Copyright � 2010 by the Genetics Society of AmericaDOI: 10.1534/genetics.109.112060

Drosophila Rab23 Is Involved in the Regulation of the Number and PlanarPolarization of the Adult Cuticular Hairs

Csilla Pataki,*,1 Tamas Matusek,*,1 Eva Kurucz,* Istvan Ando,* Andreas Jenny†

and Jozsef Mihaly*,2

*Institute of Genetics, Biological Research Center, Hungarian Academy of Sciences, Szeged 6726, Hungary and †Department ofDevelopmental and Molecular Biology, Albert Einstein College of Medicine, Bronx, New York 10461

Manuscript received November 13, 2009Accepted for publication January 27, 2010

ABSTRACT

The planar coordination of cellular polarization is an important, yet not well-understood aspect ofanimal development. In a screen for genes regulating planar cell polarization in Drosophila, we identifiedRab23, encoding a putative vesicular trafficking protein. Mutations in the Drosophila Rab23 orthologresult in abnormal trichome orientation and the formation of multiple hairs on the wing, leg, andabdomen. We show that Rab23 is required for hexagonal packing of the wing cells. We found that Rab23 isable to associate with the proximally accumulated Prickle protein, although Rab23 itself does not seem todisplay a polarized subcellular distribution in wing cells, and it appears to play a relatively subtle role incortical polarization of the polarity proteins. The absence of Rab23 leads to increased actin accumulationin the subapical region of the pupal wing cells that fail to restrict prehair initiation to a single site. Rab23acts as a dominant enhancer of the weak multiple hair phenotype exhibited by the core polaritymutations, whereas the Rab23 homozygous mutant phenotype is sensitive to the gene dose of the planarpolarity effector genes. Together, our data suggest that Rab23 contributes to the mechanism that inhibitshair formation at positions outside of the distal vertex by activating the planar polarity effector system.

THE formation of properly differentiated organsoften requires the planar coordination of cell

polarization within tissues, a feature referred to asplanar cell polarity (PCP) or tissue polarity. Althoughplanar polarity is evident in many vertebrate tissues(such as fish scales, bird feathers, and cochlear epi-thelium) and it has recently been shown that PCPregulation is highly conserved throughout the animalkingdom (Strutt 2003; Fanto and McNeill 2004;Seifert and Mlodzik 2007; Simons and Mlodzik

2008), such polarization patterns are best studied in thefruitfly, Drosophila melanogaster. PCP in flies is manifest inthe mirror-image arrangement of ommatidia in the eye,in the adult cuticle, which is decorated with parallelarrays of hairs and sensory bristles, and in the wing,which is covered by distally pointing hairs (or tri-chomes) (Adler 2002). Wing hairs form during thepupal life when each cell produces a single microvillus-like prehair stiffened by actin and microtubules. Inwild-type wing cells prehairs form at the distal vertex ofthe cells and extend distally as they grow (Wong andAdler 1993).

Mutations in PCP genes result in abnormal wing hairpolarity patterns and wing hair number (Gubb andGarcia-Bellido 1982; Wong and Adler 1993). On thebasis of their cellular phenotypes (i.e., prehair initiationsite and number of hairs per cell), initial studies placedPCP genes into three groups: the first group (oftencalled the core group) includes frizzled ( fz), dishevelled(dsh), starry night (stan) (also known as flamingo), VanGogh (Vang) (also known as strabismus), prickle (pk), anddiego (dgo); the second group consists of inturned (in),fuzzy ( fy), and fritz ( frtz) (referred to as planar polarityeffectors or In group); whereas the third group includesmultiple wing hairs (mwh) (Wong and Adler 1993).Double mutant analysis demonstrated that these phe-notypic groups also represent epistatic groups, and itwas proposed that the PCP genes may act in a regulatoryhierarchy, where the core group is on the top, whereasthe In group and mwh are downstream components(Wong and Adler 1993). Subsequent work identifiedseveral other PCP genes as well. Some of these have beenplaced into the Fat/Dachsous group (Adler et al. 1998;Strutt and Strutt 2002), while another group con-sists of cytoskeletal regulators, including Rho1 and Drok(Strutt et al. 1997; Winter et al. 2001; Adler 2002).Genetic analysis of these two groups has led to models inwhich the Fat/Dachsous group acts upstream of thecore proteins (Yang et al. 2002; Ma et al. 2003), whileRho1 and Drok act downstream of Fz (Strutt et al. 1997;Winter et al. 2001). Although the existence of a single,

Supporting information is available online at http://www.genetics.org/cgi/content/full/genetics.109.112060/DC1.

1These authors contributed equally to this work.2Corresponding author: Institute of Genetics, Biological Research Center,

Hungarian Academy of Sciences, Temesvari krt. 62, Szeged H-6726,Hungary. E-mail: mihaly@brc.hu

Genetics 184: 1051–1065 (April 2010)

linear PCP regulatory pathway is debated (Casal et al.2006; Lawrence et al. 2007), it is clear that in the wing,PCP genes regulate (1) the number of prehairs, (2) theplace of prehair formation, and (3) wing hair orientation.

While the molecular mechanism that restricts prehairformation to the distal vertex of the wing cells is elusive,it has been well established that the core PCP pro-teins adopt an asymmetrical subcellular localizationwhen prehairs form (Usui et al. 1999; Axelrod 2001;Shimada et al. 2001; Strutt 2001; Tree et al. 2002;Bastock et al. 2003; Das et al. 2004), which appears to becritical for proper trichome placement. In addition, ithas recently been found that the In group of proteinsand Mwh also display an asymmetrical pattern withaccumulation at the proximal zone (Adler et al. 2004;Strutt and Warrington 2008; Yan et al. 2008). Thesestudies concluded that the core PCP proteins aresymmetrically distributed until 24 hr after prepupaformation (APF), when they become differentially en-riched until prehair formation begins at 30–32 hr APF.This transient asymmetric localization ends by 36 hrAPF (Adler 2002; Strutt 2003; Mihaly et al. 2005). Ithas recently been shown that Fz and Stan containingvesicles are transported preferentially toward the distalcell cortex in the period of 24–30 hr APF (Shimada et al.2006), and hence, polarized vesicular trafficking mightbe an important determinant of PCP protein asymme-try. Other recent studies, however, challenged the viewthat PCP protein polarization is limited to 24–32 hr APF.Instead, it has been suggested that at least a partialproximal–distal polarization is already evident at theend of larval life and during the prepupal stages (6 hrAPF). Polarity is then largely lost at the beginning of thepupal period, but becomes evident again in severalhours until hair formation begins (Classen et al. 2005).Thus, molecular asymmetries are clearly revealed dur-ing wing hair formation, yet the molecular mechanismsthat contribute to the establishment of these asymmet-rical patterns are not well understood.

In a large-scale mosaic type of mutagenesis screen, weidentified Drosophila Rab23, encoding a vesicle traffick-ing protein, as a PCP gene involved in the regulation oftrichome orientation and number in various adultcuticular structures, including the wing, abdomen,and leg. We show that Rab23 plays a modest role incortical polarization of the core PCP proteins in thewing and that Rab23 associates with at least one coreprotein, Pk. Additionally, we found that Rab23 contrib-utes to the mechanism that restricts actin accumulationand thus, prehair initiation to a single site within eachwing cell.

MATERIALS AND METHODS

Fly strains and genetics: Fly strains used are described inFlyBase, except for the different Rab23 alleles (see below) andthe Ubx-Flp stock (kindly provided by J. Knoblich, IMP, Vienna)

(Emery et al. 2005). The Rab23T69A allele has been isolatedduring an F2 FRT/Flp (Xu and Rubin 1993) mosaic muta-genesis screen. FRT82B chromosomes were mutagenized withENU (1.6 mm), mutant clones were induced by Ubx-Flp, andPCP phenotypes were analyzed on the wing and notum. TheRab2351 excision allele was generated by remobilization of theP(RS5)5-SZ-3123 (DrosDel) P-element insertion line by usingstandard techniques. The Rab23 RNAi line has been providedby the VDRC RNAi Centre (IMP-IMBA, Vienna), whereas thew; UASp-YFP-Rab23 lines were generous gifts from M. Scott(Stanford University). Flip out clones of UASp-YFP-Rab23 weregenerated using w, hsFLP; ActP-FRT-y1-FRT-Gal4, arm-lacZ witha 2-hr heat shock at 37� during the early third instar larvalstage. For overexpression studies we used Act5C-Gal4 or Sal-Gal4. In line with FlyBase, we use starry night instead offlamingo, and Van Gogh instead of strabismus.

Cuticle preparation: Abdominal cuticles and first thoraciclegs were prepared according to a protocol by Duncan (1982).

DNA techniques and tissue culture: DNA constructs fortransgenic flies and transfection experiments were created bystandard cloning techniques. Detailed cloning procedures canbe obtained from the authors upon request. To create a Rab23genomic rescue construct, a 12.3-kb XbaI restriction fragmentfrom BACR03P13 (BACPAC Resources) was cloned into apCaSpeR4 vector. To clone the Rab2351 transcript, total RNAwas isolated from homozygous mutant adults with Trizolreagent (Invitrogen), cDNA was synthesized with Revert Aid(Fermentas), and RT–PCR was carried out with primersspecific to the first and third exon of Rab23, respectively.Primers used were as follows: Rab23Ex3 59-CCCGGCCACAACATCATTAG-39 and Rab23Ex1 59-TCAAATTTCGATCGCGACGAG-39. Molecular cloning of the RT–PCR product revealedthat Rab2351 encodes a hybrid transcript consisting of the first(noncoding) exon of Rab23, part of the first intron of Rab23,followed by 443 bp from the 59 end of the P(RS5)5-Sz-3123insertion precisely until the end of ORF0 (O’Hare and Rubin

1983) that is fused to the third exon of Rab23 (Figure 1B). Iftranslation from this fusion transcript begins with the ATG ofORF0, the predicted fusion protein is entirely devoid of Rab23sequences because ORF0 and the third exon of Rab23 are notin the same phase. If a downstream ATG were used, M53 ofRab23 is the next starting point. However, translationalinitiation from M53 would result in a mutant Rab23 proteinthat is lacking 52 N-terminal amino acids, including 15 of thehighly conserved and functionally essential part of the GTPasedomain (Santos and Nebreda 1989). Thus, these resultssuggest that Rab2351 encodes a functionally strongly, if notentirely, impaired protein.

For transfection experiments we used the following con-structs: pAVW-Rab23, pPHW-Rab23, pPVW-Rab23, BLMT-RFP-Rab23, pAVW-Rab23Q96A, pPVW-Rab23Q96A, pPHW-Rab23T69A,BLMT-Rab23Q96A-EGFP, pAC5.1-Fz-Myc, pAC5.1-Dsh-EGFP,pAC5.1-HA-Dgo, pAC5.1-6xMyc-Pk, pRME-Vang-HA (a kindgift from T. Wolff). When necessary, pActin5c-Gal4 wascotransfected to drive expression from UAS promoters. Ex-pression from the BLMT vectors was induced by 1 mm CuSO4

for 4 hr or 500 mm CuSO4 overnight.Drosophila S2 cells were transfected with Effectene (Qiagen)

and incubated in Drosophila Schneider’s medium (Lonza) for24 hr before fixation.

Antibody production: Rab23 antibody was raised in mouseinjected with a bacterially produced, His-tagged, full-lengthRab23 protein. After four boosts, the crude serum was tested forimmunostainings and Western blot analysis. As a strong back-ground staining was evident in our immunostainings, the serumsubsequently was only used for biochemical experiments.

Immunohistochemistry: For pupal wing analysis whiteprepupae were collected, aged, and dissected at the desired

1052 C. Pataki et al.

time and fixed in 4% formaldehyde in PBS. Stainings weredone according to standard procedures (Wong and Adler

1993). Drosophila S2 cells were fixed as described in Matusek

et al. (2008).For immunostainings we used the following primary anti-

bodies: mouse anti-Rab23 1:100, mouse anti-b-gal 1:1000(Promega), rabbit anti-b-gal 1:1000 (Molecular Probes),mouse anti-GFP 1:200 (DSHB), rabbit anti-GFP 1:1000 (SantaCruz Biotechnology), rabbit anti-Vang 1:500 (Rawls andWolff 2003), rabbit anti-Dgo 1:200 (Feiguin et al. 2001),mouse anti-In 1:1000 (Adler et al. 2004), rabbit anti-Pk 1:2000(Tree et al. 2002), mouse anti-Myc 1:400 (Roche), mouse anti-HA 1:400 (Roche), rabbit anti-HA (Sigma), mouse anti-Fz 1:10(DSHB), and mouse anti-Stan 1:10 (DSHB). Actin was stainedwith Rhodamine-Phalloidine 1:100 (Molecular Probes). Bright-field images were collected using Zeiss Axiocam MOT2.Confocal images were collected with an Olympus FV1000LSM microscope. Images were edited with Adobe Photoshop7.0CE and Olympus Fluoview.

Immunoprecipitation and Western blot analysis: For im-munoprecipitation experiments 100 wild-type or Rab2351

homozygous mutant pupae (28–30 hr APF) were lysed in1 ml lysis buffer (0.1% SDS, 0.2% NaDoc, 0.5% NP-40, 150 mm

NaCl, 50 mm TrisHCl, pH 8.0) for 1 hr at 4�. Insolublematerials were pelleted with a centrifugation step and the clearsupernatant was used for further studies. A portion of thecleared lysate was used as Western blot control. The remaininglysate was then preincubated with 100 ml IgG free Protein-ASepharose (CL-4B, Pharmacia) beads for 1 hr at RT to depletenonspecifically binding proteins. For immunoprecipitation60 ml of CL-4B Sepharose beads were incubated with 30 ml anti-Rab23 antibody for 2 hr at RT in 1 ml final volume in lysisbuffer (precomplex). A portion of precomplex was kept asbead control. Immunoprecipitation was carried out overnightat 4�. SDS–PAGE and Western blot analysis were carried outaccording to standard protocols.

RESULTS

To identify new planar polarity genes, we employedthe FRT/Flp mosaic system to induce homozygousmutant clones on the wing and notum of mutagenizedflies by Ubx-Flp (Emery et al. 2005). Genetic mapping ofone of the new PCP mutants, displaying multiple hairsin mutant clones, showed that the mutation affected theDrosophila Rab23 ortholog. Rab23 belongs to the Rabfamily of small GTPases known to play a role in vesicularmembrane transport (Zerial and McBride 2001).Sequence analysis of the new allele revealed a pointmutation (T69A) in the Switch I region of the GTPasedomain of Rab23, affecting an amino acid residue that isinvariant in the whole small GTPase superfamily (Figure1A) (Vetter and Wittinghofer 2001). This allele,designated Rab23T69A, is semilethal, but homozygousmutant animals occasionally survive till adulthood anddisplay a strong multiple wing hair phenotype as well ashair orientation defects (Figure 2, C, D, and F).Although these trichome orientation defects are rela-tively mild compared to fz or dsh, they are exhibited inevery Rab23 mutant wing. Whereas in wing sectors B andC hair orientation is mildly affected, the deflection fromwild-type orientation is obvious in sectors A, D, and E

(Figure 2, A–F and data not shown), clearly indicating arequirement in the regulation of hair orientation.Because no other Rab23 alleles were available, wegenerated independent Rab23 alleles by remobilizingthe P(RS5)5-Sz-3123 P-element insertion sitting in thefirst intron of Rab23 (Figure 1B). One of these alleles,Rab2351, is homozygous viable and displays wing hairphenotypes identical to that of Rab23T69A (Figure 2G).Rab23T69A is viable over Rab2351 and similarly, bothmutations are viable over deficiency chromosomesuncovering Rab23. Each of these mutant combinationsdisplays a strong multiple hair phenotype and weak hairorientation defects (Figure 2, H and I; supportinginformation, Figure S3). As the severity of the wingphenotypes is identical in the Rab23 transheterozygous,homozygous, and hemizygous mutants, by genetic crite-ria, both Rab23 alleles behave as strong loss-of-function(LOF) or null alleles. In agreement with this, it hasalready been reported that impairment of the threo-nine residue (equivalent to that of T69 in Rab23) inother small GTPases leads to a strong loss of function(Spoerner et al. 2001). Moreover, we revealed thatRab2351 encodes a hybrid transcript that cannot betranslated into a functional protein (see details inmaterials and methods) suggesting that Rab2351 is anull allele.

The wing hair phenotypes of both Rab23 alleles can befully rescued by providing a single wild-type copy of theRab23 gene (Figure 2K and data not shown). Hence,the LOF analysis and the rescue experiments togetherindicate that Rab23 in Drosophila is not essential forviability; however, it is involved in the regulation of winghair number, and to a lesser degree, wing hair orienta-tion. Consistently, silencing of Rab23 by RNAi usingubiquitously expressed, as well as wing-specific drivers,resulted in a moderately strong multiple hair phenotype(Figure 2J), without affecting viability (data not shown).

Rab23 impairs hair polarity and number on the adultcuticle: Because PCP is also manifest in tissues otherthan the wing, we examined the eyes, notum, legs, andabdomen of Rab23 mutant flies. We found that omma-tidial polarity and the orientation of the epidermalbristles on the notum, legs, and abdomen are un-affected by Rab23, and these mutants also lack theduplicated tarsal joint phenotype typical of the core PCPmutants (Figure 3 and data not shown). Conversely,Rab23 affects the orientation and number of trichomescovering the legs and abdomen (Figure 3). The orien-tation defects are equally evident on the legs, tergites,sternites (Figure 3), and pleural regions of the abdom-inal cuticle (not shown). Interestingly, hair polaritylooks randomized all over the tergites (Figure 3F), thatis different from other PCP mutations exhibiting polar-ity reversions or orientation defects affecting only cer-tain areas along the anterior–posterior axes of thetergites (Casal et al. 2002; Lawrence et al. 2004).Additionally, the formation of multiple hairs is also

Rab23 Regulates Hair Number and Polarization 1053

obvious on the leg and the abdominal cuticle (Figure 3).Together, these data suggest that Rab23 identifies aunique class of PCP genes that is specifically requiredfor the regulation of trichome orientation and numberin all body regions examined without affecting theplanar polarization of multicellular structures such asommatidia or sensory bristles.

Rab23 impairs prehair initiation and hexagonalpacking of pupal wing cells: To investigate Rab23function at the cellular level, we subsequently focusedour analysis on the wing. First, we examined prehairinitiation in Rab2351 homozygous mutant pupal wingsand Rab23T69A mutant clones at 31 hr APF. In theabsence of Rab23, apical actin accumulation is notrestricted to the distal vertex of the cells. Instead, wedetected large amounts of diffusely organized actin

filaments in the apical region of the wing cells (Figure 4,A–C$). Presumably as a consequence of the failure torestrict the site of actin accumulation, many cellsdeveloped more than one prehair that, similarly tomutations of the In group, formed in abnormal posi-tions around the cell periphery (Figure 4, B and D).Additionally, we noted that prehair initiation is some-what delayed in Rab23T69A mutant clones (Figure 4, D–D$). This delay is particularly obvious in clones inducedin the proximal half of wing sector B and C (in 85.7% ofthe clones, n¼ 23), conversely, the delay is less frequentin clones from distal B and C sectors (25%, n ¼ 20).Clones with delayed prehair initiation are more ran-domly distributed in wing sectors D and E in �50% ofclones examined (n ¼ 36). Beyond these effects, theclonal analysis also revealed that Rab23 acts in a cell

Figure 1.—Conservation of the Rab GTPasesand the structure of the Drosophila Rab23 gene.(A) Multiple alignment of small GTPase familymembers; the highly conserved GTPase domainis indicated with a black line above the align-ment. The Rab23T69A mutation affects a threonineresidue that is invariant in the small GTPase su-perfamily. (B) The Rab23 genomic region isindicated on top with the position of theP(RS5)5-Sz-3123 insertion used to generate theRab2351 allele. The structure of the RH23273EST clone, corresponding to a full-lengthRab23 cDNA, is shown below (black boxes indi-cate coding region; white ones, untranslated re-gions). Further below we indicate the genomicstructure of the Rab2351 allele, note the 1.7-kbforeign DNA from the P-element insertion line.The possible transcript of the Rab2351 allele isshown at the bottom of the figure. Translationfrom this transcript cannot result in a functionalRab23 protein because of premature termina-tion.

1054 C. Pataki et al.

autonomous manner, as the presence of multiple hairswas always restricted to the mutant cells (Figure 4, E–E$).

Recent studies uncovered that the wing epithelialcells are irregularly shaped throughout larval and earlypupal stages, but most of them become hexagonallypacked shortly before prehair formation (Classen et al.2005). It has also been shown that the core PCPmutations partly interfere with normal cellular packing(Classen et al. 2005). While analyzing Rab23 mutantpupal wings, we noticed that some of the Rab23 mutantwing cells also fail to adopt a hexagonal shape at �30–32 hr APF (Figure 5B). We quantified this effect in the Dregion of the wing distal to the posterior cross vein(Figure 2A), and found that in wild-type wings the ratioof the nonhexagonally shaped cells is�11%, whereas inRab23T69A homozygous mutant wings this ratio is in-creased to 27% (Figure 5C). As a comparison, wemeasured the packing defects in two core PCP mutants,Vang6 and dsh1, where the ratio of the nonhexagonalcells was 30 and 28%, respectively (Figure 5C). Becausethe strength of the Rab23-induced packing defects iscomparable to the effect of the core PCP mutations,these data suggest that Rab23 might also be an impor-tant determinant of cellular packing.

Next we addressed whether the packing defectsrevealed in Rab23 mutant wings correlate with the hairorientation defects and/or the formation of multiplehairs, also exhibited upon loss of Rab23. However, wefound no obvious correlation with the hair orientationdefects (not shown). Similarly, when we compared theaverage prehair number of hexagonally and nonhexag-onally packed Rab23 mutant wing cells, we failed toreveal any correlation between the presence of multiplehairs and the cellular packing defects (Figure 5D). Toextend this analysis, we investigated the packing defectsof other mutations causing multiple hairs, such as in,frtz, and mwh. These mutants exhibit somewhat weakerpacking defects than Rab23 (Figure 5C and data notshown), but as for Rab23, the formation of multiplehairs does not appear to correlate with irregular cellshape (Figure 5D). Taken together, these results dem-onstrate that cell packing has no direct effect on thenumber of prehair initiation sites in the wing epithelialcells.

Rab23 weakly impairs late cortical polarization of thePCP proteins: Given that the absence of asymmetricallylocalized core PCP proteins alters hexagonal packing ofthe wing cells (Classen et al. 2005), and packing was

Figure 2.—Rab23 mutant wingsdisplay PCP defects. (A) A sche-matized wing indicating the mainwing blade regions (A–E) and thetwo cross veins (acv, anteriorcross vein; pcv, posterior crossvein). Distal is on the right andanterior is on top on all panels.Wild-type wing cells exhibit asingle, distally pointing hair(B and E), whereas Rab23T69A

(C, D, and F) and Rab2351 (G)mutant wings exhibit multiplehairs and orientation defects inevery wing regions. Photomicro-graphs were taken from the A re-gion (B and C), the C region (D)and the D region (E–L) of thewing (dashed areas in panel A).Wings of Rab23T69A/Df(3R)BSC47(H), Rab2351/Df(3R)BSC47 (I),and wings in which Rab23 is si-lenced by RNAi ( J), also displaymultiple hairs and hair orienta-tion defects. (K and L) The wingPCP phenotypes induced byRab23 can be fully rescued by asingle copy of the Rab23 genomicrescue construct (Rab23-GR)(K), and by Act-Gal4 driven ex-pression of UAS-YFPTRab23 (L).

Rab23 Regulates Hair Number and Polarization 1055

altered in Rab23 mutant wings, we asked whether Rab23affects PCP protein localization. To this end, we exam-ined the localization pattern of several core PCPproteins in Rab23 mutant pupal wings. These experi-ments showed that the initial polarization pattern ofStan in prepupal wings is essentially identical in wild-type and Rab2351 mutant wings (Figure S1). However, by24–30 hr APF, although the core PCP proteins stillaccumulate into apicolateral complexes, they fail topolarize properly in a Rab23 mutant tissue (Figure 6, A–C9). Most notably, the PCP proteins are often not ascompletely removed from the anterior–posterior (A–P)cell boundaries as in wild type (Figure 6B). Also

sometimes even if intracellular PCP polarity is obvious,the angle of polarity is not well correlated from cell tocell (Figure 6A). Interestingly, we found that in Rab23mutant clones that exhibited a delay in actin accumu-lation, mislocalization was always evident (n ¼ 26) inmost parts of the mutant tissue (Figure 4D9) indepen-dent of the wing region the clone was from. In clonesthat did not exhibit a delay in actin accumulation,localization was strongly affected in the distal C regionin every case (n ¼ 7), whereas in other wing sectors theeffect was restricted to a smaller region within the clone(n ¼ 24). Thus, instead of a regional specificity, itappears that altered PCP protein localization correlateswith the delay in actin accumulation, i.e., localizationdefects are more evident in delayed clones than inothers. In agreement with this, we have been unable tofind a clear regional specificity in Rab23 homozygousmutant wings. Instead, smaller or larger areas in whichthe ‘‘zigzag’’ pattern was altered were evident in all wingregions (n¼ 25). Consistent with previous findings thatIn localization depends on core PCP protein localiza-tion (Adler et al. 2004), we found that the lack of Rab23impairs In localization as well (Figure 6, D and D9).Together, these results suggest that Rab23 contributes tothe late (24–30 hr APF) apical, cortical polarization ofthe PCP proteins.

As we pointed out above, Rab23 is also required forhexagonal packing of the wing cells; therefore, it wasnecessary to clarify whether the abnormal-looking PCPprotein localization is not simply the consequence ofirregular cell packing. To address this issue, we carefullyexamined Pk localization in wild-type and Rab23 mutantwings in such hexagonal cells whose six neighbors werehexagonally shaped as well and therefore packingdefects could not interfere with localization (Figure7). This analysis revealed that a Rab23 mutant hexagonalcell has a significantly higher chance of displayingimpaired PCP protein localization than a wild-type cell(Figure 7, A–C). Because impaired PCP protein locali-zation is also associated with cellular packing defects inthe core PCP mutants, we also quantified Pk localizationin dsh1 mutant hexagonal wing cells. In agreement withprevious work that provided a general assessment of Pklocalization in dsh mutants (Tree et al. 2002), we foundthat dsh strongly disrupts the asymmetric accumulation ofPk in the hexagonal cells (Figure 7C). Hence, it appearsthat the core PCP mutations have a much stronger effecton PCP protein polarization than that of Rab23.

The subcellular distribution of Rab23 in pupal wingcells: To gain insight into the mechanism wherebyRab23 contributes to core PCP protein localization, weexamined the subcellular localization of Rab23 indeveloping pupal wings. Given that the anti-Rab23serum we raised does not appear to be suitable forimmunohistochemical analysis (see materials and

methods), we expressed a YFPTRab23 fusion proteinto assess localization in wing cells. Note that YFPTRab23

Figure 3.—Rab23 affects hair polarity on the leg and theabdominal cuticle. (A and B) Photomicrographs of the femur(male first leg), the third sternite (female) (C and D), and thefourth tergite (female) (E and F). Proximal is on left, distal ison right on A and B, anterior is on top, posterior is on bottomon C–F. (A) A wild-type leg exhibiting distally pointing hairsand bristles. (B) A Rab23 mutant leg is covered by distallypointing bristles; however, trichome orientation defects,and occasionally the formation of multiple hairs are evident.(C) Hairs and bristles of a wild-type sternite and (E) tergitepoint posteriorly. In Rab23 mutants sternite (D) and tergite(F) hairs display largely randomized orientations and the ap-pearance of multiple hairs is also evident, whereas, bristle ori-entation is not significantly altered.

1056 C. Pataki et al.

fully rescues the Rab23 mutant phenotypes (Figure 2L),indicating that it is a functional protein. Between 24 and32 hr APF, when the core PCP proteins repolarize, theYFPTRab23 protein expressed uniformly in the wingexhibits a strong plasma membrane association in thejunctional zone, and also in the basolateral zone (Figure8, A–B$). Additionally, we detected an elevated YFP-TRab23 level in the subapical cytoplasmic domain withoccasional accumulations in punctate structures (Fig-ure 8, A–A$). A very similar localization pattern wasfound before 24 hr APF (data not shown), and at 36 hrAPF. The expression of the YFPTRab23 fusion proteinin flip-out clones also led to similar observations, and,importantly, it confirmed the lack of any apparentproximal–distal polarization (Figure 8, C–D$). Thus,Rab23 itself does not display a proximodistally polarizeddistribution from the onset of pupal development untilshortly after trichome placement. As a comparison, weexamined the localization pattern of YFPTRab23Q96L, aconstitutively active and YFPTRab23S51N, a dominantnegative (DN) form of Rab23. The overall patterns weresimilar to wild-type YFPTRab23, but the activated formdisplayed stronger membrane association, whereas theDN exhibited higher accumulation in the cytoplasmthan wild type (Figure S2). Because this is in fullagreement with the membrane cycle model of Rab

GTPases (Behnia and Munro 2005), we conclude thatalthough our localization studies are based on over-expression, they are likely to reflect the normal locali-zation of Rab23 to a great extent.

Next, we addressed the question whether increasedamounts of Rab23 would affect wing development andPCP protein localization. However, overexpression ofYFPTRab23 had no effect on wing development, andthe protein did not impair PCP protein localization(data not shown). The overexpression of the activatedform had no effect on pupal wing development either,while the DN induced a weak multiple hair phenotypethat was however much weaker than that of Rab2351

(data not shown) and therefore this allele was not usedfor further studies.

Rab23 associates with the Pk protein: As Rab23 seemsto be required for PCP protein polarization, we won-dered whether core protein(s) might be directly regu-lated by, or associated with Rab23. Because YFPTRab23displays a strong, uniform membrane association whenexpressed in pupal wing cells, and the core PCP proteinsare also enriched in membrane complexes, our initialcolocalization studies in pupal wings were not informa-tive with regard to the identification of potential Rab23partners. To get around this problem, we used culturedS2 cells in which Rab23 and the core PCP proteins are

Figure 4.—Rab23 mutations impair prehairinitiation. (A–A$) Prehair initiation in wild-typepupal wings at 31 hr APF. Note that each wingcell develops an actin-rich prehair at its distal ver-tex. (B–B$) Apical actin level is increased inRab2351 mutant pupal wing cells, many of whichfail to restrict prehair initiation to a single site.(C–C$) Z axis projections of the wing shown inpanel B, projection is shown along the white lineindicated on B–B$. Note that Stan and actin areaccumulated near the apical cell surface. (D–D$)Prehair initiation is delayed in a Rab23T69A mu-tant clone (derived from the proximal half ofwing sector C, and marked by the absence ofb-gal staining, in blue) as prehairs are shorterin the mutant tissue than in the surroundingwild-type tissue. Note that some cells initiate mul-tiple hairs that form around the cell periphery.Moreover, the P-D accumulation of Stan is partlyimpaired within the mutant tissue. (E–E$) Mul-tiple hairs are not seen outside of Rab23T69A

mutant clones indicating that Rab23 acts cell au-tonomously. In all panels, cell borders are la-beled with Stan staining (green in A, A9, B, B9,C, C9, D, D9, E, and E9), actin is labeled in redin A–E9, and in white in A$, B$, C$, D$, andE$. Bar, 10 mm.

Rab23 Regulates Hair Number and Polarization 1057

normally not expressed, or only expressed at a moderatelevel (Bhanot et al. 1996; and our unpublished results).We compared the subcellular distribution of the coreproteins in cultured S2 cells transfected with the appro-priately tagged version of the Fz, Dsh, Pk, Dgo, and Vangproteins in the presence and in the absence of Rab23.Additionally, we examined whether PCP protein local-ization was altered by Rab23 in Fz/Dsh and Vang/Pkcotransfected cells. Together, these studies led to theconclusion that the presence of Rab23 does not modifythe subcellular distribution of the core proteins, in-cluding the membrane localization of Fz and Fz/Dsh.However, we noticed that while Fz, Dsh and Dgo do notcolocalize with Rab23 (Figure 9, A–B$ and not shown)(for Fz 0% of the cells exhibited colocalization, n¼ 29),Vang and Pk partially colocalize with Rab23 (Figure 9,C–D$) (partial colocalization for Pk was evident in 96%of the cells, n ¼ 54). Because the Vang and Pk proteinsdid not show cell membrane localization when ex-pressed alone or together (data not shown), it was notpossible to test whether Rab23-dependent vesicle traf-ficking affects the subcellular distribution of Vang andPk. Nonetheless, these results indicate that Rab23 mightaffect core protein localization by regulating Vang and/or Pk distribution. Consistent with this possibility, theactivated form of Rab23 (Rab23Q96A) displayed astrong colocalization with Pk (Figure 9, E–E$), whereasthe T69A mutant version exhibited a very low level ofcolocalization (in 1 out of 26 cells) (Figure 9, F–F$).

To support the relevance of the observations made inS2 cells, and to verify whether Rab23 associates with

Vang and Pk in vivo, co-immunoprecipitation experi-ments were carried out. Western blot analysis of S2 cellstransfected with HA-tagged Rab23 demonstrated thatHA-Rab23 is specifically recognized by our anti-Rab23serum, but not by the preimmune serum (Figure 9G).The same was true for the purified His-tagged protein(data not shown), therefore the anti-Rab23 serumappeared suitable for biochemical experiments. In-deed, anti-Rab23 co-immunoprecipitated Pk fromwild-type but not from Rab2351 mutant pupal proteinextracts (28–30 hr APF) (Figure 9H). In parallel, wewere unable to detect Vang or Stan in the Rab23complex (Figure 9H), despite the fact that Vang isknown to bind Pk (Bastock et al. 2003; Jenny et al.2003). Thus, our data suggest that Rab23 interacts withPk that could explain the localization problems of Pkand, indirectly, the other core PCP proteins. However,because the PCP complexes might be sensitive to bio-chemical manipulations or might undergo very dy-namic changes, we cannot exclude that Rab23 directlyregulates the distribution of Vang or some of the otherPCP proteins as well.

Rab23 cooperates with the core PCP and In group ofgenes to regulate wing hair number: If the sole functionof Rab23 were to modulate the cortical polarization ofPk, and possibly some other core PCP proteins, it wouldbe expected that Rab23 mutants exhibit similar pheno-typic defects as pk mutants or mutations of the coregroup. Indeed, even if not as strongly as typical for thecore PCP mutants, Rab23 impairs wing hair orientation.However, beyond that effect, the strong multiple hair

Figure 5.—Rab23 impairs hexagonal packingof the pupal wing cells. (A) DE-cadherin stainingin wild type and (B) Rab2351 mutant pupal wingsat 30 hr APF. In wild-type wings by this stage ofdevelopment most cells acquire a hexagonalshape, whereas Rab23 wings display an increasedratio of nonhexagonal cells (indicated with whitedots). (C) Quantification of the ratio of the non-hexagonally shaped cells in wild type, Rab23 mu-tant and other PCP mutant wings. We examined10–25 wings for each genotype and counted atleast 100 cells per wing. To test for significancewe applied the t-test as a statistical method, Pindicates probability. (D) Quantification of theratio of cells that exhibit multiple hairs in hexag-onal (open columns) and in nonhexagonal(shaded columns) cells in Rab2351, in1, and frtz1

mutants. Bars, 5 mm.

1058 C. Pataki et al.

phenotype of Rab23 is different from the defectsexhibited by the core PCP mutants. Thus, Rab23 ap-pears to have two distinct activities during the establish-ment of tissue polarity in the wing. The first is a role inlate PCP protein polarization, while the second is torestrict actin accumulation and prehair initiation to asingle site. Consistent with a specific role in the re-striction of prehair formation, we found that Rab23dominantly enhances the weak multiple hair phenotypeof the core mutations (without affecting the hairorientation defects) (Figure 10, C, D, H, I, and P), whilethe Rab23 homozygous mutant phenotype is sensitive to

the gene dose of the In group and mwh (Figure 10, L andQ, and Figure S3) known to play a role in the regulationof wing hair number. In contrast, Rho1 and Drokmutations, affecting two cytoskeletal regulators of tri-chome placement, do not exhibit dominant geneticinteraction with Rab23 (Figure 10Q). Similar to in,Rab23 enhances the multiple hair phenotype inducedby late hs-Fz overexpression (Figure S4) (Krasnow andAdler 1994; Lee and Adler 2002), whereas Rho1 andDrok suppressed it (Strutt et al. 1997; Winter et al.2001). Hence, the dominant interaction studies suggestthat Rab23 cooperates with the core PCP and In group ofgenes, but not the Rho pathway, during the regulationof wing hair number.

To further probe the relationship between Rab23 andthe PCP genes, we examined double mutant combina-tions of Rab23 and mutations of the core group (fz21,Vang6, pkpk-sple13, and pkpk30), the In group (frtz1 and in1)and mwh1. The double mutants with fz, Vang, and pkpk-sple

displayed the type of hair orientation defects found inthe corresponding single core PCP mutants (Figure 10,E and G, and data not shown). However, they exhibiteda synergistic interaction with regard to their multiplehair phenotype, which is stronger than the sum of theindividual mutants (Figure 10, B, C, E–G, and R), pkpk-sple

having the strongest effect (Figure 10R). In the pkpk30;Rab23 combination we noted a partial suppression of

Figure 6.—PCP protein polarization in Rab23 mutant wingclones. (A–D9) Rab23T69A mutant clones at 30 hr APF, markedby the absence of b-gal staining (blue). The proximal–distalpolarization of Pk (in green) (A and A9), Vang (in green)(B and B9), Stan (in green) (C and C9), and In (in green)(D and D9) is impaired in many of the mutant cells. Some-times the angle of polarity is not well correlated fromcell to cell (best visible in A9 and D9), or the PCP proteincomplexes are not efficiently removed from the anterior–posterior cell boundaries and the ‘‘zig-zag’’ pattern is lost (ob-vious in B9 and C9). Photomicrographs were taken from the Cregion (B and C) and the D region (A and D) of the wing. A9,B9, C9, and D9: single images of the green channel of A, B, C,and D. Bar, 10 mm.

Figure 7.—Pk is mislocalized in hexagonally packed Rab23and dsh mutant wing cells. (A and B) Pk staining on wild type(A) and Rab2351 mutant wings (B). Pk localization was exam-ined in those hexagonally shaped cells whose all six neighborswere hexagonal as well, white dots represent cells with im-paired Pk localization. Mislocalization is assigned to thosecells where Pk localization is not restricted to the P-D cellboundaries but appears at the anterior–posterior cell bound-aries as well. Note that Rab2351 mutant wings exhibit a higherratio of cells with localization defects than wild type. Proximalis on left, anterior is up on A and B. (C) Quantification of hex-agonal cells with impaired Pk localization in wild type, Rab2351

mutant, and dsh1 mutant wings. A total of 6–15 wings wereanalyzed for each genotype. To test for significance we ap-plied the t-test as a statistical method, P indicates probability.Bars, 5 mm.

Rab23 Regulates Hair Number and Polarization 1059

the very strong and stereotyped hair orientation defectstypically exhibited by the pkpk30 single mutant (Figure 10,H and J). Additionally, with respect to wing hair number,we observed an even stronger synergistic effect thanwith fz and Vang, since the pkpk30; Rab23 double mutantexhibited a very high number of multiple hairs (Figure10, J and R). Overall, the phenotype was almost identicalto that of frtz1 or in1 (Figure 10K and Figure S3). Thefrtz1; Rab2351 and in1, Rab23T69A combinations displayedessentially identical PCP defects as the frtz1 or in1 singlemutants (Figure 10M and Figure S3), except that theaverage trichome number per cell was somewhat higherin the double mutants than in the corresponding singlemutant (Figure S5). Finally, the mwh, Rab23 doublemutants displayed an identical PCP phenotype to that ofmwh single mutants (Figure 10, N and O).

With respect to the determination of prehair initia-tion site, the double mutant analysis revealed that in,frtz, and mwh are epistatic to Rab23, and therefore theyare likely to act downstream of, or later than that ofRab23. However, the core PCP proteins appear to coop-erate with Rab23 in a more complex manner. In thedouble mutant assay they exhibit a synergistic interac-tion in regard of wing hair number, suggesting that theyact in parallel and/or redundant signaling pathways.Yet, the core PCP mutations not only exhibit a dominantgenetic interaction with that of Rab23, but Rab23 seemsto bind Pk and appears to play a role in corticalpolarization of the PCP proteins, which makes it un-likely that they only act through completely indepen-dent pathways.

Drosophila Rab23 is not required for Hedgehogsignaling: It was previously shown that the mouse Rab23ortholog is an essential negative regulator of Sonichedgehog (Shh) signaling during neural patterning of

the mouse embryo (Eggenschwiler et al. 2001). Con-trasting to that, we found that Drosophila Rab23 is notexpressed in the embryonic CNS (data not shown), andit follows that CNS defects were not detected in Rab23mutant embryos, therefore Rab23 is unlikely to play arole in neural development in flies. Nevertheless,Drosophila hedgehog (hh) is required for the properdevelopment of many different tissues from the embry-onic stages to adulthood (Ingham and McMahon

2001). Thus, to determine further whether DrosophilaRab23 plays a role in Hh signaling, we analyzed Rab23homozygous mutant embryos and adult tissues. Wefound no evidence for a Rab23 requirement in Hhsignaling, as for example, the embryonic cuticle patternor anterior–posterior patterning of the wing remainednormal in Rab23 mutants (data not shown). Moreover, ifRab23 were a negative regulator of Hh signal trans-duction in flies, the loss of Rab23 should activate the Hhpathway (such as the loss of patched) and would beexpected to lead to phenotypic effects similar to theoverexpression of Hh. However, in the case of theabdominal cuticle, Hh overexpression induces reversedhair polarity (Struhl et al. 1997; Lawrence et al. 1999),that is clearly distinct from the effect of Rab23, whichinduces randomized polarity and multiple hairs (Figure3). Together, these findings indicate that although theRab23 protein appears to be highly conserved through-out evolution (Guo et al. 2006), its role in Hh signaling islikely to be restricted only to vertebrates.

DISCUSSION

Here we have shown that Drosophila Rab23 is re-quired for the planar organization of the adult cuticularhairs covering the epidermis of the wing, leg, and

Figure 8.—The subcellular distribution ofYFPTRab23 in pupal wing cells. (A–A$) In thejunctional zone the YFPTRab23 protein (ingreen) is enriched in the plasma membrane ofpupal wing cells and displays a diffuse stainingin the cytoplasm. In the basolateral zone (B–B$) YFPTRab23 is also enriched in the plasmamembrane, but staining in the cytoplasm is muchweaker than in junctional zone. Pupal wings de-picted on A–B$ were stained 39 hr APF at 22�(corresponding to 31 hr APF at 25�), actin is la-beled in red in these panels. A and B are two dif-ferent optical planes of the same image, thedistance between the two sections is 7 mm. (C–D$) The subcellular localization of YFPTRab23(green) expressed in flip-out clones. DE-cadherin(red) labels the junctional zone. Note the lack ofany obvious proximal–distal polarization (C–C$),and the strong colocalization with DE-cadherinin the junctional zone (D–D$). D–D$ display aZ section of the wing along the white line indi-cated on C–C$. Proximal is on left, distal is onright on each panels. Bars, 10 mm in A and Band 20 mm in C.

1060 C. Pataki et al.

abdomen. Rab23 appears to regulate two main aspects oftrichome development that are hair orientation andhair number. In pupal wing cells, the absence of Rab23leads to increased actin accumulation in the subapicalregion and the formation of multiple hairs. In addition,Rab23 mutations impair hexagonal packing of the wingcells, and to a lesser degree, affect cortical polarizationof the PCP proteins. Although, Rab23 does not appearto exhibit a polarized distribution in wing cells, wefound that it associates with Pk that normally accumu-lates in the proximal cortical domain.

Careful comparison of the Rab23 mutant phenotypewith that of the other PCP mutations reveals that the

phenotypic effect of Rab23 differs from all of the knownPCP genes. Most notably, Rab23 has a specific require-ment in the development of one particular type ofsubcellular structure (i.e., the cuticular hair) in everybody region we examined. However, it does not appearto play any role in the planar orientation of multicellularunits such as ommatidia in the eye or the sensory bristlesof the adult epidermis. In contrast to this, other PCPgenes typically exhibit a tissue specific, but not structurespecific, requirement, or, such as mutations of the coregroup, affect the polarization of every tissue andstructure, regardless of whether they are hairs, bristles,or unit eyes. Focusing on the wing, loss of Rab23 results

Figure 9.—Rab23 associates with Pk. (A–F$) S2 cells cotransfected with Rab23 and Fz (A–A$), Rab23 and Dsh (B–B$), Rab23 andVang (C–C$), Rab23 and Pk (D–D$), Rab23Q96A and Pk (E–E$), and Rab23T69A and Pk (F–F$). Rab23 is labeled in green, core PCPproteins are labeled in red. Fz and Dsh do not show a significant colocalization with Rab23 (A$ and B$), whereas Vang and Pk display apartial overlap with that of Rab23 (C$ and D$). Rab23Q96A, the activated form of Rab23, exhibits a strong colocalization with Pk (E–E$), while theRab23T69Amutation reducescolocalizationwith Pk (F–F$). (G)Westernblot analysisof nontransfected S2 cells, andS2cells transfected with HA-tagged Rab23. Rab23-HA is specifically recognized by anti-Rab23 and anti-HA, but not by the preimmuneserum. The predicted molecular weight of wild-type Rab23 is 30 kDa, whereas HA-Rab23 is�40 kDa. (H) Immunoprecipitations fromlysates of wild-type and Rab2351 homozygous mutant 30-hr pupae using anti-Rab23 and probed with anti-Rab23 (upper left), anti-Pk(upper right), anti-Vang (lower left), and anti-Stan (lower right). Rab23 co-immunoprecipitates Pk but not Vang and Stan from wild-type pupae, whereas Rab23 and Pk could not be precipitated from Rab2351 mutants. Bar, 5 mm.

Rab23 Regulates Hair Number and Polarization 1061

in weak trichome orientation defects and a relativelystrong multiple hair phenotype (mostly double hairs).This is clearly different from the core PCP phenotypes(strong hair orientation defects and few multiple hairs),

or the phenotypes of the In group and mwh (strongorientation defects and multiple hairs in almost everycell). As compared to Rho1 and Drok, Rab23 displays asimilar adult wing hair phenotype in mutant clones with

Figure 10.—Genetic interaction studies and double mutant analysis with Rab23 and other PCP mutations. (A) Wild-type wingwith single distally pointing hairs. (B) Rab2351 mutant wing exhibiting multiple hairs and modest hair orientation defects. (C) fz21,(F) Vang6, and (H) pkpk30 wings displaying very few if any multiple hairs and strong hair orientation defects. Rab2351 significantlyincreases the number of multiple hairs in the fz21, Rab2351/fz21 (D), and pkpk30; Rab2351/1 (I) mutant combinations. (L) frtz1/1;Rab2351 mutant wings display a higher number of multiple hairs than a Rab2351 homozygous mutant wing (compare L to B), in-dicating that frtz1 is a dominant enhancer of Rab23. The double mutant combinations fz21, Rab2351 (E), Vang6; Rab2351 (G), andpkpk30; Rab2351 (J) display a very high number of multiple hairs and strong hair orientation defects. The orientation defects in pkpk30;Rab2351 seem to be weaker than in pkpk30 (compare J to H). frtz1 (K) mutant wings display a very high number of multiple hairs andstrong orientation defects, such as the frtz1; Rab2351 (M) double mutants. Note that the pkpk30; Rab2351 double mutant exhibits anidentical PCP phenotype as frtz1 (compare J to K). The mwh1 single mutant wing hair phenotype (N) is essentially identical to theone of a mwh1, Rab2351 (O) double mutant. Photomicrographs were taken from the D region of the wing and were positioned thesame way as the ones in Figure 2. (P) Quantification of wing hair numbers in core PCP and Rab23 mutant combinations. Rab23dominantly enhances the number of multiple wing hairs exhibited by core PCP mutations. (Q) Quantification of wing hair num-bers in mutant combinations that were homozygous for Rab23 and heterozygous for an In group mutation or mwh or a mutation ofthe Rho pathway. frtz1, fy3, and mwh1 dominantly enhance the number of multiple wing hairs exhibited by Rab2351 whereas Drok2

and RhoA72F had no significant effect. The in1 allele also had no effect in this wing region but in other wing regions it significantlyenhanced the number of multiple hairs exhibited by Rab2351/Rab23T69A (see also Figure S3). (R) Quantification of wing hair num-bers in double homozygous mutant combinations of Rab23 and core PCP mutations. The numbers indicate a synergistic geneticinteraction. During the quantitative analysis, hairs were counted in the proximal half of wing sector A, we examined 10–22 wingsfor each genotype. To test for significance we applied the t-test as a statistical method; P indicates probability.

1062 C. Pataki et al.

respect to multiple hairs, while the orientation defectsare less clear in Rho1 and Drok mutants (Strutt et al.1997; Winter et al. 2001) than in Rab23. Moreover, asignificant difference exists at the molecular level,because, unlike Rab23, Rho1 and Drok do not play a rolein cortical polarization of the core PCP proteins (C.Pataki and J. Mihaly, unpublished results). Given thatour Rab23 alleles genetically behave as strong LOF ornull alleles, Rab23 identifies a unique class of PCPgenes dedicated to the regulation of trichome planarpolarization.

Although some recent data suggested that the estab-lishment of properly polarized cortical domains is notan absolute requirement for correct trichome polarityin the wing (Strutt and Strutt 2007), asymmetricaccumulation of the PCP proteins is thought to serve asa critical cue for cell polarization. Thus, the Rab23-induced weak alterations in wing hair polarity are bestexplained by the similarly modest effect on PCP proteinasymmetries. Because Rab23 is able to associate with Pk,it follows that Rab23 is likely to play a role in theproximal accumulation of Pk. Given that the Rab familyof proteins is known to control membrane trafficking,our results provide further support for models suggest-ing that polarized membrane transport is an importantmechanism for the asymmetric accumulation of thePCP proteins (Shimada et al. 2006). Although Rab23showed a specific interaction with Pk, technical limi-tations might have prevented the detection of interac-tions with other core PCP proteins, and hence it ispossible that the mechanism whereby Rab23 con-tributes to cortical polarization is not limited to Pkregulation. One additional candidate is the transmem-brane protein Vang that partly colocalizes with Rab23 inS2 cells (this work) and has been shown to bind Pk(Bastock et al. 2003; Jenny et al. 2003). Thus, throughbinding to Pk, Rab23 might affect Vang localization orsignaling capacity. Irrespective of whether Rab23 di-rectly affects the localization of only one or more PCPproteins, in the wing Rab23 has a relatively modest effecton protein localization, and, as a consequence, on hairorientation, indicating that Rab23 has a minor or largelyredundant role in this tissue. Interestingly, however,Rab23 induces much stronger trichome orientationdefects on the abdominal cuticle. Although it is notproven formally, genetic analysis suggests that asymmet-ric PCP protein accumulation (or at least polarizedactivation) is likely to occur in the abdominal histoblastcells as well. Hence, with respect to protein polarizationRab23 may act in a tissue-specific manner playing alargely dispensable role in the wing, but having a criticalrole in the abdominal epidermis.

Correct trichome placement at a single distallylocated site is clearly a crucial step in planar polarizationof the wing cells. Current models suggest that prehairinitiation is controlled by an inhibitory cue localizedproximally in a Vang-dependent manner, and by a

Fz-dependent cue that positively regulates hair forma-tion at the distal vertex (Strutt and Warrington

2008). Whereas it is not clear how the distal cues work,with regard to the proximal cues it is known that Vangand Pk colocalize with the effector proteins In, Fy andFrtz that control the localization and activity of Mwh,which is thought to regulate prehair initiation directlyby interfering with actin bundling in the subapicalregion of cells (Strutt and Warrington 2008; Yan

et al. 2008). We found that Rab23 severely impairstrichome placement in the wing leading to the forma-tion of multiple hairs, which indicates a role in therepression of ectopic hair initiation. Where does Rab23fit into the regulatory hierarchy of trichome placement?Our double mutant analysis suggests that Rab23 isupstream of the In group and mwh, and acts at the samelevel as the core PCP genes. The synergistic geneticinteraction between Rab23 and the core PCP mutationsindicates that they function in parallel pathways duringthe restriction of prehair initiation. Remarkably, thepkpk; Rab23 double mutants exhibit an almost identicalphenotype to mutations of the In group, suggestingthat, unless we assume the existence of an In indepen-dent restriction system, Pk and Rab23 together are bothnecessary and sufficient to fully activate the In complex.In pk single mutants the proximal accumulation of In isseverely impaired, yet multiple hairs rarely develop,indicating that proper In localization plays only a minorrole in the restriction mechanism. Conversely, in Rab23single mutants In localization is weakly affected, butmultiple hairs often form, suggesting that the majorfunction of Rab23 is related to In activation. Thus, itappears that the proximally restricted activation of In onthe one hand is ensured by Pk, that mainly plays a role inproper In localization, and on the other hand by Rab23,that seems to be required for In activation. At present,the molecular function of the In system is unknown, andit is therefore also unclear how Rab23 might contributeto the activation of the In complex. Nevertheless,because Rab23 has a weaker multiple hair phenotypethan in, but the pkpk; Rab23 double mutant is nearly asstrong as in, it is conceivable that In activation is notexclusively Rab23 dependent but, beyond a role inprotein localization, Pk has a partial requirement aswell.

The regulation of cellular packing is an interesting,yet only lately appreciated aspect of wing development.It has been reported by Classen et al. (2005) that thewing epithelium is irregularly packed throughout larvaland prepupal stages, but shortly before hair formation itbecomes a quasihexagonal array of cells. Hexagonalrepacking depends on the activity of the core PCPproteins (Classen et al. 2005). However, defects inpacking geometry do not appear to directly perturb hairpolarity in core PCP mutant wing cells. The possibleexception to this rule is pk that exhibits very strong hairorientation defects and induces the strongest packing

Rab23 Regulates Hair Number and Polarization 1063

defects within the core PCP group (Classen et al. 2005;Lin and Gubb 2009). Additionally, another studyrevealed that irregularities in cell geometry are associ-ated with polarity defects in the case of fat mutant clones(Ma et al. 2008). Thus, cell geometry is not the directdeterminant of cell polarity, but in some instances cellpacking seems to impact on PCP signaling and hairorientation. Here we have shown that in the wing Rab23is predominantly involved in the regulation of wing hairnumber, and it is also required for hexagonal packing ofthe wing epithelium. Do these packing defects correlatewith the severity of the multiple hair phenotype? Ourdata argue against this idea for the case of Rab23, andalso for the cases of other strong multiple hair mutants,such as in, frtz, and mwh. Therefore, cell shape has nodirect effect on the regulation of the number of prehairinitiation sites, and Rab23 appears to regulate hexago-nal packing and hair number independently.

As Rab23 and Pk are both required for cellularpacking, and Rab23 associates with Pk, it is possiblethat they cooperate during the regulation of packing.This is in agreement with the observation that pkpk;Rab23 double mutant wings do not show strongerpacking defects than a pkpk single mutant (data notshown). However, other interpretations are also possi-ble, hence further investigations will be required tounderstand how Rab23 and Pk regulates cellular pack-ing and to clarify the impact of packing geometry onPCP establishment in the wing.

Unlike the vertebrate orthologs, Drosophila Rab23 isnot an essential gene and does not appear to regulateHedgehog signaling. Given that Rab GTPases are thoughtto regulate vesicular transport and that mouse Rab23localizes to endosomes (Evans et al. 2003), it was ex-pected that Rab23 regulates the trafficking of vesicle-associated Hedgehog signaling components. However,in the mammalian systems no clear link betweenendocytosis, Rab23, and the subcellular localization ofHedgehog signaling elements has been identified(Evans et al. 2003; Eggenschwiler et al. 2006; Wang

et al. 2006). Our finding that Rab23 associates with Pksuggests that Rab23 might be directly involved in theregulation of Pk trafficking, and therefore Pk could bethe first known direct target of Rab23. Interestingly,there is a significant overlap reported in the embryonicexpression domains of the vertebrate Pk and Rab23genes in the region of the dorsal neural ectoderm, thesomites, and the limb buds (Eggenschwiler et al. 2001;Wallingford et al. 2002; Takeuchi et al. 2003; Veeman

et al. 2003; Li et al. 2007; Cooper et al. 2008). Moreover, itis also known that blocking of Rab23 or Pk function invertebrate embryos can both lead to a spina bifidaphenotype (Eggenschwiler et al. 2001; Wallingford

et al. 2002; Takeuchi et al. 2003; Li et al. 2007). Theseobservations raise the possibility that, unlike the Rab23involvement in Hedgehog signaling, the Rab23–Pkregulatory connection is evolutionarily conserved.

We thank Paul Adler, Barry Dickson, Suzanne Eaton, David Gubb,Jurgen Knoblich, Matthew Scott, Rita Sinka, Tadashi Uemura, TanyaWolff, the Vienna Drosophila RNAi Centre (VDRC), the Bloomingtonand Szeged stock centers, and the Developmental Studies HybridomaBank (DSHB) for providing fly stocks and reagents. We are grateful toHenrik Gyurkovics and Rita Sinka for critical reading and helpfulcomments on the manuscript. We thank the technical assistance ofAnna Rehak, Edina Ordog, Aniko Berente, Olga Kovalcsik, Margit Pal,and Andor Udvardy, and Balazs Dobrovich for help with antibodypreparation. T.M. and J.M. are supported by the Hungarian ScientificResearch Fund (OTKA) (NK 72341), and this work was also supportedby a European Molecular Biology Organization/Howard HughesMedical Institute (EMBO/HHMI) scientist grant (to J.M.).

LITERATURE CITED

Adler, P. N., 2002 Planar signaling and morphogenesis in Drosophila.Dev. Cell 2: 525–535.

Adler, P. N., J. Charlton and J. Liu, 1998 Mutations in the cadher-in superfamily member gene dachsous cause a tissue polarityphenotype by altering frizzled signaling. Development 125:959–968.

Adler, P. N., C. Zhu and D. Stone, 2004 Inturned localizes to theproximal side of wing cells under the instruction of upstream pla-nar polarity proteins. Curr. Biol. 14: 2046–2051.

Axelrod, J. D., 2001 Unipolar membrane association of Dishev-elled mediates Frizzled planar cell polarity signaling. GenesDev. 15: 1182–1187.

Bastock, R., H. Strutt and D. Strutt, 2003 Strabismus is asym-metrically localised and binds to Prickle and Dishevelled duringDrosophila planar polarity patterning. Development 130: 3007–3014.

Behnia, R., and S. Munro, 2005 Organelle identity and the sign-posts for membrane traffic. Nature 438: 597–604.

Bhanot, P., M. Brink, C. H. Samos, J. C. Hsieh, Y. Wang et al.,1996 A new member of the frizzled family from Drosophilafunctions as a Wingless receptor. Nature 382: 225–230.

Casal, J., G. Struhl and P. A. Lawrence, 2002 Developmentalcompartments and planar polarity in Drosophila. Curr. Biol.12: 1189–1198.

Casal, J., P. A. Lawrence and G. Struhl, 2006 Two separatemolecular systems, Dachsous/Fat and Starry night/Frizzled, actindependently to confer planar cell polarity. Development 133:4561–4572.

Classen, A. K., K. I. Anderson, E. Marois and S. Eaton,2005 Hexagonal packing of Drosophila wing epithelial cellsby the planar cell polarity pathway. Dev. Cell 9: 805–817.

Cooper, O., D. Sweetman, L. Wagstaff and A. Munsterberg,2008 Expression of avian prickle genes during early develop-ment and organogenesis. Dev. Dyn. 237: 1442–1448.

Das, G., A. Jenny, T. J. Klein, S. Eaton and M. Mlodzik,2004 Diego interacts with Prickle and Strabismus/Van Goghto localize planar cell polarity complexes. Development 131:4467–4476.

Duncan, I. M., 1982 Polycomblike: a gene that appears to be re-quired for the normal expression of the bithorax and antenna-pedia gene complexes of Drosophila melanogaster. Genetics 102:49–70.

Eggenschwiler, J. T., O. V. Bulgakov, J. Qin, T. Li and K. V. Anderson,2006 Mouse Rab23 regulates hedgehog signaling from smooth-ened to Gli proteins. Dev. Biol. 290: 1–12.

Eggenschwiler, J. T., E. Espinoza and K. V. Anderson,2001 Rab23 is an essential negative regulator of the mouseSonic hedgehog signalling pathway. Nature 412: 194–198.

Emery, G., A. Hutterer, D. Berdnik, B. Mayer, F. Wirtz-Peitz et al.,2005 Asymmetric Rab 11 endosomes regulate delta recyclingand specify cell fate in the Drosophila nervous system. Cell122: 763–773.

Evans, T. M., C. Ferguson, B. J. Wainwright, R. G. Parton and C.Wicking, 2003 Rab23, a negative regulator of hedgehog signal-ing, localizes to the plasma membrane and the endocytic pathway.Traffic 4: 869–884.

1064 C. Pataki et al.

Fanto, M., and H. McNeill, 2004 Planar polarity from flies to ver-tebrates. J. Cell Sci. 117: 527–533.

Feiguin, F., M. Hannus, M. Mlodzik and S. Eaton, 2001 The an-kyrin repeat protein Diego mediates Frizzled-dependent planarpolarization. Dev. Cell 1: 93–101.

Gubb, D., and A. Garcia-Bellido, 1982 A genetic analysis of the de-termination of cuticular polarity during development in Dro-sophila melanogaster. J. Embryol. Exp. Morphol. 68: 37–57.

Guo, A., T. Wang, E. L. Ng, S. Aulia, K. H. Chong et al., 2006 Openbrain gene product Rab23: expression pattern in the adultmouse brain and functional characterization. J. Neurosci. Res.83: 1118–1127.

Ingham, P. W., and A. P. McMahon, 2001 Hedgehog signaling inanimal development: paradigms and principles. Genes Dev. 15:3059–3087.

Jenny, A., R. S. Darken, P. A. Wilson and M. Mlodzik, 2003 Prickleand Strabismus form a functional complex to generate a correctaxis during planar cell polarity signaling. EMBO J. 22: 4409–4420.

Krasnow, R. E., and P. N. Adler, 1994 A single frizzled protein hasa dual function in tissue polarity. Development 120: 1883–1893.

Lawrence, P. A., J. Casal and G. Struhl, 1999 Hedgehog and en-grailed: pattern formation and polarity in the Drosophila abdo-men. Development 126: 2431–2439.

Lawrence, P. A., J. Casal and G. Struhl, 2004 Cell interactionsand planar polarity in the abdominal epidermis of Drosophila.Development 131: 4651–4664.

Lawrence, P. A., G. Struhl and J. Casal, 2007 Planar cell polarity:One or two pathways? Nat. Rev. Genet. 8: 555–563.

Lee, H., and P. N. Adler, 2002 The function of the frizzled pathwayin the Drosophila wing is dependent on inturned and fuzzy.Genetics 160: 1535–1547.

Li, N., J. N. Volff and A. Wizenmann, 2007 Rab23 GTPase is ex-pressed asymmetrically in Hensen’s node and plays a role inthe dorsoventral patterning of the chick neural tube. Dev.Dyn. 236: 2993–3006.

Lin, Y. Y., and D. Gubb, 2009 Molecular dissection of DrosophilaPrickle isoforms distinguishes their essential and overlappingroles in planar cell polarity. Dev. Biol. 325: 386–399.

Ma, D., C. H. Yang, H. McNeill, M. A. Simon and J. D. Axelrod,2003 Fidelity in planar cell polarity signalling. Nature 421:543–547.

Ma, D., K. Amonlirdviman, R. L. Raffard, A. Abate, C. J. Tomlin

et al., 2008 Cell packing influences planar cell polarity signal-ing. Proc. Natl. Acad. Sci. USA 105: 18800–18805.

Matusek, T., R. Gombos, A. Szecsenyi, N. Sanchez-Soriano, A.Czibula et al., 2008 Formin proteins of the DAAM subfamilyplay a role during axon growth. J. Neurosci. 28: 13310–13319.

Mihaly, J., T. Matusek and C. Pataki, 2005 Diego and friends playagain. FEBS J. 272: 3241–3252.

O’Hare, K., and G. M. Rubin, 1983 Structures of P transposable el-ements and their sites of insertion and excision in the Drosophilamelanogaster genome. Cell 34: 25–35.

Rawls, A. S., and T. Wolff, 2003 Strabismus requires Flamingo andPrickle function to regulate tissue polarity in the Drosophila eye.Development 130: 1877–1887.

Santos, E., and A. R. Nebreda, 1989 Structural and functionalproperties of ras proteins. FASEB J. 3: 2151–2163.

Seifert, J. R., and M. Mlodzik, 2007 Frizzled/PCP signalling: aconserved mechanism regulating cell polarity and directed mo-tility. Nat. Rev. Genet. 8: 126–138.

Shimada, Y., T. Usui, S. Yanagawa, M. Takeichi and T. Uemura,2001 Asymmetric colocalization of Flamingo, a seven-pass trans-membrane cadherin, and Dishevelled in planar cell polarization.Curr. Biol. 11: 859–863.

Shimada, Y., S. Yonemura, H. Ohkura, D. Strutt and T. Uemura,2006 Polarized transport of Frizzled along the planar microtu-bule arrays in Drosophila wing epithelium. Dev. Cell 10: 209–222.

Simons, M., and M. Mlodzik, 2008 Planar cell polarity signaling:from fly development to human disease. Annu. Rev. Genet. 42:517–540.

Spoerner, M., C. Herrmann, I. R. Vetter, H. R. Kalbitzer and A.Wittinghofer, 2001 Dynamic properties of the Ras switch Iregion and its importance for binding to effectors. Proc. Natl.Acad. Sci. USA 98: 4944–4949.

Struhl, G., D. A. Barbash and P. A. Lawrence, 1997 Hedgehogorganises the pattern and polarity of epidermal cells in the Dro-sophila abdomen. Development 124: 2143–2154.

Strutt, D., 2003 Frizzled signalling and cell polarisation in Dro-sophila and vertebrates. Development 130: 4501–4513.

Strutt, D., and H. Strutt, 2007 Differential activities of the coreplanar polarity proteins during Drosophila wing patterning. Dev.Biol. 302: 181–194.

Strutt, D., and S. J. Warrington, 2008 Planar polarity genes inthe Drosophila wing regulate the localisation of the FH3-domainprotein Multiple Wing Hairs to control the site of hair produc-tion. Development 135: 3103–3111.

Strutt, D. I., 2001 Asymmetric localization of frizzled and the es-tablishment of cell polarity in the Drosophila wing. Mol. Cell7: 367–375.

Strutt, D. I., U. Weber and M. Mlodzik, 1997 The role of RhoA intissue polarity and Frizzled signalling. Nature 387: 292–295.

Strutt, H., and D. Strutt, 2002 Nonautonomous planar polaritypatterning in Drosophila: dishevelled-independent functions offrizzled. Dev. Cell 3: 851–863.

Takeuchi, M., J. Nakabayashi, T. Sakaguchi, T. S. Yamamoto, H.Takahashi et al., 2003 The prickle-related gene in vertebratesis essential for gastrulation cell movements. Curr. Biol. 13: 674–679.

Tree, D. R., J. M. Shulman, R. Rousset, M. P. Scott, D. Gubb et al.,2002 Prickle mediates feedback amplification to generate asym-metric planar cell polarity signaling. Cell 109: 371–381.

Usui, T., Y. Shima, Y. Shimada, S. Hirano, R. W. Burgess et al.,1999 Flamingo, a seven-pass transmembrane cadherin, regu-lates planar cell polarity under the control of Frizzled. Cell 98:585–595.

Veeman, M. T., D. C. Slusarski, A. Kaykas, S. H. Louie and R. T.Moon, 2003 Zebrafish prickle, a modulator of noncanonicalWnt/Fz signaling, regulates gastrulation movements. Curr. Biol.13: 680–685.

Vetter, I. R., and A. Wittinghofer, 2001 The guanine nucleotide-binding switch in three dimensions. Science 294: 1299–1304.

Wallingford, J. B., T. Goto, R. Keller and R. M. Harland,2002 Cloning and expression of Xenopus Prickle, an ortho-logue of a Drosophila planar cell polarity gene. Mech. Dev.116: 183–186.

Wang, Y., E. L. Ng and B. L. Tang, 2006 Rab23: What exactly does ittraffic? Traffic 7: 746–750.

Winter, C. G., B. Wang, A. Ballew, A. Royou, R. Karess et al.,2001 Drosophila Rho-associated kinase (Drok) links Frizzled-mediated planar cell polarity signaling to the actin cytoskeleton.Cell 105: 81–91.

Wong, L. L., and P. N. Adler, 1993 Tissue polarity genes of Dro-sophila regulate the subcellular location for prehair initiationin pupal wing cells. J. Cell Biol. 123: 209–221.

Xu, T., and G. M. Rubin, 1993 Analysis of genetic mosaics in devel-oping and adult Drosophila tissues. Development 117: 1223–1237.

Yan, J., D. Huen, T. Morely, G. Johnson, D. Gubb et al., 2008 Themultiple wing hairs gene encodes a novel GBD-FH3 domain con-taining protein that functions both prior to and after wing hairinitiation. Genetics 180: 219–228.

Yang, C. H., J. D. Axelrod and M. A. Simon, 2002 Regulation ofFrizzled by fat-like cadherins during planar polarity signalingin the Drosophila compound eye. Cell 108: 675–688.

Zerial, M., and H. McBride, 2001 Rab proteins as membrane or-ganizers. Nat. Rev. Mol. Cell. Biol. 2: 107–117.

Communicating editor: W. M. Gelbart

Rab23 Regulates Hair Number and Polarization 1065

Supporting Information http://www.genetics.org/cgi/content/full/genetics.109.112060/DC1

Drosophila Rab23 Is Involved in the Regulation of the Number and Planar Polarization of the Adult Cuticular Hairs

Csilla Pataki, Tamás Matusek, Éva Kurucz, István Andó, Andreas Jenny and József Mihály

Copyright © 2010 by the Genetics Society of America 10.1534/genetics.109.112060

C. Pataki et al. 2 SI

FIGURE S1.—Rab23 is not required for early cortical polarization of the PCP proteins. (A-B’) Early cortical polarization

of Stan at 6 hours APF is not affected by Rab23. Wild type (A) and Rab2351

(B) pupal wing cells display a similar level of

polarization of Stan at 6 hours APF, schematized in panels A’ and B’, respectively (black lines are drawn between cells with

coherent Stan polarity). Scale bar: 10 μm.

C. Pataki et al. 3 SI

FIGURE S2.—The subcellular distribution of DN and CA YFP::Rab23 in pupal wing cells. (A) In the junctional zone the

wild type YFP::Rab23 protein (in green) is enriched in the plasma membrane of pupal wing cells and displays a weak

diffused staining in the cytoplasm as well. (B) The dominant negative (DN) form of YFP::Rab23 (in green) is partly

membrane associated, but displays a stronger accumulation in the cytoplasm than the wild type protein. (C) The

constitutively active (CA) form of YFP::Rab23 (in green) is enriched in the plasma membrane, whereas the cytoplasmic

staining is somewhat weaker than that of the wild type protein. A-C displays apical sections, while Z axis reconstructions

(made along the white lines indicated on A-C) are shown below these panels. Scale bar: 10 μm.

C. Pataki et al. 4 SI

FIGURE S3.—Genetic interaction studies and double mutant analysis with Rab23 and in. (A) A wild type wing exhibiting

distally pointing trichomes. (B) A Rab23T69A

and a Rab23T69A

/Rab2351

(C) mutant wing exhibiting multiple hairs and

moderately strong hair orientation defects. (D) in1, (E) in

1, Rab23

T69A and (F) in

1, Rab23

T69A/Rab23

51 mutant wings. in

1

dominantly enhances the number of multiple hairs exhibited by Rab23T69A

/Rab2351

(compare C to F), whereas the PCP

defects in regard of wing hair number and orientation are identical in in1 and in

1, Rab23

T69A (compare D to E).

Photomicrographs were taken from the D region of the wing and were positioned the same way as the ones in Fig. 2.

Quantification of wing hair numbers is presented in Fig. 10, P-R.

C. Pataki et al. 5 SI

FIGURE S4.—Rab23 increases the number multiple hairs induced by hs-Fz expression. (A) Schematized wing, multiple

wing hairs were counted in the proximal half of the A region on the dorsal side of the wing (a square with dashed lines

indicates the area). Quantification of wing hair numbers in wings that express hs-Fz is shown on the right. Note that both

Rab23 alleles increase the number of multiple hairs produced. We examined 12-17 wings for each genotype. To test for

significance we applied the t-test as a statistical method, P indicates probability.

C. Pataki et al. 6 SI

FIGURE S5.—Quantification of the average number of trichomes per cell in Rab23, in and frtz mutant combinations. (A)

Schematized wing, trichomes were counted in the C region (area bordered by dashed lines) immediately distal to the anterior

cross vein (acv) until the line of the posterior cross vein (pcv). (B) Rab2351

, (C) frtz1 and (D) frtz

1; Rab23

51 mutant wings.

Wing cells with more than two hairs are indicated with red circles, note that the number of such cells is higher in the double

homozygous mutant combination than in the single mutants. (E) Quantification of the average number of trichomes per cell

in Rab23 and In group mutant combinations. In double mutant wings the average number of trichomes per cell is higher that

in the appropriate single mutants.