BIOLOGICAL OVERVIEW

Starry night (Stan), also known as Flamingo (Fmi), is a protein involved in the establishment of tissue polarity. Described almost simultaneously by three research groups (Usui, 1999; Lu, 1999 and Chae, 1999), Starry night possesses a huge protocadherin domain containing nine cadherin motifs, four EGF-like motifs, and two laminin G motifs. The dependence of proper Starry night localization on Frizzled (Fz) activity suggests that Stan functions downstream of Fz in controlling planar polarity. Stan protein is localized in specific domains of the plasma membranes in polarizing cells and controls planar cellular polarity. During a restricted interval prior to prehair outgrowth, Stan distribution is polarized along the proximal distal (P/D) axis; Stan molecules are present predominantly at proximal and distal cell boundaries rather than at anterior and posterior ones. Stan mutation disrupts the orientation of the axis of cell division of sensory organ precursors. stan phenotypes and biology call attention to the relationship between establisment of the axis of cell division and the resulting tissue polarity trait. It also ties together two heretofore disparate pathways: the asymmetric cell division trait in which Numb functions as a downstream factor, and the Frizzled pathway for the establisment of tissue polarity (Usui, 1999; Lu, 1999, and Chae, 1999).

In an attempt to pursue functional relationships between stan and previously discovered tissue polarity genes, a study was carried out to see whether Stan distribution is altered in various polarity mutants, particularly in fz complete loss-of-function mutants. In the total absence of Fz protein (fz^D21/fz^K21) Stan is not redistributed, as it normally is, toward the proximal/distal (P/D) cellular boundaries at 24 or 30 hr after puparium formation (APF). At the onset of prehair formation (30 hr APF), bright staining at cell boundaries is greatly reduced in length, and the fragmented signals are not necessarily restricted to the P/D boundaries, indicating that generation of the normal Stan pattern is strongly dependent on Fz. Residual boundary signals become even less prominent at later stages, leaving only fine dots both in the cytoplasm and along cell borders. Along the apicobasal cell axis, these intracellular particles are present from near the apical surface to the basolateral level. dsh¹ (an allele of dishevelled) is a genetic null allele for planar polarity. Wings with this mutation also show a decrease in intensity of Stan staining at cell boundaries, and the distribution appears to be much less polarized than that in wild type. Mutations of other genes involved in tissue polarity do not necessarily disrupt the Stan distribution. For example, in mutant cells of the multiple wing hair (mwh) gene, which is currently considered to be further downstream in the tissue polarity pathway, Stan molecules are present predominantly at P/D boundaries. Therefore, Fz and one downstream component, Dsh, are thought to be necessary to accomplish the normal distribution of Stan (Usui, 1999).

In the preceding experiment, Stan distribution was studied in pupal wings where all the cells had lost fz expression. To dissect how Fz-dependent intercellular communication controls Stan localization, two approaches were adopted to juxtapose cells with different fz expression levels and see how those conditions affect Stan distribution. One approach generated fz mutant clones, and the other expressed fz in a gradient fashion. Clones were made of cells homozygous for a strong fz allele, fz^R52, which produces a truncated polypeptide at a very low level. Strikingly, the fz mutant cells appear to decide where to localize Stan molecules in a neighbor-dependent manner. Along clone borders, Stan accumulates at almost all the interfaces between fz⁺ (fz^R52/+ or +/+) and fz^R52/fz^R52 cells, whether the interface is a P/D cell boundary or not. In contrast, Stan is never localized intensely at boundaries between outermost mutant cells; in other words, those outermost mutant cells always restrict the distribution of Stan to contact sites with fz⁺ cells. These results indicate that every cell has a system to monitor fz expression levels across each boundary and that if there is an imbalance across a certain boundary, the cell deposits Stan molecules preferentially at this particular cell-cell contact site (Usui, 1999).

Inner mutant cells that do not contact the clone borders display fuzzy Stan signals both at interfaces between themselves and in the cytoplasm, and the distribution of the boundary signals is not polarized. These abnormal patterns are reminiscent of those in wings of the fz null mutant. This observation of the fz clone confirms that the fz gene is necessary to concentrate Stan to cell-cell boundaries and, in addition, to bias the distribution of Stan toward P/D boundaries (Usui, 1999).

Several recessive stan alleles have been isolated on the basis of the wing hair polarity phenotype seen in a wing in an FLP/FRT based F1 screen. Thus, it is clear that the presence of wild-type neighboring cells do not rescue all of the mutant cells in a clone. Several of the tissue polarity genes display domineering nonautonomy in wing clones -- that is, the presence of mutant cells in a clone alters the development of wild-type cells that are near the clone. To see if stan clones also display domineering nonautonomy, mosaic wings were generated where stan clones were marked with the hair marker pwn. Several different alleles were tested including the putative null allele stan²⁴, the recessive lethal allele stan²¹ and the recessive viable allele stan³, and similar results were obtained. In all cases, the majority of clones behaved cell autonomously. Further, the extent of domineering nonautonomy for those clones scored as showing such nonautonomy was typically much weaker than seen with fz or Vang (Van Gogh). It is concluded from these experiments that stan principally functions cell autonomously (Chae, 1999).

For an in vivo assay for fz pathway function, the domineering non-autonomy of fz clones was used. To do this, fz^R52 strb clones were induced in stan³ wings. In a wild-type wing, more than 80% of fz clones show distal domineering non-autonomy. That is, cells distal (and in part anterior/posterior) but not proximal to the clone show altered polarity that extends to cells that do not border the clone. fz^R52 strb clones were induced in regions of stan³ wings, where the polarity was consistent enough to be able to score the clones for domineering non-autonomy. Out of 54 clones, forty two clones behave cell autonomously and only 12 clones showed evidence of domineering non-autonomy. Further, the extent of domineering non-autonomy in these 12 clones was modest. Thus, stan appears to be a suppressor of the domineering non-autonomy of fz. That there remains some fz domineering non-autonomy in stan³ wings may reflect the fact that stan³ is not a null allele. The ability of a stan mutation to suppress this fz phenotype argues that stan is downstream of fz and is required for the cell non-autonomous function of the fz pathway (Chae, 1999).

As a second in vivo assay for fz pathway function, the ability of a gradient of fz expression, with its highest point near the distal tip of the wing, was used to reverse the normal distal polarity of wing hairs. This result argues that cells can 'sense' the fz activity of neighboring cells and respond to this information. The production of a region of reversed polarity is likely to require both cell non-autonomous (e.g., a fz-dependent intercellular signal) and cell autonomous functions (e.g., transduction of the fz-dependent signal). stan³ was found to completely block the ability of a gradient of fz expression to reorganize wing hair polarity. Hence it is concluded that stan functions downstream of fz and is required either for the cells to be able to sense the fz activity of neighboring cells or to respond to this information (Chae, 1999).

The overexpression of fz just prior to prehair initiation causes the formation of large numbers of multiple hair cells that are a phenocopy of the in-like mutations. This fz gain-of-function phenotype has been used as a test to identify genes that are downstream of and required for the transduction of the fz signal. The function of the dsh gene, which is thought to function downstream of fz, is indeed required for this phenocopy. However, the function of several other tissue polarity genes, pk, ds and Vang, is not required. To determine if stan is required for the transduction of the fz signal, stan;hs-fz flies were constructed and fz expression was induced just prior to prehair initiation. The stan³ does not block the ability of fz overexpression to induce cells to form multiple hairs. Rather, it appears to slightly enhance the ability of fz overexpression to induce multiple hair cells (Chae, 1999).

Advantage was taken of the GAL4-UAS system to appose cells with different levels of fz expression and an examination was carried out of how the distribution of Stan was altered. The patched (ptc)-GAL4 driver was used to generate a short-range gradient of fz expression along the A/P axis of the wing imaginal disc: this ectopic expression made the wing hairs point from high to low levels of Fz. In contrast to the normal zigzag patterns of Stan at P/D cell boundaries, Stan was concentrated primarily at A/P cell boundaries in the presence of the Fz gradient along the A/P axis. This result shows that under this experimental condition, juxtaposition of cells with different Fz levels is sufficient to accumulate Stan at the interface (Usui, 1999).

Effects of fz overexpression are known to be suppressed by mutations in the dsh gene, which acts downstream of fz; this suggests that fz overexpression mimics activation of the signaling pathway. Given that levels of fz gene expression correlate with those of Fz protein activity, the results of the gradient expression and the clonal analyses allow two conclusions to be drawn: (1) a cell is able to communicate with its neighbors to monitor the amplitude of Fz activity across each cell-cell boundary; (2) enrichment of Stan at one boundary is a hallmark of a difference of Fz activity in between the two juxtaposed cells (Usui, 1999).

Results described so far suggest a model showing (1) that in the wild-type wing, Fz activity is made unequal across every P/D cell boundary; (2) that this difference causes bilateral assembly of Stan molecules at the P/D boundary, and (3) that those assembled Stan molecules play an important role in initiating prehair formation in the vicinity of the P/D boundary. However, bilateral distribution of Stan per se does not explain how the distal edge, not the proximal one, is selected to reorganize the cytoskeleton to form a prehair. An experiment of stan overexpression, described next, provides an insight into how Stan breaks cellular symmetry. As is also the case with fz, not only loss-of-function mutations but also overexpression of stan could disrupt planar polarity. By using a number of GAL4 drivers, the effects of overproduction of Stan has been compared with those of overproduction of Fz. The most striking result is obtained when a gradient of Stan is induced with the ptc-Gal4 driver. The phenotype induced by the Stan gradient presents a sharp contrast to that caused by the Fz gradient, that is, hairs are oriented toward the highest point of Stan levels. Cells within the gradient tend to accumulate overproduced Stan molecules in cytoplasm. This polarization along the Stan gradient requires the Stan ectodomain, because gradient production of the DeltaEX form has no effect on the hair polarity (Usui, 1999).

These results strongly suggest that when cells with equivalent P/D positional information are given the graded expression of stan, they perceive the difference in the expression level across the cell boundaries and direct hairs toward their neighbors, which produce higher levels of Stan. If overproduced Stan molecules activate a hypothetical downstream signaling cascade, whether they are located in plasma membranes or in cytoplasm, it follows that wing hairs point toward cells with stronger activity of Stan. This assumption permits a model that explains generation of distally oriented hairs in the wild-type wing, where neither Fz nor Stan is distributed in a gradient fashion along the P/D axis. In this model, the distal side of the P/D boundary, that is, the proximal edge of every cell, has a stronger Stan activity in spite of the apparent symmetrical distribution (Usui, 1999).

Polarization of the Stan distribution can be explained by two mechanisms that are not mutually exclusive. One is the sorting of Stan preferentially to P/D cell boundaries, and the other is involved in selective retention and degradation of Stan at P/D and A/P boundaries, respectively. Fz is able to recruit a signal transducer, Dsh, from cytoplasmic vesicles to cell-cell interfaces in a heterologous system (Axelrod, 1998). This finding may be suggestive of the intracellular sorting event. This Fz-dependent translocation of Dsh and aberrant localization of Stan in the dsh mutant imply a possibility that the sorting of Stan could be mediated by either direct or indirect interaction between Stan and Dsh. Whichever mechanism works, the observations made here suggest that this operation is initiated after 18 hr APF. Consistently, temperature-shift experiments using a cold-sensitive fz mutation suggest that fz function is required between approximately 15 hr APF at 25°C and the start of prehair morphogenesis. One obvious question that has not been rigorously answered is where Fz is localized within the cell. Staining for endogenous Fz proteins is challenging because of their low abundance; it is sometimes possible to detect Fz signals at both the apical-free surfaces and apical portions of cell boundaries, irrespective of the stage of prehair development. A more difficult task is visualizing the sites where Fz is activated; one reason for this difficulty is that the ligand for Fz, which is considered to be the polarizing signal, has not been identified (Usui, 1999).

Apart from Stan, the cell fate determinant Numb has been the only protein reported that displays biased distribution along a planar axis; its localization is also under the control of Fz signaling (Gho, 1998). Numb is a membrane-associated intracellular protein and forms a crescent that overlies one of the two spindle poles of cells that undergo asymmetric divisions. A search for proteins interacting with Numb led to the identification of Pon, which colocalizes with Numb during mitosis and directs Numb-asymmetric localization (Lu, 1998). Although polarization and depolarization of the Stan distribution occurs in postmitotic cells, a hunt for binding partners may help to disclose the molecular machinery that regulates Stan location (Usui, 1999).

Why do cells employ a system that is so complicated and requires spatially distinct activation of the two receptor species instead of Fz activation alone? What is Stan's role in breaking cellular symmetry along the P/D axis? It is speculated that the initial bias of the Fz activity is too subtle to drive cytoskeletal reorganization and that Stan is responsible for making this bias stronger. If active Stan molecules at the proximal edge downregulate activity of Fz molecules in the same domain of the plasma membrane, this inhibitory effect of Stan against Fz could enhance the difference in Fz activity across the P/D boundary. The antagonistic interplay between the two receptors could make the imbalance of Fz activity exceed a certain threshold to initiate prehair formation at the distal cell vertex. Planar polarity is also reflected in the arrangement of photoreceptor cells in the Drosophila eye, and it has been proposed that Fz sets up an initial small bias, which is amplified by the Delta-Notch pathway. Stan is also required for polarity formation in the eye; therefore, it is intriguing to investigate where Stan is localized in ommatidia and whether Stan is involved in the connection between Fz and Notch signaling pathways (Usui, 1999).

To test the function of Stan in regulating spindle orientation and Numb protein localization during the SOP pI division, the effects were analyzed of both the loss of function and the overexpression of stan on the pupal notum. In stan (fmi^E59/fmi⁷¹) transheterozygous mutant pupae, the Numb crescent is randomly positioned within the epithelial plane during the SOP pI division. The adapter protein Partner of Numb (Pon), which controls Numb localization during SOP division (Lu, 1998) is also mispositioned, but the two proteins remain colocalized. Staining for tubulin reveals that spindle orientation and Numb crescent positioning are still tightly coupled. During the SOP pI division in apterous-Gal4;UAS-fmi pupae, the Numb crescent is also mispositioned and the mitotic spindle misoriented within the epithelial plane, but they remain aligned with each other. Therefore, loss of function and overexpression of stan both disrupts the cellular process that regulates mitotic spindle orientation and protein localization during the SOP pI division (Lu, 1999).

To gain a better insight into the mechanism of Stan function, its subcellular localization was examined in SOP cells using an antibody against the ectodomain of Stan. Consistent with Stan being a seven-transmembrane cell-adhesion molecule, Stan is localized to cell-cell boundaries in both the SOPs and their surrounding epithelial cells. There was also a low level of punctate cytoplasmic staining, suggesting that cytoplasmic Stan may be associated with intracellular vesicles. The Stan staining in SOP cells is stronger than that in the surrounding epithelial cells, indicating elevated expression or stability of Stan in the SOPs. No apparent polarized distribution of Stan is observed in mitotic SOPs. In nota of apterous-Gal4;UAS-fz flies, the localization of Stan in the SOPs is similar to that in wild-type flies. In nota of fz^r54/fz^KD4a mutant flies, Stan staining at the cell-cell boundary is reduced, whereas cytoplasmic Stan staining is increased. A similar effect is observed in the notum of dsh¹ mutant flies. In the wing epithelia, localization of Stan at the cell-cell boundary is also affected in fz and dsh mutant backgrounds. Therefore, the proper recruitment of Stan from the cytoplasm to the cell-cell boundary depends on Fz signaling. It has been shown that Dsh can be selectively recruited to the membrane by Fz but not by Dfz2. Whether a similar mechanism is involved in recruiting Stan and Dsh to the membrane remains to be determined (Lu, 1999).

These data indicate that Stan functions largely in a cell-autonomous manner to control the planar polarity of sensory bristles on the Drosophila notum. The similar polarity phenotypes caused by overexpression and loss of function of stan or fz, together with the dependence of proper Stan localization on Fz activity, suggest that Stan functions downstream of Fz in the planar polarity pathway. It should be noted that, although loss of stan function affects both bristle polarity and the positioning of Numb crescents, the correlation between these two phenotypes is not strict. Although the orientation of the Numb crescent is largely random in the stan mutant, this is not the case for bristle polarity. It is possible that there exist some other mechanical constraints in the developing imaginal epithelia that influence the orientation of bristles or that Fz-independent cues can direct bristle orientation. The tight coupling of misoriented mitotic spindle and mislocalized Numb crescent during SOP pI divisions suggests that a downstream activity which coordinates these two processes is still intact when stan or fz are overexpressed or inactive. During neuroblast division, Inscuteable acts downstream of Bazooka to coordinate spindle orientation and Numb localization. It is unlikely, however, that Inscuteable is the activity required to couple these two processes during the SOP pI division. It is possible that, in the SOP lineage, Stan/Fz functions like Bazooka to regulate an Inscuteable-like activity, which in turn couples spindle orientation and protein localization. Despite the difference between the planar polarity pathway and the Bazooka/Inscuteable pathway for the orthogonal apical-basal polarity, both pathways appear to regulate the localization of Numb through Pon. Further studies will help clarify how the two pathways impinge on Pon to control Numb localization (Lu, 1999).

PROTEIN STRUCTURE

Amino Acids - 3589

Structural Domains

The predicted translational product of stan has 3575 amino acids. The Stan ectodomain has nine cadherin repeats, three cysteine-rich domains (Cys-rich), and two laminin A globular domains (LmA-G). A combination of Cys-rich and LmA-G domains is a common feature of many invertebrate members and some vertebrate molecules of the cadherin superfamily. In contrast to those of classic-type cadherins, the carboxy-terminal intracellular tail of Stan does not possess catenin-binding sequences, nor does it bind to catenins. Thus, Stan can be classified into the nonclassic-type subfamily. Compared with functionally characterized members of this superfamily, Stan is structurally unusual, as it is predicted to be a seven-pass transmembrane (7TM) protein. Sequences of the 7TM region show similarity to those of one particular family of G protein-coupled receptors: the first such protein to be isolated is a receptor for a peptide hormone, secretin. Many proteins of this secretin receptor family have been shown to increase the intracellular levels of cAMP and/or inositol phosphates upon ligand binding. Whether Stan is coupled to G proteins remains to be demonstrated. The stan gene is conserved across species: a Caenorhabditis elegans counterpart gene is present in cosmid clones F15B9 and W07G4 (GenBank), and two paralogs of mammalian receptors have been reported, that is, mouse Celsr1 (Hadjantonakis, 1998) and rat MEGF2 (Nakayama, 1998). In addition, mouse cDNA clones have been isolated that encode the entire protein of a third paralog, mouse Flamingo1 (Usui, 1999).

EVOLUTIONARY HOMOLOGS

Mcelsr1 encodes a protein of 3034 amino acids predicted to contain seven membrane spanning domains having homology to a group of peptide hormone binding G-protein coupled receptors. Its extracellular domain comprises epidermal growth factor-like repeats, laminin A G-domains and cadherin repeats. Homologous genes have been identified in C. elegans and D. melanogaster suggesting that the Celsr gene family is ancient. mCelsr1 mRNA expression precedes gastrulation, is subsequently restricted primarily to ectodermal derivatives and is tightly regulated in the developing central nervous system (CNS). Segmentally-restricted gene expression in the developing hindbrain and in the spinal cord dynamic dorso-ventrally restricted 'stripes' of expression is observed (Hadjantonakis, 1998).

To identify large proteins with an EGF-like-motif in a systematic manner, a computer-assisted method called motif-trap screening has been developed. The method exploits 5'-end single-pass sequence data obtained from a pool of cDNAs whose sizes exceed 5 kb. Using this screening procedure, five known and nine new genes for proteins with multiple EGF-like-motifs were identified from 8000 redundant human brain cDNA clones. These new genes were found to encode a novel mammalian homolog of Drosophila Fat protein; two seven-transmembrane proteins containing multiple cadherin and EGF-like motifs; two mammalian homologs of Drosophila Slit protein; an unidentified LDL receptor-like protein, and three totally uncharacterized proteins. The organization of the domains in the proteins, together with their expression profiles and fine chromosomal locations, has indicated their biological significance, demonstrating that motif-trap screening is a powerful tool for the discovery of new genes that have been difficult to identify by conventional methods (Nakayama, 1998).

cDNAs have been isolated for three members of a family of seven-pass transmembrane cadherins in mouse (Celsr1, 2 and 3). These three genes represent vertebrate homologs of flamingo/starry night, recently identified as an essential component of the Drosophila planar cell polarity pathway and for the correct formation of dendritic fields within the Drosophila peripheral nervous system. Each member of the mouse Celsr family exhibits distinct patterns of expression within a range of different tissues within the developing embryo. Celsr1 and Celsr2 expression is observed during gastrulation and within the developing nervous system. Celsr3 transcripts, however, are found only at sites of active neurogenesis (Formstone, 2001).

Celsr1 is expressed in the vicinity of the primitive streak during gastrulation. Further analysis within the late gastrulating embryo reveals that the Celsr1 expression domain lies predominately within the primitive streak. In contrast, Celsr2 expression is within anterior neural ectoderm. The contrasting expression patterns exhibited by Celsr1 and Celsr2 persist within the developing embryo at later stages. At the 7 somite stage, Celsr2 transcripts are up-regulated within the anterior ventral midline and restricted to rhombomere (r) 3 within the developing hindbrain. By 9.5 dpc, Celsr2 expression is uniformly expressed within the caudal neural tube and is absent from the isthmus region. Celsr2 transcripts are, however, up-regulated within r1 and are restricted to rhombomere boundaries. Celsr1 expression is also observed within rhombomere boundaries at this stage. At 9.5 dpc, Celsr3 exhibits punctate dorsal neural tube expression anterior to the level of the developing forelimb bud. Celsr3 expression is additionally observed within the ventral neural tube with expression restricted to r2 and r4 within the early developing hindbrain. Celsr3 is also expressed within the early peripheral nervous system. Further striking patterns of Celsr expression are revealed during spinal cord development. Within this tissue, the Celsr family appear to define different neuroepithelial cell populations. A complementary pattern of expression is clearly evident between Celsr2 and Celsr3 at 12 dpc. The sequential expression of the family within the spinal cord may reflect a progression in the maturation of neuronal precursors. Additionally, RNA transcripts for all three members of the Celsr family are observed within the dorsal root ganglia (Formstone, 2001).

Drosophila Flamingo is a 7-pass transmembrane cadherin that is necessary for dendritic patterning and axon guidance. How it works at the molecular level and whether homologs of Flamingo play similar roles in mammalian neurons or not have been unanswered questions. Loss-of-function analysis using an RNAi system and organotypic brain slice cultures have been performed to address the role of a mammalian Flamingo homolog, Celsr2. Knocking down Celsr2 results in prominent simplification of dendritic arbors of cortical pyramidal neurons and Purkinje neurons, and this phenotype seemed to be due to branch retraction. Cadherin domain-mediated homophilic interaction appears to be required for the maintenance of dendritic branches. Furthermore, expression of various Celsr2 forms elicits distinct responses that are dependent on an extracellular subregion outside the cadherin domains and on a portion within the carboxyl intracellular tail (Shima, 2004).

To gain mechanistic insight into the molecular function of Celsr2, whether or not the siRNA-induced phenotype could be rescued by coexpressing each of several deletion forms of Celsr2A was addressed. Expression constructs that contained the target recognition sequence of the siRNA were made from Celsr2A^A2088T cDNA, which had a silent point mutation within the target sequence of siRNA2078. The extracellular region of Celsr2 was dissected into two subregions: a string of tandemly repeated cadherin domains (CR) and the more membrane-proximal subregion that contains motifs such as EGF-like domains, laminin G domains, and a hormone receptor domain (HRM). This membrane-proximal subregion was designated as the EGF-HRM region. Expression of ?EGFHRM-A, an A form without the EGF-HRM region, rescued the knockdown phenotypes of pyramidal neurons and Purkinje neurons, whereas ?CR-A and ?EX-A, in which the cadherin domains were totally deleted, did not (Shima, 2004).

Two forms were studied that had modified intracellular or transmembrane domains. One was Celsr2B, which lacked part of the carboxyl intracellular tail that includes residues conserved among the Fmi/Celsr family, and the other was Ex-1TM, in which the entire 7-pass transmembrane domain and the carboxyl tail were substituted with the single-pass domain of N-cadherin. Neither Celsr2B nor Ex-1TM recovered the effect of the siRNA. The results of this structure-function analysis show that the cadherin repeats as well as the carboxyl intracellular portion are required for rescuing the siRNA-induced dendritic malformation. In all of these attempts to rescue the knockdown phenotypes described in this study, dendritic morphology as observed at 3 days in vitro (DIV). In parallel with the coexpression of each Celsr2 form with the siRNA, each form alone was also expressed; expression of any form by itself did not give rise to statistically significant morphological effects when compared with the control protein expression at 3 DIV (Shima, 2004).

Frizzled3a and Celsr2 function in the neuroepithelium to regulate migration of facial motor neurons in the developing zebrafish hindbrain

Migration of neurons from their birthplace to their final target area is a crucial step in brain development. This study shows that expression of the off-limits/frizzled3a (olt/fz3a) and off-road/celsr2 (ord/celsr2) genes in neuroepithelial cells maintains the facial (nVII) motor neurons near the pial surface during their caudal migration in the zebrafish hindbrain. Celsr2 (for cadherin, EGF-like, LAG-like and seven-pass receptor), is a vertebrate homolog of Drosophila Flamingo. In the absence of olt/fz3a expression in the neuroepithelium, nVII motor neurons extended aberrant radial processes towards the ventricular surface and mismigrated radially to the dorsomedial part of the hindbrain. These findings reveal a novel role for these genes, distinctive from their already known functions, in the regulation of the planar cell polarity (i.e. preventing integration of differentiated neurons into the neuroepithelial layer). This contrasts markedly with their reported role in reintegration of neuroepithelial daughter cells into the neuroepithelial layer after cell division (Wada, 2006).

The present finding that neuroepithelial cells are involved in positioning specific neurons near the pial surface suggests a fundamental role for the neuroepithelium in brain development. In the mammalian cortex, neurons are generated in ventricular germinal zones and migrate radially towards the pial surface to form architectural layered structures. In mouse embryos, Reelin signaling regulates the positioning of neurons during layer formation of the cerebrum, and is essential for radial migration of the nVII motor neurons. These data suggest that similar mechanisms regulate the proper positioning of both the hindbrain motor neurons and the cortical layer neurons (Wada, 2006).

In the mouse cerebral cortex, many wnt and frizzled family genes are expressed in gene-specific regional and lamina patterns. Such patterned expression suggests the possibility that these genes are involved in other aspects of brain development. Recent studies have shown that functional fzd3 and celsr3 genes are required for the development of the anterior commissure, and the cortico-subcortical, thalamocortical and corticospinal tracts. It is possible that the mouse fzd3 and celsr3 genes regulate neuroepithelial cells to guide these axonal tracts to the proper region in a similar manner to that by which the zebrafish fz3a and celsr genes act in neuroepithelial cells to restrict the migrating nVII motor neurons near the pial surface of the hindbrain. The demonstration of a role for neuroepithelial cells in preventing integration of differentiated neurons into the neuroepithelial layer may provide new insights into the general mechanisms underlying the formation of layered structures in the mammalian brain, such as in the cerebral cortex (Wada, 2006).

date revised: 6 May 2000

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.