Development and Morphogenesis

159

Cell and tissue mechanics in zebrafish gastrulation. C-P. Heisenberg1; 1IST Austria, Klosterneuburg, Austria

Tissue morphogenesis during embryonic development is brought about by mechanical forces which are generated by the specific biophysical and motility properties of its constituent cells. It

has also been suggested that embryonic tissues behave like immiscible liquids with a given surface tension and that differences in surface tension between tissues determine their spatial configuration during embryogenesis. To understand how single cell biophysical and motility properties regulate tissue surface tension and how tissue surface tension controls tissue organization in development, we are studying the specific function of germ layer progenitor cell adhesion, cell cortex tension and motility in determining germ layer organization during zebrafish gastrulation. We found that the combinatorial activity of progenitor cell adhesion, cortex tension and motility determines germ layer tissue surface tension and that differences in germ layer tissue surface tension influence germ layer organization during gastrulation. We will discuss these findings in the light of different hypotheses explaining how single cell biophysical properties determine tissue morphogenesis in development.

160

Dynamic interactions between PAR proteins modulate the cycling of actomyosin networks during Drosophila apical constriction.

D. J. David1, T. J. Harris1; 1Cell & Systems Biology, University of Toronto, Toronto, ON, Canada Apical constriction is a major mechanism underlying morphogenesis. Previously thought to be mediated by contractile actomyosin belts anchored to cell junctions, recent studies including ours have shown that many cell types apically constrict via cyclical pulses of dynamic apical actomyosin networks. What governs the cycling of these networks in vivo remains unclear. Using Drosophila amnioserosa cells during dorsal closure as a model to study apical constriction, we previously found that amnioserosa contractility is based on cyclic assembly and disassembly of apical actomyosin networks. We also found that the PAR complex, composed of the apical polarity regulators Bazooka (Baz, Drosophila PAR-3), PAR-6, and atypical protein kinase C (aPKC) regulate distinct phases of the actomyosin cycles: Baz enhances network duration whereas PAR-6/aPKC promote the lull time between pulses. We hypothesized that Baz may enhance actomyosin network durations by inhibiting PAR-6/aPKC and that dynamic interactions between PAR proteins may be coupled to the cycling of the actomyosin networks. A later change in the actomyosin networks allowed us to test this hypothesized mechanism. During later dorsal closure, the actomyosin networks have decreased cycling and increased persistence; a change that occurs concomitantly with a shift in PAR proteins away from cell junctions and towards the apical surface. This shift appears to involve a rebalancing of mechanisms influencing junctional and apical surface PAR localization, with the latter being more closely linked to the apical actomyosin networks. To test whether increasing apical surface Baz attenuates aPKC-mediated inhibition of the actomyosin networks, we overexpressed mutant forms of Baz which either abolish or mimic its aPKC phosphorylation site (S980A and S980E, respectively). BazS980A is known to have increased affinity for aPKC whereas BazS980E has decreased affinity for aPKC. Overexpressed BazS980A precociously localized to the apical surface, stabilized and colocalized with apical surface aPKC, and promoted precocious apical constriction. In contrast, BazS980E had no such effects despite higher expression levels. We propose that the normal apical shift of the PAR proteins leads to increased Baz-aPKC interaction and to increased competitive inhibition of aPKC kinase activity towards its other substrates. This increased Baz-aPKC interaction may relieve aPKC-mediated inhibition of the actomyosin cytoskeleton leading to reduced cycling, more persistent networks, and greater apical constriction.

161

Cellular mechanisms and evolution of morphogenesis in Volvox and related algae. I. Nishii1; 1Temasek Life Sciences Laboratory, Singapore, Singapore

Most metazoa first build a hollow sphere through successive cell cleavages and then this spherical multicellular sheet repeatedly folds - invaginate and protrude - to form unique shapes of organs and individual. How did such supercellular morphogenetic processes arise during the evolution of multicellular organisms from their unicellular ancestors? Volvox and their close relatives serve as a good model for investigating such issue. This group ranges from unicellular Chlamydomonas, through several related species increasing cell numbers, to Volvox with thousands of cells. The key morphogenetic process in Volvox development is called inversion, in which the spherical embryo turns inside out. The pre-inversion embryo is a cellular monolayer in which neighboring cells are linked to one another at the mid-level of their cell bodies by cytoplasmic bridges. The single place where such linkages are missing is across the opening where the inversion movement will start. In onset, the cells become flask shaped by extending long, thin stalks from their outermost ends, showing analogy with a characteristic cell shape seen in metazoan morphogenesis such as gastrulation. Next, cells near the opening migrate relative to their cytoplasmic bridges until they are linked to their neighbors only at the outermost tips of their stalks, which causes the cell sheet to turn outward. Transposon-tagged inversionless mutants were isolated and five different loci were characterized. Intriguingly, these genes were involved in cell shape changes and cell migration in inversion described above but also related to ECM. invD and invE genes, which encoded a novel phospho-protein and a MAP kinase, respectively, were required for cell elongation. Phosphorylated InvE proteins were abundant before inversion but they were rapidly dephosphorylated after inversion. In addition, phosphorylation of InvD was dependent on InvE. Next, invA encoded a kinesin motor protein that was localized near the cytoplasmic bridges (Nishii et al. 2003) and required for the cell migration. Furthermore, invB and invC encoded a sugar nucleotide transporter and a glycosyl transferase, respectively, both have roles in processing of glycoproteins involved in the size control of ECM surrounding embryo (Ueki & Nishii 2008, 2009). Finally, regarding evolution, all five homologous genes in C. reinhardii were found. Intriguingly, invD and its ortholog were quite divergent in sequence compared to other pairs, suggesting that the evolution of InvD from an ancestral protein might have been an important role for elongated flask-shaped cell formation. We also discuss tracing such cellular basis in inversion and localization of InvA in related species from evolutionary aspects.

162

Titration of histones sets the DNA threshold for activating transcription at the mid- blastula transition in Xenopus.

A. Amodeo1, A. Straight2, J. Skotheim1; 1Biology, Stanford University, Stanford, CA, 2_{Biochemistry, Stanford University, Stanford, CA}

In many species, the early post-fertilization divisions occur rapidly and synchronously without growth phases, cell cycle checkpoints, or transcription. These early embryos are almost entirely transcriptionally inactive and rely on maternally supplied RNAs. At the Mid-blastula Transition (MBT), embryos initiate large-scale transcription from the zygotic genome (also known as the Maternal to Zygotic Transition or MZT), add growth phases, and gain checkpoints. Previous work suggested that the MBT is initiated by the increased DNA-to-cytoplasmic ratio resulting from repeated rounds of DNA replication and cell division without cell growth. This led to the hypothesis that the progressive titration of an inhibitory factor present in the embryo allows the initiation of zygotic transcription. However, the molecular identity of the titratable factor remained unknown. To determine the inhibitory molecule, and thus prove the titratable factor hypothesis,

we have developed a cell-free system in which to study the MBT. We found that Xenopus egg extracts initiate transcription at threshold DNA concentrations similar to those present at the MBT in intact embryos. Using this cell free system we have biochemically isolated histones H3 and/or H4 as the proteins responsible for inhibiting transcription in response to the DNA-to- cytoplasmic ratio. We have demonstrated that exogenous, bacterially produced, H3/H4 tetramers can increase the DNA threshold for transcriptional onset in a dose dependant manner. The identification of an abundant, high affinity DNA binding molecule that can inhibit transcription in a dose dependant manner strongly supports the titration model for MBT initiation. In addition, these findings suggest a molecular mechanism in which titrating out histones progressively opens chromatin structure to allow transcription.

163

Asymmetric inheritance of primary ciliary membrane in dividing neural progenitors. J. T. Paridaen1, M. Wilsch-Bräuninger1, W. B. Huttner1; 1Max Planck Institute of Molecular Cell Biology and Genetics, Dresden, Germany

The primary cilium is a microtubule-based protrusion of the cell surface that is involved in mediating extracellular signals, such as Sonic Hedgehog (Shh). The primary cilium is nucleated by the centrosome containing the eldest centriole (mother centriole). Prior to mitosis, the primary cilium is assumed to be disassembled in order for the centrosomes to be free to serve as the poles of the mitotic spindle.

Recent findings show that asymmetric inheritance of the centrosome containing the mother centriole is related to cell fate decisions during mammalian neurogenesis. However, it is unknown whether centrosome-associated structures such as the primary cilium are involved in this asymmetrical regulation of cell fate. Therefore, we set out to investigate the fate of the primary cilium upon cell division of developing neural progenitors and the possible involvement of the primary cilium in subsequent cell fate decisions in the developing mouse neocortex. We find that in mitotic neural progenitors and cultured cell lines, a membrane structure containing the ciliary small GTPase Arl13b is associated with the centrosome that contains the mother centriole. By surface biotinylation of the apical membrane in explanted mouse telencephalic hemispheres, we show that this structure contains ciliary membrane and is derived from the primary cilium that was present at the cell surface prior to mitosis. Using live imaging, we find that upon completion of mitosis, this centrosome-associated ciliary membrane is asymmetrically inherited by one of the daughter cells. Furthermore, we show that inheritance of the ciliary membrane causes earlier reestablishment of the primary cilium on the cell surface after division. Interestingly, the association of the ciliary membrane with one centrosome in mitotic neural progenitors decreases as neurogenesis proceeds.

We hypothesize that the observed asymmetry in primary cilium reassembly after mitosis might differentially expose daughter cells to extracellular signals, such as Shh. Therefore, we speculate that inheritance of the ciliary membrane affects cell fate decisions during mammalian neurogenesis.

164

Contact-mediated long distance signaling by Drosophila cytonemes.

S. Roy1, T. B. Kornberg1; 1Cardiovascular Research Institute, University of California, San Francisco, CA

How cells communicate with each other at long distances is key to understanding how cells cooperate to form organized tissues during development and why cells in various disease states lose or escape normal controls. Although much progress has been made identifying signaling molecules that mediate these communications – proteins such as Hedgehog, Wingless,

Decapentaplegic (Dpp, a BMP homolog), Fibroblast Growth Factor and Epidermal Growth Factor – the mechanism by which these proteins move with target specificity and in regulated amount through and across tissues remains unproven. Several proposed models postulate that some form of diffusion moves these signaling proteins through extracellular spaces. My work has investigated a different “direct delivery” mechanism whereby specialized filopodia (cytonemes) transfer signaling proteins between cells at sites of direct contact (Roy and Kornberg, Sci Signal. 2011; 4:pt8). Cytonemes are types of filopodia first identified in the Drosophila wing imaginal disc that were proposed to be involved in long distance signaling during development. My work shows that same group of cells emanate different types of cytonemes that can be distinguished by their specific response to different signaling ligands depending on the presence or absence of different protein receptors in them (Roy et al. Science. 2011; 332:354-358). I then show, using the GRASP GFP reconstitution method, that cytonemes from epithelial cells make direct contact with distal target cells, and also show that contacting cytonemes exchange, receive and transport ligands from source cells to recipient cells in receptor dependent manner. Finally, I show that genetic conditions that reduce cytonemes also reduce signal transduction. These findings establish that even non-neuronal cells can make direct long distance contacts for target specific and regulated signal exchange and strongly support the model of cytoneme-based movement of signaling proteins as a novel and essential mechanism for cell-cell communication.