- Email: [email protected]
Role of the Crumbs complex in the regulation of junction formation in Drosophila and mammalian epithelial cells
Biology of the Cell 94 (2002) 305–313 www.elsevier.com/locate/bicell
Original article
Role of the Crumbs complex in the regulation of junction formation in Drosophila and mammalian epithelial cells Emmanuelle Médina, Céline Lemmers, Lydie Lane-Guermonprez, André Le Bivic * NMDA-IBDM, UMR 6156, Campus de Luminy, Université de la Méditerranée, case 907, 13288 Marseille cedex 09, France Received 9 September 2002; accepted 11 September 2002
Abstract The formation of a belt-like junctional complex separating the apical from the lateral domain is an essential step in the differentiation of epithelial cells. Thus protein complexes regulating this event are of first importance for the development of cell polarity and physiological functions of epithelial tissues. In Drosophila, the discovery of a gene, crb, controlling the coalescence of the spots of zonula adherens (ZA) into a adhesive ring around the cells was a major step. We know now that Crumbs, the product of crb is an apical transmembrane protein conserved in mammals and that it interacts by its cytoplasmic domain with two cortical modular proteins, Stardust (Sdt) and Discs lost (Dlt) that are also essential for the correct assembly of the ZA. These two proteins are also conserved in mammals and it is most likely that the Crumbs complex plays a similar role in very different species. Recently, we have shown that Crumbs interacts with the cortical cytoskeleton made of DMoesin and bheavy-Spectrin and this connection could explain in part the role of Crumbs in building the ZA. Future work will help to understand several aspects of the Crumbs complex that are still unknown, like the role of the large extracellular domain or the precise function of Sdt and Dlt in the building of the ZA. Finding an answer to these questions will help to find new therapies for Retinitis pigmentosa and other retina degeneration in which CRB1, the human homologue of crb, has been involved. © 2002 Éditions scientifiques et médicales Elsevier SAS. All rights reserved. Keywords: Epithelial cells; Polarity, Junctions; Retina; Moesin
1. Epithelial polarity Multicellular organisms could develop with the apparition of epithelial cells that physically separate the outside world from the inside of the body. With this invention, organisms could ensure the homeostasis of the liquid bathing their internal tissues, select nutrients and for some of them, isolate their nervous system. The making of epithelial cells, crucial for evolution, involved the creation of several mechanisms like the formation of specialized junctions between cells and the sorting and recycling of plasma membrane components. The organization of epithelial cells is now well known and these cells stand out from other cells by the division of their plasma membrane in at least two separate domains, the apical (facing the outside) and the basolateral (in contact with the
* Corresponding author. Tel.: +33-4-91-26-97-41; fax: +33-4-91-26-97-48. E-mail address: [email protected] (A. Le Bivic) © 2002 Éditions scientifiques et médicales Elsevier SAS. All rights reserved. PII: S0248-4900(02)00004-7
other cells and the internal medium) domains. These domains are first established by cell–cell contacts that allow adhesion molecules to cluster and nucleate other scaffold proteins that in turn recruit other transmembrane proteins (Yeaman et al., 1999). Then, junctions that physically separate the plasma membrane domains are formed and reinforced to allow the targeting of newly synthesized components to the recently formed apical and basolateral membranes (Mostov et al., 2000). Transmembane proteins that are able to nucleate and build networks of cytoplasmic proteins are called extracellular cell surface organizers (ECSOs). These proteins give cues to the cells about their position and how to orient their future polarity. In mammalian epithelial cells, integrins and cadherins play such a role (Hynes, 1992) but their precise contribution during the process of establishing and maintaining epithelial polarity in vivo is still hard to appreciate. Most of the work done on these mammalian ECSOs was performed on epithelial cell cultures that however do not always recreate the in vivo complexity (Rodriguez-Boulan and Nelson, 1989). On the other hand, Drosophilamelano-
306
E. Médina et al. / Biology of the Cell 94 (2002)
Fig. 1. Comparison of invertebrate and vertebrate apical junctional complexes: ZA is common to vertebrate and invertebrate epithelial cells. The components of this junction are well conserved among the species with the transmembrane protein E-cadherin (DE-Cad and E-Cad, in (A) and (B), respectively), connected to the actin cytoskeleton through a- and b-catenin (Da-Cat and Arm in (A) and a-Cat and b-Cat in (B), respectively). In Drosophila (A), the sub-apical complex (SAC) is apical to the ZA whereas in mammals (B), the tight junction (TJ) is located at this position. In the SAC, the transmembrane protein Crumbs (Crb) has been shown to be essential for the ZA morphogenesis. This epithelial cell surface organizer (ECSO) plays a crucial role in epithelial morphogenesis through its interaction with Stardust (Sdt) and Discs lost (Dlt), and is linked to cortical cytoskeleton through DMoesin (Dmoe). In Drosophila (A), another complex essential for polarity (the Bazooka (Baz)-aPKC-DPar6 complex) is also found in this region. In mammals (B), the homologues of the SAC components are located at the TJ level and are in close relation with the TJ components (the transmembrane proteins Claudin (Clau) and Occludin (Occ), and the MAGUKs ZO-1/ZO-2/ZO-3). In Drosophila (A), the septate junction (SJ) lies basal to the ZA and plays a role in controlling paracellular diffusion. This junction is formed by the transmembrane protein Neurexin IV (NrxIV) and several proteins important for epithelial polarity such as Discs large (Dlg), Lethal giant larvae (Lgl) and Scribble. In mammals (B), the Dlg, Lgl and Scribble homologues are also found below the ZA and colocalize with the HER2/Erbin complex.
gaster provides a powerful and important complementary genetic model. In Drosophila, epithelial cell organization is close to the one observed in mammals with, however, some differences in the structures that build the junctional complex separating the apical from the lateral domain (Fig. 1 ). In mammalian cells the most apical junction is the zonula occludens (ZO) made of polytopic membrane proteins like claudins and occludins. This structure provides both the control of paracellular transport and of cell–cell adhesion and is located just above the zonula adherens (ZA), a junction common to vertebrates and invertebrates (Fig. 1). The ZA is made up of a calcium-dependent adhesion protein, E-cadherin, which is connected to an intracellular network of
actin through its partners a- and b-catenins (Vasioukhin and Fuchs, 2001). The molecular components of this junction are extremely well conserved between Drosophila and mammals, suggesting that its formation is a crucial step for epithelial morphogenesis (Nagafuchi, 2001; Tepass, 2002). In Drosophila, below the ZA, septate junctions (SJs) play a role in preventing and controlling paracellular diffusion (Carlson et al., 2000). These junctions can have different aspects when visualized by electron microscopy techniques since they can be pleated or smooth depending on the tissue observed (Baumgartner et al., 1996). Molecularly, these junctions contain a transmembrane protein, Neurexin IV, which is a homologue of mammalian paranodin, involved in the formation of
E. Médina et al. / Biology of the Cell 94 (2002)
Ranvier node (Bellen et al., 1998). Several proteins, Discs large (Dlg), lethal giant larvae (Lgl) and Scribble, important for epithelial organization are associated to SJs, indicating that these junctions also play a role in epithelial polarity besides the fence function (Bilder et al., 2000). In Drosophila, recent findings have pointed out the crucial role of the region located just above the ZA, called the sub-apical complex (SAC) (Fig. 1). In fact, many of the proteins localized in this region, have now been shown to be essential for the positioning of the ZA and thus for epithelial polarity in general. In particular, the first identified transmembrane component of this region, Crumbs, has become a central piece for the organization of the whole junctional complex comprising the ZA.
2. An apical organizer, Crumbs In the last 10 years, several crucial components for epithelial morphogenesis and polarity have been identified using Drosophila powerful genetic screens (Perrimon et al., 1996). One of these genes was called crumbs (crb) because the cuticle of embryos bearing this mutation was reduced to small patches reminding of bread crumbs (Jürgen et al., 1984). Embryos with crb loss of function mutations die shortly after gastrulation, a stage involving intensive epithelial movement and reorganization in Drosophila(Tepass et al., 1990). This early lethal phenotype is the resultant of a loss of epithelial organization of the ectodermal layer in which there is a rounding of epidermal cells with a high degree of apoptosis (Leptin, 1999; Tepass et al., 1990). The protein encoded by crb and named Crumbs was identified by the group of E. Knust and is a transmembrane protein with a small cytoplasmic domain of 37 amino acids and a large extracellular domain made of 30 EGF like repeats and four laminin G-like repeats (Tepass et al., 1990) (Fig. 2 ). Crumbs is expressed in all the primary and secondary epithelia derived from the ectoderm and, interestingly, one of the loss of function mutants of crb, the crb8F105 allele, is a stop mutation in the cytoplasmic tail of Crumbs leaving only half of it (Wodarz et al., 1993). This was the first indication of a crucial role of the Crumbs cytoplasmic domain in its epithelial function and it was later confirmed by the fact that the expression of a transgene, deleted of the part coding for the extracellular domain (Myc-intrawt), was able to rescue the epithelial organization of the ectoderm (Wodarz et al., 1995). A more precise analysis of the crb phenotype showed that ZAs, which first appear as spots on the lateral membrane can not coalesce and form a belt-like structure around the cells in crb mutants, leading to the hypothesis that Crumbs is necessary for their positioning and integrity (Tepass, 1996; Grawe et al., 1996). Meanwhile, the over-expression of crumbs or Myc-intrawt was shown to provoke an expansion of the apical domain and a disorganization of the epithelial layer confirming that Crumbs is involved in the correct positioning and assembly of ZAs and also regulates the shape of the apical
307
membrane by a still unknown mechanism (Wodarz et al., 1995).
3. The Crumbs complex How Crumbs is involved in the correct positioning and assembly of the ZA through its cytoplasmic domain is a major question and a beginning of an answer was brought by the identification of two proteins, Discs lost (Dlt) and Stardust (Sdt), that interact with Crumbs (Fig. 2). Stardust (sdt) was the name given to Drosophila mutants that showed a phenotype very similar to crb. sdt is a zygotic gene like crb and, in sdt mutants, epidermal cells loose polarity, become round and the epithelial layer is disorganized. Furthermore, sdt could be positioned downstream in the crb cascade since over-expression of sdt could restore some of crb defects (Tepass and Knust, 1993). While the sdt mutation was described in the early 1990s, identification of the protein encoded by the sdt locus was only published recently (Bachmann et al., 2001; Hong et al., 2001). Sdt belongs to the Maguk (membrane-associated guanylate kinase) family like ZO-1 (zonula occludens 1) or PSD-95 (post synaptic density 95) in mammals (Fanning and Anderson, 1999). Two isoforms of Sdt are found in D. melanogaster and both have a single PDZ (PSD-95, Discs large, ZO-1) domain, a SH3 (Src homology region 3) domain and a guanylate kinase (GUK) domain (Bachmann et al., 2001; Hong et al., 2001). Sdt colocalizes with Crumbs to the SAC and interacts directly with the last four amino acids ERLI of Crumbs cytoplasmic tail (Fig. 3 ) but it was not shown whether this interaction is mediated through the PDZ domain of Sdt (Bachmann et al., 2001; Hong et al., 2001), as it is likely to occur. Another protein was shown to interact with Crumbs and this protein is the product of the Discs lost (dlt) gene. dlt is a gene maternally expressed as opposed to crb and sdt and thus was not discovered in the same genetic screen. In fact, Dlt was isolated by a two-hybrid screen (Bhat et al., 1999), using the cytoplasmic tail of Neurexin IV (NRX IV), a protein of the septate junctions (Baumgartner et al., 1996). Curiously, in mature epithelial epidermal cells of Drosophila embryos, Dlt colocalizes to the SAC with Crumbs and Sdt but not with Nrx IV questioning the physiological relevance of this interaction in fully polarized epithelial cells (Bhat et al., 1999). Given the co-localization with Crumbs and the fact that Dlt is a cortical protein made of four PDZ domains, it was hypothesized that the two proteins interact directly but this hypothesis has yet to be demonstrated. More biochemical work on the interaction between Crumbs and Dlt in Drosophila should be performed in order to answer this crucial point. In fact, if Dlt binds directly to the PDZ-binding motif (ERLI) of Crumbs, there must be some regulation to control its competition with Sdt for the same binding site. The fact that Dlt is already expressed during cellularization and that overexpression of Crumbs in fly embryos leads to redistribution of Dlt all around epithelial cells, while Sdt is depleted from
308
E. Médina et al. / Biology of the Cell 94 (2002)
Fig. 2. Comparison of mammalian homologues of Drosophila Crumbs, Discs lost and Stardust. (A) Drosophila Crumbs (dCRB) has three human homologues: CRB1, CRB2 and CRB3. The extracellular domains are composed by a signal peptide (black box), 30 EGF like-repeats (red boxes) and four Laminin-A G domain-like motifs (green boxes) for dCRB; EGF-like repeats and three Laminin-A G domain-like motifs for CRB1 and 15 EGF-like repeats and two Laminin A G domain-like motifs. CRB3 has a very short extracellular domain and a signal peptide. All Crumbs proteins share a small cytoplasmic tail composed of 37 amino acids for dCRB, CRB1 and CRB2. CRB3 has 40 amino acids in its cytoplasmic part. Human MUPP1 and PATJ are both similar to Drosophila Dlt. They share a MAGUK recruitment domain (MRE, triangles) and 13, 8 and 4 PDZ domains, respectively (ovals). Mus musculus Pals1 is the Stardust homologue. They have multiple protein–protein interaction domains including one or two Lin-2/lin-7 domains, respectively (L27, L27 N and L27 C), a PDZ domain (oval), a SH3 domain (square), a protein 4.1-binding domain (circle) and a GuK domain (hexagon). (B) Sequence alignment of the conserved cytoplasmic domain of Drosophila Crumbs (dCRB), human CRB1 (hCRB1), CRB2 (hCRB2), CRB3 (hCRB3), C. elegans CRB1 (CeCRB1) and CRB-like (CeCRL1). Identical and conserved amino acids are in black and gray, respectively.
the apical membrane (Bachmann et al., 2001; Hong et al., 2001) indicate that these two proteins Dlt and Sdt have a different regulation of their association to the plasma membrane. In that case, Dlt could bind to another apical transmembrane protein through one of its four PDZ domains, while Sdt would follow ectopic Crumbs and thus be somehow depleted from the apical side. Our current knowledge on the composition of the Crumbs complex in mammals does not favor a direct interaction between Crumbs and Dlt (see below).
4. Crumbs connection to the cortical cytoskeleton It is now well established that Crumbs binds to Sdt by its last four amino acids ERLI but there are other residues in Crumbs cytoplasmic tail that are also crucial for its function. In particular, it was shown that there is a tyrosine at position 10 (Y10) and a glutamic acid at position 16 (E16) which are essential to rescue a normal phenotype in Crumbs mutants (Klebes and Knust, 2000). This portion (GTY) of Crumbs cytoplasmic domain containing Y10 forms a consensus motif
E. Médina et al. / Biology of the Cell 94 (2002)
309
Fig. 3. Comparison of invertebrate and vertebrate Crumbs complexes: (A) In Drosophila, Crumbs (Crb) interacts by its four last amino acids ERLI with the proteins Stardust (Sdt) and Discs lost (Dlt). The interaction between Crb and Sdt is direct but it is not yet known what domain of Sdt is involved. The interaction with Dlt has not been extensively studied but it does not seem to be direct. It has also been shown that Crb interacts with DMoesin and bheavy-Spectrin (bH). Little is known about how Crb recruits these proteins and whether they are or not in a same complex but two amino acids of the Crb cytoplasmic domain, R7 and Y10, are essential for this process. These two proteins, DMoesin and bH provide a link between the Crb complex and actin. (B) In mammals, hCRB interacts directly by its four last amino acids ERLI with the PDZ domain of Pals1, the Sdt homologue. Pals1 directly interacts by its L27 N domain with the MRE domain of PATJ, a Dlt homologue. PATJ links the hCRB complex to the TJ components through direct interactions with Claudin1 and ZO-3. These interactions are mediated by the C-terminal amino acids of ZO-3 and Claudin1 and the PDZ domains 6 and 8 of PATJ, respectively.
for the binding of proteins of the 4.1 family of proteins (Klebes and Knust, 2000), including the ERM proteins (Ezrin/radixin/moesin) (Tepass et al., 2001). It has thus been hypothesized that Crb could bind proteins of this family providing a linkage to a spectrin/actin cytoskeleton (Klebes and Knust, 2000). In fact, in Drosophila, a protein of the 4.1 family, DMoesin is accumulated in the SAC zone and we have shown recently that DMoesin and Crumbs are coimmunoprecipitated together with Dlt from fly embryos indicating that DMoesin could be a likely candidate to bind to the GTY motif of Crumbs. In fact, mutations of theY10 or the R7 just upstream of the GTY motif impaired the connection to DMoesin as measured by capping experiments (Medina et al., 2002). In addition, over-expression of Crumbs in epidermal cells induced a redistribution of both DMoesin and actin confirming that Crumbs might be linked to the cortical cytoskeleton (Medina et al., 2002). Furthermore, it was shown that bheavy-Spectrin, another protein of the cortical cytoskeleton, was redistributed during Crumbs over-expression (Wodarz et al., 1995) suggesting that it could associate to the
same complex. There are no published DMoesin mutants and thus it is difficult to know if DMoesin plays a role in the regulation of ZA formation. Mutants for bheavy-Spectrin are called karst and these mutants show clear defects of the ZA formation during embryogenesis linking it to the Crb phenotype (Zarnescu and Thomas, 1999). In Crb mutants, bheavySpectrin is mislocalized while in karst mutants Crumbs is normally concentrated in the SAC region suggesting that Crb acts upstream of karst(Medina et al., 2002). bheavy-Spectrin has no known domain to bind directly to DMoesin and it is still a challenge to understand how the molecular connection between Crumbs and bheavy-Spectrin occurs and is regulated. Interestingly, Crumbs could also be connected to the cortical cytoskeleton through Sdt. Sdt possesses a 4.1 protein family binding domain and thus could interact with DMoesin (Bachmann et al., 2001; Hong et al., 2001). This interaction would reinforce Crumbs connection to the cytoskeleton and to actin, building a cytoskeleton around the ZA. Future work will provide a new understanding on how the regulation of these multiple connections between the SAC and the ZA is operated.
310
E. Médina et al. / Biology of the Cell 94 (2002)
5. The Crumbs complex in mammals Until the discovery of a human locus involved in a human disease and coding for a protein homologue to Crumbs, in 1999, there was no evidence for a conservation of Crumbs between invertebrates and vertebrates. Thus, the discovery that a gene mutated in Retinitis pigmentosa group 12 (RP12) was very similar to Drosophila crb was a major step (den Hollander et al., 1999) (Fig. 2). This gene was called CRB1 and it produces two isoforms, one secreted or membraneassociated containing the extracellular domain and a putative transmembrane domain and one very similar to Crumbs with a transmembrane and a cytoplasmic domain sharing 63% of identity with Crb (den Hollander et al., 2001b) (Fig. 2). CRB1 is a bona fide Crumbs, since expression of its intracellular domain in Drosophila mutants for crb can rescue the cuticle phenotype to the same extent as transgenic Crumbs (den Hollander et al., 2001b). In its extracellular domain, CRB1 is made of 19 EGF-like and three LG domains (Fig. 2) and most of the mutations found in patients, map to this extracellular domain (den Hollander et al., 1999). CRB1 is also mutated in a more severe form of retina degeneration, the Leber Congenital Amaurosis syndrome (den Hollander et al., 2001a). CRB1 expression is restricted to retina and some parts of the brain both in human and mouse (den Hollander et al., 2002), suggesting that in vertebrates a tissue specialization of Crumbs might have occurred. With the sequencing of the human genome, there are now two additional genes, CRB2 and CRB3 that code for proteins with a highly conserved cytoplasmic tail (67 and 54% of identity with Crumbs, respectively) (Fig. 2) (Lemmers et al., 2002). CRB2 has a large extracellular domain with 15 EGF-like and two LG domains while CRB3 exhibits a very short extracellular domain with no recognizable protein domain in it (Fig. 2). So far the tissue and cellular distribution of these two gene products is not known and it will be of first interest to see if one of them is present in epithelial cells and has a conserved function in the control of epithelial polarity. Mouse and human homologues of Sdt and Dlt have been described recently suggesting that the whole Crumbs complexis conserved in vertebrates. The orthologue of Sdt is a protein called Pals1 (protein associated to lin 7) that also belongs to the MAGUK family (Kamberov et al., 2000). Pals1 binds to a domain of mlin7 called L27 and possesses two L27 domains called L27 N-terminal and L27 C-terminal domains (Fig. 2). Pals1 is concentrated at the level of TJs and binds through its PDZ domain to the ERLI motif of CRB1 and to one homologue of Dlt by its L27 N-terminal domain (Roh et al., 2002a, 2002b). One homologue of Dlt was called PATJ (protein associated to tight junctions or Pals1associated tight junction) and it is a membrane associated protein with at least eight PDZ domains (Lemmers et al., 2002; Roh et al., 2002a, 2002b) (Fig. 2). PATJ binds to Pals1 by its N-terminal domain and does not seem to bind directly to CRB1 or 3 (Lemmers et al., 2002; Roh et al., 2002a, 2002b). PATJ is highly concentrated at the level of the ZO
and is also found associated to the apical membrane confirming that it is a likely homologue of Dlt (Lemmers et al., 2002). Since the endogenous CRB is not yet identified in epithelial cells, CRB1 was exogenously expressed and was also concentrated at the level of the TJs, strongly indicating that a conserved Crumbs complex regulates the formation of TJs in mammals. Indeed, over-expression of PATJ induced a loss of ZO-3 from the TJs (Lemmers et al., 2002) and it was recently shown that the PDZ domain 6 of PATJ is able to interact with ZO-3 providing a direct link with ZO (Roh et al., 2002a, 2002b) (Fig. 3). By RT-PCR we have found that CRB3 is detectable in human intestinal Caco-2 cells and thus could interact with Pals1 (Lemmers et al., 2002). This is reinforced by the fact that CRB3 is able to coimmunoprecipitate PATJ, indicating that CRB3 could be a good candidate for being an epithelial CRB in mammals. This hypothesis will need to be investigated more deeply. It is striking that over-expression of PATJ or Pals1 does not disrupt dramatically the organization of TJs as opposed to what is observed in Drosophila ectoderm. In that model, overexpression of Dlt or Crumbs leads to disruption of the ZA and disorganization of the monolayer (Bhat et al., 1999; Wodarz et al., 1995). This difference of phenotype could be explained by the fact that in mammalian epithelial cells, occludin and claudin have a structural role in establishing and reinforcing the ZO (Tsukita and Furuse, 1999). It is thus possible that the Crumbs complex in mammals is a remnant of an ancestor complex essential for the building of the ZA and its function became more regulatory when proteins like claudins were produced to establish the ZO. Experiments using loss of function of members of the Crumbs complex (RNA interference for example) will help to understand the precise role of this complex in regulating the formation of TJs in mammals.
6. Role of the Crumbs complex in the organization of the retina The fact that CRB1 is mutated in some cases of retina degeneration points at a role in maintaining the cellular organization or the normal physiology of rods and cones. Very little is known about the localization of CRB1 in retina: CRB1 mRNA is expressed by the photoreceptors and some of the inter-neurons but not by the retina pigmented epithelial (RPE) cells (den Hollander et al., 2001b; den Hollander et al., 2002) and the protein seems to be accumulated above the junctions making the outer limiting membrane (OLM) (Pellikka et al., 2002). This structure is an alignment of junctions resembling ZAs and provides adhesion between photoreceptors and a specialized type of glial cells that are called Muller cells (Williams et al., 1990). In general, a perturbation of these ZAs leads to disorganization of the photoreceptors and to retina degeneration since the Muller cells are supportive cells, essential for the maintenance of photoreceptor architecture and function (Bunt-Milam et al., 1985). Thus if the
E. Médina et al. / Biology of the Cell 94 (2002)
311
Fig. 4. Comparison of cell polarity determinants of fly photoreceptor and epithelial cell: (A) The photoreceptor cell can be considered as the equivalent of an epithelial cell with a 90° turn. The rhabdomere is a subregion of the apical membrane made up of many microvillae. Although it constitutes the most apical membrane, it is however, devoid of Crumbs (Crb). ZA (in blue) is equivalent in the two kinds of cells. The stalk region (red dotted area) is adjacent and apical to ZA and supports and delimits the rhabdomere. Because Crb, Discs lost (Dlt) and bheavy-Spectrin (bH) are located in this zone, the stalk region seems to be the equivalent of the marginal zone (MZ) of epithelial cell (B).
localization of CRB1 to the membrane just apical of the ZAs of the rods and cones is confirmed by more detailed immunoelectron microscopy studies, CRB1 could play the same role as crumbs in Drosophila photoreceptors. The eye in Drosophila is made of ommatidia and the photoreceptors differentiate from the epithelial cells of the imaginal disc. During the pupal stage, the apical membrane rotates inwardly towards the center of the cluster of photoreceptors and there is a tremendous elongation of this apical domain to form the rhabdomere with flat microvilli (Izaddoost et al., 2002) (Fig. 4 ). The stalk region surrounding this rhabdomere, contains crumbs and Dlt and is located just above the ZAs (Izaddoost et al., 2002; Pellikka et al., 2002). In crb mutants, there is a disorganization of these ZAs and the structure of the rhabdomeres is also affected suggesting that crumbs plays a role in the morphogenesis of both structures (Pellikka et al., 2002). Indeed these two recent studies showed that the cytoplasmic domain of crumbs is necessary for the proper formation of the ZAs while the extracellular domain is essential for the elongation of the rhabdomere. This together with the fact that in human patients, mutation in the extracellular domain of CRB1 leads to photoreceptor degeneration, suggest a
previously unknown role of the extracellular domain of crumbs in either the addition of new membrane or in the building of a stabilizing scaffold in the apical photoreceptor membrane. Thus it seems that Drosophila and mammalian photoreceptors have a lot in common and that the function of the crumbs complex is also conserved in these cells. It is, however, still difficult to attribute retina degeneration in human to either a defect in the proper renewal of the outer segment membrane or in the organization of the ZAs. The fact that, in RP12 patients, degeneration occurs after several years of life suggests more a cumulative defect in membrane addition or recycling of the outer segment. A mouse model deficient in mCRB1 should be very useful to understand the first steps of RP12 and will help to elucidate the respective roles of ZAs maintenance and outer segment renewal in this disease.
7. Perspectives and models It is now well established that Crumbs and the proteins belonging to its complex, like Sdt and Dlt, are essential for
312
E. Médina et al. / Biology of the Cell 94 (2002)
the normal development of a belt of ZAs at the most apical level of the lateral membrane of ectodermally derived epithelial cells and consequently for the normal development of cell polarity in Drosophila. There is, however, little understanding on how the Crumbs complex can regulate the coalescence of the spots of ZAs during gastrulation. There must be a connection between Crumbs and the DEcadherin/catenins complex making up the ZAs to coordinate their assembly but so far no direct link between the SAC where Crumbs is concentrated and the ZAs has been found. Even the identification of Sdt and Dlt as cytoplasmic partners of Crumbs did not provide a link with the ZA since these proteins co-localize with Crumbs and not with DE-cadherin (Tanentzapf et al., 2000). Both Sdt and Dlt can build a scaffold extending to the ZA but unfortunately no partners (except Neurexin IV for Dlt, see above) of these two proteins have been described in Drosophila. Sdt by its SH3 and Hook domains can interact with proline-rich and the 4.1 family proteins, respectively, while Dlt has four PDZ domains that could interact with transmembrane or cortical proteins besides Neurexin IV (Tanentzapf et al., 2000). In mammals, the homologue of Dlt, PATJ, binds to claudin-1 and to ZO-3, two proteins of the ZO but this interaction was not confirmed in vivo (Roh et al., 2002a, 2002b). Unfortunately claudin-like proteins, which are essential for the making of the strands of the ZO, have not been identified in Drosophila and the homologue of ZO-3 is still unknown thus rendering difficulty in extrapolating these data to Drosophila. Pals1, on the other hand, was identified by its binding to mlin7 but nothing is known about the role of this protein in Drosophila epithelial cells. While the identification of the molecular network built around Sdt and Dlt will likely shed light on how Crumbs can regulate the formation of the ZAs, there is now evidence that a connection between the SAC and the ZA exists through the cytoskeleton. Crumbs binds to a specialized spectrin tetramer made of bheavy-Spectrin (Medina et al., 2002) and this protein is also associated with the ZA (Thomas, 2001). Furthermore, in mutants for bheavy-Spectrin, there are defects in the building of the ZAs (Zarnescu and Thomas, 1999) and we showed that Crumbs is necessary for the correct organization of bheavy-Spectrin at the level of the sub apical region (Medina et al., 2002). Thus Crumbs could act in part by concentrating and building the bheavy-Spectrin cortical skeleton in the region of the SAC/ZA, linking the two structures and coordinating the delimitation between the apical and the basolateral membrane. This hypothesis is reinforced by the interaction between Crumbs and DMoesin (Medina et al., 2002) providing a link to the actin cytoskeleton that is also crucial for the development of ZAs and cell polarity (Wodarz, 2002). Since Sdt can potentially interact with DMoesin by its Hook domain (Bachmann et al., 2001; Hong et al., 2001) this would even strengthen the link between the SAC and the cytoskeleton. It is of some importance to note that a mammalian homologue of bheavy-Spectrin has been identified recently and that this protein, called bV-spectrin, is highly enriched in photoreceptors (Stabach and Morrow, 2000). Thus CRB1
could control the development of the ZAs present in photoreceptors as Crumbs regulates their formation in Drosophila. One important question remains about the role of Crb in regulating the formation of the ZA in this organism: why some epithelial structures found in Drosophila embryos (like the gut) are not affected in Crb mutants (Tepass et al., 1990)? A detailed study of the expression and localization of the proteins interacting with Crumbs should help to understand why in these cells, Crumbs is not essential for epithelial polarity. This together with the identification of new proteins belonging to the Crumbs complex might explain the relation between the building of the ZA and the SAC.
Acknowledgements We would like to thank J.P. Arsanto for critically reading this manuscript and members of the Crumbs network for the sharing of unpublished information. This work was supported by Centre National de la Recherche Scientifique CNRS 6156, Université de la Méditerranée, Institut de Biologie du Développement de Marseille, Fondation de France and Association pour la Recherche sur le Cancer 9297, and a EC grant (Crumbs therapeutics) to A. Le Bivic.
References Bachmann, A., Schneider, M., Theilenberg, E., Grawe, F., Knust, E., 2001. Drosophila Stardust is a partner of Crumbs in the control of epithelial cell polarity. Nature 414, 638–643. Baumgartner, S., Littleton, J.T., Broadie, K., Bhat, M.A., Harbecke, R., Lengyel, J.A., Chiquet-Ehrismann, R., Prokop, A., Bellen, H.J., 1996. A Drosophila neurexin is required for septate junction and blood-nerve barrier formation and function. Cell 87, 1059–1068. Bellen, H.J., Lu,Y., Beckstead, R., Bhat, M.A., 1998. Neurexin IV, caspr and paranodin--novel members of the neurexin family: encounters of axons and glia. Trends Neurosci. 21, 444–449. Bhat, M.A., Izaddoost, S., Lu, Y., Cho, K.O., Choi, K.W., Bellen, H.J., 1999. Discs Lost, a novel multi-PDZ domain protein, establishes and maintains epithelial polarity. Cell 96, 833–845. Bilder, D., Li, M., Perrimon, N., 2000. Cooperative regulation of cell polarity and growth by Drosophila tumor suppressors [see comments]. Science 289, 113–116. Bunt-Milam, A.H., Saari, J.C., Klock, I.B., Garwin, G.G., 1985. Zonulae adherentes pore size in the external limiting membrane of the rabbit retina. Invest. Ophthalmol. Vis. Sci. 26, 1377–1380. Carlson, S.D., Juang, J.L., Hilgers, S.L., Garment, M.B., 2000. Blood barriers of the insect. Annu. Rev. Entomol. 45, 151–174. den Hollander, A.I., ten Brink, J.B., de Kok, Y.J., van Soest, S., van den Born, L.I., van Driel, M.A., van de Pol, D.J., Payne, A.M., Bhattacharya, S.S., Kellner, U., Hoyng, C.B., Westerveld, A., Brunner, H.G., Bleeker-Wagemakers, E.M., Deutman, A.F., Heckenlively, J.R., Cremers, F.P., Bergen, A.A., 1999. Mutations in a human homologue of Drosophila crumbs cause retinitis pigmentosa (RP12). Nat. Genet. 23, 217–221.
E. Médina et al. / Biology of the Cell 94 (2002) den Hollander, A.I., Heckenlively, J.R., van den Born, L.I., de Kok, Y.J., van der Velde-Visser, S.D., Kellner, U., Jurklies, B., van Schooneveld, M.J., Blankenagel, A., Rohrschneider, K., Wissinger, B., Cruysberg, J.R., Deutman, A.F., Brunner, H.G., Apfelstedt-Sylla, E., Hoyng, C.B., Cremers, F.P., 2001a. Leber congenital amaurosis and retinitis pigmentosa with Coats-like exudative vasculopathy are associated with mutations in the crumbs homologue 1 (CRB1) gene. Am. J. Hum. Genet. 69, 198–203. den Hollander, A.I., Johnson, K., de Kok, Y.J., Klebes, A., Brunner, H.G., Knust, E., Cremers, F.P., 2001b. CRB1 has a cytoplasmic domain that is functionally conserved between human and Drosophila. Hum. Mol. Genet. 10, 2767–2773. den Hollander, A.I., Ghiani, M., de Kok, Y.J., Wijnholds, J., Ballabio, A., Cremers, F.P., Broccoli, V., 2002. Isolation of Crb1, a mouse homologue of Drosophila crumbs, and analysis of its expression pattern in eye and brain. Mech. Dev. 110, 203–207. Fanning, A.S., Anderson, J.M., 1999. Protein modules as organizers of membrane structure. Curr. Opin. Cell Biol. 11, 432–439. Grawe, F., Wodarz, A., Lee, B., Knust, E., Skaer, H., 1996. The Drosophila genes crumbs and stardust are involved in the biogenesis of adherens junctions. Development 122, 951–959. Hong, Y., Stronach, B., Perrimon, N., Jan, L.Y., Jan, Y.N., 2001. Drosophila Stardust interacts with Crumbs to control polarity of epithelia but not neuroblasts. Nature 414, 634–638. Hynes, R.O., 1992. Integrins: versatility, modulation, and signaling in cell adhesion. Cell 69, 11–25. Izaddoost, S., Nam, S.C., Bhat, M.A., Bellen, H.J., Choi, K.W., 2002. Drosophila Crumbs is a positional cue in photoreceptor adherens junctions and rhabdomeres. Nature 416, 178–183. Jürgens, G., Wieschaus, E., Nüsslein-Volhard, C., Kluding, H., 1984. Mutations affecting the pattern of the larval cuticle in Drosophila melanogaster. Roux’s Arch. Dev. Biol. 193, 283–295. Kamberov, E., Makarova, O., Roh, M., Liu, A., Karnak, D., Straight, S., Margolis, B., 2000. Molecular cloning and characterization of Pals, proteins associated with mLin-7. J. Biol. Chem. 275, 11425–11431. Klebes, A., Knust, E., 2000. A conserved motif in Crumbs is required for E-cadherin localisation and zonula adherens formation in Drosophila. Curr. Biol. 10, 76–85. Lemmers, C., Medina, E., Delgrossi, M.H., Michel, D., Arsanto, J.P., Le Bivic, A., 2002. hINADl/PATJ, a homolog of discs lost, interacts with crumbs and localizes to tight junctions in human epithelial cells. J. Biol. Chem. 277, 25408–25415. Leptin, M., 1999. Gastrulation in Drosophila: the logic and the cellular mechanisms. Embo J. 18, 3187–3192. Medina, E., Williams, J., Klipfell, E., Zarnescu, D., Thomas, G., Le Bivic, A., 2002. Crumbs interacts with moesin and {beta}Heavy-spectrin in the apical membrane skeleton of Drosophila. J. Cell Biol. 158, 941–951. Mostov, K.E., Verges, M., Altschuler, Y., 2000. Membrane traffic in polarized epithelial cells. Curr. Opin. Cell Biol. 12, 483–490. Nagafuchi, A., 2001. Molecular architecture of adherens junctions. Curr. Opin. Cell Biol. 13, 600–603. Pellikka, M., Tanentzapf, G., Pinto, M., Smith, C., McGlade, C.J., Ready, D.F., Tepass, U., 2002. Crumbs, the Drosophila homologue of human CRB1/RP12, is essential for photoreceptor morphogenesis. Nature 416, 143–149.
313
Perrimon, N., Lanjuin, A., Arnold, C., Noll, E., 1996. Zygotic lethal mutations with maternal effect phenotypes in Drosophila melanogaster. II. Loci on the second and third chromosomes identified by P-elementinduced mutations. Genetics 144, 1681–1692. Rodriguez-Boulan, E., Nelson, W.J., 1989. Morphogenesis of the polarized epithelial cell phenotype. Science 245, 718–725. Roh, M.H., Liu, C.J., Laurinec, S., Margolis, B., 2002a. The carboxyl terminus of zona occludens-3 binds and recruits a mammalian homologue of discs lost to tight junctions. J. Biol. Chem. 277, 27501–27509. Roh, M.H., Makarova, O., Liu, C.J., Shin, K., Lee, S., Laurinec, S., Goyal, M., Wiggins, R., Margolis, B., 2002b. The Maguk protein, Pals1, functions as an adapter, linking mammalian homologues of Crumbs and Discs Lost. J. Cell Biol. 157, 161–172. Stabach, P.R., Morrow, J.S., 2000. Identification and characterization of beta V spectrin, a mammalian ortholog of Drosophila beta H spectrin. J. Biol. Chem. 275, 21385–21395. Tanentzapf, G., Smith, C., McGlade, J., Tepass, U., 2000. Apical, Lateral, and Basal Polarization Cues Contribute to the Development of the Follicular Epithelium during Drosophila Oogenesis. J. Cell Biol. 151, 891–904. Tepass, U., Theres, C., Knust, E., 1990. Crumbs encodes an EGF-like protein expressed on apical membranes of Drosophila epithelial cells and required for organization of epithelia. Cell 61, 787–799. Tepass, U., Knust, E., 1993. Crumbs and stardust act in a genetic pathway that controls the organization of epithelia in Drosophila melanogaster. Dev. Biol. 159, 311–326. Tepass, U., 1996. Crumbs, a component of the apical membrane, is required for zonula adherens formation in primary epithelia of Drosophila. Dev. Biol. 177, 217–225. Tepass, U., Tanentzapf, G., Ward, R., Fehon, R., 2001. Epithelial cell polarity and cell junction in Drosophila. Annu. Rev. Genet. 35, 747–784. Tepass, U., 2002. Adherens junctions: new insight into assembly, modulation and function. Bioessays 24, 690–695. Thomas, G., 2001. Spectrin: The ghost in the machine. Bioessays 23, 152–160. Tsukita, S., Furuse, M., 1999. Occludin and claudins in tight-junction strands: leading or supporting players?. Trends Cell Biol. 9, 268–273. Vasioukhin, V., Fuchs, E., 2001. Actin dynamics and cell-cell adhesion in epithelia. Curr. Opin. Cell Biol. 13, 76–84. Williams, D.S., Arikawa, K., Paallysaho, T., 1990. Cytoskeletal components of the adherens junctions between the photoreceptors and the supportive Muller cells. J. Comp. Neurol. 295, 155–164. Wodarz, A., Grawe, F., Knust, E., 1993. CRUMBS is involved in the control of apical protein targeting during Drosophila epithelial development. Mech. Dev. 44, 175–187. Wodarz, A., Hinz, U., Engelbert, M., Knust, E., 1995. Expression of crumbs confers apical character on plasma membrane domains of ectodermal epithelia of Drosophila. Cell 82, 67–76. Wodarz, A., 2002. Establishing cell polarity in development. Nat. Cell Biol. 4, E39–44. Yeaman, C., Grindstaff, K.K., Hansen, M.D., Nelson, W.J., 1999. Cell polarity: Versatile scaffolds keep things in place. Curr. Biol. 9, R515–517. Zarnescu, D.C., Thomas, G.H., 1999. Apical spectrin is essential for epithelial morphogenesis but not apicobasal polarity in Drosophila. J. Cell Biol. 146, 1075–1086.
Original article
Role of the Crumbs complex in the regulation of junction formation in Drosophila and mammalian epithelial cells Emmanuelle Médina, Céline Lemmers, Lydie Lane-Guermonprez, André Le Bivic * NMDA-IBDM, UMR 6156, Campus de Luminy, Université de la Méditerranée, case 907, 13288 Marseille cedex 09, France Received 9 September 2002; accepted 11 September 2002
Abstract The formation of a belt-like junctional complex separating the apical from the lateral domain is an essential step in the differentiation of epithelial cells. Thus protein complexes regulating this event are of first importance for the development of cell polarity and physiological functions of epithelial tissues. In Drosophila, the discovery of a gene, crb, controlling the coalescence of the spots of zonula adherens (ZA) into a adhesive ring around the cells was a major step. We know now that Crumbs, the product of crb is an apical transmembrane protein conserved in mammals and that it interacts by its cytoplasmic domain with two cortical modular proteins, Stardust (Sdt) and Discs lost (Dlt) that are also essential for the correct assembly of the ZA. These two proteins are also conserved in mammals and it is most likely that the Crumbs complex plays a similar role in very different species. Recently, we have shown that Crumbs interacts with the cortical cytoskeleton made of DMoesin and bheavy-Spectrin and this connection could explain in part the role of Crumbs in building the ZA. Future work will help to understand several aspects of the Crumbs complex that are still unknown, like the role of the large extracellular domain or the precise function of Sdt and Dlt in the building of the ZA. Finding an answer to these questions will help to find new therapies for Retinitis pigmentosa and other retina degeneration in which CRB1, the human homologue of crb, has been involved. © 2002 Éditions scientifiques et médicales Elsevier SAS. All rights reserved. Keywords: Epithelial cells; Polarity, Junctions; Retina; Moesin
1. Epithelial polarity Multicellular organisms could develop with the apparition of epithelial cells that physically separate the outside world from the inside of the body. With this invention, organisms could ensure the homeostasis of the liquid bathing their internal tissues, select nutrients and for some of them, isolate their nervous system. The making of epithelial cells, crucial for evolution, involved the creation of several mechanisms like the formation of specialized junctions between cells and the sorting and recycling of plasma membrane components. The organization of epithelial cells is now well known and these cells stand out from other cells by the division of their plasma membrane in at least two separate domains, the apical (facing the outside) and the basolateral (in contact with the
* Corresponding author. Tel.: +33-4-91-26-97-41; fax: +33-4-91-26-97-48. E-mail address: [email protected] (A. Le Bivic) © 2002 Éditions scientifiques et médicales Elsevier SAS. All rights reserved. PII: S0248-4900(02)00004-7
other cells and the internal medium) domains. These domains are first established by cell–cell contacts that allow adhesion molecules to cluster and nucleate other scaffold proteins that in turn recruit other transmembrane proteins (Yeaman et al., 1999). Then, junctions that physically separate the plasma membrane domains are formed and reinforced to allow the targeting of newly synthesized components to the recently formed apical and basolateral membranes (Mostov et al., 2000). Transmembane proteins that are able to nucleate and build networks of cytoplasmic proteins are called extracellular cell surface organizers (ECSOs). These proteins give cues to the cells about their position and how to orient their future polarity. In mammalian epithelial cells, integrins and cadherins play such a role (Hynes, 1992) but their precise contribution during the process of establishing and maintaining epithelial polarity in vivo is still hard to appreciate. Most of the work done on these mammalian ECSOs was performed on epithelial cell cultures that however do not always recreate the in vivo complexity (Rodriguez-Boulan and Nelson, 1989). On the other hand, Drosophilamelano-
306
E. Médina et al. / Biology of the Cell 94 (2002)
Fig. 1. Comparison of invertebrate and vertebrate apical junctional complexes: ZA is common to vertebrate and invertebrate epithelial cells. The components of this junction are well conserved among the species with the transmembrane protein E-cadherin (DE-Cad and E-Cad, in (A) and (B), respectively), connected to the actin cytoskeleton through a- and b-catenin (Da-Cat and Arm in (A) and a-Cat and b-Cat in (B), respectively). In Drosophila (A), the sub-apical complex (SAC) is apical to the ZA whereas in mammals (B), the tight junction (TJ) is located at this position. In the SAC, the transmembrane protein Crumbs (Crb) has been shown to be essential for the ZA morphogenesis. This epithelial cell surface organizer (ECSO) plays a crucial role in epithelial morphogenesis through its interaction with Stardust (Sdt) and Discs lost (Dlt), and is linked to cortical cytoskeleton through DMoesin (Dmoe). In Drosophila (A), another complex essential for polarity (the Bazooka (Baz)-aPKC-DPar6 complex) is also found in this region. In mammals (B), the homologues of the SAC components are located at the TJ level and are in close relation with the TJ components (the transmembrane proteins Claudin (Clau) and Occludin (Occ), and the MAGUKs ZO-1/ZO-2/ZO-3). In Drosophila (A), the septate junction (SJ) lies basal to the ZA and plays a role in controlling paracellular diffusion. This junction is formed by the transmembrane protein Neurexin IV (NrxIV) and several proteins important for epithelial polarity such as Discs large (Dlg), Lethal giant larvae (Lgl) and Scribble. In mammals (B), the Dlg, Lgl and Scribble homologues are also found below the ZA and colocalize with the HER2/Erbin complex.
gaster provides a powerful and important complementary genetic model. In Drosophila, epithelial cell organization is close to the one observed in mammals with, however, some differences in the structures that build the junctional complex separating the apical from the lateral domain (Fig. 1 ). In mammalian cells the most apical junction is the zonula occludens (ZO) made of polytopic membrane proteins like claudins and occludins. This structure provides both the control of paracellular transport and of cell–cell adhesion and is located just above the zonula adherens (ZA), a junction common to vertebrates and invertebrates (Fig. 1). The ZA is made up of a calcium-dependent adhesion protein, E-cadherin, which is connected to an intracellular network of
actin through its partners a- and b-catenins (Vasioukhin and Fuchs, 2001). The molecular components of this junction are extremely well conserved between Drosophila and mammals, suggesting that its formation is a crucial step for epithelial morphogenesis (Nagafuchi, 2001; Tepass, 2002). In Drosophila, below the ZA, septate junctions (SJs) play a role in preventing and controlling paracellular diffusion (Carlson et al., 2000). These junctions can have different aspects when visualized by electron microscopy techniques since they can be pleated or smooth depending on the tissue observed (Baumgartner et al., 1996). Molecularly, these junctions contain a transmembrane protein, Neurexin IV, which is a homologue of mammalian paranodin, involved in the formation of
E. Médina et al. / Biology of the Cell 94 (2002)
Ranvier node (Bellen et al., 1998). Several proteins, Discs large (Dlg), lethal giant larvae (Lgl) and Scribble, important for epithelial organization are associated to SJs, indicating that these junctions also play a role in epithelial polarity besides the fence function (Bilder et al., 2000). In Drosophila, recent findings have pointed out the crucial role of the region located just above the ZA, called the sub-apical complex (SAC) (Fig. 1). In fact, many of the proteins localized in this region, have now been shown to be essential for the positioning of the ZA and thus for epithelial polarity in general. In particular, the first identified transmembrane component of this region, Crumbs, has become a central piece for the organization of the whole junctional complex comprising the ZA.
2. An apical organizer, Crumbs In the last 10 years, several crucial components for epithelial morphogenesis and polarity have been identified using Drosophila powerful genetic screens (Perrimon et al., 1996). One of these genes was called crumbs (crb) because the cuticle of embryos bearing this mutation was reduced to small patches reminding of bread crumbs (Jürgen et al., 1984). Embryos with crb loss of function mutations die shortly after gastrulation, a stage involving intensive epithelial movement and reorganization in Drosophila(Tepass et al., 1990). This early lethal phenotype is the resultant of a loss of epithelial organization of the ectodermal layer in which there is a rounding of epidermal cells with a high degree of apoptosis (Leptin, 1999; Tepass et al., 1990). The protein encoded by crb and named Crumbs was identified by the group of E. Knust and is a transmembrane protein with a small cytoplasmic domain of 37 amino acids and a large extracellular domain made of 30 EGF like repeats and four laminin G-like repeats (Tepass et al., 1990) (Fig. 2 ). Crumbs is expressed in all the primary and secondary epithelia derived from the ectoderm and, interestingly, one of the loss of function mutants of crb, the crb8F105 allele, is a stop mutation in the cytoplasmic tail of Crumbs leaving only half of it (Wodarz et al., 1993). This was the first indication of a crucial role of the Crumbs cytoplasmic domain in its epithelial function and it was later confirmed by the fact that the expression of a transgene, deleted of the part coding for the extracellular domain (Myc-intrawt), was able to rescue the epithelial organization of the ectoderm (Wodarz et al., 1995). A more precise analysis of the crb phenotype showed that ZAs, which first appear as spots on the lateral membrane can not coalesce and form a belt-like structure around the cells in crb mutants, leading to the hypothesis that Crumbs is necessary for their positioning and integrity (Tepass, 1996; Grawe et al., 1996). Meanwhile, the over-expression of crumbs or Myc-intrawt was shown to provoke an expansion of the apical domain and a disorganization of the epithelial layer confirming that Crumbs is involved in the correct positioning and assembly of ZAs and also regulates the shape of the apical
307
membrane by a still unknown mechanism (Wodarz et al., 1995).
3. The Crumbs complex How Crumbs is involved in the correct positioning and assembly of the ZA through its cytoplasmic domain is a major question and a beginning of an answer was brought by the identification of two proteins, Discs lost (Dlt) and Stardust (Sdt), that interact with Crumbs (Fig. 2). Stardust (sdt) was the name given to Drosophila mutants that showed a phenotype very similar to crb. sdt is a zygotic gene like crb and, in sdt mutants, epidermal cells loose polarity, become round and the epithelial layer is disorganized. Furthermore, sdt could be positioned downstream in the crb cascade since over-expression of sdt could restore some of crb defects (Tepass and Knust, 1993). While the sdt mutation was described in the early 1990s, identification of the protein encoded by the sdt locus was only published recently (Bachmann et al., 2001; Hong et al., 2001). Sdt belongs to the Maguk (membrane-associated guanylate kinase) family like ZO-1 (zonula occludens 1) or PSD-95 (post synaptic density 95) in mammals (Fanning and Anderson, 1999). Two isoforms of Sdt are found in D. melanogaster and both have a single PDZ (PSD-95, Discs large, ZO-1) domain, a SH3 (Src homology region 3) domain and a guanylate kinase (GUK) domain (Bachmann et al., 2001; Hong et al., 2001). Sdt colocalizes with Crumbs to the SAC and interacts directly with the last four amino acids ERLI of Crumbs cytoplasmic tail (Fig. 3 ) but it was not shown whether this interaction is mediated through the PDZ domain of Sdt (Bachmann et al., 2001; Hong et al., 2001), as it is likely to occur. Another protein was shown to interact with Crumbs and this protein is the product of the Discs lost (dlt) gene. dlt is a gene maternally expressed as opposed to crb and sdt and thus was not discovered in the same genetic screen. In fact, Dlt was isolated by a two-hybrid screen (Bhat et al., 1999), using the cytoplasmic tail of Neurexin IV (NRX IV), a protein of the septate junctions (Baumgartner et al., 1996). Curiously, in mature epithelial epidermal cells of Drosophila embryos, Dlt colocalizes to the SAC with Crumbs and Sdt but not with Nrx IV questioning the physiological relevance of this interaction in fully polarized epithelial cells (Bhat et al., 1999). Given the co-localization with Crumbs and the fact that Dlt is a cortical protein made of four PDZ domains, it was hypothesized that the two proteins interact directly but this hypothesis has yet to be demonstrated. More biochemical work on the interaction between Crumbs and Dlt in Drosophila should be performed in order to answer this crucial point. In fact, if Dlt binds directly to the PDZ-binding motif (ERLI) of Crumbs, there must be some regulation to control its competition with Sdt for the same binding site. The fact that Dlt is already expressed during cellularization and that overexpression of Crumbs in fly embryos leads to redistribution of Dlt all around epithelial cells, while Sdt is depleted from
308
E. Médina et al. / Biology of the Cell 94 (2002)
Fig. 2. Comparison of mammalian homologues of Drosophila Crumbs, Discs lost and Stardust. (A) Drosophila Crumbs (dCRB) has three human homologues: CRB1, CRB2 and CRB3. The extracellular domains are composed by a signal peptide (black box), 30 EGF like-repeats (red boxes) and four Laminin-A G domain-like motifs (green boxes) for dCRB; EGF-like repeats and three Laminin-A G domain-like motifs for CRB1 and 15 EGF-like repeats and two Laminin A G domain-like motifs. CRB3 has a very short extracellular domain and a signal peptide. All Crumbs proteins share a small cytoplasmic tail composed of 37 amino acids for dCRB, CRB1 and CRB2. CRB3 has 40 amino acids in its cytoplasmic part. Human MUPP1 and PATJ are both similar to Drosophila Dlt. They share a MAGUK recruitment domain (MRE, triangles) and 13, 8 and 4 PDZ domains, respectively (ovals). Mus musculus Pals1 is the Stardust homologue. They have multiple protein–protein interaction domains including one or two Lin-2/lin-7 domains, respectively (L27, L27 N and L27 C), a PDZ domain (oval), a SH3 domain (square), a protein 4.1-binding domain (circle) and a GuK domain (hexagon). (B) Sequence alignment of the conserved cytoplasmic domain of Drosophila Crumbs (dCRB), human CRB1 (hCRB1), CRB2 (hCRB2), CRB3 (hCRB3), C. elegans CRB1 (CeCRB1) and CRB-like (CeCRL1). Identical and conserved amino acids are in black and gray, respectively.
the apical membrane (Bachmann et al., 2001; Hong et al., 2001) indicate that these two proteins Dlt and Sdt have a different regulation of their association to the plasma membrane. In that case, Dlt could bind to another apical transmembrane protein through one of its four PDZ domains, while Sdt would follow ectopic Crumbs and thus be somehow depleted from the apical side. Our current knowledge on the composition of the Crumbs complex in mammals does not favor a direct interaction between Crumbs and Dlt (see below).
4. Crumbs connection to the cortical cytoskeleton It is now well established that Crumbs binds to Sdt by its last four amino acids ERLI but there are other residues in Crumbs cytoplasmic tail that are also crucial for its function. In particular, it was shown that there is a tyrosine at position 10 (Y10) and a glutamic acid at position 16 (E16) which are essential to rescue a normal phenotype in Crumbs mutants (Klebes and Knust, 2000). This portion (GTY) of Crumbs cytoplasmic domain containing Y10 forms a consensus motif
E. Médina et al. / Biology of the Cell 94 (2002)
309
Fig. 3. Comparison of invertebrate and vertebrate Crumbs complexes: (A) In Drosophila, Crumbs (Crb) interacts by its four last amino acids ERLI with the proteins Stardust (Sdt) and Discs lost (Dlt). The interaction between Crb and Sdt is direct but it is not yet known what domain of Sdt is involved. The interaction with Dlt has not been extensively studied but it does not seem to be direct. It has also been shown that Crb interacts with DMoesin and bheavy-Spectrin (bH). Little is known about how Crb recruits these proteins and whether they are or not in a same complex but two amino acids of the Crb cytoplasmic domain, R7 and Y10, are essential for this process. These two proteins, DMoesin and bH provide a link between the Crb complex and actin. (B) In mammals, hCRB interacts directly by its four last amino acids ERLI with the PDZ domain of Pals1, the Sdt homologue. Pals1 directly interacts by its L27 N domain with the MRE domain of PATJ, a Dlt homologue. PATJ links the hCRB complex to the TJ components through direct interactions with Claudin1 and ZO-3. These interactions are mediated by the C-terminal amino acids of ZO-3 and Claudin1 and the PDZ domains 6 and 8 of PATJ, respectively.
for the binding of proteins of the 4.1 family of proteins (Klebes and Knust, 2000), including the ERM proteins (Ezrin/radixin/moesin) (Tepass et al., 2001). It has thus been hypothesized that Crb could bind proteins of this family providing a linkage to a spectrin/actin cytoskeleton (Klebes and Knust, 2000). In fact, in Drosophila, a protein of the 4.1 family, DMoesin is accumulated in the SAC zone and we have shown recently that DMoesin and Crumbs are coimmunoprecipitated together with Dlt from fly embryos indicating that DMoesin could be a likely candidate to bind to the GTY motif of Crumbs. In fact, mutations of theY10 or the R7 just upstream of the GTY motif impaired the connection to DMoesin as measured by capping experiments (Medina et al., 2002). In addition, over-expression of Crumbs in epidermal cells induced a redistribution of both DMoesin and actin confirming that Crumbs might be linked to the cortical cytoskeleton (Medina et al., 2002). Furthermore, it was shown that bheavy-Spectrin, another protein of the cortical cytoskeleton, was redistributed during Crumbs over-expression (Wodarz et al., 1995) suggesting that it could associate to the
same complex. There are no published DMoesin mutants and thus it is difficult to know if DMoesin plays a role in the regulation of ZA formation. Mutants for bheavy-Spectrin are called karst and these mutants show clear defects of the ZA formation during embryogenesis linking it to the Crb phenotype (Zarnescu and Thomas, 1999). In Crb mutants, bheavySpectrin is mislocalized while in karst mutants Crumbs is normally concentrated in the SAC region suggesting that Crb acts upstream of karst(Medina et al., 2002). bheavy-Spectrin has no known domain to bind directly to DMoesin and it is still a challenge to understand how the molecular connection between Crumbs and bheavy-Spectrin occurs and is regulated. Interestingly, Crumbs could also be connected to the cortical cytoskeleton through Sdt. Sdt possesses a 4.1 protein family binding domain and thus could interact with DMoesin (Bachmann et al., 2001; Hong et al., 2001). This interaction would reinforce Crumbs connection to the cytoskeleton and to actin, building a cytoskeleton around the ZA. Future work will provide a new understanding on how the regulation of these multiple connections between the SAC and the ZA is operated.
310
E. Médina et al. / Biology of the Cell 94 (2002)
5. The Crumbs complex in mammals Until the discovery of a human locus involved in a human disease and coding for a protein homologue to Crumbs, in 1999, there was no evidence for a conservation of Crumbs between invertebrates and vertebrates. Thus, the discovery that a gene mutated in Retinitis pigmentosa group 12 (RP12) was very similar to Drosophila crb was a major step (den Hollander et al., 1999) (Fig. 2). This gene was called CRB1 and it produces two isoforms, one secreted or membraneassociated containing the extracellular domain and a putative transmembrane domain and one very similar to Crumbs with a transmembrane and a cytoplasmic domain sharing 63% of identity with Crb (den Hollander et al., 2001b) (Fig. 2). CRB1 is a bona fide Crumbs, since expression of its intracellular domain in Drosophila mutants for crb can rescue the cuticle phenotype to the same extent as transgenic Crumbs (den Hollander et al., 2001b). In its extracellular domain, CRB1 is made of 19 EGF-like and three LG domains (Fig. 2) and most of the mutations found in patients, map to this extracellular domain (den Hollander et al., 1999). CRB1 is also mutated in a more severe form of retina degeneration, the Leber Congenital Amaurosis syndrome (den Hollander et al., 2001a). CRB1 expression is restricted to retina and some parts of the brain both in human and mouse (den Hollander et al., 2002), suggesting that in vertebrates a tissue specialization of Crumbs might have occurred. With the sequencing of the human genome, there are now two additional genes, CRB2 and CRB3 that code for proteins with a highly conserved cytoplasmic tail (67 and 54% of identity with Crumbs, respectively) (Fig. 2) (Lemmers et al., 2002). CRB2 has a large extracellular domain with 15 EGF-like and two LG domains while CRB3 exhibits a very short extracellular domain with no recognizable protein domain in it (Fig. 2). So far the tissue and cellular distribution of these two gene products is not known and it will be of first interest to see if one of them is present in epithelial cells and has a conserved function in the control of epithelial polarity. Mouse and human homologues of Sdt and Dlt have been described recently suggesting that the whole Crumbs complexis conserved in vertebrates. The orthologue of Sdt is a protein called Pals1 (protein associated to lin 7) that also belongs to the MAGUK family (Kamberov et al., 2000). Pals1 binds to a domain of mlin7 called L27 and possesses two L27 domains called L27 N-terminal and L27 C-terminal domains (Fig. 2). Pals1 is concentrated at the level of TJs and binds through its PDZ domain to the ERLI motif of CRB1 and to one homologue of Dlt by its L27 N-terminal domain (Roh et al., 2002a, 2002b). One homologue of Dlt was called PATJ (protein associated to tight junctions or Pals1associated tight junction) and it is a membrane associated protein with at least eight PDZ domains (Lemmers et al., 2002; Roh et al., 2002a, 2002b) (Fig. 2). PATJ binds to Pals1 by its N-terminal domain and does not seem to bind directly to CRB1 or 3 (Lemmers et al., 2002; Roh et al., 2002a, 2002b). PATJ is highly concentrated at the level of the ZO
and is also found associated to the apical membrane confirming that it is a likely homologue of Dlt (Lemmers et al., 2002). Since the endogenous CRB is not yet identified in epithelial cells, CRB1 was exogenously expressed and was also concentrated at the level of the TJs, strongly indicating that a conserved Crumbs complex regulates the formation of TJs in mammals. Indeed, over-expression of PATJ induced a loss of ZO-3 from the TJs (Lemmers et al., 2002) and it was recently shown that the PDZ domain 6 of PATJ is able to interact with ZO-3 providing a direct link with ZO (Roh et al., 2002a, 2002b) (Fig. 3). By RT-PCR we have found that CRB3 is detectable in human intestinal Caco-2 cells and thus could interact with Pals1 (Lemmers et al., 2002). This is reinforced by the fact that CRB3 is able to coimmunoprecipitate PATJ, indicating that CRB3 could be a good candidate for being an epithelial CRB in mammals. This hypothesis will need to be investigated more deeply. It is striking that over-expression of PATJ or Pals1 does not disrupt dramatically the organization of TJs as opposed to what is observed in Drosophila ectoderm. In that model, overexpression of Dlt or Crumbs leads to disruption of the ZA and disorganization of the monolayer (Bhat et al., 1999; Wodarz et al., 1995). This difference of phenotype could be explained by the fact that in mammalian epithelial cells, occludin and claudin have a structural role in establishing and reinforcing the ZO (Tsukita and Furuse, 1999). It is thus possible that the Crumbs complex in mammals is a remnant of an ancestor complex essential for the building of the ZA and its function became more regulatory when proteins like claudins were produced to establish the ZO. Experiments using loss of function of members of the Crumbs complex (RNA interference for example) will help to understand the precise role of this complex in regulating the formation of TJs in mammals.
6. Role of the Crumbs complex in the organization of the retina The fact that CRB1 is mutated in some cases of retina degeneration points at a role in maintaining the cellular organization or the normal physiology of rods and cones. Very little is known about the localization of CRB1 in retina: CRB1 mRNA is expressed by the photoreceptors and some of the inter-neurons but not by the retina pigmented epithelial (RPE) cells (den Hollander et al., 2001b; den Hollander et al., 2002) and the protein seems to be accumulated above the junctions making the outer limiting membrane (OLM) (Pellikka et al., 2002). This structure is an alignment of junctions resembling ZAs and provides adhesion between photoreceptors and a specialized type of glial cells that are called Muller cells (Williams et al., 1990). In general, a perturbation of these ZAs leads to disorganization of the photoreceptors and to retina degeneration since the Muller cells are supportive cells, essential for the maintenance of photoreceptor architecture and function (Bunt-Milam et al., 1985). Thus if the
E. Médina et al. / Biology of the Cell 94 (2002)
311
Fig. 4. Comparison of cell polarity determinants of fly photoreceptor and epithelial cell: (A) The photoreceptor cell can be considered as the equivalent of an epithelial cell with a 90° turn. The rhabdomere is a subregion of the apical membrane made up of many microvillae. Although it constitutes the most apical membrane, it is however, devoid of Crumbs (Crb). ZA (in blue) is equivalent in the two kinds of cells. The stalk region (red dotted area) is adjacent and apical to ZA and supports and delimits the rhabdomere. Because Crb, Discs lost (Dlt) and bheavy-Spectrin (bH) are located in this zone, the stalk region seems to be the equivalent of the marginal zone (MZ) of epithelial cell (B).
localization of CRB1 to the membrane just apical of the ZAs of the rods and cones is confirmed by more detailed immunoelectron microscopy studies, CRB1 could play the same role as crumbs in Drosophila photoreceptors. The eye in Drosophila is made of ommatidia and the photoreceptors differentiate from the epithelial cells of the imaginal disc. During the pupal stage, the apical membrane rotates inwardly towards the center of the cluster of photoreceptors and there is a tremendous elongation of this apical domain to form the rhabdomere with flat microvilli (Izaddoost et al., 2002) (Fig. 4 ). The stalk region surrounding this rhabdomere, contains crumbs and Dlt and is located just above the ZAs (Izaddoost et al., 2002; Pellikka et al., 2002). In crb mutants, there is a disorganization of these ZAs and the structure of the rhabdomeres is also affected suggesting that crumbs plays a role in the morphogenesis of both structures (Pellikka et al., 2002). Indeed these two recent studies showed that the cytoplasmic domain of crumbs is necessary for the proper formation of the ZAs while the extracellular domain is essential for the elongation of the rhabdomere. This together with the fact that in human patients, mutation in the extracellular domain of CRB1 leads to photoreceptor degeneration, suggest a
previously unknown role of the extracellular domain of crumbs in either the addition of new membrane or in the building of a stabilizing scaffold in the apical photoreceptor membrane. Thus it seems that Drosophila and mammalian photoreceptors have a lot in common and that the function of the crumbs complex is also conserved in these cells. It is, however, still difficult to attribute retina degeneration in human to either a defect in the proper renewal of the outer segment membrane or in the organization of the ZAs. The fact that, in RP12 patients, degeneration occurs after several years of life suggests more a cumulative defect in membrane addition or recycling of the outer segment. A mouse model deficient in mCRB1 should be very useful to understand the first steps of RP12 and will help to elucidate the respective roles of ZAs maintenance and outer segment renewal in this disease.
7. Perspectives and models It is now well established that Crumbs and the proteins belonging to its complex, like Sdt and Dlt, are essential for
312
E. Médina et al. / Biology of the Cell 94 (2002)
the normal development of a belt of ZAs at the most apical level of the lateral membrane of ectodermally derived epithelial cells and consequently for the normal development of cell polarity in Drosophila. There is, however, little understanding on how the Crumbs complex can regulate the coalescence of the spots of ZAs during gastrulation. There must be a connection between Crumbs and the DEcadherin/catenins complex making up the ZAs to coordinate their assembly but so far no direct link between the SAC where Crumbs is concentrated and the ZAs has been found. Even the identification of Sdt and Dlt as cytoplasmic partners of Crumbs did not provide a link with the ZA since these proteins co-localize with Crumbs and not with DE-cadherin (Tanentzapf et al., 2000). Both Sdt and Dlt can build a scaffold extending to the ZA but unfortunately no partners (except Neurexin IV for Dlt, see above) of these two proteins have been described in Drosophila. Sdt by its SH3 and Hook domains can interact with proline-rich and the 4.1 family proteins, respectively, while Dlt has four PDZ domains that could interact with transmembrane or cortical proteins besides Neurexin IV (Tanentzapf et al., 2000). In mammals, the homologue of Dlt, PATJ, binds to claudin-1 and to ZO-3, two proteins of the ZO but this interaction was not confirmed in vivo (Roh et al., 2002a, 2002b). Unfortunately claudin-like proteins, which are essential for the making of the strands of the ZO, have not been identified in Drosophila and the homologue of ZO-3 is still unknown thus rendering difficulty in extrapolating these data to Drosophila. Pals1, on the other hand, was identified by its binding to mlin7 but nothing is known about the role of this protein in Drosophila epithelial cells. While the identification of the molecular network built around Sdt and Dlt will likely shed light on how Crumbs can regulate the formation of the ZAs, there is now evidence that a connection between the SAC and the ZA exists through the cytoskeleton. Crumbs binds to a specialized spectrin tetramer made of bheavy-Spectrin (Medina et al., 2002) and this protein is also associated with the ZA (Thomas, 2001). Furthermore, in mutants for bheavy-Spectrin, there are defects in the building of the ZAs (Zarnescu and Thomas, 1999) and we showed that Crumbs is necessary for the correct organization of bheavy-Spectrin at the level of the sub apical region (Medina et al., 2002). Thus Crumbs could act in part by concentrating and building the bheavy-Spectrin cortical skeleton in the region of the SAC/ZA, linking the two structures and coordinating the delimitation between the apical and the basolateral membrane. This hypothesis is reinforced by the interaction between Crumbs and DMoesin (Medina et al., 2002) providing a link to the actin cytoskeleton that is also crucial for the development of ZAs and cell polarity (Wodarz, 2002). Since Sdt can potentially interact with DMoesin by its Hook domain (Bachmann et al., 2001; Hong et al., 2001) this would even strengthen the link between the SAC and the cytoskeleton. It is of some importance to note that a mammalian homologue of bheavy-Spectrin has been identified recently and that this protein, called bV-spectrin, is highly enriched in photoreceptors (Stabach and Morrow, 2000). Thus CRB1
could control the development of the ZAs present in photoreceptors as Crumbs regulates their formation in Drosophila. One important question remains about the role of Crb in regulating the formation of the ZA in this organism: why some epithelial structures found in Drosophila embryos (like the gut) are not affected in Crb mutants (Tepass et al., 1990)? A detailed study of the expression and localization of the proteins interacting with Crumbs should help to understand why in these cells, Crumbs is not essential for epithelial polarity. This together with the identification of new proteins belonging to the Crumbs complex might explain the relation between the building of the ZA and the SAC.
Acknowledgements We would like to thank J.P. Arsanto for critically reading this manuscript and members of the Crumbs network for the sharing of unpublished information. This work was supported by Centre National de la Recherche Scientifique CNRS 6156, Université de la Méditerranée, Institut de Biologie du Développement de Marseille, Fondation de France and Association pour la Recherche sur le Cancer 9297, and a EC grant (Crumbs therapeutics) to A. Le Bivic.
References Bachmann, A., Schneider, M., Theilenberg, E., Grawe, F., Knust, E., 2001. Drosophila Stardust is a partner of Crumbs in the control of epithelial cell polarity. Nature 414, 638–643. Baumgartner, S., Littleton, J.T., Broadie, K., Bhat, M.A., Harbecke, R., Lengyel, J.A., Chiquet-Ehrismann, R., Prokop, A., Bellen, H.J., 1996. A Drosophila neurexin is required for septate junction and blood-nerve barrier formation and function. Cell 87, 1059–1068. Bellen, H.J., Lu,Y., Beckstead, R., Bhat, M.A., 1998. Neurexin IV, caspr and paranodin--novel members of the neurexin family: encounters of axons and glia. Trends Neurosci. 21, 444–449. Bhat, M.A., Izaddoost, S., Lu, Y., Cho, K.O., Choi, K.W., Bellen, H.J., 1999. Discs Lost, a novel multi-PDZ domain protein, establishes and maintains epithelial polarity. Cell 96, 833–845. Bilder, D., Li, M., Perrimon, N., 2000. Cooperative regulation of cell polarity and growth by Drosophila tumor suppressors [see comments]. Science 289, 113–116. Bunt-Milam, A.H., Saari, J.C., Klock, I.B., Garwin, G.G., 1985. Zonulae adherentes pore size in the external limiting membrane of the rabbit retina. Invest. Ophthalmol. Vis. Sci. 26, 1377–1380. Carlson, S.D., Juang, J.L., Hilgers, S.L., Garment, M.B., 2000. Blood barriers of the insect. Annu. Rev. Entomol. 45, 151–174. den Hollander, A.I., ten Brink, J.B., de Kok, Y.J., van Soest, S., van den Born, L.I., van Driel, M.A., van de Pol, D.J., Payne, A.M., Bhattacharya, S.S., Kellner, U., Hoyng, C.B., Westerveld, A., Brunner, H.G., Bleeker-Wagemakers, E.M., Deutman, A.F., Heckenlively, J.R., Cremers, F.P., Bergen, A.A., 1999. Mutations in a human homologue of Drosophila crumbs cause retinitis pigmentosa (RP12). Nat. Genet. 23, 217–221.
E. Médina et al. / Biology of the Cell 94 (2002) den Hollander, A.I., Heckenlively, J.R., van den Born, L.I., de Kok, Y.J., van der Velde-Visser, S.D., Kellner, U., Jurklies, B., van Schooneveld, M.J., Blankenagel, A., Rohrschneider, K., Wissinger, B., Cruysberg, J.R., Deutman, A.F., Brunner, H.G., Apfelstedt-Sylla, E., Hoyng, C.B., Cremers, F.P., 2001a. Leber congenital amaurosis and retinitis pigmentosa with Coats-like exudative vasculopathy are associated with mutations in the crumbs homologue 1 (CRB1) gene. Am. J. Hum. Genet. 69, 198–203. den Hollander, A.I., Johnson, K., de Kok, Y.J., Klebes, A., Brunner, H.G., Knust, E., Cremers, F.P., 2001b. CRB1 has a cytoplasmic domain that is functionally conserved between human and Drosophila. Hum. Mol. Genet. 10, 2767–2773. den Hollander, A.I., Ghiani, M., de Kok, Y.J., Wijnholds, J., Ballabio, A., Cremers, F.P., Broccoli, V., 2002. Isolation of Crb1, a mouse homologue of Drosophila crumbs, and analysis of its expression pattern in eye and brain. Mech. Dev. 110, 203–207. Fanning, A.S., Anderson, J.M., 1999. Protein modules as organizers of membrane structure. Curr. Opin. Cell Biol. 11, 432–439. Grawe, F., Wodarz, A., Lee, B., Knust, E., Skaer, H., 1996. The Drosophila genes crumbs and stardust are involved in the biogenesis of adherens junctions. Development 122, 951–959. Hong, Y., Stronach, B., Perrimon, N., Jan, L.Y., Jan, Y.N., 2001. Drosophila Stardust interacts with Crumbs to control polarity of epithelia but not neuroblasts. Nature 414, 634–638. Hynes, R.O., 1992. Integrins: versatility, modulation, and signaling in cell adhesion. Cell 69, 11–25. Izaddoost, S., Nam, S.C., Bhat, M.A., Bellen, H.J., Choi, K.W., 2002. Drosophila Crumbs is a positional cue in photoreceptor adherens junctions and rhabdomeres. Nature 416, 178–183. Jürgens, G., Wieschaus, E., Nüsslein-Volhard, C., Kluding, H., 1984. Mutations affecting the pattern of the larval cuticle in Drosophila melanogaster. Roux’s Arch. Dev. Biol. 193, 283–295. Kamberov, E., Makarova, O., Roh, M., Liu, A., Karnak, D., Straight, S., Margolis, B., 2000. Molecular cloning and characterization of Pals, proteins associated with mLin-7. J. Biol. Chem. 275, 11425–11431. Klebes, A., Knust, E., 2000. A conserved motif in Crumbs is required for E-cadherin localisation and zonula adherens formation in Drosophila. Curr. Biol. 10, 76–85. Lemmers, C., Medina, E., Delgrossi, M.H., Michel, D., Arsanto, J.P., Le Bivic, A., 2002. hINADl/PATJ, a homolog of discs lost, interacts with crumbs and localizes to tight junctions in human epithelial cells. J. Biol. Chem. 277, 25408–25415. Leptin, M., 1999. Gastrulation in Drosophila: the logic and the cellular mechanisms. Embo J. 18, 3187–3192. Medina, E., Williams, J., Klipfell, E., Zarnescu, D., Thomas, G., Le Bivic, A., 2002. Crumbs interacts with moesin and {beta}Heavy-spectrin in the apical membrane skeleton of Drosophila. J. Cell Biol. 158, 941–951. Mostov, K.E., Verges, M., Altschuler, Y., 2000. Membrane traffic in polarized epithelial cells. Curr. Opin. Cell Biol. 12, 483–490. Nagafuchi, A., 2001. Molecular architecture of adherens junctions. Curr. Opin. Cell Biol. 13, 600–603. Pellikka, M., Tanentzapf, G., Pinto, M., Smith, C., McGlade, C.J., Ready, D.F., Tepass, U., 2002. Crumbs, the Drosophila homologue of human CRB1/RP12, is essential for photoreceptor morphogenesis. Nature 416, 143–149.
313
Perrimon, N., Lanjuin, A., Arnold, C., Noll, E., 1996. Zygotic lethal mutations with maternal effect phenotypes in Drosophila melanogaster. II. Loci on the second and third chromosomes identified by P-elementinduced mutations. Genetics 144, 1681–1692. Rodriguez-Boulan, E., Nelson, W.J., 1989. Morphogenesis of the polarized epithelial cell phenotype. Science 245, 718–725. Roh, M.H., Liu, C.J., Laurinec, S., Margolis, B., 2002a. The carboxyl terminus of zona occludens-3 binds and recruits a mammalian homologue of discs lost to tight junctions. J. Biol. Chem. 277, 27501–27509. Roh, M.H., Makarova, O., Liu, C.J., Shin, K., Lee, S., Laurinec, S., Goyal, M., Wiggins, R., Margolis, B., 2002b. The Maguk protein, Pals1, functions as an adapter, linking mammalian homologues of Crumbs and Discs Lost. J. Cell Biol. 157, 161–172. Stabach, P.R., Morrow, J.S., 2000. Identification and characterization of beta V spectrin, a mammalian ortholog of Drosophila beta H spectrin. J. Biol. Chem. 275, 21385–21395. Tanentzapf, G., Smith, C., McGlade, J., Tepass, U., 2000. Apical, Lateral, and Basal Polarization Cues Contribute to the Development of the Follicular Epithelium during Drosophila Oogenesis. J. Cell Biol. 151, 891–904. Tepass, U., Theres, C., Knust, E., 1990. Crumbs encodes an EGF-like protein expressed on apical membranes of Drosophila epithelial cells and required for organization of epithelia. Cell 61, 787–799. Tepass, U., Knust, E., 1993. Crumbs and stardust act in a genetic pathway that controls the organization of epithelia in Drosophila melanogaster. Dev. Biol. 159, 311–326. Tepass, U., 1996. Crumbs, a component of the apical membrane, is required for zonula adherens formation in primary epithelia of Drosophila. Dev. Biol. 177, 217–225. Tepass, U., Tanentzapf, G., Ward, R., Fehon, R., 2001. Epithelial cell polarity and cell junction in Drosophila. Annu. Rev. Genet. 35, 747–784. Tepass, U., 2002. Adherens junctions: new insight into assembly, modulation and function. Bioessays 24, 690–695. Thomas, G., 2001. Spectrin: The ghost in the machine. Bioessays 23, 152–160. Tsukita, S., Furuse, M., 1999. Occludin and claudins in tight-junction strands: leading or supporting players?. Trends Cell Biol. 9, 268–273. Vasioukhin, V., Fuchs, E., 2001. Actin dynamics and cell-cell adhesion in epithelia. Curr. Opin. Cell Biol. 13, 76–84. Williams, D.S., Arikawa, K., Paallysaho, T., 1990. Cytoskeletal components of the adherens junctions between the photoreceptors and the supportive Muller cells. J. Comp. Neurol. 295, 155–164. Wodarz, A., Grawe, F., Knust, E., 1993. CRUMBS is involved in the control of apical protein targeting during Drosophila epithelial development. Mech. Dev. 44, 175–187. Wodarz, A., Hinz, U., Engelbert, M., Knust, E., 1995. Expression of crumbs confers apical character on plasma membrane domains of ectodermal epithelia of Drosophila. Cell 82, 67–76. Wodarz, A., 2002. Establishing cell polarity in development. Nat. Cell Biol. 4, E39–44. Yeaman, C., Grindstaff, K.K., Hansen, M.D., Nelson, W.J., 1999. Cell polarity: Versatile scaffolds keep things in place. Curr. Biol. 9, R515–517. Zarnescu, D.C., Thomas, G.H., 1999. Apical spectrin is essential for epithelial morphogenesis but not apicobasal polarity in Drosophila. J. Cell Biol. 146, 1075–1086.