PAR3 acts as a molecular organizer to define the

Research Article

4293

PAR3 acts as a molecular organizer to define the apical domain of chick neuroepithelial cells Cristina Afonso1,2 and Domingos Henrique1,2,* 1

Instituto Gulbenkian de Ciência, Ap. 14, Oeiras, Portugal Instituto de Medicina Molecular, Faculdade de Medicina de Lisboa, Av. Prof. Egas Moniz, 1649-028 Lisboa, Portugal

2

*Author for correspondence (e-mail: [email protected])

Journal of Cell Science

Accepted 17 July 2006 Journal of Cell Science 119, 4293-4304 Published by The Company of Biologists 2006 doi:10.1242/jcs.03170

Summary Neural progenitors in the vertebrate nervous system are fully polarized epithelial cells, with intercellular junctions at the apical region. These progenitor cells remain within the neuroepithelium throughout neurogenesis, and will ultimately give rise to all the neurons in the mature nervous system. We have addressed the role of the PAR polarity complex in vertebrate neuroepithelial polarity and show that PAR3 functions as the initial scaffold to assemble and organize the PAR complex at the apical region of neuroepithelial cells, coordinating also the recruitment of additional polarity complexes and junction-associated proteins to the same region, while restricting other polarity

Introduction The establishment of cell polarity involves the asymmetric localization of proteins and other molecules at distinct membrane domains and cellular structures, organizing the cell along a defined polarity axis. The PAR proteins are involved in organizing this polarity and were first identified in C. elegans, playing a fundamental role in polarizing the one-cell stage embryo (Kemphues, 2000). One of these proteins, PAR3, is conserved in nematodes, insects and vertebrates and has been shown to associate with two other proteins, PAR6 and atypical protein kinase C (aPKC), forming the core of a multiprotein complex designated the PAR polarity complex (Ohno, 2001; Henrique and Schweisguth, 2003). In mammalian and Drosophila epithelial cells, the PAR complex localizes to cellular junctions at the apical-most tip of the lateral membrane, where it regulates a network of molecular complexes that act together to define the apicobasal polarity of epithelial cells (Macara, 2004; Suzuki and Ohno, 2006). In addition, the PAR complex controls the asymmetric divisions of early C. elegans embryos and Drosophila neuroblasts (Wodarz, 2005), also playing an important role during specification of axonal processes in mammalian neurons (Arimura and Kaibuchi, 2005). During these processes, the correct localization and activity of the complex seem to depend on PAR3, which has been suggested to function as a scaffold to promote the assembly of the whole complex. A subsequent interaction between PAR6 and CDC42-GTP can activate aPKC by relieving it from PAR6mediated inhibition (Yamanaka et al., 2001). PAR3 is itself a target of aPKC and dissociates from the complex when phosphorylated (Nagai-Tamai et al., 2002), allowing aPKC to

proteins to the basolateral membrane. We propose that PAR3 acts as a molecular organizer to connect the acquisition of apico-basal polarity with the positioning and formation of junctional structures in neuroepithelial cells, a function of upmost importance for the morphogenesis of embryonic neural tissue and the process of neurogenesis. Supplementary material available online at http://jcs.biologists.org/cgi/content/full/119/20/4293/DC1 Key words: PAR polarity complex, Neuroepithelium, Apical junctional complexes, Cell adhesion

act on other substrates, in a cascade of signalling events that results in the segregation of different molecules to distinct membrane domains and in the correct organization of junctional structures within epithelial cells. Interference with the normal function of the PAR complex, as happens in Drosophila bazooka (encoding the fly PAR3 protein) and par6 mutants (Wodarz et al., 2000; Hutterer et al., 2004; Harris and Peifer, 2005) or in MDCK cells expressing a dominantnegative mutant of aPKC (Suzuki et al., 2001), leads to disruption of cellular junctions and profound alterations in epithelial tissue architecture. The molecular mechanisms underlying these activities of the PAR complex are not completely understood but other protein complexes are known to interact functionally with the PAR complex during the establishment of epithelial polarity. One of these is composed by the proteins Crumbs (CRB), Stardust (PALS1) and PALS-associated tight junction protein (PATJ), and is also localized to epithelial junctions, with PALS1 directly associating with PAR6 in a CDC42-dependent manner (Macara, 2004). A different group of proteins is localized basolaterally in epithelial cells, including Scribble (SCRIB), Discs Large (DLG) and Lethal Giant Larvae (LGL), which also contribute to the establishment of epithelial cell polarity and adhesion (Bilder, 2004). The LGL protein can interact directly with PAR6 and this might result in its inactivation by aPKCmediated phosphorylation (Betschinger et al., 2003; Yamanaka et al., 2003). Furthermore, LGL controls the asymmetric localization of Numb during divisions of neural precursors in Drosophila (Ohshiro et al., 2000; Peng et al., 2000; Langevin et al., 2005), in a process where PAR3 seems to also play a crucial role (Bellaiche et al., 2001; Roegiers et al., 2001).

Journal of Cell Science

4294

Journal of Cell Science 119 (20)

A dynamic crosstalk between these polarity complexes is involved in the establishment of epithelial polarity in both Drosophila and mammalian cells (Macara, 2004; Suzuki and Ohno, 2006). Although the biochemical details are still incomplete, genetic data indicate that the SCRIB group of proteins antagonizes the apical polarizing activity initiated by the PAR complex, whereas the CRB complex counteracts the activity of the SCRIB group. To maintain the apical membrane character (Bilder et al., 2003; Tanentzapf and Tepass, 2003). These interactions lead to the segregation of the various polarity proteins to distinct membrane domains, with LGL, DLG and SCRIB being excluded from the more apical domain where the PAR and CRB complexes reside. During neurogenesis in the Drosophila embryo, some epithelial cells are selected to become neuroblasts, which subsequently delaminate from the ectoderm and divide asymmetrically to generate various neural cell types (Betschinger and Knoblich, 2004). Although neuroblasts are no longer epithelial cells, they maintain apico-basal polarity, with the PAR complex localized apically. This complex serves as polarity cue to orient neuroblast divisions, leading to the unequal segregation of determinants between the two daughter cells and the consequent acquisition of distinct fates. Thus, the PAR complex in Drosophila ensures that the distinctive polarity of epithelial cells is translated into molecular asymmetries within delaminating neuroblasts, providing a functional link between epithelial polarity and neurogenesis. In vertebrates, neurogenesis occurs in a pseudostratified epithelium and, in contrast to Drosophila neuroblasts, neural progenitors are fully polarized epithelial cells, with apically localized intercellular junctions and basal endfeet contacting the basal lamina (Chenn and McConnell, 1995; Aaku-Saraste et al., 1996). Apical Junctional Complexes (AJCs) in vertebrate neuroepithelial (NEP) cells seem to be composed mainly of cadherin-based adherens junctions (AJs) (Aaku-Saraste et al., 1996), which play a structural role in maintaining epithelial integrity and contribute to organize the characteristic apicobasal polarity of these cells (Chenn et al., 1998). The presence of PAR3, PAR6 and aPKC proteins at the apical region of mouse and zebrafish NEP cells (Manabe et al., 2002; Geldmacher-Voss et al., 2003; Wei et al., 2004) suggests that the PAR polarity complex participates also in the establishment of apico-basal polarity in the vertebrate neuroepithelium. Furthermore, loss-of-function mutations in aPKC or morpholino-based inactivation of pard3 in zebrafish embryos (Wei et al., 2004; Koike et al., 2005; Imai et al., 2006) have been shown to disturb apico-basal polarity and cause disassembly of cellular junctions in NEP cells, leading to alterations in neural tissue architecture. Although these experiments show that the PAR complex is necessary for the establishment of NEP polarity and formation of junctional structures, the severe disruption of the neuroepithelium has precluded a detailed analysis of the molecular mechanisms by which the PAR complex regulates polarity of NEP cells and how it contributes to the normal architecture of embryonic neural tissue. Here, we address the role of the PAR complex in the chick embryo, using a gain-of-function approach to follow the dynamic activity of various members of the complex, when ectopically expressed in NEP cells. Our results show that cPAR3 plays a central role in directing the recruitment of various polarity and junction-associated proteins to the cortical

plasma membrane of NEP cells, excluding other polarity proteins from the same membrane region. We propose that cPAR3 functions as a molecular organizer that connects the acquisition of apico-basal polarity with the formation of junctional structures in NEP cells, thus contributing for the proper morphogenesis of neural tissue. Results The PAR polarity complex is localized at the apical region of the chick embryonic neuroepithelium We have cloned and characterized two chick par3 genes, named as cpar3 and cpar3L (Fig. 1A). A cpar3 full-length cDNA clone encodes a 1,352 amino acid protein highly homologous to the described mammalian PAR3 proteins (Izumi et al., 1998; Joberty et al., 2000; Lin et al., 2000). Only a partial cDNA could be obtained for cpar3L, encoding the N-terminal 860 amino acids of cPAR3L, with a high homology to human PAR3L (Gao et al., 2002; Kohjima et al., 2002). The two cpar3 genes have partial overlapping expression patterns in the embryo, with cpar3 showing higher expression in the developing NE (see supplementary material Fig. S1). This, and the fact that we could not obtain a fulllength cDNA for cpar3L, led us to focus our studies on cpar3. A full-length cpar6 cDNA was also characterized, which encodes a protein with high homology to mouse PAR6B (Joberty et al., 2000). To determine the subcellular localization of the PAR complex in embryonic NEP cells, a combination of specific antibodies and low-level expression of fluorescent protein fusions was used. To validate this approach, we first established that in ovo electroporation of the chick neural tube (NT) with a vector encoding a cPAR3::GFP protein at low DNA concentration (0.1-0.2 ␮g ␮l–1), results in a subcellular distribution of the fusion protein identical to that revealed by immunofluorescence (IF) with anti-PAR3 antibodies (Fig. 1Ba,b), without causing any obvious cellular defects. Using this approach, we found that cPAR3, cPAR6 and aPKC are localized in the apical-most region (close to the luminal surface) of NEP cells (Fig. 1Ba-e), similar to what has been described in mouse embryos (Manabe et al., 2002). However, confocal microscopy analysis reveals that there is no overlap between the localization of cPAR3 and its binding partners cPAR6 or aPKC: while colocalization between cPAR6 and aPKC is always found (Fig. 1Be), cPAR3 is present in a more basal domain relative to the duo cPAR6/aPKC (Fig. 1Bad). These results are different from those reported for the mouse embryonic telencephalon, where mPAR3 was described as colocalizing with mPAR6 and aPKC at the apical region of NEP cells (Manabe et al., 2002). This disparity might be due to a lower microscopic resolution of the mouse data or, alternatively, might indicate that the localization of the PAR complex proteins is different between mouse and chick embryos. We have also analyzed the distribution of other polarity complexes in chick NEP cells, using the same approach. Our results show that cPALS1 and cCRB3 are located at the apical region of the neuroepithelium (Fig. 1Da,b), colocalizing with cPAR6 (Fig. 1Dc) above the site of cPAR3 accumulation. By contrast, LGL1 is located to the lateral membrane of NEP cells and is excluded from the apical domain where cPAR3 accumulates (Fig. 1Ea).

PAR3 and neuroepithelial polarity

4295

Journal of Cell Science

Fig. 1. The PAR polarity complex is localized at the apical membrane of NEP cells. (A) Deduced protein sequence of cPAR3, depicting the aPKC, KIF3A and TIAM1 binding domains (aPKC-BD; KIF3A-BD; TIAM1-BD). (B) Endogenous cPAR3 localizes apically in NEP cells (a). cPAR3 localizes basally to cPAR6 (b,c) and aPKC (d). cPAR6 colocalizes with aPKC (e). (C) cPAR3 colocalizes with N-Cadherin, Afadin and ␤Catenin at the apical tip of NEP cells (a-c). cPAR6 and aPKC are located apically relative to NCadherin (d,e). (D) cPAR3 is localized basally relative to both cPALS1 and cCRB3 (a,b). cPAR6 colocalizes apically with cCRB3 (c). (E) LGL1 localizes basolaterally and is absent from the apical cPAR3 accumulation domain (a). (F) Schematic representation of the localization of polarity proteins in NEP cells. L, lumen of the NT. Bars in (Ba,b,d,e;C;Db,c;E), 2 ␮m; bar in (Bc), 3 ␮m; bar in (Da), 1 ␮m.

Together, these results are consistent with a direct interaction between cPALS1, cCRB3 and cPAR6 in NEP cells, as previously described for other cell types (Hurd et al., 2003; Lemmers et al., 2004). Additionally, the mutually exclusive distribution of cPAR3 and LGL1 is consistent with published biochemical evidence, which shows that PAR3 and LGL compete for binding to PAR6/aPKC (Yamanaka et al., 2003). The localization of all these proteins in chick NEP cells is summarized in Fig. 1F. cPAR3 is present at adherens junctions in neuroepithelial cells To determine how the two distinct apical domains of PAR protein accumulation are related to junctional structures in NEP cells, the distribution of various AJ components was analyzed and compared with that of cPAR3, cPAR6 and aPKC. Our results show that the AJ proteins N-Cadherin, ␤-Catenin and Afadin are concentrated at the apical tip of NEP cells, in the same domain as cPAR3 (Fig. 1Ca-c), but with no detectable colocalization with cPAR6/aPKC (Fig. 1Cd,e). This suggests that cPAR3 is a component of AJs in NEP cells, while the duo cPAR6/aPKC occupies a more apical membrane domain. Furthermore, co-IP assays reveal that N-Cadherin and cPAR3 are present in the same macromolecular complex (see supplementary material Fig. S1D), confirming that cPAR3 is an AJ-associated protein.

cPAR3 colocalizes with cNumb at the apical region of neuroepithelial cells In Drosophila neuroblasts and SOPs, the LGL protein controls the asymmetric accumulation of Numb, in a process that depends on the activity of the PAR polarity complex (Betschinger et al., 2003; Langevin et al., 2005). In the mouse brain, Numb was reported to be present at the apical surface of NEP cells (Zhong et al., 1996), a localization that seems to depend on the activity of the mouse lgl homolog, mlgl1 (Klezovitch et al., 2004). However, in the chick neural tube, Numb has been described to accumulate in a basal crescent in mitotic NEP cells, with no detectable apical accumulation (Wakamatsu et al., 1999). Our experiments, using different fixation conditions and several distinct antibodies, could not confirm the basal accumulation of cNumb in mitotic cells, revealing instead a clear apical localization in NEP cells, as well as accumulation in cytoplasmic vesicles (Fig. 2A,B). Confocal microscopy analysis reveals colocalization of cNumb with cPAR3 at the apical membrane of NEP cells (Fig. 2C-C⬙), raising the hypothesis that the two proteins are present in the same complex. To test this, IP assays were performed with antiNumb and anti-PAR3 antibodies, which revealed that the two proteins are part of the same protein complex (Fig. 2D,E). Furthermore, immunostaining with antibodies against NCadherin and ␤-Catenin show that apical cNumb colocalizes

4296

Journal of Cell Science 119 (20)

Fig. 2. cPAR3 colocalizes with cNumb at intercellular junctions. cNumb accumulates at the apical membrane (arrowhead in A) and at cytoplasmic vesicles (arrowheads in B). (C-C⬙) cPAR3 colocalizes with cNumb at the apical membrane of NEP cells. (D) cPAR3 is coimmunoprecipitated with cNumb in NEP protein lysates. CRMP2 is used as control. Conversely, cNumb co-immunoprecipitates with cPAR3 (E). (F) cNumb accumulates at the apical membrane together with ␤Catenin and N-Cadherin (G). L, lumen of the NT. Magnification in (A), 20⫻. Bar in (B), 6 ␮m; bars in (C-C⬙), 3 ␮m; bars in (F,G), 2 ␮m.

Journal of Cell Science

with both proteins (Fig. 2F,G), suggesting that cNumb is also an AJ-associated protein. Ectopic cPAR3 accumulates in punctate clusters at the cortical plasma membrane of neuroepithelial cells To study the function of the PAR complex in NEP cells, a gain-of-function approach was used to generate ectopic localization of the cPAR3 and cPAR6 proteins, and evaluate the effects on the cellular architecture of the neuroepithelium. Electroporation of chick embryos with high concentrations of the cPAR3::GFP or GFP::cPAR6 vectors allowed us to follow the spatio-temporal accumulation of the ectopic proteins in NEP cells and the effects they have on the dynamics of endogenous polarity proteins. While ectopic cPAR6 accumulates diffusely in the cell and has no detectable effect on the localization of endogenous polarity proteins (data not shown), ectopic cPAR3 accumulates in discrete clusters along the lateral plasma membrane of NEP cells (Fig. 3A,B) and causes a

Fig. 3. Ectopic cPAR3 accumulates in punctate clusters at the lateral membrane of NEP cells. (A) 12 hours post-electroporation, cPAR3::GFP accumulates in clusters along the plasma membrane of NEP cells (arrows) and also apically (arrowhead). (B) A single electroporated cell with multiple clusters of cPAR::GFP accumulation along the plasma membrane. (C) At 24 hours, cells with high levels of cPAR3 accumulate at the basal region of the NT (arrow) and ectopic mitotic cells (red, pHistH3positive) are also present in this region (arrowheads in D). (E) At 48 hours, cPAR3 accumulates in rosettelike structures at the basal part of the NE (square), with punctate clusters in other cells (arrow). (F,F⬘) Detail of a rosette-like group of cells [outlined by the dotted line; enlarged from square in (D)], with cPAR3 accumulating at the center (arrow). (G) The cPAR3-⌬N::GFP mutant still localizes apically (arrowhead), but does not form punctate clusters. L, lumen of the NT. Magnification in (A,D), 40⫻; in (C,E,G), 63⫻. Bars in (B) and (F,F⬘), 10 ␮m.

redistribution of other polarity proteins (see below). Cells with ectopic cPAR3::GFP clusters concentrate in the basal region of the neuroepithelium 24 hours postelectroporation (Fig. 3C) and some of these cells undergo mitosis (Fig. 3D). Later, at 48 hours, cells with cPAR3::GFP punctates are still accumulated basally (Fig. 3E) but, in addition, we could detect the formation of rosette-like groups of cells in this region of the neuroepithelium. cPAR3::GFP concentrates at the center of these rosettes (Fig. 3F,F⬘) and is almost absent from the remaining plasma membrane. Overexpression of untagged cPAR3, followed by detection with anti-PAR3 antibodies, gave essentially the same results (data not shown). To address whether formation of cPAR3 punctate clusters results from protein oligomerization (Benton and St Johnston, 2003; Mizuno et al., 2003), a form of cPAR3 without the CR1 conserved region (cPAR3-N::GFP) was overexpressed in the chick embryonic neural tube. The results show that this truncated protein is still partially localized at the apical tip of

PAR3 and neuroepithelial polarity

Journal of Cell Science

NEP cells (Fig. 3G), but is unable to form punctate accumulations at the membrane, being distributed homogeneously in the cytoplasm (compare Fig. 3G with 3C,D). Thus, the CR1 region is essential to promote cPAR3 oligomerization, although it is dispensable for the apical localization. Cells overexpressing cPAR3 are misplaced within the neuroepithelium During neurogenesis, NEP cells undergo a characteristic interkinetic nuclear movement (INM) (Fujita, 1960), associated with the various phases of cell cycle. At S phase, the nucleus is located in a more basal position and, as cells enter G2, it moves apically and reaches the luminal surface, where cells divide. After division, the nucleus of the two daughter cells moves basally, as these cells enter S phase. Surprisingly, cPAR3-overexpressing cells in mitosis can be detected at the basal region of the neuroepithelium, contrasting with the normal, non-electroporated half of the neural tube, where mitosis always occur close to the lumen. The number of mitotic cells is very similar in the two sides of the neural tube (data not shown), discarding the hypothesis that cPAR3 overexpression might cause cell cycle alterations. Another hypothesis is that cPAR3 overexpression might arrest the normal INM of NEP cells, resulting in cells entering mitosis at ectopic positions in the basal region of the neuroepithelium. To test this, we followed cPAR3-overexpressing cells by live-

Fig. 4. cPAR3 overexpression affects positioning of dividing cells within the NE. (A) Normal NEP cells, expressing a membranetethered GFP, move their nucleus apically and always divide near the luminal surface (red and white arrowheads). (B) Illustrative example (n=12 cells) of a GFP-gpi NEP cell overexpressing cPAR3 dividing away from the apical surface (red arrow). Bar in (A), 10 ␮m; (B), 20 ␮m.

4297

imaging confocal analysis. A neural tube slice culture system was developed that allows visualization of NEP cells along the various phases of their INM, through incorporation of a GPIanchored GFP (GFP-gpi) at the plasma membrane. To assess the effects of cPAR3 overexpression, a group of embryos (n=12) was electroporated with a mixture of GFP-gpi vector and a vector encoding a non-tagged cPAR3 protein, while control embryos (n=12) were electroporated with the GFP-gpi vector only. Our results show that, in control embryos, NEP cells undergo normal INM and always enter mitosis at the luminal surface (Fig. 4A, see supplementary material Movie 1). However, in embryos electroporated with cPAR3, overexpressing cells show increased adhesiveness and seem ‘trapped’ within the NE, unable to accomplish INM. As a consequence, when these cells enter mitosis, the nucleus is still basally located, far from the luminal surface (Fig. 4B, see supplementary material Movie 2). cPAR3 drives local accumulation of junction-associated proteins at the neuroepithelial cell membrane This behavior of cPAR3-overexpressing cells might involve the formation of new intercellular adhesion sites, primed by the ectopic accumulation of cPAR3 and subsequent recruitment of other proteins that control the assembly of junctional complexes. In support of this hypothesis, ectopic N-Cadherin is readily detectable in cPAR3-overexpressing cells 12 hours

Journal of Cell Science

4298

Journal of Cell Science 119 (20)

Fig. 5. Ectopic cPAR3 is sufficient to mislocalize polarity and junction-associated proteins in NEP cells. At the electroporated side of the NT, N-Cadherin accumulates ectopically (white bracket in A,A⬘), colocalizing with cPAR3::GFP punctates (arrowheads in B,B⬘). The same applies to ␤-Catenin (arrowheads in C-D⬘) and aPKC (arrowheads in E-F⬘). Triple-labelling of cPAR3, cPAR6 and aPKC shows colocalization at ectopic clusters (G,G⬘). Members of the CRB complex also mislocalize to cPAR3 punctates (arrowheads in H-K⬘). (L) LGL1 is excluded from the ectopic sites of cPAR3 accumulation at the lateral membrane (arrowheads). b, basal side of the NE. Magnification in (A,A⬘,C,C⬘,E,E⬘,H,H⬘,I,I⬘), 40⫻. Bars in (B,B⬘,G,G⬘), 2 ␮m; bars in (D,D⬘,F,F⬘,J,J⬘,K,K⬘), 10 ␮m; bar in (L), 1 ␮m.

after electroporation, being distributed along the cell membrane (Fig. 5A-B⬘), instead of normally concentrated at the apical tip of NEP cells. Confocal analysis shows a clear correspondence between cPAR3::GFP punctates and NCadherin accumulation (Fig. 5B,B⬘), suggesting that cPAR3 is able to recruit N-Cadherin. Consistent with this, coIP studies show that N-Cadherin and cPAR3 are normally present in the same macromolecular complex in NEP cells (see supplementary material Movie 1). In addition, ␤-Catenin is also seen to accumulate ectopically in cPAR3-overexpressing cells, showing a good correlation with ectopic cPAR3::GFP localization (Fig. 5C-D⬘). However, not all cPAR3 puncta show ectopic accumulation of ␤-Catenin (and of other recruited proteins, see below), which might be due to limited amounts of available endogenous proteins. Together, these results indicate that cPAR3 overexpression interferes with the distribution of junction-associated proteins and drives the formation of ectopic accumulations of both NCadherin and ␤-Catenin, thereby triggering the formation of ectopic adhesion structures in NEP cells. cPAR3 coordinates the recruitment of other polarity proteins to the membrane of neuroepithelial cells The recruitment of junctional proteins to ectopic positions and the subsequent establishment of new adhesion sites is likely to

involve the concerted activity of other polarity proteins which function together with PAR3 to control apico-basal polarity in epithelial cells. We therefore asked whether PAR6, aPKC, and members of other polarity complexes are also present at ectopic sites of cPAR3 accumulation. Our results reveal that both aPKC and PAR6 are ectopically distributed in cPAR3overexpressing cells, overlapping with some of the membrane associated cPAR3::GFP punctates (Fig. 5E-G⬘). Similarly, cCRB3 and cPALS1 are also mislocalized in cPAR3overexpressing cells (Fig. 5H-I⬘), with a subcellular distribution that coincides with ectopic cPAR3 clusters (Fig. 5J-K⬘). We thus conclude that cPAR3 is able to trigger the assembly of apical polarity complexes in NEP cells, implying that it normally functions as a scaffold to coordinate the ordered recruitment of various polarity proteins to the cortical membrane. This function is unique to cPAR3 and depend on its capacity to oligomerize, as overexpression of cPAR6B or cPAR3-N::GFP in NEP cells does not cause any of the molecular or cellular defects due to ectopic cPAR3 (data not shown). Other polarity proteins are known to be involved in defining the basolateral membrane domain of epithelial cells, like LGL (Wodarz, 2005). As described above, LGL1 is normally distributed along the lateral membrane of chick

Journal of Cell Science

PAR3 and neuroepithelial polarity

4299

Fig. 6. PAR complex formation and cNumb subcellular distribution are affected by the presence of excess cPAR3. (A) Western blot shows that overexpression (OE) produces an 80⫻ increase in of cPAR3 levels. (B) Endogenous cPAR6 and aPKC levels are unaffected when cPAR3 is overexpressed. (C,D) Overexpression of cPAR3 recruits more aPKC molecules to the PAR complex. (E) cPAR6/aPKC complex formation is increased in the presence of excess cPAR3. (F) cLGL1 and cNumb protein levels are unaffected when cPAR3 is overexpressed. (G-G⬙) At 24 hours post-electroporation, cNumb accumulates at the apical region (arrowhead) and in ectopic punctates (arrow). (H,H⬘) Higher magnification of square in (G⬙) shows that cNumb colocalizes with ectopic cPAR3 punctates (arrowheads), with cNumb vesicles accumulating at the cell periphery (upper half of the image). In the bottom half of the image, showing non-electroporated cells, cNumb-positive vesicles are uniformly distributed in the cytoplasm. Magnification in (G-G⬙), 20⫻. Bars in (H,H⬘), 6 ␮m.

NEP cells and is excluded from the apical cPAR3 domain. In cPAR3-overexpressing cells, we observed that LGL1 is also excluded from the ectopic cPAR3 punctates at the cortical plasma membrane (Fig. 5L). This effect is likely mediated by the activity of aPKC at the ectopic cPAR3 clusters, with LGL1 being phosphorylated and excluded from the membrane, as it happens in Drosophila epithelial cells (Hutterer et al., 2004). Taken together, these results suggest that ectopic activity of cPAR3 leads to the establishment of new membrane domains with apical character, at the sites where ectopic apical polarity complexes assemble and from where LGL is excluded. cPAR3 promotes the formation of ectopic cPAR3/cPAR6/aPKC complexes To address whether the effects on the localization of apical and basolateral proteins are due to an increase in their synthesis or to a redistribution of the pre-existing pool, the various proteins were quantified by western blot. Our results show that there is no significant increase in the amount of endogenous polarity proteins, despite an 80-fold increase in cPAR3 levels (Fig. 6A,B). Nevertheless, the presence of more cPAR3 molecules in NEP cells seems to promote formation of additional PAR complexes, as shown by IP with anti-PAR6 antibody, followed by western blot with the anti-aPKC antibody (Fig. 6E). This is

further supported by IP experiments with anti-PAR3 and antiaPKC antibodies, which reveal the presence of more aPKC molecules in a complex with cPAR3 (Fig. 6C,D). Taken together, these results indicate that excess cPAR3 drives formation of additional PAR complexes in NEP cells, and show that ectopic localization of cPAR6 and aPKC is a consequence of protein mislocalization, involving recruitment by cPAR3. Similarly, western blot data confirm that LGL1 mislocalization (and cNumb, see below) is due to protein redistribution in NEP cells (Fig. 6F). Ectopic cPAR3 causes a redistribution of cNumb in neuroepithelial cells LGL phosphorylation by aPKC is important to position Numb correctly at the basal pole of Drosophila neuroblasts (Betschinger et al., 2003). Furthermore, the asymmetric localization of Numb during SOP division is controlled by BAZ (Bellaiche et al., 2001; Roegiers et al., 2001), with the LGL protein being also involved in this process (Langevin et al., 2005). Given the defects in LGL1 distribution caused by cPAR3 overexpression, we asked whether cPAR3 also controls cNumb localization in NEP cells. Analysis of embryonic neural tubes where cPAR3::GFP was overexpressed revealed that there is a clear alteration in cNumb distribution, which becomes highly concentrated at the apical membrane (Fig. 6G-

4300

Journal of Cell Science 119 (20)

G⬙), and appears also as ectopic punctates in cells at the basal region of NE (Fig. 6G-G⬙). Furthermore, confocal analysis reveals a striking accumulation of cNumb at the membrane of cPAR3-overexpressing cells, with a clear overlap with ectopic cPAR3 clusters (Fig. 6H,H⬘). Concurrently, a reduction of cytoplasmic cNumb is observed (compare upper and bottom halves of Fig. 6H,H⬘) which, since the overall intracellular concentration of cNumb is not affected (Fig. 6F), suggests that ectopic cPAR3 causes a redistribution of cNumb from cytoplasmic vesicles to the cell membrane. Taken together, these results show that cPAR3 regulates the subcellular localization of cNumb in NEP cells. This might involve an interaction between the two proteins, a hypothesis supported by our finding that cPAR3 and cNumb can be detected in the same molecular complex in NEP cells (Fig. 2D,E).

Journal of Cell Science

cPAR3 overexpression causes the formation of neuroepithelial rosettes Forty eight hours after electroporation, cPAR3-overexpressing embryos show a dramatic disruption of the neuroepithelium.

Ectopic mitosis is often detected and BrdU incorporation indicates the presence of several proliferating cells in the basal region, which normally should contain only post-mitotic neurons (Fig. 7A,A⬘). The most striking phenotype is the appearance of rosettes in the basal region of the NE (Fig. 3E), formed by groups of cells organized concentrically around a lumen (Fig. 3F,F⬘). In rosette cells, cPAR3 concentrates in the region close to the ‘luminal’ surface, together with aPKC and cPAR6 (Fig. 7B-B⬙). This suggests that a defined apical domain is present in rosette cells, which is supported by the presence of centrosomes close the same membrane domain (Fig. 7C,C⬘). N-Cadherin and ␤-Catenin also occupy the region closest to the luminal space, colocalizing with cPAR3 (Fig. 7D-E⬘), implying that junctional complexes are present at the apical region of rosette cells and might function to organize the cells around the central lumen. In this aspect, rosette cells behave like normal NEP cells, forming polarized structures similar to neural tubes. To test whether rosette cells are also engaged in neurogenesis, the expression of various members of the Notch

Fig. 7. Characterization of rosette-like structures in the NE of cPAR3-overexpressing embryos. (A,A⬘) Cycling cells (BrdU+) are present outside the VZ (bars). (B-B⬙) cPAR3 is localized in a central ring inside the rosettes, together with cPAR6/aPKC. (C,C⬘) Rosette cells (dotted line) contain centrosomes close to the central lumen (dotted circle). (D,D⬘) ␤-Catenin (arrowhead) and N-Cadherin (arrowhead in E,E⬘) localize with cPAR3::GFP at the center of rosettes. (F,G) cPAR3 overexpression (left side of the NT) causes an expansion of the cNotch1 and ches5-1 expression domains (white brackets). (H) Ectopic cDelta1 expressing cells can be detected in the basal region of the cPAR3-electroporated NT (white brackets). (I,I⬘) Rosette cells express ches5-1. (J,J⬘) TUJ1- and HU-positive neurons (K,K⬘) accumulate outside rosettes. Magnification in (A,A⬘), 40⫻; magnification in (F-H), 20⫻. Bars in (B-B⬙), 4 ␮m; bars in (C-D⬘,J,J⬘), 10 ␮m; bars in (I,I⬘), 5 ␮m; bars in (K,K⬘), 20 ␮m.

PAR3 and neuroepithelial polarity

Journal of Cell Science

pathway was examined in cPAR3-overexpressing embryos, using in situ hybridization. Expression of cNotch1 and ches51, two genes active in neural progenitors (Fior and Henrique, 2005), could be detected in numerous cells in ectopic positions outside the ventricular zone (VZ) (Fig. 7F,G), including rosette cells (Fig. 7I,I⬘). In addition, the cDelta1 gene, which is expressed in nascent neurons, is also transcribed in some rosette cells (Fig. 7H). Together, this indicates that the Notch pathway is active in rosette cells, some of which stop cycling and start to differentiate. This is confirmed by the presence of neuronal markers (HU and TUJ1) in cells at the periphery of rosettes (Fig. 7J-K⬘), some of them still maintaining an apical process (Fig. 7J). In conclusion, these results show that rosette cells are engaged in neurogenesis, with newborn neurons accumulating outside the rosette environment, in the same way that differentiating neurons in the neural tube migrate out of the VZ and accumulate in the mantle layer. Discussion In this work, we address the central role of PAR3 in defining the apico-basal polarity of NEP cells in chick embryos. Our data support a model in which PAR3 functions as the initial scaffold to assemble the PAR polarity complex at the apical region of NEP cells, coordinating also the recruitment of additional polarity complexes and junction-associated proteins to the same region, while restricting other polarity proteins to the basolateral membrane. We propose that PAR3 acts as a molecular organizer to connect the acquisition of apico-basal polarity with the positioning and formation of junctional structures in NEP cells, a function of upmost importance for the correct morphogenesis of embryonic neural tissue and the subsequent process of neurogenesis. cPAR3 as a molecular organizer of apico-basal polarity in neuroepithelial cells The role of the PAR complex in controlling the establishment and maintenance of cell polarity in different cell types, from various organisms, is well established (Henrique and Schweisguth, 2003; Macara, 2004). The main biochemical output of the PAR complex is provided by aPKC activity, whereas PAR6 regulates aPKC in a CDC42-dependent manner and PAR3 is usually seen as a structural scaffold. However, analysis of protein localization during polarization of the C. elegans one-cell embryo suggests that PAR3 might play a more active role in localizing the PAR complex (Cuenca et al., 2003), and recent results in Drosophila reveal that PAR3 acts upstream of aPKC and PAR6 as an essential polarity cue during epithelial cellularization (Harris and Peifer, 2004; Harris and Peifer, 2005). Furthermore, in MDCK cells PAR3 has been shown to act independently of aPKC/PAR6 during epithelial polarization, regulating the activity of the GTP exchange factor TIAM1 to control actin dynamics (Chen and Macara, 2005). Thus, the contribution of PAR3 to the mechanisms that generate cellular polarity is still not totally known. Although loss-of-function experiments in zebrafish embryos indicate that PAR3 is necessary for the correct morphogenesis of neural tissue, analysis of these mutants could not reveal the specific role of PAR3 in the polarity mechanisms, due to the severe disorganization of the neuroepithelium and widespread NEP cell death (Geldmacher-Voss et al., 2003; Wei et al., 2004).

4301

The data presented in this paper, using ectopic expression of cPAR3 in chick NEP cells, unveil a new perspective on the function of PAR3 during establishment of NEP polarity. We show that PAR3 is able to oligomerize and form punctate clusters at the cortical plasma membrane of NEP cells, where it can nucleate new PAR complexes by recruiting endogenous cPAR6 and aPKC. Biochemical analysis confirms that cPAR3 ectopic expression results in higher number of PAR complexes within the cell and that this is due to a redistribution of endogenous cPAR6 and aPKC. Recruitment of aPKC to ectopic cPAR3 sites is most likely mediated by cPAR6, since overexpression of a mutant cPAR3 without the putative aPKCbinding domain still results in the accumulation of aPKC (and cPAR6) at cPAR3 punctates (data not shown), also indicating that the ability of cPAR3 to nucleate the PAR complex is independent of its phosphorylation state. The PAR complex seems to be highly dynamic, disassembling quickly after its initial formation so that the constituent proteins can subsequently interact with other polarity complexes. This is suggested by our finding that cPAR3 and the duo cPAR6/aPKC are localized to distinct apical membrane domains in NEP cells, with the assembled PAR complex being detected only by the more sensitive coIP assays (and at very low levels). Similar observations have been recently reported in the forming epithelia of the Drosophila embryo (Harris and Peifer, 2005) and imply that the interaction between PAR3 and the duo PAR6/aPKC is indeed very transient, with the proteins localizing to different membrane domains after dissociation of the complex. Phosphorylation of cPAR3 by aPKC might be responsible for such dynamic behavior of the PAR complex, as phosphorylated PAR3 is known to have reduced affinity for aPKC and can dissociate from the complex (Nagai-Tamai et al., 2002). After dissociation, either cPAR3 or the duo cPAR6/aPKC (or both) could then quickly relocate and occupy a distinct subcellular localization. Experiments using cPAR3 proteins without the aPKC phosphorylation site reveal that unphosphorylated cPAR3 occupies the same position as endogenous cPAR3 in relation to the duo cPAR6/aPKC (data not shown). This implies that the final localization of cPAR3 is independent of the interaction with aPKC and subsequent phosphorylation, thus excluding the hypothesis that cPAR3 moves basally following aPKC-mediated phosphorylation. Our results are more compatible with the alternative model in which the duo cPAR6/aPKC achieve its final localization after interaction with cPAR3 at the apical membrane, followed by phosphorylation of this protein and release of PAR6/aPKC to a more apical position. There, the duo PAR6/aPKC might then serve to localize the CRB complex, which we have shown to be present in the same apical domain. Our finding that endogenous cCRB3 and cPALS1 are also recruited to sites of cPAR3 ectopic accumulation is consistent with this model. In cPAR3-overexpressing cells, after the initial recruitment of cPAR6/aPKC to ectopic cPAR3 punctates and subsequent cPAR3 phosphorylation, the duo cPAR6/aPKC might relocate and become available to bind members of the CRB complex, thereby mimicking the cascade of protein assembly that normally occurs at the apical membrane. These results also raise the hypothesis that cPAR3 is able to confer an apical character to the membrane regions where it accumulates, which is further supported by our finding that the

4302

Journal of Cell Science 119 (20)

Journal of Cell Science

basolateral LGL protein is excluded from the sites of ectopic cPAR3 accumulation. Strikingly, overexpression of cPAR6 failed to cause similar reorganization of polarity complexes in NEP cells (data not shown), emphasizing the unique and central role of cPAR3 in this process. In conclusion, our data highlight the central role of cPAR3 in coordinating the spatio-temporal organization of the various polarity complexes in NEP cells, triggering a series of dynamic interactions that lead to the segregation of different molecular domains in the plasma membrane. cPAR3 and assembly of intercellular junctions Acquisition of apico-basal polarity and formation of cellular junctions are connected processes in epithelial cells, but it is generally accepted that the establishment of cell-cell contacts is the initiating cue for polarity generation (Nelson, 2003; Perez-Moreno et al., 2003). The process starts with the assembly of the AJC at the apical end of the lateral membrane, which mediates cell adhesion and triggers a series of biochemical events that will lead to the delimitation of distinct apical and basolateral membrane domains in epithelial cells. However, recent work in Drosophila demonstrated that this is not a universal mechanism and that the activity of BAZ precedes AJC formation, accumulating at the apical domain of epithelial cells even in the absence of junctional complexes (Harris and Peifer, 2004). Our data in the chick neuroepithelium point to a similar hierarchy in polarity mechanisms, with cPAR3 acting upstream of AJCs during establishment of apico-basal polarity in NEP cells. Indeed, we show that cPAR3 is able to direct the formation of new AJCs at the cell membrane, as revealed by the accumulation of N-Cadherin and ␤-Catenin at the sites of ectopic cPAR3 localization. The consequent higher adhesiveness of cPAR3-overexpressing cells might be responsible for the appearance of mitotic cells at abnormal positions within the NE. As our time-lapse analysis reveals, the INM of cPAR3-overexpressing cells is arrested, probably due to additional adhesion sites with their neighbors, and this results in cells entering division at ectopic positions. The ability of cPAR3 to promote the assembly of AJCs at the membrane of NEP cells might involve various mechanisms. cPAR3 might contribute to the positioning of AJCs through its organizing activity of the apical polarity complexes, which results in the definition of a boundary between membrane regions with apical and basolateral characteristics, where AJCs will form and develop into mature junctions. In addition, cPAR3 might have an active role in recruiting AJC components to the sites where it first accumulates, as suggested by our overexpression experiments and by the fact that cPAR3 and N-Cadherin can be detected in the same macromolecular complex. This is also consistent with recent findings in the Drosophila embryonic epithelium (Harris and Peifer, 2005), which demonstrated that BAZ can form complexes with DEcadherin and ␤-Catenin, directing their apical recruitment during polarity establishment. An additional mechanism might involve the function of cNumb, which we show to localize at AJs in NEP cells and to be recruited to the sites of ectopic cPAR3 accumulation. Numb has been linked to the ARF6 membrane recycling pathway (Smith et al., 2004), which is a key modulator of Cadherin endocytosis from AJs (Palacios et al., 2001), so it is

tempting to speculate that apical cNumb acts downstream of cPAR3 as part of the mechanism that regulates AJ dynamics in NEP cells. The primary role of PAR3 as polarity cue, upstream of AJC, in the mechanisms that generate apico-basal polarity of epithelial cells, raises the question of how is the protein targeted to the plasma cell membrane in the first place. During cellularization of the Drosophila embryo, dynein-mediated transport along microtubules and subsequent capture by the apical actin scaffold have been shown to be involved in the initial targeting of BAZ (Harris and Peifer, 2005). The question has not yet been addressed in vertebrate epithelial cells, but a good candidate to mediate the transport of PAR3 to the apical membrane is the kinesin motor KIF3, which can interact with PAR3 (Nishimura et al., 2004) and is involved in the localization of the PAR complex to the cilia of MDCK cells (Fan et al., 2004) and to the axonal tip of polarizing hippocampal neurons (Shi et al., 2004). The KIF3 motor is also involved in N-Cadherin transport and localization to the apical region of mouse NEP cells, and a mutation of Kap3, which encodes an essential component of the KIF3 motor, causes a severe disruption of the NE’s polarized architecture (Teng et al., 2005). Although these defects were attributed to the lack of N-Cadherin function at AJCs, it is also possible that the apical targeting of PAR3 is affected in Kap3 mutants, and this might have an important contribution to the defects observed in NEP polarity. cPAR3, neuroepithelial polarity and cell fate determination Another cellular phenotype due to cPAR3 overexpression in chick neuroepithelium was the late formation of aggregates of NEP cells organized around a small lumen (rosettes), which were often found in the basal part of the NT. Within these rosettes, ectopic cPAR3 concentrates at the region of the membrane close to the lumen, where other apical polarity proteins also accumulate, indicating that this region constitutes the apical domain of rosette cells. This is further supported by the finding that AJ proteins, like N-Cadherin and ␤-Catenin, are also concentrated at the same region, and that centrosomes are present in the cytoplasm just below this apical domain, like in normal NEP cells. Taken together, these results indicate that cPAR3-overexpressing cells are able to reorganize their polarity and define a single apical domain, close to the luminal surface, with associated junctional structures allowing cells to adhere to their neighbors and form rosettes. In addition, rosette cells are progressing through the cell cycle and engaged in neurogenesis, indicating that these structures behave like small ectopic neural tubes. The function of PAR3 in organizing the apical domain and correctly positioning intercellular junctions in NEP cells is therefore of upmost importance for neurogenesis, as it allows cells to establish intercellular communication and participate in the normal process of cell fate determination. This is supported by the finding that a partial loss-of-function of the zebrafish par3D gene causes a profound disorganization of the retinal neuroepithelium, a block of neurogenesis and a consequent reduction in neuronal differentiation (Wei et al., 2004). In Drosophila, the PAR3 homolog BAZ also plays a central role in defining neuroblast polarity, underlying the asymmetric divisions that give rise to the various neuronal and glial cells

PAR3 and neuroepithelial polarity that make up the fly nervous system (Wodarz, 2005). Thus, although vertebrate and Drosophila neurogenesis are apparently very different processes, with neural progenitors possessing distinct polarity characteristics (epithelial in vertebrates and mesenchymal in Drosophila), the function of PAR3 in coordinating polarity establishment is essential in both types of progenitors for the normal process of neural fate determination. Materials and Methods Plasmids Plasmids for expression of GFP-gpi (Keller et al., 2001), mRFP::cPAR6, cPAR3, cPAR3::GFP and cPAR3-N::GFP (corresponding to a deletion of the first 228 amino acids of cPAR3) were constructed using the pCAGGS vector (Niwa et al., 1991). Accession numbers for cPAR3, cPAR3L and cPAR6B are, respectively, DQ683186, DQ683187 and DQ683188. The mLGL1 plasmid (mLGL1pFLAGCMV2) was a kind gift from Dr Tony Pawson, and was used at a final concentration of 1 ␮g ␮l–1. The mRFP cDNA (Campbell et al., 2002) was kindly provided by Dr Roger Tsien.

Journal of Cell Science

In ovo electroporation of chick embryos Embryos at HH11-12 were microinjected in the NT with plasmid DNA and electroporated using a BTX ECM830 electroporator. Embryos were incubated at 38°C and then harvested and processed for criostat sectioning. BrdU labeling was done as previously described (Fior and Henrique, 2005). Plasmids were electroporated at 0.1-0.4 ␮g ␮l–1 for localization analysis and at 2 ␮g ␮l–1 for gainof-function studies.

Immunofluorescence on cryostat sections Embryos were fixed in either 1% or 4% paraformaldehyde, for 15 minutes or 2 hours at room temperature, respectively. After embedding in 7.5% gelatin, 8 ␮m criostat sections were prepared for IF as described previously (Manabe et al., 2002). Results were analyzed in a Leica DMR microscope, with both 20⫻ (PlanApo NA 0.7) and 40⫻ (PL Fluotar NA 0,75) objectives, and photographed with a Leica DC350F digital camera. Confocal analysis was done using a Zeiss LSM510 microscope with 63⫻ (Plan-Apochromat; NA 1.4) objectives and LSM 510 AIM software. Adobe Photoshop 6.0 was used to adjust input levels, brightness and contrast.

Antibodies The antibodies used in this work were: rabbit anti-mPAR3 and anti-phosphohistone H3 (Upstate); rabbit anti-aPKC␨ and goat anti-actin (SCB); rabbit anti-PALS1 and rabbit anti-CRB3 (Makarova et al., 2003); rabbit anti-mLGL1 (Musch et al., 2002); mouse anti-N-cadherin, mouse anti-␥-tubulin, mouse anti-BrdU and rabbit anti-␤Catenin (Sigma); mouse anti-Numb (CMNB1, DSHB) mouse anti-HU (Molecular Probes); mouse anti-TUJ-1 (BabCo); rabbit anti-PAR6␤ (Yamanaka et al., 2003); rabbit anti-CRMP2 (Nishimura et al., 2003). Secondary antibodies were HRPconjugated anti-rabbit IgG and HRP-conjugated anti-mouse IgG (Promega); HRPconjugated anti-goat IgG (SCB): Alexa594-goat anti-rabbit, Alexa594-donkey antimouse and Alexa488-goat anti-rabbit (Molecular Probes).

Live imaging of NT slices

Embryonic NT was electroporated in ovo with the GFP-gpi vector (0.4 ␮g ␮l–1), alone or in combination with the cPAR3-vector (2 ␮g ␮l–1), and embryos were harvested 4 hours later. Slices (100 ␮m) were prepared using a McILTwain Tissue Chopper, and embedded in 1% low-melting agarose in embryo media (Ham’s/F12 with 4 mM glutamine, 1 mM sodium pyruvate, 5% fungizone, 50 U ml–1 penicillin and 8 mg ml–1 streptomycin). Slices were mounted in a glassbottom Petri dish, covered with embryo media and overlayed with mineral oil. Dishes were then incubated at 38°C in an inverted microscope and Z-series images were collected at 1.5, 2, 2.3 or 3.3 ␮m-steps using a laser confocal microscope Zeiss LSM510, at various time intervals, with a 20⫻ objective (Plan-NeoFluor; NA 0.5). Images were processed using the Zeiss LSM AIM software. ImageJ and Jasc Animation Shop2 software were used to convert image sequences to QuickTime movies. A total of 18 slices (six with GFP-gpi alone and 12 with a mix of GFP-gpi and cPAR vectors) was followed for periods between 8 and 24 hours. We analyzed a total of 12 cPAR3overexpressing cells in detail, all with a similar behavior (Fig. 4 and see supplementary material Movie 2).

In situ hybridization Chicken embryos were collected and fixed in 4% paraformaldehyde/PBS at 4°C. Whole-mount in situ hybridization was done as described (Fior and Henrique, 2005). For hybridization on cryostat sections, fixed embryos were cryoprotected in 15% sucrose in PBS, embedded in 7.5% gelatin/15% sucrose/PBS and

4303

cryosectioned (12 ␮m). Hybridization on cryostat sections was done as previously described (Fior and Henrique, 2005), with Fast-Red (Roche) as Alkaline Phosphatase substrate, producing a red fluorescent precipitate. Detailed protocols are available upon request.

Immunoprecipitation NTs from electroporated embryos were dissected and sectioned longitudinally along the midline. Protein lysates from each half, or from entire NTs of HH18 and HH23 embryos, were prepared by homogenizing the tissue in cold Lysis Buffer containing 150 mM NaCl, 1% Triton, 50 mM HEPES pH 7.5, 2 mM EDTA, 10% glycerol, 50 mM NaF, 1 mM Na3VO4 and protease inhibitors (Boehringer). Protein extracts were incubated with antibodies overnight at 4°C. Protein G Sepharose (Amersham) or ProteinA/G PLUS-Agarose IP Reagent (SCB) beads were added to the extract and incubated for 3 hours at 4°C. After washing in Lysis Buffer, extracts were subjected to western blot analysis with different antibodies. Signal quantification was performed by densitometry using ImageQuant® software (Molecular Dynamics).

We thank S. Ohno, B. Margolis, A. LeBivic, T. Pawson, R. Tsien, K. Simons, A. Musch, K. Kaibuchi and the Developmental Studies Hybridoma Bank for providing antibodies and DNA constructs. We thank Alina Costa for excellent technical help. We thank P. Ingham, D. Ish-Horowicz, F. Schweisguth and O. Pourquié for critical reading of the manuscript. This work was supported by FCT (POCTI/SAU/48783). C.A. was recipient of a FCT fellowship.

References Aaku-Saraste, E., Hellwig, A. and Huttner, W. B. (1996). Loss of occludin and functional tight junctions, but not ZO-1, during neural tube closure–remodeling of the neuroepithelium prior to neurogenesis. Dev. Biol. 180, 664-679. Arimura, N. and Kaibuchi, K. (2005). Key regulators in neuronal polarity. Neuron 48, 881-884. Bellaiche, 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/Discslarge complex is required to establish planar polarity during asymmetric cell division in Drosophila. Cell 106, 355-366. Benton, R. and St Johnston, D. (2003). A conserved oligomerization domain in drosophila Bazooka/PAR-3 is important for apical localization and epithelial polarity. Curr. Biol. 13, 1330-1334. Betschinger, J. and Knoblich, J. A. (2004). Dare to be different: asymmetric cell division in Drosophila, C. elegans and vertebrates. Curr. Biol. 14, R674-R685. Betschinger, J., Mechtler, K. and Knoblich, J. A. (2003). The Par complex directs asymmetric cell division by phosphorylating the cytoskeletal protein Lgl. Nature 422, 326-330. Bilder, D. (2004). Epithelial polarity and proliferation control: links from the Drosophila neoplastic tumor suppressors. Genes Dev. 18, 1909-1925. Bilder, D., Schober, M. and Perrimon, N. (2003). Integrated activity of PDZ protein complexes regulates epithelial polarity. Nat. Cell Biol. 5, 53-58. Campbell, R. E., Tour, O., Palmer, A. E., Steinbach, P. A., Baird, G. S., Zacharias, D. A. and Tsien, R. Y. (2002). A monomeric red fluorescent protein. Proc. Natl. Acad. Sci. USA 99, 7877-7882. Chen, X. and Macara, I. G. (2005). Par-3 controls tight junction assembly through the Rac exchange factor Tiam1. Nat. Cell Biol. 7, 262-269. Chenn, A. and McConnell, S. K. (1995). Cleavage orientation and the asymmetric inheritance of Notch1 immunoreactivity in mammalian neurogenesis. Cell 82, 631641. Chenn, A., Zhang, Y. A., Chang, B. T. and McConnell, S. K. (1998). Intrinsic polarity of mammalian neuroepithelial cells. Mol. Cell Neurosci. 11, 183-193. Cuenca, A. A., Schetter, A., Aceto, D., Kemphues, K. and Seydoux, G. (2003). Polarization of the C. elegans zygote proceeds via distinct establishment and maintenance phases. Development 130, 1255-1265. Fan, S., Hurd, T. W., Liu, C. J., Straight, S. W., Weimbs, T., Hurd, E. A., Domino, S. E. and Margolis, B. (2004). Polarity proteins control ciliogenesis via kinesin motor interactions. Curr. Biol. 14, 1451-1461. Fior, R. and Henrique, D. (2005). A novel hes5/hes6 circuitry of negative regulation controls Notch activity during neurogenesis. Dev. Biol. 281, 318-333. Fujita, S. (1960). Mitotic pattern and histogenesis of the central nervous system. Nature 185, 702-703. Gao, L., Macara, I. G. and Joberty, G. (2002). Multiple splice variants of Par3 and of a novel related gene, Par3L, produce proteins with different binding properties. Gene 294, 99-107. Geldmacher-Voss, B., Reugels, A. M., Pauls, S. and Campos-Ortega, J. A. (2003). A 90-degree rotation of the mitotic spindle changes the orientation of mitoses of zebrafish neuroepithelial cells. Development 130, 3767-3780. Harris, T. J. and Peifer, M. (2004). Adherens junction-dependent and -independent steps in the establishment of epithelial cell polarity in Drosophila. J. Cell Biol. 167, 135147. Harris, T. J. and Peifer, M. (2005). The positioning and segregation of apical cues during epithelial polarity establishment in Drosophila. J. Cell Biol. 170, 813-823.

Journal of Cell Science

4304

Journal of Cell Science 119 (20)

Henrique, D. and Schweisguth, F. (2003). Cell polarity: the ups and downs of the Par6/aPKC complex. Curr. Opin. Genet. Dev. 13, 341-350. Hurd, T. W., Gao, L., Roh, M. H., Macara, I. G. and Margolis, B. (2003). Direct interaction of two polarity complexes implicated in epithelial tight junction assembly. Nat. Cell Biol. 5, 137-142. Hutterer, A., Betschinger, J., Petronczki, M. and Knoblich, J. A. (2004). Sequential roles of Cdc42, Par-6, aPKC, and Lgl in the establishment of epithelial polarity during Drosophila embryogenesis. Dev. Cell 6, 845-854. Imai, F., Hirai, S., Akimoto, K., Koyama, H., Miyata, T., Ogawa, M., Noguchi, S., Sasaoka, T., Noda, T. and Ohno, S. (2006). Inactivation of aPKC{lambda} results in the loss of adherens junctions in neuroepithelial cells without affecting neurogenesis in mouse neocortex. Development 133, 1735-1744. Izumi, Y., Hirose, T., Tamai, Y., Hirai, S., Nagashima, Y., Fujimoto, T., Tabuse, Y., Kemphues, K. J. and Ohno, S. (1998). An atypical PKC directly associates and colocalizes at the epithelial tight junction with ASIP, a mammalian homolog of Caenorhabditis elegans polarity protein PAR-3. J. Cell Biol. 143, 95-106. Joberty, G., Petersen, C., Gao, L. and Macara, I. G. (2000). The cell-polarity protein Par6 links Par3 and atypical protein kinase C to Cdc42. Nat. Cell Biol. 2, 531-539. Keller, P., Toomre, D., Diaz, E., White, J. and Simons, K. (2001). Multicolour imaging of post-Golgi sorting and trafficking in live cells. Nat. Cell Biol. 3, 140-149. Kemphues, K. (2000). PARsing embryonic polarity. Cell 101, 345-348. Klezovitch, O., Fernandez, T. E., Tapscott, S. J. and Vasioukhin, V. (2004). Loss of cell polarity causes severe brain dysplasia in Lgl1 knockout mice. Genes Dev. 18, 559571. Kohjima, M., Noda, Y., Takeya, R., Saito, N., Takeuchi, K. and Sumimoto, H. (2002). PAR3beta, a novel homolog of the cell polarity protein PAR3, localizes to tight junctions. Biochem. Biophys. Res. Commun. 299, 641-646. Koike, C., Nishida, A., Akimoto, K., Nakaya, M. A., Noda, T., Ohno, S. and Furukawa, T. (2005). Function of atypical protein kinase C lambda in differentiating photoreceptors is required for proper lamination of mouse retina. J. Neurosci. 25, 10290-10298. Langevin, J., Le Borgne, R., Rosenfeld, F., Gho, M., Schweisguth, F. and Bellaiche, Y. (2005). Lethal giant larvae controls the localization of notch-signaling regulators numb, neuralized, and Sanpodo in Drosophila sensory-organ precursor cells. Curr. Biol. 15, 955-962. Lemmers, C., Michel, D., Lane-Guermonprez, L., Delgrossi, M. H., Medina, E., Arsanto, J. P. and Le Bivic, A. (2004). CRB3 binds directly to Par6 and regulates the morphogenesis of the tight junctions in mammalian epithelial cells. Mol. Biol. Cell 15, 1324-1333. Lin, D., Edwards, A. S., Fawcett, J. P., Mbamalu, G., Scott, J. D. and Pawson, T. (2000). A mammalian PAR-3-PAR-6 complex implicated in Cdc42/Rac1 and aPKC signalling and cell polarity. Nat. Cell Biol. 2, 540-547. Macara, I. G. (2004). Parsing the polarity code. Nat. Rev. Mol. Cell Biol. 5, 220-231. Makarova, O., Roh, M. H., Liu, C. J., Laurinec, S. and Margolis, B. (2003). Mammalian Crumbs3 is a small transmembrane protein linked to protein associated with Lin-7 (Pals1). Gene 302, 21-29. Manabe, N., Hirai, S., Imai, F., Nakanishi, H., Takai, Y. and Ohno, S. (2002). Association of ASIP/mPAR-3 with adherens junctions of mouse neuroepithelial cells. Dev. Dyn. 225, 61-69. Mizuno, K., Suzuki, A., Hirose, T., Kitamura, K., Kutsuzawa, K., Futaki, M., Amano, Y. and Ohno, S. (2003). Self-association of PAR-3-mediated by the conserved Nterminal domain contributes to the development of epithelial tight junctions. J. Biol. Chem. 278, 31240-31250. Musch, A., Cohen, D., Yeaman, C., Nelson, W. J., Rodriguez-Boulan, E. and Brennwald, P. J. (2002). Mammalian homolog of Drosophila tumor suppressor lethal (2) giant larvae interacts with basolateral exocytic machinery in Madin-Darby canine kidney cells. Mol. Biol. Cell 13, 158-168. Nagai-Tamai, Y., Mizuno, K., Hirose, T., Suzuki, A. and Ohno, S. (2002). Regulated protein-protein interaction between aPKC and PAR-3 plays an essential role in the polarization of epithelial cells. Genes Cells 7, 1161-1171. Nelson, W. J. (2003). Adaptation of core mechanisms to generate cell polarity. Nature 422, 766-774. Nishimura, T., Fukata, Y., Kato, K., Yamaguchi, T., Matsuura, Y., Kamiguchi, H.

and Kaibuchi, K. (2003). CRMP-2 regulates polarized Numb-mediated endocytosis for axon growth. Nat. Cell Biol. 5, 819-826. Nishimura, T., Kato, K., Yamaguchi, T., Fukata, Y., Ohno, S. and Kaibuchi, K. (2004). Role of the PAR-3-KIF3 complex in the establishment of neuronal polarity. Nat. Cell Biol. 6, 328-334. Niwa, H., Yamamura, K. and Miyazaki, J. (1991). Efficient selection for highexpression transfectants with a novel eukaryotic vector. Gene 108, 193-199. Ohno, S. (2001). Intercellular junctions and cellular polarity: the PAR-aPKC complex, a conserved core cassette playing fundamental roles in cell polarity. Curr. Opin. Cell Biol. 13, 641-648. Ohshiro, T., Yagami, T., Zhang, C. and Matsuzaki, F. (2000). Role of cortical tumoursuppressor proteins in asymmetric division of Drosophila neuroblast. Nature 408, 593596. Palacios, F., Price, L., Schweitzer, J., Collard, J. G. and D’Souza-Schorey, C. (2001). An essential role for ARF6-regulated membrane traffic in adherens junction turnover and epithelial cell migration. EMBO J. 20, 4973-4986. 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. Perez-Moreno, M., Jamora, C. and Fuchs, E. (2003). Sticky business: orchestrating cellular signals at adherens junctions. Cell 112, 535-548. Roegiers, F., Younger-Shepherd, S., Jan, L. Y. and Jan, Y. N. (2001). Two types of asymmetric divisions in the Drosophila sensory organ precursor cell lineage. Nat. Cell Biol. 3, 58-67. Shi, S. H., Cheng, T., Jan, L. Y. and Jan, Y. N. (2004). APC and GSK-3beta are involved in mPar3 targeting to the nascent axon and establishment of neuronal polarity. Curr. Biol. 14, 2025-2032. Smith, C. A., Dho, S. E., Donaldson, J., Tepass, U. and McGlade, C. J. (2004). The cell fate determinant numb interacts with EHD/Rme-1 family proteins and has a role in endocytic recycling. Mol. Biol. Cell 15, 3698-3708. Suzuki, A. and Ohno, S. (2006). The PAR-aPKC system: lessons in polarity. J. Cell Sci. 119, 979-987. Suzuki, A., Yamanaka, T., Hirose, T., Manabe, N., Mizuno, K., Shimizu, M., Akimoto, K., Izumi, Y., Ohnishi, T. and Ohno, S. (2001). Atypical protein kinase C is involved in the evolutionarily conserved par protein complex and plays a critical role in establishing epithelia-specific junctional structures. J. Cell Biol. 152, 1183-1196. Tanentzapf, G. and Tepass, U. (2003). Interactions between the crumbs, lethal giant larvae and bazooka pathways in epithelial polarization. Nat. Cell Biol. 5, 46-52. Teng, J., Rai, T., Tanaka, Y., Takei, Y., Nakata, T., Hirasawa, M., Kulkarni, A. B. and Hirokawa, N. (2005). The KIF3 motor transports N-cadherin and organizes the developing neuroepithelium. Nat. Cell Biol. 7, 474-482. Wakamatsu, Y., Maynard, T. M., Weston, J. A. and Jones, S. A. (1999). NUMB localizes in the basal cortex of mitotic avian neuroepithelial cells and modulates neuronal differentiation by binding to NOTCH-1. Neuron 23, 71-81. Wei, X., Cheng, Y., Luo, Y., Shi, X., Nelson, S. and Hyde, D. R. (2004). The zebrafish Pard3 ortholog is required for separation of the eye fields and retinal lamination. Dev. Biol. 269, 286-301. Wodarz, A. (2005). Molecular control of cell polarity and asymmetric cell division in Drosophila neuroblasts. Curr. Opin. Cell Biol. 17, 475-481. Wodarz, A., Ramrath, A., Grimm, A. and Knust, E. (2000). Drosophila atypical protein kinase C associates with Bazooka and controls polarity of epithelia and neuroblasts. J. Cell Biol. 150, 1361-1374. Yamanaka, T., Horikoshi, Y., Suzuki, A., Sugiyama, Y., Kitamura, K., Maniwa, R., Nagai, Y., Yamashita, A., Hirose, T., Ishikawa, H. et al. (2001). PAR-6 regulates aPKC activity in a novel way and mediates cell-cell contact-induced formation of the epithelial junctional complex. Genes Cells 6, 721-731. Yamanaka, T., Horikoshi, Y., Sugiyama, Y., Ishiyama, C., Suzuki, A., Hirose, T., Iwamatsu, A., Shinohara, A. and Ohno, S. (2003). Mammalian Lgl forms a protein complex with PAR-6 and aPKC independently of PAR-3 to regulate epithelial cell polarity. Curr. Biol. 13, 734-743. Zhong, W., Feder, J. N., Jiang, M. M., Jan, L. Y. and Jan, Y. N. (1996). Asymmetric localization of a mammalian numb homolog during mouse cortical neurogenesis. Neuron 17, 43-53.