the conserved discs-large binding partner banderuola ... · in sop cells, the planar polarity...

The Conserved Discs-large

Current Biology 24, 1811–1825, August 18, 2014 ª2014 Elsevier Ltd All rights reserved http://dx.doi.org/10.1016/j.cub.2014.06.059

Article

Binding Partner Banderuola RegulatesAsymmetric Cell Division in Drosophila

Federico Mauri,1,2 Ilka Reichardt,1

Jennifer L. Mummery-Widmer,1 Masakazu Yamazaki,1,3

and Juergen A. Knoblich1,*1Institute of Molecular Biotechnology of the Austrian Academyof Sciences, Vienna 1030, Austria

Summary

Background: Asymmetric cell division (ACD) is a key processthat allows different cell types to be generated at preciselydefined times and positions. In Drosophila, neural precursorcells rely heavily on ACD to generate the different cell typesin the nervous system. A conserved protein machinery thatregulates ACD has been identified in Drosophila, but howthis machinery acts to allow the establishment of differentialcell fates is not entirely understood.Results: To identify additional proteins required for ACD, wehave carried out an in vivo live imaging RNAi screen for genesaffecting the asymmetric segregation of Numb in Drosophilasensory organ precursor cells. We identify Banderuola (Bnd),an essential regulator of cell polarization, spindle orientation,and asymmetric protein localization in Drosophila neural pre-cursor cells. Genetic and biochemical experiments show thatBnd acts together with the membrane-associated tumor sup-pressor Discs-large (Dlg) to establish antagonistic cortical do-mains during ACD. Inhibiting Bnd strongly enhances the dlgphenotype, causing massive brain tumors upon knockdownof both genes. Because the mammalian homologs of Bnd andDlg are interacting as well, Bnd function might be conservedin vertebrates, and it might also regulate cell polarity in higherorganisms.Conclusions:Bnd is a novel regulator of ACD in different typesof cells. Our data place Bnd at the top of the hierarchy of thefactors involved in ACD, suggesting that its main function isto mediate the localization and function of the Dlg tumor sup-pressor. Bnd has an antioncogenic function that is redundantwith Dlg, and the physical interaction between the two proteinsis conserved in evolution.

Introduction

Although most cell divisions are symmetric, some cells candivide asymmetrically into two daughter cells that assumedifferent fates [1–3]. During development, asymmetric cell divi-sion (ACD) allows specific cell types to be generated at preciselocations relative to surrounding tissues. To achieve this, theaxis of ACD needs to be coordinated with the architectureand polarity of the developing organism. Over the past years,a conserved protein machinery for ACD has been identified,but how this machinery connects to the organism architectureis less clear.

2Present address: IRIBHM, Universite Libre de Bruxelles, Brussels 1070,

Belgium3Present address: Research Center for Biosignal, Akita University, Akita

010-8543, Japan

*Correspondence: [email protected]

The fruit fly Drosophila melanogaster is one of the best-understoodmodel systems for ACD. In particular, the develop-ment of the Drosophila CNS and peripheral nervous systemrelies heavily on ACD and has contributed much to our currentunderstanding of this process. In the peripheral nervous sys-tem, external sensory (ES) organs are formed by two outercells (hair and socket) and two inner cells (neuron and sheath).The four cell types arise from a single sensory organ precursor(SOP) cell, which divides asymmetrically into an anterior pIIbcell and a posterior pIIa cell. In a second round of ACD, pIIaand pIIb generate the outer or inner cells of the ES organ,respectively [4]. The difference between pIIa and pIIb cellsarises from different levels of Notch signaling in the twodaughter cells. This difference is established by the asym-metric segregation of the Notch inhibitor Numb into the pIIbcell. Numb is known to regulate endocytosis, but how it inhibitsNotch signaling is not precisely understood [5, 6].In SOP cells, the polarity axis is coordinated with the ante-

rior-posterior planar polarity axis of the overlying epithelium[7, 8]. Planar polarity involves the localization ofmutually inhib-itory components of a well-characterized machinery to theanterior or posterior plasma membrane [8, 9]. In SOP cells,the planar polarity protein Strabismus (Stbm) localizes to theanterior cortex and initiates the reorganization of plasmamembrane domains to establish the axis of ACD [9]. One ofthe most upstream events of this process is the recruitmentof the membrane-associated guanylate kinase (MAGUK)Discs-large (Dlg) to the anterior cortex [10]. This may involvea direct interaction of Dlg with the planar polarity proteinStbm [11]. Dlg was originally identified as a tumor suppressorinvolved in the regulation of epithelial cell polarity [12, 13] andlater shown to play a role in ACD [14, 15] and synaptogenesis[16, 17]. Despite its widespread functions, the biochemicalpathways regulated by Dlg in those various cell types are notentirely understood.In SOP cells, Dlg associates with the adaptor protein Pins to

direct the protein Bazooka (Baz) to the basal-posterior side ofthe dividing SOP cell [10]. Together with Par-6 and aPKC, Bazforms the so-called Par protein complex that plays a pivotalrole during ACD in many different cell types [18]. Eventually,the kinase aPKC phosphorylates Numb, mediating its releasefrom the posterior plasma membrane and thereby causing itsaccumulation to the anterior side [19–21].To ensure the asymmetric segregation of Numb to the ante-

rior pIIb cell, the mitotic spindle has to be oriented along thepolarity axis [22]. This function is mediated by Pins throughthe binding of the microtubule binding protein Mushroombody defect (Mud), which forms a cortical attachment site forastral microtubules, aligning the spindle into the correct orien-tation [23–25]. The binding to Pins requires the heterotrimericG protein Gai, which associates with Pins to mediate itsrecruitment to the anterior plasma membrane and switches itto an open conformation in which Pins can bind Mud [26, 27].The same protein machinery directs ACD in neuroblasts, the

stem cell-like progenitors of the Drosophila CNS. Neuroblastsdivide asymmetrically into self-renewing daughter neuroblastsand smaller ganglion mother cells (GMCs) that generate twodifferentiating neurons through a terminal symmetric division

http://dx.doi.org/10.1016/j.cub.2014.06.059


mailto:[email protected]

http://crossmark.crossref.org/dialog/?doi=10.1016/j.cub.2014.06.059&domain=pdf

Current Biology Vol 24 No 161812

[28]. The asymmetric segregation of the cell fate determinantsNumb, Prospero (Pros), and Brat into the GMC is required forproper differentiation. The asymmetric partitioning of Prosand Brat is mediated by the adaptor protein Miranda [29–32],and the asymmetric localization of both Miranda and Numbdepends on phosphorylation by aPKC [21, 33]. Mutations inany of the three segregating determinants lead to the genera-tion of excessive numbers of neuroblasts and ultimately causethe formation of lethal, transplantable brain tumors [34]. As inSOP cells, Pins, Dlg, and Baz are required for ACD in neuro-blasts, but they act in a characteristically different manner.First, neuroblast divisions are oriented along the apical-basalaxis and not the planar polarity axis. Second, Pins, Dlg, andBaz colocalize apically in neuroblasts while they occupy oppo-site domains in SOP cells [35]. In part, those differences can beexplained by the recruitment of the adaptor protein Inscute-able (Insc) in the apical complex [36]. Insc is not expressedin SOP cells, but in neuroblasts, it coordinates cortical polarityand spindle alignment by connecting Pins to Baz, ensuringthe correct segregation of cell fate determinants in the differ-entiating daughter cell [37–40]. In addition, Dlg has a neuro-blast-specific role in mediating spindle orientation, actingdownstream of Pins to align the spindle pole through theinteraction with the kinesin motor Khc-73 [41]. Pins, Dlg, andKhc-73 also regulate a pathway called ‘‘telophase rescue’’that corrects ACD defects during late mitotic stages [42].This pathway realigns cortical polarity along the spindle axisindependently of the Par complex through a Dlg cortical clus-tering mechanism to ensure that determinants eventuallysegregate asymmetrically and daughter cell fates are correctlyspecified. How Dlg performs those seemingly divergent rolesin SOPs and neuroblasts is currently unclear.

Because our knowledge about ACD is evidently incomplete,we performed several RNAi screens to identify additionalplayers required for the correct establishment of daughtercell fates [43–45]. Here, we used the results from one of thosescreens [43] to identify Banderuola (Bnd), a new key regulatorof ACD that acts both in neuroblasts and in SOP cells. Wedemonstrate that Baz, Pins, and Dlg are all mislocalized inbnd mutant SOP cells, placing Bnd at the top of the hierarchyfor ACD. In bnd mutant neuroblasts, the asymmetric segrega-tion of cell fate determinants is disrupted because aPKC andDlg fail to accumulate apically. Importantly, Bnd interactsphysically and genetically with Dlg, suggesting that it supportsDlg in performing its divergent functions in various cell types.Because Bnd is conserved in evolution, our data identify anew member of the universal machinery for ACD that mightdirect cell polarity in vertebrates as well.

Results

A Screen for GFP-Pon Localization Identifies banderuola

A genome-wide RNAi screen has identified 134 genes that arepotentially involved in ACD because they cause defects in theSOP lineage [43]. To identify genes regulating Numb localiza-tion, we used an enhanced green fluorescent protein (EGFP)fusion of theNumbbinding partner Pon expressed in SOP cellsand their descendants from the phyllopod promoter [43, 46].This construct allowed live imaging of asymmetric proteinsegregationwhile simultaneously expressing upstreamactiva-tion sequence (UAS)-RNAi lines using the pnr-Gal4 driver (Fig-ure 1A). Three RNAi lines caused defects in EGFP::Pon.LDlocalization and targeted genes not previously implicated inACD. TID38988 targets APC4 (CG32707), a subunit of the

anaphase promoting complex; TID31742 targets san (separa-tion anxiety,CG12352), an acetyltransferase required for sisterchromatid cohesion; and TID27759 targets CG45058, a genenot previously characterized in flies (Figure 1B). Becauseonly CG45058 RNAi causes defects in the establishment ofpolarity, whereas the other two lines appear to be requiredfor maintaining polarity (data not shown), we focused on thisline for further analysis.Lineage staining shows that RNAi of CG45058 causes the

formation of ES organs that contain two neurons and twosheath cells, indicating a cell fate transformation of pIIa intopIIb (Figure 1C). CG45058 encodes a 180 kDa protein contain-ing a low complexity region, two Ankyrin repeat-containingdomains (ANK), a fibronectin type III domain (FN3), and aC-ter-minal Ras association (RA) domain (Figure 1D). CG45058 isconserved throughout most eukaryotes, but not in fungi. Thepredicted human ortholog (InParanoid, [47]) is ANKFN1-201(Ensembl gene ID ENSG00000153930). The most conservedportion of the protein includes the ANK domains, the FN3domain and a large sequence following those domains thatwe defined as the ‘‘CG45058 Motif’’ (Figure 1D). The RAdomain is not conserved in mammalian orthologs. We namedthe gene banderuola (bnd; the Italian word for ‘‘weathercock,’’indicating that crescent formation is variable like the blowingwind) to describe the randomized crescent formation in themutants.

Bnd Is Required for Asymmetric Cell DivisionTo understand the role of Bnd in asymmetric cell division, weanalyzed and quantified the EGFP::Pon.LD localization de-fects. Before anaphase onset, EGFP::Pon.LD was completelysymmetric in 11% of the mitotic SOPs and weakly asymmetricin 55% of the mitotic SOPs. In 6%, EGFP::Pon.LD localizationwas severely delayed (Figures 1B and S1 available online), andonly 28% of the SOPs normally localized the protein to theanterior cell cortex at the onset of metaphase. Remarkably,most of these defects were corrected later in mitosis, and, inanaphase/telophase, 45% of the SOP cells showed no pheno-type. In 33% of those late mitotic cells, EGFP::Pon was asym-metric, but the crescent was not aligned with the mitoticspindle (Figure 1B, arrowhead), whereas 11% of those latemitotic cells showed split or nonanterior crescents (Figure S1).In only 11%, EGFP::Pon.LD remained symmetric throughoutmitosis and segregated equally into both daughter cells (n =30 for control; n = 18 for bnd RNAi). Thus, Bnd is required forinitiating asymmetric protein localization in prometaphasebut seems to be less important during later mitotic stages.To verify the RNAi phenotype, we generated a bnd loss-of-

function mutant by Flp-mediated recombination of twoP elements that carry flippase recognition target (FRT) sitesflanking the gene (Figure S2) [48]. The resulting recombinantswere expected to be complete loss-of-function allelesbecause the entire bnd coding region was removed. Homozy-gousmutants were pupal lethal, with only a few adult escapers(<1% of the progeny) that displayed severe locomotion de-fects and died within hours after eclosion (data not shown).In mutant SOP cells, Numb was not asymmetric or was onlyvery weakly asymmetric (27%, 7 of n = 26), or the Numb cres-cent was not aligned with the mitotic spindle (15.4%, 5 of n =26; Figures 2A–2D). Again, those defects were less obviousduring anaphase and telophase (data not shown). Live imagingof RFP::Pon.LD together with the mitotic spindle markerZeus::GFP [49] (Figures 2E and 2F) showed that the asym-metric accumulation of RFP::Pon.LD was generally delayed

A B

C D

Figure 1. Identification of CG45058 through a Live Imaging Assay

(A) Setup of the live imaging assay. The notum of living pupae was exposed and imaged at the confocal microscope (top). The Numb localization reporter

EGFP::Pon.LD was fused to the SOP-specific phyl promoter while expressing the RNAi lines in the notum with pnr-Gal4. This allowed us to score for Numb

localization defects in mitotic SOP cells upon gene knockdown (bottom, frame from a live movie).

(B) Identification of novel regulators of ACD through live imaging. Anterior is up. In the first row, images from a control movie showing EGFP::Pon.LD local-

ization to the anterior side are shown. In the second and third rows, images from APC4 and san RNAi movies showing defects in the maintenance of polarity

are shown: the asymmetric localization of the reporter is lost by anaphase onset. In the last row, images of CG45058 RNAi displaying defects in the estab-

lishment of polarity and its coordination with spindle alignment (arrowhead) are shown. Time is expressed in min:s and counted considering nuclear enve-

lope breakdown (NEBD) as reference point (t = 0:00).

Scale bars represent 5 mm. n R 18 mitotic SOP cells from 4–5 pupae for each RNAi line, recording mitosis in its full length.

(C) Lineage staining showing cell fate transformations caused by CG45058 knockdown (asterisk): only inner cells (two neuron and two sheath cells) are

detectable. Su(H) (green) marks the socket cell. Pros (blue) marks the sheath cell. Cut (red) marks all cell types.

(D) Structure and evolutionary conservation of CG45058. The CG45058 protein consists of two ankyrin repeat-containing domains (ANK), a FN3 domain, and

a RA domain. The conservation of the protein is shown as percentage of sequence identity and similarity, calculated with pairwise alignment among the

depicted species on the whole sequence and on the most conserved portion of the protein (ANK domains, FN3 domain, and CG45058 motif).

Banderuola Regulates Asymmetric Cell Division1813

and less striking, and the mitotic spindle displayed abnormalrotation movements during the alignment to the planar cellpolarity (PCP) axis (Figures 2E–2G). In particular, whereasthe crescent of PON was clearly detectable at prometaphase(t = 0:00) in wild-type SOP cells, in the bnd loss-of-function(LOF) background, it was either undetected or weak, and itbecame clear only at later time points (Figure 2H). In telophase,

however, the segregation of PON::RFP into the pIIb cell wasalmost always correct (data not shown). In the bnd heterozy-gous background, the spindle became aligned with the planarpolarity axis early in mitosis (3–4 min before the metaphase-anaphase transition) and displayed limited rotation move-ments (Figures 2E and 2G). In the bnd null background, howev-er, there were extensive rotation movements carrying on

D

E

F

G H

Figure 2. Numb Is Mislocalized, and Spindle Positioning Is Affected upon bnd LOF

(A–C0) Numb localization is affected upon bnd LOF. In bnd heterozygous flies (A), Numb (green) is normally segregated to the anterior side during SOP cells

division. Upon bnd LOF (B and C), about half of the cells display defects in Numb localization, either due to misalignment of the Numb crescent and the

spindle (yellow arrow in B0) or due to very weak or absent polarization of Numb (C). Sanpodo (Spdo, red) marks SOP cells, pH3 (white) marks mitotic cells,

Centrosomin staining (Cnn, green) shows the position of the spindle, and DAPI marks DNA. Anterior is left. The schematics to the right of the figure describe

the corresponding Numb localization phenotypes. Scale bars represent 5 mm.

(D) Summary of Numb localization phenotypes. Heterozygous bnd has been used as control (bnd D/+ n = 18; bnd D/D n = 26).

(E and F) Time-lapse imaging of Numb localization and spindle alignment in dividing SOP cells in the bnd deletion background. SOP cells were identified

based on morphology, position, timing of division, and asymmetric segregation of PON::RFP at cytokinesis, as done in [9]. The microtubule marker

Zeus::GFP is expressed ubiquitously and used to image themitotic spindle, whereas the expression of the Numb localization reporter RFP::Pon.LD is driven

in the notum by the pnr-Gal4 driver. Anterior is up. t = 0:00 corresponds to the formation of the mitotic spindle (prometaphase). Metaphase-anaphase tran-

sition takes place at t = 7:50 in the heterozygous and t = 8:30 in the homozygous mutant. In the heterozygous background (bnd D/+; E), RFP::Pon.LD asym-

metry is established already at prometaphase (t = 0:00), whereas in the bnd LOF background (bnd D/D; F), its accumulation is much weaker, and just at the

metaphase-anaphase transition (t = 8:30), the reporter is clearly segregated to the anterior side. Also, the spindle displays abnormal rotation movements

(arrows) that aremore intense and prolonged in the homozygousmutant, continuing throughoutmetaphase. Scale bars represent 5 mm. Time is expressed in

min:s.

(G) Quantification of the spindle rotationmovementsmeasured between prometaphase and anaphase. Each bar corresponds to the division of one SOPcell.

The extent of rotationmovements is expressed in degrees and quantified in a cumulative way over time. The blue part of the bar shows the angle value of the

early rotation movements that are happening between t = 0:00 and midmetaphase, which is defined as the midpoint between the formation of the mitotic

spindle and the metaphase-anaphase transition. The red part of the bar gives the value of late rotation movements, which are happening between midme-

taphase and the metaphase-anaphase transition. In the bnd D/D background, the spindle rotation movements are more intense, especially the later ones.

(H) Quantification of RFP::Pon.LD asymmetric accumulation over time. In the bnd homozygous mutants, PON crescent is formed later during mitosis

compared to the heterozygous counterpart. Each point corresponds to the division of one SOP cell (sample Homozygous 5 has been excluded because

it never formed a PON crescent).


throughout metaphase (Figures 2F and 2G). A similar pheno-type has already been described for other polarity proteinssuch as Dlg [10]. Thus, Bnd is required for both spindlepositioning and asymmetric protein localization in mitosis,indicating that it might be involved in setting up the polarityaxis.

Because the asymmetric localization of Numb in SOP cellsrequires the Par protein complex [10, 20, 50, 51], we tested

the localization of Baz and aPKC.While both proteins normallyaccumulated on the posterior side of the SOP cells in hetero-zygous control animals, either they were not asymmetricallylocalized, or their accumulation was not aligned with themitotic spindle in about half of the SOP cells in homozygousbnd mutant animals (Figures 3A–3F).Baz asymmetry is established by the Dlg/Pins complex that

connects the axis of ACD to planar polarity [10] and also


mediates the alignment of the mitotic spindle with the polarityaxis through the interaction with Gai and Mud [26, 27, 52]. Wetherefore assessed the localization of Pins, Gai, andMud uponloss of bnd. All three proteins accumulated asymmetrically atthe anterior cortex in heterozygous control animals (Figures3G, 3J, and 3M). In bnd null mutants, however, Pins, Gai, andMud showed localization defects similar to those observedfor the Par proteins and Numb (Figures 3H, 3K, and 3N). Cres-cents were weak or not aligned with the polarity axis in morethan half of the mutant cells (Figures 3I, 3L, and 3O). The ante-rior enrichment of Dlg [10] was also affected in a similar way(Figures 3P–3R). Because Dlg is the most upstream compo-nent for establishing polarity in SOP cells and because wedid not observe planar polarity defects [43] (Figure S3 anddata not shown), we conclude that Bnd acts between theplanar polarity machinery and Dlg to establish cortical polarityand align it with the mitotic spindle in SOP cells.

Bnd Localizes to the Centrosomes and the Cell Cortex

To further characterize the role of Bnd, we analyzed its subcel-lular localization in SOP cells by live-cell imaging and immu-nofluorescence. Multiple attempts to generate a specificantibody that reliably detects the endogenous protein failed.Therefore, we expressed a GFP-tagged version of the proteinin SOP cells together with Histone::RFP using the neuralized-Gal4 driver. Bnd overexpression does not affect polarity andspindle orientation in dividing SOP cells (data not shown).The GFP-tagged protein is functional because its expressionusing the insc-Gal4 driver can rescue the viability of bnd mu-tants (Figures S4A andS4B) and the loss-of-neuroblast pheno-type (see below) we observe in bndmutants (Figures S4C andS4D). During interphase, Bnd::GFP is localized diffusely in thecytoplasm and enriched at the apical cortex (Figure S5A;Movie S2). At the onset of mitosis, the protein accumulatesat centrosomes (Figure 4A; Figure S5B; Movies S1 and S2).During metaphase and anaphase, it is enriched on the mitoticspindle aswell, and, upon cytokinesis, it is found concentratedon the central spindle (Figures 4A and S5; Movies S1 and S2).In addition, the protein shows transient accumulation onvarious domains along the cell cortex throughout mitosis,although those domains are not always correlated with theknown anterior or posterior polarity proteins (Figure 4A; MovieS1). The Bnd localization patternwas confirmed by coexpress-ing a CFP-tagged version of the protein (Bnd::CFP) withZeus::GFP [49] using pnr-Gal4 (Figures 4B–4E). Bnd::CFPcolocalizes with the spindle microtubules and centrosomesduring mitosis (Figures 4B and 4C) and also recapitulates thetransient cortical enrichment observed with the GFP fusion(Figures 4B–4D). Centrosomal and cortical enrichment werealso confirmed by anti-GFP staining of SOP cells expressingBnd::GFP from the neur-Gal4 driver (Figure 4F). Thus, the sub-cellular localization of Bnd is consistentwith a role in establish-ing cortical polarity and orienting the mitotic spindle.

bnd Is Required for Self-Renewal and Asymmetric CellDivision in Neuroblasts

Because most regulators of ACD are conserved between SOPcells and neuroblasts, we analyzed the bndmutant phenotypein the CNS. Staining for the neuroblast markers Miranda andDeadpan and the neuronal marker Prospero revealed a strongloss of neuroblasts in bndmutant brains, which was becomingmore severe during larval development (Figures 5A, 5B, 5H,and 5I). Consistently, there was a strong decrease in thenumber of phospho-Histone H3 (pH3)-positive mitotic cells

(Figures 5D, 5E, and 5J) in bnd mutant larval brains, whereaswe did not detect any increase in the levels of activated proap-optotic Caspase 3 (Figures 5F and 5G). Because this is a fairlynonspecific defect that could have been caused by a secondsite mutation as well, we confirmed the specificity by rescuingthe phenotype with a bacmid carrying the bnd locus (Figures5C and 5H) andwith aUAS-bnd::GFP transgene under the con-trol of the insc-Gal4 driver (Figures S4C and S4D). Interest-ingly, the number of optic lobe neuroblasts was unaffected,and the overall structure of the brain was normal, althoughhomozygous mutant brains were slightly smaller on average.Thus, we conclude that the loss of bnd causes premature dif-ferentiation of larval neuroblasts.To test whether the loss of neuroblasts correlates with de-

fects in asymmetric cell division, we analyzed the establish-ment of cortical domains. The remaining neuroblasts in bndmutants were smaller than their wild-type counterparts. Thebasal localization of both the adaptor protein Miranda (Fig-ures 6A–6I) and the cell fate determinant Numb (Figures 6E,6F, and 6L) was deficient, and the apical localization of thePar complex member aPKC and of Inscuteable (Figures 6A–6D, 6J, and 6K) was also abnormal in bndmutant neuroblasts.In particular, aPKC did not form apical crescents and oftendisplayed a uniform cortical localization (Figure 6B). In addi-tion, the apical enrichment of Dlg that has been described inneuroblasts was also essentially absent (Figures 6G, 6H,and 6M). We observed similar defects for all the polarity pro-teins we analyzed, such as Pins, Baz, and Mud (Figure S6 anddata not shown). Similar to what we observed in SOP cells,the localization of the basal cell fate determinants was oftenrescued during later mitotic stages, even in the absence ofthe apical complex (Figures S6G and S6H), presumably bythe action of the so-called telophase rescue pathway [42].Notably, in neuroblasts, we did not observe any defect inthe coordination between the polarity axis and the mitoticspindle. Thus, Bnd is required to ensure neuroblast self-renewal, and it exerts its function in the early stages of polar-ity establishment in multiple types of asymmetrically dividingcells.

Banderuola Interacts with Discs-large

To define the possible mechanism of action of Bnd, we per-formed coimmunoprecipitation assays of FLAG-tagged Bndin transfected Drosophila S2 cells with proteins involved inthe establishment of cortical polarity and the alignment ofthe mitotic spindle. Among other proteins, we focused onDlg because dlg mutant alleles display similar polarity andspindle rotation phenotypes [10–12]. We could coimmunopre-cipitate Bnd::FLAG with Dlg::GFP in S2 cells (Figure 7A), andthe biochemical interaction between Bnd and Dlg wasconfirmed in vivo using Drosophila embryos that express aBnd::FLAG transgene under the control of a heat shock induc-ible promoter. In these embryos, Dlg coimmunoprecipitatedwith Bnd::FLAG (Figure 7B), and, similarly, Bnd::FLAGwas de-tected upon immunoprecipitation of Dlg (data not shown).Moreover, Dlg and Bnd showed transient colocalization individing neuroblasts and SOP cells (Figure S7), further sug-gesting that there is a functional interaction between the twoproteins. To test whether this interaction is conserved in evo-lution, we expressed themammalian homologs of Bnd and Dlg(ANKFN1 and Dlg1, respectively) in 293T cells. Indeed, immu-noprecipitation of GFP-tagged ANKFN1 coprecipitated MYC-tagged Dlg1 from those cells (Figure 7C). Bnd is therefore aninteraction partner of Dlg, and because this interaction is

C

F

I

L

O

R

Figure 3. Loss of bnd Affects Polarity Establishment and Spindle Alignment in SOP Cells

(A–F) The Par complex proteins aPKC and Baz aremislocalized in bnd LOF. In the bnd heterozygous control, aPKC (A, green) and Baz (D, green) accumulate

to the posterior side of dividing SOP cells. In the bnd null background, both aPKC (B, green) and Baz (E, green) display localization defects. In more than half

of mitotic SOP cells (summarized in C and F), crescent formation is weak or absent, the crescent is not aligned with the spindle poles (misaligned crescent),

or the crescent is correctly polarized, but the spindle is not aligned along the anterior-posterior axis (PCP defects).

(G–O) Proteins involved in the spindle alignment are mislocalized in bnd LOF. The localization of Pins (G and H, green; summarized in I), Gai (J and K, green;

summarized in L), and Mud (M and N, green; summarized in O) is abnormal in 60%–70% of the mitotic SOP cells in the bnd LOF background, while in the

heterozygous background they are correctly segregated to the anterior side. PCPdefects describe cases inwhich the protein is polarized to the correct side,

but the spindle is not aligned with the polarity of the tissue.

(legend continued on next page)



conserved in evolution, we assume that it is important for Bndfunction.

To verify the close functional connection between Dlg andBnd, we analyzed their genetic interactions. For this purpose,we expressed bnd and dlg RNAi lines in all neuroblasts usinginsc-Gal4 [29] (Figures 7D–7G). Knockdown of dlg alonedoes not cause an obvious change in neuroblast numbers (Fig-ures 7E and 7M), but it does induce defects in the basal segre-gation of Miranda, consistent with what has been reportedpreviously [14, 15, 53] (Figure 7I). Knockdown of bnd alonealso causes defects in Miranda localization but no changesin neuroblast number (Figures 7F, 7J, and 7M), presumablybecause RNAi results in a partial loss-of-function phenotype.Strikingly, however, the simultaneous knockdown of both dlgand bnd causes a massive overproliferation of neuroblaststhat is accompanied by increased mitotic activity (Figure 7G).To confirm this data, we performed genetic interaction exper-iments using dlgsw and dlgHF321 hypomorphic alleles [54, 55](Figures 7K and 7L and data not shown). Although dlgsw mu-tants per se did not display neuroblast overproliferation phe-notypes (Figures 7K and 7M), the combination with bnd LOFcaused the onset of massive brain tumors (Figure 7L). Thesame results were obtained with the dlgHF321 allele (data notshown). These data establish a strong genetic interactionbetween dlg and bnd, and this, together with their physicalinteraction, supports the model that the genes are functionallyconnected. In addition, these data suggest that Bnd mighthave additional functions that go beyond simply localizingDlg because the combined knockdown phenotype is evenstronger than the dlg null mutant phenotype. We concludethat Bnd acts together with Dlg to establish polarity in bothasymmetrically dividing neuroblasts and SOP cells. In addi-tion, our data establish that Bnd has antioncogenic functions,acting redundantly with Dlg to prevent the formation of braintumors in Drosophila.

Discussion

Our results establish Bnd as a new component of the machi-nery for asymmetric cell division. We show that bnd RNAi orloss-of-function mutations cause defects in the establishmentof polarity and the positioning of the mitotic spindle in mitoticSOP cells. We also demonstrate that bnd is required for ACDand continued self-renewal activity in Drosophila larval neuro-blasts. Because Bnd interacts both biochemically and geneti-cally with the tumor suppressor protein Dlg, we propose that itexerts its function during ACD by regulating the function ofDlg. Moreover, the spindle rotation phenotype we observedin mitotic SOP cells in bnd mutants is very similar to that ofdlgsw mutants [10], further strengthening the possibility thatthe two proteins are functionally connected. Because themammalian homologs of these two proteins also interact,this function might be conserved in higher organisms as well.

Banderuola Regulates Asymmetric Cell DivisionThe process of ACD involves the establishment of a polarityaxis, the orientation of the mitotic spindle, the polarizeddistribution of cell fate determinants, and, ultimately, the

(P–R) The enrichment of Dlg to the anterior cortex is disrupted. Although Dlg ant

is defective upon bnd LOF (Q, white; summarized in R).

Spdo (red) labels the SOP cells, pH3 (white) marks themitotic cells, Cnn staining

left. aPKC: n = 17 (bnd D/+) and n = 38 (bnd D/D); Baz: n = 18 (bnd D/+) and n = 1

and n = 24 (bnd D/D); Mud: n = 30 (bnd D/+) and n = 30 (bnd D/D); Dlg: n = 35

establishment of different daughter cell fates. In SOP cells,the axis of polarity is established when Dlg and Pins interactwith components of the planar polarity pathway to concentrateanteriorly [9]. Because Bnd binds to Dlg and is required forPins and Dlg localization, but not for planar polarity (Figures3 and S3), our data indicate that it acts at the very top of thishierarchy. Because Dlg is alsomislocalized in bndmutant neu-roblasts, the role of Bnd in this tissue appears to be similar.Nevertheless, because the defect in asymmetry establishmentis not completely penetrant, it is plausible that bnd function ispartially redundant. Alternatively, it might also be that the re-sidual protein derived from maternal contribution is sufficientto maintain, at least partially, the asymmetric partitioning ofdeterminants. Further experiments will be needed to addressthese issues and clarify the instructive role of Bnd in establish-ing cell asymmetry.How could Bnd perform its function on a molecular level?

Bnd::GFP localizes at the centrosomes, on the spindle, and,transiently, at the cell cortex. Because Bnd contains both An-kyrin repeats and an FN3 domain, it could mediate protein-protein interactions leading to the anterior localization of Dlgdownstream of the PCP pathway. The localization of Dlg andPins to the anterior side of dividing SOP cells is regulated byStrabismus (Stbm) and Dishevelled (Dsh) [9]. It is thoughtthat Dsh excludes Dlg/Pins from the posterior side, whereasStbm binds Dlg at the anterior cortex, promoting the associa-tion with Pins. This hypothesis is reinforced by the fact that Dlginteracts directly with the PDZ binding motif (PBM) of Stbm inDrosophila embryos [11]. However, Pins is localized to theanterior cortex in stbm mutant SOP cells expressing a Stbmprotein lacking the PBM domain [9]. Hence, the localizationof Dlg/Pins can be regulated independently of a direct bindingto Stbm. It is tempting to speculate that Bnd could be a medi-ator between the PCP pathway and the establishment of theasymmetry axis in mitotic SOP cells.Alternatively, however, Bnd could also affect the function of

Dlg and other cortical proteins through its RA domain. RAdomains mediate binding to small GTPases and regulate theiractivity. Small GTPases are involved in the modification of theactomyosin network, and the establishment of polarity is influ-enced by myosin activity and by the contractility of the acto-myosin mesh [2, 56, 57]. In particular, Cdc42, a small GTPaseof the Rho family, plays a central role in the establishment ofpolarity in a wide variety of biological contexts, including thelocalization of Par6/aPKC to the apical cortex of neuroblasts[58]. More recent data have also implicated small Ras-likeGTPases in regulating cortical polarity and spindle orientation.The Rap1/Rgl/Ral signaling network was shown to mediatethose events through the regulation of the PDZ domain proteinCanoe, which is a known binding partner of Pins [59, 60]. It isintriguing to hypothesize that Bnd could be part of a similarsignaling network impinging on Dlg. Because the RA domainof Banderuola is not conserved in higher organisms, however,we favor the first hypothesis that rests on the conserveddomains of the protein (ANK domains, FN3 domain, and Bndmotif). Hence, Bnd could act as an adaptor that mediates pro-tein-protein interactions and regulates the function of bindingpartners such as Dlg.

erior localization is not affected in the heterozygous background (P, white), it

(green) shows the alignment of the spindle, and DAPImarks DNA. Anterior is

6 (bnd D/D); Pins: n = 27 (bnd D/+) and n = 15 (bnd D/D); Gai: n = 27 (bnd D/+)

(bnd D/+) and n = 24 (bnd D/D). Scale bars represent 5 mm.

A

F

Figure 4. Bnd Localizes at the Centrosomes, at the Spindle, and at Cortical Domains during Mitosis

(A) Time-lapse confocal images of Bnd::GFP localization. Bnd::GFP (green) is initially diffuse in the cytoplasm, but at prophase, it gets enriched at the

centrosomes (arrowheads) before NEBD. At metaphase, it associates with the centrosomes and spindle microtubules (arrowheads). It is also transiently

enriched at cortical domains during mitosis (arrow). neur-Gal4 drives the simultaneous expression of Bnd::GFP and His::RFP (red). Time is expressed in

min:s and counted considering NEBD as reference point (t = 0:00).

(B–E0 0) Time-lapse confocal images showing the colocalization of Bnd::CFP (blue) and Zeus::GFP (green), a microtubule marker. RFP::Pon.LD (red) marks

Numb localization. Bnd localizes at the spindle frommetaphase to anaphase (B–D) and is enriched at the cleavage furrow at telophase (E). Bnd::CFP is also



H

I

J

Figure 5. banderuola LOF Affects Neuroblast Self-Renewal

(A–C0 0 0) Staining of larval brain hemispheres with the neuroblast markers Miranda (Mira, green) and Deadpan (Dpn, red). In the bnd LOF background, the

number of neuroblasts is reduced (B) compared to heterozygous condition (A). The number of neurons (Pros-positive cells, blue) is not affected. A genomic

rescue of the bnd locus is able to restore normal neuroblast numbers in bnd mutant flies, as shown by the increase in Mira/Dpn-positive neuroblasts (C).

(D–E0) In bnd LOF (E), there is a decrease in the number ofmitotic neuroblasts (pH3,magenta) in the central brain area (D and E, yellow dashed line) compared

to the heterozygous control (D). Note that the optic lobe neuroblasts are unaffected.

(F–G0) There is no increase of proapoptotic Caspase3 activation (red) upon bnd LOF (G) compared to the heterozygous control (F).

(H) Quantification of the number of neuroblasts in bnd heterozygous mutants and homozygous mutants and upon insertion in the bnd D/D background of a

transgene carrying the bnd locus. Error bars show SD.

(I) Quantification of neuroblast numbers in bnd heterozygous and homozygous mutants during larval development. Although the number of neuroblasts is

comparable until the II instar stage, it starts decreasing progressively during the III instar stage. Error bars show SD.

(J) Neuroblast mitotic index quantification. The percentage of mitotic neuroblasts is lower in bnd D/D larvae compared to bnd D/+ throughout all the larval

stages examined. Error bars show SD.

Scale bars represent 50 mm.


Why Bnd is also found at centrosomes and at the spindleis harder to explain. In fact, Bnd is the only known proteinapart from Mud that localizes to both the centrosome andthe cell cortex during ACD. It could help in promoting thealignment of the spindle through the interaction with thePins/Gai/Mud complex, but this cannot explain the entirephenotype because microtubules are not strictly requiredfor polarity establishment during ACD [37, 42, 61, 62].Although we could not detect a biochemical interaction

localized at the centrosomes and at the cortex (B–D).

(F) Bnd::GFP expressed using neur-Gal4 colocalizes with the centrosomal mar

cortical enrichment on the anterior side. Sanpodo (red) marks SOP cells.

Anterior is up in (A)–(E) and left in (F). Scale bars represent 5 mm. See also Figu

between Bnd and Pins, Gai, or Mud, this interaction couldbe transient, or it could depend on polymerized microtu-bules. It will be compelling to verify the localization of theendogenous protein because this would consolidate thedata derived from the protein overexpression experiments.Furthermore, this could allow us to unravel in detail the dy-namics of Bnd cortical localization and its alignment withthe SOP polarity axis, which could be concealed in overex-pression conditions.

ker g-tubulin (blue) in fixed tissue staining. It is also possible to see areas of

re S5 and Movies S1 and S2.

I

J

K

L

M

Figure 6. banderuola LOF Affects Polarity Establishment in Mitotic Neuroblasts

(A–B0 0) aPKC (green) and Miranda (red) do not localize asymmetrically in mitotic bnd D/D neuroblasts (B) compared to heterozygous ones (A). In particular,

aPKC is either weakly localized at the cortex or uniformly cortical (B and B0). Cnn staining (green) marks the position of the mitotic spindle, and pH3 (white)

marks the mitotic cells.




The Antioncogenic Function of banderuolaIn neuroblasts, Bnd is required for self-renewal and asym-metric protein segregation and has an antioncogenic functionthat is redundant with Dlg. In bndmutants, we observe defectsleading to neuroblast loss. The remaining neuroblasts aremisshapen, displaying abnormalities in the asymmetric pro-tein segregation and reduced mitotic activity. Although theFRT site remaining in the bnd mutants prevents us from ad-dressing this question through a clonal analysis, we favor thehypothesis that the phenotype is cell autonomous and is dueto premature differentiation of neuroblasts. Indeed, this isconsistent with the phenotype in the SOP lineage because ge-netic manipulations resulting in a pIIa to pIIb transformation(like Numb overexpression or Notch loss of function) oftencause neuroblasts to divide symmetrically into two differenti-ating daughter cells.

The localization of both the basal determinants and Dlg itselfare affected in bndmutants. Dlg is known to mediate the basallocalization of cell fate determinants inDrosophila neuroblasts[14, 15, 53]. The abnormal localization of aPKC in bnd mutantneuroblasts could also be explained as an effect of dlg LOFbecause aPKC localization is affected in dlg mutants [63].Thus, the various protein mislocalization phenotypes in bndmutant neuroblasts could be explained by a model in whichBnd exerts its function solely by localizing Dlg.

The tumor phenotypes, on the other hand, suggest that thetwo genes act in parallel. Overproliferation phenotypes areobserved only upon LOF of both genes, and bnd LOF en-hances the dlg RNAi phenotype. In fact, this type of geneticinteraction has been described forpins and lglbefore: whereaspins mutant neuroblasts underproliferate due to self-renewalfailure, pins lgl double mutants have a massive overprolifera-tion of neuroblasts due to an aberrant self-renewal programtriggered by aPKC [63]. A similar mechanism could underliethe overproliferation we observe upon double RNAi of bndand dlg. An alternative explanation for the double knockdownphenotype is provided by the additional role that Dlg has in thetelophase rescue pathway, which might be independent frombnd. This pathway is known to mediate the establishment ofPins/Gai cortical polarity, even in the absence of the Par com-plex, through a Dlg-dependent mechanism [42]. The pathwayis active in wild-type neuroblasts but becomes essential onlywhen components of the apical Par complex are missing[35]. It is possible that the telophase rescue pathway ensuresthe asymmetric segregation of cell fate determinants uponbndRNAi. When dlg is inhibited as well, however, this pathwaycould be compromised, resulting in overproliferation andtumor formation.

Possible Conservation of the Interaction with Dlg

Dlg has four mammalian homologs. Like the Drosophila pro-tein, they localize at the basolateral cortex in epithelia andhave been shown to regulate cell polarity in various cell types[64, 65]. During rat astrocyte migration, for example, Dlg1 is

(C–D0 0) Inscuteable (green) and Miranda (red) do not localize asymmetrically i

mitotic spindle, and pH3 (white) marks the mitotic cells.

(E–F0 0) Numb (green) and Miranda (red) are not correctly segregated to the basa

heterozygous (E). pH3 (white) marks the mitotic cells.

(G–H0 0) Upon bnd LOF, Dlg (green) is not enriched at the apical cortex of mitot

mised (H, red). No defects are visible in the heterozygous controls (G). pH3 sta

Scale bars of (A)–(H) represent 10 mm.

(I–M) Quantification of the polarity phenotypes displayed in (A)–(H). aPKC: n = 1

Numb: n = 57 (bnd D/+) and n = 20 (bnd D/D); Dlg: n = 54 (bnd D/+) and n = 42

required in association with APC for the polarization of themicrotubule cytoskeleton at the leading edge of the migratingcell [66, 67]. Dlg-mediated polarity can be also considered agatekeeper against tumor progression: Dlg1 is a target of on-coviral proteins and is often mislocalized or downregulatedin late-stage tumors, implicating a causal connection betweenDlg1 and cancer [68, 69]. As the interaction between Bnd andDlg is conserved, Banderuola could be an evolutionarilyconserved regulator of Dlg activity, and our studies may there-fore be relevant for a variety of biological processes in higherorganisms as well.

Experimental Procedures

Fly Strains

Flies were raised on standard Drosophila media at 25�C. The following fly

strains were used: pnr-Gal4 (MD237) (Bloomington Drosophila Stock Center

[BDSC]); pnr-Gal4, phyllopod>>EGFP::Pon.LD [43]; neur-Gal4, UAS-H2RFP

[21]; P{XP}07120 and P{RB}01794 (Exelixis); UAS-Dcr2; insc-Gal4, UAS-

CD8::GFP [44]; insc-Gal4 [29]; Zeus::GFP (BL6836); yw;; Zeus-GFP, pnr-

Gal4 mw+, UAS-mRFP1::Pon.LD mw+/TM3, Sb. RNAi lines were obtained

from the Vienna Drosophila RNAi Center (VDRC). The dlgsw and dlgHF321

mutant alleles were described in [54, 55] and were obtained from Y. Bel-

laiche (y dlgsw/Bascy) and from the BDSC (BL36278), respectively.

Antibodies and Immunohistochemistry

Primary antibodies used in this studywere as follows: mouse anti-cut (2B10,

1:500; Developmental Studies Hybridoma Bank [DSHB]), rat anti-Su(H)

(1:100) (made following [70]), mouse anti-prospero (1:10; DSHB MR1A),

rabbit anti-prospero (1:1,000) [71], mouse anti-phospho-histone H3 (Ser10)

(1:1,000; Cell Signaling), rabbit anti-Centrosomin (1:1,000; gift from T. Kauf-

man), guinea pig anti-Sanpodo (1:1,000) [72], rabbit anti-Numb (1:100) [38],

rabbit anti-aPKCz (1:500; sc-216; Santa Cruz), rabbit anti-Baz (1:200) [21],

rabbit anti-Gai (1:100) [26], rabbit anti-Mud (1:500) [73], mouse anti-Discs

Large (1:200 immunofluorescence/1:1,000 western blot; DSHB), rabbit

anti-Strabismus (1:200) [9], mouse anti-gtubulin (1:1,000; GTU-88; Abcam),

rabbit anti-Miranda (1:200) [29], mouse anti-Prospero (1:10; DSHB), guinea

pig anti-Dpn (1:500) [74], rabbit anti-cleaved Caspase 3 (1:500; 5A1E; Cell

Signaling), rabbit anti-GFP (1:2,000; Abcam), mouse anti-Flag (1:2,000;

Sigma M2), rabbit anti-Myc (1:2,000; ab9106; Abcam). Secondary anti-

bodies were conjugates of Alexa Fluor 488, Alexa Fluor 568, Alexa Fluor

647, and Alexa Fluor 405 (1:500; Invitrogen).

For immunofluorescence, pupaewere dissected in 8%paraformaldehyde

(PFA), fixed for 20–30 min on ice, and stained as described [75]. Pupae were

collected at 0 hr after puparium formation (APF) and aged for 15 hr at 25�C.For lineage analysis, pupae were collected at 0 hr APF and aged for 23–26 hr

at 29�C. For neuroblast stainings, third instar larvae were dissected in PBS,

fixed for 15 min in 5% PFA with 0.1% Triton X-100, and processed as

described [29]. Live imaging of SOP cells was performed essentially as pre-

viously described [8]. Images were acquired on a Zeiss LSM510 Meta

confocal microscope, and 3D movies and three-channel movies were ac-

quired using a LSM710 Spectral confocal microscope.

Immunoprecipitation

For immunoprecipitations, embryos or cells were lysed in RIPA buffer

(150 mM NaCl, 50 mM Tris-HCl [pH 8], 1 mM EDTA, 1% NP40, 0.5% Na-

deoxycholate) supplemented with 10 mg/ml PMSF and Complete Protease

Inhibitor Cocktail (Roche), cleared by centrifugation, and incubated with

antibodies for 1–2 hr at 4�C. The immunocomplexes were then precipitated

by using Protein G Sepharose 4B beads (Amersham) or GFP-Trap

n the bnd LOF background. Cnn staining (green) marks the position of the

l side in mitotic neuroblasts in the bnd LOF background (F) compared to the

ic neuroblasts, and the segregation of Miranda to the basal side is compro-

ining (red) marks mitotic cells.

17 (bnd D/+) and n = 50 (bnd D/D); Insc: n = 65 (bnd D/+) and n = 50 (bnd D/D);

(bnd D/D); Miranda: n = 164 (bnd D/+) and n = 51 (bnd D/D).

A B C

M

Figure 7. Bnd Interacts with Dlg, and the Two Proteins Are Functionally Connected

(A) Coimmunoprecipitation (coIP) of Bnd and Dlg in S2 cells. Bnd::Flag and Dlg::GFP were cotransfected in Drosophila S2 cells. Bnd::Flag coimmunopre-

cipitates with Dlg::GFP upon GFP immunoprecipitation.

(B) CoIP of Dlg and Bnd in Drosophila embryos. Bnd::Flag is overexpressed using a heat shock promoter, and endogenous Dlg coimmunoprecipitates with

Bnd::Flag in embryos.

(C) The interaction between Bnd and Dlg is evolutionarily conserved as rat Dlg1::MYC coimmunoprecipitates with GFP-tagged mouse Bnd (ANKFN1) when

overexpressed in 293T cells.

(D–G0 0) bnd and dlg are displaying additive phenotypes. The number of neuroblasts (Mira-positive cells, red) is unaffected upon dlg RNAi (E) or bnd RNAi (F)

compared to control (D), although the asymmetric partitioning of Miranda is defective (H–J). Double RNAi of bnd and dlg (G) causes massive neuroblast




(Chromotek) in the HEK293T cells experiments, washed three times with

lysis buffer, and eluted by boiling in SDS sample buffer.

Cell Culture

S2 cells were cultured at 27�C in Schneider’s Drosophila medium (Gibco),

with 10% scomplemented FCS, penicillin (50 U/ml), and streptomycin

(50 mg/ml). Cells were transfected with Cellfectin II reagent (Invitrogen),

according to the manufacturer’s protocol. UAS constructs were expressed

by cotransfection with actin-Gal4. HEK293T cells were cultured at 37�C in

DMEM, with 10% FBS, 1 mM L-Gln, penicillin (100 U/ml), and streptomycin

(100 mg/ml). Cells were transfected using Turbofect (Fermentas), according

to manufacturer’s protocol.

Constructs and Fly Transgenics

The banderuola coding region (starting at position 256 of CG45058-RB) was

amplified by PCR from embryonic cDNA (0–9 hr) and verified by DNA

sequencing. bnd expression constructs for cell culture and fly transgenesis

expression were generated by subcloning the bnd CDS into destination

vectors using Gateway technology (Invitrogen). For rescue experiments,

the CH322-118J18 clone was used (BACPAC Resources Center). Trans-

genic flies were derived using standard methodology. ANKFN1 CDS was

subcloned from ORFeome clone 100016073 (Source BioScience Life Sci-

ences) using the Gateway technology. Rat Dlg constructs were obtained

from S. Etienne-Manneville [76].

Supplemental Information

Supplemental Information includes seven figures and two movies and

can be found with this article online at http://dx.doi.org/10.1016/j.cub.

2014.06.059.

Acknowledgments

We wish to thank Elke Kleiner, Sara Farina Lopez, and Joseph Francis Gok-

cezade for technical assistance; all members of the J.A.K. laboratory for dis-

cussions; T. Kaufman, F. Schweisguth, Y. Bellaiche, S. Etienne-Manneville,

and Yuh Nung Yan for sharing reagents; Maria Novatchkova for performing

protein sequence alignment and conservation analysis; Marko Repic for

help with the HEK293T cells experiments; and Christoph Juschke, Spyros

Goulas, and Frederik Wirtz-Peitz for comments on the manuscript. Work

in the J.A.K. laboratory is supported by the Austrian Academy of Sciences,

the EU Seventh Framework Programme network EuroSyStem, the Austrian

Science Fund (grants I_552-B19 and Z_153_B09), and an advanced grant of

the European Research Council.

Received: September 12, 2012

Revised: March 8, 2014

Accepted: June 23, 2014

Published: July 31, 2014

References

1. Knoblich, J.A. (2008). Mechanisms of asymmetric stemcell division. Cell

132, 583–597.

2. Gonczy, P. (2008). Mechanisms of asymmetric cell division: flies and

worms pave the way. Nat. Rev. Mol. Cell Biol. 9, 355–366.

3. Doe, C.Q. (2008). Neural stem cells: balancing self-renewal with differ-

entiation. Development 135, 1575–1587.

4. Gho, M., Bellaıche, Y., and Schweisguth, F. (1999). Revisiting the

Drosophila microchaete lineage: a novel intrinsically asymmetric cell

division generates a glial cell. Development 126, 3573–3584.

overproliferation phenotypes and increased mitotic activity (pH3 staining, blue

CD8-GFP (green). The reduction of the levels of Dlg protein (blue) is evident up

(H–J0) Close-up images of dividing neuroblasts from panels (D)–(F). While Miran

neuroblasts (H), upon dlg (I) or bnd (J) RNAi, Miranda cortical enrichment and a

cells.

(K–L0 0) dlg and bnd are interacting genetically. The number of neuroblasts (Mi

affected (K), while the simultaneous loss of bnd causes the onset of massive b

(M) Quantification of the neuroblast phenotypes displayed in (D)–(L). Error bar

Scale bars of (D)–(G), (K), and (L) represent 50 mm. Scale bars of (H)–(J) repres

5. Coumailleau, F., Furthauer, M., Knoblich, J.A., and Gonzalez-Gaitan, M.

(2009). Directional Delta andNotch trafficking in Sara endosomes during

asymmetric cell division. Nature 458, 1051–1055.

6. Couturier, L., Vodovar, N., and Schweisguth, F. (2012). Endocytosis

by Numb breaks Notch symmetry at cytokinesis. Nat. Cell Biol. 14,

131–139.

7. Gho, M., and Schweisguth, F. (1998). Frizzled signalling controls orien-

tation of asymmetric sense organ precursor cell divisions in Drosophila.

Nature 393, 178–181.

8. Bellaıche, Y., Gho, M., Kaltschmidt, J.A., Brand, A.H., and Schweisguth,

F. (2001). Frizzled regulates localization of cell-fate determinants and

mitotic spindle rotation during asymmetric cell division. Nat. Cell Biol.

3, 50–57.

9. Bellaıche, Y., Beaudoin-Massiani, O., Stuttem, I., and Schweisguth, F.

(2004). The planar cell polarity protein Strabismus promotes Pins ante-

rior localization during asymmetric division of sensory organ precursor

cells in Drosophila. Development 131, 469–478.

10. Bellaıche, Y., Radovic, A., Woods, D.F., Hough, C.D., Parmentier, M.L.,

O’Kane, C.J., Bryant, P.J., and Schweisguth, F. (2001). The Partner of

Inscuteable/Discs-large complex is required to establish planar polarity

during asymmetric cell division in Drosophila. Cell 106, 355–366.

11. Lee, O.K., Frese, K.K., James, J.S., Chadda, D., Chen, Z.H., Javier, R.T.,

and Cho, K.O. (2003). Discs-Large and Strabismus are functionally

linked to plasma membrane formation. Nat. Cell Biol. 5, 987–993.

12. Woods, D.F., Hough, C., Peel, D., Callaini, G., and Bryant, P.J. (1996).

Dlg protein is required for junction structure, cell polarity, and prolifera-

tion control in Drosophila epithelia. J. Cell Biol. 134, 1469–1482.

13. Bilder, D., Li, M., and Perrimon, N. (2000). Cooperative regulation of cell

polarity and growth by Drosophila tumor suppressors. Science 289,

113–116.

14. Ohshiro,T.,Yagami,T.,Zhang,C., andMatsuzaki,F. (2000).Roleofcortical

tumour-suppressor proteins in asymmetric division of Drosophila neuro-

blast. Nature 408, 593–596.

15. Peng, C.Y., Manning, L., Albertson, R., and Doe, C.Q. (2000). The

tumour-suppressor genes lgl and dlg regulate basal protein targeting

in Drosophila neuroblasts. Nature 408, 596–600.

16. Thomas, U., Kim, E., Kuhlendahl, S., Koh, Y.H., Gundelfinger, E.D.,

Sheng, M., Garner, C.C., and Budnik, V. (1997). Synaptic clustering of

the cell adhesion molecule fasciclin II by discs-large and its role in the

regulation of presynaptic structure. Neuron 19, 787–799.

17. Tejedor, F.J., Bokhari, A., Rogero, O., Gorczyca, M., Zhang, J., Kim, E.,

Sheng, M., and Budnik, V. (1997). Essential role for dlg in synaptic clus-

tering of Shaker K+ channels in vivo. J. Neurosci. 17, 152–159.

18. Suzuki, A., and Ohno, S. (2006). The PAR-aPKC system: lessons in

polarity. J. Cell Sci. 119, 979–987.

19. Nishimura, T., and Kaibuchi, K. (2007). Numb controls integrin endocy-

tosis for directional cell migration with aPKC and PAR-3. Dev. Cell 13,

15–28.

20. Smith, C.A., Lau, K.M., Rahmani, Z., Dho, S.E., Brothers, G., She, Y.M.,

Berry, D.M., Bonneil, E., Thibault, P., Schweisguth, F., et al. (2007).

aPKC-mediated phosphorylation regulates asymmetric membrane

localization of the cell fate determinant Numb. EMBO J. 26, 468–480.

21. Wirtz-Peitz, F., Nishimura, T., and Knoblich, J.A. (2008). Linking cell cy-

cle to asymmetric division: Aurora-A phosphorylates the Par complex to

regulate Numb localization. Cell 135, 161–173.

22. Siller, K.H., and Doe, C.Q. (2009). Spindle orientation during asymmetric

cell division. Nat. Cell Biol. 11, 365–374.

23. Izumi, Y., Ohta, N., Hisata, K., Raabe, T., and Matsuzaki, F. (2006).

Drosophila Pins-binding proteinMud regulates spindle-polarity coupling

and centrosome organization. Nat. Cell Biol. 8, 586–593.

24. Bowman, S.K., Neumuller, R.A., Novatchkova, M., Du, Q., and Knoblich,

J.A. (2006). The Drosophila NuMA Homolog Mud regulates spindle

orientation in asymmetric cell division. Dev. Cell 10, 731–742.

). RNAi is driven by the neuroblast-specific driver insc-Gal4, expressing also

on RNAi (E and G).

da (red) gets asymmetrically localized to the basal side in dividing wild-type

symmetric partitioning are compromised. pH3 staining (blue) marks mitotic

ra-positive cells and Dpn-positive cells) in dlg homozygous mutants is not

rain tumors (L). pH3 (blue) marks the mitotic cells.

s show SD.

ent 10 mm.




25. Siller, K.H., Cabernard, C., and Doe, C.Q. (2006). The NuMA-relatedMud

protein binds Pins and regulates spindle orientation in Drosophila

neuroblasts. Nat. Cell Biol. 8, 594–600.

26. Schaefer, M., Petronczki, M., Dorner, D., Forte, M., and Knoblich, J.A.

(2001). Heterotrimeric G proteins direct two modes of asymmetric cell

division in the Drosophila nervous system. Cell 107, 183–194.

27. Nipper, R.W., Siller, K.H., Smith, N.R., Doe, C.Q., and Prehoda, K.E.

(2007). Galphai generates multiple Pins activation states to link cortical

polarity and spindle orientation in Drosophila neuroblasts. Proc. Natl.

Acad. Sci. USA 104, 14306–14311.

28. Ito, K., and Hotta, Y. (1992). Proliferation pattern of postembryonic

neuroblasts in the brain of Drosophila melanogaster. Dev. Biol. 149,

134–148.

29. Betschinger, J., Mechtler, K., and Knoblich, J.A. (2006). Asymmetric

segregation of the tumor suppressor brat regulates self-renewal in

Drosophila neural stem cells. Cell 124, 1241–1253.

30. Ikeshima-Kataoka, H., Skeath, J.B., Nabeshima, Y., Doe, C.Q., and

Matsuzaki, F. (1997). Miranda directs Prospero to a daughter cell during

Drosophila asymmetric divisions. Nature 390, 625–629.

31. Lee, C.Y., Wilkinson, B.D., Siegrist, S.E., Wharton, R.P., and Doe, C.Q.

(2006). Brat is a Miranda cargo protein that promotes neuronal differen-

tiation and inhibits neuroblast self-renewal. Dev. Cell 10, 441–449.

32. Shen, C.P., Jan, L.Y., and Jan, Y.N. (1997). Miranda is required for the

asymmetric localization of Prospero during mitosis in Drosophila. Cell

90, 449–458.

33. Atwood, S.X., and Prehoda, K.E. (2009). aPKC phosphorylates Miranda

to polarize fate determinants during neuroblast asymmetric cell divi-

sion. Curr. Biol. 19, 723–729.

34. Caussinus, E., and Gonzalez, C. (2005). Induction of tumor growth by

altered stem-cell asymmetric division in Drosophila melanogaster.

Nat. Genet. 37, 1125–1129.

35. Knoblich, J.A. (2010). Asymmetric cell division: recent developments

and their implications for tumour biology. Nat. Rev. Mol. Cell Biol. 11,

849–860.

36. Kraut, R., and Campos-Ortega, J.A. (1996). inscuteable, a neural precur-

sor gene of Drosophila, encodes a candidate for a cytoskeleton adaptor

protein. Dev. Biol. 174, 65–81.

37. Kraut, R., Chia,W., Jan, L.Y., Jan, Y.N., and Knoblich, J.A. (1996). Role of

inscuteable in orienting asymmetric cell divisions in Drosophila. Nature

383, 50–55.

38. Schober, M., Schaefer, M., and Knoblich, J.A. (1999). Bazooka recruits

Inscuteable to orient asymmetric cell divisions in Drosophila neuro-

blasts. Nature 402, 548–551.

39. Wodarz, A., Ramrath, A., Kuchinke, U., and Knust, E. (1999). Bazooka

provides an apical cue for Inscuteable localization in Drosophila neuro-

blasts. Nature 402, 544–547.

40. Yu, F., Morin, X., Cai, Y., Yang, X., and Chia, W. (2000). Analysis of part-

ner of inscuteable, a novel player of Drosophila asymmetric divisions,

reveals two distinct steps in inscuteable apical localization. Cell 100,

399–409.

41. Johnston, C.A., Hirono, K., Prehoda, K.E., and Doe, C.Q. (2009).

Identification of an Aurora-A/PinsLINKER/Dlg spindle orientation

pathway using induced cell polarity in S2 cells. Cell 138, 1150–1163.

42. Siegrist, S.E., and Doe, C.Q. (2005). Microtubule-induced Pins/Galphai

cortical polarity in Drosophila neuroblasts. Cell 123, 1323–1335.

43. Mummery-Widmer, J.L., Yamazaki, M., Stoeger, T., Novatchkova, M.,

Bhalerao, S., Chen, D., Dietzl, G., Dickson, B.J., and Knoblich, J.A.

(2009). Genome-wide analysis of Notch signalling in Drosophila by

transgenic RNAi. Nature 458, 987–992.

44. Neumuller, R.A., Richter, C., Fischer, A., Novatchkova, M., Neumuller,

K.G., and Knoblich, J.A. (2011). Genome-wide analysis of self-renewal

in Drosophila neural stem cells by transgenic RNAi. Cell Stem Cell 8,

580–593.

45. Carney, T.D., Miller, M.R., Robinson, K.J., Bayraktar, O.A., Osterhout,

J.A., and Doe, C.Q. (2012). Functional genomics identifies neural stem

cell sub-type expression profiles and genes regulating neuroblast

homeostasis. Dev. Biol. 361, 137–146.

46. Pi, H., Huang, S.K., Tang, C.Y., Sun, Y.H., and Chien, C.T. (2004). phyllo-

pod is a target gene of proneural proteins in Drosophila external sensory

organ development. Proc. Natl. Acad. Sci. USA 101, 8378–8383.

47. Remm, M., Storm, C.E., and Sonnhammer, E.L. (2001). Automatic clus-

tering of orthologs and in-paralogs from pairwise species comparisons.

J. Mol. Biol. 314, 1041–1052.

48. Parks, A.L., Cook, K.R., Belvin, M., Dompe, N.A., Fawcett, R., Huppert,

K., Tan, L.R., Winter, C.G., Bogart, K.P., Deal, J.E., et al. (2004).

Systematic generation of high-resolution deletion coverage of the

Drosophila melanogaster genome. Nat. Genet. 36, 288–292.

49. Morin, X., Daneman, R., Zavortink, M., and Chia, W. (2001). A protein

trap strategy to detect GFP-tagged proteins expressed from their

endogenous loci in Drosophila. Proc. Natl. Acad. Sci. USA 98, 15050–

15055.

50. Roegiers, F., Younger-Shepherd, S., Jan, L.Y., and Jan, Y.N. (2001).

Bazooka is required for localization of determinants and controlling

proliferation in the sensory organ precursor cell lineage in Drosophila.

Proc. Natl. Acad. Sci. USA 98, 14469–14474.

51. Betschinger, J., Mechtler, K., and Knoblich, J.A. (2003). The Par com-

plex directs asymmetric cell division by phosphorylating the cytoskel-

etal protein Lgl. Nature 422, 326–330.

52. Segalen, M., Johnston, C.A., Martin, C.A., Dumortier, J.G., Prehoda,

K.E., David, N.B., Doe, C.Q., and Bellaıche, Y. (2010). The Fz-Dsh planar

cell polarity pathway induces oriented cell division via Mud/NuMA in

Drosophila and zebrafish. Dev. Cell 19, 740–752.

53. Albertson, R., and Doe, C.Q. (2003). Dlg, Scrib and Lgl regulate neuro-

blast cell size and mitotic spindle asymmetry. Nat. Cell Biol. 5, 166–170.

54. Woods, D.F., and Bryant, P.J. (1989). Molecular cloning of the lethal(1)

discs large-1 oncogene of Drosophila. Dev. Biol. 134, 222–235.

55. Woods, D.F., and Bryant, P.J. (1991). The discs-large tumor suppressor

gene of Drosophila encodes a guanylate kinase homolog localized at

septate junctions. Cell 66, 451–464.

56. Barros, C.S., Phelps, C.B., and Brand, A.H. (2003). Drosophila non-

muscle myosin II promotes the asymmetric segregation of cell fate de-

terminants by cortical exclusion rather than active transport. Dev. Cell 5,

829–840.

57. Petritsch, C., Tavosanis, G., Turck, C.W., Jan, L.Y., and Jan, Y.N. (2003).

The Drosophila myosin VI Jaguar is required for basal protein targeting

and correct spindle orientation in mitotic neuroblasts. Dev. Cell 4,

273–281.

58. Etienne-Manneville, S. (2004). Cdc42—the centre of polarity. J. Cell Sci.

117, 1291–1300.

59. Speicher, S., Fischer, A., Knoblich, J., and Carmena, A. (2008). The PDZ

protein Canoe regulates the asymmetric division of Drosophila neuro-

blasts and muscle progenitors. Curr. Biol. 18, 831–837.

60. Carmena, A., Makarova, A., and Speicher, S. (2011). The Rap1-Rgl-Ral

signaling network regulates neuroblast cortical polarity and spindle

orientation. J. Cell Biol. 195, 553–562.

61. Knoblich, J.A., Jan, L.Y., and Jan, Y.N. (1995). Asymmetric segregation

of Numb and Prospero during cell division. Nature 377, 624–627.

62. Broadus, J., and Doe, C.Q. (1997). Extrinsic cues, intrinsic cues and

microfilaments regulate asymmetric protein localization in Drosophila

neuroblasts. Curr. Biol. 7, 827–835.

63. Lee, C.Y., Robinson, K.J., and Doe, C.Q. (2006). Lgl, Pins and aPKC

regulate neuroblast self-renewal versus differentiation. Nature 439,

594–598.

64. Dow, L.E., Brumby, A.M., Muratore, R., Coombe, M.L., Sedelies, K.A.,

Trapani, J.A., Russell, S.M., Richardson, H.E., and Humbert, P.O.

(2003). hScrib is a functional homologue of the Drosophila tumour sup-

pressor Scribble. Oncogene 22, 9225–9230.

65. Humbert, P., Russell, S., and Richardson, H. (2003). Dlg, Scribble and

Lgl in cell polarity, cell proliferation and cancer. Bioessays 25, 542–553.

66. Etienne-Manneville, S., Manneville, J.B., Nicholls, S., Ferenczi, M.A.,

andHall, A. (2005). Cdc42 andPar6-PKCzeta regulate the spatially local-

ized association of Dlg1 and APC to control cell polarization. J. Cell Biol.

170, 895–901.

67. Osmani, N., Vitale, N., Borg, J.P., and Etienne-Manneville, S. (2006).

Scrib controls Cdc42 localization and activity to promote cell polariza-

tion during astrocyte migration. Curr. Biol. 16, 2395–2405.

68. Royer, C., and Lu, X. (2011). Epithelial cell polarity: a major gatekeeper

against cancer? Cell Death Differ. 18, 1470–1477.

69. Humbert, P.O., Grzeschik, N.A., Brumby, A.M., Galea, R., Elsum, I., and

Richardson, H.E. (2008). Control of tumourigenesis by the Scribble/Dlg/

Lgl polarity module. Oncogene 27, 6888–6907.

70. Gho, M., Lecourtois, M., Geraud, G., Posakony, J.W., and Schweisguth,

F. (1996). Subcellular localization of Suppressor of Hairless in

Drosophila sense organ cells during Notch signalling. Development

122, 1673–1682.

71. Vaessin, H., Grell, E., Wolff, E., Bier, E., Jan, L.Y., and Jan, Y.N. (1991).

prospero is expressed in neuronal precursors and encodes a nuclear


protein that is involved in the control of axonal outgrowth in Drosophila.

Cell 67, 941–953.

72. Hutterer, A., and Knoblich, J.A. (2005). Numb and alpha-Adaptin regu-

late Sanpodo endocytosis to specify cell fate in Drosophila external

sensory organs. EMBO Rep. 6, 836–842.

73. Yu, J.X., Guan, Z., and Nash, H.A. (2006). The mushroom body defect

gene product is an essential component of the meiosis II spindle appa-

ratus in Drosophila oocytes. Genetics 173, 243–253.

74. Eroglu, E., Burkard, T.R., Jiang, Y., Saini, N., Homem, C.C., Reichert, H.,

and Knoblich, J.A. (2014). SWI/SNF complex prevents lineage reversion

and induces temporal patterning in neural stem cells. Cell 156, 1259–

1273.

75. Rhyu, M.S., Jan, L.Y., and Jan, Y.N. (1994). Asymmetric distribution of

numbprotein during division of the sensory organ precursor cell confers

distinct fates to daughter cells. Cell 76, 477–491.

76. Manneville, J.B., Jehanno, M., and Etienne-Manneville, S. (2010). Dlg1

binds GKAP to control dynein association with microtubules, centro-

some positioning, and cell polarity. J. Cell Biol. 191, 585–598.

the conserved discs-large binding partner banderuola ... · in sop cells, the planar polarity...

Documents