京都大学 / 0048 / 新制・課程博士 / 博士(医学) / 甲第21262号 / 医博第4380号 / 新制||医||1029(附属図書館) / 京都大学大学院医学研究科医学専攻 / (主査)教授 平井 豊博, 教授 伊達 洋至, 教授 渡邊 直樹 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

<p>Organelle transport in eukaryotic cells is regulated by a precisely coordinated activity of phosphoinositide lipids, small GTPases, and molecular motors. Despite the extensive study of functional activities of individual regulators, how these activities promote precise deliveries of particular membrane proteins to specific cellular locations remained unclear. Ankyrin-B, which is previously well recognized as a plasma membrane adaptor that assembles diverse specialized plasma membrane domains, exhibited an unexpected role in intracellular transport. This thesis establishes ankyrin-B as a master integrator of the polarized long range organelle transport via direct interactions with Rab GTPase Activating Protein 1 Like (RabGAP1L), phosphatidylinositol 3-phosphate (PI3P) and dynactin 4. In Chapter 2, I identified an ankyrin-B death domain binding partner, RabGAP1L, that specifically interacts with ankyrin-B on intracellular organelles and requires ankyrin-B for its proper localization. In Chapter 3, I demonstrated that ankyrin-B is a PI3P-effector in mouse embryonic fibroblasts (MEFs) and promotes the polarized transport of associated organelles in migrating cells in a RabGAP1L-dependent manner. I continued to investigate what membranes/membrane-associated proteins utilize the ankyrin-B/RabGAP1L pathway in Chapter 4 and identified α5β1-integrin as a cargo whose polarized transport and recycling are ankyrin-B-dependent. I further presented that ankyrin-B, through recruiting RabGAP1L to PI3P-positive/Rab22A-associated endosomes containing α5β1-integrin, promotes polarized recycling of α5β1-integrin in migrating mouse embryonic fibroblasts. In collaboration with James Bear (UNC Chapel Hill), we further demonstrated that this ankyrin-B/RabGAP1L-mediated pathway is required for haptotaxis along fibronectin gradients. In Chapter 5, I elucidated the in vivo interaction between ankyrin-B and RabGAP1L. I demonstrated that ankyrin-B specifically interacts with RabGAP1L at long axon tracts in the brain and at costameres in the skeletal muscle. I also show the phenotypic analysis of ankyrin-B floxDD mice as an initial attempt to determine the physiological function of the ankyrin-B death domain in vivo. Together, this thesis dissects an ankyrin-B-mediated molecular mechanism for polarized endosomal trafficking and α5β1-integrin recycling during directional cell migration, and provides new insights into how phosphoinositide lipids, Rab GTPases, and molecular motor activities are coordinated to control the directional transport of specialized membrane cargos.</p> / Dissertation

Thesis advisor: David R. Burgess / Polarity established by the first cleavages in sea urchin embryos was investigated in this thesis revealing precocious embryonic polarity. Studies of embryonic polarity have focused on protostomes such as <italics>C. elegans</italics>, and those on deuterostomes have focused on later developmental stages. I find asymmetries in the sea urchin membrane cell cortex as early as the first division after fertilization as a result of new membrane addition in the cleavage furrow. Membrane domains and the polarity determinants Par6, aPKC, and Cdc42 are polarized to the apical, or free, cell surface, while the cell-cell contact site remains distinct. Using immunofluorescence, fluorescence recovery after photobleaching (FRAP), and specific inhibitor treatments, myosin filaments were identified as the major regulator of membrane cortex polarity. However, membrane domains and cortical polarity determinants are differentially regulated with respect to blastomere dissociation. These asymmetries are required for proper spindle alignment and cleavage plane determination and are responsible for polarized fluid phase endocytosis. The work in this thesis and future studies addressing the connection between the membrane cortex and myosin filaments has and will lead to a greater understanding of the maintenance of embryonic polarity in cleavage stage sea urchin embryos. / Thesis (PhD) — Boston College, 2009. / Submitted to: Boston College. Graduate School of Arts and Sciences. / Discipline: Biology.

The establishment of cell polarity is a vital requirement for cellular processes such as proper cell division, growth and movement. Cell polarization relies on different internal and external cues in order to reorient the cell growth machinery along the axis of polarity. The core mechanisms involved in establishment of polarized growth are highly conserved from yeast to humans. Cells of the fission yeast Schizosaccharomyces pombe grow in a highly polarized fashion, with cell growth restricted to the cell tips, making fission yeast an excellent model system to study polarized growth. Here I describe a system for long-term live-cell imaging of fission yeast polarized growth that is stress free, physiological and accessible to media change and drug addition. I use this improved imaging system along with yeast genetics and drug perturbations to address how cell polarity is established and maintained in fission yeast. I have shown that fission yeast growth polarity depends on competition and cooperation among three distinct internal polarity cues: 1) A microtubule-based cue involving Tea1/Tea4 polarity proteins positively regulates polarized growth, initially at the “old” cell end (i.e., the end that pre-existed in the mother cell) and later at the “new” cell end (i.e. the end that is generated by septation), in order to initiate the transition from monopolar to bipolar growth (also known as New End Take-Off, or “NETO”). 2) An actin cable-based cue “clears” polarity proteins from the new end immediately after cytokinesis thereby reinforcing old-end growth. As a result perturbation of actin cable-based transport by either deleting actin cable nucleator For3 or cable-based transporter Myo52 results in premature bipolar growth. 3) A novel “memory-based” growth polarity cue helps to establish polarized growth in the absence of the microtubule-based cue. This memory-based cue is dependent on the predicted transmembrane proteins Rax1/Rax2. In the absence of both Tea1/Tea4 cue and Rax1/Rax2 cue, cells depend on septation cue and grow exclusively from the cell ends generated by septation. Furthermore, both Tea1/Tea4 and Rax1/Rax2 cue are important to maintain polarized growth under various environmental stresses. In fission yeast, during interphase, nucleus is positioned at the centre of the cell and this precise positioning of nucleus, which is important for defining the position of cytokinetic ring is thought to be exclusively MT-dependent. Here I show that MT-independent nuclear movement exists in fission yeast and this nuclear movement is mediated by actin cables and type myosin myo52. Furthermore, I show that actin cable might be important for buffering the pushing forces generated by MTs on the nucleus. In this way both microtubules and actin cables are involved in nuclear movement in fission yeast.

The Wnt signaling network has critical roles in embryonic development and is implicated in human disease. One of the outputs of the Wnt network, called the planar cell polarity (PCP) pathway, regulates tissue polarity and directs cell migration. Core PCP components (Frizzled, Dishevelled, Prickle, Vangl, Celsr) localize asymmetrically in polarized cells and establish polarity across the tissue through protein interactions between adjacent cells. The core PCP component activate tissue-specific "effectors" which translate the signal into morphological changes. PCP is related to several disease conditions, including neural tube defects, cystic kidney disease, and cance metastasis. However, mechanisms of the PCP underlying physiological and disease-related conditions are not well understood. Here, I explore functions of the core PCP component Pk, and its relationship to disease, in the zebrafish model system. Mutations in Pk1 and Pk2 have been identified in human progressive myoclonic epilepsy patients. Pk coodinate cell movement, neuronal migration and axonal outgrowth during embryonic development. Yet, how dysfunctions of pk relates to epilepsy is unknown. Here, I show that knockdown of pk1a sensitizes the zebrafish larva to convulsant drug. To model the defects in central nervous system, I examine neurogenesis in the retina and find that both pk1a and pk2 are required for proper dendritic outgrowth in the retinal inner plexiform layer. Furthermore, I characterize the epilepsy-related mutant forms of Pk1a and Pk2. The mutant Pk1a forms show reduced ability to suppress the retinal neurogenesis defects compared to the wild-type, as well as differential ubiquitination levels. Pk2 mutant forms also show differential activities in overexpression assays and seemingly more stable proteins relative to the wild-type. Taken together, pk1a and pk2 may contribute to epilepsy by affecting neuronal patterning and thus signal processing. Another aspect of PCP function has been implicated in cilia and cilia-related disorders, also called ciliopathy. PCP effectors have been shown to modulate ciliogenesis and core PCP proteins (Vang and Dvl) regulate cilia orientation. On the other hand, cilia are not required for PCP signaling, especially asymmetric core PCP protein localization. These findings leave open the question what is the precise relationship between PCP and cilia. The Bardet Biedl Syndrome (BBS) is a type of ciliopathy that leads to obesity, retinitis pigmentosa, polydactyly, mental retardation and other symptons. A subset of BBS genes share similar knockdown phenotype in cell migration as seen in PCP knockdown embryos. Shared pehnotypes have led some to proposethat PCP and bbs genes may interact. Yet a direct relationship has yet to be established. I examine the interaction between pk2 and a central Bbs gene, bbs7. By analyzing shared phenotypes in double knockdown embryos, I find no synergistic interaction between the two, suggesting they act in distinct pathways. Bbs regulate ciliary trafficking and in zebrafish, knockdown of bbs genes leads to delayed retrograde melanosome transport. Interestingly, I find knockdown of pk2 suppresses this retrograde transport delay. Additionally, pk2 knockdown embryos show a delay in anterograde melanosome transport. These findings highlight a new role for pk2 in intracellular transport and clarifies the relationship between PCP and BBS. In summary, my work here strengthens the link between pk mutations and human epilepsy and identifies functions of pk in retinal neurogenesis and in intracellular transport. To what extent the role of neurogenesis and intracellular transport are related is worth future study. Yet, this new information provides insights into potential mechanisms of epilepsy and the relationship between PCP and BBS.

During neuronal development, the axonal and dendritic projections are polarized and oriented along specific body axis. To further explore the molecular basis of neuritogenesis in vivo, we used the nematode Caenorhabditis elegans as a developmental model and performed a forward genetic screen to identify genes that specify the polarity of neurite outgrowth. We examined the VC4 and VC5 neurons, members of the six VC motor neurons using the Pcat-1::gfp transgene cyIs4. The VC motor neurons are ventrally located neurons that extend two processes. VC1, VC2, VC3, and VC6 extend axons along the anterior-posterior (A/P) axis; VC4 and VC5 extend axons around the vulva along a mediolateral left-right (L/R) axis perpendicular to the A/P axis. We identified and showed that vang-1/Van Gogh, a core component of planar cell polarity (PCP) signalling pathway, acts cell-autonomously in VC4 and VC5 neurons and non-autonomously from the epithelial cells to restrict neurite formation along the A/P axis. vang-1 mutant animals display ectopic neurites along the A/P axis. Using a candidate gene approach, we further identified and revealed two additional core members of PCP signalling, Prickle (PRKL-1) and Dishevelled (DSH-1), to play a role in A/P-directed neurite suppression. We also showed prkl-1 and dsh-1 genetically interact with vang-1 and VANG-1 is required to suppress A/P-directed neurite outgrowth from larval stage 4 to adulthood. Overexpression of VANG-1 results in a loss-of-function (lof) phenotype, suggesting that an appropriate level of VANG-1 activity is important. Additionally, vang-1/prkl-1, and dsh-1 may interact in parallel pathways. Our findings implicate PCP genes to play a previously unidentified role in maintaining polarized neuronal morphology by inhibiting neuronal outgrowth responses to environmental cues.

Baz/PAR3 is a key regulator of epithelial cell polarity (ECP). To identify proteins functioning with Baz, I completed a baz genetic interaction screen by localizing 15 GFP-tagged candidates. Then I tested the role of a top candidate, dASAP (Drosophila Arf GTPase-activating protein with SH3 domain, Ankyrin repeat and PH domain), in Drosophila ECP. To determine whether dASAP might interact with polarity players, I defined the localization of dASAP throughout embryogenesis with GFP-tagged proteins and an anti-dASAP antibody. To study how loss of dASAP function affects ECP, I generated a deletion allele by imprecise P-element excision. To evaluate how each of the six domains of dASAP contributes to its localization and functions, I generated constructs deleting each domain. I found associations between dASAP, actin and the apical domain. The six domains may act redundantly to localize dASAP, although interactions between domains may affect the degree of membrane association.

Baz/PAR3 is a key regulator of epithelial cell polarity (ECP). To identify proteins functioning with Baz, I completed a baz genetic interaction screen by localizing 15 GFP-tagged candidates. Then I tested the role of a top candidate, dASAP (Drosophila Arf GTPase-activating protein with SH3 domain, Ankyrin repeat and PH domain), in Drosophila ECP. To determine whether dASAP might interact with polarity players, I defined the localization of dASAP throughout embryogenesis with GFP-tagged proteins and an anti-dASAP antibody. To study how loss of dASAP function affects ECP, I generated a deletion allele by imprecise P-element excision. To evaluate how each of the six domains of dASAP contributes to its localization and functions, I generated constructs deleting each domain. I found associations between dASAP, actin and the apical domain. The six domains may act redundantly to localize dASAP, although interactions between domains may affect the degree of membrane association.

Cell polarity is fundamental to numerous cellular processes including migration, molecular transport, and cell division. The establishment and organization of polarity is crucial to the maintenance of cellular homeostasis in mammalian systems. Deregulation of cell polarity is observed in disease states, including cancer. Numb is an adaptor protein that functions in regulating endocytic trafficking events. Numb was originally identified in Drosophila as an asymmetrically localized cell fate determinant, and was subsequently found to be conserved in vertebrates. In mammalian polarized epithelial cells, Numb is distributed asymmetrically along the basolateral membrane domain. The work herein describes phosphorylation of Numb by the Par complex protein, atypical Protein Kinase C (aPKC), as a means of regulating membrane localization and asymmetric distribution of Numb. A mutant of Numb that cannot be phosphorylated by aPKC accumulates on the plasma membrane and localizes to both apical and basolateral membranes. In aPKC-depleted cells, endogenous Numb is unable to achieve polarized distribution and localizes around the entire cell cortex. We demonstrate that this mechanism is conserved in Drosophila as mutation of the corresponding phosphorylation sites disrupts Numb asymmetric localization in dividing sensory organ precursor cells. In polarized epithelial cells, one function of Numb is to promote epithelial morphology when cells are challenged with external stimuli that disrupt cell-cell adhesion. For example, depletion of Numb results in enhanced sensitivity of cells to lose cell-cell contacts when treated with calcium chelating agents. Loss of Numb potentiates hepatocyte growth factor (HGF)-induced lamellipodia formation and cell dispersal – early steps in epithelial-mesenchymal transition (EMT). In Numb-depleted cells, Rac1-GTP loading is enhanced, which corresponds with increased rate in loss of cell-cell adhesion and increased lamellipodia formation, following depletion of extracellular calcium and HGF stimulation, respectively. Together, this work identifies a mechanism that regulates polarized distribution of Numb and provides insight into its function in polarized epithelial cells.

Cell polarity is fundamental to numerous cellular processes including migration, molecular transport, and cell division. The establishment and organization of polarity is crucial to the maintenance of cellular homeostasis in mammalian systems. Deregulation of cell polarity is observed in disease states, including cancer. Numb is an adaptor protein that functions in regulating endocytic trafficking events. Numb was originally identified in Drosophila as an asymmetrically localized cell fate determinant, and was subsequently found to be conserved in vertebrates. In mammalian polarized epithelial cells, Numb is distributed asymmetrically along the basolateral membrane domain. The work herein describes phosphorylation of Numb by the Par complex protein, atypical Protein Kinase C (aPKC), as a means of regulating membrane localization and asymmetric distribution of Numb. A mutant of Numb that cannot be phosphorylated by aPKC accumulates on the plasma membrane and localizes to both apical and basolateral membranes. In aPKC-depleted cells, endogenous Numb is unable to achieve polarized distribution and localizes around the entire cell cortex. We demonstrate that this mechanism is conserved in Drosophila as mutation of the corresponding phosphorylation sites disrupts Numb asymmetric localization in dividing sensory organ precursor cells. In polarized epithelial cells, one function of Numb is to promote epithelial morphology when cells are challenged with external stimuli that disrupt cell-cell adhesion. For example, depletion of Numb results in enhanced sensitivity of cells to lose cell-cell contacts when treated with calcium chelating agents. Loss of Numb potentiates hepatocyte growth factor (HGF)-induced lamellipodia formation and cell dispersal – early steps in epithelial-mesenchymal transition (EMT). In Numb-depleted cells, Rac1-GTP loading is enhanced, which corresponds with increased rate in loss of cell-cell adhesion and increased lamellipodia formation, following depletion of extracellular calcium and HGF stimulation, respectively. Together, this work identifies a mechanism that regulates polarized distribution of Numb and provides insight into its function in polarized epithelial cells.