C 2014 Wiley Periodicals, Inc. V

genesis 52:387–398 (2014)

RESEARCH ARTICLE

Centralspindlin is Required for Thorax Development During Drosophila Metamorphosis Michael Sfregola* Department of Molecular, Cellular and Developmental Biology, University of Colorado, Boulder, Colorado Received 3 October 2013; Revised 26 March 2014; Accepted 31 March 2014

Summary: Epithelial morphogenesis is an essential process in all metazoans during both normal development and pathological processes such as wound healing. The coordinated regulation of cell shape, cell size, and cell adhesion during the migration of epithelial sheets ultimately gives rise to the diversity of body plans among different organisms as well as the diversity of cellular structures and tissues within an organism. Metamorphosis of the Drosophila pupa is an excellent system to study these transformative events. During pupal development, the cells of the wing imaginal discs migrate dorsally and fuse to form the adult thorax. Here I show centralspindlin, a protein complex well known for its role in cytokinesis, is essential for migration of wing disc cells and proper thorax closure. I show the subcellular localization of centralspindlin is important for its function in thorax development. This study demonstrates the emerging role of centralspindlin in regulating cell migration and cell adhesion in addition to its previously known function during cytokinesis. genesis C 2014 Wiley Periodicals, Inc. 52:387–398, 2014. V Key words: morphogenesis; thorax closure; epithelium; cytokinesis; Tumbleweed; Pavarotti

INTRODUCTION Epithelial morphogenesis allows the transformation of cells and tissues into adult organisms with complex and diverse body plans and is fundamental to metazoan development. This transformative process requires coordinated cell migration, cell shape changes, and the formation and release of intercellular adhesive contacts. An understanding of these processes has been derived from studying the morphogenetic events in the model organism Drosophila melanogaster where the transformation from a syncytial embryo to an adult fly requires

numerous separate but coordinated morphogenetic movements. Dorsal closure (DC) during Drosophila embryogenesis has provided an excellent model to study these dynamic tissue movements (Kiehart et al., 2000). DC requires adhesive contacts between the migrating epithelia and the underlying substrate cells called the amnioserosa (Bloor and Kiehart, 2002; Gorfinkiel and Arias, 2007). Amnioserosa cells undergo pulsed contractions of the cytoskeleton gradually “pulling” the two leading edges of the overlying migrating epithelium together (Blanchard et al., 2010; David et al., 2010; Solon et al., 2009). In a similar process during Drosophila pupal development, the epithelial sheets of the wing imaginal discs migrate over the underlying amnioserosa and fuse at the dorsal midline to form the continuous epithelium of the adult thorax in a process called thorax closure (TC; Agne`s et al., 1999; Martın-Blanco et al., 2000; Usui and Simpson, 2000; Zeitlinger and Bohmann, 1999). The thorax migration and fusion process relies on many of the same molecular pathways and mechanisms as DC such as JNK and Dpp signaling (Agne`s et al., 1999; Martın-Blanco et al., 2000; Zeitlinger and Bohmann, 1999). Additionally, both DC and TC require the reorganization of the cytoskeleton and plasma membrane through the regulation of Rho family small GTPases, including Rho and Rac (Blanchard et al., 2010; Bloor Additional Supporting Information may be found in the online version of this article. * Correspondence to: Michael Sfregola, Department of Molecular, Cellular and Developmental Biology, University of Colorado, Boulder, CO 80309 E-mail: [email protected] Contract grant sponsor: NINDS, Contract grant number: 5R03NS075 458 (to M.S.) Published online 2 April 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/dvg.22777

388

SFREGOLA

and Kiehart, 2002; Harden et al., 1999). Several upstream regulators of Rho family proteins have been implicated in epithelial morphogenesis. Pebble (Pbl), a RhoGEF, was discovered to be important for completion of both DC and TC by an RNAi screen (Jankovics et al., 2011; Mummery-Widmer et al., 2009). Pbl was originally recognized as having an essential function for cytokinesis during embryonic cell divisions. In this process, Pbl is required to activate Rho at the cell membrane to allow cleavage furrows to form (Dechant and Glotzer, 2003). Pbl is recruited to the cell membrane by a complex called “centralspindlin” (Yuce et al., 2005). Centralspindlin is composed of two proteins: Tumbleweed (Tum) and Pavarotti (Pav; Mishima et al., 2002). Pav (MKLP-1 in humans) has microtubule motor activity allowing transport of the centralspindlin complex to the cell cortex (Somers and Saint, 2003; Zavortink et al., 2005). In addition to recruiting Pbl, centralspindlin also inhibits Rac through the GAP (GTPase activating protein) activity of Tum (MgcRacGAP in humans; Bastos et al., 2012; Canman et al., 2008; D’Avino et al., 2004; Ratheesh et al., 2012). These dual functions of centralspindlin allow coordinated regulation of both Rho and Rac activity. Similar to Pbl, Tum was also identified in a genomewide RNAi screen for genes required for proper patterning of sensory bristles in the pupal thorax (MummeryWidmer et al., 2009). These studies demonstrate the importance of Pbl and Tum in epithelial morphogenesis but it is still unclear what function they serve during TC. Because of their well-established roles in cytokinesis, developmental defects associated with Tum or Pbl depletion have been proposed to be secondary consequences of cytokinesis failure. This view was fueled further by the finding that Tum, Pav, and Pbl proteins contain nuclear localization sequences and are predominantly nuclear during interphase (Jones et al., 2010; Minestrini et al., 2002; Saito et al., 2003). In fact, nuclear sequestration of Pbl was thought to be essential to prevent Rho activation and ectopic furrow formation during interphase (Kamijo et al., 2006; Saito et al., 2003). However, the view that the function of centralspindlin and Pbl in Drosophila development is due solely to their role in cytokinesis is challenged by two recent lines of evidence. First, a cytoplasmic pool of Pbl and centralspindlin has been detected in interphase cells, which could allow a direct role in epithelial morphogenesis that is distinct from a role in cell division. Centralspindlin and Pbl can localize to cell-cell junctions in nondividing cells (Murray et al., 2012; Ratheesh et al., 2012) as well as axons and dendrites of neuronal cells in addition to their predominantly nuclear localization (Goldstein et al., 2005). Second, in addition to their role in cytokinesis, centralspindlin and Pbl have recently been implicated as important regulators of Rho signal-

ing at the plasma membrane of interphase cells (Ratheesh et al., 2012). Indeed, Pbl and Tum interact with E-cadherin and alpha-catenin (Ratheesh et al., 2012), two proteins essential for cell adhesion. Pbl and Tum also interact directly with lipids of the plasma membrane—Pbl through a PH domain (Murray et al., 2012) and Tum through a C1 domain (Lekomtsev et al., 2012). To understand the role of centralspindlin in epithelial morphogenesis, I studied the role of Tum and Pav, the subunits of the centralspindlin complex, during Drosophila thorax formation. I found that centralspindlin is required for proper closure of the pupal thorax and also for proper division and differentiation of sensory organs that populate the thoracic epithelium. I found that the regulation of centralspindlin’s nuclear versus cytoplasmic localization is important for thorax morphogenesis such that ectopic cytoplasmic sequestration of these proteins results in thoracic closure defects similar to what results from their depletion. These results support the model that cytoplasmic centralspindlin is active and that nuclear sequestration is important for its function in epithelial morphogenesis. RESULTS Centralspindlin is Required for TC To test the role of centralspindlin in TC, I used the pnr-GAL4 driver to express dsRNA constructs under the control of an upstream activating sequence (UAS) in the developing thorax. During larval development, pnrGAL4 is expressed in the migratory region of the wing imaginal disc (Heitzler et al., 1996). This consists of the peripodial stalk and the adjacent cells in the hinge region of the wing imaginal disc. Stalk cells form the leading edge of the migrating wing disc and pull the imaginal cells of the presumptive notum dorsally (Agne`s et al., 1999; Zeitlinger and Bohmann, 1999). This driver was also used in a genome-wide RNAi screen for regulators of sensory bristle formation in the pupal thorax because its expression is limited to the thoracic epithelium (Mummery-Widmer et al., 2009). Additionally this GAL4 insertion in the pnr gene region is known to cause mild thoracic cleft phenotypes (Heitzler et al., 1996; Mummery-Widmer et al., 2009; Ma et al., 2014). Knockdown of Tum using pnr-GAL4 to drive a UAStum-dsRNA construct led to a severe cleft thorax phenotype and a 95% reduction in the number of sensory bristles compared to either wild type or pnr-GAL4 alone (Fig. 1a–c). These phenotypes are similar to what was reported for tum-dsRNA in a genome-wide screen for regulators of sensory organ development (MummeryWidmer et al., 2009). These defects are dose dependent; doubling the copy number of the tum-dsRNA construct resulted in a severe thoracic cleft and partial

CENTRALSPINDLIN IN THORAX DEVELOPMENT

389

FIG. 1. Depletion of centralspindlin components Tum or Pav by RNAi results in bristle loss and TC defects. (a-f) UAS-dsRNA constructs were expressed using the pnr-GALl4 driver. (a) w1118. (b) Quantification of remaining bristles in the pnr-GAL4 expression domain after knockdown of Tum or Pav. Counts are normalized to the average bristle numbers for pnr-GAL4/1 (control) flies. (c) pnr-GAL4/UAS-tumdsRNA. (c0 ) pnr-GAL4, UAS-tum-dsRNA/UAS-tum-dsRNA. (d) pnr-GAL4/1; UAS-pav-dsRNA/1. (d0 ) UAS-pav-dsRNA/UAS-pav-dsRNA; pnr-GAL4/1. (e,f) Immunofluorescence images (e) wt and (f,f0 ) UAS-tum-dsRNA thoraxes dissected 24 h APF. Tissues were stained for DAPI, Elav (green, neuron cell marker), and Su(H) (red, socket cell marker) to highlight positioning of sensory organs.

lethality (Fig. 1c0 ). To ask if only Tum or the whole centralspindlin complex is required for proper thorax formation, I knocked down Pav and examined the consequences. I found that Pav depletion produced defects nearly identical to Tum depletion (Fig. 1c,d). The defects were also dose dependent (Fig. 1d0 ). Quantitation of these phenotypes is reported in Supporting Information Table S1. To characterize the cellular basis for thoracic defects I dissected pupae 24 h after pupal formation and examined gross thorax morphology as well as sensory organ lineages. In wild-type animals, the migrating sheets of wing disc cells have fused by this time and show regularly spaced vertical rows of differentiated sensory organs (stained for Elav, a neural cell marker and Su(H), a socket cell marker, Fig. 1e). In the thorax of Tum

depleted animals, the leading edge of migrating wing discs appear to be arrested midmigration and have failed to fuse at the dorsal midline. The resulting gaps are visible as areas devoid of differentiating sensory organs (Fig. 1f,f0 ). Similar defects were seen after Pav depletion (not shown). I conclude that depletion of either component of the centralspindlin complex causes a failure in TC. Next I asked if the GAP activity of Tum is required for thorax development. Expression of a Tum mutant transgene whose GAP activity has been disrupted (TumDGAP) resulted in thorax and bristle defects (Fig. 2b). Additionally I found expression of a Tum-MYC transgene in the background of tum-dsRNA is able to completely rescue TC defects and partially restore wildtype bristle number (Fig. 2c,e). However, expression of

390

SFREGOLA

FIG. 2. Tum mutant lacking GAP activity is unable to rescue thorax phenotypes in Tum depleted cells. (a–d) GAP deficient and wild-type Tum were expressed in a tum-dsRNA background and bristle numbers and thorax morphology were examined. All constructs were driven by pnr-GAL4. (a) UAS-tum-dsRNA. (b) UASTumDGAP. (c) UAS-tum-dsRNA/UAS-Tum-MYC. (d) UAS-tumdsRNA/UAS-TumDGAP. (e) Quantification of bristle numbers for the above conditions.

the TumDGAP transgene in the background of tumdsRNA is unable to rescue TC defects or restore bristle number (Fig. 2d,e). Together these experiments demonstrate Tum’s GAP activity is essential for thorax development and formation of sensory organs. Centralspindlin is Required for Proper Sensory Organ Formation A closer examination of the sensory organs through staining for the neuron cell marker Elav or the socket

cell marker Su(H) showed that knockdown of Tum or Pav resulted in cell division and differentiation defects. As described above, neurons and socket cells were absent or greatly reduced in the dorso-central region of the developing pupal notum (Fig. 1e,f). This is consistent with the failure to fully close the thoracic epithelium. Additionally, sensory organs in the remaining epithelia show misspecification of cell types. Sensory organ cells expressing dsRNA displayed missing or duplicated cell types (Fig. 3a). The number of neurons ranged from one to four in these sensory organs and often accompany missing socket and bristle cells. Sheath cell numbers also varied from zero to three. I did see evidence of cytokinesis failure as indicated by multinucleate cells in some differentiating sensory organs (Fig 3b, middle row). This result was expected considering centralspindlin’s essential role in cytokinesis. Cytokinesis failure, however, could not fully account for misspecification of cell types because some sensory organ clusters appeared to have successfully completed cytokinesis and yet showed multiple cells with neuronal fates (Fig. 3b, bottom row). Sensory organ differentiation relies on cell polarity and asymmetric division. Centralspindlin has been shown to be asymmetric during the division of Drosophila neuroblasts and may be involved in the regulation of cell polarity (Cabernard et al., 2010). To examine cell polarity during asymmetric division of pupal sensory organ cells, I followed the localization of Partner of Numb tagged with green fluorescent protein (Pon-GFP) in real time during mitosis. Pon-GFP is a known asymmetric determinant and forms a crescent at the anterior cortex of the dividing sensory organ precursor, or pI cell (Lu et al., 1998; this manuscript, Fig. 4a). As mitosis progresses Pon-GFP is asymmetrically segregated exclusively to one of the daughter cells, the pIIb cell (Fig. 4a). Upon depletion of Tum or Pav, Pon-GFP is still able to form an anterior crescent during the early stages of mitosis, indicating cell polarity remains intact despite the disruption of centralspindlin (Fig. 4b,c). However, in cells expressing tum-dsRNA Pon-GFP eventually equalizes between the pIIa and pIIb cells at the final stages of cytokinesis (Fig. 4a vs. 4b). In pav-dsRNA treated cells, Pon-GFP asymmetry is lost as cell division breaks down early in mitosis. Pon-GFP then becomes mislocalized around the entire cell cortex (Fig. 4c). This failure to segregate Pon-GFP to a single daughter cell in centralspindlin depleted cells may be an underlying cause of cell fate transformation. Misexpression/Localization of Centralspindlin Causes Thorax Defects Because depletion of centralspindlin leads to TC phenotypes, I investigated the consequences of overexpression of Tum and Pav. Pnr-GAL4 was used to drive the

CENTRALSPINDLIN IN THORAX DEVELOPMENT

391

FIG. 3. Depletion of Tum or Pav by RNAi leads to cell fate transformation and cytokinesis defects in sensory organ lineages. (a) Cell fate transformations of tum- and pav-dsRNA expressing sensory organs. Sensory organ clusters from 24 h APF pupal nota stained for Elav (green, neuron cell marker), Su(H) (red, socket cell marker, top row), and Prospero (red, sheath cell marker, bottom row). Both tum-dsRNA and pav-dsRNA result in multiple neurons and sheath cells compared to wild type. (b) Cytokinesis defects in tum-dsRNA treated sensory organs. Cells were stained for Elav (green) and spectrin (red, plasma membrane marker). The control (w1118, top row) shows a wild-type sensory organ. Tum-dsRNA treated cells (middle row) display cell division and cell fate defects demonstrated by a sensory organ containing two binucleate cells. Two nuclei stain positive for the neuron cell marker. Tum-dsRNA expressing cells (bottom row) can divide successfully but still display cell fate defects. The sensory organ consists of four mononucleate cells three of which stain positive for the neuron cell marker. All scale bars, 5 mm.

expression of previously described inducible Pav-GFP and Tum-MYC transgenes in the developing thorax (Goldstein et al., 2005; Minestrini et al., 2002). Published Pav-GFP transgenes were driven from either UASp (low expressing) or UAST (high expressing) regulatory regions (Minestrini et al., 2002, 2003). Expression of Pav-GFP from a low-expressing UASp promoter

did not result in noticeable defects in the adult notum. Pav-GFP was confined to the nucleus in interphase cells of the pupal thorax in these experiments (Fig. 5a). Expression of Pav-GFP from a high-expressing UAST promoter did produce TC defects (Fig. 5a). Under these conditions, Pav-GFP was found in both the nucleus and in the cytoplasm at the cell periphery (Fig. 5a, arrow).

392

SFREGOLA

FIG. 4. Pon-GFP localization during sensory organ cell division in wild-type and centralspindlin depleted flies. Pon-GFP localization was followed in live cells during asymmetric division of the sensory organ precursor also called the pI cell. (a) In wild-type cells, Pon-GFP is asymmetrically localized as the pI cell begin mitosis. Following division, Pon-GFP is segregated exclusively to the anterior daughter cell, or pIIb. (b) In tum-dsRNA cells Pon-GFP is segregated to the anterior cell normally. In the final stages of cytokinesis, Pon-GFP is detected in both the pIIb and pIIa daughter cells. (c) After Pav depletion Pon-GFP initially localizes to the anterior cell. However, as cytokinesis fails, Pon localization spreads along the cell cortex in the arrested cell.

Overexpression of Tum-MYC in the developing notum leads to a minor thoracic cleft defect (Fig. 5b). TumMYC was detected in the nucleus in interphase cells of the pupal thorax in these experiments (Fig. 5b). Expression of Tum with its nuclear localization signal (NLS) deleted (TumDnls-MYC, Jones et al., 2010) resulted in nearly complete lethality due to a failure to eclose from the pupal case. The few escapers that manage to develop into adults display severe thoracic clefts (Fig. 5b). Examination of TumDnls-MYC localization reveals it is largely excluded from the nucleus and localizes to the cell periphery (Fig. 5b). These results suggested an interesting hypothesis that defects due to overexpression of Pav and Tum occurred when the resulting protein was localized to the cytoplasm. To test this hypothesis I expressed Pav without the NLS (PavDnlsGFP, Minestrini et al., 2003) using the UAST promoter. This resulted in Pav localization at the cell periphery. Importantly, the resulting adult nota show TC defects (Fig. 5c). Previous experiments have shown that expression of one component of centralspindlin can influence the subcellular localization of the other. PavDnls-GFP is

constitutively cytoplasmic. However, expression of TumMYC with an intact NLS is able to restore wild-type localization of PavDnls-GFP by “pulling” it back into the nucleus (Goldstein et al., 2005). I took advantage of this characteristic to ask if thoracic defects caused by PavDnls-GFP could be rescued by sequestering the protein in the nucleus. Coexpression of Tum-MYC resulted in sequestration of PavDnls-GFP in the nucleus in the developing pupal thorax (Fig. 5c). Importantly, the resulting adult thorax shows rescue. In fact, Tum activity was dispensable for the rescue; Tum lacking a GAP domain was sufficient to sequester PavDnls-GFP in the nucleus and rescue the adult thorax phenotypes (Fig. 5c). Quantification of these rescue experiments is reported in Supporting Information Table S1. Altering Centralspindlin Expression Levels or Localization Disrupts Epithelial Organization and Integrity Examination of the TC phenotype caused by PavDnlsGFP expression reveals holes or tears in the developing thoracic epithelium that are absent in wild type. These

FIG. 5. Cytoplasmic accumulation of Tum or Pav results in TC failure. Tum and Pav transgenes were expressed with pnr-GAL4. (a) Overexpressed Pav-GFP localizes to cytoplasm and causes TC defects. Pav-GFP at low levels (top row) is sequestered in the nucleus and does not affect thorax morphology (wild type thorax is shown). Pav-GFP at high levels (bottom row) localizes to nuclei and cell-cell junctions causing TC defects. (b) Tum-MYC (top row) localizes primarily to nuclei and causes mild TC defects. TumDnls-MYC (bottom row, expressed with ptc-GAL4) is constitutively cytoplasmic and is found primarily at the cell periphery. pnr-GAL4 expressed TumDnls-MYC causes lethality and severe thoracic closure defects. (c) PavDnls-GFP is constitutively cytoplasmic and causes thorax morphology defects (top row). Removal of PavDnls-GFP from the cytoplasm with expression of Tum-MYC (middle row) or TumDGAP (bottom row) restores nuclear localization of PavDnls-GFP and restores wild-type thorax morphology.

394

SFREGOLA

FIG. 6. RNAi or overexpression of centralspindlin disrupts organization and morphology of the thorax epithelium. Tum and Pav transgenes or RNAi constructs were expressed with pnr-GAL4. (a) Thoraces expressing PavDnls-GFP (dashed lines) were stained for GFP (green) and DAPI (blue). Presence of PavDnls-GFP in the cytoplasm causes holes or tears (arrow and box) in the thoracic epithelium, indicating tissue integrity defects. Zooming in on the boxed area shows one of these holes contains no sensory organs (Elav, red) consistent with the sensory organ reduction seen in adults. (b) Control and UAS-tum-dsRNA thoraces were stained with spectrin (red) and DAPI (blue) to visualize cell morphology. The control sample shows regular hexagonally shaped cells while depletion of Tum results in misshaped cells and a disorganized epithelium. (c) Overexpression of Tum-MYC and Pav-GFP results in a severe thoracic cleft. (d) Tum-MYC and Pav-GFP overexpression disrupts epithelial integrity and causes formation of filopodia-like structures (arrow). Cells were stained for Rho (red) to visualize membrane structures and Tum (green) to examine centralspindlin localization.

holes vary in size and severity (Fig. 6a, arrow and inset). Similarly, depletion of centralspindlin components can disrupt the thorax epithelium. The wild-type thoracic epithelium consists of regularly spaced hexagonal cells with interspersed sensory organ clusters while Tum depletion results in disorganization and misshaped cells of irregular size (Fig. 6b). These cell shape differences are visualized by staining for spectrin, a protein known to localize to the plasma membrane, and indicate an underlying reorganization of the cytoskeleton. To further demonstrate the role of centralspindlin in the formation of the thorax epithelium, I overexpressed TumMYC and Pav-GFP and examined cell morphology. Simultaneous overexpression of both centralspindlin components results in a severe thoracic cleft phenotype (Fig. 6c). Consistent with this severe cleft phenotype I observe a severe disruption to the epithelium and localization of Tum-MYC and Pav-GFP (not shown) to the cytoplasm, specifically at filopodia-like membrane protrusions (Fig. 6d, arrow). These protrusions are also marked by Rho staining, a plasma membrane marker and centralspindlin interacting protein. DISCUSSION Previous studies have demonstrated the importance of centralspindlin and the recruitment of Pbl in the regulation of Rho family small GTPases during cytokinesis (Mishima et al., 2002; Somers and Saint, 2003). Epithelial morphogenesis requires the coordinated regulation

of cell adhesion and cell migration, two processes known to rely on Rho and Rac signaling at the plasma membrane (Bastos et al., 2012; McCormack et al., 2013; Yamada and Nelson, 2007). In the present study, I have shown centralspindlin is essential for the regulation of epithelial morphogenesis during thoracic development in addition to the division and differentiation of sensory organ cells. Cytoplasmic accumulation of centralspindlin or depletion of centralspindlin levels can result in similar thoracic cleft defects. This demonstrates the importance of regulating the levels of cytoplasmic centralspindlin and its activity during TC. Depleting either component of centralspindlin by dsRNA prevents proper wing disc migration and causes a failure of TC in a dose-dependent manner and results in a disorganized epithelium. These phenotypes are likely due to the failure of centralspindlin to recruit Pbl to the plasma membrane and activate Rho signaling. In addition to TC defects, depletion of centralspindlin caused cell fate transformations within sensory organ lineages. Interestingly, centralspindlin depleted cells maintain normal spindle organization and polarity. However, as cell division progresses depletion of Tum or Pav results in the delay or arrest of cell division and a disruption of cell polarity possibly contributing to the cell fate transformation phenotype. Timing of cell division is known to be important for sensory organ cell fate specification and could also contribute to the observed cell fate transformation phenotype (Remaud et al., 2008). Recently, a nearly identical cell fate and

CENTRALSPINDLIN IN THORAX DEVELOPMENT

cell division defect has been observed in sensory organs after mutation of another cytokinesis protein Diaphanous (formin), a protein involved in cytoskeletal regulation (Founounou et al., 2013). Together these results indicate cell fate transformation is likely a secondary consequence of defects or delays in cell division. In support of this model, overexpression of Tum or Pav does not adversely affect cell division and consequently does not lead to cell fate transformation. While depletion of centralspindlin during wing disc morphogenesis causes TC defects, so does overexpression of either centralspindlin component. The predominantly nuclear localization of centralspindlin during interphase raised the important question of how a nuclear protein complex could regulate epithelial migration. Difficulties in detecting native centralspindlin components outside of the nucleus contributed to the confusion. However, a recent study found native centralspindlin can localize to the plasma membrane during interphase where it physically interacts with adherens junction components (Ratheesh et al., 2012). This localization pattern is supported by my finding that Pav-GFP localizes to cell-cell junctions when overexpressed. This association with cell junctions is likely due to overwhelming the nuclear import of centralspindlin resulting in a significant amount that remains cytoplasmic. Cytoplasmic accumulation of Tum and Pav causes constitutive recruitment of centralspindlin and its associated proteins to the plasma membrane resulting in disruption to epithelial organization. Cells display filopodial extensions reminiscent of those seen during normal TC and may represent the arrest of this process at an early stage. The altered cell shape and plasma membrane structures indicate defects in epithelial migration and/or tissue integrity due to loss of cell adhesion. These defects are likely due to the misregulation of Rho and/or Rac signaling and ultimately lead to the failure of TC. Given the importance of cytoplasmic centralspindlin in regulation of cell morphology, it is not surprising that the cleft thorax phenotype is enhanced when Tum and Pav are constitutively sequestered in the cytoplasm through the deletion of the NLS’s. While the presence of either component in the cytoplasm results in thoracic defects similar to centralspindlin depletion, the TumDnls-MYC phenotype is much more severe. This difference in severity is easily explained by Pav’s lack of direct regulatory activity toward either Rac or Rho. Constitutive cytoplasmic localization of centralspindlin’s regulatory GAP activity (Tum) seems to be more detrimental than constitutive expression of its transport and localization module (Pav). The TC defects resulting from the presence of PavDnls-GFP in the cytoplasm can be rescued by restoring the protein’s ability to re-enter the nucleus. Expression of a Tum transgene with an intact NLS restores nuclear sequestration of Pav and wild type thorax devel-

395

opment. This is true even with the expression of a TumDGAP mutant and indicates that the rescue is facilitated by the nuclear sequestration of the centralspindlin complex rather than by restoring its activity. These results demonstrate the importance of regulating centralspindlin’s cytoplasmic activity by altering its subcellular localization during thorax development. Together the results indicate centralspindlin and Pbl are likely sequestered in the nucleus during interphase. As cells begin to migrate a subset of centralspindlin is released into the cytoplasm to regulate dynamic changes in cell adhesion and cell shape. Constitutive cytoplasmic localization on the other hand causes ectopic Rho activation and disrupts the dynamic regulation of the small GTPase. As cells prepare for mitosis they undergo a morphological change, detaching from neighboring cells and becoming round in shape. This drastic change in cell shape requires the concurrent reorganization of cortical actin/myosin and release of adhesive contacts from neighboring cells. The “rounding” process relies on the transport of Pbl and centralspindlin from the nucleus to the plasma membrane to activate Rho (Matthews et al., 2012). Similar to cell rounding, cytoplasmic centralspindlin is essential for regulating plasma membrane morphology during interphase as well. Migrating epithelial sheets require the coordinated dynamic regulation of cell shape and cell adhesion. Together these studies have demonstrated centralspindlin is a multifunctional protein complex coordinating various cellular events including cytokinesis, cell migration, and cell adhesion. MATERIALS AND METHODS Drosophila Stocks Used The following fly stocks were used: Pnr-GAL4 (Heitzler et al., 1996), ptc-GAL4 (Speicher et al., 1994), UASp-Pav-GFP (Minestrini et al., 2003), UAST-PavDnlsGFP (GFP-PavNLS(4–7)*, Minestrini et al., 2003), UASTPav-GFP (GFP-PavNLS5*, Minestrini et al., 2003), UASpav-dsRNA (VDRC, 46137, Dietzl et al., 2007), UASTum-MYC (Goldstein et al., 2005), UAS-TumDnls-MYC (Jones et al., 2010), UAS-TumDGAP (Sotillos and Campuzano, 2000), UAS-tum-dsRNA (Valium, TRiP.JF01639, Ni et al., 2007), and UAS-Pon-GFP (Lu et al., 1998). Flies were maintained at 27 C unless otherwise specified. Immunohistochemistry Pupal nota were dissected in PBS 24 h after pupal formation. Nota were fixed in PBST (PBS 1 0.3% Triton X100) 1 5% formaldehyde for 20 min. Tissues were then blocked in PBST 1 10% normal goat serum (Jackson ImmunoResearch) for 1 h then incubated with primary antibody overnight. The following primary antibodies

396

SFREGOLA

were used: anti-Elav, 7E8A10 (rat, 1:1000, DHSB), antiProspero, MR1A (mouse, 1:500, DHSB), anti-Su(H) (rabbit, 1:500, Santa Cruz), anti-Spectrin, 3A9 (mouse, 1:500, DHSB), anti-c-Myc (chicken, 1:500, Molecular Probes), anti-GFP, ab13970 (chicken, 1:500, Abcam), anti-Tumbleweed (rabbit, 1:200, Somers and Saint, 2003). For all immunostaining, the appropriate Alexafluor conjugated secondary antibody was used (Invitrogen). Bristle Defect and TC Quantification Pnr-GAL4 was used to drive the following UAS-dsRNA constructs in the dorso-central region of the adult notum: UAS-tum-dsRNA and UAS-pav-dsRNA. Total remaining microchaete contained within the pnr expression domain were counted for individual flies. n > 100 for all genotypes except for control (n 5 30). Average bristle counts for each genotype were normalized to the average bristle counts for pnr-GAL4/1 (control) flies. For rescue experiments, pnr-GAL4 was used to drive either UAS-tum-dsRNA or UAS-PavDnls-GFP in combination with the following constructs: UASTum-MYC and UAS-TumDGAP. To measure the penetrance of TC defect bristles were counted for individual flies. As bristle number was closely correlated to the presence of a thoracic cleft, flies with less than 40 remaining bristles (approximately half that of the pnrGAL4/1 control) were defined as having a TC defect. These data are presented in Supporting Information Table S1. Live Imaging of SOP Divisions Live cell imaging of pupal nota was carried out according to published protocols (Zitserman and Roegiers, 2011). Pnr-GAL4 was used to drive UAS-Pon-GFP. Image stacks were taken through the dividing SOP cells throughout the time course. Image Processing and Analysis All imaging was performed with an inverted fluorescence microscope (TE2000-U; Nikon) equipped with an electron-multiplying charge-coupled device camera (Cascade II; Photometrics) and a Yokogawa spinning disc confocal system (CSU-Xm2; Nikon). Images were acquired using Metamorph (version 7.0; MDS Analytical Technologies) and adjusted and merged using Photoshop (Adobe) or ImageJ (NIH). Scale bars were generated using ImageJ. Fixed pupal sensory organ images as well as live movie frames are maximum intensity zprojections of image stacks assembled in ImageJ. Adult nota were mounted in glycerol and imaged using an Olympus SZX12 microscope equipped with a SPOT Insight 2 CCD camera and SPOT imaging software. Images from multiple focal planes were combined using Helicon Focus software.

ACKNOWLEDGMENTS The author thanks A. Bejsovec, A. Carpenter, P. D’Avino, D. Glover, R. Saint, Y. Jan, the Bloomington Drosophila Stock Center (Indiana University), the Vienna Drosophila RNAi Center (VDRC), and the Developmental Studies Hybridoma Bank (University of Iowa) for fly stocks and/ or antibody reagents. He would like to thank M. Winey and T.T. Su for support, advice, and critical reading of the manuscript. LITERATURE CITED Agne`s F, Suzanne M, Noselli S. 1999. The Drosophila JNK pathway controls the morphogenesis of imaginal discs during metamorphosis. Development 126: 5453–5462. Bastos RN, Penate X, Bates M, Hammond D, Barr FA. 2012. CYK4 inhibits Rac1-dependent PAK1 and ARHGEF7 effector pathways during cytokinesis. J Cell Biol 198:865–880. Blanchard GB, Murugesu S, Adams RJ, Martinez-Arias A, Gorfinkiel N. 2010. Cytoskeletal dynamics and supracellular organisation of cell shape fluctuations during dorsal closure. Development 137:2743– 2752. Bloor JW, Kiehart DP. 2002. Drosophila RhoA regulates the cytoskeleton and cell-cell adhesion in the developing epidermis. Development 129:3173–3183. Cabernard C, Prehoda KE, Doe CQ. 2010. A spindleindependent cleavage furrow positioning pathway. Nature 467:91–94. Canman JC, Lewellyn L, Laband K, Smerdon SJ, Desai A, Bowerman B, Oegema K. 2008. Inhibition of Rac by the GAP activity of centralspindlin is essential for cytokinesis. Science 322:1543–1546. D’Avino PP, Savoian M, Glover DM. 2004. Mutations in sticky lead to defective organization of the contractile ring during cytokinesis and are enhanced by Rho and suppressed by Rac. J Cell Biol 166:61–71. David DJV, Tishkina A, Harris TJC. 2010. The PAR complex regulates pulsed actomyosin contractions during amnioserosa apical constriction in Drosophila. Development 137:1645–1655. Dechant R, Glotzer M. 2003. Centrosome separation and central spindle assembly act in redundant pathways that regulate microtubule density and trigger cleavage furrow formation. Dev Cell 4:333–344. Dietzl G, Chen D, Schnorrer F, Su K-C, Barinova Y, Fellner M, Gasser B, Kinsey K, Oppel S, Scheiblauer S, Couto A, Marra V, Keleman K, Dickson BJ. 2007. A genome-wide transgenic RNAi library for conditional gene inactivation in Drosophila. Nature 448: 151–156. Founounou N, Loyer N, Le Borgne R. 2013. Septins regulate the contractility of the actomyosin ring to

CENTRALSPINDLIN IN THORAX DEVELOPMENT

enable adherens junction remodeling during cytokinesis of epithelial cells. Dev Cell 24:242–255. Goldstein AY, Jan Y-N, Luo L. 2005. Function and regulation of Tumbleweed (RacGAP50C) in neuroblast proliferation and neuronal morphogenesis. Proc Natl Acad Sci USA 102:3834–3839. Gorfinkiel N, Arias AM. 2007. Requirements for adherens junction components in the interaction between epithelial tissues during dorsal closure in Drosophila. J Cell Sci 120:3289–3298. Harden N, Ricos M, Ong YM, Chia W, Lim L. 1999. Participation of small GTPases in dorsal closure of the Drosophila embryo: Distinct roles for Rho subfamily proteins in epithelial morphogenesis. J Cell Sci 112:273–284. Heitzler P, Haenlin M, Ramain P, Calleja M, Simpson P. 1996. A genetic analysis of pannier, a gene necessary for viability of dorsal tissues and bristle positioning in Drosophila. Genetics 143:1271.  , Vilmos P, Kiss N, Erdelyi Jankovics F, Henn L, Bujna A M. 2011. A functional genomic screen combined with time-lapse microscopy uncovers a novel set of genes involved in dorsal closure of Drosophila embryos. PLoS ONE 6:e22229. Jones WM, Chao AT, Zavortink M, Saint R, Bejsovec A. 2010. Cytokinesis proteins Tum and Pav have a nuclear role in Wnt regulation. J Cell Sci 123:2179– 2189. Kamijo K, Ohara N, Abe M, Uchimura T, Hosoya H, Lee J-S, Miki T. 2006. Dissecting the role of Rhomediated signaling in contractile ring formation. Mol Biol Cell 17:43–55. Kiehart DP, Galbraith CG, Edwards KA, Rickoll WL, Montague RA. 2000. Multiple forces contribute to cell sheet morphogenesis for dorsal closure in Drosophila. J Cell Biol 149:471–490. Lekomtsev S, Su K-C, Pye VE, Blight K, Sundaramoorthy S, Takaki T, Collinson LM, Cherepanov P, Divecha N, Petronczki M. 2012. Centralspindlin links the mitotic spindle to the plasma membrane during cytokinesis. Nature 492:276–279. Lu B, Rothenberg M, Jan LY, Jan YN. 1998. Partner of Numb colocalizes with Numb during mitosis and directs Numb asymmetric localization in Drosophila neural and muscle progenitors. Cell 95: 225–235. Ma X, Li W, Yu H, Yang Y, Li M, Xue L, Xu T. 2014. Bendless modulates JNK-mediated cell death and migration in Drosophila. Cell Death Differ 21:407–415. Martın-Blanco E, Pastor-Pareja JC, Garcıa-Bellido A. 2000. JNK and decapentaplegic signaling control adhesiveness and cytoskeleton dynamics during thorax closure in Drosophila. Proc Natl Acad Sci 97:7888–7893. Matthews HK, Delabre U, Rohn JL, Guck J, Kunda P, Baum B. 2012. Changes in Ect2 localization couple

397

actomyosin-dependent cell shape changes to mitotic progression. Dev Cell 23:371–383. McCormack J, Welsh NJ, Braga VMM. 2013. Cycling around cell-cell adhesion with Rho GTPase regulators. J Cell Sci 126:379–391. Minestrini G, Harley AS, Glover DM. 2003. Localization of Pavarotti-KLP in living Drosophila embryos suggests roles in reorganizing the cortical cytoskeleton during the mitotic cycle. Mol Biol Cell 14:4028– 4038. Minestrini G, Mathe E, Glover DM. 2002. Domains of the Pavarotti kinesin-like protein that direct its subcellular distribution: Effects of mislocalisation on the tubulin and actin cytoskeleton during Drosophila oogenesis. J Cell Sci 115:725–736. Mishima M, Kaitna S, Glotzer M. 2002. Central spindle assembly and cytokinesis require a kinesin-like protein/RhoGAP complex with microtubule bundling activity. Dev Cell 2:41–54. Mummery-Widmer JL, Yamazaki M, Stoeger T, Novatchkova M, Bhalerao S, Chen D, Dietzl G, Dickson BJ, Knoblich JA. 2009. Genome-wide analysis of notch signalling in Drosophila by transgenic RNAi. Nature 458:987–992. Murray MJ, Ng MM, Fraval H, Tan J, Liu W, Smallhorn M, Brill JA, Field SJ, Saint R. 2012. Regulation of Drosophila mesoderm migration by phosphoinositides and the PH domain of the Rho GTP exchange factor Pebble. Dev Biol 372:17–27. Ni J-Q, Markstein M, Binari R, Pfeiffer B, Liu L-P, Villalta C, Booker M, Perkins L, Perrimon N. 2007. Vector and parameters for targeted transgenic RNA interference in Drosophila melanogaster. Nat Methods 5: 49–51. Ratheesh A, Gomez GA, Priya R, Verma S, Kovacs EM, Jiang K, Brown NH, Akhmanova A, Stehbens SJ, Yap AS. 2012. Centralspindlin and a-catenin regulate Rho signalling at the epithelial zonula adherens. Nat Cell Biol 14:818–828. Remaud S, Audibert A, Gho M. 2008. S-phase favours notch cell responsiveness in the Drosophila bristle lineage. PloS One 3:e3646. Saito S, Liu X-F, Kamijo K, Raziudden R, Tatsumoto T, Okamoto I, Xiaoyan C, Lee C-C, Lorenzi M, Ohara N, Miki T. 2003. Deregulation and mislocalization of the cytokinesis regulator ECT2 activate the Rho signaling pathways leading to malignant transformation. J Biol Chem 279:7169–7179. Solon J, Kaya-C¸opur A, Colombelli J, Brunner D. 2009. Pulsed forces timed by a ratchet-like mechanism drive directed tissue movement during dorsal closure. Cell 137:1331–1342. Somers WG, Saint R. 2003. A RhoGEF and Rho family GTPase-activating protein complex links the contractile ring to cortical microtubules at the onset of cytokinesis. Dev Cell 4:29–39.

398

SFREGOLA

Sotillos S, Campuzano S. 2000. DRacGAP, a novel Drosophila gene, inhibits EGFR/Ras signalling in the developing imaginal wing disc. Development 127: 5427–5438. Speicher SA, Thomas U, Hinz U, Knust E. 1994. The serrate locus of Drosophila and its role in morphogenesis of the wing imaginal discs: Control of cell proliferation. Development 120:535–544. Usui K, Simpson P. 2000. Cellular basis of the dynamic behavior of the imaginal thoracic discs during Drosophila metamorphosis. Dev Biol 225:13–25. Yamada S, Nelson WJ. 2007. Localized zones of Rho and Rac activities drive initiation and expansion of epithelial cell cell adhesion. J Cell Biol 178:517–527.

Yuce O, Piekny A, Glotzer M. 2005. An ECT2centralspindlin complex regulates the localization and function of RhoA. J Cell Biol 170:571–582. Zavortink M, Contreras N, Addy T, Bejsovec A, Saint R. 2005. Tum/RacGAP50C provides a critical link between anaphase microtubules and the assembly of the contractile ring in Drosophila melanogaster. J Cell Sci 118:5381–5392. Zeitlinger J, Bohmann D. 1999. Thorax closure in Drosophila: Involvement of Fos and the JNK pathway. Development 126:3947–3956. Zitserman D, Roegiers F. 2011. Live-cell imaging of sensory organ precursor cells in intact Drosophila pupae. J Vis Exp 51:2706.