candidate region (Figure 2.2) were designed using Primer3 (Untergasser et al., 2007) (Table 2.2).  29 Table 2.2 CG2904 E1 F 5’-CAAATTAATCCTGGATGGAGCC-3’ CG2904 E1 R 5’-TAGCCCATTACGGCCAATTC-3’ CG2904 E5 F 5’-CCCAGCCAGCATCGAAATATAT-3’ CG2904 E5 R 5’-TCGTACTCGATTGCGTTGTCA-3’ SD15943 F 5’-CTGCATCGCTAAGCTTTAGCAC-3’ SD15943 R 5’-TCAATGCCAATGTTCCCAAG-3’ roX1_3’ F 5’-CCAGTCCCCTTCCTTGACTTT-3’ roX1_3’ R 5’-GCCTTTTAATGCGTTTTCCG-3’ roX1_5’ F 5’-ACACAGCTTCAAGCGCATTACT-3’ roX1_5’ R 5’-TAAGAATTGCAGCTCAGGCG-3’ RH68894 F 5’-CGGAATATGGTCAAGACGGAAG-3’ RH68894 R 5’-TAATCGATGGCTAATGCGGC-3’ Yin F 5’-TGGCACGTTGTGGAATAAGG-3’ Yin R 5’-CCATTTTCATCTCAAGACCCG-3’ rp49 F 5'-AGTCGGATCGATATGCTAAGCT-3' rp49 R 5'-AGATACTGTCCCTTGAAGCGG-3’  Table 2.2 | RT-PCR forward (F) and reverse (R) primer sequences  30 Amplicons were approximately 250 base pairs in size.  All amplifications were performed in triplicate on an Applied Biosystems 7900 Sequence Detection System and melting curve analysis was employed to ensure that a single product of the expected melting temperature was obtained. Results were standardized against CT values of a reference primer set (Drosophila rp49).  2.2.5) Computational analysis of protein sequence data Protein sequences of potential homologues were identified using BLAST ( (Altschul et al., 1990; Altschul et al., 1997) and multiply aligned using ClustalW ( (Larkin M.A., 2007).   31 2.3) Results 2.3.1) The echinus mutation is localized to the 3F2-3F3 region of the X chromosome Echinus had previously been mapped genetically (1-5.5) (Lawlor, 1980) and cytologically (3E8-F1) (Lefevre and Wilkins, 1966; Oliver et al., 1988) to a 200-250kb region of the X chromosome, an area that contained approximately 15 genes of both known and unknown function.  To refine the echinus candidate region I conducted an examination of all available deficiency mapping data (Figure 2.1).  This analysis suggested that echinus might instead reside in the 3F2-3F3 region.  This region is approximately 70 kilobases in size and contained 8 known genes (Figure 2.2). The ec19R transposon insertion within the gene roX1 (and the associated ecΔ9 deletion of roX1) in the 3F3 region supported the hypothesis that the echinus phenotype is caused by mutations in roX1 or a nearby gene.  Since mutations in roX1 have no observable phenotype and mutations in yin, the only known gene near roX1, have no phenotype in the retina, I hypothesized that echinus was an uncharacterized gene in the region nearby to roX1.    32 Figure 2.1   Figure 2.1 | Map of deficiencies in the echinus candidate region.  Schematic shows the location of breakpoints of genomic deficiencies mapped against echinus (Flybase, 2002). Deficiencies that complement echinus are shown in green and those that fail to complement echinus are in red. Dotted lines at deficiency ends indicate that they extend outside this genomic region.  Scale shows the cytological band position.  The double-arrowed line represents uncertainty in the distal breakpoint of Df(1)cho19.   These data suggest that echinus is located in the 3F2-3F3 region of the X chromosome (vertical dotted lines).  33 Fi gu re  2 .2    Fi gu re  2 .2  |  O rg an iz at io n  o f th e ec hi nu s c an di da te  r eg io n . Th e re gi on  o f th e ve rs io n 2. 0 re fe re nc e ge no m e (J an . 2 00 3)  p re di ct ed  to  co nt ai n ec hi nu s d is pl ay ed  in  th e U C SC  G en om e Br ow se r (h tt p: // ge no m e. uc sc .e du ) (K en t e t a l.,  2 00 2) . Th e ca nd id at e re gi on  c ov er s ap pr ox im at el y 70 kb  a nd  c on ta in s 8 kn ow n ge ne s.   T he  p os it io ns  o f pr ot ei n- co di ng  a nd  n on -c od in g ge ne s ar e in di ca te d in  b lu e an d G en sc an  ( Bu rg e an d K ar lin , 1 99 7)  g en e pr ed ic ti on s ar e sh ow n in  b ro w n.   T he  tw o sp lic e fo rm s of  C G 29 04  a re  b as ed  o n co m pu ta ti on  pr ed ic ti on s us in g G en ie  ( A da m s et  a l.,  2 00 0;  R ee se  e t a l.,  2 00 0) . Th e lo ca ti on s of  r ep et it iv e el em en ts  a nd  lo w -c om pl ex it y D N A , a s pr ed ic te d by  R ep ea tM as ke r (S m it , 1 99 6- 20 04 ),  a re  in di ca te d in  b la ck .  34 2.3.2) Two transcripts are present on the opposite strand to CG2904 In addition to the fully sequenced Drosophila genome, numerous gene expression analyses have provided a wealth of data for novel gene discovery (Matthews et al., 2005).  To further characterize the genomic region surrounding roX1, available gene expression and EST data were analyzed.  Adjacent to roX1 is a gene prediction (CG2904) based on ORF analysis and homology to proteins containing ubiquitin hydrolase enzymatic domains.  In addition to this predicted gene, my analysis of EST data within the candidate echinus region indicated that two transcripts (RH68894, SD15943) were present in an antisense orientation to the predicted gene CG2904.  Numerous examples of partially or completely overlapping transcripts exist in both the Drosophila and mammalian genomes and some of these sense-antisense pairs can interact in complex ways to alter gene expression and translation (Numata et al., 2007).  I obtained cDNA clones corresponding to RH68894 and SD15943, sequenced their ends and aligned these sequences to the genome (Figure 2.3).  Comparison of the cDNA size to the resulting alignments indicated that both genes likely contained no introns and sequence analysis showed that neither contained an open reading frame with significant similarity to known proteins of greater than 50 amino acids in size, suggesting that neither codes for a functional protein.  Only a small region (53bp) separates the two transcripts.  The presence of these overlapping transcripts with no apparent protein-coding potential raised the possibility that they had a functional role in modulating CG2904 expression levels and might be related to echinus function.   35 Figure 2.3   Figure 2.3 | QRT-PCR analysis of the four genes in the echinus candidate region.  Schematic of the transcripts targeted for expression analysis by quantitative RT-PCR.  Gene models are based on version 2.0 of the Drosophila genome assembly and arrows indicate the direction of transcription.  The CG2904 prediction consists of five exons and the antisense transcripts RH68894 and SD15943 overlap the majority of the gene.  Locations of regions amplified for realtime RT-PCR analysis of transcripts within the echinus candidate region are indicated above. Dotted lines indicate the approximate locations of the roX1ex6 and ecΔ9 deletions and the triangle marks the location of the transposon insertion in ec19R.  An approximate scale with marks in kilobases is shown below.  (SAR = scaffold attachment region repetitive element).  36 2.3.3) The gene CG2904 in the echinus candidate region has altered expression levels in some echinus mutants One possible cause of the defect in echinus mutants is an alteration in the expression level of a gene within the candidate region.  Quantitative realtime RT-PCR was employed to determine if any transcripts within this region had altered expression in echinus mutants.  The predicted gene CG2904 was targeted at both the predicted first exon (E1) and the fifth exon (E5), roX1 at both the 5’ and 3’ end, and both antisense cDNAs.  All tested transcripts were present in OreR, ec1, ec3c3 and ec64d and transcripts in the echinus mutant strains vary up to 2- to 3- fold from wild type OreR (Figure 2.4).  All tested transcripts were absent in ec19R and ecΔ9 except for the first exon of CG2904 in ec19R (Figure 2.4).  These data suggested that the ecΔ9 deletion was larger than expected and removed some or all of CG2904.  As well, the ec19R insertion appeared to disrupt expression of CG2904 and the antisense transcripts in addition to roX1.  An updated gene model based on cDNA data revealed that amplification with the SD15943 primers (Figure 2.3) will additionally target the CG2904 3’ UTR and so two-step (strand specific) realtime RT-PCR was used to assess the relative abundance of the overlapping transcripts (Figure 2.5).  These data indicated that CG2904 and both antisense transcripts were absent in the ecΔ9 strain.  These transcripts were present in ec1, ec64d and ec3c3, with modest differences in relative levels of expression.  In parallel with this approach, deletion analysis and candidate gene sequencing were conducted on multiple echinus mutants and these analyses were successful in identifying echinus (see below).   37 Figure 2.4 0 0.5 1 1.5 2 2.5 roX1-3' roX1-5' cg2904 E1 cg2904 E5 SD15943 RH68894 Primer set R el at iv e ex pr es si on  le ve l OreR ec[1] ec[64d] ec[D9] ec[19R] roX1[ex6]  Figure 2.4 | Quantitative RT-PCR analysis of transcripts within the echinus candidate region.  Primer sets correspond to those indicated in Figure 2.3.  Transcript expression levels are represented as relative to OreR control RNA (from 26hr APF pupae).  roX1 RNA is present in ec1and ec64d but absent in both ecΔ9 and ec19R as well as the roX1ex6 deletion mutant.  All tested transcripts were absent from ec19R and ecΔ9 except for the first exon of CG2904 in the ec19R strain.     38 Figure 2.5 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 cg2904 E5 cg2904 E5 (F) cg2904 E5 (R) SD15943 SD15943 (F) SD15943 (R) RH68894 RH68894 (F) RH68894 (R) Primer set R el at iv e ex pr es si on OreR ec[3c3] ec[1] roX1[ex6] ec[64d] ec[D9]  Figure 2.5 | Strand-specific realtime RT-PCR of CG2904 and antisense transcripts. cDNA was transcribed using a single primer (forward (F) or reverse (R)) before amplification by PCR and quantitation.  Reverse primers are expected to measure the abundance of the targeted transcript and forward primers the abundance of transcripts on the opposite strand.  RH68894 and CG2904-E5 forward primers detected no overlapping transcripts.  The SD15943-F amplification measures CG2904 transcript levels due to an overlap with the 3’ UTR of CG2904.   The SD15943-R amplification showed significantly increased expression in ec3c3 and ec64d.  All transcripts are absent in the ecΔ9 strain.  39 2.3.4) Mutations in the predicted gene CG2904 are associated with the echinus phenotype The ecΔ9 mutation was generated via P-element-mediated excision of a genomic region surrounding the ec19R transposon insertion.  This deletion was predicted to be approximately 3kb in size and centered on the 5’ region of roX1 (H. Kramer, personal communication). However the absence of detectable CG2904 RNA in this mutant suggested that the deletion also encompassed some part of this gene as well.  To better define the extent of the ecΔ9 deletion, I generated a series of PCR amplicons across the echinus candidate region.  This analysis indicated that all of CG2904 and much of roX1 had been lost (Figure 2.6) in both the ec19R and ecΔ9 strains.  Since roX1 is a redundant gene with no known phenotype when deleted, these data implicated CG2904 as the gene responsible for the echinus phenotype.  Although CG2904 is transcribed at normal levels in other echinus mutants, those alleles arose via spontaneous mutation and may result from single-base changes that disrupt the protein-coding potential of the gene without altering transcript abundance.  Sequence analysis of CG2904 in the lines ec56 and ec3c3, which carried mutant alleles of echinus generated via EMS, and ec1, an allele that spontaneously arose within a wild-type stock, revealed point mutations within the coding region of CG2904.  The ec1 chromosome also contains a copia transposon insertion 240bp downstream of the CG2904 3’ UTR (Figure 2.7), although the effect of this insertion upon CG2904 function is unknown. These point mutations results in stop codons that lead to substantial truncations of the CG2904 protein.   40 Figure 2.6   Figure 2.6 | PCR amplification of genomic DNA reveals a 22.5 kb deletion in ecΔ9 and ec19R. Genomic DNA from ecΔ9 (lanes 1-5), OreR (top, lanes 6-10) and ec19R (lanes 11-15) were amplified with five sets of primers (Table 2.1 and Figure 2.7) to assess the presence of genomic sequence within the echinus candidate region (amplicon 1, lanes 1,6,11; amplicon 2, lanes 2,7,12; amplicon 3, lanes 3,8,13; amplicon 4, lanes 4,9,14; amplicon 5, lanes 5,10,15) . PCR amplicons (arrows) were obtained with all primer sets in OreR, indicating that this region remains intact in the wild type strain.  Only the most distal amplicon was obtained with DNA from ec19R and ecΔ9 indicating that the other regions have been lost.  Genomic sequencing was used to confirm these data and precisely map the 22.5 kilobase deletion in these strains (Figure 2.7).   41 Fi gu re  2 .7    Fi gu re  2 .7  |  L oc at io n  o f C G 29 04  m u ta ti on s in  e ch in us  m u ta n t st ra in s.   T he  lo ca ti on s of  m ut at io ns  a re  in di ca te d he re  r el at iv e to  a  sc he m at ic  o f ec hi nu s s pl ic e fo rm  2  ( se e be lo w ).   A st er is ks  in di ca te  s in gl e- ba se  c ha ng es  le ad in g to  n on se ns e m ut at io ns . R el at iv e to  th e C G 29 04  s pl ic e fo rm  2  c od in g re gi on  th es e ar e ec 56  C A G Æ TA G  ( E9 1Æ st op ),  ec 1  C G A Æ TG A  ( R 26 1Æ st op ) an d ec 3c 3  A A G Æ TA G  (K 75 8Æ st op ).   T he  c op ia  in se rt io n w hi ch  li es  2 40 b do w ns tr ea m  o f th e C G 29 04  3 ’ U TR  in  ec 1  i s al so  s ho w n,  a s is  th e Pl ac Z in se rt io n up st re am  o f th e sp lic e fo rm  2  5 ’ U TR .  T he  a rr ow ed  li ne  in di ca te s th e ex te nt  o f th e 19 ,8 20 bp  g en om ic  d el et io n in  ec 19 R  a nd  ec Δ 9  a nd  th e nu m be re d op en  b ox es  a bo ve  s ho w  th e lo ca ti on  o f PC R  a m pl ic on s us ed  to  m ap  th e di st al  e nd  o f th is  d el et io n.    42 2.3.5) echinus has multiple splice forms Echinus was initially predicted to consist of two splice forms of either four or five exons (Figure 2.2), but my analysis of available cDNA data indicated that these predictions were incorrect.  Instead, three distinct splice forms of echinus were evident from sequenced cDNA clones (Figure 2.8), all of which had three exons in common but differed in the number and location of the exons in the 5’ region of the gene.  Splice form 2 (Ec-SF2) and splice form 3 (Ec-SF3) were initially isolated and sequenced from cDNA libraries by J. Copeland (Copeland et al, 2007).  My analysis of genomic sequence upstream of echinus revealed an open reading frame that contained homology to a portion of the ubiquitin hydrolase domain.  PCR primers were designed from the predicted start site using genomic sequence and a cloned cDNA of this splice form was generated via PCR amplification from a pupal cDNA pool.  This cDNA corresponds to echinus splice form 1 (Ec-SF1), the longest and the only form that contained all residues required for an active ubiquitin hydrolase domain (see below).  Subsequent to this observation a 5’ EST sequence was submitted to Genbank (Exelixis, Genbank Acc: CO186832) that confirmed expression from this locus and correctly described the 5’ UTR of this splice variant.   43 Fi gu re  2 .8    Fi gu re  2 .8  |  e ch in us  e n co de s m u lt ip le  s pl ic e fo rm s.   T he  g en om ic  r eg io n co nt ai ni ng  ec hi nu s i n th e cu rr en t a ss em bl y (B D G P R el ea se  5 .0 ),  di sp la ye d in  th e U C SC  G en om e Br ow se r (h tt p: // ge no m e. uc sc .e du ) (K en t et  a l.,  2 00 2) . Th e po si ti on s of  p ro te in -c od in g an d no n- co di ng  g en es  a re  in di ca te d in  b lu e an d N -S C A N  ( A ru m ug am  e t a l.,  2 00 6)  g en e pr ed ic ti on s ar e sh ow n in  g re en . Th e th re e kn ow n sp lic e fo rm s of  ec hi nu s a re  in di ca te d.   E ac h sp lic e fo rm  h as  a  u ni qu e 5’  e nd  b ut  a ll sh ar e si m ila r 3’  e xo ns  a nd  3 ’ U TR . A lig nm en ts  o f pu bl ic  E ST  da ta  a re  c ol la ps ed  in to  a  s in gl e tr ac k an d di sp la ye d in  b la ck , a s ar e th e lo ca ti on s of  R ep ea tm as ke r (S m it , 1 99 6- 20 04 ) pr ed ic ti on s of  re pe ti ti ve  D N A  e le m en ts  a nd  lo w -c om pl ex it y D N A .   44 2.3.6) Features of echinus inferred protein sequence To gain insight into echinus function, the inferred protein sequence was examined for known protein domains and homology to previously characterized proteins.  The 6.5 kilobase echinus transcript encodes a large protein (Ec-SF1: 1746aa; Ec-SF2: 1712aa; Ec-SF3: 1765aa) (Figure 2.9) with homology to ubiquitin hydrolase enzymes.  All residues known to be required for enzymatic activity of this domain (Hu et al., 2002) are present in Ec-SF1.  Ec-SF2 and Ec-SF3 both lack a key cysteine residue in the N-terminal portion of the domain, and Ec- SF3 additionally lacks a conserved asparagine residue nearby.  These differences suggest that Ec-SF1 encodes a functional enzyme while Ec-SF2 and Ec-SF3 encode non-functional hydrolase enzymes.  However, an in vitro assay using a standard substrate (ubiquitin-β-galactosidase) failed to show deubiquitinating activity for the Ec-SF1 protein and in vivo expression of any echinus variant is sufficient for functional rescue (Copeland et al, 2007), suggesting that either all forms are normally inactive in vivo or that enzymatic activity is not required for Echinus function. The C-terminus of Echinus contains a region rich in glutamine.  This region is 156 amino acids in length and contains 77 glutamine residues (49%), including a polyglutamine repeat 34 amino acids in length.  Also in this region of the protein are a proline-rich motif (1107FPPPPP1112) that matches the canonical binding site for EVH1 protein domains that are known to mediate protein-protein interactions (Niebuhr et al., 1997; Renfranz and Beckerle, 2002; Krause et al., 2003) and a putative caspase-3 cleavage motif (1277DEVD1280) (Villa et al., 1997).  All three of these features are conserved in the insect homologues, albeit with variable polyglutamine repeat lengths.   A human homologue, USP53, also contains a close match to the EVH1-binding motif near its C-terminus (920LPPP923).  45  Figure 2.9   Figure 2.9 | Protein structure of Echinus.  Echinus encodes a protein 1746 (Ec-SF 1), 1712 (Ec-SF2) or 1765 (Ec-SF3) amino acids in length.  The N-terminus contains a ubiquitin hydrolase domain, or at least sequence with homology suggestive of this function.  The Ec-SF1 form of the protein contains all amino acids thought to be required for enzymatic activity (shaded boxes in alignment) (Hu et al., 2002), while the Ec-SF2 and EcSF3 variants are missing a necessary catalytic cysteine residue (arrow).  Homologues in insect and the mammalian USP53 and USP54 genes show conservation of the active site residues as well as surrounding sequences. A proline-rich motif identical to the binding sequence of EVH1 domains is found in the second half of the protein, as well as a consensus sequence for caspase-3 proteases.  The C-terminal portion of the gene contains a large glutamine-rich region including a polyglutamine stretch 34 residues in length.  Asterisks show the location of putative clathrin-binding motifs.  * * *  46 2.4) Discussion The initial aim of this study was to identify the gene underlying the echinus mutant phenotype. Using available deficiency mapping data and partially characterized echinus mutants the echinus mutation was localized to a 70kb candidate region.  Analysis of genome structure and sequence in this region, in addition to expression analysis of genes within the region, revealed that the gene CG2904 is mutated in echinus strains.  Quantitative RT-PCR was also able to demonstrate that this gene is expressed during the developmental period that retinal PCD is observed.  These data indicated that CG2904 encodes echinus.  Subsequent to this study’s identification of CG2904 as echinus, another group reported similar results (J. Copeland, 2005). They demonstrated further that insertion of a CG2904 transgene can rescue the echinus mutation, and that an in vivo RNAi construct complimentary to echinus can produce a similar phenotype (Copeland et al, 2007), which conclusively demonstrate that CG2904 is echinus. Homologues of echinus can be found in many insect species including all 11 other sequenced Drosophila spp., mosquitoes (Anopheles gambaie), honeybee (Apis mellifera), the wasp Nasonia vitripennis and the flour beetle Tribolium castaneum.  Outside of the insect lineage the closest homologues are to the mammalian ubiquitin hydrolases USP53 and USP54 (Quesada et al., 2004), with the majority of homology occurring within the catalytic domain (USP53 shows 48% identity over 353/1073aa).  The presence of multiple splice forms in other insect species is suggested by conservation of open reading frames corresponding to the 5’ region of the three known echinus transcripts.  As well, recent cDNA sequence data indicates the presence of a similar gene structure in both USP53 and USP54.  All these genes display an arrangement of a splice variant with a transcriptional start far upstream of the gene, and coding for an active enzymatic domain, as well as one or more variants that start closer to the remainder of the gene but code for an inactive domain. However, despite this conservation of gene structure, the functional differences between the three splice forms of echinus, if any, remain unclear. All  47 three are capable of rescuing the rough eye phenotype of echinus (Copeland et al, 2007) despite the lack of key catalytic residues in the ubiquitin hydrolase domain of splice forms 2 and 3. This result suggests that the differences may have a subtle or negligible effect on normal development, at least within the retina. That the ubiquitin hydrolase domain apparently does not require activity for the normal function of echinus raises questions about its role.  In addition to proteins that convert ubiquitin polymers into monomers, this domain can also be found in proteins that act as deubiquitinating (DUBs) enzymes (Soboleva and Baker, 2004).  These DUBs play a role in the removal of ubiquitin modifications from specific substrates.  An example of this is the Drosophila protein fat facets, which has been shown to remove monoubiquitin modifications from the epsin-like protein liquid facets, saving it from degradation via the proteasome (Overstreet et al., 2004).  Together fat facets and liquid facets control the rate of endocytosis of the cell surface protein Delta and thus its activity.  Cell-surface proteins may also be conjugated to ubiquitin directly, signaling them for clathrin-mediated endocytosis.  This process has been observed with both the EGF receptor (Stang et al., 2004) and E-cadherin (Fujita et al., 2002; Bryant and Stow, 2004).  However, a catalytically inactive DUB such as echinus would be unable to remove ubiquitin modifications from substrates, and so an alternate mode of function must be hypothesized.  Aside from the lack of active-site residues, the ubiquitin hydrolase domains in all three splice forms are highly similar to each other and to the sequence of known ubiquitin hydrolases.  This would suggest that while the domain may lack enzymatic activity it retains the ability to reversibly bind ubiquitin monomers.  If correct this can suggest several possible roles for the protein.  One possibility is that the ubiquitin hydrolase domain of Echinus facilitates binding to a monoubiquitinated substrate hindering the ability of other proteins to act upon the modified substrate.  For example, modification of a cell surface protein can stimulate the binding of proteins such as specific endocytic adapters, resulting in  48 internalization of the protein.  Binding of an inactive DUB to the modified substrate may hinder the endocytic process and allow the protein to remain at the cell surface.  In this way an inactive DUB could act to regulate the concentration or activity of a cell surface protein. Another possible function of an inactive DUB may be to alter the trafficking of a protein through the cell.  Eukaryotic cells have a complex system of cellular trafficking through a series of internal structures and transport vesicles, directed by a complex arrangement of proteins bound to these structures (Zerial and McBride, 2001; Jordens et al., 2005).  Ubiquitination has been demonstrated to play an important role in the route proteins take through this system (Hicke and Dunn, 2003).  Since the ultimate destination of this trafficking can be as diverse as lysosomal degradation, recycling to the cell surface or return to the Golgi for modification, control of this process can have a profound affect on the abundance and activity of the substrate.  In this way echinus may play a role in regulating the abundance or subcellular location of a specific substrate. In the absence of demonstrated enzymatic activity, and a requirement of that activity for echinus function, the role of this domain is hypothesized to be that of an ubiquitin interacting motif to bind echinus to the monoubiquitinated form of a specific substrate.  The identity of that substrate, and the function of the interaction, remain to be elucidated.   49 Chapter 3 | Mutations in echinus disrupt normal development of the Drosophila retina1  3.1) Introduction Flies carrying mutations in echinus have defective development of the retina, giving a rough and disordered appearance to the adult eye.  A rough eye can result from failure in any one of several morphogenetic processes, such as programmed cell death during pupal development (Hay et al., 1994), cell sorting within the interommatidial lattice (Wolff and Ready, 1991b; Reiter et al., 1996; Carthew, 2007) or even an alteration in the number of photoreceptor neurons (Carthew and Rubin, 1990; Basler and Hafen, 1991).  Previous characterizations of echinus have identified insufficient cell death in the cells of the interommatidial lattice as the cause of the rough eye phenotype in these mutants as additional cells remain and disrupt the normal patterning of the ommatidia (Wolff and Ready, 1991b; Reiter et al., 1996). Developmental apoptosis occurs in the retina following specification of the photoreceptors, cone cells (CC) and primary pigment cells (PPC) of the ommatidium and sorting of the interommatidial cells (IOC) within the surrounding lattice (reviewed in Brachmann and Cagan, 2003).  A subset of IOCs undergoes apoptosis, leaving the exact number required to surround each ommatidium in the mature retina. The cell sorting process appears to be necessary but not sufficient for the initiation of this apoptosis within the cells of   1 A portion of the data in this chapter has been published.  echinus expression analysis in the pupal retina and identification of the cell sorting defect in echinus mutants is contained in: Copeland, J. M., Bosdet, I., Freeman, J. D., Guo, M., Gorski, S. M. and Hay, B. A. (2007). echinus, required for interommatidial cell sorting and cell death in the Drosophila pupal retina, encodes a protein with homology to ubiquitin-specific proteases. BMC Dev Biol 7, 82.  50 the lattice.  Sorting can occur relatively normally in cells unable to undergo apoptosis (Hay et al., 1994) but defective cell sorting can substantially reduce the extent of cell death (Wolff and Ready, 1991b; Reiter et al., 1996; Brachmann and Cagan, 2003). The mechanisms controlling IOC sorting are incompletely understood but recent work has highlighted the importance of cell-cell adhesion (reviewed in Tepass and Harris, 2007) and remodeling of the cytoskeleton (Mege et al., 2006; Cordero et al., 2007; Tepass and Harris, 2007; Johnson et al., 2008).  The cell adhesion molecules Roughest, expressed in the IOCs, and Hibris, expressed in the PPCs, play an essential role in promoting reorganization of the cells within the lattice via heterophilic interactions that encourage IOC-PPC contacts (Reiter et al., 1996; Bao and Cagan, 2005; Carthew, 2007).  A central component of the epithelial adherens junction is the transmembrane protein DE-cadherin, which mediates cell adhesion via homophilic interactions to neighboring cells.  Protein complexes bound to DE-cadherin’s cytoplasmic tail, including the Drosophila β-catenin Armadillo, anchor the adherens junction to the actin cytoskeleton.  During the period of cell sorting, DE-cadherin strongly labels the junctions between IOCs and PPCs and appears reduced at IOC-IOC junctions in a dynamic process regulated by the interaction of Roughest and Hibris (Bao and Cagan, 2005; Grzeschik and Knust, 2005).  The decapentaplegic pathway, along with the actin cytoskeleton regulator rhoA (Magie et al., 2002), has recently been identified as regulators of this interaction between Roughest and DE-cadherin (Cordero et al., 2007).  Additionally, the Drosophila CD2AP/CIN85 homologue cindr has been identified as a functional link between cell adhesion mediated by DE-cadherin and Roughest and components of the actin cytoskeleton (Johnson et al., 2008). Another regulator of the actin cytoskeleton in Drosophila is the Ena/VASP (Kwiatkowski et al., 2003) protein Enabled.  Enabled promotes unbranched actin filament formation at barbed ends by interfering with the function of actin capping proteins (Pasic et al., 2008). Enabled activity and localization are regulated in part by the nonreceptor tyrosine kinase  51 Abelson (Abl) (Van Etten, 1999; Grevengoed et al., 2003) and Enabled has been shown to regulate the density of actin filaments associated with the AJ (Scott et al., 2006). These recent results demonstrate that the cell sorting process depends upon selective and dynamic cellular adhesion and directed cell migration coordinated by a diverse series of cell signaling pathways.  A better understanding of how this process is accomplished and regulated will shed light on the mechanisms required for tissue patterning during development and how defects can contribute to the pathology of diseases such as cancer. This section of the thesis investigates the function of echinus through characterization of defects observed in the developing retina of echinus mutants.  Echinus was found to be required for sorting of interommatidial cells, possibly through its role as a regulator of DE-cadherin accumulation within the adherens junction.   My genetic analyses suggest that echinus is a positive regulator of wingless signaling and that it interacts with the actin cytoskeleton regulator enabled.  52 3.2) Materials and Methods 3.2.1) Drosophila lines and genetics Drosophila strains were maintained at room temperature on standard yeast- extract/cornmeal agar.  Crosses for interaction testing were performed at 25C.  Male flies carrying mutant alleles on the second and third chromosomes were crossed to the hypomorphic echinus allele ecPlacZ (gift - J. Copeland and B. Hay) and male progeny of the F1 generation were observed for changes in the extent of disorder in the adult eye by comparison to F1 males carrying the balancer chromosome.  For interaction testing of X-linked mutations, male GMR-miEcSF1 flies, which carry an echinus-specific RNAi construct, were crossed to females carrying the mutant allele and female progeny in the F1 generation were observed.  All positive interactions were performed in duplicate and at least ten F1 males of each genotype were examined from each cross.  Interactions were assessed by visible changes in the severity of ommatidial row disruptions across the adult eye and scored using a qualitative scale (+++/++/+/NE/-/--/--- representing strong/moderate/weak enhancer, no effect, weak/moderate/strong suppressor).  Images provided are from representative animals.  3.2.2) Pupal retina cell counts Pupal cell counts were performed at 42hrs APF by counting non-bristle cells that surround two 1° pigment cells.  Five separate areas from at least three animals were counted for each genotype and P-values were calculated using an unpaired Student’s t-test.   53 3.2.3) Immunohistochemistry Pupal retinas were dissected at the indicated developmental timepoint in PBS supplemented with 1mM CaCl2 and fixed for 30 minutes in 3% paraformaldehyde. Immunostaining was carried out in PBT (PBS + 0.1% Triton-X100 + 1mM CaCl2) containing 10% fetal calf serum and primary antibody.  Primary antibodies used were mouse anti- Armadillo N2 7A1 (1:50, Developmental Studies Hybridoma Bank (DSHB)), rat anti-DE- cadherin (1:25, DSHB), mouse anti-beta-galactosidase 40-1a (1:50, DSHB), mouse anti-Rst Mab24A5.1 (1:25, (Schneider et al., 1995)), mouse anti-discs-large 4F3 (1:50, DSHB), mouse anti-Enabled 5G2 (1:50, DSHB) and rabbit anti-Echinus (1:100,  gift - J. Copeland and B. Hay).  Secondary antibodies used were anti-mouse or anti-rabbit IgG conjugated to Alexa 488 or 546 (1 ug/mL; Invitrogen) or anti-Rat IgG conjugated to Cy3 (Jackson Laboratories). Fluorescently labeled tissues were mounted in Antifade reagent (SlowFade Light Antifade Kit, Molecular Probes), viewed with a Zeiss Axioplan 2 and images were captured using a Retiga 1350EX digital camera (Qimaging Corp.) and Northern Eclipse software (Empix Imagin, Inc.). For each antibody at each timepoint, at least eight retinas from at least five different animals were examined over at least two experiments and representative images are presented unless otherwise noted.  3.2.4) TUNEL staining Pupal retinas were dissected in PBS and fixed for 30 minutes in 3% paraformaldehyde. TUNEL labeling with fluorescein-dUTP was performed using the In situ Cell Death Detection Kit (Roche Applied Science).  Tissues were incubated at 37C for 1 hour in a 1:9 mixture of enzyme and label solution, and then rinsed in PBS.  Labeled tissues were mounted in Antifade reagent (SlowFade Light Antifade Kit, Molecular Probes), viewed with a Zeiss Axioplan 2 and  54 images were captured using a Retiga 1350EX digital camera (Qimaging Corp.) and Northern Eclipse software (Empix Imagin, Inc.).  For each genotype and timepoint at least five retinas from different animals were examined and representative images are presented.  3.2.5) RNA in situ staining A ~1.2 kilobase region within echinus (nucleotide position corresponding to base pairs 1,532 to 2,691 of splice form 2) was amplified by PCR using T3/T7-tailed primers. Digoxigenin-labeled RNA probes were prepared for both sense and antisense strands (Roche). In situ hybridization to pupal retinas was performed essentially as described in Tautz and Pfeifle (1989).   55 3.3) Results 3.3.1) echinus mutants have excess interommatidial cells The previously reported echinus mutant phenotype is developmental defects in the cells of the retina, which cause a disordered arrangement of ommatidia and result in a rough appearance of the adult eye.  The cause of this phenotype was identified (Wolff and Ready, 1991b) as a failure of developmental apoptosis within the cells of the ommatidial lattice.  To visualize these ectopic cells, pupal retinas were dissected at 42hrs APF and stained for the Drosophila beta-catenin protein Armadillo (Hatzfeld, 1999; Bienz, 2005; Brembeck et al., 2006) to visualize cell boundaries (Figure 3.1).  To determine the average number of additional interommatidial cells in the retinas of selected echinus mutant lines, cell counts were performed on 42hr APF retinas (Table 3.1).  Analysis of 42hr APF retinas shows varying levels of ectopic cells within the interommatidial lattice of all echinus alleles.  Additional cells per ommatidia range from only a few in presumed hypomorphic alleles (ec1, ecPlacZ) to five or more in the amorphic allele ecΔ9.  These additional cells disorder the regular spacing of the ommatidia and result in the rough appearance.  The shape and size of the observed apical profile of the secondary and tertiary cells suggests that the extent of development is generally comparable to wild-type at this timepoint. The GMR promoter directs expression within the cells of the developing retina, starting in the late third instar larval stage and continuing throughout the pupal stage (Ellis et al., 1993; Hay et al., 1994). A line over-expressing echinus (splice form 1) under control of the GMR promoter (GMR-ec) displays only occasional ectopic interommatidial cells, usually in a position of contact with the cone cell cluster (see below). The relative differences in ectopic cell number between echinus alleles are generally consistent with the resulting roughness of the adult eye.   56 Figure 3.1   Figure 3.1 | echinus mutant retina have excess interommatidial cells.  Pupal retina of the indicated genotype dissected at 42hrs APF.  Staining for the adherens junction protein Armadillo (β-catenin) outlines the cells.  Cells of a single ommatidia are outlined in (A): cones (yellow), primary pigment cells (green), secondary pigment cells (blue) and tertiary pigment cells (orange).  Disc is shown with the anterior to the left and dorsal side up.  Individual cone cells are labeled: anterior (a), posterior (p), polar (pl) and equatorial (e).  echinus mutants have additional secondary and tertiary pigments cells within the interommatidial lattice (some indicated by arrows).  Primary pigment and cone cell numbers are largely unaffected by mutations in echinus, although ectopic cone cells are observed at low frequency (* in B).  Some sorting defects are apparent (arrowheads) resulting from a failure of interommatidial cells to align end-to-end in a single row during the cell-sorting period that precedes cell death.  57 Table 3.1 Genotype Average IOC Number OreR 9.0 ec1 10.4 +/- 1.4 * ecPlacZ 11.8 + /- 1.1 * ec3c3 12.7 +/- 1.2 * ecΔ9   14.4 +/- 1.6 * GMR-ec 9.3 +/- 0.2  Table 3.1 | echinus mutants have excess interommatidial cells.  Cell counts from retinas of the indicated genotypes.  (n = 5 areas from at least three animals for each genotype).  echinus alleles result in between one (ec1, a weak loss-of-function allele) and five (ecΔ9, a strong loss-of- function allele) additional lattice cells per ommatidia.  Flies over-expressing echinus under control of the GMR promoter display a low frequency of additional lattice cells.  (Asterisks indicate significant results (P < 0.0001)).  58 3.3.2) echinus mutations show decreased developmental apoptosis in the Drosophila retina The ectopic interommatidial cells observed in echinus mutants may result from either defective cell death or increased cell proliferation.  Previous studies of echinus pupal retinas at this stage have identified a decrease in the number of cells stained with acridine orange, a vital dye that is excluded from living cells (Wolff and Ready, 1991b).  To assess the level of cell death in pupal retinas we used a TUNEL assay that specifically labels the nuclei of cells undergoing apoptosis (Figure 3.2A). TUNEL analysis of 28hr APF retinas shows that mutations in echinus result in a substantially decreased amount of apoptosis in the retina at this developmental time point (Figure 3.2B). This identifies the source of ectopic cells in echinus retinas as the reduction in apoptosis of the interommatidial cells following cell sorting. Staining with an antibody specific for activated caspase-3, an effector of apoptosis, shows a similar reduction in apoptotic cells at this time point (Copeland et al., 2007).  59 Figure 3.2   Figure 3.2 | echinus mutations decrease apoptosis during retinal development.  (A) TUNEL staining labels cells undergoing apoptosis in OreR 28hr APF retinas. (B) The ecPlacZ mutant retinas have substantially-reduced levels of apoptosis at this timepoint compared to wild type.  The entire retina is displayed for each genotype.  60  3.3.3) echinus is expressed in all epithelial cells of the developing retina Mutations in echinus primarily affect the cells of the interommatidial lattice, although defects in cone cell number are observed at a low frequency.   I examined the pattern of echinus expression in the developing retina to determine if echinus expression was restricted to a subset of retinal cells.  The ecPlacZ mutant allele was generated by insertion of a P-element upstream of the transcriptional start site of echinus splice form 2 (Figure 2.7).  This P-element contains the B-galactosidase gene and functions as an enhancer trap to allow detection of expression at this locus.   Retina from ecPlacZ pupae were dissected at 28hrs AFP and stained with anti-beta- galactosidase antibody.  All cells examined expressed beta-galactosidase at this time point, including cone cells, primary pigment cells and interommatidial cells (Figure 3.3).  This result was supported by RNA in situ staining using a probe complementary to all splice forms of echinus.  Approximately 1.2 kilobase digoxygen-labeled antisense and sense (control) RNA probes were generated and hybridized to fixed 28hr APF wild-type retinas.  Using this method echinus transcript was detected in the interommatidial cells and the primary pigment cells, although expression in the cone cells could not be conclusively demonstrated (Figure 3.4). These data show expression of echinus in all epithelial cells of the developing retina at 28hr APF.   61 Figure 3.3   Figure 3.3 | echinus is expressed in epithelial cells of the developing ommatidia.  Anti-B-gal immunostaining of retina from the ecPlacZ enhancer trap line indicates the echinus promoter is active in cone cells (A), 1° pigment cells (B) and interommatidial cells (C) at 28hr APF. Images represent different focal planes of the same location on a single retina.  62 Figure 3.4   Figure 3.4 | RNA in situ staining of wild-type retinas.   RNA in situ staining of 28hr APF wild-type retina with a DIG-labeled RNA probe complementary to all echinus splice forms shows the presence of echinus transcript in primary pigment (A) and interommatidial cells (B). Arrows in (A) and (B) indicate representative examples of primary pigment cell and interommatidial cell staining, respectively.  Expression in cone cells was not conclusively observed.  The RNA sense strand control probe showed no staining (A, inset).   63 3.3.4) echinus mutations cause defective cell sorting During normal development the interommatidial cells sort into an end-to-end arrangement that results in a lattice one cell wide.  It is following this cell sorting process that excess cells within the lattice are removed.  echinus mutations have previously been characterized as disrupting the cell death process but not the cell sorting that precedes it (Wolff and Ready, 1991b; Reiter et al., 1996).  To determine if the interommatidial cells also sorted correctly in the newly characterized echinus alleles, we examined echinus mutant retina for evidence of sorting defects.  My analysis of echinus retinas subsequent to the cell death phase of development shows they contain ectopic interommatidial cells that are in a side-to-side arrangement (arrowheads, Figure 3.1).  This has not been reported previously during investigation of echinus mutants and is most likely due to the strong hypomorphic and amorphic alleles being examined here.  In contrast, in retinas where apoptosis has been blocked by expression of the caspase inhibitor P35, the IOCs retain the ability to sort correctly (Hay et al., 1994), demonstrating that these processes are separable.  These data suggest that echinus plays a role in the earlier process of cell sorting. A molecule known to play an important role in cell sorting is the cell-surface protein Roughest (Reiter et al., 1996).  Mutations in Roughest result in severe defects in cell sorting and an almost complete block of the subsequent apoptosis.  This indicates that correct sorting is a requirement for cell death to occur and suggests the possibility that echinus could primarily function to facilitate cell sorting and the observed apoptosis defects in loss-of-function mutants are a downstream effect of disrupted sorting.  The distribution of Roughest protein was examined in echinus mutants to determine whether this could explain the mechanism whereby echinus effects sorting and apoptosis.  In wild-type retinas during the cell-sorting period of development (approximately 21-25hrs APF) Roughest is found almost exclusively at the  64 contacts between IOCs and PPC (Figure 3.5A) as well as in intracellular vesicles.  Examination of Roughest distribution in echinus loss-of-function alleles shows similar accumulation at IOC- PPC contacts but also a small amount of inappropriate localization of the protein to IOC:IOC contacts (arrows, Figure 3.5B).  This abnormal localization remains present at 26hrs APF, suggesting that it is not caused by a developmental delay in echinus mutants.  In contrast to the loss-of-function phenotype, overexpression of echinus results in accumulation of Roughest that is properly restricted to the IOC-PPC contacts (Figure 3.5C).  However the cell-cell contacts in GMR-ec retinas are abnormal, appearing thicker than wild-type and lacking the smooth, scalloped appearance normal for this developmental timepoint. Roughest staining is also observed in a punctate pattern throughout some IOCs and PPCs.  These structures have been observed previously and are thought to be multivesicular bodies (Vishnu et al., 2006).  However their small size and dispersed nature makes quantification and confirmation of their number and size in the different genetic backgrounds difficult.     65 Figure 3.5  Figure 3.5 | Roughest protein distribution is altered in echinus mutants.  Retina were dissected at 24hrs (A-C) and 26hrs (A’-B’) APF and stained with anti-Roughest antibody to observe its distribution. At 24hr APF, a developmental period at which the interommatidial cells are undergoing sorting, Roughest normally accumulates at the borders between interommatidial cells and the primary pigment cells, as well as within intracellular vesicles (A). In echinus loss-of-function mutants Roughest protein is also observed at IOC-IOC borders (B, arrows).  This inappropriate distribution remains apparent at 26hrs APF (B’) indicating it is likely not due to a developmental delay in ecΔ9.  (C) Overexpression of echinus maintains Roughest localization to the IOC-PPC border.  Note the IOC-PPC contacts in (C) appear thicker and more irregular than in (A) or (B) (arrowheads).  66 3.3.5) echinus is a positive regulator of DE-cadherin accumulation in retinal epithelial cell adherens junctions The observations that cell sorting is disrupted in echinus loss-of-function mutants and that cell junctions in GMR-ec overexpression retinas appeared irregular prompted an examination of the integrity of the adherens junctions at cell-cell contacts. I investigated whether echinus affects components of the adherens junction by examining the localization of DE-cadherin.  At 24hr APF, a developmental stage at which lattice cells are sorting, DE- cadherin is normally distributed to the IOC-PPC cell contacts (Figure 3.6A).  Reduced labeling is also observed at the IOC-IOC border.  DE-cadherin is also largely absent from the contacts between cone cells where DN-cadherin is expressed to promote tight cone cell clustering (Hayashi et al., 2004).  In echinus mutants there is a reduced level of DE-cadherin in the adherens junctions of the interommatidial cells and the junctions appear faint and discontinuous (arrows, Figure 3.6B).  Overexpression of echinus using a GMR-ec construct resulted in strong localization of DE-cadherin to the adherens junctions (Figure 3.6C).  Similar to the Roughest staining results, DE-cadherin staining reveals adherens junctions to be slightly irregular in thickness as well as abnormally straight and elongated compared to the wild type. These data demonstrate that echinus regulates the amount of DE-cadherin associated with the adherens junctions at this developmental stage.  67 Figure 3.6   Figure 3.6 | echinus promotes DE-cadherin accumulation in the adherens junction of retinal cells.  Retinas were dissected at 24hr APF and stained with rat anti-DE-cadherin to visualize adherens junctions.  (A) Wild-type retina display continuous, well-defined adherens junctions between IOCs and primary pigment cells.  IOC:IOC contacts are visible but less distinct.  (B) echinus loss-of-function mutants have discontinuous adherens junctions – arrows indicate examples of contacts that are indistinct or devoid of DE-cadherin immunostaining. Some DE-cadherin appears to be mislocalized in IOCs – either to the apical surface or distributed diffusely in the cytoplasm (compare A and B).  (C) echinus over-expression results in adherens junctions that stain strongly for DE-cadherin.  Cell profiles are elongated and junctions are abnormally straight, suggesting increased IOC-POC adhesion compared to (A). DE-cadherin staining in the bristle cells can be used to make approximate comparisons in staining intensity between retinas. (A’-C’) Enlargement of boxed regions in (A-C) highlights discontinuous DE-cadherin staining in echinus loss-of-function mutants (B’, arrows) and increased diffuse staining in interommatidial cells (*) compared to primary pigment cells (**).   68 3.3.6) Septate junctions are intact in echinus mutants The septate junction is another epithelial cell-cell contact structure that is located basal to the adherens junction and functions as a physical barrier to prevent diffusion of solutes between cells and to provide structural strength to the cell (Banerjee et al., 2006).  To determine if changes in the adherens junctions of echinus mutants were reflected in the septate junction, 24hr APF retinas were stained with an antibody to Discs-large (dlg), a component of the septate junction.  As expected, in wild-type discs-large was found in a continuous band around the cell perimeter (Figure 3.7A).  This distribution appeared unchanged in both echinus loss-of-function (Figure 3.7B) and overexpression lines (Figure 3.7C), suggesting that the septate junction is unaffected by mutations in echinus.  An analysis of additional echinus alleles at a later developmental timepoint reached a similar conclusion (Copeland, 2007)  69 Figure 3.7  Figure 3.7 | Septate junctions appear intact in echinus mutants.   Retinas (n=3 for each genotype) were dissected at 24hr APF and stained with anti-discs-large to visualize the septate junctions.  Unlike the adherens junctions, no obvious differences were observed between (A) wild type and (B) echinus mutants or (C) overexpression lines.  70 3.3.7) echinus is a positive regulator of Enabled accumulation in the adherens junctions of retinal epithelial cells Enabled functions to regulate actin dynamics by opposing the action of actin capping proteins and permitting elongation of actin fibres (Renfranz and Beckerle, 2002; Grevengoed et al., 2003).  The stability of the adherens junction is affected in part by the strength of its association with the actin cytoskeleton, and disruption of this association is one mechanism for removal of DE-cadherin from the adherens junction and downregulation of cell-cell adhesion. To determine if Enabled function is altered in echinus mutants, we examined its distribution by staining 24hr APF retina with an anti-Enabled antibody.  In wild type retina at this time point, the distribution and overall concentration of Enabled is largely similar to that of DE-cadherin and beta-catenin; staining is strong at IOC-PPC borders and weaker at the contacts between IOCs (Figure 3.8A).  Some differences are apparent, however, including a similar level of staining in the contacts between cone cells and strong puncta of labeling at points of contact between three cells (inset, Figure 3.8A).  In echinus loss-of function alleles, Enabled appears to maintain a similar distribution pattern to wild type but accumulates to dramatically reduced levels (Figure 3.8B).  The reduction is more apparent in contacts of the IOCs than in those of the cone cells.  Conversely, overexpression of echinus results in robust accumulation of Enabled at cell junctions, particularly those between IOCs and PPC (Figure 3.8C).  All retinas (n = 8 for each genotype) were processed in parallel and replicated in duplicate experiments.  Images were collected using similar exposures and representative examples are shown.  The patterning of Enabled distribution is similar to that of other adherens junction components in both wild type and echinus mutants, suggesting that enabled is required for correct assembly of adherens junctions during this period of retinal development.   71 Figure 3.8   Figure 3.8 | echinus regulates association of the actin stimulating protein Enabled with adherens junctions in the developing retina.  Retinas at 24hrs APF were dissected and stained with anti-Enabled to visualize its distribution.  (A) In wild-type tissue Enabled accumulates preferentially at all cell-cell contacts although reduced labeling is observed at IOC:IOC junctions.  Intense labeling is observed at points of contact between three cells, most notably the 8 three-cell contacts in the cone cluster (A, inset).  (B) echinus loss-of-function tissue has reduced and discontinuous labeling of Enabled.  (C) echinus over-expression results in a relative increase in Enabled staining within all cell junctions, most notably those between IOCs and PPCs.  Contacts also appear thickened and more irregular than in wild-type tissue.  Arrow indicates an ommatidia with five cone cells where all tricellular junctions display an increased point of Enabled labeling.  72 3.3.8) Over-expression of echinus in the retina results in ommatidial rotation defects and alterations in cone cell morphology An examination of retinas expressing echinus under the control of the GMR promoter revealed several defects in patterning.  The 90-degree rotation of ommatidia during development is a process tightly controlled by multiple signaling pathways.  Subtle rotation defects in echinus loss-of-function mutants have been described previously (Montrasio et al., 2007), but GMR-ec retina show substantial variation in the extent of rotation, with some ommatidia even displaying a perpendicular orientation to adjacent ommatidia (asterisk, Figure 3.9A).  In addition, the arrangement of cone cells is altered in many ommatidia.  At this point in the development of wild-type and echinus mutant retinas, the two 1° pigment cells completely encircle the four cone cells, which cluster together with the polar and equatorial cones contacting each other in the middle of the cluster (Figure 3.9B).  However, in GMR-ec retinas many ommatidia contain an interommatidial cell that maintains an inappropriate contact with one of the cone cells (arrows, Figure 3.9A).   Many cone cell clusters also have an abnormal arrangement whereby the anterior and posterior cone cells are in direct contact with each other and separate the polar and equatorial cones (arrowheads, Figure 3.9A).  These alterations in ommatidial patterning show that ectopic expression of echinus using the GMR promoter can disrupt the normal patterning process of the developing retina. The patterning of GMR-ec retinas was also observed to be abnormal with respect to the overall arrangement of ommatidia across the retina.  The wild-type arrangement, also observed in echinus mutants, is that of hexagonal packing (inset, Figure 3.9B) whereas in GMR-ec retinas the ommatidia are in a pattern that more closely resembles square packing (inset, Figure 3.9A).  This altered arrangement remains present later in development (not shown) suggesting that it is not the result of a developmental delay in these retinas.  73 Figure 3.9   Figure 3.9 | Over-expression of echinus causes defects in patterning, ommatidial rotation and cone cell morphology.  Retinas were dissected at 24hrs APF and stained with anti-DE- cadherin to outline cell boundaries.  Ommatidial rotation is a tightly regulated process that ensures all ommatidia have an identical orientation. (A) Retinas over-expressing echinus have variable rotation of ommatidia (asterisks, A – ommatidia have perpendicular orientations). Defects in cone cell arrangement and contacts are also observed.  In many ommatidia, primary pigment cells incompletely encircle the cone cell cluster, resulting in cone cells that maintain inappropriate contact with IOCs (arrows).  Abnormal cone cell arrangements (arrowheads) where the anterior and posterior cones make direct contact with each other are also present. The arrangement of ommatidia in GMR-ec retinas is disorganized and suggestive of square packing instead of hexagonal (insets in A and B; dotted lines represent the orientation of ommatidia in displayed retina).  These phenotypes show variable penetrance, both between animals and within an individual retina, and an extreme example is shown. (B) echinus loss-of- function retinas display normal cone cell arrangements.  Ectopic cone cells and minor defects in ommatidial rotation occur at low frequencies (not shown).   74 3.3.9) Echinus appears within IOCs and in association with the IOC/PPC junction at an early stage of pupal retinal development A polyclonal antibody directed against the ubiquitin hydrolase domain of Echinus was used to examine the distribution pattern of this protein in the developing retina during the period of cell sorting.  At 21hrs APF, this antibody shows diffuse staining throughout the interommatidial cell lattice in addition to small puncta of intense staining (Figure 3.10A).  At 24hrs APF, the distribution is similar although larger punctations appear reduced in number and staining can be observed at increased concentration adjacent to the IOC/PPC border (Figure 3.10B, arrows).  At both time points Echinus seems either weakly expressed or entirely absent in cones cells.  These data suggest that Echinus is expressed in the retina during this period of retinal development.  Based on the pattern of staining Echinus appeared to be most concentrated in the interommatidial cells.  To confirm this hypothesis a co-stain to reveal cell contacts was attempted but resulted in poor anti-Echinus immunostaining quality.   75 Figure 3.10   Figure 3.10 | Echinus protein is found in interommatidial cells.  OreR retinas at (A) 21hrs and (B) 24hrs APF were dissected and stained with anti-Echinus antibody to visualize its distribution in the developing retina.  At 21hrs APF Echinus protein is distributed diffusely within interommatidial cells as well as in small intracellular punctations.  At 24hrs APF the distribution is similar although increased concentration can be observed at the presumptive IOC-PPC borders (arrows).  (C) Staining is absent in control ecΔ9 retinas at 24hrs APF.  76 3.3.10) echinus mutants have abnormal development of some mechanosensory bristles In echinus loss-of-function mutants a subset of mechanosensory bristles (macrochaetes) on the notum are either duplicated or missing (Figure 3.11).  To better define this phenotype, bristle numbers in the dorsocentral and scutellar positions were scored in adults carrying different echinus alleles (Table 3.2).  Wild-type flies were absolutely invariant in bristle number at these locations, reflecting the highly regulated development of these sensory organs.  In contrast, echinus mutant showed significant changes in bristle numbers.  Between 14% (ecPlacZ) and 82% (ec1) of flies showed altered bristle number at one or more locations.  The most common change was loss of the posterior dorsocentral bristle.  The number and position of macrochaetes other than the dorsocentral and scutellar bristles appear unaffected by mutations in echinus.   77 Figure 3.11   Figure 3.11 | The dorsocentral and scutellar bristles develop abnormally in echinus mutants. (A) Wild-type adult notum showing the stereotypical location of eight dorsocentral and scutellar bristles: anterior dorsocentral (aDC), posterior dorsocentral (pDC), anterior scutellar (aSc) and posterior scutellar (pSc) on each of the right and left sides.  (B) echinus mutants show abnormal numbers of notum bristles with both the addition of ectopic bristles (arrows; two or more bristles where one is expected) and the loss of bristles (arrowhead; absence of bristle at a typical location).  The number and pattern of microchaete hairs appears unaffected by echinus mutations.  78 Table 3.2  % wild type aDC pDC aSc pSc   + - + - + - + - ec1 18 88 4 5 19 1 0 0 0 ecdelta9   56 7 8 0 34 3 0 0 1 ec3c3 59 2 2 0 13 39 2 2 0 ecPlacZ 86 1 1 0 7 7 1 0 0 OreR 100 0 0 0 0 0 0 0 0   Table 3.2 | Dorsocentral and scutellar bristle numbers are altered in echinus mutants.  Adult flies were scored for bristle number at eight locations: right/left anterior dorsocentral (aDC), right/left posterior dorsocentral (pDC), right/left anterior scutellar (aSc) and right/left posterior scutellar (pSc) (see Figure 3.16).  The proportion of flies with wild type bristles (one at each location) is indicated (% wild type). For each location the number of ectopic (+) or missing (-) bristles was recorded.  n=100 animals for each genotype.   79 3.3.11) echinus interacts genetically with enabled, wingless and components of the fat tumor suppressor pathway To better understand the functional role of echinus in retinal development a number of mutant lines were crossed to echinus mutants in an attempt to find enhancers or suppressors of the rough eye phenotype (Figure 3.12).   Most mutants, in genes known to effect programmed cell death, retinal patterning and bristle morphogenesis, failed to modify the echinus loss-of-function phenotype in a detectable manner (Table 3.3).  A small number, however, showed modest but reproducible modification of the adult rough eye phenotype.  A number of components of the fat tumor suppressor pathway were tested and one, expanded (ex), showed an obvious suppression of the echinus rough eye using two different alleles (Fig. 3.13D- E’).  Alleles of the genes fat (ft) and dachsous (ds) also appeared to show a suppression of the rough eye (Fig. 3.13C, F-F’) but the effect is too subtle to conclusively demonstrate with this assay.  Four-jointed (fj), a negative regulator of the fat pathway, showed a weak enhancement of echinus (Fig. 3.13I-I’).  Stronger enhancement was observed in crosses of echinus to mutant alleles of enabled (ena) (Fig. 3.13G-H’) and wingless (wg) (J-K’).  These mutations caused an obvious increase in the disorder of the adult retina.      80 Figure 3.12  Figure 3.12 | Schematic representation of genetic crosses employed for interaction testing. Examples of (A) first and (B) second chromosome genetic crosses are shown.  (A) For crosses to mutant alleles on the first chromosome (Table 3.3) males carrying an RNAi construct (GMR-miEcSF1) were crossed to virgin females carrying a balanced copy of the mutation.  In the F1 generation, the retinas of females carrying the mutant allele were compared to those carrying the balancer. (B) For crosses to mutant alleles on the second and third chromosomes, virgin ecPlacZ females were crossed to males carrying balanced copies of the mutation.  In the F1 generation, the retinas of males carrying the mutant allele were compared to those carrying the balancer. (* = mutant allele; bal = balancer chromosome)     81 Table 3.3 X 2 or 3 β-catenin (arm3, arm4) casein kinase (dco3) clathrin (chc4) delta (Dl6B) notch (Nfa-g) echinoid ed1 dynamin (shi1) EGF receptor (EGFRE1, EGFRF2)  extra machrochaetes (emcD)  fat (ft1)  hairless (H1)  grim (GMR-Grim)  hid (GMR-Hid)  reaper (GMR-Rpr)  nemo (nemoj147-1)  neuralized (neur11)  pannier (pnrMD237)  rack1 (Rack1Ey00128)  DE-cadherin (shg2, shgk03401)  flamingo (stanfrz3)  Table 3.3 | Mutant alleles that failed to modify the echinus rough-eye phenotype.  Flies carrying mutant alleles were crossed to echinus mutants and F1 progeny were examined for changes in the extent of disorder in the adult retina.  The listed alleles failed to cause detectible modification of the rough eye phenotype in echinus mutants.          82  Figure 3.13   83 Figure 3.13 | echinus interacts with enabled, wingless and components of the fat tumor suppressor pathway.  Virgin females carrying the hypomorphic echinus loss-of-function allele ecPlacZ (B) were crossed to males with mutations in the following genes:  fat (C), expanded (ex, D-E’), dachsous (ds, F, F’), enabled (ena, G- H’), four-jointed (fj, I, I’), and wingless (wg, H-H’, J-K’). Interactions were assessed by changes in the frequency and severity of ommatidial row disruptions across the adult eye in males of the F1 generation.  The rough-eye phenotype is suppressed weakly by mutations in expanded, and possibly fat and dachsous, and is enhanced weakly by mutant alleles of four-jointed, enabled and wingless.  84 3.4) Discussion A previous study of Drosophila eye development identified excess interommatidial cells as the cause of rough eyes in adults carrying mutations in echinus, and used acridine orange staining to show a decrease in cell death as the probable source of these cells (Wolff and Ready, 1991b).  In this study the use of TUNEL staining demonstrated a decrease in cells displaying DNA fragmentation in echinus mutants, indicating a reduction in the number of cells undergoing apoptosis (Figs. 3.2).  Immunostaining for activated caspase-3 also pointed to a similar result (Copeland et al., 2007).   These data suggest that echinus plays a direct role in caspase-dependant death in the eye; however extensive genetic interaction screens have failed to support this hypothesis. My analysis of multiple echinus alleles, some of which are amorphs, indicated that in addition to a decrease in apoptosis, mutations in echinus also cause defects in the cell sorting process that precedes it (Fig. 3.1).  This cell sorting process, in which interommatidial cells rearrange their positions so that they occupy the space between ommatidia in a one cell wide end-to-end orientation, is separable from, and necessary for, the period of apoptosis that follows.  Analysis of mutant retinas expressing the caspase inhibitor P35 indicates that interommatidial cells retain the ability to sort correctly in the absence of apoptosis. Additionally, mutations in genes associated with cell sorting, such as roughest, block the round of apoptosis that normally follows this process (Wolff and Ready, 1991b; Reiter et al., 1996). Considering these data, it appears likely that the defect in apoptosis observed in echinus mutants is related at least in part to defects in the upstream process of cell sorting.  This hypothesis partially contradicts the previously characterized nature of echinus, which is that of defects in cell death but not cell sorting (Reiter et al., 1996).  Efforts to outcross the common echinus allele ec1 may explain this contradiction, since the severity of the phenotype associated with this allele increased considerably once it was moved to a new genetic background  85 (Copeland et al., 2007).  This result points to the accumulation of one or more modifiers of echinus in the original ec1 stock that may have partially suppressed the cell-sorting defect. An examination of the expression pattern of echinus within the developing retina showed expression in all interommatidial cells as well as primary pigment cells and cone cells (Figures 3.3, 3.4).  Although defects are generally only observed in the interommatidial cells this result is not entirely unexpected.  Early work in retinal development indicated that signaling from the primary and cone cells is required for correct patterning of the interommatidial lattice (Miller and Cagan, 1998).  While some cone defects can be observed in echinus mutants, in general the cone and primary pigment cells are normal in appearance. This suggests that echinus is not required for their correct development, perhaps because it is required for a process that at this developmental stage only occurs in the lattice cells. Alternatively, other genes expressed within the cone and primary pigment cells may have functions redundant to that of echinus.  However, the cell autonomy of echinus function remains unclear.  Restricting expression of an echinus-targeting RNAi construct to the lattice cells is sufficient to generate ectopic interommatidial cells, and yet expression of echinus with the same driver in a null-mutant background is unable to rescue the phenotype (Copeland et al., 2007). This indicates that expression of echinus within the cone and pigment cells may be necessary for correct lattice formation but it is not sufficient for this process to occur; echinus expression in the lattice is also required for correct development of those cells.  Such an arrangement can possibly be explained by hypothesizing that echinus is required not only for the successful transmission of a signal from the cone or primary pigment cells but also for the detection and reaction to that signal by the interommatidial cells. Defects in the cell sorting process in echinus mutants prompted an analysis of the distribution of Roughest in these flies.  Roughest mediates cell sorting in the interommatidial lattice via heterophilic interactions with the PPC-specific protein Hibris, and Roughest is  86 normally found almost exclusively at cell contacts between IOCs and PPCs.   Analysis of echinus mutants identified a small amount of Roughest at the contacts between IOCs (Fig 3.6). Whether this redistribution is sufficient to cause the sorting defect in echinus mutants is unknown.  The majority of Roughest remains localized to IOC-PPC contacts, however, so it seems more likely that this mis-localization might prolong or complicate the sorting process without significantly disrupting it.  The cause of this change in Roughest distribution is also unclear.  A block in roughest expression can profoundly disrupt cell sorting but Roughest protein levels appear relatively normal in echinus mutants.  Redistribution of Roughest to IOC:IOC contacts has previously been observed as a result of constitutive overexpression of the adherens junction protein DE-cadherin, alterations in DE-cadherin localization or disruption of Notch function at an earlier stage of retinal development (Grzeschik and Knust, 2005).  Ectopic expression of hibris within the IOCs is also capable of causing a similar redistribution of Roughest (Bao and Cagan, 2005).  A recent study revealed that disruption of the protein Cindr, which functionally links DE-cadherin and Roughest-containing junctions to the actin cytoskeleton, can also localize Roughest to IOC-IOC contacts (Johnson et al., 2008).  These examples of disruption in Roughest distribution are all associated with changes to the cell adherens junctions, which suggests a possible mechanism for the same changes observed in echinus mutants. An examination of the adherens junctions of retinal cells in echinus mutants during the period of cell sorting showed a noticeable decrease in the total content of DE-cadherin at these junctions and also the contiguity of the adherens junction (Fig. 3.6).  Junctions in echinus mutants appear faint and in areas discontinuous, the polarized distribution of DE-cadherin is less apparent with an essentially equivalent concentration on all sides of the IOCs and some appears to be distributed either to the apical cell surface or diffusely within the cell. Conversely, ectopic expression of echinus resulted in an apparent increase in the concentration  87 of DE-cadherin within the adherens junctions.  These results suggest that echinus can regulate the concentration of DE-cadherin within the adherens junction.  Since precise control over adherens junctions is required for the cell sorting process (Bao and Cagan, 2005; Grzeschik and Knust, 2005; Cordero et al., 2007; Johnson et al., 2008), these defective cell junctions may be responsible for the cell sorting defects observed in echinus mutants and may also explain the observed changes in Roughest localization. Another cell contact important for epithelial cell development and function is the septate junction, which is located basally to the adherens junction. Although not directly implicated in the cell sorting process, the septate junction is a vital structure for epithelial integrity and cell polarity.  Changes to the septate junction in echinus mutants would suggest a broad role for echinus in regulating cell-cell contacts.  However, alteration of the expression level of echinus had no observable effect on the distribution of the septate junction protein Discs large (Fig. 3.7).  This result suggests that the septate junctions are without gross defect in echinus mutants and point toward a more specific role of regulating adherens junction stability. A regulator of actin filament formation known to associate with the adherens junction is the Ena/VASP protein Enabled.  Analysis of Enabled distribution in the retina shows a strong association with the adherens junction and at presumptive tricellular contacts (Figure 3.8), reflecting similar observations in the developing embryo (Stevens et al., 2008).  As seen with other AJ components, the band of Enabled staining around each interommatidial cell in echinus mutants was reduced and discontinuous, and the concentration at tricellular junctions also appears similarly reduced or absent.  Similar to results for DE-cadherin, Enabled staining in the contacts between cone cells is less affected by the loss of echinus than the IOC-IOC contacts, perhaps suggesting a relative decrease in the importance of echinus for correct AJ formation in these cells.  GMR-ec retinas displayed a dramatic increase in Enabled staining in AJs, most notably at IOC-PPC contacts which appear thickened and rougher compared to wild-type.  88 These data reinforce previous observations that echinus functions to increase the accumulation of adherens junction proteins at cell-cell contacts, possibly resulting in increased adhesive strength between cells. Close examination of GMR-ec retinas reveals defects in the rotation of some ommatidia (Figure 3.9).  During normal development of the retina the cluster of photoreceptor cells at each ommatidium undergo a highly coordinated 90-degree rotation.  Multiple signaling pathways combine to regulate this process, including EGFR, Notch and the Frizzled (planar cell polarity) pathways (Strutt and Strutt, 2003), and the MAP kinase nemo is also known to contribute at later stages (Fiehler and Wolff, 2008).  The relative expression level of DE- and DN-cadherin also contributes to the regulation of this rotation (Mirkovic and Mlodzik, 2006), suggesting the possibility that changes in the DE-cadherin content in GMR-ec retinas are the cause of this phenotype.   Thus, numerous mechanisms can be hypothesized to cause the rotation defect but it is notable that a similar phenotype can be observed in retinas that overexpress Crumbs (Grzeschik and Knust, 2005) and in certain mutants in the fat tumor suppressor pathway (Silva et al., 2006).  Defects in ommatidial rotation have not been previously observed in roughest mutant retinas or in modifier screens of roughest (Tannenbaum, 2000), suggesting that this phenotype does not directly result from defects in IOC sorting or abnormal distribution of Roughest protein. Interestingly, in this study similar defects in ommatidial rotation were not observed in echinus loss-of-function mutants despite the reduction in DE-cadherin observed in those mutants.  This is possibly due to the nature of the GMR promoter, which promotes high-level and consistent expression of the echinus transgene beginning in the third-instar larval stage and throughout pupal development in both the photoreceptors and epithelial cells of the retina. Thus the loss of echinus may affect the level of AJ-associated DE-cadherin minimally or only at a development stage subsequent to that of ommatidial rotation, whilst the expressed transgene  89 may alter adherens junctions at this rotation stage and throughout eye development.  A recent study of ommatidial rotation in the eye did observe defects in this process in echinus loss-of- function mutants (Montrasio et al., 2007).  The likely cause of this discrepancy is that the authors of that study performed their analysis in the larval retina by directly observing the orientation of the photoreceptor cluster.  Using this method they were able to observe and quantitate subtle changes in rotation that are less apparent when viewing the accessory cells of the retina. The cones cells of each ommatidia adopt a stereotypical arrangement that is governed by the relative strength of cell-cell contacts mediated by DE- and DN-cadherin (Hayashi, 2004).  GMR-ec retinas display two abnormal cone cell phenotypes (Figure 3.9).  Firstly, some cone clusters show an abnormal arrangement where the anterior and posterior cone cells contact each other.  This differs from the wild type morphology where the polar and equatorial cells are in contact.  Similar defects in arrangement have been observed in mutant cone cells expressing high levels of DE-cadherin (Hayashi, 2004).  The interface between the cone cells and the primary pigment cells is shaped by the relative adhesiveness of the interfaces, as mediated by DE-cadherin and DN-cadherin.  The shape of cone-PPC interfaces in GMR-ec retinas appears normal suggesting that DN-cadherin expression pattern is unaffected. My preliminary examination of echinus loss-of-function and echinus overexpression retinas indicates that echinus has no obvious effect on DN-cadherin distribution at 24hrs APF.  The second observed cone cell defect is that polar and equatorial cones sometimes contact a cell within the IOC lattice.  A possible cause for this may be the increased concentration of DE- cadherin within the AJs of all retinal cells, resulting in an increased adherence between cones and the lattice cells that inhibits the ability of the primary pigment cells to intercalate and contact each other.  Similar examples of cone-contact cells have been observed previously in a screen for modifiers of roughest (Tannenbaum, 2000).  Mutant retinas expressing P35 (Hay,  90 1998) display this phenotype as well, suggesting that it can also occur from a block in apoptosis during development. The frequency of these patterning defects in GMR-ec retinas varies, both between animals and within an individual retina.  The cause of this variation awaits further investigation.  Inter-animal variation may be due to changes in the activity of the GMR promoter used to drive echinus expression, or to changes in the activity of other genes required for ectopic echinus to affect change on retinal development.  Intra-retina variation may be associated with the location of the affected ommatidia within the retina.  Defects often appear more frequent and severe at the perimeter of the retina than in its centre.  This may hint at the mechanism by which echinus is causing these effects, since certain signaling pathways such as wingless have gradients of activity across the retina during development. Distribution of Echinus was examined in the retina at 21hrs and 24hrs APF (Figure 3.10).  At 21hrs APF Echinus appears diffusely throughout the cell or plasma membrane and to large punctations, while at the later timepoint smaller puncta are more common and localization near the IOC-PPC border is more prevalent.  Echinus may also be present in cone cells and primary pigment cells in reduced levels.  This pattern identifies the lattice cells as the likely location of echinus function during this period of development, supporting previous cell autonomy experiments (Copeland et al., 2007), and the possible concentration of Echinus at IOC-PPC cell contacts is consistent with the changes in those structures observed in echinus mutants.  However the function and other constituents of these punctate structures remain to be elucidated. In addition to defects in the eye, echinus mutants have been shown to have aberrant development of a subset of mechanosensory bristles on the notum (Figure 3.11 and Table 3.2).  These defects can arise from a failure of asymmetric cell division within the neural precursor cells or from failure of these precursors to be properly defined by the prepattern  91 genes (Tomoyasu et al., 1998; Calleja et al., 2002).  While the number and location of these bristles is invariant in wild-type flies, echinus mutants displayed a pattern of ectopic and absent dorsocentral and scutellar bristles.  Interestingly, while the frequency and pattern of bristle gain and loss was generally consistent between animals of a similar genotype there existed obvious differences between different echinus alleles.   The phenotype was highly penetrant in ec1 flies, moderately so in ecΔ9 and ec3c3, and only occasionally observed in ecPlacZ.  ec1 flies most commonly carried ectopic anterior dorsocentrals, while in ecΔ9 loss of the posterior dorsocentral was the primary defect and ec3c3 flies often developed ectopic anterior scutellars. These differences do not appear to correlate with the severity of the mutation in each echinus allele, although the ec1 stock is known to carry one or more modifiers of the echinus phenotype in its genetic background (Copeland et al., 2007).  The difference in phenotypes between the amorph ecΔ9 and ec3c3, which produces a truncated protein, is particularly intriguing.  One interpretation of these results is that in echinus mutants the cells of the developing notum are impaired in their ability to generate or respond to gradients in wingless or dpp signaling. Alternatively, echinus mutant cells may have defects in their ability to undergo the complex divisions necessary to create a mature bristle.  It is also important to note that this is a non- retinal phenotype and indicates that echinus is required for normal development in more than a single tissue. A genetic screen for modifiers of a mutant phenotype is a powerful tool for associating genes with cellular processes (Simon et al., 1991).  A search for enhancers and suppressors of the echinus rough eye phenotype was undertaken in an attempt to better-define its functional role in the cell.  This screen identified three genes that show consistent, although mild, interaction with echinus (Figure 3.13).  One, expanded, is a suppressor and two, wingless and enabled, are enhancers.  92 Mutants of expanded were able to suppress the echinus eye to an almost wild-type appearance with only minor deviations in the rows of ommatidia.  Expanded is a positive regulator of the fat tumor suppressor pathway (Cho et al., 2006; Jaiswal et al., 2006; Silva et al., 2006; Willecke et al., 2006) and can regulate cell cycle and differentiation in the eye (Tyler and Baker, 2007).  It also has been implicated in the turnover of cell surface receptors and, interestingly, is a negative regulator of wingless in the eye (Pellock et al., 2007).  Additional components of the fat pathway were tested with inconclusive results.  Fat and dachsous are both positive regulators of this pathway and appeared to have a subtle suppressive effect on the echinus rough eye.  The golgi-resident protein four-jointed, a negative regulator of fat, acted as an enhancer of echinus.   Confidence in this result, however, awaits confirmation with a second allele of four-jointed.  The role of the fat pathway as a negative regulator of wingless signaling in the eye is consistent with the enhancement of echinus by two alleles of wingless.  These results support a hypothesis that echinus functions as a positive regulator of wingless signaling.  The role of wingless in regulating expression and accumulation of DE-cadherin (Jaiswal et al., 2006; Wodarz et al., 2006) suggests a possible link to changes observed in the developing retina. The other gene found to interact with echinus is enabled.  Enabled stimulates actin fibre formation and its accumulation in the adherens junction of retinal cells was found to be sensitive to the dosage of echinus.  Interestingly, enabled has previously been shown to interact genetically with four-jointed (Buckles et al., 2001).  The sorting defect in echinus mutants has above been hypothesized to result from weakened adherens junctions, and so it is tempting to speculate that a reduction in enabled function further weakens these junctions and enhances the phenotype.  Additionally, as one of six proteins in the Drosophila genome that encodes an EVH1 domain, there exists the possibility this interaction is mediated through Enabled binding directly to the proline-rich motif present in the C-terminus of Echinus.  93 The interaction with Enabled suggests a link between echinus function and the actin cytoskeleton.  This link is particularly interesting in light of a recent publication describing the function of the Drosophila CD2AP/CIN85 homologue cindr (Johnson et al., 2008).  Previous work has associated CIN85 with endocytosis of cell surface receptors (Dikic, 2003) through association with the ubiquitin ligase Cbl.  Cindr was demonstrated to be required for normal localization of DE-cadherin and Roughest within the adherens junction as well as regulation of the cytoskeleton through interaction with actin capping proteins, whose activity Enabled is known to oppose.    The common themes observed in the function of cindr and echinus encourage a future investigation of possible interactions between these two genes. Despite the positive results, all observed genetic interactions had modest effects on the disorder of the echinus adult eye and a number of possibilities can be suggested to explain this. The simplest explanation is that echinus has a strong interaction with a gene that has not yet been tested, and continued screening could be employed to test this hypothesis.  Additionally, while the mutant alleles tested were generally strong hypomorphs or amorphs, the observed F1 progeny are heterozygous for these loci.  As such the interaction screen tested for the sensitivity of the echinus phenotype to changes in the dosage of these genes.  This decreased dosage may have been insufficient to produce visible changes in the disorder of the adult eye. Another consideration is the nature of the ecPlacZ allele used for interaction testing.  Cell counts in the pupal eye suggest that it is a hypomorphic allele and thus it was considered the best candidate for detection of both suppressors and enhancers.  It may, however, result from more complex changes such as disruption of certain splice forms of the echinus transcript. Alternatively, while disorder in the adult eye has been shown to be a very sensitive detector of changes in cell number it may be insensitive to relative changes in cell sorting once wild-type development is disrupted.  Finally, the small magnitude of the observed changes may result from the existence of genes with redundant function to that of echinus, rendering it dispensable  94 for the majority of cellular processes.  Indeed, a combination of these last two possibilities seems the most likely.  Searches for genes interacting with echinus have been extensive. A previous study of echinus function (J. Copeland, personal communication) tested over 75 mutant alleles for interaction with echinus gain- and loss-of-function alleles without a significant result.  A recent examination of ommatidial rotation defects associated with a novel echinus allele tested 36 mutant alleles, the majority of which were negative (Montrasio et al., 2007).  These results, combined with the negative results in this study, cover a wide range of known cell death, signaling and patterning genes in Drosophila, suggesting that echinus has a limited ability to interact with most loci.  Additionally, the analysis of the rotation defect in the larval eye identified interactions with the genes nemo and shotgun and components of the EGFR signaling pathway, all of which appeared negative when examining the adult eye.  These data argue that future examinations of interactions may best be performed in the larval or pupal retina, where small changes in rotation, cell number or patterning can be better assessed.  This approach could also be employed to further validate the genetic interactions already identified here. These results identify a requirement for echinus in normal adherens junction formation in the epithelial cells of the developing retina and lead to the hypothesis that these defective junctions are responsible for the inability of interommatidial cells to correctly sort.  This block in cell sorting is then suggested to result in the decrease in apoptosis in these interommatidial cells.  However, the precise mechanism by which echinus effects adherens junctions and causes these defects in retinal development remains to be characterized.  Any proposed mechanism must consider the relatively subtle nature of the phenotypes observed in flies homozygous for echinus amorphic alleles.  The relatively mild phenotype of echinus mutants argues against an essential role for this gene in the signaling pathways known to be required for initial specification and formation of the eye.  One possibility is that genes exist with largely  95 redundant functions to that of echinus.  Another possibility, although perhaps unlikely, is that echinus mutants have profound effects on the development or function of tissues that have yet to be observed or characterized in echinus mutants.  Finally, echinus could function to refine the signaling of one or more pathways in a subtle or transient way, effecting the final precise steps in tissue patterning.    96 Chapter 4 | Echinus is a vesicle-associated protein that interacts with Clathrin and Enabled  4.1) Introduction Identification of echinus and characterization of its role in retinal development led to questions regarding its function within the cell.  Analysis of the protein structure was unable to provide much insight, particularly in light of the apparent lack of enzymatic activity of the hydrolase domain.  Similarly, the apparent regulation of adherens junctions revealed by analysis of echinus alleles was informative for understanding the cause of developmental defects in the retina, but those results and the described genetic interactions were unable to characterize the molecular role of Echinus.  To better understand the molecular function of echinus, a screen for potential interacting proteins was undertaken. The molecular processes and signaling pathways within the cell often require proteins to assemble as complexes associated by direct or secondary interactions (Alberts, 1998; Vidal, 2005).  It is generally believed that these complexes contain proteins of similar function, and this property can be used to predict protein function and gain insight into the processes a protein may influence – a principle referred to as “guilt by association” (Choudhary and Grant, 2004).  Challenges in identification of these interacting proteins include the often transient nature of the interactions, the similarity that exists between proteins and protein isoforms and the sometimes vast differences in protein expression levels that can make detection difficult. Two methods have been widely used for analysis of protein-protein interactions: the yeast two-hybrid assay (Phizicky et al., 2003) and immunoprecipitation followed by mass spectrometry analysis (Aebersold and Mann, 2003).  The yeast two –hybrid assay is a genetic method that identifies interactions between proteins expressed within the nucleus of a yeast cell.  Applications of this method have successfully identified a large number of potential  97 interactions within the Drosophila genome (Giot et al., 2003; Formstecher et al., 2005), although often with conflicting results.  Another method, and one employed in this study, is immunoprecipitation followed by identification by LC-MS/MS (liquid chromatography, tandem mass spectrophotometry) analysis (Elias et al., 2005; Chang, 2006; Mueller et al., 2007).  Using this method (outlined in Figure 4.1), the protein of interest is immunoprecipitated from a cell lysate along with any proteins that may interact.  The resulting protein complexes are then separated by SDS-PAGE, fractionated by trypsin digestion and analyzed by mass spectrometry.  Tandem MS analysis identifies proteins by fragmenting the products of this digestion and accurately sizing the resulting peptides.  Bioinformatic analysis of the peptide masses then permits identification of the original protein.  This method enjoys some advantages over two hybrid analysis.  A primary advantage is that it allows expression of the target protein (bait) within the native cell type, resulting in proper post-translational modifications and localization to the native cellular compartment.  It is also capable of identifying protein complexes, resulting in a richer data set than the binary two-hybrid interactions.  There remain caveats, however, which must be considered when interpreting data from this type of assay.  The target protein is generally over-expressed, which may alter cellular function or promote artifactual interactions, and interactions within tissue culture cells may not necessarily reflect interactions normally found in vivo.  Additionally the process requires the interaction to be maintained upon cell lysis and immunoprecipitation, thus reducing the ability to detect weak or transient interactions. To screen for potential protein interaction partners, Echinus was fused to a FLAG affinity tag and expressed in Drosophila S2 cell culture.  Immunoprecipitation with an anti-FLAG antibody allowed recovery of this construct and the proteins associated with it, which were then identified by mass spectrometry.  These data, and an analysis of Echinus distribution  98 within the cell, indicated that Echinus is a vesicle-associated protein that interacts directly with the vesicle coat protein Clathrin and with the actin filament regulator Enabled.   99 Figure 4.1  Figure 4.1 | Overview of protein interaction assay by immunoprecipitation followed by LC-MS/MS analysis.  Echinus was fused to a FLAG affinity tag (1) and expressed in Drosophila S2 cell culture (2).  Immunoprecipitation of Echinus-FLAG (3) allowed recovery of interacting proteins (4), which were then identified by mass-spectrometry (5).  100 4.2) Materials and Methods  4.2.1) cDNA cloning into GatewayTM vectors Expression constructs were assembled using the GATEWAYTM system (Invitrogen). echinus (splice form 1) cDNA was generated from total RNA by reverse transcription using Superscript III (Invitrogen) and a gene-specific primer, followed by amplification with Pfx Polymerase (Invitrogen) and gene-specific primers modified to include AttB1 and AttB2 sequences.  The echinus cDNA was sequenced to ensure that it was free of non-conservative base changes in the coding region. cDNA constructs RE74715(RACK1), LD43101 (clathrin) and LD04938 (dco) (DGRC) were used as template for PCR amplification using Pfx Polymerase (Invitorgen) and gene-specific primers modified to include AttB1 and AttB2 sequences.  PCR products were cloned into pDONR221TM vector (Invitrogen) using BP Clonase (Invitrogen) to create entry clones, which were then used to transfer the cloned insert into expression vectors using LR Clonase (Invitrogen).  Expression vectors used were pAFW (N-terminal 3X FLAG), pAWF (C-terminal 3X FLAG) and pAMW (N-terminal 6xMyc), which employ the actin5C promoter to drive expression.  These vectors allow the expression in Drosophila cell culture of the cloned open reading frame fused to three copies of the FLAG peptide (DYKDDDDK) or 6 copies of the Myc antigen (EQKLISEEDL).  4.2.2) Cell culture and transfection Drosophila S2 cells (Invitrogen) were maintained in ESF921 serum free medium (Expression Systems) in 25cm2 flasks (Sarstedt) at 25°C without CO2.  Cells were passaged every 7 days or 4 days before transfection.  For cell transfection, 1ug of plasmid DNA and 10uL of Cellfectin (Invitrogen) were combined in 200uL Grace serum-free medium (Invitrogen), incubated at room temperature for 30 minutes and then added to 3x106 cells in  101 800uL Grace medium.  Cells were incubated overnight (16 hours) at 25°C in one well of a 24 well suspension plate (Sarstedt).  After incubation cells were split and 2mL of ESF921 media was added followed by an additional incubation of 24 hours.  4.2.3) Immunofluorescence Transfected cells were resuspended gently and 100uL of the suspension was added to each well of an 8-well CC2 coated chamber slide (Nunc) and incubated for at least one hour at room temperature unless otherwise noted.  Media was decanted and cells were fixed with 100uL of 4% paraformaldehyde in PBS + 1mM Ca2+ for 30 minutes.  Fixed cells were washed with PBS + 1mM Ca2+, permeablized with 0.2% Triton X-100 for 5 minutes and blocked with 1% BSA for 30 minutes.  A standard immunostaining protocol was followed thereafter. Primary antibodies used were: mouse anti-FLAG (1:2000, Sigma), rabbit anti-FLAG (1:2000, Sigma), mouse anti-Myc (1:1000, Roche), mouse anti-Armadillo N2 7A1 (1:50, DSHB), rat anti-DE-cadherin (1:25, DSHB), and mouse anti-Enabled 5G2 (1:50, DSHB).  Secondary antibodies used were anti-mouse or anti-rabbit IgG conjugated to Alexa 488 or 546 (1 ug/mL; Invitrogen) or anti-Rat IgG conjugated to Cy3 (Jackson Laboratories).  Where indicated, rhodamine phalloidin (6.6μM, Invitrogen) was included with the secondary antibody at a dilution of 1:100.  After immunostaining, cells were mounted with Slowfade Gold with DAPI (Invitrogen) and viewed with a Zeiss Axioplan 2 and 100X objective.  Images were captured using a Retiga 1350EX digital camera (Qimaging Corp.) and Northern Eclipse software (Empix Imagin, Inc.).  Confocal images were obtained with a 63X objective on a Nikon C1 instrument.  Images were analyzed with EZ-C1 V3.00 software and represent a single Z slice of 0.10μm.  102 The dextran endocytosis assay was performed using the conditions described above but with the following changes.  Uncoated glass 8-well chamber slides (Nunc) were used to minimize adhesion of the dextran to the slide surface.  100uL of cells were added to each chamber and incubated for 1 hour at room temperature to encourage adhesion.  Anionic Alexa488-labeled 10,000mw dextran (Invitrogen) was added to the culture medium at a final concentration of 1mM and incubated at room temperature for the indicated time.  Media was then carefully decanted and the cells were washed 3X with 200uL ESF921 media before processing as above.  4.2.4) Co-immunoprecipitation and MS/MS analysis [LC-MS/MS operation and initial informatic processing of peptide data using Mascot (Matrix Sciences, London, UK) and X!tandem software were performed by Dr. M. Kuzyk.]  For large-scale immunoprecipitation (IP) experiments, all steps were performed on ice or at 4°C unless otherwise indicated.  For each expression construct, 18x9mL of transfected cell culture was combined into 50mL polypropylene tubes, washed with 1X cold PBS and resuspended in 18mL of lysis buffer (20mM Tris pH7.5, 150mM NaCl, 1mM EDTA, 1% NP-40, 10mM β- glycerophosphate, 2mM sodium orthovanadate, 1mM AEBSF, 10μg/mL pepstatin A, 10μg/ml leupeptin and 10μg/ml aprotinin).  Cells were disrupted by 5 times passage through a 21G syringe and lysates were incubated with agitation for 30 minutes.  Cell extracts were clarified by centrifugation at 20,000g for 30min followed by filtration with a 0.45uM nylon syringe filter.  Clarified extracts were incubated with 180uL of a 50% slurry of Sepharose 4B (Sigma) and agitated gently for 1 hour.  Sepharose was removed by centrifugation and supernatant transferred to new tubes containing 60uL of 50% slurry of anti-FLAG-M2 agarose resin, followed by incubation at room temperature for 1 hour.  Anti-FLAG resin was recovered via  103 centrifugation and washed 5X with lysis buffer and once with 50mM ammonium bicarbonate. Bound proteins were eluted twice by resuspension in 50uL of 400ug/mL 3xFLAG peptide (Sigma).  Eluates were combined, vacuum dried, resuspended in protein sample buffer (Invitrogen) and separated by SDS-PAGE using a 10-20% NuPAGE gradient gel (Invitrogen) and 1X MES buffer.  Protein bands were visualized with colloidal Coomassie stain and each lane was cut into 16 equal sections.  Gel slices were transferred into a 96 well plate, reduced with 10 mM dithiothreitol (DTT), S-alkylated with 100 mM iodoacetamide and then subjected to in-gel trypsin digestion with 20μl of 20ng/μl trypsin per well overnight at 37°C. Resulting peptides were extracted under basic and acidic (50% v/v acetonitrile, 5% v/v formic acid) conditions. Peptide mixtures were subjected to LC-MS/MS analysis on a Finnigan LCQ (PTRL West) or 4000QTRAP (Applied Biosystems) ion trap mass spectrometers via reversed phase HPLC nano-electrospray ionization. All MS/MS spectra were queried against Drosophila Ensembl release 43 sequence databases using Mascot (Matrix Sciences, London, UK) or X!tandem algorithms. An in-house web-based database called SpecterWeb (Sun M, Kurzyk, M, and Morin GM, unpublished) was developed to store and process raw mass spectrometric protein identifications. Nonspecifically binding proteins that were found in the IPs from cells transfected with vector only (negative control) were subtracted using SpecterWeb software from the proteins that were identified in the cells transfected with Echinus bait constructs. Stringent criteria were employed to assign MS/MS spectra to peptides sequences. Protein identifications require 2 or more unique peptides assigned to high quality (at least e-value <- 3), manually validated MS/MS spectra. For reciprocal co-expressed coIP experiments, 4 x 9mL of transfected culture was used for each condition and processed as above using 4mL of lysis buffer and a similarly reduced volume of other reagents.  Proteins separated on SDS-PAGE were subsequently subjected to Western blot analysis.  104 4.2.5) Western blotting IP eluates were loaded onto 10-20% NuPAGE gradient SDS-PAGE gels (Invitrogen) and separated with 1X MES buffer (Invitrogen).  Proteins were transferred to methanol-wetted PVDF membrane (Millipore) and processed according to the standard protocol for the Li-COR system: membranes were blocked for 1 hour in Odyssey buffer (Li-COR Biosciences), primary antibody was added and incubated with shaking overnight at 4°C.  Membranes were washed with 1xTBS and secondary antibodies were diluted in Odyssey buffer and added.  Secondary antibodies used were goat anti-mouse IR800 (Rockland Immunochemicals) and goat anti- rabbit IR700 (Rockland Immunochemicals) for 1 hour at room temperature.  Membranes were imaged using an Odyssey Infrared Imaging System (Li-COR Biosciences).  4.2.6) Computational analysis of protein sequences Protein sequences were obtained in FASTA format from NCBI and Flybase.  Analysis was performed using Perl 5.8.8 running on the Linux operating system v2.6.23.  105 4.3) Results 4.3.1) Echinus protein localizes to intracellular vesicles Defining its subcellular localization often facilitates functional characterization of a protein.  To determine the potential cellular domains in which it may function, Echinus was fused to a C-terminal FLAG peptide and expressed in Drosophila S2 cell culture.  The Echinus- FLAG fusion localizes to numerous intracellular vesicles (Figure 4.2).  Large vesicles are observed in the perinuclear region and small punctate staining, possibly small vesicles, can be observed in the cell periphery and in association with the plasma membrane.  The circular staining pattern of the large vesicles (arrowhead, Figure 4.2D) suggests that Echinus is associated with the vesicular membrane and not contained within the vesicular space. Small punctations often appear more concentrated in presumptive leading edges and lamellae, suggesting they contribute to, or respond to, cell motility.  Similar results were obtained when FLAG was fused to the N-terminus of Echinus (Figure 4.4A’) and an anti-Echinus polyclonal antibody displayed a largely similar staining pattern in transfected cells.  These data indicate that in S2 cells Echinus protein may associate with the membranes of intracellular vesicles.   106 Figure 4.2   Figure 4.2 | Echinus-FLAG localizes to intracellular vesicles.  A C-terminal FLAG fusion to Echinus expressed in Drosophila S2 cell culture.  Staining for (A) actin and (B) anti-FLAG shows localization of Echinus-FLAG to intracellular vesicles.  (C) Merge of phalloidin staining (red) and anti-FLAG (green). Cell nuclei are stained with DAPI (blue).  Large vesicles are observed in the perinuclear region while in the cell periphery small punctate staining suggests smaller vesicles.  Association with the periphery often appears polarized, with accumulation concentrating to regions of the cell opposite to cell contacts.  (D) Echinus-FLAG labeling appears around the perimeter of large vesicles, suggesting it is associated with the vesicle membrane (D, arrowhead).  Small punctate structures are occasionally observed in radial lines throughout the cell periphery (D, arrow).  107 4.3.2) The C-terminal region of Echinus is required for localization to vesicles Echinus is a large protein that possibly contains multiple functional domains.  To better understand the region of the protein required for localization to intracellular vesicles several subclones of Echinus were constructed (Figure 4.3).  These subclones were expressed as N- terminal FLAG fusion proteins in Drosophila S2 cells and visualized by immunofluorescence (Figure 4.4).  Echinus subclones that lack the C-terminal portion of the protein were spread diffusely throughout the cell and possibly also in association with the plasma membrane (Figure 4.4B’, C’, D’).  A subclone containing the C-terminus of the protein, however, displayed obvious association with vesicles (Figure 4.4E’).  These results indicate that a domain within the C-terminal region of Echinus is required for its association with intracellular vesicles.        108 Figure 4.3   Figure 4.3 | Echinus subclones constructed for subcellular localization analysis.  A schematic of Echinus segments cloned into N-terminal FLAG fusion vector for expression in S2 cell culture.  The locations of putative functional domains and motifs are indicated within each subclone.  Ec-S34 and Ec-S4 failed to transfect efficiently and were not used in later analyses. Numbers above indicate amino acid position relative to splice form 1.  Asterisks show the location of putative clathrin-binding motifs.   109 Figure 4.4   110 Figure 4.4 | The Echinus C-terminus is required for localization to vesicles. (A-A”) Full- length Echinus localizes to vesicles (arrowhead) and the outer membrane (arrows). (B-D”) Expression of a FLAG fusion to the N-terminal region of Echinus subclones lacking the C- terminus (see Figure 4.3) results in diffuse cytoplasmic localization and possible association with the outer membrane.  (E-E”) Expression of an Echinus segment including the C-terminus results in localization of the protein to vesicle structures.  (A”-E”) Merge of phalloidin staining (red) and anti-FLAG (green). Cell nuclei are stained with DAPI (blue).  111 4.3.3) Echinus vesicles do not co-localize with internalized dextran or Rab5 The roles of vesicles within the cell are diverse.  A well-characterized role of small vesicles, particularly those observed to associate with the outer membrane of the cell, is the endocytosis of cell surface receptors.  Small Echinus-positive vesicles often appear in radial patterns that are suggestive of active transport along the microtubles of the cytoskeleton (arrow Figure 4.2D; arrow, Figure 4.5 A’). To determine if Echinus vesicles were endocytic in origin, transfected cells were stained with an antibody against the early endosomes marker Rab5.  Rab5 staining showed no detectible overlap with staining for Echinus-FLAG (Figure 4.5 B-B”) indicating that small Echinus-containing vesicles are not early endosomes.  Another marker of endocytosis is the internalization of fluorescently labeled dextran from the culture medium.   Transfected cells were immersed in a 1mM solution of Alexa 488 labeled dextran and incubated for 30 minutes prior to fixing.  Internalized dextran was not observed in association with vesicles containing Echinus (Figure 4.5 C-C”).  (Cells remained spherical in this assay due to the use of uncoated glass slides that fail to encourage cell adhesion.)  These results neither support nor conclusively refute the hypothesis that vesicles containing Echinus are endocytic in nature.  However, they do suggest that Echinus vesicles may be associated with some other process.   112 Figure 4.5   Figure 4.5 | Echinus vesicles do not co-localize with the early endosome marker Rab5 or an endocytosis marker.  (A-A”) Small Echinus-FLAG positive structures are often observed in a radial pattern and in association with the cell leading edge, suggesting they are actively trafficking (arrow, A’).  (B-B”) Echinus-FLAG vesicles (arrow, red in B”) do not co-localize with the early endosomes marker Rab5 (arrowhead, green in B”).  (C-C”) Fluorescently labeled dextran in the culture medium is internalized non-specifically during endocytosis. Fluorescent dextran (arrowhead, green in C”) is not observed within Echinus-FLAG vesicles (arrow, red in C”) 30 minutes after addition to the culture medium.  Cell nuclei in are stained with DAPI (blue in A”, B”).  113  4.3.4) Echinus protein interacts with the vesicle coat protein Clathrin, Casein kinase Iε and the protein kinase C scaffolding protein RACK1 To better understand the role that echinus plays within the cell we conducted a protein interaction screen using a co-immunoprecipitation methodology followed by tandem MS analysis to identify interacting proteins.  N- and C-terminally FLAG-tagged Echinus was in Drosophila S2 cell culture under the control of the actin promoter and cell lysate was immunoprecipitated with anti-FLAG antibody.  Immunoprecipitates were analyzed by LC- MS/MS for protein detection and identification.  Fifteen candidate interacting proteins were observed to interact with both the N- and C-terminal Echinus-FLAG fusion constructs (Table 4.1) but not with either negative control (vector only).  These candidates included the vesicle coat protein clathrin, the microtubule protein β- tubulin and Drosophila casein kinase Iε (dco – discs overgrown kinase).  A large number of additional proteins were observed to interact with only one of the two FLAG fusions (Table 4.2 and Supplementary Table 1).  These candidates included the coatomer proteins β-COP and γ-COP, the scaffolding protein RACK1 and a number of proteins of unknown function.  Analysis of candidate protein sequences revealed a relative overabundance of proteins containing homopolymers of the amino acids glycine or alanine (asterisks, Table 4.1, 4.2 and Supplementary Table 1).  While this phenomenon might be reflective of Echinus function, a similar overabundance of poly-glycine containing proteins was observed in the negative control and similar results have been observed previously under similar IP conditions (G. Cheng, unpublished).  These proteins were thus identified as possible artifacts of the immunoprecipitation process and considered with caution during subsequent analyses. A subset of the high confidence candidate interacting proteins was selected for confirmation by reciprocal immunoprecipitation analysis.  The vesicular localization of  114 Echinus-FLAG prompted a confirmation of the Clathrin-Echinus interaction.  Similarly, the association of PKC (a RACK1 binding partner) and casein kinase I with regulation of E- cadherin in the cell membrane (Le et al., 2002; Dupre-Crochet et al., 2007) provided a rationale for confirmation of RACK1-Echinus and Dco-Echinus interactions.  Clathrin, RACK1 and Dco were cloned into N-terminal FLAG expression vectors and each was co-transfected with a vector generating Echinus fused to a N-terminal Myc peptide.  Transfected cells were lysed and immunoprecipitated with anti-FLAG agarose to isolate the FLAG fusions and any associated proteins.  Western blotting with anti-FLAG and anti-Myc revealed that each FLAG fusion was capable of interacting with Myc-Echinus (Figure 4.6).  These reciprocal immunoprecipitation experiments are currently undergoing confirmation with additional biological replicates. Analysis of co-transfected cells by immunofluorescence was used to examine the distribution patterns of the FLAG fusions relative to Myc-Echinus (Figure 4.7).  RACK1-FLAG and Dco-FLAG primarily displayed a diffuse cytoplasmic distribution while clathrin appeared concentrated in irregular structures within the perinuclear region.  Clathrin and Echinus staining patterns appear to have significant overlap and Myc-Echinus localization to vesicles appeared reduced or blocked in these cells (Figure 4.7 D’-D”).  Staining with a native antibody to clathrin gave poor results in this immunofluorescence assay but suggested a similar overlap of distribution exists between Echinus-FLAG and endogenous clathrin in S2 cells.      115 Table 4. 1 Accession Name (molecular function if known) Mass (kDa) FLAG fusion log(E) Unique Peptides  CG8937-PA Hsc70-1 (ATPase activity; unfolded protein binding;) 70.6 C -67.3 7    N -27.8 4 CG9012-PA Clathrin heavy chain (vesicle coat protein) 191 C -61 7    N -3.2 1 CG11154-PA* ATP synthase beta chain, mitochondrial precursor (EC 54.1 C -47.7 6    N -12.4 2 CG9888-PA* Fibrillarin (rRNA processing) 34.6 C -42.8 6    N -2.6 1 CG1913-PA Tubulin alpha-1 chain (structural constituent of cytoskeleton) 49.9 C -31.2 4    N -12.8 2 CG2216-PA ferritin 1 heavy chain (ferric iron binding; oxidoreductase activity) 23.1 C -30.3 4    N -3.5 1 CG2048-PA* Discs overgrown kinase  (Casein kinase I epsilon - EC  47.9 N -25.7 4    C -2.1 1 CG14648-PA* growl (5-formyltetrahydrofolate cyclo-ligase)  59.7 C -21.9 3    N -13.8 2 CG4463-PA Heat shock protein 23 (actin binding, response to heat) 20.6 C -20.9 3    N -2.4 1 CG18212-PB aluminum tubes 95 C -20.7 3    N -9.3 2 CG6603-PA Hsc70Cb (chaperone binding; ATP binding) 88.4 C -19.5 3    N -9.7 2 CG1548-PA cathepsin D (aspartic peptidase) 42.4 C -14.4 2    N -11.3 2 CG4169-PA CG4169 (ubiquinol-cytochrome-c reductase) 45.4 C -12.6 2    N -9.9 2 CG8996-PA walrus (electron transfer flavoprotein) 34.1 C -12.1 2    N -3.1 1 CG9748-PA belle (ATP-dependent RNA helicase) 85 C -10.1 2    N -3.4 1  Table 4.1 | Proteins identified to interact with both N- and C-terminal Echinus FLAG. Fifteen proteins were identified as interactors of both N- and C-terminal FLAG fusion proteins. The molecular function of each protein is indicated when known, as is the identification confidence (log(E)) and number of unique peptides observed.  Bolded results highlight low- confidence interactions (unique peptides < 2 and/or log(E) > -3.0).  An asterisk adjacent to the accession number indicates that the protein contains a homopolymer of glycine or alanine.   116 Table 4.2 Accession    Name (molecular function if known)  Mass (kDa) FLAG fusion  log(E) Unique Peptides CG12701-PB * CG12701 170.2 N -109 14 CG9817-PA * CG9817 222.1 N -101 13 CG5394-PA * Bifunctional aminoacyl-tRNA synthetase  189.3 C -64 8 CG30122-PB * CG30122 140.4 C -63.8 8 CG6223-PA  beta-COP (Coatomer beta subunit) 107.3 C -54.7 7 CG3523-PA  Fatty acid synthase 266.3 C -46.2 5 CG1840-PA * CG1840-PA  11.8 N -44.2 6 CG7111-PA  RACK1 (Receptor of activated protein kinase C )  35.6 C -43.3 5 CG10067-PA  Actin-57B (structural constituent of cytoskeleton) 41.8 C -38.6 5 CG18572-PA  rudimentary (de novo pyrimidine base biosynthesis) 246.5 C -35.7 5 CG7752-PA * Z4 (DNA binding) 105 C -35 5 CG6453-PA  80K-H (PKC substrate) 61.5 C -31.1 4 CG1528-PA  gamma-COP (Coatomer gamma subunit) 97.2 C -30.4 4 CG6692-PA  Cathepsin L (aspartic protease) 37.9 C -30.1 4  Table 4.2 | Top scoring proteins identified to interact with either N- or C-terminal Echinus FLAG.  Many potential protein interactors were observed in association with only one of the FLAG fusions (a complete list is included in Supplementary Table 1).  The molecular function of each protein is indicated when known, as is the identification confidence (log(E)) and number of unique peptides observed.  An asterisk adjacent to the accession number indicates that the protein contains a homopolymer of glycine or alanine.    117 Figure 4.6  Figure 4.6 | Myc-Echinus co-immunoprecipitates with Dco-FLAG, RACK1-FLAG and Clathrin-FLAG.  (A) A Western blot of proteins immunoprecipitated with anti-FLAG agarose and probed for anti-Myc (top) or anti-FLAG (bottom).  Expression of Dco-FLAG, RACK1-FLAG or Clathrin-FLAG fusion was sufficient to immunoprecipitate Myc-Echinus. Myc-Echinus was not precipitated in the absence of FLAG-tagged interaction partners (lane 1).  (B) Lysates from transfected cells demonstrating the presence of FLAG fusions and Myc-Echinus in each cell culture lysate before immunoprecipitation.  118 Figure 4.7  Figure 4.7 | Co-expression of Dco-FLAG, RACK1-FLAG and Clathrin-FLAG with Myc- Echinus in Drosophila S2 cell culture.  Drosophila S2 cells transfected with Myc-Echinus (A-A”), Dco-FLAG and Myc-Echinus (B-B”), RACK1-FLAG and Myc-Echinus (C-C”) or Clathrin-FLAG and Myc-Echinus (D-D”).  Dco-FLAG and RACK1-FLAG display a cytoplasmic distribution (B, C) and do not appear concentrated in Myc-Echinus containing vesicles.  Clathrin-FLAG and Myc-Echinus have similar patterns of distribution and are localized primarily to irregularly shaped structures surrounding the nucleus (D-D”).  (A”-D”) Merge of phalloidin staining (red) and anti-FLAG (green). Cell nuclei are stained with DAPI (blue).  119 4.3.5) Echinus co-localizes and interacts with the Ena/VASP protein Enabled The genetic interaction between echinus and enabled suggests that these two genes functionally interact, and the presence in Echinus of a canonical binding motif for the EVH1 domain suggests that the two proteins physically interact.  To determine if the two proteins are co-localized within the cell we examined their relative distributions by immunofluorescence. Staining of Echinus-FLAG transfected cells with anti-FLAG and anti-Enabled antibodies demonstrated that these two proteins co-localize within Drosophila S2 cells (Figure 4.8 A-D). These data suggest that Echinus and Enabled interact physically, despite the fact that Enabled was not identified in the MS/MS dataset as an interaction partner.  This hypothesis was tested by immunoprecipitation of Echinus-FLAG followed by Western blot analysis with anti-Enabled antibody.  This analysis identified two Enabled-positive bands of similar molecular weights to the two known isoforms of Enabled (Figure 4.9) and is compatible with the notion that Echinus and Enabled interact within the S2 cells.  120 Figure 4.8   Figure 4.8 | Enabled co-localizes with Echinus-FLAG in Drosophila S2 cells.  Analysis by light (A-A”) and confocal microscopy (B-B”) show co-localization of Echinus and Enabled. Drosophila S2 cells expressing Echinus-FLAG are stained with anti-FLAG (A,B) and anti-Enabled (A’,B’) and show substantial overlap in localization (A”,B”).  Scale bar in (B-B”) is 10μm.  121 Figure 4.9   Figure 4.9 | Enabled co-immunoprecipitates with Echinus-FLAG.  A Western blot of an immunoprecipitation of Echinus-FLAG expressed in Drosophila S2 cells identifies Enabled as an interacting protein.  Two bands, corresponding to the two isoforms of Enabled, are detected using an anti-Enabled monoclonal antibody.  Enabled was not precipitated in the absence of Echinus-FLAG (lane 1).  122 4.4) Discussion The identification of echinus and characterization of its effects on retinal development stimulated a desire to understand its role within the cell.  To this end, the subcellular localization and interacting proteins were investigated to gain insight into the cellular processes in which echinus participates.  These analyses identified an association of Echinus with a series of vesicles within the cell, suggesting it may play a role in vesicle trafficking. Echinus-FLAG apparently localizes to a multitude of intracellular vesicles of various sizes (Figure 4.2).  Large vesicles are generally localized in the perinuclear region and small vesicles are in the periphery, although the location of large vesicles may be constrained to a certain extent by the “fried-egg” shape of cells as they adhere to the slide surface.  In addition to small vesicles at the cell surface, Echinus itself appears to associate with the outer membrane.  It often appears to be localized at or near the leading edge of lamellar structures, although this could be an artifact of cell shape and reflect association with the entire outer membrane.  Echinus staining appears as a ring around the vesicle (Figure 4.2D and Figure 4.7A’), implying that it is associated with the vesicle membrane and not contained within the vesicular space.  These vesicles are often very numerous within transfected cells, suggesting that perhaps high-level expression of echinus stimulates an increase in vesicle production, perhaps by stimulating budding or scission.  Small vesicles can often be observed in association with the outer membrane, indicating that they may be involved in transport of cell surface proteins.  They can also be seen in radial lines throughout the cell, which again is suggestive of active transport perhaps along microtubules. Whether these vesicles are involved in endocytosis or exocytosis is an important question for understanding echinus function, and one that has not yet been answered.  Initial experiments examining co-localization with the early endosome marker Rab5 and with internalized dextran both suggest that these vesicles are not endocytic in origin (Figure 4.5).  123 Attempts to determine Echinus localization relative to the lysosomal marker LAMP1 have been hindered by poor performance of available polyclonal antibodies with the immunofluorescence assay, and so this possibility warrants future attention because co- localization would link these vesicles to degradation of cellular proteins. Due to the effects of echinus mutations on the distribution of DE-cadherin and Armadillo (β-catenin) in retinal adherens junctions, Echinus-positive vesicles in S2 cells were examined by immunofluorescence for evidence of these proteins (data not shown).  However, no significant accumulation of these proteins was detected in Echinus-FLAG positive vesicles. Since S2 cells do not assemble adherens junctions these proteins are generally present at a low level within S2 cells, and so there may not be sufficient quantities to be detected using this assay.  An alternative explanation is that Echinus is involved in transport of a signaling molecule and it is the downstream effect of changes in a signaling pathway that result in the changes to adherens junction structure. Does Echinus instead function in the trafficking of cell surface signaling molecules?  A recent study provides evidence that Echinus is able to regulate the TGF-β signaling pathway. That study used expression of Punt, Tkv and MAD-FLAG constructs in S2 cell culture, followed by addition of Dpp to the culture media to induce MAD phosphorylation and translocation to the nucleus (Xu et al., 2007).  A whole-genome RNAi screen was employed to identify genes necessary for accumulation of the MAD-FLAG construct within the nucleus.  RNAi knockdown of echinus strongly disrupted the nuclear translocation of MAD-FLAG in this system.  However, as designed, the screen was unable to differentiate between failure of MAD phosphorylation and failure of phosphorylated FLAG-MAD to translocate to the nucleus.  From membrane to nucleus, Dpp signaling is known to require passage through vesicle intermediates (Gonzalez- Gaitan, 2003; Jekely and Rorth, 2003; Affolter and Basler, 2007) and it is possible that echinus is required for this signal trafficking.  An alternate possibility, and one that is perhaps more  124 harmonious with the results presented in this study, is that echinus knockdown disrupts the ability of the cell to transport the receptors Punt and Thickvein to the cell surface.  A decreased concentration of these receptors in the outer membrane would render the cell insensitive to the Dpp ligand in the assay culture medium and produce the observed result.  Examining the location of Punt and Tkv in the transfected cells, as well as the phosphorylation state of the FLAG-MAD construct, could test this possibility. Examining the localization of Echinus segments revealed that the C-terminal region of the protein is required for association with vesicles (Figure 4.4), possibly through its direct association with the vesicle coat protein clathrin (Figure 4.6).  This would suggest that the C- terminus of Echinus contains a clathrin-binding motif.  The consensus sequence for clathrin binding motifs is Lϕpϕ(-) (Dell'Angelica, 2001), where L represents leucine, ϕ denotes a bulky hydrophobic residue, p is a polar residue and (-) is a negatively charged residue. Echinus does contain a motif that matches this consensus sequence but it is located immediately adjacent to the active site within the ubiquitin hydrolase domain: 95LWHLD99 (Figure 2.7).  This location, at the N-terminus and within an enzymatic protein fold, makes it unlikely that this motif is responsible for association with clathrin.  Functional clathrin- binding motifs have a moderate degree of variability, however, and closely matching sequences can be found within the C-terminus: 1213LLKLR1217 and 1531LQQLD1535.  Of these, the second is within the C-terminal portion of Echinus required for vesicle association (Figures 4.3 and 4.4), and so this motif may be responsible for the localization of Echinus to vesicles through a direct interaction with clathrin. Echinus was also demonstrated to interact with RACK1 (Receptor for Activated C Kinase 1), a scaffolding protein that interacts with protein kinase C (PKC) and a myriad of other proteins.  RACK1 has a β-propeller-like structure, formed by seven WD-repeat sequences.  RACK1 binds the BII isoform of PKC as well proteins such as Src tyrosine kinase,  125 integrin, and phosphodiesterase (Schechtman and Mochly-Rosen, 2001). Its extensive list of interaction partners has implicated RACK1 in processes as diverse as cell adhesion and migration (Cox et al., 2003; Kiely et al., 2006), apoptosis (Mourtada-Maarabouni et al., 2005) and protein translation (Nilsson et al., 2004).  The role of RACK1 and PKC in regulating the concentration of E-cadherin in cell-cell contacts (Mourton et al., 2001; Le et al., 2002) also may be relevant to its interaction with Echinus, as is its possible association with Abl (Huang et al., 2008), which is a known regulator of Enabled. The interaction between Echinus and Dco (casein kinase Iε) suggests a possible link between the molecular function of Echinus and its described role in the eye.  In addition to its role as a structural protein in the adherens junction, beta-catenin is a transcriptional activator in the wingless (Wnt) signaling pathway.  The casein kinase I (CKI) family of serine/threonine kinases is evolutionarily conserved from yeast to mammals (Knippschild et al., 2005) and these proteins are known to play both a positive and negative role in wingless signaling (Swiatek et al., 2006) by regulating β-catenin (Armadillo) stability.  Casein kinase Iε (CKIε) appears to have a pleiotropic effect on wingless signaling.  It can act as a positive regulator by destabilizing the complex of proteins that degrade beta-catenin (Sakanaka et al., 1999; Sakanaka, 2002) and as a negative regulator by phosphorylating the Wnt co-receptors LRP5 and LRP6 (Swiatek et al., 2006).  Another role of CKIε that has overlap with Echinus function is its ability to regulate E-cadherin within the adherens junction. Phosphorylation of E- cadherin by CKIε decreases the interaction between E-cadherin and β-catenin and enhances endocytosis of E-cadherin (Dupre-Crochet et al., 2007).  Despite the interaction of Echinus and Dco the genes do not show a genetic interaction (Chapter 3).  This discrepancy may be due to the specific nature of the dco3 allele tested in the genetic interaction experiment or may result from artifactual data in the protein interaction assay.  126 echinus has been shown to interact genetically with enabled and to regulate its association with the adherens junction in the developing retina.  The EVH1 protein domain in Enabled mediates interactions with proteins containing a proline-rich motif (L/FPPPP).  Based on the presence of such a motif within Echinus, these proteins were hypothesized to interact.  This prediction was confirmed by co-immunoprecipitation and co-localization experiments. The role of this interaction awaits further investigation.  The Ena/VASP family of proteins inhibits the function of actin capping proteins, resulting in extension of actin fibres (Krause et al., 2003; Kwiatkowski et al., 2003).   A possible role of this interaction with Enabled is to promote motility of Echinus-containing vesicles.  The pathogen Listeria utilizes Ena/VASP proteins for intracellular motility in a process known as actin rocketing (Laurent et al., 1999; Reinhard et al., 2001).  A similar process has been described previously in vesicle movement (Allen, 2003; Southwick et al., 2003) and may be utilized by binding of Echinus to Enabled. Another possible role of this interaction may be to localize Enabled to targeted regions of the plasma membrane, which may help the cell to promote actin filament formation at specific cell contacts.  Finally, it is possible that the observed interaction between these two proteins is non-specific and that the overexpression of Echinus makes it an artificial ligand.  However the genetic interaction between these two genes suggests that this is not the case. The N-terminus of Echinus contains a ubiquitin hydrolase domain but despite the presence of all known active site residues, the protein lacks enzymatic function in an in vitro assay (Copeland et al., 2007).  This result led to the hypothesis that the function of this domain in Echinus was as a protein interaction domain to facilitate binding to a protein modified by addition of an ubiquitin monomer (Chapter 2).  Analysis of proteins interacting with Echinus was unable to support or refute this prediction as no proteins known to be monoubiquitinated (as a means of regulation) were observed in association with Echinus. Thus, the functional role of the ubiquitin hydrolase domain, if any, remains unknown.  127 These results demonstrate that Echinus is a vesicle-associated protein and reveal interactions between it and four other proteins: Clathrin, RACK1, Casein kinase Iε and Enabled.  A better understanding of the nature of these vesicles and the role interacting proteins play in Echinus function will help to explain how Echinus is able to regulate adherens junction stability in the developing retina.  Defining the basic function of the Echinus-positive vesicles, whether it is endocytic, exocytic, degrative or recycling, is an important next step. Few clues are contained within the list of protein interactors, although if confirmed the interaction with the golgi coat proteins β-COP and γ-COP would suggest these vesicles are involved in protein export.   Another potential protein interactor worthy of further investigation is 80K-H, which in a mammalian system has been shown to promote the trafficking and fusion of GLUT4 vesicles to the apical plasma membrane in response to insulin signaling (Brule et al., 2000; Hodgkinson et al., 2005).  Perhaps then, Echinus acts to promote trafficking and fusion of vesicles to a specific location within the outer membrane.   128 Supplemental Table 1 FLAG fusion  Accession     log(E)  Mass (kDa) Unique Peptides Description N CG12701-PB * -109.2 170 14 CG12701 N CG9817-PA * -101.4 222 13 CG9817 C CG8937-PA  -67.3 70.6 7 Hsc70-1 C CG5394-PA * -64 189 8 Bifunctional aminoacyl-tRNA synthetase C CG30122-PB * -63.8 140 8 CG30122 C CG9012-PA  -61 191 7 Clathrin heavy chain. C CG6223-PA  -54.7 107 7 Coatomer beta subunit (Beta-COP) C CG11154-PA * -47.7 54.1 6 ATP synthase beta chain, mitochondrial precursor (EC C CG3523-PA  -46.2 266 5 Fatty acid synthase N CG1840-PA * -44.2 11.8 6 CG1840-PA C CG7111-PA  -43.3 35.6 5 RACK1 (Receptor of activated protein kinase C ) C CG9888-PA * -42.8 34.6 6 Fibrillarin. C CG10067-PA  -38.6 41.8 5 Actin-57B. C CG18572-PA  -35.7 247 5 CAD protein (Protein rudimentary) C CG7752-PA * -35 105 5 CG7752-PA C CG6453-PA  -31.1 61.5 4 PKC substrate 80K-H C CG1528-PA  -30.4 97.2 4 Coatomer gamma subunit (gamma-COP) C CG2216-PA  -30.3 23.1 4 Ferritin C CG6692-PA  -30.1 37.9 4 Cathepsin L C CG4954-PA  -29.1 106 4 eIF3-S8 C CG11943-PA  -28.8 233 4 Nucleoporin (Nup205) C CG7762-PA * -28.3 102 4 CG7762-PA C CG2238-PA * -28.1 94.4 4 Elongation factor 2 (EF-2). N CG8937-PA  -27.8 70.6 4 Hsc70-1 C CG11198-PA  -26.6 278 4 CG11198-PB, isoform B N CG17489-PB  -25.8 34 4 CG17489-PD.3 N CG2048-PA * -25.7 47.9 4 Discs overgrown kinase (EC (Protein double-time). N CG14142-PA * -24.3 45.8 4 CG14142-PA C CG12030-PA  -22.8 38.7 3 UDP-glucose 4-epimerase (EC C CG14648-PA * -21.9 59.7 3 CG14648-PB, isoform B C CG18212-PB  -20.7 95 3 alt C CG6603-PA  -19.5 88.4 3 CG6603-PA, isoform A C CG3989-PA  -19.4 47.2 3 Multifunctional protein ADE2 (Adenosine-5) N CG11949-PA  -19.2 184 3 Protein 4.1 homolog (Coracle protein). C CG33547-PA * -19.1 315 3 Rim CG33547-PA N CG10019-PA  -18.1 126 3 CG10019-PA C CG32626-PA  -17.9 89.4 3 CG32626-PD, isoform D C CG6143-PB * -17.8 78 3 Zinc finger protein on ecdysone puffs. N CG8276-PB * -17.3 146 3 bicoid-interacting protein 3 CG8276-PB, isoform B C CG12202-PA  -17.3 103 3 Nat1 CG12202-PA C CG17246-PB  -17.3 72.3 3 Succinate dehydrogenase [ubiquinone] flavoprotein subunit, C CG4429-PA * -17.2 35 3 RNA-binding protein 2 CG4429-PC, isoform C C CG4357-PA  -17 129 3 sodium chloride cotransporter 69 CG4357-PB  129 N CG10706-PC * -16.9 101 3 small conductance calcium-activated potassium channel C CG11458-PA * -16.8 8.1 3 CG11458-PA C CG1548-PA  -14.4 42.4 2 cathD CG1548-PA N CG14648-PA * -13.8 59.7 2 CG14648-PB, isoform B C CG4169-PA  -12.6 45.4 2 CG4169-PA C CG5474-PA  -12.6 21.2 2 Signal sequence receptor CG5474-PA C CG18102-PG  -12.5 97.7 2 Dynamin (EC (dDyn) (Protein shibire). N CG11154-PA * -12.4 54.1 2 ATP synthase beta chain, mitochondrial (EC N CG17291-PC  -12.3 65.4 2 Protein phosphatase PP2A 65 kDa regulatory subunit A C CG8996-PA  -12.1 34.1 2 walrus CG8996-PB, isoform B C CG2331-PA  -11.9 88.8 2 CG2331-PB, isoform B C CG13185-PA  -11.5 605 2 Midasin C CG3751-PA  -11.5 15 2 RpS24 C CG9373-PA * -11.4 66.7 2 Hrp59 - RNA binding N CG1548-PA  -11.3 42.4 2 cathD CG1548-PA N CG7831-PA  -11.3 77.4 2 Claret segregational protein. C CG14066-PC  -11.2 151 2 La-related protein (dlarp). N CG32779-PA  -10.8 41.5 2 CG32779-PA C CG1516-PE  -10.7 131 2 CG1516-PL, isoform L C CG6778-PA  -10.7 75.7 2 Glycyl-tRNA synthetase CG6778-PB, isoform B C CG8542-PA  -10.6 74 2 Heat shock 70 kDa protein cognate 5. N CG2984-PA  -10.6 154 2 Protein phosphatase 2C CG2984-PA C CG7803-PA  -10.6 61.9 2 Regulatory protein zeste. C CG5371-PA  -10.6 91.9 2 Ribonucleoside-diphosphate reductase (EC N CG4290-PA  -10.5 150 2 CG4290-PA N CG15729-PA  -10.4 35.3 2 CG15729-PA C CG8772-PA  -10.3 78.5 2 no extended memory CG8772-PB, isoform B C CG5170-PA  -10.2 144 2 CG5170-PC, isoform C N CG6846-PA  -10.1 17.3 2 CG6846-PA C CG9748-PA  -10.1 85 2 CG9748-PA N CG1404-PB  -10.1 24.7 2 GTP-binding nuclear protein Ran. N CG1063-PA  -10 318 2 Inositol 1,4,5-trisphosphate receptor (InsP3 receptor) (InsP3R) C CG5670-PA  -10 116 2 Sodium/potassium-transporting ATPase alpha chain (EC N CG4169-PA  -9.9 45.4 2 CG4169-PA C CG9176-PB  -9.9 202 2 CNG channel-like CG9176-PC, isoform C N CG6603-PA  -9.7 88.4 2 CG6603-PA, isoform A N CG12175-PB  -9.6 45 2 toothrin CG12175-PB C CG3905-PA  -9.5 146 2 Protein suppressor 2 of zeste (Protein posterior sex combs). C CG9354-PB  -9.4 18.4 2 CG9354-PA, isoform A C CG4125-PA  -9.4 82.9 2 Roughest N CG18212-PB  -9.3 95 2 CG18212-PF, isoform F C CG9281-PB  -9.3 69.4 2 CG9281-PC, isoform C N CG5953-PA  -9.2 73.9 2 CG5953-PA, isoform A N CG5462-PA  -9.1 190 2 Protein lap4 (Protein scribble) (Protein smell-impaired). N CG1975-PA  -9.1 51.8 2 Rep2 CG1975-PA N CG32019-PA  -9 1001 2 bent CG32019-PE, isoform E C CG14471-PB  -9 131 2 CG14471-PA, isoform A C CG32206-PB  -8.9 138 2 CG32206-PC, isoform C  130 C CG33087-PC  -8.9 525 2 CG33087-PC N CG10223-PA  -8.9 164 2 DNA topoisomerase 2 (EC (DNA topoisomerase II) C CG12262-PA  -8.8 45.8 2 medium-chain specific acyl-CoA dehydrogenase(EC N CG32239-PA  -8.7 215 2 Guanine nucleotide exchange factor GEF64C CG32239-PA C CG9764-PA  -8.7 105 2 yurt CG9764-PA C CG2048-PA * -2.1 47.9 1 Discs overgrown kinase (EC (Protein double-time). N CG9012-PA  -3.2 191 1 Clathrin heavy chain.  Supplemental Table 1 | Complete list of proteins identified by LC-MS/MS analysis as interactors of Echinus  131 Chapter 5 | Summary and Future Research   This study sought to identify echinus and characterize its role in Drosophila eye development.  Echinus was successfully identified as the gene CG2904 and its role in regulating adherens junctions in the developing retina has been discovered and described. Genetic interactions with wingless, expanded and enabled were also observed.  Examination of Echinus function in cell culture revealed its association with intracellular vesicles and interactions with Clathrin, Dco, RACK1 and Enabled.  These results have answered many general questions about echinus and have raised many new more specific ones. What is the function of the ubiquitin hydrolase domain within Echinus?  The absence of a requirement for enzymatic activity in retinal development points to this domain being utilized instead for a possible interaction to a specific monoubiquitinated protein, possibly a cargo within the membranes of Echinus-positive vesicles.  Ubiquitination is also used for directing vesicle trafficking and Echinus may help to sense these signals.  An examination of the content of ubiquitinated proteins within Echinus-positive vesicles may provide clues to the role of this domain. What are the subcellular compartments in which Echinus is found?  The Echinus- positive vesicles remain incompletely characterized and defining the origin and function of these structures may enhance understanding of Echinus function by suggesting a subset of cellular processes that Echinus participates in.  This could perhaps be accomplished in vitro by comparing Echinus distribution to that of markers specific for structures such as the Golgi, ER, recycling endosome and lysosome. What is the significance of the genetic interactions with wingless and expanded?  Wingless is secreted from the apical cell surface of polarized cells (Marois et al., 2006) and its receptors  132 (Wu et al., 2004) and even its transcript are apically localized (Simmonds et al., 2001), as is the TGF-β receptor Thickvein (Bokel et al., 2006).  Expanded, in addition to its role in the fat tumor suppressor pathway, promotes turnover of cell surface receptors (Maitra et al., 2006). Perhaps Echinus functions to deliver proteins to specific regions of the cell membrane by regulation of vesicle trafficking, or instead “rescues” surface receptors that have been endocytosed by other means and returns them to the surface.  Determining the concentration of Thickvein, the Wingless receptors Arrow and DFrizzled2 (Rives et al., 2006) and other potential target proteins in the cell membrane of echinus mutants, and understanding the source and destination of Echinus-positive vesicles, would help to test these possibilities. What is the role of the interaction with Enabled?  Echinus interacts both genetically and physically with Enabled, suggesting the interaction is direct and relevant.  Dynamic regulation of actin fibre formation is known to be important for modulating the adhesion of cell contacts and perhaps this is how Echinus exerts its effect on the adherens junction.  Alternatively, Enabled may play a role in the motility of Echinus-containing vesicles.  An Echinus construct lacking the FPPPP motif would likely lose its ability to interact with Enabled.  In vivo expression of this construct in an echinus mutant background could be employed to test the role of the Echinus-Enabled interaction in adherens junction formation in the retina.  An in vitro examination of the effect of this construct on Echinus vesicle localization may also reveal a role for Enabled in the dynamics of these structures. A recent report on the gene Cindr indicates that it shares some common phenotypes with echinus, including regulation of DE-cadherin and Roughest and interaction with actin cytoskeleton regulators (Johnson et al., 2008).  This commonality in function encourages a future examination of possible interactions between echinus and cindr, which maybe lead to a better understanding of how they both contribute to adherens junction regulation.  133 Future studies will be required before an experimentally validated model of Echinus function can be proposed, and yet it is difficult to resist the urge to speculate with what is currently available.  Thus, a hypothetical model of Echinus function is presented for consideration (Figure 5.1).    134 Figure 5.1   Figure 5.1 | Speculative model of Echinus function.  (A) Echinus associates with vesicles trafficking to the cell surface.  Discs-overgrown kinase (Dco) and Enabled (Ena) are bound to Echinus, as well as a possible cell surface protein within the vesicle membrane.  80K-H is localized to regions of high Roughest concentration.  (B) Binding of Echinus to 80K-H results in vesicle fusion with the outer membrane, delivering cell surface proteins and localizing Dco and Ena near Roughest for downregulation of DE-cadherin and stimulation of the actin cytoskeleton.   135 This model suggests that Echinus associates with vesicles trafficking to the outer membrane.  At the cell membrane the cytoplasmic tail of Roughest has been demonstrated to interact with the adapter protein Dmint1 (Vishnu et al., 2006).  Mint1--Munc18 and Munc18- -80K-H interactions have been demonstrated previously (Okamoto and Sudhof, 1997; Biederer and Sudhof, 2000; Hodgkinson et al., 2005), and both are thought to be associated with the interior of the plasma membrane and mediate vesicle fusion.  Association of Echinus with 80K-H promotes vesicle fusion with the plasma membrane, possibly inserting a cell surface receptor or adhesion molecule and localizing Dco and Enabled to this region of the membrane.  Roughest has been shown to alter adherens junctions via downregulation of DE- cadherin and stimulation of actin fibre formation, processes which perhaps Dco and Ena contribute to.  While this model incorporates much experimental data and is compatible with other studies, an obvious deficiency is the proposed negative regulation of DE-cadherin, an effect opposite to that observed in this study.  This may be due to the dynamic nature of DE- cadherin regulation during cell sorting processes or may simply reflect a need for further consideration. Echinus was initially thought to promote apoptosis in the developing retina but instead as a consequence of my work it has emerged as a regulator of adherens junction stability. Recently, adherens junctions have become recognized as key components of the cell sorting process and future analysis of echinus will undoubtedly contribute to a better understanding of this process.  In addition, the regulation of cell-cell adhesion is a key component of both development and the pathology of diseases such as cancer.  In this regard, Drosophila has again shown its worth as a model system for investigating and understanding complex cellular processes.  136 Chapter 6 | References  Abrams, J. M. (1999). An emerging blueprint for apoptosis in Drosophila. Trends Cell Biol 9, 435-40. Abrams, J. M., White, K., Fessler, L. I. and Steller, H. (1993). Programmed cell death during Drosophila embryogenesis. Development 117, 29-43. Adams, M. D. Celniker, S. E. Holt, R. A. Evans, C. A. Gocayne, J. D. Amanatides, P. G. Scherer, S. E. Li, P. W. Hoskins, R. A. Galle, R. F. et al. (2000). The genome sequence of Drosophila melanogaster. Science 287, 2185-95. Adams, M. D. and Sekelsky, J. J. (2002). From sequence to phenotype: reverse genetics in Drosophila melanogaster. Nat Rev Genet 3, 189-98. Aebersold, R. and Mann, M. (2003). Mass spectrometry-based proteomics. Nature 422, 198- 207. Affolter, M. and Basler, K. (2007). The Decapentaplegic morphogen gradient: from pattern formation to growth regulation. Nat Rev Genet 8, 663-74. Alberts, B. (1998). The cell as a collection of protein machines: preparing the next generation of molecular biologists. Cell 92, 291-4. Allen, P. G. (2003). Actin filament uncapping localizes to ruffling lamellae and rocketing vesicles. Nat Cell Biol 5, 972-9. Alnemri, E. S., Livingston, D. J., Nicholson, D. W., Salvesen, G., Thornberry, N. A., Wong, W. W. and Yuan, J. (1996). Human ICE/CED-3 protease nomenclature. Cell 87, 171. Altschul, S. F., Gish, W., Miller, W., Myers, E. W. and Lipman, D. J. (1990). Basic local alignment search tool. J Mol Biol 215, 403-10. Altschul, S. F., Madden, T. L., Schaffer, A. A., Zhang, J., Zhang, Z., Miller, W. and Lipman, D. J. (1997). Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25, 3389-402. Andersen, K. M., Hofmann, K. and Hartmann-Petersen, R. (2005). Ubiquitin-binding proteins: similar, but different. Essays Biochem 41, 49-67. Aravind, L., Dixit, V. M. and Koonin, E. V. (1999). The domains of death: evolution of the apoptosis machinery. Trends Biochem Sci 24, 47-53. Aravind, L., Dixit, V. M. and Koonin, E. V. (2001). Apoptotic molecular machinery: vastly increased complexity in vertebrates revealed by genome comparisons. Science 291, 1279-84.  137 Arumugam, M., Wei, C., Brown, R. H. and Brent, M. R. (2006). Pairagon+N-SCAN_EST: a model-based gene annotation pipeline. Genome Biol 7 Suppl 1, S5 1-10. Baehrecke, E. H. (2002). How death shapes life during development. Nat Rev Mol Cell Biol 3, 779 - 787. Baker, N. E. (2001). Cell proliferation, survival, and death in the Drosophila eye. Semin Cell Dev Biol 12, 499-507. Baker, N. E. and Yu, S. Y. (2001). The EGF receptor defines domains of cell cycle progression and survival to regulate cell number in the developing Drosophila eye. Cell 104, 699-708. Banerjee, S., Sousa, A. D. and Bhat, M. A. (2006). Organization and function of septate junctions: an evolutionary perspective. Cell Biochem Biophys 46, 65-77. Bangs, P., Franc, N. and White, K. (2000). Molecular mechanisms of cell death and phagocytosis in Drosophila. Cell Death Differ 7, 1027-34. Bangs, P. and White, K. (2000). Regulation and execution of apoptosis during Drosophila development. Dev Dyn 218, 68-79. Bao, S. and Cagan, R. (2005). Preferential adhesion mediated by Hibris and Roughest regulates morphogenesis and patterning in the Drosophila eye. Dev Cell 8, 925 - 935. Baonza, A. and Freeman, M. (2005). Control of cell proliferation in the Drosophila eye by Notch signaling. Dev Cell 8, 529-39. Basler, K. and Hafen, E. (1991). Specification of cell fate in the developing eye of Drosophila. Bioessays 13, 621-31. Biederer, T. and Sudhof, T. C. (2000). Mints as adaptors. Direct binding to neurexins and recruitment of munc18. J Biol Chem 275, 39803-6. Bienz, M. (2005). beta-Catenin: a pivot between cell adhesion and Wnt signalling. Curr Biol 15, R64-7. Bokel, C., Schwabedissen, A., Entchev, E., Renaud, O. and Gonzalez-Gaitan, M. (2006). Sara endosomes and the maintenance of Dpp signaling levels across mitosis. Science 314, 1135-9. Bonini, N. M. and Fortini, M. E. (1999). Surviving Drosophila eye development: integrating cell death with differentiation during formation of a neural structure. Bioessays 21, 991-1003. Brachmann, C. B. and Cagan, R. L. (2003). Patterning the fly eye: the role of apoptosis. Trends Genet 19, 91-6. Bray, S. J. (2006). Notch signalling: a simple pathway becomes complex. Nat Rev Mol Cell Biol 7, 678 - 689.  138 Breitkreutz, B. J., Stark, C., Reguly, T., Boucher, L., Breitkreutz, A., Livstone, M., Oughtred, R., Lackner, D. H., Bahler, J., Wood, V. et al. (2008). The BioGRID Interaction Database: 2008 update. Nucleic Acids Res 36, D637-40. Brembeck, F. H., Rosario, M. and Birchmeier, W. (2006). Balancing cell adhesion and Wnt signaling, the key role of beta-catenin. Curr Opin Genet Dev 16, 51-9. Brule, S., Rabahi, F., Faure, R., Beckers, J. F., Silversides, D. W. and Lussier, J. G. (2000). Vacuolar system-associated protein-60: a protein characterized from bovine granulosa and luteal cells that is associated with intracellular vesicles and related to human 80K-H and murine beta-glucosidase II. Biol Reprod 62, 642-54. Bryant, D. M. and Stow, J. L. (2004). The ins and outs of E-cadherin trafficking. Trends Cell Biol 14, 427 - 434. Buckles, G. R., Rauskolb, C., Villano, J. L. and Katz, F. N. (2001). Four-jointed interacts with dachs, abelson and enabled and feeds back onto the Notch pathway to affect growth and segmentation in the Drosophila leg. Development 128, 3533-42. Burge, C. and Karlin, S. (1997). Prediction of complete gene structures in human genomic DNA. J Mol Biol 268, 78-94. Cagan, R. L. and Ready, D. F. (1989a). The emergence of order in the Drosophila pupal retina. Dev Biol 136, 346-62. Cagan, R. L. and Ready, D. F. (1989b). Notch is required for successive cell decisions in the developing Drosophila retina. Genes Dev 3, 1099-112. Calleja, M., Renaud, O., Usui, K., Pistillo, D., Morata, G. and Simpson, P. (2002). How to pattern an epithelium: lessons from achaete-scute regulation on the notum of Drosophila. Gene 292, 1-12. Carthew, R. W. (2007). Pattern formation in the Drosophila eye. Curr Opin Genet Dev 17, 309- 13. Carthew, R. W. and Rubin, G. M. (1990). seven in absentia, a gene required for specification of R7 cell fate in the Drosophila eye. Cell 63, 561-77. Celniker, S. E., Wheeler, D. A., Kronmiller, B., Carlson, J. W., Halpern, A., Patel, S., Adams, M., Champe, M., Dugan, S. P., Frise, E. et al. (2002). Finishing a whole-genome shotgun: release 3 of the Drosophila melanogaster euchromatic genome sequence. Genome Biol 3, RESEARCH0079. Chang, I. F. (2006). Mass spectrometry-based proteomic analysis of the epitope-tag affinity purified protein complexes in eukaryotes. Proteomics 6, 6158-66.  139 Chen, P. and Abrams, J. M. (2000). Drosophila apoptosis and Bcl-2 genes: outliers fly in. J Cell Biol 148, 625-7. Chen, P., Nordstrom, W., Gish, B. and Abrams, J. M. (1996). grim, a novel cell death gene in Drosophila. Genes Dev 10, 1773-82. Cho, E., Feng, Y., Rauskolb, C., Maitra, S., Fehon, R. and Irvine, K. D. (2006). Delineation of a Fat tumor suppressor pathway. Nat Genet 38, 1142-50. Choudhary, J. and Grant, S. G. (2004). Proteomics in postgenomic neuroscience: the end of the beginning. Nat Neurosci 7, 440-5. Colussi, P. A., Quinn, L. M., Huang, D. C., Coombe, M., Read, S. H., Richardson, H. and Kumar, S. (2000). Debcl, a proapoptotic Bcl-2 homologue, is a component of the Drosophila melanogaster cell death machinery. J Cell Biol 148, 703-14. Copeland, J. M., Bosdet, I., Freeman, J. D., Guo, M., Gorski, S. M. and Hay, B. A. (2007). echinus, required for interommatidial cell sorting and cell death in the Drosophila pupal retina, encodes a protein with homology to ubiquitin-specific proteases. BMC Dev Biol 7, 82. Cordero, J., Jassim, O., Bao, S. and Cagan, R. (2004). A role for wingless in an early pupal cell death event that contributes to patterning the Drosophila eye. Mech Dev 121, 1523 - 1530. Cordero, J. B., Larson, D. E., Craig, C. R., Hays, R. and Cagan, R. (2007). Dynamic decapentaplegic signaling regulates patterning and adhesion in the Drosophila pupal retina. Development 134, 1861 - 1871. Cox, E. A., Bennin, D., Doan, A. T., O'Toole, T. and Huttenlocher, A. (2003). RACK1 regulates integrin-mediated adhesion, protrusion, and chemotactic cell migration via its Src- binding site. Mol Biol Cell 14, 658-69. Dell'Angelica, E. C. (2001). Clathrin-binding proteins: got a motif? Join the network! Trends Cell Biol 11, 315-8. Dikic, I. (2003). Mechanisms controlling EGF receptor endocytosis and degradation. Biochem Soc Trans 31, 1178-81. Duffy, J. B. (2002). GAL4 system in Drosophila: a fly geneticist's Swiss army knife. Genesis 34, 1-15. Dupre-Crochet, S., Figueroa, A., Hogan, C., Ferber, E. C., Bialucha, C. U., Adams, J., Richardson, E. C. N. and Fujita, Y. (2007). Casein kinase 1 is a novel negative regulator of E- cadherin-based cell-cell contacts. Mol Cell Biol 27, 3804 - 3816. Elias, J. E., Haas, W., Faherty, B. K. and Gygi, S. P. (2005). Comparative evaluation of mass spectrometry platforms used in large-scale proteomics investigations. Nat Methods 2, 667-75.  140 Ellis, M. C., O'Neill, E. M. and Rubin, G. M. (1993). Expression of Drosophila glass protein and evidence for negative regulation of its activity in non-neuronal cells by another DNA- binding protein. Development 119, 855-65. Evan, G. and Littlewood, T. (1998). A matter of life and cell death. Science 281, 1317-22. Fiehler, R. W. and Wolff, T. (2008). Nemo is required in a subset of photoreceptors to regulate the speed of ommatidial rotation. Dev Biol 313, 533-44. Flockhart, I., Booker, M., Kiger, A., Boutros, M., Armknecht, S., Ramadan, N., Richardson, K., Xu, A., Perrimon, N. and Mathey-Prevot, B. (2006). FlyRNAi: the Drosophila RNAi screening center database. Nucleic Acids Res 34, D489-94. Flybase. (1994). FlyBase--the Drosophila database. The FlyBase Consortium. Nucleic Acids Res 22, 3456-8. Flybase. (2002). The FlyBase database of the Drosophila genome projects and community literature. Nucleic Acids Res 30, 106-8. Formstecher, E., Aresta, S., Collura, V., Hamburger, A., Meil, A., Trehin, A., Reverdy, C., Betin, V., Maire, S., Brun, C. et al. (2005). Protein interaction mapping: a Drosophila case study. Genome Res 15, 376-84. Freeman, M. (1994). The spitz gene is required for photoreceptor determination in the Drosophila eye where it interacts with the EGF receptor. Mech Dev 48, 25-33. Freeman, M. (1996). Reiterative use of the EGF receptor triggers differentiation of all cell types in the Drosophila eye. Cell 87, 651-60. Freeman, M. (1997). Cell determination strategies in the Drosophila eye. Development 124, 261- 70. Fujita, Y., Krause, G., Scheffner, M., Zechner, D., Leddy, H. E., Behrens, J., Sommer, T. and Birchmeier, W. (2002). Hakai, a c-Cbl-like protein, ubiquitinates and induces endocytosis of the E-cadherin complex. Nat Cell Biol 4, 222-31. Gates, J. and Peifer, M. (2005). Can 1000 reviews be wrong? Actin, alpha-Catenin, and adherens junctions. Cell 123, 769 - 772. Giot, L., Bader, J. S., Brouwer, C., Chaudhuri, A., Kuang, B., Li, Y., Hao, Y. L., Ooi, C. E., Godwin, B., Vitols, E. et al. (2003). A protein interaction map of Drosophila melanogaster. Science 302, 1727-36. Gonzalez-Gaitan, M. (2003). Endocytic trafficking during Drosophila development. Mech Dev 120, 1265-82.  141 Gorski, S. and Marra, M. (2002). Programmed cell death takes flight: genetic and genomic approaches to gene discovery in Drosophila. Physiol Genomics 9, 59-69. Gorski, S. M., Brachmann, C. B., Tanenbaum, S. B. and Cagan, R. L. (2000). Delta and notch promote correct localization of irreC-rst. Cell Death Differ 7, 1011-3. Goyal, L., McCall, K., Agapite, J., Hartwieg, E. and Steller, H. (2000). Induction of apoptosis by Drosophila reaper, hid and grim through inhibition of IAP function. Embo J 19, 589-97. Grevengoed, E. E., Fox, D. T., Gates, J. and Peifer, M. (2003). Balancing different types of actin polymerization at distinct sites: roles for Abelson kinase and Enabled. J Cell Biol 163, 1267-79. Grzeschik, N. A. and Knust, E. (2005). IrreC/rst-mediated cell sorting during Drosophila pupal eye development depends on proper localisation of DE-cadherin. Development 132, 2035- 45. Gumbiner, B. M. (2005). Regulation of cadherin-mediated adhesion in morphogenesis. Nat Rev Mol Cell Biol 6, 622-34. Hatzfeld, M. (1999). The armadillo family of structural proteins. Int Rev Cytol 186, 179-224. Hay, B. A. (2000). Understanding IAP function and regulation: a view from Drosophila. Cell Death Differ 7, 1045-56. Hay, B. A., Wolff, T. and Rubin, G. M. (1994). Expression of baculovirus P35 prevents cell death in Drosophila. Development 120, 2121-9. Hayashi, A., Ohnishi, H., Okazawa, H., Nakazawa, S., Ikeda, H., Motegi, S.-i., Aoki, N., Kimura, S., Mikuni, M. and Matozaki, T. (2004). Positive regulation of phagocytosis by SIRPbeta and its signaling mechanism in macrophages. J. Biol. Chem. 279, 29450 - 29460. Held, L. I. (2002). Imaginal Discs: The Genetic and Cellular Logic of Pattern Formation Cambridge University Press. Hicke, L. and Dunn, R. (2003). Regulation of membrane protein transport by ubiquitin and ubiquitin-binding proteins. Annu. Rev. Cell Dev. Biol. 19, 141 - 172. Hochstrasser, M. (1996). Ubiquitin-dependent protein degradation. Annu Rev Genet 30, 405-39. Hodgkinson, C. P., Mander, A. and Sale, G. J. (2005). Identification of 80K-H as a protein involved in GLUT4 vesicle trafficking. Biochem J 388, 785-93. Hu, M., Li, P., Li, M., Li, W., Yao, T., Wu, J. W., Gu, W., Cohen, R. E. and Shi, Y. (2002). Crystal structure of a UBP-family deubiquitinating enzyme in isolation and in complex with ubiquitin aldehyde. Cell 111, 1041-54.  142 Huang, C. C., Liu, C. H. and Chuang, N. N. (2008). An enhanced association of RACK1 with Abl in cells transfected with oncogenic ras. Int J Biochem Cell Biol 40, 423-31. Hurley, J. H., Lee, S. and Prag, G. (2006). Ubiquitin-binding domains. Biochem J 399, 361-72. Igney, F. H. and Krammer, P. H. (2002). Death and anti-death: tumour resistance to apoptosis. Nat Rev Cancer 2, 277 - 288. Jaiswal, M., Agrawal, N. and Sinha, P. (2006). Fat and Wingless signaling oppositely regulate epithelial cell-cell adhesion and distal wing development in Drosophila. Development 133, 925 - 935. Jekely, G. and Rorth, P. (2003). Hrs mediates downregulation of multiple signalling receptors in Drosophila. EMBO Rep 4, 1163 - 1168. Johnson, R. I., Seppa, M. J. and Cagan, R. L. (2008). The Drosophila CD2AP/CIN85 orthologue Cindr regulates junctions and cytoskeleton dynamics during tissue patterning. J Cell Biol 180, 1191-204. Jordens, I., Marsman, M., Kuijl, C. and Neefjes, J. (2005). Rab proteins, connecting transport and vesicle fusion. Traffic 6, 1070-7. Kelley, R. L., Meller, V. H., Gordadze, P. R., Roman, G., Davis, R. L. and Kuroda, M. I. (1999). Epigenetic spreading of the Drosophila dosage compensation complex from roX RNA genes into flanking chromatin. Cell 98, 513-22. Kent, W. J., Sugnet, C. W., Furey, T. S., Roskin, K. M., Pringle, T. H., Zahler, A. M. and Haussler, D. (2002). The human genome browser at UCSC. Genome Res 12, 996-1006. Kerr, J. F., Wyllie, A. H. and Currie, A. R. (1972). Apoptosis: a basic biological phenomenon with wide-ranging implications in tissue kinetics. Br J Cancer 26, 239-57. Kiely, P. A., O'Gorman, D., Luong, K., Ron, D. and O'Connor, R. (2006). Insulin-like growth factor I controls a mutually exclusive association of RACK1 with protein phosphatase 2A and beta1 integrin to promote cell migration. Mol Cell Biol 26, 4041-51. Knippschild, U., Gocht, A., Wolff, S., Huber, N., Lohler, J. and Stoter, M. (2005). The casein kinase 1 family: participation in multiple cellular processes in eukaryotes. Cell Signal 17, 675-89. Krammer, P. H. (2000). CD95's deadly mission in the immune system. Nature 407, 789-95. Krause, M., Dent, E. W., Bear, J. E., Loureiro, J. J. and Gertler, F. B. (2003). Ena/VASP proteins: regulators of the actin cytoskeleton and cell migration. Annu Rev Cell Dev Biol 19, 541- 64.  143 Kumar, J. P. and Moses, K. (2001). The EGF receptor and notch signaling pathways control the initiation of the morphogenetic furrow during Drosophila eye development. Development 128, 2689-97. Kurada, P. and White, K. (1998). Ras promotes cell survival in Drosophila by downregulating hid expression. Cell 95, 319-29. Kwiatkowski, A. V., Gertler, F. B. and Loureiro, J. J. (2003). Function and regulation of Ena/VASP proteins. Trends Cell Biol 13, 386-92. Lai, E. C. (2004). Notch signaling: control of cell communication and cell fate. Development 131, 965 - 973. Larkin M.A., B. G., Brown N.P., Chenna R., McGettigan P.A., McWilliam H.*, Valentin F.*, Wallace I.M., Wilm A., Lopez R.*, Thompson J.D., Gibson T.J. and Higgins D.G. (2007). ClustalW and ClustalX version 2. Bioinformatics 23, 2947-2948. Larson, D. E., Liberman, Z. and Cagan, R. L. (2008). Cellular behavior in the developing Drosophila pupal retina. Mech Dev 125, 223-32. Laurent, V., Loisel, T. P., Harbeck, B., Wehman, A., Grobe, L., Jockusch, B. M., Wehland, J., Gertler, F. B. and Carlier, M. F. (1999). Role of proteins of the Ena/VASP family in actin- based motility of Listeria monocytogenes. J Cell Biol 144, 1245-58. Lawlor, T. A. (1980). Genetic and cytological localization of mei-9. D.R.I.S. 55, 81--82. Le, B. R. (2006). Regulation of Notch signalling by endocytosis and endosomal sorting. Curr Opin Cell Biol 18, 213 - 222. Le, B. R., Bardin, A. and Schweisguth, F. (2005). The roles of receptor and ligand endocytosis in regulating Notch signaling. Development 132, 1751 - 1762. Le, T. L., Joseph, S. R., Yap, A. S. and Stow, J. L. (2002). Protein kinase C regulates endocytosis and recycling of E-cadherin. Am J Physiol Cell Physiol 283, C489-99. Lee, W. R. (1989). Addition of molecular methods to mutation studies with Drosophila melanogaster. Environ Mol Mutagen 14 Suppl 16, 99-104. Lefevre and Wilkins. (1966). Cytogenetic studies on the white locus in Drosophila melanogaster. Genetics 53, 175--187. Liu, Q. A. and Hengartner, M. O. (1999). The molecular mechanism of programmed cell death in C. elegans. Ann N Y Acad Sci 887, 92-104. Magie, C. R., Pinto-Santini, D. and Parkhurst, S. M. (2002). Rho1 interacts with p120ctn and alpha-catenin, and regulates cadherin-based adherens junction components in Drosophila. Development 129, 3771-82.  144 Maitra, S., Kulikauskas, R. M., Gavilan, H. and Fehon, R. G. (2006). The tumor suppressors Merlin and Expanded function cooperatively to modulate receptor endocytosis and signaling. Curr Biol 16, 702 - 709. Marois, E., Mahmoud, A. and Eaton, S. (2006). The endocytic pathway and formation of the Wingless morphogen gradient. Development 133, 307-17. Martin, S. J. (2002). Destabilizing influences in apoptosis: sowing the seeds of IAP destruction. Cell 109, 793-6. Matthews, K. A., Kaufman, T. C. and Gelbart, W. M. (2005). Research resources for Drosophila: the expanding universe. Nat Rev Genet 6, 179-93. Mege, R. M., Gavard, J. and Lambert, M. (2006). Regulation of cell-cell junctions by the cytoskeleton. Curr Opin Cell Biol 18, 541-8. Meier, P., Finch, A. and Evan, G. (2000). Apoptosis in development. Nature 407, 796-801. Meller, V. H., Wu, K. H., Roman, G., Kuroda, M. I. and Davis, R. L. (1997). roX1 RNA paints the X chromosome of male Drosophila and is regulated by the dosage compensation system. Cell 88, 445-57. Metzstein, M. M., Stanfield, G. M. and Horvitz, H. R. (1998). Genetics of programmed cell death in C. elegans: past, present and future. Trends Genet 14, 410-6. Miller, D. T. and Cagan, R. L. (1998). Local induction of patterning and programmed cell death in the developing Drosophila retina. Development 125, 2327-35. Mirkovic, I. and Mlodzik, M. (2006). Cooperative activities of drosophila DE-cadherin and DN-cadherin regulate the cell motility process of ommatidial rotation. Development 133, 3283 - 3293. Miyoshi, J. and Takai, Y. (2008). Structural and functional associations of apical junctions with cytoskeleton. Biochim Biophys Acta 1778, 670-91. Monserrate, J. P. and Brachmann, C. B. (2007). Identification of the death zone: a spatially restricted region for programmed cell death that sculpts the fly eye. Cell Death Differ 14, 209 - 217. Montrasio, S., Mlodzik, M. and Fanto, M. (2007). A new allele uncovers the role of echinus in the control of ommatidial rotation in the Drosophila eye. Dev Dyn 236, 2936 - 2942. Morris, J. R., Chen, J., Filandrinos, S. T., Dunn, R. C., Fisk, R., Geyer, P. K. and Wu, C. (1999). An analysis of transvection at the yellow locus of Drosophila melanogaster. Genetics 151, 633-51.  145 Mourtada-Maarabouni, M., Kirkham, L., Farzaneh, F. and Williams, G. T. (2005). Functional expression cloning reveals a central role for the receptor for activated protein kinase C 1 (RACK1) in T cell apoptosis. J Leukoc Biol 78, 503-14. Mourton, T., Hellberg, C. B., Burden-Gulley, S. M., Hinman, J., Rhee, A. and Brady-Kalnay, S. M. (2001). The PTPmu protein-tyrosine phosphatase binds and recruits the scaffolding protein RACK1 to cell-cell contacts. J Biol Chem 276, 14896-901. Mueller, D. R., Voshol, H., Waldt, A., Wiedmann, B. and Van Oostrum, J. (2007). LC- MALDI MS and MS/MS--an efficient tool in proteome analysis. Subcell Biochem 43, 355-80. Muro, I., Hay, B. A. and Clem, R. J. (2002). The Drosophila DIAP1 protein is required to prevent accumulation of a continuously generated, processed form of the apical caspase DRONC. J Biol Chem 22, 22. Niebuhr, K., Ebel, F., Frank, R., Reinhard, M., Domann, E., Carl, U. D., Walter, U., Gertler, F. B., Wehland, J. and Chakraborty, T. (1997). A novel proline-rich motif present in ActA of Listeria monocytogenes and cytoskeletal proteins is the ligand for the EVH1 domain, a protein module present in the Ena/VASP family. Embo J 16, 5433-44. Nijman, S. M., Luna-Vargas, M. P., Velds, A., Brummelkamp, T. R., Dirac, A. M., Sixma, T. K. and Bernards, R. (2005). A genomic and functional inventory of deubiquitinating enzymes. Cell 123, 773-86. Nilsson, J., Sengupta, J., Frank, J. and Nissen, P. (2004). Regulation of eukaryotic translation by the RACK1 protein: a platform for signalling molecules on the ribosome. EMBO Rep 5, 1137- 41. Nordstrom, W. and Abrams, J. M. (2000). Guardian ancestry: fly p53 and damage-inducible apoptosis. Cell Death Differ 7, 1035-8. Nordstrom, W., Chen, P., Steller, H. and Abrams, J. M. (1996). Activation of the reaper gene during ectopic cell killing in Drosophila. Dev Biol 180, 213-26. Numata, K., Okada, Y., Saito, R., Kiyosawa, H., Kanai, A. and Tomita, M. (2007). Comparative analysis of cis-encoded antisense RNAs in eukaryotes. Gene 392, 134-41. Okamoto, M. and Sudhof, T. C. (1997). Mints, Munc18-interacting proteins in synaptic vesicle exocytosis. J Biol Chem 272, 31459-64. Oliver, B., Perrimon, N. and Mahowald, A. (1988). Genetic evidence that the sans fille locus is involved in Drosophila sex determination. Genetics 120, 159--171. Overstreet, E., Fitch, E. and Fischer, J. A. (2004). Fat facets and Liquid facets promote Delta endocytosis and Delta signaling in the signaling cells. Development 131, 5355 - 5366.  146 Palaga, T. and Osborne, B. (2002). The 3D's of apoptosis: death, degradation and DIAPs. Nat Cell Biol 4, E149-51. Pasic, L., Kotova, T. and Schafer, D. A. (2008). Ena/VASP Proteins Capture Actin Filament Barbed Ends. J Biol Chem 283, 9814-9. Pellock, B. J., Buff, E., White, K. and Hariharan, I. K. (2007). The Drosophila tumor suppressors Expanded and Merlin differentially regulate cell cycle exit, apoptosis, and Wingless signaling. Dev Biol 304, 102 - 115. Perez-Moreno, M., Jamora, C. and Fuchs, E. (2003). Sticky business: orchestrating cellular signals at adherens junctions. Cell 112, 535-48. Phizicky, E., Bastiaens, P. I., Zhu, H., Snyder, M. and Fields, S. (2003). Protein analysis on a proteomic scale. Nature 422, 208-15. Pickart, C. M. (2001). Mechanisms underlying ubiquitination. Annu Rev Biochem 70, 503-33. Potter, C. J., Turenchalk, G. S. and Xu, T. (2000). Drosophila in cancer research. An expanding role. Trends Genet 16, 33-9. Quesada, V., Diaz-Perales, A., Gutierrez-Fernandez, A., Garabaya, C., Cal, S. and Lopez- Otin, C. (2004). Cloning and enzymatic analysis of 22 novel human ubiquitin-specific proteases. Biochem Biophys Res Commun 314, 54-62. Ready, D. F., Hanson, T. E. and Benzer, S. (1976). Development of the Drosophila retina, a neurocrystalline lattice. Dev Biol 53, 217-40. Reese, M. G., Kulp, D., Tammana, H. and Haussler, D. (2000). Genie--gene finding in Drosophila melanogaster. Genome Res 10, 529-38. Reinhard, M., Jarchau, T. and Walter, U. (2001). Actin-based motility: stop and go with Ena/VASP proteins. Trends Biochem Sci 26, 243-9. Reiter, C., Schimansky, T., Nie, Z. and Fischbach, K. F. (1996). Reorganization of membrane contacts prior to apoptosis in the Drosophila retina: the role of the IrreC-rst protein. Development 122, 1931-40. Renfranz, P. J. and Beckerle, M. C. (2002). Doing (F/L)PPPPs: EVH1 domains and their proline-rich partners in cell polarity and migration. Curr Opin Cell Biol 14, 88-103. Richardson, H. and Kumar, S. (2002). Death to flies: Drosophila as a model system to study programmed cell death. J Immunol Methods 265, 21-38. Rives, A. F., Rochlin, K. M., Wehrli, M., Schwartz, S. L. and DiNardo, S. (2006). Endocytic trafficking of Wingless and its receptors, Arrow and DFrizzled-2, in the Drosophila wing. Dev Biol 293, 268 - 283.  147 Rodriguez, A., Chen, P., Oliver, H. and Abrams, J. M. (2002). Unrestrained caspase- dependent cell death caused by loss of Diap1 function requires the Drosophila Apaf-1 homolog, Dark. Embo J 21, 2189-97. Rodriguez, A., Oliver, H., Zou, H., Chen, P., Wang, X. and Abrams, J. M. (1999). Dark is a Drosophila homologue of Apaf-1/CED-4 and functions in an evolutionarily conserved death pathway. Nat Cell Biol 1, 272-9. Ross, C. A. (2002). Polyglutamine pathogenesis: emergence of unifying mechanisms for Huntington's disease and related disorders. Neuron 35, 819-22. Sakanaka, C. (2002). Phosphorylation and regulation of beta-catenin by casein kinase I epsilon. J Biochem 132, 697-703. Sakanaka, C., Leong, P., Xu, L., Harrison, S. D. and Williams, L. T. (1999). Casein kinase iepsilon in the wnt pathway: regulation of beta-catenin function. Proc Natl Acad Sci U S A 96, 12548-52. Sambrook, J., Fritsch, E. F. and Maniatis, T. (1989). Molecular Cloning: A Laboratory Manual: Cold Spring Harbor Laboratory Press. Schechtman, D. and Mochly-Rosen, D. (2001). Adaptor proteins in protein kinase C- mediated signal transduction. Oncogene 20, 6339-47. Schneider, T., Reiter, C., Eule, E., Bader, B., Lichte, B., Nie, Z., Schimansky, T., Ramos, R. G. and Fischbach, K. F. (1995). Restricted expression of the irreC-rst protein is required for normal axonal projections of columnar visual neurons. Neuron 15, 259-71. Scott, J. A., Shewan, A. M., den Elzen, N. R., Loureiro, J. J., Gertler, F. B. and Yap, A. S. (2006). Ena/VASP proteins can regulate distinct modes of actin organization at cadherin- adhesive contacts. Mol Biol Cell 17, 1085-95. Silva, E., Tsatskis, Y., Gardano, L., Tapon, N. and McNeill, H. (2006). The tumor-suppressor gene fat controls tissue growth upstream of expanded in the hippo signaling pathway. Curr Biol 16, 2081 - 2089. Simmonds, A. J., dosSantos, G., Livne-Bar, I. and Krause, H. M. (2001). Apical localization of wingless transcripts is required for wingless signaling. Cell 105, 197-207. Simon, M. A., Bowtell, D. D., Dodson, G. S., Laverty, T. R. and Rubin, G. M. (1991). Ras1 and a putative guanine nucleotide exchange factor perform crucial steps in signaling by the sevenless protein tyrosine kinase. Cell 67, 701-16. Smit, A., Hubley, R & Green, P. (1996-2004). RepeatMasker Open-3.0  148 Soboleva, T. A. and Baker, R. T. (2004). Deubiquitinating enzymes: their functions and substrate specificity. Curr Protein Pept Sci 5, 191-200. Song, Z. and Steller, H. (1999). Death by design: mechanism and control of apoptosis. Trends Cell Biol 9, M49-52. Southwick, F. S., Li, W., Zhang, F., Zeile, W. L. and Purich, D. L. (2003). Actin-based endosome and phagosome rocketing in macrophages: activation by the secretagogue antagonists lanthanum and zinc. Cell Motil Cytoskeleton 54, 41-55. Spencer, S. A., Powell, P. A., Miller, D. T. and Cagan, R. L. (1998). Regulation of EGF receptor signaling establishes pattern across the developing Drosophila retina. Development 125, 4777-90. St Johnston, D. (2002). The art and design of genetic screens: Drosophila melanogaster. Nat Rev Genet 3, 176-88. Stang, E., Blystad, F. D., Kazazic, M., Bertelsen, V., Brodahl, T., Raiborg, C., Stenmark, H. and Madshus, I. H. (2004). Cbl-dependent ubiquitination is required for progression of EGF receptors into clathrin-coated pits. Mol Biol Cell 15, 3591-604. Stevens, T. L., Rogers, E. M., Koontz, L. M., Fox, D. T., Homem, C. C., Nowotarski, S. H., Artabazon, N. B. and Peifer, M. (2008). Using bcr-abl to examine mechanisms by which abl kinase regulates morphogenesis in Drosophila. Mol Biol Cell 19, 378-93. Strutt, H. and Strutt, D. (2003). EGF signaling and ommatidial rotation in the Drosophila eye. Curr Biol 13, 1451-7. Sturtevant, A. H. (1970). Studies on the bristle pattern of Drosophila. Dev Biol 21, 48-61. Sun, L. and Chen, Z. J. (2004). The novel functions of ubiquitination in signaling. Curr Opin Cell Biol 16, 119 - 126. Swiatek, W., Kang, H., Garcia, B. A., Shabanowitz, J., Coombs, G. S., Hunt, D. F. and Virshup, D. M. (2006). Negative regulation of LRP6 function by casein kinase I epsilon phosphorylation. J Biol Chem 281, 12233-41. Tanenbaum, S. B., Gorski, S. M., Rusconi, J. C. and Cagan, R. L. (2000). A screen for dominant modifiers of the irreC-rst cell death phenotype in the developing Drosophila retina. Genetics 156, 205-17. Tepass, U. and Harris, K. P. (2007). Adherens junctions in Drosophila retinal morphogenesis. Trends Cell Biol 17, 26-35. Tepass, U., Tanentzapf, G., Ward, R. and Fehon, R. (2001). Epithelial cell polarity and cell junctions in Drosophila. Annu Rev Genet 35, 747-84.  149 Thornberry, N. A. and Lazebnik, Y. (1998). Caspases: enemies within. Science 281, 1312-6. Tissot, M. and Stocker, R. F. (2000). Metamorphosis in drosophila and other insects: the fate of neurons throughout the stages. Prog Neurobiol 62, 89-111. Tomoyasu, Y., Nakamura, M. and Ueno, N. (1998). Role of dpp signalling in prepattern formation of the dorsocentral mechanosensory organ in Drosophila melanogaster. Development 125, 4215-24. Tyler, D. M. and Baker, N. E. (2007). Expanded and fat regulate growth and differentiation in the Drosophila eye through multiple signaling pathways. Dev Biol 305, 187 - 201. Untergasser, A., Nijveen, H., Rao, X., Bisseling, T., Geurts, R. and Leunissen, J. A. (2007). Primer3Plus, an enhanced web interface to Primer3. Nucleic Acids Res 35, W71-4. Van Etten, R. A. (1999). Cycling, stressed-out and nervous: cellular functions of c-Abl. Trends Cell Biol 9, 179-86. Vaux, D. L. and Korsmeyer, S. J. (1999). Cell death in development. Cell 96, 245-54. Vernooy, S. Y., Copeland, J., Ghaboosi, N., Griffin, E. E., Yoo, S. J. and Hay, B. A. (2000). Cell death regulation in Drosophila: conservation of mechanism and unique insights. J Cell Biol 150, F69-76. Vidal, M. (2005). Interactome modeling. FEBS Lett 579, 1834-8. Vidal, M. and Cagan, R. L. (2006). Drosophila models for cancer research. Curr Opin Genet Dev 16, 10 - 16. Villa, P., Kaufmann, S. H. and Earnshaw, W. C. (1997). Caspases and caspase inhibitors. Trends Biochem Sci 22, 388-93. Vishnu, S., Hertenstein, A., Betschinger, J., Knoblich, J. A., Gert de Couet, H. and Fischbach, K. F. (2006). The adaptor protein X11Lalpha/Dmint1 interacts with the PDZ- binding domain of the cell recognition protein Rst in Drosophila. Dev Biol 289, 296-307. Wang, Q. and Margolis, B. (2007). Apical junctional complexes and cell polarity. Kidney Int 72, 1448-58. Weis, W. I. and Nelson, W. J. (2006). Re-solving the cadherin-catenin-actin conundrum. J Biol Chem 281, 35593-7. Weissman, A. M. (2001). Themes and variations on ubiquitylation. Nat Rev Mol Cell Biol 2, 169- 78. Willecke, M., Hamaratoglu, F., Kango-Singh, M., Udan, R., Chen, C.-L., Tao, C., Zhang, X. and Halder, G. (2006). The fat cadherin acts through the hippo tumor-suppressor pathway to regulate tissue size. Curr Biol 16, 2090 - 2100.  150 Wodarz, A., Stewart, D. B., Nelson, W. J. and Nusse, R. (2006). Wingless signaling modulates cadherin-mediated cell adhesion in Drosophila imaginal disc cells. J. Cell Sci. 119, 2425 - 2434. Wolff, T. and Ready, D. F. (1991a). The beginning of pattern formation in the Drosophila compound eye: the morphogenetic furrow and the second mitotic wave. Development 113, 841- 50. Wolff, T. and Ready, D. F. (1991b). Cell death in normal and rough eye mutants of Drosophila. Development 113, 825-39. Wolff, T. and Ready, D. F. (1993). Pattern formation in the Drosophila retina. In The Development of Drosophila melanogaster,  (ed. M. Bate and A. Martinez Arias), pp. 1277-1325: Cold Spring Harbor Laboratory Press. Wu, J., Klein, T. J. and Mlodzik, M. (2004). Subcellular localization of frizzled receptors, mediated by their cytoplasmic tails, regulates signaling pathway specificity. PLoS Biol 2, E158. Xu, L., Yao, X., Chen, X., Lu, P., Zhang, B. and Ip, Y. T. (2007). Msk is required for nuclear import of TGF-{beta}/BMP-activated Smads. J Cell Biol 178, 981-94. Yu, S. Y., Yoo, S. J., Yang, L., Zapata, C., Srinivasan, A., Hay, B. A. and Baker, N. E. (2002). A pathway of signals regulating effector and initiator caspases in the developing Drosophila eye. Development 129, 3269-78. Zakeri, Z. and Lockshin, R. A. (2002). Cell death during development. J Immunol Methods 265, 3-20. Zerial, M. and McBride, H. (2001). Rab proteins as membrane organizers. Nat Rev Mol Cell Biol 2, 107-17. Zimmermann, K. C., Ricci, J. E., Droin, N. M. and Green, D. R. (2002). The role of ARK in stress-induced apoptosis in Drosophila cells. J Cell Biol 156, 1077-87.  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items